1 INFO-VAX	Tue, 30 Apr 2002	Volume 2002 : Issue 237       Contents: Re: $QIO and SYSPRV  Re: $QIO and SYSPRV  Re: $QIO and SYSPRV  Re: $QIO and SYSPRV  Re: $QIO and SYSPRV  Re: $QIO and SYSPRV  Re: $QIO and SYSPRV & Re: (GCC) GNU C/C++ compiler for Alpha	 AIM HACK? = Re: AlphaServer 1000 Hobbyist (was: Re: What have I got here) + Re: Another Microsoft breaking of standards + Re: Another Microsoft breaking of standards " Anticipating the HP court decision& Re: Anticipating the HP court decision& Re: Anticipating the HP court decision& Re: Anticipating the HP court decision, Re: Anyone using CHARON-VAX for development? Re: Can't edit a file  Re: Can't edit a file  Re: Compaq HPS times Re: Compaq HPS times Re: Compaq HPS times2 Re: comparison chart: pros and cons of WWW servers2 Re: COPY/RCP fails to remote node (UCX4.2 VMS 7.1)2 Re: COPY/RCP fails to remote node (UCX4.2 VMS 7.1)2 Re: COPY/RCP fails to remote node (UCX4.2 VMS 7.1)D dBASE IV package/license for VMS VAX, product status?  Transferable?+ Re: DCL labels out of scope after ON ERROR? : Re: DECUS/EncompassUS Survey -- no longer assuming Windows: Re: DECUS/EncompassUS Survey -- no longer assuming Windows Re: Digital 7-year plan... Re: Digital 7-year plan... Re: Digital 7-year plan... Re: download VMS Re: download VMS Re: download VMS Re: download VMS Re: EDT or EVE Re: EDT or EVE Re: exceed escape characters Exceed to get VMS display  Re: Exceed to get VMS display  Re: Exceed to get VMS display   Re: Fix for EDT emulation in EVE  Re: Fix for EDT emulation in EVE Re: Gold key Gold key Re: Gold key HPS Times is now availableP INFONOVA and SyntheSys Secure Technologies, Inc. Partner To Globally Address PKI Re: Itanium troubles  Re: MAIL on VMS and attachments. Re: Netscape and Mozilla Re: Network Delay? Re: Network Delay?- Re: OpenVMS MAIL (was: Re: Wither ALL-IN-1 ?)  Re: OT Bobby Vinton # Re: Pathworks 6.1 exhausting KNBCBs P Re: put up or shut up (was: RE: AMD To Announce Microsoft Support For New Chip FP Re: put up or shut up (was: RE: AMD To Announce Microsoft Support For New Chip F  Re: Reduce interupt time on CPU0, Re: removing extra carriage return/line feed Re: Security question....  Re: Security question.... > Re: somewhat off-topic: is there a DSL router with a console?!
 TCPIP$C_SHARE * Re: UK/EU OpenVMS job market: non-existant* Re: UK/EU OpenVMS job market: non-existant* Re: UK/EU OpenVMS job market: non-existant* Re: UK/EU OpenVMS job market: non-existant Re: VAX/Alpha CI* Re: VAXstation 4000 90 and serial console?* Re: VAXstation 4000 90 and serial console?! virtual i/o cache v7.2-1 and v7.3 ' Re: What does "affordable" mean to you? ' Re: What does "affordable" mean to you? ' Re: What does "affordable" mean to you? ' Re: What does "affordable" mean to you? ' Re: What does "affordable" mean to you?  What happened to Uwe Zessin?1 Re: will memory leak program caused system crash?  Wither ALL-IN-1 ? 
 Re: X session   F ----------------------------------------------------------------------  # Date: Mon, 29 Apr 2002 19:00:56 GMT 2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman) Subject: Re: $QIO and SYSPRV8 Message-ID: <ILgz8.46$1_.696357@cacnews.cac.cpqcorp.net>  ` In article <3CC88BA9.5080902@wasd.vsm.com.au>, Mark Daniel <Mark.Daniel@wasd.vsm.com.au> writes:" :Thanks for your involvement Hoff. :  :Hoff Hoffman wrote: :8< snip 8< K :>   ASTs are not particularly relevent, as only DECthreads (and then only  J :>   with the per-threads security enabled, and this is available only on M :>   recent OpenVMS releases) operate with privileges specific to the thread  M :>   of execution -- ASTs operate within process context, and either have or  I :>   do not have the privileges.  The ASTs do not preserve the privilege  J :>   context for later use, the privileges check(s) are made when the AST 
 :>   runs. :8< snip 8<  : 3 :Which privilege checks are made when the AST runs? I :I'd say the privilege checks on what the AST code demands.  Now the AST  I :routine may be doing something totally unrelated to the processing that   :triggered it.  J   The privilege checks are entirely and directly based on the contents of '   the code executed by the AST routine.   6 :What about privileges required for driver processing?     That depends on the driver.   J :I'm either being too obtuse in my explanations or in my understanding of I :the responses (or both I suppose).  I understand the AST routine is not  M :relevant, at least not to the intent of my question.  Let me try to clarify  5 :(I won't pursue the point after this one last time).   L :When an asynchronous $QIO is initiated does any required privilege need to E :*remain* enabled until the AST routine is *activated* (i.e. for the  G :duration of the driver handling the processing outside of the process  E :context), or once the asynchronous $QIO is initiated and returns to  H :initiating code can it be disabled immediately.  The latter is what my = :reading of McCoy's description of the XQP subsystem implies.   H   I re-wrote the AST chapter in the programming concepts manual a couple'   of releases ago.  Please start there.   E   The XQP is another whole discussion, as that is a large and complex F   environment.  In general, you only need the privileges or the accessE   when you initiate the file operation, and you can safely ignore any 5   of the various ASTs that fly around within the XQP.   E : > In the chapter The XQP and I/O processing, section on $QIO System E : > Service Dispatching (page 271 in my copy) it actually states that D : > blah-blah, the IRP is filled in, then *validated against processI : > privilege*, blah-blah, *then* the IRP is given to the driver.  So, my L : > reading of this would be that regardless of the $QIO (RMS) call I shouldH : > be able to disable a required privilege immediately the asynchronousE : > $QIO returns regardless of whether the operation is anywhere near $ : > completion at that stage or not.  
   Correct.  H :Whether or not the AST routine requires any extended privileges is not J :relevant to my query.  It is whether the privilege required for the $QIO M :needs to remain enabled during the driver processing, which, if it requires  B :some truly asynchronous activity such as device I/O, may be some " :(relatively) considerable period.  D   The privilege checks precede the operation -- that said, there are>   buffer checks when a read operation completes; OpenVMS againA   confirms that a read operation will only write to a buffer that    the user has access to.   >   Drivers are fully privileged, and can do whatever they want.B   Thus the relevent interface is the FDT routine within the driver@   code, and the $qio code within OpenVMS.  These two are used to>   control access.  This is not to say that the driver will not>   probe memory access and even process ownership of the memory@   when an asychronous read operation completes; the driver does A   need to verify that the memory is still accessable to the user.   G :I guess this can now be extended to RMS' use of $QIOs.  Do privileges  I :required for various RMS activities (such as file creation in otherwise  G :protected areas) need to remain enabled during the period between the  I :asynchronous service call and AST initiation?  If this is the case then  G :enabling an extended privilege for only a subset of all activities of  L :AST-driven code would need to prevent intervening (perhaps already queued) L :AST deliveries executing with that extended privilege, requiring some sort F :of serialization to be employed, blocking these deliveries until the K :privileged activity is complete.  The only way to provide this would seem  K :to be via a synchronous use of the service for that subset of activities.  M :This would imply that for this subset the application could not continue be  I :event-driven, it would stall for the duration of whatever that activity  
 :required.  6   You are getting XQP and ASTs mixed together.  Don't.  G   When you initiate an XQP operation, all XQP-related privilege checks     are performed.  D   You have "honked around" long enough here without explaining what D   you are really up to.  Please consider posting some background on D   the particular problem you seek to solve.  (There are ways you canD   get into deep trouble here, and among the more obvious ways is to E   try to implement privilege checks within your own application code. D   This is far tougher than it looks.  There are also ways to greatlyA   simplify this problem, such as subsystem identifier mechanism.)   N  ---------------------------- #include <rtfaq.h> -----------------------------N       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  % Date: Tue, 30 Apr 2002 11:09:31 +0930 / From: Mark Daniel <Mark.Daniel@wasd.vsm.com.au>  Subject: Re: $QIO and SYSPRV/ Message-ID: <3CCDF5D3.1D40D2D5@wasd.vsm.com.au>    Hoff Hoffman wrote: b > In article <3CC88BA9.5080902@wasd.vsm.com.au>, Mark Daniel <Mark.Daniel@wasd.vsm.com.au> writes:$ > :Thanks for your involvement Hoff. > :  > :Hoff Hoffman wrote:
 > :8< snip 8< M > :>   ASTs are not particularly relevent, as only DECthreads (and then only  L > :>   with the per-threads security enabled, and this is available only on O > :>   recent OpenVMS releases) operate with privileges specific to the thread  O > :>   of execution -- ASTs operate within process context, and either have or  K > :>   do not have the privileges.  The ASTs do not preserve the privilege  L > :>   context for later use, the privileges check(s) are made when the AST  > :>   runs.
 > :8< snip 8<  > : 5 > :Which privilege checks are made when the AST runs? K > :I'd say the privilege checks on what the AST code demands.  Now the AST  K > :routine may be doing something totally unrelated to the processing that   > :triggered it. > L >   The privilege checks are entirely and directly based on the contents of ) >   the code executed by the AST routine.   ! See comment on AST chapter below.    8 > :What about privileges required for driver processing? >  >   That depends on the driver.  > L > :I'm either being too obtuse in my explanations or in my understanding of K > :the responses (or both I suppose).  I understand the AST routine is not  O > :relevant, at least not to the intent of my question.  Let me try to clarify  7 > :(I won't pursue the point after this one last time).  > N > :When an asynchronous $QIO is initiated does any required privilege need to G > :*remain* enabled until the AST routine is *activated* (i.e. for the  I > :duration of the driver handling the processing outside of the process  G > :context), or once the asynchronous $QIO is initiated and returns to  J > :initiating code can it be disabled immediately.  The latter is what my ? > :reading of McCoy's description of the XQP subsystem implies.  > J >   I re-wrote the AST chapter in the programming concepts manual a couple) >   of releases ago.  Please start there.   B Been there, done that.  The Programming Concepts document has beenG steadily improving right through V7 ... thankyou.  This chapter is more H about the AST mechanism whereas I still think my query (poorly explainedC as it obviously is) is about from which point to which point in the F processing does a required privilege need to be enabled during the useE of asynchronous services, in particular $QIO and RMS which is largely & built on it (see code examples below).   G >   The XQP is another whole discussion, as that is a large and complex H >   environment.  In general, you only need the privileges or the accessG >   when you initiate the file operation, and you can safely ignore any 7 >   of the various ASTs that fly around within the XQP.  > G > : > In the chapter The XQP and I/O processing, section on $QIO System G > : > Service Dispatching (page 271 in my copy) it actually states that F > : > blah-blah, the IRP is filled in, then *validated against processK > : > privilege*, blah-blah, *then* the IRP is given to the driver.  So, my N > : > reading of this would be that regardless of the $QIO (RMS) call I shouldJ > : > be able to disable a required privilege immediately the asynchronousG > : > $QIO returns regardless of whether the operation is anywhere near & > : > completion at that stage or not. >  >   Correct.  0 But can't be generalized to all $QIO processing?  J > :Whether or not the AST routine requires any extended privileges is not L > :relevant to my query.  It is whether the privilege required for the $QIO O > :needs to remain enabled during the driver processing, which, if it requires  D > :some truly asynchronous activity such as device I/O, may be some $ > :(relatively) considerable period. > F >   The privilege checks precede the operation -- that said, there are@ >   buffer checks when a read operation completes; OpenVMS againC >   confirms that a read operation will only write to a buffer that  >   the user has access to.  > @ >   Drivers are fully privileged, and can do whatever they want.D >   Thus the relevent interface is the FDT routine within the driverB >   code, and the $qio code within OpenVMS.  These two are used to@ >   control access.  This is not to say that the driver will not@ >   probe memory access and even process ownership of the memoryB >   when an asychronous read operation completes; the driver does C >   need to verify that the memory is still accessable to the user.  > I > :I guess this can now be extended to RMS' use of $QIOs.  Do privileges  K > :required for various RMS activities (such as file creation in otherwise  I > :protected areas) need to remain enabled during the period between the  K > :asynchronous service call and AST initiation?  If this is the case then  I > :enabling an extended privilege for only a subset of all activities of  N > :AST-driven code would need to prevent intervening (perhaps already queued) N > :AST deliveries executing with that extended privilege, requiring some sort H > :of serialization to be employed, blocking these deliveries until the M > :privileged activity is complete.  The only way to provide this would seem  M > :to be via a synchronous use of the service for that subset of activities.  O > :This would imply that for this subset the application could not continue be  K > :event-driven, it would stall for the duration of whatever that activity   > :required. > 8 >   You are getting XQP and ASTs mixed together.  Don't.  G I'm not sure that I am (except perhaps when trying to explain myself) I G was using AST delivery as the obvious end-point of the $QIO processing, D trying to ask for what duration from service initation (start point)@ until AST delivery (end point) do required privileges need to be enabled.   I >   When you initiate an XQP operation, all XQP-related privilege checks   >   are performed. > F >   You have "honked around" long enough here without explaining what F >   you are really up to.  Please consider posting some background on F >   the particular problem you seek to solve.  (There are ways you canF >   get into deep trouble here, and among the more obvious ways is to G >   try to implement privilege checks within your own application code. F >   This is far tougher than it looks.  There are also ways to greatlyC >   simplify this problem, such as subsystem identifier mechanism.)   F Both of the following examples are executing at user-mode AST delivery; level and so cannot be preempted by their own AST routines.   1 This is an *example* of the current construct ...       if (!rqptr->DeleteOnClose)     {&       /* asynchronous close service */.       tkptr->AccessFab.fab$l_fop |= FAB$M_ASY;;       sys$close (&tkptr->AccessFab, &SsiAccessesFileClosed, <                                     &SsiAccessesFileClosed);
       return;     }    else     {:       /* use SYSPRV to ensure deletion-on-close of file */(       sys$setprv (1, &SysPrvMask, 0, 0);B       /* synchronous close service (careful when using SYSPRV!) *//       tkptr->AccessFab.fab$l_fop &= ~FAB$M_ASY; .       tkptr->AccessFab.fab$l_fop |= FAB$M_DLT;*       sys$close (&tkptr->AccessFab, 0, 0);(       sys$setprv (0, &SysPrvMask, 0, 0);0       SsiAccessesFileClosed (&tkptr->AccessFab);
       return;e    }  0 This is what I would like to replace it with ...  /    /* "temporary" file, delete on closing it */p    if (rqptr->DeleteOnClose)    {.       tkptr->AccessFab.fab$l_fop |= FAB$M_DLT;(       sys$setprv (1, &SysPrvMask, 0, 0);    }+    tkptr->AccessFab.fab$l_fop |= FAB$M_ASY;a8    sys$close (&tkptr->AccessFab, &SsiAccessesFileClosed,9                                  &SsiAccessesFileClosed);i?    if (rqptr->DeleteOnClose) sys$setprv (0, &SysPrvMask, 0, 0);O  H Further; if I wanted to do a similar style of construct a little furtherH down (no code example at present), say with the ACP-QIO interface, would the same considerations apply?  3 My regrets if this is a little like drilling teeth..& Thanks again for your time (everyone).   ------------------------------  % Date: Mon, 29 Apr 2002 22:28:31 -0400V- From: JF Mezei <jfmezei.spamnot@videotron.ca>T Subject: Re: $QIO and SYSPRV, Message-ID: <3CCE014B.9F4F748C@videotron.ca>   Mark Daniel wrote:I > I'm not sure that I am (except perhaps when trying to explain myself) InI > was using AST delivery as the obvious end-point of the $QIO processing,-F > trying to ask for what duration from service initation (start point)B > until AST delivery (end point) do required privileges need to be
 > enabled.  M In the example you gave of SYS$CREATE, I think that Hoff gave a subtle answerp (perhaps he can confirm):t  L When you call SYS$CREATE, it checks the privs first, and then dispatches theN XQP call to do the actual stuff. XQP is at the driver level (is that correct?)M and thus is immune from "privileges". Once the XQP has completed the work, itaJ then queues your AST (if desired), and that AST executes in the context of2 your process with whatever privs your process has.  N Hoff explained that upon completion, the driver/XQP will check to make sure itK has write access to whatever memory buffers you supplied, but your "SYSPRV"  issue wouldn't be affected.r    / So based on what Hoff said, it would seem that:e   Setpriv(SYSPRV)iJ SYS$CREATE(new file in diretory you normally don't have access, with a AST completion routine)o Setpriv(NOsysprv)s  K would work fine. The only hitch is that your AST may execute with SYSPRV orkK without SYSPRV depending on if it executes before or after the disabling ofM the sysprv.    ------------------------------  # Date: Tue, 30 Apr 2002 02:40:46 GMTI2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman) Subject: Re: $QIO and SYSPRV: Message-ID: <Ounz8.54$Fc1.1323104@cacnews.cac.cpqcorp.net>  a In article <3CCDF5D3.1D40D2D5@wasd.vsm.com.au>, Mark Daniel <Mark.Daniel@wasd.vsm.com.au> writes:o2 :This is an *example* of the current construct ... :h :   if (!rqptr->DeleteOnClose) :   {t' :      /* asynchronous close service */ / :      tkptr->AccessFab.fab$l_fop |= FAB$M_ASY;-< :      sys$close (&tkptr->AccessFab, &SsiAccessesFileClosed,= :                                    &SsiAccessesFileClosed);0 :      return; :   }@ :   else :   {s; :      /* use SYSPRV to ensure deletion-on-close of file */ ) :      sys$setprv (1, &SysPrvMask, 0, 0);>C :      /* synchronous close service (careful when using SYSPRV!) */D0 :      tkptr->AccessFab.fab$l_fop &= ~FAB$M_ASY;/ :      tkptr->AccessFab.fab$l_fop |= FAB$M_DLT;n+ :      sys$close (&tkptr->AccessFab, 0, 0);n) :      sys$setprv (0, &SysPrvMask, 0, 0); 1 :      SsiAccessesFileClosed (&tkptr->AccessFab);  :      return; :   }  :c1 :This is what I would like to replace it with ...f :t0 :   /* "temporary" file, delete on closing it */ :   if (rqptr->DeleteOnClose)h :   { / :      tkptr->AccessFab.fab$l_fop |= FAB$M_DLT;d) :      sys$setprv (1, &SysPrvMask, 0, 0);t :   }t, :   tkptr->AccessFab.fab$l_fop |= FAB$M_ASY;9 :   sys$close (&tkptr->AccessFab, &SsiAccessesFileClosed,e: :                                 &SsiAccessesFileClosed);@ :   if (rqptr->DeleteOnClose) sys$setprv (0, &SysPrvMask, 0, 0); :nI :Further; if I wanted to do a similar style of construct a little further I :down (no code example at present), say with the ACP-QIO interface, would  :the same considerations apply?i    C   You are digging yourself a deep hole here.  Why not use subsystemdE   identifiers here?  If you do, then you use ACLs and do not need anynF   particular privileges to open or to close the file, so most of this H   whole discussion disappears into the proverbial puff of greasy orange G   smoke.  :-)  Privileges are comparatively powerful magic, and should tD   be avoided when simpler and more targeted access mechanisms exist.  E   Oh, and barring changes that can be possible via XABPRO and similarsD   mechanisms -- ownership changes and such that be written out to a B   file during close, or such cases as marking a file for delete onH   close when the caller does not have delete access to it -- the $close F   operation should not normally encounter a privilege error.  I'd not G   typically expect to need to enable any privileges in order to $close a
   a file.   ,   In other works, you can probably just use:  ,     RetStat = sys$close (&tkptr->AccessFab, 6       &SsiAccessesFileClosed, &SsiAccessesFileClosed);(     if (!$VMS_STATUS_SUCCESS( RetStat ))       failure-case...u  G   The example code also unfortunately omits the status checks that are oD   necessary on the system service calls -- this is one of the usual F   programming bugs, please see Ask The Wizard topic (1661) for a more E   complete list of common coding bugs...  AST re-entrancy issues and  G   sychronization bugs also figure prominently among the list of bugs...hF   I will assume this omission is merely to shorten up and to simplify A   the example code, and that the actual code involved checks the d'   condition values that are returned...h  H   And even if you get the RMS$_PRV error on the $close (due to an errantH   delete-on-close, for instance), you can use the error routine argumentG   on $close to log this or to go mess with the failure or to alter the n   process privileges.h  N  ---------------------------- #include <rtfaq.h> -----------------------------N       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  % Date: Tue, 30 Apr 2002 12:04:02 +0930n/ From: Mark Daniel <Mark.Daniel@wasd.vsm.com.au>s Subject: Re: $QIO and SYSPRV/ Message-ID: <3CCE029A.5A312806@wasd.vsm.com.au>    JF Mezei wrote: 
 8< snip 8<1 > So based on what Hoff said, it would seem that:  >  > Setpriv(SYSPRV)SL > SYS$CREATE(new file in diretory you normally don't have access, with a AST > completion routine)o > Setpriv(NOsysprv)o > M > would work fine. The only hitch is that your AST may execute with SYSPRV orrM > without SYSPRV depending on if it executes before or after the disabling of 
 > the sysprv.   G With my examples further down and this sceario, it is already executingaG at AST delivery level and so SYSPRV will be disabled before the RMS ASTeH is delivered (which is what I want), and more importantly to the rest ofE the application no other ASTs can be executed while SYSPRV is enabled B (which is what I want), and I get asynchronous (important for goodH granularity) RMS access via S:RWED into the bargain (which is again what$ I want).  A win-win-win scenario :^)   ------------------------------  % Date: Tue, 30 Apr 2002 12:40:13 +0930l/ From: Mark Daniel <Mark.Daniel@wasd.vsm.com.au>c Subject: Re: $QIO and SYSPRV/ Message-ID: <3CCE0B15.E8794F29@wasd.vsm.com.au>n   Hoff Hoffman wrote:v
 8< snip 8<E >   You are digging yourself a deep hole here.  Why not use subsystempG >   identifiers here?  If you do, then you use ACLs and do not need anyuG >   particular privileges to open or to close the file, so most of thisnI >   whole discussion disappears into the proverbial puff of greasy orange H >   smoke.  :-)  Privileges are comparatively powerful magic, and shouldF >   be avoided when simpler and more targeted access mechanisms exist.  F Let's assume for the moment that this needs to be performed on volumesC where protected subsystems are not enabled or where ACLs may not beeE applied.  After reading Programming Concepts and Security on them I'dcC agree that the mechanism is far more elegant but for a "grafted-on"iD rather than an "integrated" situation they may not be practicable or	 possible.t  iG >   Oh, and barring changes that can be possible via XABPRO and similar>E >   mechanisms -- ownership changes and such that be written out to aeD >   file during close, or such cases as marking a file for delete on  A There are none in this example.  It SYS$CREATE _CIFs a file usingo> ownership and protections inherited from the parent directory,E reads/write/updates (as required by the SYS$CREATE status returned) arF counter quadword, and then closes it again.  The _DLT only occur for a small (but necessary) subset.0  I >   close when the caller does not have delete access to it -- the $close2G >   operation should not normally encounter a privilege error.  I'd notT  B It's doesn't unless the _DLT is required.  Then it requires SYSPRV' because it's doesn't own the directory.y  H >   typically expect to need to enable any privileges in order to $close >   a file.l > . >   In other works, you can probably just use: > - >     RetStat = sys$close (&tkptr->AccessFab, 8 >       &SsiAccessesFileClosed, &SsiAccessesFileClosed);* >     if (!$VMS_STATUS_SUCCESS( RetStat )) >       failure-case...e > H >   The example code also unfortunately omits the status checks that areE >   necessary on the system service calls -- this is one of the usual8  F They are performed in the AST routine.  I'm assuming that because I amB providing an 'err' AST address it will always be delivered and the status appraised there.   G >   programming bugs, please see Ask The Wizard topic (1661) for a more F >   complete list of common coding bugs...  AST re-entrancy issues andI >   sychronization bugs also figure prominently among the list of bugs...t  G As mentioned it's happening at user-mode AST delivery level and so this E serializes the code in the application.  What is happening outside isiH happening outside.  If the deletion were to fail for some reason then itB just remains there.  Too bad.  There is a limit to which a generalC application can cater for unexpected events, and always a practical ? cutoff point.  What must *not* happen though is the applicationnH continues on, or process ASTs, with SYSPRV enabled.  That would be *bad*F because it may provide access to resources otherwise protected from to, the fairly bog-standard application account.  G >   I will assume this omission is merely to shorten up and to simplifynB >   the example code, and that the actual code involved checks the) >   condition values that are returned...g  D "Well, I don't think there is any question about it.  It can only beF attributable to human error.  This sort of thing has cropped up before+ and it has always been due to human error."e    Yup.  J >   And even if you get the RMS$_PRV error on the $close (due to an errantJ >   delete-on-close, for instance), you can use the error routine argumentH >   on $close to log this or to go mess with the failure or to alter the >   process privileges.p  E You seem to be confirming my comment about the checking of results ini the 'err' routine.  
