1 INFO-VAX	Mon, 12 Dec 2005	Volume 2005 : Issue 691       Contents:5 Re: Alphaservers and VMS v. NT mentioned in a webblog " Re: calloc() fails, no more memory" Re: calloc() fails, no more memory" Re: calloc() fails, no more memory" Re: calloc() fails, no more memory" Re: calloc() fails, no more memory DHV11 DHQ11 in a VAX4000 QBUS ! Re: DHV11 DHQ11 in a VAX4000 QBUS 1 Re: DVNETRTG (Phase IV routing) Hobbyist license? 6 Re: Extreme backup slowdown after ECO updates (V7.3-1) Re: HP's 4Q FY05 results RE: HP's 4Q FY05 results Re: HP's 4Q FY05 results Re: INSTALLing images J Re: mistake in/suggested enhancements for HELP LEXICALS F$GETDVI ARGUMENTS) Re: Need model number for Infoserver-1000 - Re: propagation of write bitmaps for minicopy A Re: Question about SPAWN/INPUT table in the DCL Dictionary manual A Re: Question about SPAWN/INPUT table in the DCL Dictionary manual A Re: Question about SPAWN/INPUT table in the DCL Dictionary manual A Re: Question about SPAWN/INPUT table in the DCL Dictionary manual A Re: Question about SPAWN/INPUT table in the DCL Dictionary manual " Re: saving recovery data with PCSI" Re: saving recovery data with PCSI Re: shadowing questions 0 Re: Timeout strategy: terminal vs Telnet drivers0 Re: Timeout strategy: terminal vs Telnet drivers0 Re: Timeout strategy: terminal vs Telnet drivers0 RE: Timeout strategy: terminal vs Telnet driversC Re: USB on OpenVMS (was: Re: DS10L hardware configuration question) C Re: USB on OpenVMS (was: Re: DS10L hardware configuration question)   F ----------------------------------------------------------------------  % Date: Mon, 12 Dec 2005 14:04:05 +0000 ) From: Tom Wade <nospam@picard.eurokom.ie> > Subject: Re: Alphaservers and VMS v. NT mentioned in a webblog0 Message-ID: <439D8355.5080208@picard.eurokom.ie>  
 > Microsoft's G > solution to the problem had been to reduce the time it took for NT to  > re-boot!    @ It is certainly true that Windows machines are extremely fast to@ complete a boot.  This is probably because they have had so much more practice at it.  9 --------------------------------------------------------- @ Tom Wade                 | EMail: tee dot wade at eurokom dot ie3 EuroKom                  | Tel:   +353 (1) 296-9696 3 A2, Nutgrove Office Park | Fax:   +353 (1) 296-9697 L Rathfarnham              | Disclaimer:  This is not a disclaimer            G Dublin 14                | Tip:   "Friends don't let friends do Unix !"  Ireland    ------------------------------    Date: 12 Dec 2005 05:29:10 -08003 From: "Joshua Lehrer" <usenet_vms@lehrerfamily.com> + Subject: Re: calloc() fails, no more memory C Message-ID: <1134394150.749178.295150@g44g2000cwa.googlegroups.com>   C In what language are you compiling?  Is it a 32-bit app or a 64-bit G app?  If it is a 32-bit app, you could be out of virtual address space, A which happens in C/C++ when you reach around 1-gig of allocation.   
 joshua lehrer  http://www.lehrerfamily.com/   ------------------------------    Date: 12 Dec 2005 07:50:47 -0600; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler) + Subject: Re: calloc() fails, no more memory 3 Message-ID: <QCVrBHxVCgCI@eisner.encompasserve.org>   o In article <Ca4nf.2690$3Z.2035@newsread1.news.atl.earthlink.net>, Jack Patteeuw <jjpatteeuw@nospam.net> writes:  > 1 > Wouldn't an Alpha require 8 bytes per pointer ?  >   F    Only if you compile for 64 bit addressing.  By default you're using>    32 bit addressing dynamically mapped to 64 bits by the load    instruction.    ------------------------------  % Date: Mon, 12 Dec 2005 08:59:55 -0500  From: "Jilly" <jilly@hp.com>+ Subject: Re: calloc() fails, no more memory , Message-ID: <439d825c$1@usenet01.boi.hp.com>  K If the process is running out PGFLQUOTA then the VMS status for the calloc  M should be INSVIRMEM but if the process is exceeding P0 virtual address space  I then the VMS status should be VASFULL.  Have you pointed the user to the  M Programming Concepts manual for an example on how to allocate memory from P2   space?    0 "Hans Blom" <tocum2@yahoo.com> wrote in message D news:1134303970.bcd67978e90e81688cc82000703ebe4c@fe2.teranews.com... > Hello all,I > I'm running OpenVMS 7.3-2 on an Alphaserver. A programmer has developed I > an application that, in order to work with speed, wants to keep as much I > as possible of the data in memory. At some point he does an calloc() to F > get a memoryarea to keep about 400 000 pointers. He gets a null backF > from the call, basically OpenVMS saying - sorry sir, no more memory! > J > We have tried raising pgflquota, wsextent (in order to decrease need forI > paging) and every other conceivable quota both in sysuaf and sysgen. We I > can see that as long as the program works in memory everything is fine, F > but as soon as wsextent is hit and it has to start paging, it fails. > 1 > I'm stuck! Anybody got any ideas on what to do?  > 	 > Regards  >  > Hans Blom    ------------------------------  % Date: Mon, 12 Dec 2005 10:49:49 -0500 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> + Subject: Re: calloc() fails, no more memory , Message-ID: <439D9C1C.402D8DF3@teksavvy.com>   Jilly wrote: > L > If the process is running out PGFLQUOTA then the VMS status for the callocN > should be INSVIRMEM but if the process is exceeding P0 virtual address space( > then the VMS status should be VASFULL.  F calloc is C function. It doesn't return a status code, it returns NULL9 when something didn't go right and the allocation failed.      Another possible suggestion:  E calloc a small amount of memory, and then use realloc to grow it to a  the huge chunk you need.   ------------------------------  # Date: Mon, 12 Dec 2005 17:32:44 GMT # From: hoff@hp.nospam (Hoff Hoffman) + Subject: Re: calloc() fails, no more memory 0 Message-ID: <0vinf.483$Lv3.122@news.cpqcorp.net>  o In article <1134303970.bcd67978e90e81688cc82000703ebe4c@fe2.teranews.com>, Hans Blom <tocum2@yahoo.com> writes: H :I'm running OpenVMS 7.3-2 on an Alphaserver. A programmer has developedH :an application that, in order to work with speed, wants to keep as muchH :as possible of the data in memory. At some point he does an calloc() toE :get a memoryarea to keep about 400 000 pointers. He gets a null back E :from the call, basically OpenVMS saying - sorry sir, no more memory!  ..  A   What this programmer is very likely encountering is the classic ?   address space limit -- PDP-11/RSX-11 programmers are familiar =   with all the "fun" that went into fitting into 16 bits, and =   OpenVMS VAX progammers (and OpenVMS Alpha programmers using >   versions prior to V7.0) with fitting everything into 32 bitsC   of virtual address space.  In this particular case, OpenVMS Alpha ?   (by default) provides 32-bit addressing, of which 31 comprise A   the process address space, and that is most certainly what your >   programmer is encountering here.  OpenVMS Alpha also offers @   64-bit addressing, allowing programmers to access rather more    virtual address space.  B   You can't fit more than one gigabyte (1GB) into P0 space nor canB   you fit more than 1GB of stack and process control data and such@   into P1 space -- if you are looking to use more than that (as C   certainly appears to be the case here), do seriously consider the B   use the 64-bit addressing capabilities of OpenVMS Alpha V7.0 and%   later releases, and of OpenVMS I64.   <   Put another way, 400,000 longwords -- the default size forA   address pointers on OpenVMS -- will not fit into either P0 nor  C   into P1 space, particularly given that the image, DCL and process F   control structures, and other stack contents also occupy some space.D   If you got clever, you could split this amount of data between P0 G   and P1 spaces, but that is more work than using the 64-bit addressing    capabilities.     N  ---------------------------- #include <rtfaq.h> -----------------------------K     For additional, please see the OpenVMS FAQ -- www.hp.com/go/openvms/faq N  --------------------------- pure personal opinion ---------------------------G        Hoff (Stephen) Hoffman   OpenVMS Engineering   hoff[\0100]hp.com    ------------------------------  % Date: Mon, 12 Dec 2005 11:09:13 -0500 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> & Subject: DHV11 DHQ11 in a VAX4000 QBUS, Message-ID: <439DA0A7.8A281BCC@teksavvy.com>  G The documentation for the various flavours of the VAX 4000 mention only 2 the CXS16 CXB16 and CXY08 serial line controllers.  G Would good old DHV11 and DHQ11 (the DHQ is a half width module, the DHV G is the older one that takes up A B C and D of the Q bus slot) work in a 
 VAX 4000 ?  F I realise that there are mechanical considerations for the actual DB25E connector assemblies. Assuming I have good quality duct tape to affix > the DB25 connector assembly, would the board have any problems& interfacing with the CPU and with VMS?  B Similarly, on my all mighty Microvax II, I have a DILOG SQ739 SCSIE controller. I have a set of console mode commands that start the SCSI E controller configuration menu. (deposit a few bytes here and there to D initialise the IO subsystem and then  START at a specific address to@ branch to the controller program (all done from the >>> prompt).  G If I were to plug this one into a VAX 4000-200 or -600, would it work ? E And more importantly, would the same commands function ? (I know that   the CSR adresses would change).   @ The doc seems to indicate that it works for Microvax II and III.  ' D/P/L 20008004 80000001  ; setup IO map ) D/P/W 20001F40 20        ; enable mapping   9 D/P/W 200000DC 0         ; write IP to reset host adapter 8 D/P/W 200000DE 3FFF      ; write keyword to initiate DMA  C S 200                    ; start virtual terminal driver (the SQ739 
 config menus)     A The middle 2 instructions have memory adresses that depend on the G board's "position" on the QBUS in terms of CSR vectors etc. The first 2  and last one seem "generic".  H Are the generic instructions' memory adresses  part of the QBUS standard* on all VAX models equipped with the QBUS ?   ------------    G On a realted note, I have seem somewhat conflicting info on whether the G motherboard and memory from a VAX4000-200 can be inserted into the QBUS E of an all mighty Microvax II. Were there special boards made for such H upgrades, or can any generic 4000-200 motherboard and memory be insertedJ into the cabinet of a Microvax II (I have the big H9642 with the 2 BA213s.   ------------------------------  % Date: Mon, 12 Dec 2005 12:05:42 -0500 C From: "David Turner, Island Computers US Corp" <dbturner@icusc.com> * Subject: Re: DHV11 DHQ11 in a VAX4000 QBUS: Message-ID: <SZhnf.75538$Y82.46043@bignews4.bellsouth.net>  I If I remember correctly there was a type of Microvax III that went into a 
 BA23 or BA123 9 If that is so, I don't see why you can't go the other way   I As for the DHV11, I remember a long time ago when we sold to ticketmaster A they were using DHV11 with the KA650  Microvax III in a BA23 form    DT   --     David B Turner Island Computers US Corp 2700 Gregory St, Suite 180 Savannah GA 31404  Tel: 912 447 6622 X201 Cell: 912 447 6622 X252  Fax: 912 201 0402  Email: dbturner@icusc.com  Web: http://www.islandco.com% ===================================== < All orders are subject to the following terms and conditions. of sale. These should be read before ordering.% http://www.islandco.com/warranty.html   : "JF Mezei" <jfmezei.spamnot@teksavvy.com> wrote in message& news:439DA0A7.8A281BCC@teksavvy.com...I > The documentation for the various flavours of the VAX 4000 mention only 4 > the CXS16 CXB16 and CXY08 serial line controllers. > I > Would good old DHV11 and DHQ11 (the DHQ is a half width module, the DHV I > is the older one that takes up A B C and D of the Q bus slot) work in a  > VAX 4000 ? > H > I realise that there are mechanical considerations for the actual DB25G > connector assemblies. Assuming I have good quality duct tape to affix @ > the DB25 connector assembly, would the board have any problems( > interfacing with the CPU and with VMS? > D > Similarly, on my all mighty Microvax II, I have a DILOG SQ739 SCSIG > controller. I have a set of console mode commands that start the SCSI G > controller configuration menu. (deposit a few bytes here and there to F > initialise the IO subsystem and then  START at a specific address toB > branch to the controller program (all done from the >>> prompt). > I > If I were to plug this one into a VAX 4000-200 or -600, would it work ? G > And more importantly, would the same commands function ? (I know that ! > the CSR adresses would change).  > B > The doc seems to indicate that it works for Microvax II and III. > ) > D/P/L 20008004 80000001  ; setup IO map + > D/P/W 20001F40 20        ; enable mapping  > ; > D/P/W 200000DC 0         ; write IP to reset host adapter : > D/P/W 200000DE 3FFF      ; write keyword to initiate DMA > E > S 200                    ; start virtual terminal driver (the SQ739  > config menus)  >  > C > The middle 2 instructions have memory adresses that depend on the I > board's "position" on the QBUS in terms of CSR vectors etc. The first 2  > and last one seem "generic". > J > Are the generic instructions' memory adresses  part of the QBUS standard, > on all VAX models equipped with the QBUS ? >  > ------------ >  > I > On a realted note, I have seem somewhat conflicting info on whether the I > motherboard and memory from a VAX4000-200 can be inserted into the QBUS G > of an all mighty Microvax II. Were there special boards made for such J > upgrades, or can any generic 4000-200 motherboard and memory be insertedL > into the cabinet of a Microvax II (I have the big H9642 with the 2 BA213s.   ------------------------------  # Date: Mon, 12 Dec 2005 16:44:48 GMT 1 From: Keith Parris <keithparris_NOSPAM@yahoo.com> : Subject: Re: DVNETRTG (Phase IV routing) Hobbyist license?/ Message-ID: <4Ohnf.464$Kw.226@news.cpqcorp.net>    Bob Armstrong wrote:H >   Is there a reason that DVNETRTG is not included (among the 100 or so' > other PAKs!) in the hobbyist program?   1 If you're on Alpha, try the DVNETEXT PAK instead.    ------------------------------    Date: 12 Dec 2005 08:57:43 -0800( From: "Rich Jordan" <jordan@ccs4vms.com>? Subject: Re: Extreme backup slowdown after ECO updates (V7.3-1) B Message-ID: <1134406663.052977.86980@g49g2000cwa.googlegroups.com>  C Yes, we did.  Used saved feedback data from before the first reboot , (about 70 days worth if I recall correctly).  E New note: just found the V8.2A (and I) Fibre-SCSI V1 patch references G the same error we're seeing on verification.  The error is not noted in E any of the V7.3-1 patches.  The only other references I've found were  in a 1987 post to COV...  4  %BACKUP-E-READERR, error reading MKA500:120905.BCK;/  -BACKUP-E-SHORTBLOCK, save-set block top short   ? Each is followed by a series of block verification errors for a G particular file which generally has not changed (or even been accessed) 
 in months.  F HP has the call, sent to the backup group.  Hopefully they can come up with a fix.    Rich   ------------------------------  % Date: Mon, 12 Dec 2005 11:12:04 +0100 + From: Karsten Nyblad <nospam@nospam.nospam> ! Subject: Re: HP's 4Q FY05 results = Message-ID: <439d4cd0$0$78284$157c6196@dreader1.cybercity.dk>    JF Mezei wrote: G > In that CNET article about that IA64 thing, Marcello said that 90% of 6 > those leaving HP-UX on Pa-Risc choose to go on IA64.  C The 90% is such a high number, that sounds more like the number of  D people voting of the president in a dictatorship than the the whole A truth.  Marcello did not say the customers were moving all their  F applications to IA64.  Thus, if a customer runs 10 applications on an D HP-UX server and choses to move 9 to Linux on x86 and 1 to HP-UX on , IA64, then that customer is part of the 90%.  E > VMS has already suffered much of the attrition due to the "high end C > servers only" negative policy begun under Digital's self-destruct F > procedures. But for HP-UX, being relegated to high end only is a newI > policy, and there is a risk of losing a large portion of installed base H > below that "high end server" line since HP-US still has a large number  > of customers in those markets.  G Whether you like it or not:  VMS does not have the cost structure that  G will allow it to compete with Linux and Windows on the low-end servers  I where high availability is not needed.  And as far as I can see from the  B numbers published from HP, VMS is not a high end OS.  Most of the ? installations are clusters of 1-2 CPU machines.  VMS is a high   availability OS.   ------------------------------  % Date: Mon, 12 Dec 2005 08:08:23 -0500 ' From: "Main, Kerry" <Kerry.Main@hp.com> ! Subject: RE: HP's 4Q FY05 results R Message-ID: <FD827B33AB0D9C4E92EACEEFEE2BA2FB7734EF@tayexc19.americas.cpqcorp.net>   > -----Original Message-----7 > From: Karsten Nyblad [mailto:nospam@nospam.nospam]=20 ! > Sent: December 12, 2005 5:12 AM  > To: Info-VAX@Mvb.Saic.Com # > Subject: Re: HP's 4Q FY05 results  >=20 > JF Mezei wrote: @ > > In that CNET article about that IA64 thing, Marcello said=20
 > that 90% of 8 > > those leaving HP-UX on Pa-Risc choose to go on IA64. >=20G > The 90% is such a high number, that sounds more like the number of=20 H > people voting of the president in a dictatorship than the the whole=20E > truth.  Marcello did not say the customers were moving all their=20 J > applications to IA64.  Thus, if a customer runs 10 applications on an=20H > HP-UX server and choses to move 9 to Linux on x86 and 1 to HP-UX on=20. > IA64, then that customer is part of the 90%. >=20G > > VMS has already suffered much of the attrition due to the "high end E > > servers only" negative policy begun under Digital's self-destruct H > > procedures. But for HP-UX, being relegated to high end only is a new? > > policy, and there is a risk of losing a large portion of=20  > installed base@ > > below that "high end server" line since HP-US still has a=20 > large number" > > of customers in those markets. >=20< > Whether you like it or not:  VMS does not have the cost=20 > structure that=20 ; > will allow it to compete with Linux and Windows on the=20  > low-end servers=20@ > where high availability is not needed.  And as far as I can=20 > see from the=20 F > numbers published from HP, VMS is not a high end OS.  Most of the=20C > installations are clusters of 1-2 CPU machines.  VMS is a high=20  > availability OS. >=20  B So, I guess you are saying low end <4 cpu system Customers are notE interested in security or availability or virues or trojans or worms?   H Perhaps these Customers enjoy QA/testing their applications with monthlyH security patches required for Windows/Linux? Yes, that must be it ..theyE likely have no new applications or functionality testing required, so D those QA groups would have nothing to do if it were not for these OS: vendor recommended monthly security patches anyway, right?  E After all, Red Hat has only released 260+ Linux *security* patches in B 2005 alone on their public web site. Course, that is not a concern
 though right?   G Course, it may be that Customers today are not concerned that 80-90% of F their Windows x86 servers average *peak* utilization during prime timeH is 10-15%?  Perhaps they have not yet figured out what it actually costsF to manage so many distributed servers? Perhaps all those consolidation< RFP's I keep seeing are just a figment of my imagination?=20  ! Does not have the cost structure?   D Please - give me a break. Not all Customers are impressed with shiny@ stuff that looks cool. Some actually look at the big picture.=20   :-)    Regards   
