1 INFO-VAX	Thu, 17 Aug 2000	Volume 2000 : Issue 458       Contents:. Re: /float qualifier and resulting performance. Re: /float qualifier and resulting performance Adding new mail accounts& RE: Adjusting Width of Show Users Info& Re: Adjusting Width of Show Users Info Re: Caching RMS Block I/O ' Cluster transaction processing paradigm + Re: Cluster transaction processing paradigm + Re: Cluster transaction processing paradigm  Re: Copy from VMS to NT server Re: Copy from VMS to NT server$ Re: DECnet Plus copying time problem Re: Detached process question  Re: Detached process question  Re: Detached process question  Re: Detached process question  Re: Detached process question * DJE Systems DECUS Presentations on the Web. Re: DJE Systems DECUS Presentations on the Web Re: dose vms run on intel  Re: dose vms run on intel  Re: dose vms run on intel  Re: dose vms run on intel  Re: File Headers Re: File Headers Re: File Headers$ Re: FTP to/from Alpha boxes - solved! Re: Help with pagefile full error  Help with pagefile full error 2 Re: Here we go again - WTB/T/etc source listing CD2 Re: Here we go again - WTB/T/etc source listing CD2 Re: Here we go again - WTB/T/etc source listing CD4 Re: How to backup files since the last image backup.( Re: How to fake a device full situation.( Re: How to fake a device full situation.# Re: HP LJ8000 won't print correctly " If Airlines were operating systems# Re: Ken Olson's "snake oil" comment  Re: LN03 Printer Issues  Re: Mail Re: Mass adding of users?  Re: Mass adding of users?  Re: Mass adding of users?  Newbie Decnet+ Question  Re: Newbie Decnet+ Question  NFS serving and yp Re: openhack challenge and VMS  Re: OpenVMS Patches Mailing List RE: OpenVMSware  Perl on VMS  Re: Perl on VMS  Re: Perl on VMS 
 Posix support  Re: Posix support . Re: Problem setting up YAHMAIL with OSU server= Re: Programmatic access to byte count rx/tx on TCP/IP socket. = Re: Programmatic access to byte count rx/tx on TCP/IP socket. = Re: Programmatic access to byte count rx/tx on TCP/IP socket. = Re: Programmatic access to byte count rx/tx on TCP/IP socket.  Re: RD51 vs. RD53  Re: Reconfigure Network Cards  Re: Samba (was PC-NFS ?) Re: Samba (was PC-NFS ?) socket paths???  Re: socket paths??? $ Re: TCPIP - upgrade from 4.2 to 5.0a Re: UCX BUFFERS ? " VAXStation 4000 Model 60 internals& Re: VAXStation 4000 Model 60 internals VMS Mail problem Re: VMS Vs any other OS  Re: VMS Vs any other OS  Re: VMS Vs any other OS  Re: VMS Vs any other OS  Re: VMS Vs any other OS  Re: VMS Vs any other OS 
 Re: X windows   F ----------------------------------------------------------------------  % Date: Wed, 16 Aug 2000 18:27:00 -0400 + From: Tim Shoppa <shoppa@trailing-edge.com> 7 Subject: Re: /float qualifier and resulting performance 1 Message-ID: <399ADCF4.6ABEAA3F@trailing-edge.com>    Steve Lionel wrote:  > 1 > On Tue, 15 Aug 2000 15:41:14 +0200, Jouk Jansen $ > <joukj@hrem.stm.tudelft.nl> wrote: > F > >Is there any difference in performance of the resulting applicationH > >when compiled either with /float=ieee or /float=g_float, when runningI > >on a VMS7.2-1 system? Especially programs wriiten in F90 and C have my  > >interest. >  > No difference using Fortran.  : I'm pretty sure the original question (and Steve's answer); implied that the hardware platform was an Alpha.  So let me  ask a different question:   ; On modern VAXes, will /G_FLOAT give the same performance as  /NOG_FLOAT ?  For older VAXes?  B I seem to recall, at some point in the past, encountering a 11/780= that didn't have G_FLOAT hardware present, and it ran my code : (which used G_FLOAT) about a hundred times slower than the& 11/780 that did have G_FLOAT hardware.   Tim.   ------------------------------   Date: 16 Aug 2000 16:06:23 PDTT From: Fairfield@SLAC.Stanford.EDU (Ken Fairfield; SLAC: 650-926-2924; FAX: 926-3515)7 Subject: Re: /float qualifier and resulting performance 3 Message-ID: <jjU7g65aY1QA@mccdev.slac.stanford.edu>   2 In article <399ADCF4.6ABEAA3F@trailing-edge.com>, 2     	Tim Shoppa <shoppa@trailing-edge.com> writes: [...] = > On modern VAXes, will /G_FLOAT give the same performance as   > /NOG_FLOAT ?  