 Thanks again.    ------------------------------  # Date: Tue, 30 Apr 2002 04:45:09 GMTl2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman) Subject: Re: $QIO and SYSPRV: Message-ID: <pjpz8.56$qc1.1315900@cacnews.cac.cpqcorp.net>  a In article <3CCE0B15.E8794F29@wasd.vsm.com.au>, Mark Daniel <Mark.Daniel@wasd.vsm.com.au> writes:f :Hoff Hoffman wrote: :8< snip 8<iF :>   You are digging yourself a deep hole here.  Why not use subsystemH :>   identifiers here?  If you do, then you use ACLs and do not need anyH :>   particular privileges to open or to close the file, so most of thisJ :>   whole discussion disappears into the proverbial puff of greasy orangeI :>   smoke.  :-)  Privileges are comparatively powerful magic, and should G :>   be avoided when simpler and more targeted access mechanisms exist.m :iG :Let's assume for the moment that this needs to be performed on volumesaD :where protected subsystems are not enabled or where ACLs may not beF :applied.  After reading Programming Concepts and Security on them I'dD :agree that the mechanism is far more elegant but for a "grafted-on"E :rather than an "integrated" situation they may not be practicable or$
 :possible.  @   If I had a choice of configuring and giving a hunk of code an C   identifier or giving the same hunk an ALL-class privilege, well, e@   that choice would take all of a clock tick or two to decide...    H :>   Oh, and barring changes that can be possible via XABPRO and similarF :>   mechanisms -- ownership changes and such that be written out to aE :>   file during close, or such cases as marking a file for delete on  : B :There are none in this example.  It SYS$CREATE _CIFs a file using? :ownership and protections inherited from the parent directory,rF :reads/write/updates (as required by the SYS$CREATE status returned) aG :counter quadword, and then closes it again.  The _DLT only occur for a  :small (but necessary) subset. :rJ :>   close when the caller does not have delete access to it -- the $closeH :>   operation should not normally encounter a privilege error.  I'd not :rC :It's doesn't unless the _DLT is required.  Then it requires SYSPRV ( :because it's doesn't own the directory.  1   Ownership is not relevent.  Access is relevent.p  I :>   typically expect to need to enable any privileges in order to $closey :>   a file. :> a/ :>   In other works, you can probably just use:  :> e. :>     RetStat = sys$close (&tkptr->AccessFab,9 :>       &SsiAccessesFileClosed, &SsiAccessesFileClosed);e+ :>     if (!$VMS_STATUS_SUCCESS( RetStat ))  :>       failure-case... :> tI :>   The example code also unfortunately omits the status checks that arelF :>   necessary on the system service calls -- this is one of the usual :eG :They are performed in the AST routine.  I'm assuming that because I amsC :providing an 'err' AST address it will always be delivered and them :status appraised there.     Yep.  H :>   programming bugs, please see Ask The Wizard topic (1661) for a moreG :>   complete list of common coding bugs...  AST re-entrancy issues and J :>   sychronization bugs also figure prominently among the list of bugs... : H :As mentioned it's happening at user-mode AST delivery level and so this* :serializes the code in the application.    E   That was a GENERAL recommendation.  Go read the topic.  The list ofy!   bugs may or may not apply here.   K :>   And even if you get the RMS$_PRV error on the $close (due to an erranteK :>   delete-on-close, for instance), you can use the error routine argumentnI :>   on $close to log this or to go mess with the failure or to alter theo :>   process privileges. :dF :You seem to be confirming my comment about the checking of results in :the 'err' routine.e  G   I'm also trying to get you away from the chainsaw that is SYSPRV; to tF   encourage you to use simple hand-tools entirely appropriate for the 4   task at hand, tools such as subsystem identifiers.    N  ---------------------------- #include <rtfaq.h> -----------------------------N       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  # Date: Mon, 29 Apr 2002 19:49:14 GMTy4 From: "Matt Muggeridge" <Matt.Muggeridge@compaq.com>/ Subject: Re: (GCC) GNU C/C++ compiler for Alphai? Message-ID: <_shz8.65299$o66.199532@news-server.bigpond.net.au>S   Oh, I hadn't realised that.    Thanks,  Matt.y   --= -------------------------------------------------------------( OpenVMS TCP/IP Engineering Enterprise Computing Group Compaq Computer Corporation  Gold Coast, AUSTRALIAl= -------------------------------------------------------------     G "Craig A. Berry" <craig.berry@SignalTreeSolutions.com> wrote in message-7 news:7f15589f.0204290800.7ac3fed4@posting.google.com...dA > "Matt Muggeridge" <Matt.Muggeridge@compaq.com> wrote in messagel; news:<hi%y8.62030$o66.188110@news-server.bigpond.net.au>...-G > > GCC and other great tools can be found in the "GNV - GNU's Not VMS"0 package,$ > > at  http://gnv.sourceforge.net/. >YE > But there are no GNU compilers in that package, at least not that I G > can see.  The GCC tool is just a wrapper that actually uses Compaq C.l >)K > > You may also be interested in the related product for porting UNIX appss toJ > > OpenVMS.  It is called the OpenVMS Porting Library, and is located at: > >CC > > http://www.openvms.compaq.com/openvms/products/ips/porting.htmli   ------------------------------    Date: 29 Apr 2002 15:49:34 -07003 From: baltimoreplayboy@yahoo.com (BaltimorePlayboy)  Subject: AIM HACK?= Message-ID: <e179cdf4.0204291449.134a13e3@posting.google.com>u  B Can someone teach/tell me how to push someone off of AIM using the= buffer underrun thing I've been hearing about? It seems quiteaA effective for online A-holes, but I dont know how to do it. (Evenh happened to me once) Thanks!   ------------------------------  # Date: Mon, 29 Apr 2002 18:07:10 GMT(2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)F Subject: Re: AlphaServer 1000 Hobbyist (was: Re: What have I got here)8 Message-ID: <iZfz8.39$vZ.667793@cacnews.cac.cpqcorp.net>  X In article <3cc84570$1@obsidian.gov.bc.ca>, "Alder" <MUNDDGNTDYTV@spammotel.com> writes:  C   Please pick a subject line that does not look like spam, and thatiC   is relevent to the question -- this makes it most likely that youhA   will attract the attention of somebody that knows the answer tot@   the question, and also avoids ending up in the spam-killfiles.	   Thanks!c  M :It was to my great surprise today that an Alpha machine came into my hands -eJ :gratis.  I'm hoping someone here can help me understand just what is I'veJ :got.  There were no documents with the machine, but I did manage to gleanM :from the source that it is a circa-1996 Digital Alpha Server 1000 4/266 withm :Windows NT installed.  F   This is a reasonable hobbyist box, with an EV4-class microprocessor    running at 266 MHz.o  K :The reason I'm asking here is that I've recently been toying with the ideabJ :of learning more about VMS (I'm a lowly user of an Oracle Rdb database orH :two running on OpenVMS for Alpha).  Just to show you how much I have toL :learn I'll ask this question: Does the presence of Windows NT mean that theH :hardware on this thing is incapable of running an OpenVMS system on it?  H   No -- though some platforms are designated as being for use only with 3   Microsoft Windows boxes, this is not one of them.r  F   Please see the OpenVMS FAQ for the details on switching from the ARCD   or AlphaBIOS console (used by Windows NT) to the SRM console (used?   by OpenVMS), and for details on the OpenVMS hobbyist program.   F   OpenVMS will run on this AlphaServer, but may or may not support allG   of the I/O widgets present -- you need to ensure that the contents of H   the buses -- particularly the graphics controller -- are supported by *   the target operating system and version.    N  ---------------------------- #include <rtfaq.h> -----------------------------N       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  % Date: Tue, 30 Apr 2002 02:18:21 -0000 / From: Michael Zarlenga <zarlenga@conan.ids.net>M4 Subject: Re: Another Microsoft breaking of standards/ Message-ID: <ucrvndd1mouued@corp.supernews.com>;  . JF Mezei <jfmezei.spamnot@videotron.ca> wrote:
 : quick note:   M : I have noticed that more and more web sites now serve their contents with aoM : charset "windows-1252" instead of the expected iso latin 1 one  iso-8859-1.a  P : I suspect both are quite similar, but this is a real pain that microsoft can'tK : use standard character sets. Heck, I even got an email from someone usingrM : outlook express on a mac and it was sent with a charset of windows-1252 :-(e  7 Ok, I'll bite: What written standard does this violate?t   -- p -- Mike Zarlenga  '       A good friend will help you move.h.       A best friend will help you move a body.   ------------------------------  % Date: Tue, 30 Apr 2002 02:19:31 -0000m/ From: Michael Zarlenga <zarlenga@conan.ids.net> 4 Subject: Re: Another Microsoft breaking of standards/ Message-ID: <ucrvpj4ccump07@corp.supernews.com>   . JF Mezei <jfmezei.spamnot@videotron.ca> wrote:I : has it ever occured to you that it is MUCH SIMPLER to have all software L : exchange data in a very small number of character sets as opposed to everyR : piece of software having to be aware of all possible character sets that exist ?  1 Sure, that's why the world only uses 6-bit ASCII.e   -- o -- Mike Zarlenga  '       A good friend will help you move.b.       A best friend will help you move a body.   ------------------------------  # Date: Mon, 29 Apr 2002 20:41:22 GMT * From: "Bill Todd" <billtodd@metrocast.net>+ Subject: Anticipating the HP court decision C Message-ID: <Sdiz8.142747$ro5.12654012@bin6.nnrp.aus1.giganews.com>s  E OK, I was willing to wait until the markets closed today (something IaJ suspected the court might have wished to do) but am now getting impatient.J So (in Terry's spirit of the new year - since whichever way it goes likelyL will signal the start of a new era for the DEC heritage) I will, even though< not being anything like a lawyer (thank God) make a Fearless Prognostication:  J I think the judge will order a new shareholder vote and give Carly & Co. aE stern slap on the wrist.  I think he will do it because HP managementeB *grossly* crossed over the line of its fiduciary responsibility inL attempting to 'sell' the merger to its shareholders (and sticking resolutelyI to its initial rosy predictions despite growing internal concerns) rather-K than laying it out in reasonable detail, both pro and con, and letting saidwK shareholders, as the owners it's responsible to, decide on the merits.  And-L because HP subtly but clearly strong-armed Deutsche Bank to change its 3 - 1E vote against the merger into a 4 - 1 vote for the merger in the finaluH hours - again, an attempt to influence the people it should have insteadK just been informing to the best of its ability.  And because the outcome of.K the vote was still so tight that such actions could easily have changed itsuH result:  had it been a walk for Carly's crowd, I suspect she still wouldE have received the slap, but possibly not the demand for another vote.e   ------------------------------  % Date: Mon, 29 Apr 2002 20:54:42 -0400y- From: JF Mezei <jfmezei.spamnot@videotron.ca>n/ Subject: Re: Anticipating the HP court decisioni, Message-ID: <3CCDEB3A.CB1128E1@videotron.ca>   Bill Todd wrote:L > I think the judge will order a new shareholder vote and give Carly & Co. a > stern slap on the wrist.    I No way. Worse case scenario, the judge will declare that the Deutche Bank K votes are null because Deutche Bank was hired to perform some merger duties @ and was no longer "neutral" party representing its shareholders.  N And it won't change anything because Deutche Bank doesn't hold enough votes toF change the outcome of the merger votes. (nullifiying their votes won't increase the NO vote).  I As far as Carly using makeup to make the merger look better, every bloodyPL merger proposal was done with heaps of makeup to make it look palatable, andK one might even say that often, the top CEOs don't even know how much makeups3 their underlings applied. This one is no different.A  L Does anyone *really* believe any of the Carly promises ? Those who votes yesJ did so because they had some vested interest to vote yes, not because theyJ believed that carly would be abnle to realise all those promised benefits.   ------------------------------  % Date: Mon, 29 Apr 2002 21:32:29 -0500a& From: "Dave Gudewicz" <k9jdk@arrl.net>/ Subject: Re: Anticipating the HP court decisionh8 Message-ID: <3cce023b$0$3568$1dc6e903@news.corecomm.net>  J I say if H-P really did do something wrong, they should pay or be punishedJ with whatever the law of the land prescribes.  If it means that the merger is a no-go, then so be it.  H Now I'll replace my "I live in a perfect world" tee-shirt and replace itL with "apply liberal dose of reality".  Story changes to whoever has the bestJ lawyer/s money can buy, usually (not always) wins, guilty of wrongdoing or not.  2 Lets see which one wins in this case.  Or will we?   Dave...p  5 "Bill Todd" <billtodd@metrocast.net> wrote in messaget= news:Sdiz8.142747$ro5.12654012@bin6.nnrp.aus1.giganews.com...eG > OK, I was willing to wait until the markets closed today (something InL > suspected the court might have wished to do) but am now getting impatient.L > So (in Terry's spirit of the new year - since whichever way it goes likelyG > will signal the start of a new era for the DEC heritage) I will, even  though> > not being anything like a lawyer (thank God) make a Fearless > Prognostication: >tL > I think the judge will order a new shareholder vote and give Carly & Co. aG > stern slap on the wrist.  I think he will do it because HP managementyD > *grossly* crossed over the line of its fiduciary responsibility inC > attempting to 'sell' the merger to its shareholders (and stickingT
 resolutelyK > to its initial rosy predictions despite growing internal concerns) rathernH > than laying it out in reasonable detail, both pro and con, and letting saidH > shareholders, as the owners it's responsible to, decide on the merits. AndsL > because HP subtly but clearly strong-armed Deutsche Bank to change its 3 - 1rG > vote against the merger into a 4 - 1 vote for the merger in the finalAJ > hours - again, an attempt to influence the people it should have insteadJ > just been informing to the best of its ability.  And because the outcome ofI > the vote was still so tight that such actions could easily have changedA itslJ > result:  had it been a walk for Carly's crowd, I suspect she still wouldG > have received the slap, but possibly not the demand for another vote.r >m >e   ------------------------------    Date: 29 Apr 2002 23:01:56 -0500+ From: young_r@encompasserve.org (Rob Young)t/ Subject: Re: Anticipating the HP court decision 3 Message-ID: <ssMC9uPDrFX7@eisner.encompasserve.org>o  p In article <Sdiz8.142747$ro5.12654012@bin6.nnrp.aus1.giganews.com>, "Bill Todd" <billtodd@metrocast.net> writes:G > OK, I was willing to wait until the markets closed today (something I L > suspected the court might have wished to do) but am now getting impatient.L > So (in Terry's spirit of the new year - since whichever way it goes likelyN > will signal the start of a new era for the DEC heritage) I will, even though> > not being anything like a lawyer (thank God) make a Fearless > Prognostication: > L > I think the judge will order a new shareholder vote and give Carly & Co. a > stern slap on the wrist.    " 	I think I agree with this fellow:  M "There's been a lot of dirty linen aired in this thing," said Needham analystlO Charlie Wolf. But "I think Hewlett's got a very, very small chance to win. It'slO not impossible, but boy, it's low." Wolf estimates Hewlett has only a 5% to 10% , chance of getting a new vote on the merger.   C 	After all, they only thing Hewlett has going for him is rumors and  	a certain wiftiness.n  ) http://news.com.com/2100-1001-891243.htmlY  I Under cross-examination by HP attorney Steven Schatz, Hewlett had to haveeM several questions repeated, and he frequently answered with a simple "yes" or J "no." At one point, Hewlett said, "I'm sorry, I flipped out while you were asking the question."   K When asked whether Hewlett believed HP had misrepresented various facts, heoN responded, "I believe HP's numbers are unattainable, but the rest is rumors."   C 	No smoking gun.  Shoot, a stained dress couldn't shame a Presidents; 	out of office... there is a whole lot less than that here.r   				Roba   ------------------------------    Date: 29 Apr 2002 12:19:36 -0500- From: koehler@encompasserve.org (Bob Koehler)t5 Subject: Re: Anyone using CHARON-VAX for development?o3 Message-ID: <RxlXpN5YT59N@eisner.encompasserve.org>o  T In article <3CCD648B.2040901@iee.org>, "antonio.carlini" <arcarlini@iee.org> writes: > Bob Koehler wrote: > F >>    Poking a HALT in is convenient if you forgot to boot your 11/780; >>    with XDELTA.  I've seen others poke in a BRB to self.c >  > ! > I don't have the context of theo# > rest of the thread, but XDELTA ism > dynamically loadable. One of# > the DECnet debug tools can do it.o  C    In VMS 2 ?  (At the time).  I haven't seen anything in XDELTA ona>    doing a dynamic load, which document is that written up in?   ------------------------------  % Date: Mon, 29 Apr 2002 13:00:38 -0400r0 From: "Syltrem" <syltremspammenot@videotron.com> Subject: Re: Can't edit a file5 Message-ID: <eZez8.10533$a04.47290@tor-nn1.netcom.ca>e   $ SHOW SYMBOL EDIT may provide the answer   --   SyltremtI http://pages.infinit.net/syltrem (OpenVMS related web site - en franais)d> To reply to myself directly, remove .spammenot from my address  A "David Lee" <phongle@kornet.net> a crit dans le message de news:t  aajc1i$nu9$1@news1.kornet.net...H > Why can't I invoke any of the editor.  I try to do the following on my" > workstation with VMS 7.2 loaded./ > from the command line, I issued the followingo > $edit/edt login.com. > or > $edit/tpu login.comn! > But I kept getting this messagep >l> > %DCL-W-IOV...  check validity, spelling, misplacement \edit\ > / > What does that mean?  Am I missing something?a >  >t >e >    ------------------------------  % Date: Tue, 30 Apr 2002 02:23:36 -0000F/ From: Michael Zarlenga <zarlenga@conan.ids.net>! Subject: Re: Can't edit a file/ Message-ID: <ucs018pgmits6d@corp.supernews.com>e  % David Lee <phongle@kornet.net> wrote: H : Why can't I invoke any of the editor.  I try to do the following on my" : workstation with VMS 7.2 loaded./ : from the command line, I issued the followinge : $edit/edt login.come : or : $edit/tpu login.coms! : But I kept getting this messager  I Is this a system mean for user use?  At a prior job we routinely disabledpG edit, backup, etc on production machines since careless use during peak-L transaction times could stealenough cycyles to cause unacceptable processing delays.f   Speak to your OPERATOR.    -- p -- Mike Zarlenga  '       A good friend will help you move.e.       A best friend will help you move a body.   ------------------------------    Date: 29 Apr 2002 11:46:08 -0500- From: Kilgallen@SpamCop.net (Larry Kilgallen)u Subject: Re: Compaq HPS times$3 Message-ID: <gFpwDF4TtKd2@eisner.encompasserve.org>n  d In article <JHdz8.127744$nc.18243036@typhoon.tampabay.rr.com>, "John N." <JNixon@cfl.rr.com> writes:M > I just got an e-mail touting  the new Compaq HPS (High Performance Systems)mK > Times.  This will replace the OpenVMS Times, and the Tru64 times and willo > include information on Linux.c  D I looked at it with Adobe Acrobat, and what presumably were supposed* to be photographs were just black squares.   They should stick with text.  ? Knowing they won't, I should not bother looking at such things.-   ------------------------------  % Date: Mon, 29 Apr 2002 18:58:49 +0200 9 From: Jan-Erik =?iso-8859-1?Q?S=F6derholm?