 Kerry Main Senior Consultant  HP Services Canada Voice: 613-592-4660  Fax: 613-591-4477  kerryDOTmainAThpDOTcom (remove the DOT's and AT)=20  4 OpenVMS - the secure, multi-site OS that just works.   ------------------------------  # Date: Mon, 12 Dec 2005 17:31:40 GMT 1 From: Keith Parris <keithparris_NOSPAM@yahoo.com> ! Subject: Re: HP's 4Q FY05 results . Message-ID: <0uinf.482$bD3.9@news.cpqcorp.net>   Tom Linden wrote: 7 > I meant to what amount. I know it was yoy comparison.   C http://h71028.www7.hp.com/ERC/cache/251028-0-0-0-121.html?ERL=true  G indicates HP Integrity Server revenues topped $1 Billion in FY2004. So  E one could conclude that FY2005 HP Integrity Server revenue must have  $ been getting up close to $2 Billion.   ------------------------------  # Date: Mon, 12 Dec 2005 17:12:57 GMT # From: hoff@hp.nospam (Hoff Hoffman)  Subject: Re: INSTALLing images0 Message-ID: <tcinf.475$Lv3.199@news.cpqcorp.net>  e In article <1134134449.219591.263250@g14g2000cwa.googlegroups.com>, ferrariTR512m@hotmail.com writes:   J   FWIW, this is not so much an Rdb question as a generic INSTALL question.  J   Performance questions tend to be a "how high is up" category -- the onlyJ   answer that is feasible outside a "it depends" requires some idea of theI   performance characteristics and the system configuration and the usage.   K   Fully installing reduces the I/O to load the header (since it is loaded), K   and reduces the memory required for shareable portions of images running  >   in parallel to just one copy of the shareable code and data.  G   This might or might not be a net benefit, depending on the system and F   the images involved, and particularly on the system activity and the   system configuration.   I   Would I install Oracle Rdb if I had multiple database users?  Yes.  The H   images are big, and sharing large images is normally a net win.  If I I   had only infrequent users and was memory constrained, I might well look F   at other images, and I would certainly look at resolving the memory D   constraints by reducing the load, increasing memory, or by tuning F   working sets (by quota changes, by encouraging the use of batch (vs 7   interactive), or by any of various other techniques).   G   I'd also be looking at tuning the databases, which most certainly is  E   specific to Rdb, and most certainly an area where Norm Lastovica is D   expert.   (Not to imply that Norm isn't good at OpenVMS system and   performance tuning, too.)   E :1: How can I determine whether these items can be installed and what ( :options are useful/possible (CODE,DATA)  C   Run a baseline, load the images, and run another baseline, and go B   look at the delta.  (There should be various discussions of whenB   to install images in AskQ and in the manuals, and certainly here   in the newsgroups.)   D :2: What will be the memory effect of installing any of these images  D   When you load them, you will see from INSTALL LIST.  (Brute force,.   I know, by easy to do and easy to describe.)  B :3: What will be the performance effect of installing any of these :images   C   Anything from hugely negative (on memory-constrained systems, and F   where these images are not used) to hugely positive (on systems withD   free memory and that are pounding on these images as part of image   activitions.    N  ---------------------------- #include <rtfaq.h> -----------------------------K     For additional, please see the OpenVMS FAQ -- www.hp.com/go/openvms/faq N  --------------------------- pure personal opinion ---------------------------G        Hoff (Stephen) Hoffman   OpenVMS Engineering   hoff[\0100]hp.com    ------------------------------    Date: 12 Dec 2005 11:50:00 -0600. From: brooks@cuebid.zko.hp.nospam (Rob Brooks)S Subject: Re: mistake in/suggested enhancements for HELP LEXICALS F$GETDVI ARGUMENTS , Message-ID: <FtAO1OcGzR9P@cuebid.zko.hp.com>  R helbig@astro.multiCLOTHESvax.de (Phillip Helbig---remove CLOTHES to reply) writes: > ALPHA: > E >          SHDW_CATCHUP_COPYING    SHDW_COPIER_NODE SHDW_DEVICE_COUNT C >          SHDW_GENERATION  SHDW_MASTER             SHDW_MASTER_MBR D >          SHDW_MASTER_NAME        SHDW_MBR_COPY_DONE SHDW_MBR_COUNT= >          SHDW_MBR_MERGE_DONE SHDW_MBR_READ_COST SHDW_MEMBER  >  > VAX: > K >          SHDW_CATCHUP_COPYING    SHDW_MASTER             SHDW_MASTER_NAME M >          SHDW_MEMBER             SHDW_MERGE_COPYING      SHDW_NEXT_MBR_NAME  > K > Note that, on ALPHA, some arguments (like VOLCHAR) are starred.  At the   F > bottom of the list is "* Alpha only".  Shouldn't the non-VAX shadow   > arguments be shadowed as well?  7 That has already been corrected for V8.2.  To wit . . .   G PATH_NOT_RESPONDING*+   PATH_POLL_ENABLED*+     PATH_SWITCH_FROM_TIME*+ 3 PATH_SWITCH_TO_TIME*+   PATH_USER_DISABLED*+    PID 5 PREFERRED_CPU           PROT_SUBSYSTEM_ENABLED* QLEN* 3 RCK                     RCT                     REC = RECSIZ                  REFCNT                  REMOTE_DEVICE 3 RND                     ROOTDEVNAM              RTM 7 SCSI_DEVICE_FIRMWARE_REV*  SDI                  SECTORS A SERIALNUM               SERVED_DEVICE           SET_HOST_TERMINAL B SHDW_CATCHUP_COPYING    SHDW_COPIER_NODE*       SHDW_DEVICE_COUNT*@ SHDW_GENERATION*        SHDW_MASTER             SHDW_MASTER_MBR*? SHDW_MASTER_NAME        SHDW_MBR_COPY_DONE*     SHDW_MBR_COUNT* ; SHDW_MBR_MERGE_DONE*    SHDW_MBR_READ_COST*     SHDW_MEMBER B SHDW_MERGE_COPYING*     SHDW_MINIMERGE_ENABLE*  SHDW_NEXT_MBR_NAME< SHDW_READ_SOURCE*       SHDW_SITE*              SHDW_TIMEOUT   * Alpha and I64 only" + Used with the pathname argument.  J > Would anyone else appreciate at least a one-line description of each of  > the arguments?  C Given that the documentation for the DCL dictionary is available at   7 http://h71000.www7.hp.com/doc/82FINAL/9996/9996PRO.HTML   + and the System Services reference manual at   7 http://h71000.www7.hp.com/doc/82FINAL/4527/4527PRO.HTML   K I don't see anyone here spending time to augment the online help to include ! descriptions for every item code.    --    L Rob Brooks    VMS Engineering -- I/O Exec Group     brooks!cuebid.zko.hp.com   ------------------------------  # Date: Mon, 12 Dec 2005 17:20:44 GMT # From: hoff@hp.nospam (Hoff Hoffman) 2 Subject: Re: Need model number for Infoserver-10000 Message-ID: <Mjinf.477$Lv3.379@news.cpqcorp.net>  q In article <gxYg+juJRoyy@eisner.encompasserve.org>, koehler@eisner.nospam.encompasserve.org (Bob Koehler) writes:  : F :   I'm trying to get a hardware maintenance contract from HP.  One ofE :   my systems, an Infoserver-1000, has the following model number on  :   its tag: :  :      70-30343-03. A02  : < :   HP claims that "703034303A" is not a valid model number.  4   The 2-5-2 format is a classic DIGITAL part number.  ;   The serial number will begin with two characters (tied to =   a DIGITAL site code, such as WM, ZK or NI, followed by some 8   number of letters and (usually mostly or all) numbers.  B :   Is this a punctuation problem, or is HP looking for some other :   number?   