For older VAXes? > D > I seem to recall, at some point in the past, encountering a 11/780? > that didn't have G_FLOAT hardware present, and it ran my code < > (which used G_FLOAT) about a hundred times slower than the( > 11/780 that did have G_FLOAT hardware.  H         Yes,  that's  my  recollection  as  well.   Unless  you  had theH     specific  (hardware) "option" for G-float and/or H-float  installed,H     arithmetic on those datatypes  was  done  in  software  rather  thanH     hardware.   [As Tim well knows, D-float is/was the VAX native double(     precision floating point data type.]  H         Would you consider a VAXstation  4000/60  a "modern VAX"?  I canH     try  putting together a test program and see what I get.  Oh, I alsoH     have VAX 66x0's I can test on...those are pretty modern (for variousH     values of "modern :-).   But  I  thought  one  of  the  things  thatH     happened  over  the  evolution  of  the  VAX was to _remove_ variousH     instructions   from    the    hardware    and    emulate   them   inH     software/firmware.    The   various  MACRO   character   comparison/H     manipulation instructions come to mind,  but  I'm  sure  there  were     others.            -Ken --  M  Kenneth H. Fairfield            |  Internet: Fairfield@SLC.Slac.Stanford.Edu :  SLAC, 2575 Sand Hill Rd, MS 46  |  Voice:    650-926-2924:  Menlo Park, CA  94025           |  FAX:      650-926-3515N  -----------------------------------------------------------------------------B  These opinions are mine, not SLAC's, Stanford's, nor the DOE's...   ------------------------------  # Date: Wed, 16 Aug 2000 22:03:28 GMT ( From: hx_101@hotmail.com (Horse Nuts...)! Subject: Adding new mail accounts 4 Message-ID: <399b0f9b.11096145@news.accesscable.net>  2 What is the procedure to add new mail accounts????   Thanks   ------------------------------  % Date: Wed, 16 Aug 2000 16:02:47 -0500 / From: "Stuart, Ed" <Ed.Stuart@austinenergy.com> / Subject: RE: Adjusting Width of Show Users Info T Message-ID: <CB874B506A79D1118FBC006097306B8906828D0C@ohms.electric.ci.austin.tx.us>  K We would not be paying it so much attention if the display was consistently D incorrect across all the nodes in our cluster.  However, we have twoJ machines with identical configurations and on one box the port informationK is complete and on the other box it is missing characters at the beginning. G Both machines are running OpenVMS V6.2, with TCPware V5.2-3.  The other H interesting twist is that the amount of information displayed is machineK specific.  So the information from all the sessions on a particular node is L displayed completely no matter what other node in the cluster is viewing the information.  See below.  K In our configuration node Flash is functioning correctly. A Show Users/Full G command displays a completely qualified machine name for our NTA ports.   + ES_FLASH$ show users/full/node=flash stuart 7       OpenVMS User Processes at 16-AUG-2000 15:50:55.26 7     Total number of users = 1,  number of processes = 1   2  Username  Node   Process Name    PID     Terminal3  STUART    FLASH  Ed            34E52F19  NTA27:       $ (ae-stuart.electric.ci.austin.tx.us)  G Node Tesla is not functioning correctly and the same command results in  missing characters.   + ES_TESLA$ show users/full/node=tesla stuart 7       OpenVMS User Processes at 16-AUG-2000 15:52:29.72 7     Total number of users = 1,  number of processes = 1   2  Username  Node   Process Name    PID     TerminalC  STUART    TESLA  Ed            3C0079A4  NTA10:   (i.austin.tx.us)   J However, if we inquire about the sessions on Flash from Tesla the complete information is returned.    