= <aaa@aaa.com>r Subject: Re: Compaq HPS times-' Message-ID: <3CCD7BC9.77D0CC3E@aaa.com>    I'm using Acrobat reader 5.0.dC All picures both look OK both on screen and after printed on paper.p   Better check your setup...   Jan-Erik Sderholm.u     Larry Kilgallen wrote: > f > In article <JHdz8.127744$nc.18243036@typhoon.tampabay.rr.com>, "John N." <JNixon@cfl.rr.com> writes:O > > I just got an e-mail touting  the new Compaq HPS (High Performance Systems)-M > > Times.  This will replace the OpenVMS Times, and the Tru64 times and wille! > > include information on Linux.o > F > I looked at it with Adobe Acrobat, and what presumably were supposed, > to be photographs were just black squares. >  > They should stick with text. > A > Knowing they won't, I should not bother looking at such things.m   ------------------------------    Date: 29 Apr 2002 20:34:22 -0500- From: Kilgallen@SpamCop.net (Larry Kilgallen)g Subject: Re: Compaq HPS timest3 Message-ID: <twipGPtw1SYN@eisner.encompasserve.org>V  c In article <3CCD7BC9.77D0CC3E@aaa.com>, Jan-Erik =?iso-8859-1?Q?S=F6derholm?= <aaa@aaa.com> writes:a > I'm using Acrobat reader 5.0.eE > All picures both look OK both on screen and after printed on paper.  >  > Better check your setup...  B I don't need to.  I know it is not 5.0.  Somehow this problem does not occur with text.   ------------------------------    Date: 29 Apr 2002 17:18:23 -0700( From: bob@instantwhip.com (Bob Ceculski); Subject: Re: comparison chart: pros and cons of WWW servers0< Message-ID: <d7791aa1.0204291618.96975bb@posting.google.com>  | Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com> wrote in message news:<01KH12X67B068Y7GCV@sysdev.deutsche-boerse.com>...G > Does anyone have (or can anyone point me to) a GENERAL comparison of m? > various HTTP servers for VMS (OSU, CSWS, WASD, Purveyor,...)?   C also take a test drive of the rsm ... another strength of purveyor.-F it allows someone w/limited knowledge (i.e. operators) to make changesF easily to a web site w/o getting lost in .conf jungle ... anything youB need to do (except creating error pages) can be done from here ...E error pages can be referenced by a simple edit of the config file ...N  % http://vms.process.com/~rsm/main.htmla   ------------------------------  # Date: Mon, 29 Apr 2002 19:57:48 GMTv4 From: "Matt Muggeridge" <Matt.Muggeridge@compaq.com>; Subject: Re: COPY/RCP fails to remote node (UCX4.2 VMS 7.1) ? Message-ID: <0Bhz8.65301$o66.199676@news-server.bigpond.net.au>-  K I don't know why you are seeing this, but would much rather see you runningiI on a more current version of TCP/IP.  Upgrading to at least V5.1 + latesteH ECO, (V5.3 is about to hit the streets too - there was no V5.2 release).  ; To offer some alternatives, you might try FTP or DECnet/IP:-  ; o copy/ftp login.com host22"username password"::login10.comeJ o Configure DECnet over IP and use same functionality you had in the past.   Matt.n   --= -------------------------------------------------------------e OpenVMS TCP/IP Engineering Enterprise Computing Group Compaq Computer CorporationR Gold Coast, AUSTRALIAn= -------------------------------------------------------------o    8 "Simon Brown" <simonb@colourcare.co.uk> wrote in message" news:3ccd5370$1@news.star.co.uk...C > Attempting to use COPY/RCP command to copy file to remote networku > using UCX4.2 > command like;a >e6 > copy /rcp login.com host22::users:[brown]login10.com >n > fails on local node with >>e7 > INTERnet ACP AUXS failure  Status = %LOGIN-F-NOTVALIDC > . > On remote node I get an OPCOM message saying& > Message from user INTERnet on HOST22E > INTERnet ACP REXEC Accept Request from Host: 10.10.1.1        Port:h > 1738 >v > UCX>: > %%%%%%%%%%%  OPCOM  29-APR-2002 13:38:44.32  %%%%%%%%%%%* > Message from user AUDIT$SERVER on HOST22D > Security alarm (SECURITY) and security audit (SECURITY) on HOST22, > system id: 1524a1 > Auditable event:          Network login failure-3 > Event time:               29-APR-2002 13:38:44.31-$ > PID:                      0000FAE6+ > Process name:             UCX$REXE_BG2590   > Username:                 ..$ > Process owner:            [SYSTEM]
 > Image name: 3 > HOST22$DKA0:[SYS0.SYSCOMMON.][SYSEXE]LOGINOUT.EXElA > Status:                   %LOGIN-F-NOTVALID, user authorizationp	 > failure8 > # > Notice the User name is garbage ?nF > Also if I specify the username with /USER=BROWN I get the exact same > response.  >rF > I wanted to use this approach to move from DECNET to TCPIP as I haveE > 100s jobs to change which are already COPY type commands and also IiA > believe it keeps the VMS file attributes whereas RCP changes toe	 > STREAM.x >t > Any Ideas anyone ? >8 >1   ------------------------------  % Date: Mon, 29 Apr 2002 22:58:22 +0100o. From: "Simon Brown" <simonjm.brown@virgin.net>; Subject: Re: COPY/RCP fails to remote node (UCX4.2 VMS 7.1)nB Message-ID: <qkjz8.22510$%k.3882759@news6-win.server.ntlworld.com>   Matt,W  L Ideal world etc I would, But I have a couple of weeks to get 8 sites runningJ on TCPIP to connect our network to new owners network, which is ALL IP and through there strict firewall.: Using DECNET phase 4 so I believe cannot do DECNET over IPH Also do not have time or resource at the moment to go about upgrading toL TCPIP 5.1 and undergoing the learning curve for new syntax and modifying allG my com jobs, programs with submits and spawns of copy and submit tasks.   * I really need to isolate this problem now.J The RCP command in itself works to a fashion (Different syntax and changesL the file format to stream) but so much code with the traditional DECNET COPY; that would make an already huge job a little more bearable.-  9 P.S. This from Home PC so not name probably not matching.e  , Thx for any info you may be able to provide.   Simon-    ? "Matt Muggeridge" <Matt.Muggeridge@compaq.com> wrote in messager9 news:0Bhz8.65301$o66.199676@news-server.bigpond.net.au...tE > I don't know why you are seeing this, but would much rather see you. running.K > on a more current version of TCP/IP.  Upgrading to at least V5.1 + latestoJ > ECO, (V5.3 is about to hit the streets too - there was no V5.2 release). >a= > To offer some alternatives, you might try FTP or DECnet/IP:  > = > o copy/ftp login.com host22"username password"::login10.comhL > o Configure DECnet over IP and use same functionality you had in the past. >  > Matt.  >g > --? > -------------------------------------------------------------  > OpenVMS TCP/IP Engineering > Enterprise Computing Group > Compaq Computer Corporation  > Gold Coast, AUSTRALIAe? > -------------------------------------------------------------e >r > : > "Simon Brown" <simonb@colourcare.co.uk> wrote in message$ > news:3ccd5370$1@news.star.co.uk...E > > Attempting to use COPY/RCP command to copy file to remote networkr > > using UCX4.2 > > command like;l > > 8 > > copy /rcp login.com host22::users:[brown]login10.com > >  > > fails on local node with >>o9 > > INTERnet ACP AUXS failure  Status = %LOGIN-F-NOTVALIDe > >e0 > > On remote node I get an OPCOM message saying( > > Message from user INTERnet on HOST22G > > INTERnet ACP REXEC Accept Request from Host: 10.10.1.1        Port:i > > 1738 > >u > > UCX>< > > %%%%%%%%%%%  OPCOM  29-APR-2002 13:38:44.32  %%%%%%%%%%%, > > Message from user AUDIT$SERVER on HOST22F > > Security alarm (SECURITY) and security audit (SECURITY) on HOST22, > > system id: 1524t3 > > Auditable event:          Network login failure-5 > > Event time:               29-APR-2002 13:38:44.31r& > > PID:                      0000FAE6- > > Process name:             UCX$REXE_BG2590c" > > Username:                 ..& > > Process owner:            [SYSTEM] > > Image name: 5 > > HOST22$DKA0:[SYS0.SYSCOMMON.][SYSEXE]LOGINOUT.EXEaC > > Status:                   %LOGIN-F-NOTVALID, user authorizationd > > failurer > >t% > > Notice the User name is garbage ?uH > > Also if I specify the username with /USER=BROWN I get the exact same
 > > response.t > >rH > > I wanted to use this approach to move from DECNET to TCPIP as I haveG > > 100s jobs to change which are already COPY type commands and also IsC > > believe it keeps the VMS file attributes whereas RCP changes to  > > STREAM.a > >t > > Any Ideas anyone ? > >, > >y >o >P   ------------------------------  # Date: Tue, 30 Apr 2002 01:49:30 GMT,4 From: "Matt Muggeridge" <Matt.Muggeridge@compaq.com>; Subject: Re: COPY/RCP fails to remote node (UCX4.2 VMS 7.1)o? Message-ID: <KKmz8.65879$o66.202038@news-server.bigpond.net.au>a  H In the long term, I urge you to use the V5.1 ECO latest release.  If youI lock down with UCX V4.2, then you may be locking the 8 sites into supporte issues.   J > Also do not have time or resource at the moment to go about upgrading toJ > TCPIP 5.1 and undergoing the learning curve for new syntax and modifying allcI > my com jobs, programs with submits and spawns of copy and submit tasks.c  H Provided you use documented features, then the upgrade from V4.2 to V5.1I should be seamless.  No need to change command syntax or modify com jobs.tC There's a learning curve in as much as V5.1 offers new and excitinglI features, but you don't need to undertake that until you are ready to useo the new features.c  < > Using DECNET phase 4 so I believe cannot do DECNET over IP   That's correct.g   >(Different syntax and changesI > the file format to stream) but so much code with the traditional DECNET  COPY= > that would make an already huge job a little more bearable.5  E This is why I suggested FTP.  You can use exactly the same syntax forhK COPY/FTP as you do for DECnet COPY.  Also for VMS<->VMS transfers, the file ( attributes are maintained automagically.  , > I really need to isolate this problem now.  L Then you will need to raise the issue with the CSC as they have the means toK track historical problems.  I'm not sure where V4.2 is with regards supportk contracts...   Matt.t   --= -------------------------------------------------------------n OpenVMS TCP/IP Engineering Enterprise Computing Group Compaq Computer Corporationn Gold Coast, AUSTRALIAr= -------------------------------------------------------------p    9 "Simon Brown" <simonjm.brown@virgin.net> wrote in messagee< news:qkjz8.22510$%k.3882759@news6-win.server.ntlworld.com... > Matt,n >nF > Ideal world etc I would, But I have a couple of weeks to get 8 sites runningiL > on TCPIP to connect our network to new owners network, which is ALL IP and  > through there strict firewall.< > Using DECNET phase 4 so I believe cannot do DECNET over IPJ > Also do not have time or resource at the moment to go about upgrading toJ > TCPIP 5.1 and undergoing the learning curve for new syntax and modifying alleI > my com jobs, programs with submits and spawns of copy and submit tasks.  >e, > I really need to isolate this problem now.L > The RCP command in itself works to a fashion (Different syntax and changesI > the file format to stream) but so much code with the traditional DECNETn COPY= > that would make an already huge job a little more bearable.e >9; > P.S. This from Home PC so not name probably not matching.b >u. > Thx for any info you may be able to provide. >o > Simong >D >sA > "Matt Muggeridge" <Matt.Muggeridge@compaq.com> wrote in messager; > news:0Bhz8.65301$o66.199676@news-server.bigpond.net.au...dG > > I don't know why you are seeing this, but would much rather see youl	 > running(F > > on a more current version of TCP/IP.  Upgrading to at least V5.1 + latestL > > ECO, (V5.3 is about to hit the streets too - there was no V5.2 release). > >i? > > To offer some alternatives, you might try FTP or DECnet/IP:I > >,? > > o copy/ftp login.com host22"username password"::login10.com H > > o Configure DECnet over IP and use same functionality you had in the past.d > >t	 > > Matt.C > >& > > --A > > -------------------------------------------------------------  > > OpenVMS TCP/IP Engineering > > Enterprise Computing Group > > Compaq Computer Corporationu > > Gold Coast, AUSTRALIAvA > > -------------------------------------------------------------  > >  > >e< > > "Simon Brown" <simonb@colourcare.co.uk> wrote in message& > > news:3ccd5370$1@news.star.co.uk...G > > > Attempting to use COPY/RCP command to copy file to remote networke > > > using UCX4.2 > > > command like;2 > > >Y: > > > copy /rcp login.com host22::users:[brown]login10.com > > >a! > > > fails on local node with >>t; > > > INTERnet ACP AUXS failure  Status = %LOGIN-F-NOTVALID  > > >p2 > > > On remote node I get an OPCOM message saying* > > > Message from user INTERnet on HOST22I > > > INTERnet ACP REXEC Accept Request from Host: 10.10.1.1        Port:n
 > > > 1738 > > >s
 > > > UCX>> > > > %%%%%%%%%%%  OPCOM  29-APR-2002 13:38:44.32  %%%%%%%%%%%. > > > Message from user AUDIT$SERVER on HOST22H > > > Security alarm (SECURITY) and security audit (SECURITY) on HOST22, > > > system id: 1524-5 > > > Auditable event:          Network login failureo7 > > > Event time:               29-APR-2002 13:38:44.31n( > > > PID:                      0000FAE6/ > > > Process name:             UCX$REXE_BG2590 $ > > > Username:                 ..( > > > Process owner:            [SYSTEM] > > > Image name:e7 > > > HOST22$DKA0:[SYS0.SYSCOMMON.][SYSEXE]LOGINOUT.EXE E > > > Status:                   %LOGIN-F-NOTVALID, user authorizationn
 > > > failurec > > > ' > > > Notice the User name is garbage ?eJ > > > Also if I specify the username with /USER=BROWN I get the exact same > > > response.n > > >oJ > > > I wanted to use this approach to move from DECNET to TCPIP as I haveI > > > 100s jobs to change which are already COPY type commands and also I-E > > > believe it keeps the VMS file attributes whereas RCP changes tol
 > > > STREAM.p > > >@ > > > Any Ideas anyone ? > > >t > > >  > >  > >g >D >e   ------------------------------    Date: 29 Apr 2002 14:29:44 -0700& From: jordan@ccs4vms.com (Rich Jordan)M Subject: dBASE IV package/license for VMS VAX, product status?  Transferable?y< Message-ID: <cc5619f2.0204291329.59d2402@posting.google.com>  = We have a complete dBASE IV V1.1 for VAX/VMS package that wasaD purchased through some kind of promotion back in 1991.  It was neverF registered or installed, and the license PAK is a generic (not printedD to/for a specific customer) p/n QL-YE1A9-PB, DBASEIV-USER, issued by DEC, produced by Ashton-Tate.n  E I've tried calling the licensing support folks, but the hold time and,A my available time are at odds, so asking here just in case anyones: knows; I'll contact licensing later in the week if needed.  E Is this product transferable?  Would it need to be transferred, since<@ (although it was purchased) it was never registered or installedD (there is registration paperwork with the package).  Were there ever= any updates available?  When did the product cease support or B continued sales?  Searching Compaq's site can't find the SPD, even with the number (32.36.00).   @ Basically we're cleaning up in preparation for a move later thisD summer.  I've been tasked to move or pitch anything that hasn't beenB needed in the last few years; I hate throwing out usable VMS itemsD (though I have no real experience as to the actual usability of thisF package), and would like to know if this can legally be sold (probablyC on Ebay), and with or without the PAK.  Since its not an actual DEC A product, and Ashton-Tate no longer exists, and Borland won't talk.C about dBASE, and the company that now holds the license for the DOS0F versions has no information at all on VMS or dBASE for same and so far? hasn't answered queries about licensing, and since apparently a F federal court invalidated Borland's copyright on dBASE due to it beingF derived from JPL code..... for all I know it may be public domain now.  Yeah, right... :)   Rich Jordanm   ------------------------------  % Date: Mon, 29 Apr 2002 14:36:10 -0400l0 From: "Syltrem" <syltremspammenot@videotron.com>4 Subject: Re: DCL labels out of scope after ON ERROR?5 Message-ID: <Tmgz8.10545$a04.47345@tor-nn1.netcom.ca>y  D >Also, DEC BASIC will yell at you at compile time if you try that in >BASIC.c No it won't!  6 It will only complain if you GOTO INSIDE an IF...ENDIFJ And it makes sens that it does. Same way as it is complaining GOTOing into WHILE, UNTIL, FOR...  J GOTOing out of a n IF block (or any other block) is not a problem with the( exception of GOSUB which is NOT a block.   --   SyltremrI http://pages.infinit.net/syltrem (OpenVMS related web site - en franais) > To reply to myself directly, remove .spammenot from my address  L "David J. Dachtera" <djesys.nospam@fsi.net> a crit dans le message de news: 3CCB6BF3.5CC44F59@fsi.net... > Chris Olive wrote: > >eG > > Actually, yes the block it calls from is inside an "If F$Mode .EQS.4H "INTERACTIVE"" if-then-endif block.  So...  That's probably it.  Thanks. >tE > In general, I recommend that you avoid GOTO-ing out of an IF block.  >e; > Newer DCL doesn't get as confued as earlier versions did.  >nE > Also, DEC BASIC will yell at you at compile time if you try that in- > BASIC. >-9 > GOSUB or CALL within an IF block is always good, AFAIK.l >@E > ...and yes, a missing ENDIF can lead to all kinds of debugging fun!- >  > -- > David J. Dachteraa > dba DJE Systemsa > http://www.djesys.com/ >.* > Unofficial Affordable OpenVMS Home Page:! > http://www.djesys.com/vms/soho/    ------------------------------  # Date: Tue, 30 Apr 2002 02:36:44 GMT 1 From: "Terry C. Shannon" <terryshannon@attbi.com>.C Subject: Re: DECUS/EncompassUS Survey -- no longer assuming Windowss; Message-ID: <0rnz8.56401$%s3.22511873@typhoon.ne.ipsvc.net>a  > "Michael Austin" <maustin@firstdbasource.com> wrote in message, news:3CC9AEA7.4916823A@firstdbasource.com... > "Terry C. Shannon" wrote:g > > > > > "Larry Kilgallen" <Kilgallen@SpamCop.net> wrote in message1 > > news:IEnbX4I5irvC@eisner.encompasserve.org...sF > > > DECUServe discussion from Richard Copeland quoted some notice ofI > > > a survey, and for once I liked what I saw, both for privacy and forT > > > question design: > >,I > > Well, we tried! ;-} Many lessons learned from the Advocacy Tool, etc.e Not toJ > > mention the www.openvms.org surveys. We did away with evil cookies and went. > > to IP addresses to minimize double-voting. >nH > Oh good, then I can vote, reconnect my DSL (PPPoE) which gets me a new > IP address and vote again :)  K But of course you can! In fact, you can disconnect, reconnect, and vote allmF the live-long day. Gee, I might be tempted to do the same thing, but I( unfortunately have to work for a living.  D There's a Better Answer: requiring the use of a DECUS Membership ID.   ------------------------------  % Date: Mon, 29 Apr 2002 23:07:31 -0400n- From: JF Mezei <jfmezei.spamnot@videotron.ca>hC Subject: Re: DECUS/EncompassUS Survey -- no longer assuming Windows , Message-ID: <3CCE0A6E.6E5CEF03@videotron.ca>   "Terry C. Shannon" wrote:tF > There's a Better Answer: requiring the use of a DECUS Membership ID.  L If you do that, better make sure that *ALL* your announcements on newsgroupsK clearly specify that the survey is restricted to USA DECUS (or whatever the   name is this week) members only.  K It is very frustrating to be invited to fill out a survey, take the time to/D fill it out, only to be told later that you were not in fact wanted.   ------------------------------    Date: 29 Apr 2002 10:29:36 -0700+ From: Saad_alexan@hotmail.com (Saad Alexan)n# Subject: Re: Digital 7-year plan... = Message-ID: <cf525a10.0204290929.3b383d51@posting.google.com>P  _ "Frank Sapienza" <sapienza@noesys.com> wrote in message news:<a8ih310pag@enews1.newsguy.com>...oL > There's a big difference between having a web site that returns hits for aJ > search on 'RDB' and *marketing* RDB for OpenVMS.  For starters why don'tL > they at least put a button on the main Oracle web site that say's "LookingK > for Oracle RDB for OpenVMS?  Click here."  As for Metalink, only licensedCK > users with a support contract have access to that system, so again that'sm > hardly a marketing tool. >   E I think you are correct. Oracle is no longer interested in markettingt? or supporting RDB. Under the RDB page there use to be a Mission A Statement from Oracle singing the virtues of RDB and how good and D reliable the product is for mission critical applications. Oracle noA longer publish this statement for good reason, they are no longerE interested!a  = Here is the Oracle Mission Statement regarding RDB (no longert published):   ! Oracle Rdb Statement of Directionr Rdb is a SuccessB Oracle Rdb enjoys phenomenal success in environments requiring theE utmost in performance, reliability, availability and maintainability. C Since Rdb became part of Oracle Corporation in late 1994, customers > have enjoyed a significant enhancement in its capabilities and= unparalleled support for their most mission critical systems.a; Rdb is one of the software industry&#8217;s most successful-? acquisition stories. While Rdb is now over 15 years old, Oracle0B Corporation has added nearly 50% new code during the last 5 years.E This significant enhancement has enabled product successes which make%? Rdb a database of choice for high-throughput, critical database F applications on OpenVMS systems. Oracle Rdb products today host a very) significant portion of the world&#8217;s:w;  	Cellular telephone customer service and billing systems,	n