   The latter.   G   The serial number sticker is located on the bottom of the InfoServer  F   1000 itself, and clearly labeled with the model number and with the A   classic "SN" prefix indicating the serial number of the device. D   Pull the cables off the back, slide the release lever to the left,3   and pull the device out of the shelf for a look.)   E   We now have InfoServer support built directly into current versions F   of OpenVMS, as another alternative to the older InfoServer hardware.D   You can use an AlphaServer, for instance, for InfoServer services.C   (Not including the CD recording that was once provided by the old D   and optional InfoServer Scribe package, though that capability is 2   also available within OpenVMS V7.3-1 and later.)  N  ---------------------------- #include <rtfaq.h> -----------------------------K     For additional, please see the OpenVMS FAQ -- www.hp.com/go/openvms/faq N  --------------------------- pure personal opinion ---------------------------G        Hoff (Stephen) Hoffman   OpenVMS Engineering   hoff[\0100]hp.com    ------------------------------    Date: 12 Dec 2005 11:55:05 -0600. From: brooks@cuebid.zko.hp.nospam (Rob Brooks)6 Subject: Re: propagation of write bitmaps for minicopy, Message-ID: <odFryngqK$sS@cuebid.zko.hp.com>  . Ken Fairfield <my.full.name@intel.com> writes:  C >     I find the inline HELP for MOUNT/POLICY=MINICOPY and DISMOUNT # > /POLICY=MINICOPY confusing! :-(      [...]   G I've forwarded this note to the engineer who is most familiar with this  aspect of minicopy.    --    L Rob Brooks    VMS Engineering -- I/O Exec Group     brooks!cuebid.zko.hp.com   ------------------------------    Date: 12 Dec 2005 08:56:18 -0600 From: briggs@encompasserve.orgJ Subject: Re: Question about SPAWN/INPUT table in the DCL Dictionary manual3 Message-ID: <xlHiO2ZuCRsS@eisner.encompasserve.org>   j In article <1134093015.964975.170830@g14g2000cwa.googlegroups.com>, "AEF" <spamsink2001@yahoo.com> writes: > Hello, >  > OK, here's a VMS question: > F > Can someone explain to me the following excerpt from the SPAWN/INPUT& > documentation in the DCL dictionary? >  > [begin excerpt]  > F > You cannot explicitly specify non-record-oriented, process-permanentD > files (NRO PPFs) with the /INPUT qualifier. The system displays anC > error message when it encounters such a file as the value for the  > /INPUT qualifier.  > I > Note that when NRO PPFs are used as implicit input (that is, the /INPUT C > qualifier is not specified and SYS$INPUT is a NRO PPF), the SPAWN 7 > command can succeed, as shown in the following table:  > + > Process type    SYS$INPUT  Implicit Input + > --------------  ---------  -------------- ( > Interactive     NRO PPF    SYS$COMMAND( > Noninteractive  NRO PPF    Null device& > Any             Any other  SYS$INPUT > C > If SYS$INPUT is a terminal, it cannot have an associated terminal 
 > mailbox. >  > [end excerpt]  > ; > OK, what is a non-record-oriented process-permanent file?   ( Let's start with process-permanent file.  A A process permenent file is one that is pre-opened and stays open & from one image activation to the next.  C Ordinary files opened up by a program are automatically closed when 4 the program exits.  Process permanent files are not.  @ Process permanent files are referred to by an associated logicalD name such as "SYS$INPUT", "SYS$OUTPUT", "SYS$ERROR" or "JOHNS_FILE".  @ The translation of that logical name has a special 4 byte prefixE that is usually invisible (i.e. $ SHOW LOGICAL and $ SHOW TRANSLATION E suppress the prefix automatically).  The last two bytes of the prefix 3 encode a 16 bit RMS IFI (internal file identifier).   F When a program does an OPEN on a logical name in this format, the fileD is not actually opened.  Instead, the resulting RABs and FABs end up6 pointing to the file through the already open context.  E Of course, process-permanent files are process-specific.  If you have H a sub-process, the 16 bit RMS IFI from the parent process is meaninglessG to the sub-process.  The file is open in parent context, not subprocess  context.  J If you have FOO.BAR (an ordinary file on disk) open as a process permanentB file with logical name FOO, and you try to use it from both parentB and subprocess context then you have a problem -- do you feed themC both a record stream starting in mid-file?  The same record stream? @ Do you divvy out records to each processes as they ask for them?E That's the problem set that the "no non record-oriented PPFs" rule is  intended to legislate away.   % Now, what's a "record-oriented" file?   B A record-oriented file is a mailbox, or a card reader.  What otherG platforms might call a "unit record" device.  It is a sequential device   that returns a record at a time.  . I don't know whether serial terminals qualify.  D The key thing that a record oridented device buys you is the abilityC to "divvy out" input between multiple competing processes.  You can C hand them each an I/O channel to the device and let the device feed * input on a first come, first served basis.  & > How does one create such a creature?   $ OPEN FOO FOO.BAR /READ  B There.  Now you have a process permanent file accessed through the logical name FOO.   ? > Does one use the DEFINE command to make it implicit input via + > SYS$INPUT? Any special quals needed? Etc.   D DCL plays funny games with SYS$INPUT and SYS$OUTPUT.  In particular,F it opens up SYS$OUTPUT as a PPF whenever you do a $ DEFINE SYS$OUTPUT.  I > Can someone please give some concrete examples illustrating the various  > scenarios in the table?   ( > Interactive     NRO PPF    SYS$COMMAND  I <Typing away at an interactive command prompt, you invoke a command file>  $ @COMMAND_FILE.COM ( $ ! This is the text of COMMAND_FILE.COM $ SPAWN E <the spawned subprocess takes input from SYS$COMMAND rather than from  COMMAND_FILE.COM>  $ EXIT ! from COMMAND_FILE.COM  ( > Noninteractive  NRO PPF    Null device  L <In the context of a BATCH job, in BATCH_FILE.COM a command file is invoked> $ @COMMAND_FILE.COM ' $! This is the text of COMMAND_FILE.COM  $ SPAWN ? <the spawned subprocess takes input from NLA0: rather than from  COMMAND_FILE.COM>  $ EXIT ! from COMMAND_FILE.COM  & > Any             Any other  SYS$INPUT  G <You are running a DCL procedure taking input from a mailbox, MBA1234:>  $ SPAWN /NOWAIT 2 <the spawned subprocess takes input from MBA1234:>F <and the parent process continues to take input from MBA1234: as well>7 <which is one reason you might not want to use /NOWAIT>   0 > And the same questions for the output version.  / > And what is an "associated terminal mailbox"?   B One of the fancy things that you can do with a VMS terminal device? is to associate a mailbox with it.  This mailbox gets specially B formatted messages delivered to it when various events occur.  ForA instance, when there is unsolicited input (user presses a key and ; there is no read I/O active), when the modem hangs up, etc.   C You can only have one mailbox associated with a terminal at a time.   D I've never been clear on what problems would manifest if you allowedE a second process to use a channel to a terminal device when there was 1 already a mailbox association in another process.    	John Briggs   ------------------------------    Date: 12 Dec 2005 07:57:45 -0800$ From: "AEF" <spamsink2001@yahoo.com>J Subject: Re: Question about SPAWN/INPUT table in the DCL Dictionary manualC Message-ID: <1134403065.283312.274140@z14g2000cwz.googlegroups.com>    briggs@encompasserve.org wrote: l > In article <1134093015.964975.170830@g14g2000cwa.googlegroups.com>, "AEF" <spamsink2001@yahoo.com> writes:
 > > Hello, > >  > > OK, here's a VMS question: > > H > > Can someone explain to me the following excerpt from the SPAWN/INPUT( > > documentation in the DCL dictionary? > >  > > [begin excerpt]  > > H > > You cannot explicitly specify non-record-oriented, process-permanentF > > files (NRO PPFs) with the /INPUT qualifier. The system displays anE > > error message when it encounters such a file as the value for the  > > /INPUT qualifier.  > > K > > Note that when NRO PPFs are used as implicit input (that is, the /INPUT E > > qualifier is not specified and SYS$INPUT is a NRO PPF), the SPAWN 9 > > command can succeed, as shown in the following table:  > > - > > Process type    SYS$INPUT  Implicit Input - > > --------------  ---------  -------------- * > > Interactive     NRO PPF    SYS$COMMAND* > > Noninteractive  NRO PPF    Null device( > > Any             Any other  SYS$INPUT > > E > > If SYS$INPUT is a terminal, it cannot have an associated terminal  > > mailbox. > >  > > [end excerpt]  > > = > > OK, what is a non-record-oriented process-permanent file?  > * > Let's start with process-permanent file. > C > A process permenent file is one that is pre-opened and stays open ( > from one image activation to the next. > E > Ordinary files opened up by a program are automatically closed when 6 > the program exits.  Process permanent files are not. > B > Process permanent files are referred to by an associated logicalF > name such as "SYS$INPUT", "SYS$OUTPUT", "SYS$ERROR" or "JOHNS_FILE". > B > The translation of that logical name has a special 4 byte prefixG > that is usually invisible (i.e. $ SHOW LOGICAL and $ SHOW TRANSLATION G > suppress the prefix automatically).  The last two bytes of the prefix 5 > encode a 16 bit RMS IFI (internal file identifier).  > H > When a program does an OPEN on a logical name in this format, the fileF > is not actually opened.  Instead, the resulting RABs and FABs end up8 > pointing to the file through the already open context. > G > Of course, process-permanent files are process-specific.  If you have J > a sub-process, the 16 bit RMS IFI from the parent process is meaninglessI > to the sub-process.  The file is open in parent context, not subprocess 
 > context. > L > If you have FOO.BAR (an ordinary file on disk) open as a process permanentD > file with logical name FOO, and you try to use it from both parentD > and subprocess context then you have a problem -- do you feed themE > both a record stream starting in mid-file?  The same record stream? B > Do you divvy out records to each processes as they ask for them?G > That's the problem set that the "no non record-oriented PPFs" rule is  > intended to legislate away.   G OK, I knew what PPF's are, but not with that much internal detail! Good  to know. Thanks.  ' > Now, what's a "record-oriented" file?  > D > A record-oriented file is a mailbox, or a card reader.  What otherI > platforms might call a "unit record" device.  It is a sequential device " > that returns a record at a time.  E I thought record-oriented meant that it said so in the output of SHOW ? DEVICE/FULL command. I would think that files and terminals are B record-oriented devices. The only example of a non-record-orientedB device I could think of would be an unformatted file, but who uses those with SPAWN?   0 > I don't know whether serial terminals qualify. > F > The key thing that a record oridented device buys you is the abilityE > to "divvy out" input between multiple competing processes.  You can E > hand them each an I/O channel to the device and let the device feed , > input on a first come, first served basis. > ( > > How does one create such a creature? >  > $ OPEN FOO FOO.BAR /READ > D > There.  Now you have a process permanent file accessed through the > logical name FOO.   ; I know how to do that. What I don't know is how to create a D NON-record-oriented PPF. FOO.BAR is record-oriented, no? It's a file9 that contains records. Therefore I would think that it is  record-oriented.  A > > Does one use the DEFINE command to make it implicit input via - > > SYS$INPUT? Any special quals needed? Etc.  > F > DCL plays funny games with SYS$INPUT and SYS$OUTPUT.  In particular,H > it opens up SYS$OUTPUT as a PPF whenever you do a $ DEFINE SYS$OUTPUT. > K > > Can someone please give some concrete examples illustrating the various  > > scenarios in the table?  > * > > Interactive     NRO PPF    SYS$COMMAND > K > <Typing away at an interactive command prompt, you invoke a command file>  > $ @COMMAND_FILE.COM * > $ ! This is the text of COMMAND_FILE.COM	 > $ SPAWN G > <the spawned subprocess takes input from SYS$COMMAND rather than from  > COMMAND_FILE.COM>   > $ EXIT ! from COMMAND_FILE.COM  ? But SYS$INPUT and SYS$COMMAND are both record-oriented devices:    $ SH DEV/FULL SYS$INPUT   C Terminal NTA3:, device type unknown, is online, enabled as operator 	 terminal, -     record-oriented device, carriage control.   <     Error count                    0    Operations completed    2648 1     Owner process           "_NTA3:"    Owner UIC 	 [FELDMAN] 0     Owner process ID        0001F43C    Dev Prot S:RWPL,O:RWPL,G,W ;     Reference count                2    Default buffer size      132    $ SH DEV/FULL SYS$COMMAND   C Terminal NTA3:, device type unknown, is online, enabled as operator 	 terminal, -     record-oriented device, carriage control.   <     Error count                    0    Operations completed    2660 1     Owner process           "_NTA3:"    Owner UIC 	 [FELDMAN] 0     Owner process ID        0001F43C    Dev Prot S:RWPL,O:RWPL,G,W ;     Reference count                2    Default buffer size      132     * > > Noninteractive  NRO PPF    Null device > N > <In the context of a BATCH job, in BATCH_FILE.COM a command file is invoked> > $ @COMMAND_FILE.COM ) > $! This is the text of COMMAND_FILE.COM 	 > $ SPAWN A > <the spawned subprocess takes input from NLA0: rather than from  > COMMAND_FILE.COM>   > $ EXIT ! from COMMAND_FILE.COM  @ In this case, SYS$INPUT points to the command file which is also record-oriented.  ( > > Any             Any other  SYS$INPUT > I > <You are running a DCL procedure taking input from a mailbox, MBA1234:>  > $ SPAWN /NOWAIT 4 > <the spawned subprocess takes input from MBA1234:>H > <and the parent process continues to take input from MBA1234: as well>9 > <which is one reason you might not want to use /NOWAIT>   B Mailboxes are also record-oriented. How do you run a DCL procedure taking input from a mailbox?  C Apparently the confusion is coming from the NRO term. I thought all D these examples you show use record-oriented devices. So what's going7 on? I think this term, NRO, is the crux of the problem.   G BTW, thanks for responding. I posted this before and no one touched it. 9 Maybe now we can finally get to the bottom of it. Thanks!    > 2 > > And the same questions for the output version. > 1 > > And what is an "associated terminal mailbox"?  > D > One of the fancy things that you can do with a VMS terminal deviceA > is to associate a mailbox with it.  This mailbox gets specially D > formatted messages delivered to it when various events occur.  ForC > instance, when there is unsolicited input (user presses a key and = > there is no read I/O active), when the modem hangs up, etc.  > E > You can only have one mailbox associated with a terminal at a time.  > F > I've never been clear on what problems would manifest if you allowedG > a second process to use a channel to a terminal device when there was 3 > already a mailbox association in another process.    Thanks again.    >  > 	John Briggs   AEF    ------------------------------  % Date: Mon, 12 Dec 2005 16:15:22 +0000 0 From: Chris Sharman <chris.sharman@sorry.nospam>J Subject: Re: Question about SPAWN/INPUT table in the DCL Dictionary manual4 Message-ID: <dnk7mq$rbm$1$8302bc10@news.demon.co.uk>  
 AEF wrote:' >>Now, what's a "record-oriented" file?  >>D >>A record-oriented file is a mailbox, or a card reader.  What otherI >>platforms might call a "unit record" device.  It is a sequential device " >>that returns a record at a time.  D > Mailboxes are also record-oriented. How do you run a DCL procedure > taking input from a mailbox?  H You open a mailbox (with one of the tools from freeware), and use it as  sys$input - no magic.   E > Apparently the confusion is coming from the NRO term. I thought all F > these examples you show use record-oriented devices. So what's going9 > on? I think this term, NRO, is the crux of the problem.    Indeed. B I think when it says "NRO file", it means a file on an NRO device.? So your terminal is RO, but a DCL procedure (on a disk) is NRO. % Terminals, mailboxes, and NL: are RO.  Disks are NRO.   Chris    ------------------------------    Date: 12 Dec 2005 08:27:53 -0800$ From: "AEF" <spamsink2001@yahoo.com>J Subject: Re: Question about SPAWN/INPUT table in the DCL Dictionary manualB Message-ID: <1134404873.834036.55340@g47g2000cwa.googlegroups.com>   Chris Sharman wrote: > AEF wrote:) > >>Now, what's a "record-oriented" file?  > >>F > >>A record-oriented file is a mailbox, or a card reader.  What otherK > >>platforms might call a "unit record" device.  It is a sequential device $ > >>that returns a record at a time. > F > > Mailboxes are also record-oriented. How do you run a DCL procedure  > > taking input from a mailbox? > I > You open a mailbox (with one of the tools from freeware), and use it as  > sys$input - no magic.  > G > > Apparently the confusion is coming from the NRO term. I thought all H > > these examples you show use record-oriented devices. So what's going; > > on? I think this term, NRO, is the crux of the problem.  > 	 > Indeed. D > I think when it says "NRO file", it means a file on an NRO device.A > So your terminal is RO, but a DCL procedure (on a disk) is NRO. ' > Terminals, mailboxes, and NL: are RO.  > Disks are NRO.  G But how can a DCL command procedure be non-record-oriented? It contains F records! Each command consists of one or more records. How can that beB non-record-oriented? The doc says NRO PPF with F meaning file, notD device. Well, if what you say is the case, the doc is misleading. ByF that logic, the disk is in the computer which is in the room, which isG non-record-oriented, so nothing on the computer can be record-oriented!   D And what about terminals? Well, they're record-oriented, but they'reC not files! I guess what John Briggs posted is what was meant by the  doc.   Thanks for your input.   >  > Chris    ------------------------------    Date: 12 Dec 2005 12:37:08 -0600 From: briggs@encompasserve.orgJ Subject: Re: Question about SPAWN/INPUT table in the DCL Dictionary manual3 Message-ID: <TSb$i$N5SEcR@eisner.encompasserve.org>   j In article <1134403065.283312.274140@z14g2000cwz.googlegroups.com>, "AEF" <spamsink2001@yahoo.com> writes:E > Apparently the confusion is coming from the NRO term. I thought all F > these examples you show use record-oriented devices. So what's going9 > on? I think this term, NRO, is the crux of the problem.   > Disk devices are not record oriented.  They are file oriented.  I Disk EISNER$DRA3:, device type 7 Member RAID 5, is online, mounted, file- O     oriented device, shareable, available to cluster, error logging is enabled.   M Device MBA8964:, device type local memory mailbox, is online, record-oriented &     device, shareable, mailbox device.  O Terminal VTA2621:, device type VT200 Series, is online, record-oriented device,      carriage control.   F Again, as I wrote in my first response, the issue is file positioning.D If you are trying to pass an open file complete with record positionE to the subprocess, that's more difficult than simply passing a device E name.  And if you are trying to reposition the parent record position D to reflect records consumed by the child, that's more difficult thanJ simply reading the next record from the mailbox/card reader/user keyboard.  K The important distinction is that record-oriented devices are non-seekable.    ------------------------------  # Date: Mon, 12 Dec 2005 16:24:44 GMT 3 From: hammond@not@peek.ssr.hp.com (Charlie Hammond) + Subject: Re: saving recovery data with PCSI / Message-ID: <gvhnf.459$fA3.52@news.cpqcorp.net>   E In article <dnelh8$r9h$1@online.de>, helbig@astro.multiCLOTHESvax.de  2 (Phillip Helbig---remove CLOTHES to reply) writes:  = >Has anyone successfully used the PRODUCT UNDO PATCH command?   8 You can save recovery data ONLY when installing PATCHES.                   A You can save recovery data for any number of patch installations, C provided that ALL the patch installation save recovery data, and NO ) other install/remove operations intevene.   B If you install patches A, B and C, saving recovery data, that data will be lost if you then  6     - Install patch D without saving recovery data, or*     - Install or remove a layered product.  I PCSI will warn you if you perform an operation that will require removing  previously saved recovery data.   G PRODUCT DELETE RECOVERY_DATA can explicitly delete saved recovery data. . (Then you will NOT get the warning from PCSI.)  L PRODUCT UNDO PATCH can rollback one or more patches for which there is savedK recovery data.  The rollback happens in reverse chronological order -- i.e. K the reverse of the order in which the patches were installed.  If there is  G recoverdy data for more than one patch, you "undo" only the most recent L patch(es), or all of them.  However, you cannot "skip" on or more patches --" the removal must be reverse order.    If this doesn't help, ask again.   --  J       Charlie Hammond -- Hewlett-Packard Company -- Ft Lauderdale  FL  USAF           (hammond@not@peek.ssr.hp.com -- remove "@not" when replying)J       All opinions expressed are my own and not necessarily my employer's.   ------------------------------  # Date: Mon, 12 Dec 2005 16:14:53 GMT 3 From: hammond@not@peek.ssr.hp.com (Charlie Hammond) + Subject: Re: saving recovery data with PCSI 0 Message-ID: <1mhnf.457$fA3.251@news.cpqcorp.net>  E In article <pan.2005.12.11.21.18.05.324331@staffan.tjernstrom.name>,  G Staffan Tjernstrom <thisisafakeaddress@staffan.tjernstrom.name> writes:   3 >The trick is in the last digit of the PCSI file :-  >  >1 - Full (Layered Product)  >2 - Operating System J >3 - Partial (An upgrade to currently installed software). Changes product >version of installed software/ >4 - Partial (A correction to currently install K >installed software). Does not change the version of the installed software 4 >5 - Platform (Integrated set of software products)  >6 - Transition (ProductI >information used to register a product that was provided by VMSINSTAL or 