  	Lotteries,	4  	Financial exchanges,	E1  	Satellite Direct Broadcast Television systems,	t$  	Integrated circuit manufacturing,	F  	And many other extremely high volume systems which absolutely cannot tolerate failure.oD Oracle is quite proud of the contribution it has made to Rdb and theD level of service that is available to Rdb customers. Not only is RdbC enhanced by the development investment Oracle has made, but it alsotE benefits from being part of the Oracle family of products. Through ana= increasing emphasis on common technologies and product familysE standards, Rdb has been integrated into a broad selection of softwaretE for custom application development, multi-tier application deployment>E via internet technologies and even the Oracle database engine itself..F Rdb further benefits from Oracle&#8217;s investment in quality support> & consulting. As a result of all these contributions, Rdb is a+ successful business for Oracle Corporation.i
 Rdb Strengthst? The success of Rdb can be primarily attributed to its technical ' strengths. Customers rely upon Rdb for:/F  	Very high performance to deliver extraordinary real-world throughput>  	Extremely high reliability, providing both data security and database uptime F  	Extremely high availability - requirements for scheduled maintenance downtime are minimal6  	Easy maintainability so changes can be made quickly	  	Low cost per transaction	 E These technical strengths are enhanced by the significant investments-A Oracle Corporation has made in Rdb to integrate it into the largekE family of database products and development tools. Developers and endgB users are able to use common and widely acclaimed Oracle tools forF application development, ad hoc queries, decision support and internet application deployment. 
 Rdb FutureB Oracle&#8217;s plans for the future of Rdb are simple: Don&#8217;t change a successful formulam? This means we will maintain our relentless focus on quality andEA stability while enhancing the features and family compatibilities.C which our customers require to maximize the capabilities upon whichM their businesses depend.F We will continue to deliver new releases of Rdb products to fill theseD requirements. Our current focus is to build on the strengths we haveC engineered into Rdb7. With our next major release, Rdb8, we plan to . include significant new features and enhanced:B  	Performance, by making further advances in very-large-memory and numerous other areas; C  	Reliability, through a focused program of engineering code reviewS and extensive beta testing;OC  	Versatility, by incorporating many changes to further enhance the E interoperability with the Oracle database server and SQL*Net and Java  clients;D  	Ease of use, through enhanced management capabilities and numerous additions to Rdb SQL; andaE  	Quality, through countless internal changes designed to ensure evenmC greater stability, compliance with standards and compatibility with " other Oracle Corporation products.D Oracle Corporation&#8217;s commitment to Rdb and to Rdb customers isE ongoing and open-ended. As we continue to enhance and support Rdb, we-B are guided by what our customers tell us they require. Rdb ProductF Management consults with three Customer Advisory Councils in differentE parts of the world. In addition, we offer Rdb Technical Forums around3A the world where Rdb engineers and other acknowledged experts givesE detailed technical presentations about the latest features of Rdb andoD how to best deploy them. These Council meetings, Forums and frequentC customer visits and discussions provide critical feedback we use to  set Rdb development priorities.u Integration with Oracle8ifE The Oracle Rdb Team consists of extremely dedicated professionals who > are a part of the overall database development group at OracleE Corporation. While Rdb engineers are working to enhance Rdb, they areaA working side-by-side with engineers working on Oracle8i and other  related database projects.E Oracle8i enjoys the dominant position among open-system databases andyA is the most popular database in the world. It is the system which C powers the internet. Because of this popularity, most Rdb customersdA also utilize Oracle8i. Over the past 5 years, Rdb Engineering hasaA pursued a strategy of making Rdb a full participant in the OracletA environment. One result of this approach has been that many othercF Oracle technologies - for eBusiness applications, data warehousing and< custom application development - have become well integrated= participants in Rdb customer environments. Beginning with theh@ engineers and continuing to large customer installations, OracleA actively encourages this "cross pollination" of technologies. ThetB result is a stronger set of products which meet a broader range of< requirements and provide the Rdb customer with the choice of> technologies to best meet their needs today and in the future.              L > This has been an ongoing nit of mine re: Oracle RDB for the past couple ofJ > years.  First, they don't seem to want people to know that Oracle RDB isK > available for OpenVMS, much like Compaq doesn't want people to know aboutlM > OpenVMS itself.  Second, if you happen to know about Oracle RDB, or somehowPL > find out about it through some mechanism, there is no (apparent) effort to1 > inform the Oracle sales team about the product.  > M > Here's my example.  It's ok, you can try this at home.  Say for the sake of > > argument that you've stumbled across the Oracle RDB web siteL > (www.oracle.com/rdb). You want to know about licensing and pricing, right?M > On that web site someplace you'll find a note about RDB not being availablerM > through the online store, so please call Oracle sales (1-800-Oracle1).  YoudN > make the call and ask for information about Oracle RDB 7 Enterprise Edition.( > What do you expect the response to be? > J > Well, the average end-user would expect to receive pricing and licensingM > information.  The last time I tried this (1st week of March, this year) theiJ > sales person's first comment was, "Why do you want such an old version?"J > Obviously, she thought I was talking about Oracle 7i.  So I informed herM > that Oracle RDB 7 Enterprise Edition was a *different* database sold by her N > employer.  She clicks around on the keyboard and said, "I'll have to send anG > e-mail to get this information.  Give me your name, phone number, and.  > someone will get back to you." > K > Ok, it's been a month and nobody has called me back.  Worse yet, since myrM > call to 1-800-Oracle1 I attended the RDB Technical Forum in NYC a couple ofiK > weeks ago.  They asked everyone to fill out the usual questionnaire about-K > whether it was useful, blah, blah, blah.  At the bottom of the form was anK > question about whether you would like to be contacted by Oracle or Compaq L > sales.  I said yes to being contacted by Oracle.  Still no phone call.  (IN > will admit that the folks at that event told me to e-mail them directly if IK > had continued problems getting sales information, but I haven't done thatR9 > yet.  It's no longer on the front-burner, so to speak.)4 > D > Two years ago it was both much worse and much better.  Through theF > comp.databases.rdb newsgroup I grumbled and got hooked up with real,J > honest-to-goodness Oracle RDB folks who fed the information to an OracleK > telephone sales person.  When it came time to speak with that person theyTG > asked me what the clock rate was for a VAX 4106 and VAX 4300 (the twooN > systems I wanted to upgrade at the time).  Well, does anyone here know whereM > to find the clock rate for those VAX processors?  I didn't, and neither diduD > the Oracle sales guy, so that attempt at licensing RDB dead-ended. > L > You'd think that at the current price (apparently $40k per CPU + 22% for aK > support contract, though unconfirmed by any Oracle representative) Oracle L > would make a better attempt at selling the product to people that actuallyJ > ask for it by name.  Certainly, if you've got an installed base of usersH > that want to upgrade to the current release I'd think any vendor wouldM > provide a mechanism to capture that sale.  Instead, it seems that Oracle iscN > quite content collecting upgrade and support contracts from their very largeM > named accounts with the big-ticket Galaxy systems that are mission-critical N > to billion-dollar enterprises and aren't too worried about the little crumbsB > that are still quite loyal to both OpenVMS and RDB and still runA > mission-critical applications for a small/medium size business.l > @ > Oh, well, thanks for letting me vent.  I feel much better now. > :-)g   ------------------------------  % Date: Mon, 29 Apr 2002 18:12:15 -040072 From: norm lastovica <norman.lastovica@oracle.com># Subject: Re: Digital 7-year plan...,* Message-ID: <3CCDC53F.4E0C83D3@oracle.com>   Saad Alexan wrote: > G > I think you are correct. Oracle is no longer interested in markettingpA > or supporting RDB. Under the RDB page there use to be a Mission1C > Statement from Oracle singing the virtues of RDB and how good and F > reliable the product is for mission critical applications. Oracle noC > longer publish this statement for good reason, they are no longeri
 > interested!g  * 	this is, of course, nonsense.  the same  1 Oracle Rdb Statement of Direction is still on thee oracle rdb web site.   	www.oracle.com.rdb    ------------------------------  % Date: Mon, 29 Apr 2002 19:08:24 -0400e2 From: norm lastovica <norman.lastovica@oracle.com># Subject: Re: Digital 7-year plan...t) Message-ID: <3CCDD268.76B8860@oracle.com>k  ! and before anyone feels compelledv" to point it out again (!), the url! is, of course, www.oracle.com/rdbr and I fall on my sword.    norm lastovica wrote:t >  > Saad Alexan wrote: > >aI > > I think you are correct. Oracle is no longer interested in markettingeC > > or supporting RDB. Under the RDB page there use to be a MissionaE > > Statement from Oracle singing the virtues of RDB and how good andeH > > reliable the product is for mission critical applications. Oracle noE > > longer publish this statement for good reason, they are no longeri > > interested!i > 1 >         this is, of course, nonsense.  the same 3 > Oracle Rdb Statement of Direction is still on thea > oracle rdb web site. >  >         www.oracle.com.rdb   -- h> norman lastovica / oracle rdb engineering / usa / 610.696.4685   ------------------------------  # Date: Mon, 29 Apr 2002 18:15:05 GMTo2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman) Subject: Re: download VMSc8 Message-ID: <J4gz8.40$vZ.667793@cacnews.cac.cpqcorp.net>  L In article <aa9hql$e49$2@aquila.mdx.ac.uk>, david20@alpha2.mdx.ac.uk writes:` :In article <aa9c5f$odd$1@wilson.uits.indiana.edu>, bdwheele@indiana.edu (Brian Wheeler) writes:4 :>In article <3cc82bbb.16519531@news.cable.ntl.com>,: :>	peter.watkinson1@ntlworld.com (Peter Watkinson) writes: :>> J :>> I was wondering if it would be a good idea for Compaq (HP?) to let youI :>> download VMS through the internet. Particularly the Hobbyist version.oI :>> With Linux and BSD it so easy to download and install via FTP so longOF :>> as you have a fast connection. I don't think it would be neccesaryJ :>> with windows as a 5min drive to the local PC world can pick up a copy.< :>> Could be a good idea to bring VMS into the 21st century. :>>  :>>  e :>>  :>> Peter Watkinsonn! :>> peter.watkinson1@ntlworld.comh :>L :>I think it would be great, especially if they included some of the layeredI :>products.  Since you have to have a license anyway to use the binaries,s= :>making the media freely available can't hurt license sales.d :> :>Brian. : J :Many of the people in this newsgroup including myself have argued for theJ :layered products to be made available for download from the web for ages.N :Unfortunately, with a few welcome exceptions such as the C++ kits, Compaq has :always resisted.AN :However I think this is the first time I've heard of anyone suggesting CompaqO :put the OS itself up on the web. It's not a bad idea - after all if Sun can doi :it why not Compaq ? J     Microsoft doesn't, of course./  H   This is obviously a large shift in long-standing distribution and cost	   models.   F   We could certainly provide this (for those customers that have largeF   pipes and a willingness to run downloaded code), but the decision toI   change over the model would have to be made by Compaq VPs sufficiently 3G   high up in the corporate organizational food-chain.  This change-overrH   is something that has been discussed (both internally and externally) G   for a few years now, of course -- this all devolves into the businessp(   issues, which "turfs" this to the VPs.  G   I know of no plans to provide this for general access, though we are oG   working with specific customer sites using downloads of OpenVMS kits.     N  ---------------------------- #include <rtfaq.h> -----------------------------N       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  % Date: Mon, 29 Apr 2002 15:55:58 -0400u- From: JF Mezei <jfmezei.spamnot@videotron.ca>n Subject: Re: download VMSe, Message-ID: <3CCDA54A.B14AF44B@videotron.ca>   Hoff Hoffman wrote:tH >   We could certainly provide this (for those customers that have largeH >   pipes and a willingness to run downloaded code), but the decision toJ >   change over the model would have to be made by Compaq VPs sufficiently7 >   high up in the corporate organizational food-chain.r    N Perhaps the best approach would be a gradual shift and see how things evolve ?D Start with compilers, then with layered products, and then with VMS.  G "Shirley" the VMS engineers can provide the adequate security to ensurenL integrity of the kits. Wouldn't checksum provide enough security to ensure aL kit has not been tampered with and was transfered intact without losing some bits ?  G This should not only seen as a day-to-day cost saving from shipping andlN handling, but also as a great marketing issue that would put VMS on the map as+ a leader in safe distribution over the net.a   ------------------------------  % Date: Mon, 29 Apr 2002 16:08:01 -0700-' From: David Mathog <mathog@caltech.edu>. Subject: Re: download VMSe+ Message-ID: <3CCDD251.336A3231@caltech.edu>e   Hoff Hoffman wrote:e > N > In article <aa9hql$e49$2@aquila.mdx.ac.uk>, david20@alpha2.mdx.ac.uk writes:b > :In article <aa9c5f$odd$1@wilson.uits.indiana.edu>, bdwheele@indiana.edu (Brian Wheeler) writes:6 > :>In article <3cc82bbb.16519531@news.cable.ntl.com>,A > :>      peter.watkinson1@ntlworld.com (Peter Watkinson) writes:k > :>>lL > :>> I was wondering if it would be a good idea for Compaq (HP?) to let youK > :>> download VMS through the internet. Particularly the Hobbyist version.sK > :>> With Linux and BSD it so easy to download and install via FTP so longhH > :>> as you have a fast connection. I don't think it would be neccesaryL > :>> with windows as a 5min drive to the local PC world can pick up a copy.> > :>> Could be a good idea to bring VMS into the 21st century. > :>>l > :>>  > :>>t > :>> Peter Watkinsonc# > :>> peter.watkinson1@ntlworld.como > :>N > :>I think it would be great, especially if they included some of the layeredK > :>products.  Since you have to have a license anyway to use the binaries,r? > :>making the media freely available can't hurt license sales.  > :>	 > :>Briana > :oL > :Many of the people in this newsgroup including myself have argued for theL > :layered products to be made available for download from the web for ages.P > :Unfortunately, with a few welcome exceptions such as the C++ kits, Compaq has > :always resisted. P > :However I think this is the first time I've heard of anyone suggesting CompaqQ > :put the OS itself up on the web. It's not a bad idea - after all if Sun can dom > :it why not Compaq ? > ! >   Microsoft doesn't, of course.t  O That's only partially true.  While Microsoft doesn't provide installation disksnJ for their products on the net they do provide service packs.  IE and a fewL other things (the free ones) can be downloaded too.  Microsoft has bandwidthI considerations to think about that Compaq doesn't.  They would have to begJ able to handle a million people all trying to download WXP++ or Office2002H all at the same time. Sadly, a load problem that the VMS server will not
 encounter.   > J >   This is obviously a large shift in long-standing distribution and cost >   models.o >   K Sure is - from a prehistoric model to one that is considerably more modern.c  H >   We could certainly provide this (for those customers that have large4 >   pipes and a willingness to run downloaded code),  M Hard to imagine that any VMS site these days doesn't have access to that sorttK of bandwidth - other than maybe on oil platforms and JSTARS.  Neither would L be adversely affected by the new download method: the former would just haveO a CD written in the central office and shipped out on the next chopper, and theaN latter, well, I'm pretty sure that the software on one of those is not altered
 in flight!   Regards,   David  Mathog mathog@caltech.edu   ------------------------------    Date: 29 Apr 2002 20:39:53 -0500- From: Kilgallen@SpamCop.net (Larry Kilgallen)h Subject: Re: download VMSh3 Message-ID: <MX6uOFs8b0y3@eisner.encompasserve.org>h  U In article <3CCDD251.336A3231@caltech.edu>, David Mathog <mathog@caltech.edu> writes:t > Hoff Hoffman wrote:a >> mO >> In article <aa9hql$e49$2@aquila.mdx.ac.uk>, david20@alpha2.mdx.ac.uk writes:dc >> :In article <aa9c5f$odd$1@wilson.uits.indiana.edu>, bdwheele@indiana.edu (Brian Wheeler) writes:   O >> :>I think it would be great, especially if they included some of the layeredbL >> :>products.  Since you have to have a license anyway to use the binaries,@ >> :>making the media freely available can't hurt license sales. >> :> 
 >> :>Brian >> :M >> :Many of the people in this newsgroup including myself have argued for thesM >> :layered products to be made available for download from the web for ages. Q >> :Unfortunately, with a few welcome exceptions such as the C++ kits, Compaq hasm >> :always resisted.Q >> :However I think this is the first time I've heard of anyone suggesting CompaqeR >> :put the OS itself up on the web. It's not a bad idea - after all if Sun can do >> :it why not Compaq ?  >>  " >>   Microsoft doesn't, of course. > Q > That's only partially true.  While Microsoft doesn't provide installation disksfL > for their products on the net they do provide service packs.  IE and a few   Same as VMS.  N > other things (the free ones) can be downloaded too.  Microsoft has bandwidth  
 Same as CSWS.t  I >>   We could certainly provide this (for those customers that have largey5 >>   pipes and a willingness to run downloaded code),  > O > Hard to imagine that any VMS site these days doesn't have access to that sortiM > of bandwidth - other than maybe on oil platforms and JSTARS.  Neither would N > be adversely affected by the new download method: the former would just haveQ > a CD written in the central office and shipped out on the next chopper, and thebP > latter, well, I'm pretty sure that the software on one of those is not altered > in flight!  0 Which is more than one can say for the Internet.   ------------------------------    Date: 29 Apr 2002 12:03:24 -0500- From: koehler@encompasserve.org (Bob Koehler)t Subject: Re: EDT or EVEa3 Message-ID: <grEu+lX2D5pA@eisner.encompasserve.org>t  \ In article <3CC8EC01.DB472760@videotron.ca>, JF Mezei <jfmezei.spamnot@videotron.ca> writes:  @ > has anyone ever converted from EDT to TPU/EVE as main editor ?  .    Had too.  There was no EVE in the begining.  2 > Was the conversion painful ? What prompted it ?   C    The painfull part was the EVE didn't originally ship with a goodxF    predefined keypad, nor was it obvious how to compile and save a TPUD    routine.  Later the EDT keypad shipped with EVE and TPU functions3    to simply compile the current buffer were added.w  F    So, my keypad is my own.  I started with the VT100 emulation of theB    inverted-T arrow keys and added buttons as I found things I didD    repeatedly.  When I finally got a VT200 I redefined the keys that"    had been used up by the arrows.  ?    About a 1 decade later, I depend very heavily on MY keypad.  > J > However' I am curious about folks who were brought up on EDT. Seems thatL > conversions to TPU are rare and that those stick to EDT. (same applies to  > TECO).  E    My first text editor was TECO, then SOS.  EDT blew away SOS.  I'ved;    never seen a TPU emulator I've liked on other platforms.d    t   ------------------------------    Date: 29 Apr 2002 18:33:15 -0700. From: SPAMSINK2001@YAHOO.COM (Alan E. Feldman) Subject: Re: EDT or EVEe= Message-ID: <343f30ae.0204291733.3feb0cd5@posting.google.com>t  n P.Young@unsw.EDU.AU (Patrick Young) wrote in message news:<55f85d77.0204260137.6c6898c0@posting.google.com>...u > SPAMSINK2001@YAHOO.COM (Alan E. Feldman) wrote in message news:<343f30ae.0204251536.76b07dfe@posting.google.com>...ro > > Michael Austin <maustin@firstdbasource.com> wrote in message news:<3CC827BD.6181EC1E@firstdbasource.com>...  > > > "Alan E. Feldman" wrote:v > > > > P.Young@unsw.EDU.AU (Patrick Young) wrote in message news:<55f85d77.0204250058.1e93f258@posting.google.com>...	 > > > > >dL > > > > > I never did work out why EDT does not have LEARN where as KED did.: > >                                ^^^^^^^^^^^^^^^^^^^^^^^4 > >                                    ^^^^^^^^^^^^^ > J > Sheesh, you have to be really careful how you word things in this NG :-)    D Yes, because your saying that "EDT does not have LEARN" implies thatF you cannot define keys with EDT. It is misleading at best. If you saidC "you cannot see the keystrokes executed as you define the key using1+ EDT" then that would be much more accurate.     @ > EDT does *not* *have* LEARN - EDT has DEFINE KEY (aka CTRL-K). >  > Command: LEARN >      ^ > Unrecognized command  r  E Fine, so EDT calls its version of "LEARN" DEFINE KEY. So? If you typetB DIRECTORY at a Unix prompt, you won't get a listing of your files.A That doesn't mean that you can't do the VMS DIRECTORY function iniB Unix. Yes, there are differences, of course, but both systems haveA that function, just like both editors have the "define key/learn" A function. You can't say "Well, I like VMS better because it has ai directory command".     H > DEFINE KEY in EDT (even though much more along the lines of LEARN thanJ > DEFINE KEY is in EVE) gives you less WYSIWYG functionality and thereforeF > differs enough for it not to be called LEARN. Yes, it does achieve a > similar goal.c    E C'mon. It achieves *exactly* the same goal. You want to define a key. A There are differences between them, but you can define a key with, either.t  $ In your previous article, you wrote:   [CAP quote]rF EVE is more powerful and is my choice (with EDT keypad enabled) simplyE because you can define learn sequences and bind them to a key, and itt= ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^f+ supports longer than 255 character records.f [end CAP quote]     E It seemed to me that "learn sequences" was just your way of saying "a4F set of keystrokes" -- something the editor, be it EDT or EVE, "learns"B and subsequently assigns to a key. It's not clear at all from thisA that you meant "LEARN sequences" instead of "learn sequences". InoC fact, you are saying that in EDT you cannot "define learn seqeunces 4 and bind them to a key". No? Well you certainly can!  D I apologize if I appear to be nitpicking but I feel that it is quiteA likely that newbies reading your post would assume that you can'tlD define keys in EDT when you clearly can. Not as easily, perhaps, but you can do it.    G > I defined LEARN as how it functions in KED (the PDP-11 Keypad EDitor) E > which I have used under the RT-11 operating system. LEARN also fits     C Only in your second reference, only vaguely, and since the originalbE poster almost certainly doesn't know about KED, it's not of much use.     D > this description under EVE (CONTROL-R replaces GOLD-S and EVE letsA > you define multiple different learn sequences and nest them). I G > express surprise that at least the KED LEARN functionality is missingi > from EDT.o    E OK, you are surprised that you can't see your key-defining keystrokese? in EDT. OK. But that wasn't mentioned in your original message.i    E > I'm not here to argue which editor is better - (although the recordtE > size issue can be a show stopper for EDT - especially when you haveiG > users creating files using PC products over Advanced Server shares) -lG > as I use both and would be very upset if EDT was removed. My original ; > posting suggested the new user learn *both* (or TECO :-).0 > = > > > > Why doesn't EDT's Ctrl/K count as a LEARN equivalent?o > J > Because you can't see on the screen the results on the text to be editedF > as you make up your macro sequence. Using LEARN, you are essentiallyF > *testing* and *editing* your macro sequence as you are composing it.    C Yes, I understand this difference. But as I said above, you made it A sound like EDT cannot define keys at all. That was my main point.t That's all.i     Disclaimer: JMHO Alan E. Feldmant" afeldman astki gfigroup dotski com   ------------------------------   Date: 29 Apr 02 19:27:59 +0200) From: p_sture@elias.decus.ch (Paul Sture)i% Subject: Re: exceed escape charactersw) Message-ID: <dyziWss1a$Nx@elias.decus.ch>-  g In article <60fdd9c9.0204290701.54440b9a@posting.google.com>, nikki_wire@yahoo.com (nikki_wire) writes:cG > does anyone know how to use escape characters when using the terminalkF > emulator "exceed"? (CTRL V then 3 doesn't seem to work in TPU editorF > nor does CTRL 3 then 3 from EDT) I'm guessing I may need to change aC > setting in exceed. I'm running 6.2.0 of Hummingbird exceed and amr! > connecting over telnet. Thanks!r   Try one of these:   
 CTRL/V CTRL/[    <PF1>27<PF1><KP3>g  * (the last there meaning Numberic Keypad 3)   ------------------------------  % Date: Mon, 29 Apr 2002 22:28:12 +0530o5 From: "Sandeep Yelwatkar" <Sandeep_Yelwatkar@bmc.com> " Subject: Exceed to get VMS display/ Message-ID: <ucqug69074mh07@corp.supernews.com>i   Hi, J     I am trying to use Exceed to get the xterm display of remotely located, VMS machine onto my local host (W2k machine)       Just as we use the  C export DISPLAY=wints:31; /usr/openwin/bin/xterm -display @d -sb -sl A 5000 -ls -fg navyblue -bg wheat -T suntan -geometry 75X40+140+110   ; command in xstart window to get the display of Unix machineo  8 What is the corresponding command to get the VMS display   Any help is most welcomede   Thanks Sandeepa   ------------------------------  # Date: Mon, 29 Apr 2002 17:00:49 GMTl4 From: Tim Llewellyn <tim.llewellyn@blueyonder.co.uk>& Subject: Re: Exceed to get VMS display0 Message-ID: <3CCD7BB1.EBE83B45@blueyonder.co.uk>   Sandeep Yelwatkar wrote: >  > Hi,aL >     I am trying to use Exceed to get the xterm display of remotely located. > VMS machine onto my local host (W2k machine) >  >     Just as we use the > E > export DISPLAY=wints:31; /usr/openwin/bin/xterm -display @d -sb -sllC > 5000 -ls -fg navyblue -bg wheat -T suntan -geometry 75X40+140+110i >   2 $ set display/node=wints:31/transport=tcpip/create $ create/terminal/detached  8 will get you started if decterm rather than xterm is OK.9 Then you can tweak the display settings to your taste and  save as default.  9 Xterm is non included with VMS but it may be installed ont& your system, check with your sysadmin.  = This all assumes base DECwindows support is installed on yours% VMS box and configured to use TCP/IP.    Hope this helps.   Regards,    = > command in xstart window to get the display of Unix machinef > : > What is the corresponding command to get the VMS display >  > Any help is most welcomed  >  > Thanks	 > Sandeepw   -- e tim.llewellyn@blueyonder.co.uk    F * tim.llewellyn@cableinet.co.uk address will cease to work June 2002 *   ------------------------------  % Date: Mon, 29 Apr 2002 22:15:42 +0200s2 From: martin@radiogaga.harz.de (Martin Vorlaender)& Subject: Re: Exceed to get VMS display; Message-ID: <3ccda9ee.524144494f47414741@radiogaga.harz.de>i   Small correction:o  5 Tim Llewellyn (tim.llewellyn@blueyonder.co.uk) wrote:t > Sandeep Yelwatkar wrote:G > > export DISPLAY=wints:31; /usr/openwin/bin/xterm -display @d -sb -sloE > > 5000 -ls -fg navyblue -bg wheat -T suntan -geometry 75X40+140+110t >34 > $ set display/node=wints:31/transport=tcpip/create > $ create/terminal/detached   It's  ;   $ set display/node=wints/server=31/transport=tcpip/create    cu,e   Martin -- rB                         | Martin Vorlaender | VMS & WNT programmer1  OpenVMS: Where do you  | work: mv@pdv-systeme.de D  want to BE today?      |   http://www.pdv-systeme.de/users/martinv/8                         | home: martin@radiogaga.harz.de   ------------------------------  # Date: Mon, 29 Apr 2002 20:56:57 GMT ) From: DCantor@shore.net (David A. Cantor)F) Subject: Re: Fix for EDT emulation in EVE-; Message-ID: <tsiz8.57761$sI3.13434018@news02.optonline.net>-  Z In article <3CCC29A7.9416D4D7@fsi.net>, "David J. Dachtera" <djesys.nospam@fsi.net> wrote:  G >By any chance, have you some tips on how to convert EDT's "DEFINE KEY"F5 >commands (line mode, as found in EDTINI.EDT) to TPU?w  < Convert?  By hand.  I guess you _could_ write a programmatic? converter for simple cases.  Maybe someone has already done so.c: I converted all mine by hand, however.  Programming in TPU< is not that hard if you have examples to look at, and, IMAO,D there are plenty of examples in the EVE code itself in SYS$EXAMPLES.   Dave C.e   ------------------------------  # Date: Tue, 30 Apr 2002 01:19:29 GMTa1 From: "David J. Dachtera" <djesys.nospam@fsi.net>c) Subject: Re: Fix for EDT emulation in EVEh' Message-ID: <3CCDF409.7FE15E62@fsi.net>r   Nico van der Boom wrote: > > > "David J. Dachtera" <djesys.nospam@fsi.net> wrote in message# > news:3CCC29A7.9416D4D7@fsi.net...  > > "David A. Cantor" wrote: > > >e > <snip>J > > By any chance, have you some tips on how to convert EDT's "DEFINE KEY"8 > > commands (line mode, as found in EDTINI.EDT) to TPU? >  > add them in EVE$INIT.EVE ?  3 Well, that's one place to put the translation, yes.e  : The problem is getting the translation in the first place.  E I have rather an extensive EDTINI.EDT file. I cannot justify making aRE career out of converting it to TPU. Would have been nice had DEC comeH< out with an EDT -> TPU translator for your EDTINI.EDT files.  D Then again, scripting TPU/EVE in batch (like you can with EDT) would remain rather a challenge, no?   -- u David J. Dachterai dba DJE Systemsc http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/    ------------------------------  % Date: Mon, 29 Apr 2002 16:38:12 -0400a0 From: "Syltrem" <syltremspammenot@videotron.com> Subject: Re: Gold key 5 Message-ID: <d9iz8.10561$a04.47389@tor-nn1.netcom.ca>e  % lookup your terminal emulator manual.nI It might be the Num-Lock key, as this one is located in the same physicalyH position as the PF1 (gold) key on VT keyboards. This is the default with
 PowerTerm.   HTHe   --   SyltremrI http://pages.infinit.net/syltrem (OpenVMS related web site - en franais)o> To reply to myself directly, remove .spammenot from my address  J "Sandeep Yelwatkar" <Sandeep_Yelwatkar@bmc.com> a crit dans le message de* news: ucraltqve6to7b@corp.supernews.com... > Hi, : > I am a novice to VMs system and learning the EDT editor.+ > I am at present going through the manual.ME > I want to know what is "Gold Key". I am using normal keyboard whiche doesn'trJ > have "Gold" return on any of the keys. How do I simulate it using normal > keyboard?i > Any help is most appreciated.a >h > Thanks in advances	 > Sandeep  >i >d >  >H   ------------------------------  % Date: Tue, 30 Apr 2002 01:55:54 +0530q5 From: "Sandeep Yelwatkar" <Sandeep_Yelwatkar@bmc.com>e Subject: Gold keyn/ Message-ID: <ucraltqve6to7b@corp.supernews.com>f   Hi,'8 I am a novice to VMs system and learning the EDT editor.) I am at present going through the manual.oK I want to know what is "Gold Key". I am using normal keyboard which doesn'teH have "Gold" return on any of the keys. How do I simulate it using normal	 keyboard?o Any help is most appreciated.o   Thanks in advancev Sandeep    ------------------------------  # Date: Mon, 29 Apr 2002 21:18:58 GMT22 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman) Subject: Re: Gold key 9 Message-ID: <6Niz8.51$r11.854111@cacnews.cac.cpqcorp.net>n  g In article <ucraltqve6to7b@corp.supernews.com>, "Sandeep Yelwatkar" <Sandeep_Yelwatkar@bmc.com> writes:s  9 :I am a novice to VMs system and learning the EDT editor.   A   Please use TPU or other editor, EDT was retired some years ago.r  ?   When posting, please include the OpenVMS version and platforms@   involved and -- in cases such as this -- the specific hardware   involved.f  * :I am at present going through the manual.  A   With TPU and newer editors, you can get help from the keyboard.D@   (Again, EDT really isn't the best editor for a new user -- andB   this is not intended to start an editor war, as I do prefer and A   do use the EDT-style keyboard layout daily.  EDT is simply not aA   as advanced as other more recent editors -- TPU and EVE and LSE2&   all have keypad help, for instance.)  L :I want to know what is "Gold Key". I am using normal keyboard which doesn't>                                                ^^^^^^^^^^^^^^^  @   There is (unfortunately) no such thing as a "normal keyboard".  A   If this is a DIGITAL or Compaq LK-series or PC-series keyboard,dA   flip it over and look at the model number on the sticker on thei	   bottom.e  I :have "Gold" return on any of the keys. How do I simulate it using normale
 :keyboard?  E   On the OpenVMS-style LK-series keyboards (eg: LK461, LK46W, LK401, aG   LK201, LK411, etc) the Gold key is the upper left key on the numeric rG   keypad, on the cluster of keys to the far right of the main keyboard.a(   On these keyboards, it is labeled PF1.  F   On the WPS-series keyboards -- the LK402 was one example -- this PF1G   key actually is the gold-colored key on the keyboard.  And PF4 is them
   "blue" key.c  H   On a PC keyboard, the key is determined by the terminal emulator, but E   it is often the key labeled "num lock" (numb lock :-), and is in a h;   similar position in the upper left of the numeric keypad.     N  ---------------------------- #include <rtfaq.h> -----------------------------N       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  # Date: Mon, 29 Apr 2002 17:21:02 GMT 2 From: "Sue Skonetski" <susan.skonetski@compaq.com># Subject: HPS Times is now availabled8 Message-ID: <2ifz8.35$RW.555100@cacnews.cac.cpqcorp.net>  3 http://www.compaq.com/hps/news/hpstimes_issue1.htmlo   ------------------------------  # Date: Mon, 29 Apr 2002 16:52:50 GMTe2 From: "Sue Skonetski" <susan.skonetski@compaq.com>Y Subject: INFONOVA and SyntheSys Secure Technologies, Inc. Partner To Globally Address PKIn8 Message-ID: <CTez8.33$AZ.672935@cacnews.cac.cpqcorp.net>  8       Just back from vacation and I had this sent to me.  	       Suen  E       http://www.synthesysusa.com/news/releases/infonova_april23.html2  J       INFONOVA and SyntheSys Secure Technologies, Inc. Partner To Globally  Address PKI Services for OpenVMS  ?       DEERFIELD BEACH, FL-INFONOVA has awarded SyntheSys Secure@J Technologies, Inc., a wholly owned subsidiary of AUG Corp. (OTC/BB: AUGC),G the rights to port their software to Compaq's OpenVMS operating system.eI Under the agreement SyntheSys holds exclusive global rights to market andtL sell the OpenVMS version and reserves the right to brand these products with the SyntheSys name.pH       Additional terms of the agreement call for INFONOVA to construct aG Trust Center in Florida. In addition to INFONOVA's proprietary softwareeF proTRUST for certification services, applications from the proSECURITY1 SOLUTIONS product series will also be integrated.   K       In the next months, the Public Key Infrastructure (PKI) department ofgK INFONOVA's E-Security Business Unit will set up the complete infrastructure B for SyntheSys to become a Security Service Provider (SSP). BesidesH consulting services, the project includes, above all, the integration ofH INFONOVA propriety Trust Center Software proTRUST, now branded under theI SyntheSys idSURE Secure Application Suite as idTRUST, for a certificationyA service according to European specifications. Moreover, currentlyfH implemented solutions from the idSECURITY SOLUTIONS software series willK complete the project; first and foremost, the idSEND application for secure H document exchange on the Internet, and the idSAFE application for highly secure data storage.  K       "Excellent know-how in the area of E-Security and the availability ofcH a self-contained product line for PKI applications were decisive for ourJ partnership with INFONOVA," said Shashi Kapur, Chief Executive Officer and0 President of SyntheSys Secure Technologies, Inc.  E       Klaus Kersten, Business Unit Manager for E-Security at INFONOVA L stated, "This project is a successful start for our international engagementG in the area of E-Security. The market in the United States is enormous.iG There are 7 million people living just within the city limits of Miami. K Internet availability is excellent and the need for E-Security solutions is J increasing permanently. In the field of E-Security the United States needsJ to catch up with Europe. Austria has one of the most strict signature lawsJ worldwide - our software solutions fulfill not only legal demands, but areF also being used in many projects presently in operation. Together withI SyntheSys, our knowledge will be brought to markets worldwide. Florida iss only the beginning."  G       INFONOVA, an Austrian daughter company of KPMG Consulting AG, hasnD completed several successful security projects in Europe. Their coreD competencies are in the development, implementation, integration andH maintenance of networks and applications based on Internet technologies.K INFONOVA provides products and services in Internet Infrastructure, OSS/BSSaJ (Operational Support Systems/Business Support Systems), Server Technology,G Network Management, IP Telephony, E-Security, Customer Care Technology, I Interactive TV & Digital Video Broadcast as well as technology consultingo# and comprehensive training courses.       L ----------------------------------------------------------------------------       BACK TO NEWS   ------------------------------  # Date: Tue, 30 Apr 2002 01:01:37 GMTe1 From: "David J. Dachtera" <djesys.nospam@fsi.net>i Subject: Re: Itanium troubles ' Message-ID: <3CCDEFD7.9BE3DDD8@fsi.net>u   Peter da Silva wrote:r > [snip]L > I'm sorry, but I must be missing something. Would you mind explaining yourG > reasoning a little more explicitly... reading that post a second timei, > doesn't do anything to clarify your point.  & O.k. Let's look at that one more time:   Sander Vesik wrote:f > C > In comp.arch Tarjei T. Jensen <tarjei.jensen@kvaerner.com> wrote:o > >y