 >similar) ; >7 - Mandatory Update. Functionality the same as ($) above.   , *** number 4 should be PATCH. (3 is partial)     --  J       Charlie Hammond -- Hewlett-Packard Company -- Ft Lauderdale  FL  USAF           (hammond@not@peek.ssr.hp.com -- remove "@not" when replying)J       All opinions expressed are my own and not necessarily my employer's.   ------------------------------  # Date: Mon, 12 Dec 2005 16:48:47 GMT 1 From: Keith Parris <keithparris_NOSPAM@yahoo.com>   Subject: Re: shadowing questions/ Message-ID: <PRhnf.466$Kw.366@news.cpqcorp.net>    Ken Fairfield wrote:G >> Is there some way to allow a MINICOPY in the case of unplanned node   >> crashes?  ... >  >     No.  End of story.  I If the Shadowing developer has his way, the story will not end like this.    ------------------------------  % Date: Mon, 12 Dec 2005 01:57:58 -0500 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> 9 Subject: Re: Timeout strategy: terminal vs Telnet drivers , Message-ID: <439D1F69.C2EC7417@teksavvy.com>   "John Gemignani, Jr." wrote:J > 1) The $QIO/$SETIMR/$CANCEL mechanism ~does~ work and does not close the  H I have this implemented already to detect a dead link. But it isn't niceF and clean like the terminal driver interface (you need some persistantG structure specific to that link to keep the event flag number etc etc).   M > 2) You can use the TNQIO front end for the socket and get the same behavior - > that you described  with terminal timeouts.   H That got my attention. Where do I find documentation on this TNQIO beastH ????? Is it available for TCPIP Servioces 5.3 ? (the last common version for all VMS platforms).    ------------------------------    Date: 12 Dec 2005 07:54:13 -0600; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler) 9 Subject: Re: Timeout strategy: terminal vs Telnet drivers 3 Message-ID: <fWdZpaelQci0@eisner.encompasserve.org>   \ In article <439CE6E2.D659D5D0@teksavvy.com>, JF Mezei <jfmezei.spamnot@teksavvy.com> writes: > J > The TCPIP stack does not support timeouts in read QIOs. What is the bestB > way to emulate the above performance for an inbound telnet style" > communcation to an application ?  J    Post a _hughe_ read, and process according to the actual size read when    the read completes.  8    This is valid in any properly operating TCP/IP stack.   ------------------------------  % Date: Mon, 12 Dec 2005 10:48:09 -0500 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> 9 Subject: Re: Timeout strategy: terminal vs Telnet drivers , Message-ID: <439D9BB8.2403BE03@teksavvy.com>   Bob Koehler wrote:L >    Post a _hughe_ read, and process according to the actual size read when >    the read completes.  , What causes a read to complete  for TCPIP ?   G If I request a read of 500 characters and the system at the other sends . only 400, what/how/why does my read complete ?   ------------------------------  % Date: Mon, 12 Dec 2005 11:23:17 -0500 # From: "Dan Allen" <dallen@nist.gov> 9 Subject: RE: Timeout strategy: terminal vs Telnet drivers : Message-ID: <JFEPKAPBPMDFDBOIANGDGEHPHBAA.dallen@nist.gov>   > -----Original Message-----6 > From: JF Mezei [mailto:jfmezei.spamnot@teksavvy.com]* > Sent: Monday, December 12, 2005 10:48 AM > To: Info-VAX@Mvb.Saic.Com ; > Subject: Re: Timeout strategy: terminal vs Telnet drivers  >  >  > Bob Koehler wrote:N > >    Post a _hughe_ read, and process according to the actual size read when > >    the read completes. > . > What causes a read to complete  for TCPIP ?  > I > If I request a read of 500 characters and the system at the other sends 0 > only 400, what/how/why does my read complete ? >   A Set the socket non-blocking or use a timer to interrupt the read.    Dan    ------------------------------  # Date: Mon, 12 Dec 2005 16:58:17 GMT # From: hoff@hp.nospam (Hoff Hoffman) L Subject: Re: USB on OpenVMS (was: Re: DS10L hardware configuration question)0 Message-ID: <J_hnf.469$Lv3.375@news.cpqcorp.net>  b In article <dn84tr$i9v$1@pcls4.std.com>, moroney@world.std.spaamtrap.com (Michael Moroney) writes:  D :I'm kinda peeved that the USB on the DS10L seems to be deliberately6 :disabled so I can't even play with unsupported stuff.  L   I was discouraged, as well.  Unfortunately, the latent USB hardware (whichJ   was never supported, and was never part of the advertised product, BTW) K   didn't work, and a decision was made to disable it rather than to receive M   and to process the problem reports that were to be expected if the hardware K   was left visble and potentially enabled -- even if unsupported.  The PCI  L   USB card is the approach, if you want to add and use USB in an unsupported?   fashion on those AlphaServer systems that lack supported USB.   K   The USB PCI cards are variously offered for US$10 or so, and usually well J   under US$30, based on various prices I've seen.  (I am not counting whatL   I saw offered at Singapore's Sim Lim City -- PCI USB cards were absolutelyL   everywhere, as was pretty much any other electronic device you might want.G   Singapore was part of the Asia Pacific portion of this year's OpenVMS K   Technical Update Days (TUD), and Sim Lim City is certainly computer geek     heaven.  But I digress.)  I   The USB on the Integrity servers and on specific AlphaServer systems is >   supported, does work, and is part of the advertized product.    N  ---------------------------- #include <rtfaq.h> -----------------------------K     For additional, please see the OpenVMS FAQ -- www.hp.com/go/openvms/faq N  --------------------------- pure personal opinion ---------------------------G        Hoff (Stephen) Hoffman   OpenVMS Engineering   hoff[\0100]hp.com    ------------------------------  + Date: Mon, 12 Dec 2005 18:11:11 +0000 (UTC) 7 From: moroney@world.std.spaamtrap.com (Michael Moroney) L Subject: Re: USB on OpenVMS (was: Re: DS10L hardware configuration question)( Message-ID: <dnkefv$hck$1@pcls4.std.com>  % hoff@hp.nospam (Hoff Hoffman) writes:   c >In article <dn84tr$i9v$1@pcls4.std.com>, moroney@world.std.spaamtrap.com (Michael Moroney) writes:   F > :I'm kinda peeved that the USB on the DS10L seems to be deliberately8 > :disabled so I can't even play with unsupported stuff.  M >  I was discouraged, as well.  Unfortunately, the latent USB hardware (which J >  was never supported, and was never part of the advertised product, BTW)L >  didn't work, and a decision was made to disable it rather than to receiveN >  and to process the problem reports that were to be expected if the hardwareB >  was left visble and potentially enabled -- even if unsupported.  F So I guess it was like the ISDN port on the old Flamingos (or did that' work but there was never a VMS driver?)   D If it didn't work and support never offered, I'm surprised it wasn't  quickly "value engineered" away.   >  The PCI  M >  USB card is the approach, if you want to add and use USB in an unsupported @ >  fashion on those AlphaServer systems that lack supported USB.  L >  The USB PCI cards are variously offered for US$10 or so, and usually wellK >  under US$30, based on various prices I've seen.  (I am not counting what M >  I saw offered at Singapore's Sim Lim City -- PCI USB cards were absolutely M >  everywhere, as was pretty much any other electronic device you might want.   G Which ones work with the OpenVMS USB driver?  Do they all work or only  J ones with certain chips, or only certain cards or whatever?  I see severalI kinds of those cards in places like CompUSA/Best Buy/Circuit City as well  as PC shows.  4 I have another Alpha I could plug a USB PCI card in.   ------------------------------   End of INFO-VAX 2005.691 ************************