 > > [snip]K > > Reading books on usability should be a requirement for all programmers.n > >s > G > Having a clue about how FreeBSD and similar projects happen should beEC > compulsory, so nobody would ever say something as stupid as that.g  H Now, I may be wrong, but I took that to mean that Sander apparently seesC neither the necessity nor importance of usability as having as muchgE weight as the other issues involved in the development of open-sourcel	 software.a   Does this help you out?n  A If it doesn't that's o.k. - my wife frequently doesn't follow me,.1 either, and she has a much higher IQ than I have.(   -- - David J. Dachtera  dba DJE Systemse http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/,   ------------------------------  % Date: Mon, 29 Apr 2002 22:19:48 +0200t2 From: martin@radiogaga.harz.de (Martin Vorlaender)) Subject: Re: MAIL on VMS and attachments. ; Message-ID: <3ccdaae4.524144494f47414741@radiogaga.harz.de>)  ) Bob Ceculski (bob@instantwhip.com) wrote:D7 > martin@radiogaga.harz.de (Martin Vorlaender) wrote...oE > > In MultiNet and TCPware, if you MAIL/FOREIGN/TYPE=1, you will getTH > > a standard MIME encoded mail, which can be received by any compliant > > client.  >cE > what are the other types for?  I tried this with a simple text files2 > with a /subject="" and got the mail but no text!  C /TYPE=0 is the default (the standard MAIL/FOREIGN). I don't know ofIC other values. It's as "well-known undocumented" as the MAIL/FOREIGN  command itself.    cu,    Martin -- hB                         | Martin Vorlaender | VMS & WNT programmer1  OpenVMS: Where do you  | work: mv@pdv-systeme.deeD  want to BE today?      |   http://www.pdv-systeme.de/users/martinv/8                         | home: martin@radiogaga.harz.de   ------------------------------    Date: 29 Apr 2002 09:39:40 -0700: From: craig.berry@SignalTreeSolutions.com (Craig A. Berry)! Subject: Re: Netscape and Mozillak= Message-ID: <7f15589f.0204290839.1e09ed8e@posting.google.com>m  | Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com> wrote in message news:<01KH16DZ4IB88Y7GCV@sysdev.deutsche-boerse.com>...   >  Isn't Mozilla one of C > these open-source thingies?  Can't one just get the source, edit DK > COMPILE.COM and LINK.COM (or whatever the equivalent is in the way it is pJ > normally built, perhaps using MMS or whatever) to turn off debugging andJ > turn on optimisation and recompile and relink?  If so, someone could do 3 > so and let us know how much performance improves.t    E Hmm.  It's unfortunately quite a bit more complicated than that.  The C page here describes how to get the source and how to go through the  build process:  H http://www.openvms.compaq.com/openvms/products/ips/cswb/cswb_source.html  3 and says that the build takes 22 hours on a DS20!  -  E There is also a description of static builds vs. shared builds, whererF static includes most code in the main image but shared puts most of itD into a large number of shareable images.  The claim is made that, "AF static build of Mozilla starts up slightly faster and has fewer global@ section requirements than its equivalent shared build," and this, reason is given for using static by default.  A One wonders whether this is really true for systems that are lesshB memory-rich than what OVMS Engineering is testing with, i.e., on aE system with memory to burn it no doubt is more efficient to just grab B one big chunk of it at the outset, but the same trade-off does notD necessarily hold for the systems that most customers are using.  TheD only way to know for sure would be to try it, but I'm afraid I don't> have the time or massive compute resources necessary to do so.   ------------------------------  # Date: Mon, 29 Apr 2002 20:03:50 GMTy4 From: "Matt Muggeridge" <Matt.Muggeridge@compaq.com> Subject: Re: Network Delay?m? Message-ID: <GGhz8.65302$o66.199656@news-server.bigpond.net.au>x  F > The NBIO(nonblocking) flag is set for the socket descriptor, and the0 > process is delayed during the write operation.  J Is this intentional?  Since you don't know what to do with the EWOULDBLOCKB error, it sounds as though you want blocking IO, in which case the! EWOULDBLOCK error will not occur.e  H > Also, I have tried "$ netstat -p tcp".  But, the command is not found.H > %DCL-W-IVVERB, unrecognized command verb - check validity and spelling >  \NETSTAT\  4 If you are running V5.0 or later, you must first run# @sys$manager:tcpip$define_commands./  4 Did you take a look at the ECHO server example code?   Matt.d   --= -------------------------------------------------------------n OpenVMS TCP/IP Engineering Enterprise Computing Group Compaq Computer Corporationa Gold Coast, AUSTRALIAw= -------------------------------------------------------------r    , "wing" <wingwong@witty.com> wrote in message7 news:873e96d6.0204282344.52139856@posting.google.com...0H > > Your write() call returns an error?  What is the error?  Is that the causet1 > > of the delays you are seeing at the receiver.7G > I have added some logic to print the errno, and it is EWOULDBLOCK.  IlA > have checked the Compaq Sockets API and System Services Manual.t2 > EWOULDBLOCK returned with ::write() calls means, >sF > The NBIO(nonblocking) flag is set for the socket descriptor, and the0 > process is delayed during the write operation. >cG > I am a bit confused of this statement, "the process is delayed duringwF > the write operation".  Is the process is paused just as use blocking3 > socket and the waiting the recieve read the data?a >r: > Moreover, what is the cause of this error (EWOULDBLOCK)? >BH > Also, I have tried "$ netstat -p tcp".  But, the command is not found.H > %DCL-W-IVVERB, unrecognized command verb - check validity and spelling >  \NETSTAT\ >2	 > Thanks,2 >6 > Wing >rA > "Matt Muggeridge" <Matt.Muggeridge@compaq.com> wrote in messagee; news:<j7%y8.62019$o66.187957@news-server.bigpond.net.au>...23 > > You have not given much information to go on.... > > H > > Your write() call returns an error?  What is the error?  Is that the causem1 > > of the delays you are seeing at the receiver.a > >lC > > If you suspect unusual network delays then start using ping andu
 tracerouteC > > to observe your times.  Are you seeing any packet loss? Try: "$h
 netstat -pJ > > tcp".  To see the counters of the LAN driver use "$ mcr lancp show dev > > ewa0/count". > >e+ > > When was the last time you ran autogen?n > > 	 > > Matt.    ------------------------------    Date: 29 Apr 2002 20:31:19 -0700 From: wingwong@witty.com (wing)  Subject: Re: Network Delay?v= Message-ID: <873e96d6.0204291931.3d0597ea@posting.google.com>E  L > Is this intentional?  Since you don't know what to do with the EWOULDBLOCKD > error, it sounds as though you want blocking IO, in which case the# > EWOULDBLOCK error will not occur.i > C My intention is to use non-blocking socket.  Thus, the sender couldcE returns instantly from the ::write call.  If there is error, I expectsD that the message written will be lost, instead of sender delay (thisD is the requirement) if there is error writting socket.  However, now the sender process gets delay.  < Is nonblocking socket will behave as blocking socket in some= circumstances, for example too many connection? socket buffer 	 overflow?l    What is the cause of this error?   > 6 > Did you take a look at the ECHO server example code? > D Where could i get this example code, my main reference is the Compaq& Socket API and System Services Manual.  ? The following is the code, I use to create non blocking socket.l // Creating a socket0 if ((fd_=::socket(AF_INET, SOCK_STREAM, 0))<0) {     return false;- }-  ! // Set the socket to non-blocking 
 int arg=1;, if (::ioctl(fd_, FIONBIO, (char *)&arg)<0) {     close();     return false;D }d  ! // Set the socket to SO_REUSEADDRa
 int optval=1;u7 if ( setsockopt(fd_, SOL_SOCKET, SO_REUSEADDR, &optval,y sizeof(optval)) < 0 ) {i     close();     return false;  }    Thanks,k   Wing   ------------------------------  # Date: Tue, 30 Apr 2002 04:38:55 GMTr2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)6 Subject: Re: OpenVMS MAIL (was: Re: Wither ALL-IN-1 ?): Message-ID: <zdpz8.55$qc1.1315900@cacnews.cac.cpqcorp.net>  \ In article <3CCE141C.EF0442DD@videotron.ca>, JF Mezei <jfmezei.spamnot@videotron.ca> writes: ..L :Could the VMS engineers make any comments on the future of VMSmail ? Would N :any enhgancements be envisaged, especially with respect to handling fo MIME , :multiple attachements etc ?  H   I'd certainly like to see various enhancements to MAIL (not the least H   of which would be LDAP support, and better integration of the existingG   MIME support), but as a whole the OpenVMS engineering group is quite  H   busy with the port to IA-64 and with support of EV7 and EV79, as well    as with new hardware.G  I   There is and has been (related) continued work within the MIME utility vK   (which can provide the requested attachment capabilities for the OpenVMS  I   MAIL utility) and there is work on the send-from-file (SFF) stuff over  H   in the TCP/IP Services package (to permit sending MIME messages), and D   the work to add IMAP into the base operating system is going on.    H   All that said, I know of no particular plans to significantly enhance "   the OpenVMS MAIL utility itself.  E   Various tools offer mail capabilities for OpenVMS, of course -- and6F   then there is the old joke about every application developer really G   seeking to provide support for mail and messaging.  Most applicationsrH   of any standing and of any complexity seem to have at least some sort D   of mail support. :-)  Mozilla and Netscape both have mail clients.  N  ---------------------------- #include <rtfaq.h> -----------------------------N       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  # Date: Mon, 29 Apr 2002 22:09:49 GMT  From: "McEagle" <spam@spam.com>a Subject: Re: OT Bobby Vinton? Message-ID: <Nwjz8.348810$K52.56511424@typhoon.tampabay.rr.com>   G I think Bobby Vinton was known as the "Polish Prince" back in the 60's.l   Mike  5 "Bob Ceculski" <bob@instantwhip.com> wrote in message 7 news:d7791aa1.0204290829.1b97cb0b@posting.google.com... 2 > paddy.o'brien@zzz.tg.nsw.gov.au wrote in message1 news:<01KH5BI1VUNM0005TE@tgmail.tg.nsw.gov.au>...o& > > I think I've got the credit right. > >o > > Bob Ceculski wrote:tB > > > I know something that can help you ... try listening to some > > > Bobby Vinton records ... > >aI > > This was a follow-up to speaking German or Polish.  My memory of that4 era has,K > > foundered, because I do not understand what any of his songs have to do  withJ > > the other OT of speaking German or Polish.  Perhaps I'm think of Bobby Vee whor@ > > if IIRC was almost as good an American guy as was Pat Boone. > >sG > > But then the 1960's were a long time ago and it's getting harder toe remember > > my twenties. > >" > > Regards, Paddy >b > remember "My Melody of Love"?e >r   ------------------------------  % Date: Mon, 29 Apr 2002 15:16:04 -0500-1 From: "Dave Gudewicz" <david.gudewicz@abbott.com>e, Subject: Re: Pathworks 6.1 exhausting KNBCBs8 Message-ID: <aak9pj$j2e$1@fizban.fizban.pprd.abbott.com>    For us unwashed.  Whats a KNBCB?   Thanks,e   Dave...   ! "MG" <as@if.com> wrote in message-+ news:3cc7b4fe.23108106@news.videotron.ca...- > Rob,E > you can quickly check for stale sessions by observing the idle timeaG > against "$admin show sessions" for each session.  Kill the duplicatesp# > that show the highest idle times.e >iG > You might want to check with your CSC as that I believe a patch mightsE > exist for sessions being exhausted due to session leaks to the PDC.aG > This mostly affects Member servers, but a patch does exist to correct  > this., > H > Still do try to raise you session limit though since it will give your > sesions more padding.y >e > HTH, > Mike G >e >I   ------------------------------  % Date: Mon, 29 Apr 2002 17:43:40 +0100wT From: Andrew Harrison SUNUK Consultancy <andrew_nospam.remove_this.harrison@sun.com>Y Subject: Re: put up or shut up (was: RE: AMD To Announce Microsoft Support For New Chip FN& Message-ID: <3CCD783C.6040103@sun.com>   david20@alpha1.mdx.ac.uk wrote:   ^ > In article <3CCBACAE.B9BC9BEA@videotron.ca>, JF Mezei <jfmezei.spamnot@videotron.ca> writes: >  >>Bill Todd wrote: >>F >>>So unless Hammer tanks, Intel will have to field something similar. >>> N >>The wildcard here is compilers. Intel has Digital's compiler engineers. DoesN >>AMD have experienced compiler people to develop compilers that will maximize! >>the Hammer chip's performance ?h >> > P > But Itanic is far more reliant on improved compiler technology than Hammer is.L > To get the best out of EPIC requires compiler writers to solve some as yet > unsolved problems. >     ; Far more reliant is probably a huge understatement. Totallye) dependent on would be closer to the mark.t  E This may be why MS appear to be much more interested in x86-64/HammerOA than IA64. People have speculated about IA64's ability to executer: spagetti type code well. MS have their fair share of this.  @ Hammers huge advantage over IA64 is its ability to do a good job of running 32 bit x86 apps.o  C Paradoxically IA64's ability to run x86 apps at all is its achilles F heel. Its runs them but it runs them incredibly badly. The apps vendor@ many of who don't really care about performance or at least care> more about features/functions can tick the box and say it runs@ on IA64 without having to do any work. Customers running a mixed: bag of IA32 and IA64 apps on the same box will end up with: a very sub optimal platform from a performance standpoint.  A On Hammer you should not have the problem, 64bit apps will appearo? but not specifically for performance reasons and a mixed bag ofi. IA32 and x86-64 apps will execute efficiently.  > In this light IA64's requirement to have everything recompiled  looks like a huge dissadvantage.   regards, Andrew Harrisonn   ------------------------------  % Date: Mon, 29 Apr 2002 15:49:53 -0400r- From: JF Mezei <jfmezei.spamnot@videotron.ca> Y Subject: Re: put up or shut up (was: RE: AMD To Announce Microsoft Support For New Chip Fv, Message-ID: <3CCDA3DD.AF4D3938@videotron.ca>  ( Andrew Harrison SUNUK Consultancy wrote:B > on IA64 without having to do any work. Customers running a mixed< > bag of IA32 and IA64 apps on the same box will end up with< > a very sub optimal platform from a performance standpoint.  F You are forgetting the fact that in the Wintel word, many run just oneM application per machine. So if you have one app compiled for IA64, you run itrN on an IA64 wintel, and if another app is optimised for the 8086, you run it on an 8086 box.  M However, in the end, if the 8086 box costs less than IA64 box and can run thea; same application, why bother with IA64 for windows at all ?n  M Ia64 is simply replacing Alpha as the niche chip, expect that IA64 won't haveoK the performance advantage and a decade of fine tuning of compilers etc thatc* Capellas and Fiorina threw down the drain.   ------------------------------    Date: 29 Apr 2002 19:41:36 -0700) From: anickerj@comcast.net (Jeff Anicker)p) Subject: Re: Reduce interupt time on CPU0 = Message-ID: <977914c8.0204291841.64bd9144@posting.google.com>y  C Well I believe we finally located the issue.  After much time spent-F tweaking we learned that there is one database application that prettyF much the entire company connects to.  When a blank database is createdA the system flies.  Once a hundred thousand records are entered itu@ turns everything into a dog.  Once we prove this for certain theB programmers will have to rethink their app.  This one database wasF generating a tremendous amount of io thus overloading poor cpu 0.  The= more records get added to the database the slower things get.   D Thanks much for all your suggestions.  We do currently use fast pathE but with the memory channel turned on I don't think that helps us tooeC much.  We may go to vms 7.3 to allow for scsi fastpath as the nodesn> are all directly connected to a scsi hub connected to hsz70's.  C We are also implementing global buffers with good success.  After ad= fix from compaq for the monitor rms/files command I can see adB reduction in io, locking to a lesser degree, and a rather high hit rate on the global buffers.h  C I am also spending my weekends for the forseeable future convertinghB databases, many of which have not been converted since 1997!  I am? reading the guides to file applications for tuning suggestions.w  D PC sampling was able to lead us in the right direction by confirmingD that indeed we did have an io saturation of cpu0, from there it justF took a while to put the other pieces together to pinpoint the problem.   Thanks All,o Jeff   ------------------------------  % Date: Mon, 29 Apr 2002 15:28:25 -0400 - From: JF Mezei <jfmezei.spamnot@videotron.ca>e5 Subject: Re: removing extra carriage return/line feed-, Message-ID: <3CCD9ED7.E08343EA@videotron.ca>  L 1- can yo do a DIR/FULL on the VMS file in question ? (its record attributes are important).e  M 2- can you describe how the file becomes accessible to your wintel box ? (FTPe ? kermit ? NFS ?, PATHWORKS?)     K One tool you could use on VMS is DUMP/RECORD of the file (HELP DUMP /RECORDhK will tell you how you can specify just a range of records, notably only thed last few records).  N It could also be the file transfer software that adds a blank line at the end.L In which case, you could review this software's use, or perhaps use CONVERT,I providing an FDL that would create some streamlf file that you could then N transfer in binary mode to your wintel, which might eliminate any record level0 manipulation done by the file transfer software.  H Again, it really all depends on the file structure on VMS ( which is why DIR/FULL is important)   ------------------------------  # Date: Mon, 29 Apr 2002 18:50:17 GMTo2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)" Subject: Re: Security question....8 Message-ID: <JBgz8.44$1_.696357@cacnews.cac.cpqcorp.net>  i In article <8b51ed8.0204251254.1d2efcd9@posting.google.com>, vmendham@altavista.com (Vic Mendham) writes:.= :We're having a discussion at work about log file security...nE :Can a user (system or root) manipulate a logfile without anyone evere :knowing about it. If so how???i  H   So you and everyone else that reads this list is entirely trustworthy,I   and will use knowledge gleaned here only for purposes of good.  Uh-huh.u2   Right.  Would you like to rephrase the question?  E :I don't think so cause the revised date on the file would be changedtC :and it should indicate the user who last accessed the file. If thehF :file is open, you can't edit it or copy it as this is a file access ( :sharing ) violation.c  ?   If you have untrusted privileged users, you have no security.h  E   There are tools to tweak the file dates, DFU (Freeware) can do thistE   for instance.  Most of (all of?) the sensitive portions of the the oA   file headers require privileges to change -- READALL is allowedt=   to change headers because of the backup dates, but most anylC   (knowledgeable) privileged user can do whatever said user wishes.s  F   Again, if you have untrusted privileged users, you have no security.  E   OpenVMS does not use the term "root", and OpenVMS security is quiteg;   different from the platforms that do use the term "root".u  E :Anyone have any thoughts on the matter. Our security guy is freakinge :out...b  G   Your security manager will want to read through the OpenVMS Security cG   Manual.  But again, if you have untrusted privileged users, you have u   no security.    N  ---------------------------- #include <rtfaq.h> -----------------------------N       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------    Date: 29 Apr 2002 20:43:48 -0500- From: Kilgallen@SpamCop.net (Larry Kilgallen)?" Subject: Re: Security question....3 Message-ID: <sPGgH5WKFKST@eisner.encompasserve.org>n  m In article <JBgz8.44$1_.696357@cacnews.cac.cpqcorp.net>, hoffman@xdelta.zko.dec.nospam (Hoff Hoffman) writes: k > In article <8b51ed8.0204251254.1d2efcd9@posting.google.com>, vmendham@altavista.com (Vic Mendham) writes:\  G > :Anyone have any thoughts on the matter. Our security guy is freakingg	 > :out...f > I >   Your security manager will want to read through the OpenVMS Security dI >   Manual.  But again, if you have untrusted privileged users, you have n >   no security.  ; Technically, anyone to whom you grant privilege is trusted.   - The question is whether they are trustWORTHY.-  D And it is not sufficient that their honesty be worthy of your trust.= Their technical competence must also be worthy of your trust.e   ------------------------------  % Date: Mon, 29 Apr 2002 13:50:24 -0400h( From: "Mike Foley" <mike-spam@yelof.com>G Subject: Re: somewhat off-topic: is there a DSL router with a console?!n/ Message-ID: <ucr1v0teqlke0d@corp.supernews.com>a  C     I currently use it to host my site (yelof.com), my wife's radio :     program site (theroadhouse.org) and our graphic design     business (lauramfoley.com).d  <     I really don't have much interest in hosting other folks=     sites. There's alot of other places that can do it betterr;     and charge less than I would. :) And I think that's onea7     thing that Speakeasy doesn't allow unless you go to      another type of service.  .     For what I use it for, it works great tho.  <     As for using just one static IP, well, I have 2. One for9     my server that can host all the above web sites usingk8     Apache's VirtualHost feature and one for the Linksys7     router. I second the other person who said to use ao9     linksys and port forwarding if you want to be able tov9     telnet/ssh into different machines behind the router.t  H                                                                     mike< "David J. Dachtera" <djesys.nospam@fsi.net> wrote in message! news:3CCCB2B3.74A20823@fsi.net...u > Mike Foley wrote:o
 > > [snip]G > >     I have my DSL thru Speakeasy.net. They allow and encourage home 0 > >     servers and you get a static IP address. >"D > Is this perhaps something you can make a buck off of? Supplemental > income, at least?  >  > -- > David J. Dachterar > dba DJE Systems  > http://www.djesys.com/ >-* > Unofficial Affordable OpenVMS Home Page:! > http://www.djesys.com/vms/soho/i   ------------------------------    Date: 29 Apr 2002 15:41:50 -0700" From: cstranslations@msn.com (Joe) Subject: TCPIP$C_SHARE= Message-ID: <d56d1c2d.0204291441.148b35e6@posting.google.com>o   OpenVMS 7.1-1H2w
 C V6.2-003 TCPIP 5.0A - ECO 3  B What exactly does the IO$_SETMODE option TCPIP$C_SHARE do? Maybe I" should be asking how it is used...  A Basic idea: a server will be listening on a port for connections.,A "Clients" connect to the port and then send information on one ofr@ several possible "services" that they want to request. After theB server listening on the port has this information it decides if itC wants to pass the request off to a worker capable of processing thepF request (if not it drops the connection). The docs and what I've found+ on the net seem to indicate this is doable.t  C The server specifies TCPIP$C_SHARE on the IO$_ACCESS | IO$M_ACCEPT.,C The worker does the same specifying P1/P3/P5 (with no joy) with itsa> IO$_SETMODE call when it attepts to bind to the resulting port& (SS$_DUPLNAM is returned in the IOSB).  A The only way I can get the "worker" to "connect" to the resultingSD socket is to enable SHARE, use $GETDVI to get the underlying BG unit= number, call $ASSIGN and the use $QIO... I saw something thatSF indicated this was kind of a bad thing to do (something in relation toB unit numbers north of 9999) although at the moment I can't seem to find that reference.  C Any ideas on what I'm doing wrong here or code samples appreciated.u  
 As always TIA: Joe    ------------------------------  % Date: Mon, 29 Apr 2002 15:38:33 -0400V- From: JF Mezei <jfmezei.spamnot@videotron.ca>D3 Subject: Re: UK/EU OpenVMS job market: non-existantn, Message-ID: <3CCDA136.4586049A@videotron.ca>   Chris Bardell wrote:E > Any light at the end of the tunnel / recession / whatever? Or is it D > that, by & large, OpenVMS boxes just run too damn well and require
 > less staff?e  J My take is that everything is on hold pending a clear statement from HP on where VMS is or isn't going.  M The current public information from early merger  is is that "HP is committedoL to supporting its commitments to customers" means simply that it is business8 as usual with VMS ignored and not marketed (slow death).  J If HP makes a VERY VERY CLEAR statement indicating a real change in policyN (increased marketing, leverage the product, expand its target markets), then I, suspect that you'll see a nice boost in VMS.  L If HP makes a wishy washy statement of "we'll support our existing customersK and will continue the port to IA64", then VMS will continue to be lethargic P since that is business as usual and besides nobody is really interested in IA64.   ------------------------------  # Date: Mon, 29 Apr 2002 18:47:32 GMTM1 From: LESLIE@JRLVAX.HOUSTON.RR.COM (Jerry Leslie)l3 Subject: Re: UK/EU OpenVMS job market: non-existantt: Message-ID: <8zgz8.10084$My4.504021@typhoon.austin.rr.com>  5 Tim Llewellyn (tim.llewellyn@blueyonder.co.uk) wrote:  :  : Chris Bardell wrote: : > # : > If anyone has a view on this...e : > H : > Does anyone think the job market for OpenVMS tech support / sys mgrs> : > will ever take off again, specifically in the UK & Europe? : > B : Sure hope so, as it seems even more difficult to diversify into 0 : other areas in IT if you have VMS on your CV.  :  : <dream mode on>sL : Of course, if VMS were to become the reliable web platform to small/medium : buisinesses....a : <dream mode off> :  -  + Can you get Carly to share that dream ? :-)p  G : > One month into the new financial year, but hardly any upturn in theb" : > amount of vacancies available. : H : The only good thing I can try and glean from recent experience is thatD : there are few vacancies but there is turnover, ie not the too much, : of the same job cropping up over and over. :  : > ? : > Any light at the end of the tunnel / recession / whatever? v  D The Economic Policy Institute's briefing paper predicts unemployment! should start to decrease by 2004:i  2    http://www.epinet.org/briefingpapers/bp121.html;    It ain't over till it's really over (EPI Briefing Paper)b  7   "Slow growth will lead to rising unemployment in 2002a#    and high unemployment in 2003...t  	    [snip],  F    In the end, a growth rate of less than 3.0-3.5% (based on 2.0-2.5% F    productivity growth and 1.0% labor force growth) is not sufficient H    to provide jobs to new entrants and to move people from unemployment     to employment...s  B A GDP of 3.0-3.5% is in line with the forecast from Glenn Hubbard,9 Chairman of the White House Council of Economic Advisers:   =    http://www.siliconvalley.com/mld/siliconvalley/3145388.htm-    Reuters Wire | 04/26/2002 |5    White House sees GDP slowing to 3.5 percent in '02c  H   "...Hubbard, citing private sector forecasts, said U.S. GDP was likelyG    to slow to an average of 3.5 percent in the second, third and fourthNA    quarters. "I'm quite comfortable with those numbers," he said.   @    That stood in sharp contrast to the administration's officialE    forecast, issued in January, for economic growth of 0.7 percent ind+    2002 followed by 3.8 percent in 2003..."e  H : > Or is it that, by & large, OpenVMS boxes just run too damn well and  : > require less staff?a : >  : B : I'd say it is a fallacy and a real business risk to run without F : serious VMS expertise available, no matter how stable the system is.B : However, I'm not the one hiring for businesses that rely on VMS. : 6 A similar problem exists for *nix and Windows systems.  D The lack of sysadmins may be why so many Windows systems didn't haveD the patch for the Code Red worm, including Microsoft's Service Pack/J Hot Fix site, even though the patch had been available for several months.  I : > Would appreciate anyone's advice. VMS has paid my way since 1989, buteG : > it's all looking a bit terminal at the moment ...to the extent that G : > I'm practising saying "would you like fries with that, sir?" with at : > helpful smile on my face.i : E : :-). I too am looking at diversifying into non-IT areas, but that'se! : not easy at the moment either. d : 2 A lot of U.S. IT people are also opting out of IT:  H    http://www.denverpost.com/Stories/0,1413,36%257E33%257E510226,00.html5    Laid-off tech, telecom workers embrace job centers   I   "...But the job that will grow most in the future may be disappointing rK    to those white-collar workers accustomed to $80,000 to $100,000 salariess    of the dot-com heyday.   C    Most of Colorado's job growth over the next six years will be ineG    fields like retail sales, office clerks and cashiers - jobs that payb    less than $21,490 a year.  F    VoiceStream, for instance, just opened a call center in Westminster=    and plans to hire 1,000 people by midsummer. But those are.4    lower-skilled jobs, which pay $25,000 to $30,000.  ?    Many fields are in dire need of workers, including teachers,s8    principals, emergency medical technicians and nurses.  F    The drawbacks are that these positions require a firm commitment to7    switch careers, more training and lower salaries..."o    H --Jerry Leslie   leslie@clio.rice.edu  (my opinions are strictly my own)9   Note: leslie@jrlvax.houston.rr.com is invalid for email3   ------------------------------    Date: 29 Apr 2002 16:26:52 -0700- From: tessier-ashpool@usa.net (Chris Bardell)m3 Subject: Re: UK/EU OpenVMS job market: non-existanta< Message-ID: <9f261edc.0204291526.4481764@posting.google.com>  9 Cheers for the responses guys, interesting points raised.    ------------------------------  # Date: Tue, 30 Apr 2002 01:37:36 GMTe1 From: "David J. Dachtera" <djesys.nospam@fsi.net>n3 Subject: Re: UK/EU OpenVMS job market: non-existantp' Message-ID: <3CCDF847.65AEABC2@fsi.net>f   Chris Bardell wrote: > ! > If anyone has a view on this...a > F > Does anyone think the job market for OpenVMS tech support / sys mgrs< > will ever take off again, specifically in the UK & Europe? > E > One month into the new financial year, but hardly any upturn in thed  > amount of vacancies available. > E > Any light at the end of the tunnel / recession / whatever? Or is itnD > that, by & large, OpenVMS boxes just run too damn well and require
 > less staff?o > G > Would appreciate anyone's advice. VMS has paid my way since 1989, butnE > it's all looking a bit terminal at the moment ...to the extent thatyE > I'm practising saying "would you like fries with that, sir?" with as > helpful smile on my face.u  1 Well, JF makes some good - if unpopular - points.n  E My take is that until such time as OVMS Management does an abrupt andtH complete about-face with regard to advertising, pricing, packaging, etc.H (in other words, "marketing"), VMS will continue to languish in the hell- of attrition until it simply ceases to exist.2  C I should have my "alternative" computer store well established long:F before the untimely demise of that which has supported me since circa. 1983 or so..   -- c David J. Dachterae dba DJE Systemsa http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/a   ------------------------------  % Date: Mon, 29 Apr 2002 11:30:34 -0700eM From: "Kenneth H. Fairfield" <Kenneth.H.Fairfield@spam-be-gone.intel.com.net>  Subject: Re: VAX/Alpha CIe: Message-ID: <3CCD914A.472CE725@spam-be-gone.intel.com.net>   Bob Willard wrote:  - > Your memory is better than mine.  Two nits:o >o5 > 1. The CI King of ZKO spells his name Verell Boaen.d  2 Thanks for the correction and apologies to Verell.. I was afraid phonetics were not enough here...   >sC > 2. CIBCA is the real CI-BI adapter, although there was an earliers9 >    quick-time-to-market (ugly) version named the CIBCI.t > --
 > Cheers, Bobr       -Ken --6 I don't speak for Intel, Intel doesn't speak for me...  
 Ken Fairfieldg! F20 Automation VMS System Supportf kenneth.h.fairfield#intel.com    ------------------------------  % Date: Mon, 29 Apr 2002 21:31:11 +0200r' From: Lasse Jensen <rdx@worldonline.dk>h3 Subject: Re: VAXstation 4000 90 and serial console?f; Message-ID: <3ccd9f85$0$242$edfadb0f@dspool01.news.tele.dk>h   Holi - Holitska Andrs wrote:9  L > In article <3ccd276f$0$97316$edfadb0f@dspool01.news.tele.dk>, Lasse Jensen > <rdx@worldonline.dk> writes:K >> I have a digital VAXstation 4000 90 i want to connect to a PC with LinuxoJ >> and minicom, theres a MMJ serial port on the VAX and a DB9 on the PC, iI >> have tried several different console cable layouts i found on the net,-K >> but none of them worked, so what pins should i connect and what settings@K >> should i use to get the VAX running a terminal on my PC? I can't wait tog% >> get my own VMS machine up running.  >> c > C > Is the S3 switch near the HALT button (under a cover on the front B > side) in the upper position? If it's in the lower position, thenC > the VAX is in graphical console mode, but you need serial consolen > mode.u > + > Here is a figure, on how to make a cable:P > I >             ________________   [DEC  MMJ]      <25 pin>        <9  pin>bK >            /   /       /   /| ===========     ===========     ===========rK >           /   /_______/   / | pin meaning     pin meaning     pin meaningfJ >          / ____________  /  |  5     RD+  ---  2     TXD  ===  3     TXDJ >         / / / / / / / / /   |  2     TD+  ---  3     RXD  ===  2     RXDJ >        / /6/5/4/3/2/1/ /   /   4     RD-  -+-  7     GND  ===  5     GND. >       / /_/_/_/_/_/_/ |   /    3     TD-  -^ >       | |D|R|G|G|T|D| |  /I >       | |S|D|N|N|D|T| | /      6     DSR  ---  20    DTR  ===  4    DTRaI >       | |R|+|D|D|+|R| |/       1     DTR  ---  6     DSR  ===  6     SRf >       |_|_|_|_|_|_|_|_|c >        | . . |/  >        \____/8 >  > Signs: >  >   ---  just connect them >   -+-  a connection point28 >   -^   connect the left side to where the arrow points8 >   ===  these pins have equal meaning on the connectors > ) > 9600 bps, 8 bit, 1 stop bit, no parity.o  G That helped a lot! I still can't write anything to the console, but at ?+ least i now know my VAX is working. Thanks.h   --  ( Lasse Jensen [rdx at worldonline dot dk] [lasse@Earth ~]$ uname -afK Linux earth.lan 2.4.7-10smp #1 SMP Thu Sep 6 17:09:31 EDT 2001 i686 unknownt [lasse@Earth ~]$ uptime =   4:01pm  up  3:55,  4 users,  load average: 0.03, 0.03, 0.00h   ------------------------------  # Date: Mon, 29 Apr 2002 21:09:42 GMT,2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)3 Subject: Re: VAXstation 4000 90 and serial console?e9 Message-ID: <qEiz8.50$r11.854111@cacnews.cac.cpqcorp.net>o  e In article <3ccd9f85$0$242$edfadb0f@dspool01.news.tele.dk>, Lasse Jensen <rdx@worldonline.dk> writes:eH :That helped a lot! I still can't write anything to the console, but at , :least i now know my VAX is working. Thanks.  H   The pinouts and the MMJ stuff are also in the OpenVMS Frequently AskedH   Questions (FAQ) document -- please check there for details and for theG   answers to common questions.  You will also see details on how to setnG   up and run OpenVMS, on the serial port names, and related.  There aretI   also tips on X Windows keymappings for communications with Linux, once uG   you get that far -- and there are some related widgets on the OpenVMSo   Freeware.M    N  ---------------------------- #include <rtfaq.h> -----------------------------N       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------   Date: 29 Apr 2002 16:25:32 GMT- From: "Jim Strehlow" <jimSnoSpam@data911.com>-* Subject: virtual i/o cache v7.2-1 and v7.30 Message-ID: <aajs5s$o4f@dispatch.concentric.net>  4 I believe we need to expand our virtual i/o cache on OpenVMS Alpha v7.2-1.r; Our i/o intensive background server program crashes withouts> opening a sys$error file ... which we believe is caused by not2 having a large enough virtual i/o cache (a guess).  3 The default value is 3200 Kbytes (VCC_MAXSIZE=6400)m! $show memory   !on OpenVMS v7.2-1o ... @ Virtual I/O Cache (Kbytes):        Total        Free      In Use@   Cache Memory                      3200           0        3200 ...   8 The OpenVMS wizard suggests 3% to 10% of physical memory for virtual i/o cache.  5 Easily fixed in sys$system:modparams.dat and autogen.g    % My question is regarding OpenVMS v7.3h< I do not know where the virtual i/o cache appears in the new display of SHOW MEMORY.rL I can verify the value in SYSGEN; but where is it in the new memory display?  @ What is the best source of reading material for me to read up onA the new Extended File Cache in v7.3 and how the virtual i/o cache- is affected?   Thanks in advance.  . Jim Strehlow, OpenVMS Systems Manager, Data911 jimS@data911.com Alameda, CA, USA      1 $show memory/all/full      !on OpenVMS Alpha v7.3a@               System Memory Resources on 29-APR-2002 09:14:46.64  L Physical Memory Usage (pages):     Total        Free      In Use    ModifiedL   Main Memory (1024.00Mb)         131072       25496       97991        7585  B Extended File Cache  (Time of last reset: 17-APR-2002 07:33:33.29)?     Allocated (Mbytes)          474.65    Maximum size (Mbytes). 512.00?     Free (Mbytes)                 0.00    Minimum size (Mbytes)s 0.23>     In use (Mbytes)             474.65    Percentage Read I/Os 62%h8     Read hit rate                   61%   Write hit rate 0%9     Read I/O count             4971927    Write I/O count  2985640t9     Read hit count             3043001    Write hit count  0e@     Reads bypassing cache      1286401    Writes bypassing cache 2187467 =     Files cached open              542    Files cached closede 78F     Vols in Full XFC mode            0    Vols in VIOC Compatible mode 24G     Vols in No Caching mode          0    Vols in Perm. No Caching modeS 0h  1 I/O Statistics - Distributions (MAX_IO_SIZE: 127) 1 ------------------------------------------------- . Transfer Size:      Reads  Read Hits    Writes.    1 Block IO:     686791     658885     86213 ... .  127 Block IO:         44          0         0  . >127<256 Block IO:     12                    0. >255     Block IO:     17                    01 -------------------------------------------------r. Totals:           4972264    3043001   2985640  L Granularity Hint Regions (pages):  Total        Free      In Use    ReleasedL   Execlet code region               1024           0         671         353L   Execlet data region                296           0         244          52L   S0/S1 Executive data region        751           0         751           0L   Resident image code region        1024           0         864         160  L Slot Usage (slots):                Total        Free    Resident     SwappedL   Process Entry Slots                323         247          76           0L   Balance Set Slots                  321         247          74           0  / Nonpaged Dynamic Memory      (Lists + Variable)tL   Current Size (Mb)                 5.42   Current Size (Pagelets)     11120L   Initial Size (Mb)                 4.81   Initial Size (Pagelets)      9856L   Maximum Size (Mb)                20.60   Maximum Size (Pagelets)     42192L   Free Space (Mb)                   1.78   Space in Use (Mb)            3.64L   Largest Var Block (Kb)          488.81   Smallest Var Block (bytes)     64L   Number of Free Blocks             1108   Free Blocks LEQU 64 bytes     189L   Free Blocks on Lookasides          126   Lookaside Space (Kb)       136.06  / Bus Addressable Memory       (Lists + Variable) L   Current Size (Kb)                40.00   Current Size (Pagelets)        80L   Initial Size (Kb)                40.00   Initial Size (Pagelets)        80L   Free Space (Kb)                  14.87   Space in Use (Kb)           25.12L   Largest Var Block (Kb)           13.56   Smallest Var Block (Kb)      1.31L   Number of Free Blocks                2   Free Blocks LEQU 64 bytes       0L   Free Blocks on Lookasides            0   Lookaside Space (bytes)         0   Paged Dynamic MemoryL   Current Size (Mb)                 4.03   Current Size (Pagelets)      8256L   Free Space (Mb)                   1.78   Space in Use (Mb)            2.24L   Largest Var Block (Mb)            1.77   Smallest Var Block (bytes)     16L   Number of Free Blocks              127   Free Blocks LEQU 64 bytes     100   ------------------------------   Date: 29 Apr 02 19:03:28 +0200) From: p_sture@elias.decus.ch (Paul Sture)m0 Subject: Re: What does "affordable" mean to you?) Message-ID: <o$EXGyL+AC6U@elias.decus.ch>s  f In article <3CCCAA03.C7D82409@firstdbasource.com>, Michael Austin <maustin@firstdbasource.com> writes: >> f > F > When companies continually use accounting practices that satisfy theI > morons on wall street, this is what you get.  Idiots that can no longernI > calculate TCO.  I do not know how many times over the past 6 years as a-5 > consultant I have heard cnversations to the effect:LJ >  "I saved my company $1M by NOT buying xyz that would improve throughput > by 20%"  t' >   - how much downtime did you have.. @$ >   "Oh, only about 15hours or so. "J >   - So let's do a bit of math.  The cost per hour for your manufacturingG > site to be down is $300K/hr and because you did not invest wisely, ito2 > actually cost you $4.5M because you saved $1M???- >  "Wweelll, now that you put it that way..."& >  > SHEESH!!! I > No wonder businesses are in the shape they are in...  Even  my 16yr oldr? > son knows more about economics than these so-called "experts"  > @ > (not a great example, but it is late and I have had enough for > tonight....)  % Have a look at these and be dismayed:   ; http://www.timesleader.com/mld/timesleader/news/3146951.htmt; http://www.timesleader.com/mld/timesleader/news/3146952.htmu; http://www.timesleader.com/mld/timesleader/news/3124134.htmr   ------------------------------  # Date: Mon, 29 Apr 2002 18:20:55 GMT 1 From: LESLIE@JRLVAX.HOUSTON.RR.COM (Jerry Leslie)m0 Subject: Re: What does "affordable" mean to you?: Message-ID: <bagz8.11956$Q42.520674@typhoon.austin.rr.com>  * Paul Sture (p_sture@elias.decus.ch) wrote:h : In article <3CCCAA03.C7D82409@firstdbasource.com>, Michael Austin <maustin@firstdbasource.com> writes: : >> l : > H : > When companies continually use accounting practices that satisfy theK : > morons on wall street, this is what you get.  Idiots that can no longer.K : > calculate TCO.  I do not know how many times over the past 6 years as a 7 : > consultant I have heard cnversations to the effect: L : >  "I saved my company $1M by NOT buying xyz that would improve throughput
 : > by 20%"  o) : >   - how much downtime did you have.. 0& : >   "Oh, only about 15hours or so. "L : >   - So let's do a bit of math.  The cost per hour for your manufacturingI : > site to be down is $300K/hr and because you did not invest wisely, itW4 : > actually cost you $4.5M because you saved $1M???/ : >  "Wweelll, now that you put it that way..."v : > 
 : > SHEESH!!! K : > No wonder businesses are in the shape they are in...  Even  my 16yr oldnA : > son knows more about economics than these so-called "experts"e : > B : > (not a great example, but it is late and I have had enough for : > tonight....) : ' : Have a look at these and be dismayed:o : = : http://www.timesleader.com/mld/timesleader/news/3146951.htme= : http://www.timesleader.com/mld/timesleader/news/3146952.htma= : http://www.timesleader.com/mld/timesleader/news/3124134.htmi : , They don't call 'em PHBs & MGMs for nothing.  J That's a marketing opportunity for Compaq to sell (donate?) a VMSCluster, D assuming some ISV has VMS versions of similar applications to what's running on the AS/400.  H --Jerry Leslie   leslie@clio.rice.edu  (my opinions are strictly my own)9   Note: leslie@jrlvax.houston.rr.com is invalid for email,   ------------------------------    Date: 29 Apr 2002 14:43:33 -0500+ From: young_r@encompasserve.org (Rob Young) 0 Subject: Re: What does "affordable" mean to you?3 Message-ID: <gCyfWhE5Nd1e@eisner.encompasserve.org>t  n In article <bagz8.11956$Q42.520674@typhoon.austin.rr.com>, LESLIE@JRLVAX.HOUSTON.RR.COM (Jerry Leslie) writes:, > Paul Sture (p_sture@elias.decus.ch) wrote:   > : ) > : Have a look at these and be dismayed:a > : ? > : http://www.timesleader.com/mld/timesleader/news/3146951.htmT? > : http://www.timesleader.com/mld/timesleader/news/3146952.htmm? > : http://www.timesleader.com/mld/timesleader/news/3124134.htme > : . > They don't call 'em PHBs & MGMs for nothing. > L > That's a marketing opportunity for Compaq to sell (donate?) a VMSCluster, F > assuming some ISV has VMS versions of similar applications to what's > running on the AS/400. >   : 	Don't get too excited about using Wilkes-Barre based DataD 	Systems for examples of "what is wrong in IT."  "A lot" of the pay A 	scale in that area is weak, talent usually matches pay (mostly):e   Position: VMS/Programmer+ Location: Pocono Mountains of Pennsylvania u  H Duties: Data Center Environment in NE PA seeks 1-5 years VMS programmingL experience. Must work well independently. Solid VMS knowledge and excel in a fast paced environment.    Salary: $45 - 49K    Relocation Paid: Yes a      Contact:        Ann M. Silvestri#                 S-H-S Internationali1                 216 North River Street, Suite 550l+                 Wilkes-Barre, PA 18702-2594y                 717-825-3411                  717-825-7790 Fax  E 	But hey, they will relocate you ... so there is an upside I suppose!0C 	Reading the articles, there is a good amount of .. ummm.. not goodOB 	decisions which could very well tie back into the "talent issue."   				Robi   ------------------------------  # Date: Mon, 29 Apr 2002 20:44:58 GMTd1 From: LESLIE@JRLVAX.HOUSTON.RR.COM (Jerry Leslie)30 Subject: Re: What does "affordable" mean to you?: Message-ID: <ehiz8.10117$My4.514273@typhoon.austin.rr.com>  , Rob Young (young_r@encompasserve.org) wrote:= : In article <bagz8.11956$Q42.520674@typhoon.austin.rr.com>, i5 : LESLIE@JRLVAX.HOUSTON.RR.COM (Jerry Leslie) writes:f. : > Paul Sture (p_sture@elias.decus.ch) wrote: :  : > : + : > : Have a look at these and be dismayed:n : > : A : > : http://www.timesleader.com/mld/timesleader/news/3146951.htmtA : > : http://www.timesleader.com/mld/timesleader/news/3146952.htm/A : > : http://www.timesleader.com/mld/timesleader/news/3124134.htm  : > :n0 : > They don't call 'em PHBs & MGMs for nothing. : > N : > That's a marketing opportunity for Compaq to sell (donate?) a VMSCluster, H : > assuming some ISV has VMS versions of similar applications to what's : > running on the AS/400. : >  : < : 	Don't get too excited about using Wilkes-Barre based DataF : 	Systems for examples of "what is wrong in IT."  "A lot" of the pay C : 	scale in that area is weak, talent usually matches pay (mostly):g :  : Position: VMS/Programmer- : Location: Pocono Mountains of Pennsylvania S : J : Duties: Data Center Environment in NE PA seeks 1-5 years VMS programmingL : experience. Must work well independently. Solid VMS knowledge and excel in : a fast paced environment.  :  : Salary: $45 - 49K  :  : Relocation Paid: Yes . :  : " : Contact:        Ann M. Silvestri% :                 S-H-S Internationall3 :                 216 North River Street, Suite 550/- :                 Wilkes-Barre, PA 18702-2594  :                 717-825-3411" :                 717-825-7790 Fax : G : 	But hey, they will relocate you ... so there is an upside I suppose!eE : 	Reading the articles, there is a good amount of .. ummm.. not goodlD : 	decisions which could very well tie back into the "talent issue." : 	 : 				Rob   J $45 - 49K looks very appealing after months of $0.00/year here in Houston.F I'd apply if I wasn't an only child looking after an elderly parent in a nursing home here.  L A lot of the IT jobs still left have had their salary range cut by as much  > as 50%, and many employers no longer pay relocation expenses.   F Then there's the "stealth" campaign to relocate IT and IT-enabled jobs	 offshore:V  I    http://iwsun4.infoworld.com/articles/fe/xml/01/08/20/010820feindia.xml-!    The back office moves to India:  G   "...Revenue from IT-enabled services is likely to continue to grow intH    India despite -- or because of -- the U.S. slowdown, but it is likelyI    to be a low-profile activity, because most customers are wary of beingdG    named. "It doesn't make for good public relations in the U.S. marketnE    to be seen to be moving jobs outside the United States, even if iteH    makes great economic sense," says a representative for a U.S. company.    that outsources customer support to India."  F The next telemarketer that calls may be in India. There's already beenG a report in the comp.dcom.telecom.tech newsgroup that of a call selling < AT&T Wireless from an Indian telemarketer. The Message-ID is  !   : <a8tc4h$mp3$2@news.panix.com>     H --Jerry Leslie   leslie@clio.rice.edu  (my opinions are strictly my own)9   Note: leslie@jrlvax.houston.rr.com is invalid for emaili   ------------------------------  % Date: Tue, 30 Apr 2002 02:21:00 -0000s/ From: Michael Zarlenga <zarlenga@conan.ids.net> 0 Subject: Re: What does "affordable" mean to you?/ Message-ID: <ucrvsc1dek1v32@corp.supernews.com>   0 David J. Dachtera <djesys.nospam@fsi.net> wrote:G : I didn't find the message this evening, but this afternoon I saw (via E : Google) a post which suggested that "affordable" = "cheap, shitty".-   It means I can afford it.    Duh!   --   -- Mike Zarlenga  '       A good friend will help you move. .       A best friend will help you move a body.   ------------------------------  % Date: Mon, 29 Apr 2002 21:54:50 -0700 6 From: "Howard Taylor" <Howard.Taylor@pacificcoast.net>% Subject: What happened to Uwe Zessin?  Message-ID: <3cce2233@nubby2.>  J Uwe Zessin, who singlehandedly managed the port of Python to VMS, has left8 this message at his web page http://www.decus.de/~zessin  / "No software is available from here any longer.o6 The former contents of this web page have been erased./ I have not made them available somewhere else."   3 Does anybody know where Uwe went? Uwe, we miss you!a  ! Howard.Taylor at pacificcoast.netc   ------------------------------  # Date: Mon, 29 Apr 2002 17:58:47 GMTw2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman): Subject: Re: will memory leak program caused system crash?8 Message-ID: <rRfz8.37$vZ.667793@cacnews.cac.cpqcorp.net>  _ In article <873e96d6.0204250553.15fcf454@posting.google.com>, wingwong@witty.com (wing) writes:-  >   Tersely-worded queries are surprisingly difficult to answer.C   Why?  With a minimum of detail, you may see an answer posted to aXE   question that you had not intended to ask -- please see the OpenVMScE   FAQ for an introduction on how to ask questions.  (Why do you care?1F   Well, if you provide the necessary details, then you will typically H   receive an answer to your intended question, and usually more quickly.G   If you omit the details, well, we get to ask you questions about yourl   question.)  < :Will a memroy leak program cause system crash in openvms?    D   No.  User-mode code can happily leak away all its memory, and this/   will not adversely affect system integrity.  e  G   There are possibilities around over-committing system resources that WG   can lead to system hangs and such -- if the system manager configuresfD   the application to consume more pagefile than is available on the E   OpenVMS system, you can cause OpenVMS to hang.  But this should NOT-   crash OpenVMS.  H :Is there any user account setting/run option to set the max memory use  :by a program?  =   Yes, there are.  This is set on a process basis by default.   F   There are several settings involved, and details can and do vary by E   OpenVMS version and platform.  Most central here are the PGFLQUOTA rE   process quota setting and -- depending on the OpenVMS platform and eD   version -- hopefully you will now understand why the inclusion of D   details such as the version and the platform are important -- the F   system parameter VIRTUALPAGECNT.  (There is a section in the OpenVMS    FAQ on INSVIRMEM errors, too.)  A   For processes that are created by a user or an application, theaD   settings are explicit on the RUN command, while the default values/   are acquired from the PQL system parameters.    C   I would encourage a tour through the Programming Concepts Manual, B   as well as through the System Manager's Manuals -- both are partA   of the OpenVMS documentation set, and these (and other manuals).<   will provide invaluable background information on OpenVMS.    N  ---------------------------- #include <rtfaq.h> -----------------------------N       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  % Date: Mon, 29 Apr 2002 23:48:53 -0400p- From: JF Mezei <jfmezei.spamnot@videotron.ca>m Subject: Wither ALL-IN-1 ?, Message-ID: <3CCE141C.EF0442DD@videotron.ca>  B I recently got a newsletter from the All-in-1 office server group.  J They have done work on porting their IMAP server to VMS-MAIL, to appear in& TCPIP Services 5.3 to be available Q2.  L They have abandonned plans to make Office Server/All-IN-1 deal directly withK SMTP mail, giving the excuse that PMDF-MTA does teh job already (anyone whouM has looked at the costs of this solution will simply use PMDF only since onceeK you get PMDF with enough bells and whistles to interface with A1, you don't/ need ALL-IN-1 anymore.   They also have text such as : M "Office server will continue to be available and supported on OpenVMS VAX and4! Alpha for the foreseable future".3  ; No mention of IA64 in the newsletter at all. I take it that1M Office-Server/ALL-IN-1 has officially been made redundant and won't be ported5/ to IA64, or is that still under consideration ?X  I Interesting that all the underlying middleware (except message router, ofi; course) has been announced to move to IA64, but not allin1./  X Looks like Compaq is handing all messaging to Process's PMDF or to Microsoft's Exchange.  N Could the VMS engineers make any comments on the future of VMSmail ? Would anyI enhgancements be envisaged, especially with respect to handling fo MIME ,0 multiple attachements etc ?    ------------------------------  % Date: Mon, 29 Apr 2002 21:32:16 -0400-1 From: Michael Austin <maustin@firstdbasource.com>s Subject: Re: X session2 Message-ID: <3CCDF420.D6FF7267@firstdbasource.com>   "Bruin, J.M. de" wrote:n > Q > I just created a simple command procedure that needs one parameter: the name or 0 > ip-address of the PC to create the display on. > G > This is all I need to get the session manager on my PC using X-Win 32  >  > Mark >  > -----Original Message-----: > From: peter@langstoeger.at [mailto:peter@langstoeger.at]$ > Sent: Sunday, April 28, 2002 21:55 > To: Info-VAX@Mvb.Saic.Come > Subject: Re: X session > ? > In article <3CC24BC9.8090607@qsl.network>, "John E. Malmberg"0 > <wb8tyw@qsl.network> writes: > >James Wiley wrote: D > > > Can anyone tell me how to start an X session on a Open/VMS 7.1 > D > It's "OpenVMS VAX V7.1" or "OpenVMS Alpha V7.1" not "Open/VMS" ;-) > 
 > > > system?  > E > Usually (eg. for customizations) @DECW$SESSIONCOM (without .COM !!)u > L > DECW$SESSIONCOM points normally to SYS$STARTUP:DECW$STARTSM.COM which doesL > besides some setup tasks (sylogin, workingset,...) invoke DECW$SESSION.EXE > B > >            I am trying to use Starnet's X-Win32 to set up an XJ > > > session, and I don't know what command to send the VMS system to get > > > the X server going.  > >1G > >IIRC: after you use the SET DISPLAY command, then the image would bei > >SYS$SYSTEM:DECW$SESSION.EXE > E > I don't know what Starnet's X-Win32 supports/knows as host systems,iK > but even without host support, it will work (but you have to do more ;-).e > L > The recent versions of eXcursion do support VMS/DCL as OS/CLI and thereforI > know exactly how to set a display and you thus only need to specify therL > imagename - or a dollar sign and a DCL command (means "$@DECW$SESSIONCOM") > M > Previous versions of eXcursion did not set the display itself and thereforeoH > required a little work. Eg. there has been a SYS$SYSTEM:PCX$SERVER.COMF > (provided by what ? maybe DECnet/DOS or PCSA or who knows) which didJ > the work on the host. But IIRC it did it for the DECnet transport and soK > may or may not support the TCPIP transport (I beleive, it did support TCP4 > in the latest versions)- > K > It was called with parameters like version#, server#, screen#, transport,. > nodename or similar. > F > So, if X-Win32 doesn't support DCL, you need to specify the commandsF > to setup the display and invoke the image all by yourself. You couldE > use the PIPE command of newer versions of VMS (like yours) or use asE > copy of above mentioned PCX$SERVER.COM. X11 servers for Win usuallylF > do support more than one method to connect to the host, but REXEC isE > mostly the minimum (other methods are RSH, TELNET and DECnet-Task).l > & > So, via DCL PIPE, you could specify: > B > PIPE SET DIS/CR/TR=TCPIP/NO=a.b.c.d/SE=x/SC=y ; @DECW$SESSIONCOM > , > or, via PCX$SERVER.COM, you could specify: > A > @SYS$SYSTEM:PCX$SERVER.COM 4 0 0 TCP a.b.c.d "@DECW$SESSIONCOM"  > F > > > Also, years ago, when we first got our VMS system, we used a DEC; > > > product called eXcursion, but it doesn't work anymorel > G > Which version ? On what M$ application launcher (DOS, WFW, W9x,...) ?tF > What is the error message ? Please be more specific the next time... > J > >                                                          - what is the2 > > > follow on product for that, or is there one? > >i
 > >eXcursion.v > > 1 > >It is sold as part of the Pathworks32 product.a > I > eXcursion is still alive, but is now integrated/included (just like thetG > PowerTerm 525 Terminal Emulator product) in the PATHWORKS-32 product.w > 4 >         http://www.openvms.compaq.com/pathworks32/ > H > Since M$ integrated a TCPIP capability in their OS imitating products,L > the usefulness of the PATHWORKS client (DECnet for M$ and also LAT for M$)C > decreased dramatically and so DEC decided to include the formerlyoD > standalone product eXcursion into the PATHWORKS Client (previouslyM > DECnet-DOS) and named it then PATHWORKS-32 (as it was no longer 16bit then)i* > just to keep the PATHWORKS Client alive. > J > btw: Last time I checked, PATHWORKS-32 was cheaper (but not better ! eg.L > I had 10% performance advantage with NCD PCXware over eXcursion for normalN > work or 600% performance advantage for downloading fonts by twice the price)N > then a pure X11 server of the competion and/or a pure good terminal emulator > of the competition. YMMV > K > PATHWORKS-32 V7.2 is current and officially supports also NT5 (= Win2000)oM > and the euro-sign. It does also run on XP, but only in W2K compabilty mode.n > Q > PATHWORKS-32 V7.3 already entered field test & will soon officially support XP.  >  > HIHy >  > -- > Peter "EPLAN" LANGSTOEGER?' > Network and OpenVMS system specialist  > E-mail  peter@langstoeger.atR > A-1030 VIENNA  AUSTRIA              I'm looking for (a) Network _and_ VMS Job(s)    4 One addition and you will have a much easier time...  H This will return the IP address of the current process.  Login from your< PC and this should be either the IP address or the DNS name..   f$element(1," ",f$getjpi("","TT_ACCPORNAM")) (vms 7.2-1, TCPIP  V5.1) -- / Regards,  7 Michael Austin            Registered Linux User #261163t7 First DBA Source, Inc.    http://www.firstdbasource.comy Sr. Consultant 704-947-1089 (Office). 704-236-4377 (Mobile)o   ------------------------------   End of INFO-VAX 2002.237 ************************