ES_TESLA$ show users/full stuart7       OpenVMS User Processes at 16-AUG-2000 15:53:42.68 7     Total number of users = 1,  number of processes = 2   2  Username  Node   Process Name    PID     Terminal3  STUART    FLASH  Ed            34E52F19  NTA27:       $ (ae-stuart.electric.ci.austin.tx.us)C  STUART    TESLA  Ed            3C0079A4  NTA10:   (i.austin.tx.us)   L So the question is, how do we correct the other nodes in the cluster so thatG the compete terminal information is displayed?  I realize this could be K trivial, but the correct information makes it easier to locate clients, and H keeps our DCL consistent.  For example on the correct node (Flash) if weE need the TT_ACCPORNAM a DCL command returns the following information   . ES_FLASH$ port = f$getdvi("tt","tt_accpornam") ES_FLASH$ show sym port -   PORT = "ae-stuart.electric.ci.austin.tx.us"   I However, on an incorrect node (Tesla) the same DCL returns the following.   / ES_TESLA$  port = f$getdvi("tt","tt_accpornam")  ES_TESLA$ sh sym port    PORT = "i.austin.tx.us"   $ Ed Stuart                           ( Manager, Systems and Desktop Services	   Information Technology Services  City of Austin, Austin Energy  Ed.Stuart@austinenergy.com  + "Glittering prizes and endless compromises  . shatter the illusion of integrity" - Neil Pert  B *Please apply a generous amount of all the usual disclaimers here*   ------------------------------  % Date: Wed, 16 Aug 2000 16:35:54 -0500 7 From: "David J. Dachtera" <djesys.nospam@earthlink.net> / Subject: Re: Adjusting Width of Show Users Info - Message-ID: <399B093A.6DA718D9@earthlink.net>    "Stuart, Ed" wrote:  > M > We would not be paying it so much attention if the display was consistently F > incorrect across all the nodes in our cluster.  However, we have twoL > machines with identical configurations and on one box the port informationM > is complete and on the other box it is missing characters at the beginning. I > Both machines are running OpenVMS V6.2, with TCPware V5.2-3.  The other J > interesting twist is that the amount of information displayed is machineM > specific.  So the information from all the sessions on a particular node is N > displayed completely no matter what other node in the cluster is viewing the > information.  See below. > M > In our configuration node Flash is functioning correctly. A Show Users/Full I > command displays a completely qualified machine name for our NTA ports.  > - > ES_FLASH$ show users/full/node=flash stuart 9 >       OpenVMS User Processes at 16-AUG-2000 15:50:55.26 9 >     Total number of users = 1,  number of processes = 1  > 4 >  Username  Node   Process Name    PID     Terminal2 >  STUART    FLASH  Ed            34E52F19  NTA27: > & > (ae-stuart.electric.ci.austin.tx.us) > I > Node Tesla is not functioning correctly and the same command results in  > missing characters.  > - > ES_TESLA$ show users/full/node=tesla stuart 9 >       OpenVMS User Processes at 16-AUG-2000 15:52:29.72 9 >     Total number of users = 1,  number of processes = 1  > 4 >  Username  Node   Process Name    PID     TerminalE >  STUART    TESLA  Ed            3C0079A4  NTA10:   (i.austin.tx.us)  > L > However, if we inquire about the sessions on Flash from Tesla the complete > information is returned. > " > ES_TESLA$ show users/full stuart9 >       OpenVMS User Processes at 16-AUG-2000 15:53:42.68 9 >     Total number of users = 1,  number of processes = 2  > 4 >  Username  Node   Process Name    PID     Terminal2 >  STUART    FLASH  Ed            34E52F19  NTA27: > & > (ae-stuart.electric.ci.austin.tx.us)E >  STUART    TESLA  Ed            3C0079A4  NTA10:   (i.austin.tx.us)  > N > So the question is, how do we correct the other nodes in the cluster so thatI > the compete terminal information is displayed?  I realize this could be M > trivial, but the correct information makes it easier to locate clients, and J > keeps our DCL consistent.  For example on the correct node (Flash) if weG > need the TT_ACCPORNAM a DCL command returns the following information  > 0 > ES_FLASH$ port = f$getdvi("tt","tt_accpornam") > ES_FLASH$ show sym port / >   PORT = "ae-stuart.electric.ci.austin.tx.us"  > K > However, on an incorrect node (Tesla) the same DCL returns the following.  > 1 > ES_TESLA$  port = f$getdvi("tt","tt_accpornam")  > ES_TESLA$ sh sym port  >   PORT = "i.austin.tx.us"    Hhmmm...  H You may want to take this up with the TCPware people at Process Software= then. I believe it's the guts of TCPware that's providing the % information returned as TT_ACCPORNAM.    That'd be my recommendation.   --   David J. Dachtera  dba DJE Systems  http://www.djesys.com/  : Unofficial Affordable OpenVMS Home Page and Message Board: http://www.djesys.com/vms/soho/    ------------------------------   Date: 16 Aug 2000 23:09:49 GMT- From: "Phil Tregoning" <ptregoni@esoc.esa.de> " Subject: Re: Caching RMS Block I/O= Message-ID: <01c007d5$604d96e0$4b53b083@ptregoni.esoc.esa.de>    [Reordered post]  5 Rob Young <young_r@eisner.decus.org> wrote in article " <qzUI72SR+ybC@eisner.decus.org>... > 6 > In article <8nbrmc$5r3$1@pyrite.mv.net>, "Bill Todd" <billtodd@foo.mv.com> writes: G > > If write-caching is indeed acceptable on the output side (though it  shouldG > > be battery-backed cache or cache protected by a reliable UPS, since  ODS-2 E > > depends upon the ordering of disk writes for the integrity of its 	 meta-data J > > structures and lost updates there could be catastrophic), you might be ableK > > to enable disk-level read-ahead caching (an option on both SCSI and IDE I > > drives that VMS may have a means to address) to cover the input side.   C Write caching may be acceptable (with battery backup) on the output D side if nothing better can be found. I can probably arrange for this? to write to a separate disk with nothing else on it. In case of = failure I can restore from backup and redo the last few runs.   I > > The eXtended File Cache features planned for VMS V7.3 might also help  the J > > input side (I don't know the degree to which they recognize sequentialJ > > access patterns and react to them).  IIRC XFC is either in or close to a J > > beta test (which might mean you could get your hands on it).  For thatJ > > matter, check out the existing VIOC facilities (something someone saidB > > recently suggested that they may support a form of transparent large-block 2 > > access that might help, at least for reading).  ? In my experience VIOC only caches data actually read - it won't < do read-ahead caching. When you say "transparent large-block; access" do you mean the VIOC can be made to do read-ahead ?   @ XFC does sound like it might do the trick. We have a fair amount= of time to sort this out, so perhaps could wait for V7.3, but  XFC is Alpha only according to  2 http://www.openvms.digital.com/openvms/os/sdk.html  = There is, unfortunately, no chance of porting the application  to Alpha :(   I > > You didn't specify whether the output-side files were being initially I > > populated or updated in place.  If the former, experimenting with the K > > various forms of default extension quantity control could prove useful.  > >   = That may make some difference, but the overriding performance ; problem is the small size of the I/Os with all the overhead A that involves. As an example of the kind of thing the application 9 does, take a look the example "B-10: Use of Block I/O" in ; _OpenVMS_Record_Management_Services_Reference_Manual_. With ; that example, copying a 7000 block file with 2 block record ? buffers takes me about 30 seconds, with 20 block record buffers  it takes about 4 seconds.   / > 	To add to what Bill is saying (getting in on : > 	this late), you may also want to use HSG80 controllers. > : > 	Why HSG80 controllers?  Because they support sequential= > 	read-ahead caching.  Couple that with write behind caching  > 	and you can't hardly lose.   > I guess any read-ahead controller would fix the problem, but I? can't see us getting a fibre channel controller when a SCSI one > would do. The reason I mentioned the HSZ50 is that we have one> laying around that we could use. But it doesn't do read-ahead, only write-behind.   [SNIP]  = > 	Bit over the top above.. but the point is Volume Shadowing ; > 	gives you a measure of reliability and by being a mirror ; > 	copy, improves read performance.  How backed up are your @ > 	disks you are reading and writing to?  $ MONI DISK/ITEM=QUEUE  : The average queue length is about .5. The application does; a synchronous read on one disk, then a synchronous write onn= the other. The shadow read performance improvement is because ; the average rotational latency drops from half a revolutiona= to a third (the other way round for writes). I am looking forp8 ways to remove the latency (and seek time) completely by somehow getting I/Os combined.  ; One or two people also mentioned RMS buffering, but that isl: record based and doesn't work for block I/O. These are not7 RMS files (for example, the files can't be copied, onlye9 backed-up as a copy fails with "RMS-F-IRC, illegal recordE encountered").  $ The options at the moment look like:  3 1) Get a write-behind / read-ahead disk controller:o      o HSZ70 or HSZ80e       L http://www.compaq.com/products/storageworks/Array-and-SCSI-Controllers/hsz80
 index.html  
    o HSG80  4      http://www.digital.com/info/SP7047/SP7047HM.HTM  B 2) Get a third party caching product. Three suggestions were made,A    but none of the web sites explicitly mention read-ahead cache. B    I guess none of them will offer write-behind, as that is fairlyB    dangerous without some kind of hardware battery backup. So this?    option still requires a disk controller to cache the writes, 0    but we already have a HSZ50 that can do that.  <    Caching software will also (hopefully) keep the Interrupt9    processing down by issuing fewer I/Os. Hardware caches ;    won't do that as the same number of I/Os get issued, butb)    they do get satisfied from the cache. e      o RAXCO's PerfectCachee  A      http://www.raxco.com/products/performancesuite/more_info.cfmp  &    o R&D Performance Group's Disk Xcel  ?      According to Marty Kuhrt it may be possible to get this to       do read-ahead.t  1      http://www.rdperf.com/RDDOC/DXL_SPD_V50.HTMLx  %    o Executive Software's I/O Expressa  ?      http://www.execsoft-europe.com/products/openvms/io4vms.htms  F I'll get in contact with these companies for some further information.   Phil T   ------------------------------  % Date: Wed, 16 Aug 2000 17:08:02 -0400e- From: JF Mezei <jfmezei.spamnot@videotron.ca>)0 Subject: Cluster transaction processing paradigm, Message-ID: <399B02AE.DF337EA6@videotron.ca>  K I am trying to think of an application architecture which would not only be M cluster friendly (easy) but also make maximum use of clustering to distributes= processing and provide transparent operation if a node fails.O  9 Transactions go into an indexed file in a random fashion.-  L Transactions are extracted from the file in an ordered fashion and sent to aM third party (by modem) in batches of say 30 transactions at a time (or less)./M There are various third parties involved, so the reception of 30 transactionslN might mean one phone call if they are all destined to the same destination, or multiple phone calls.j  ' FEEDING TRANSACTIONS TO THE QUEUE-FILE:r  @ To feed the transactions into the queue file, I see 3 paradigms: 	oJ -client application itself opens the queue file and writes to it. However,G this still requires that the client contact some server to get a uniques transaction number.c  N -client application uses ICC to contact a single server process on the clusterJ which provides the transaction number and it stores the transaction in theU queue file. However, if the node on which that server runs fails, then you're cooked.?  I -client application submits to any number of servers on the cluster. EachnL server then provides a unique transaction number (with its node name as part+ of it) and writes to the single queue file.e  E Question: what technologies exist right now that would allow multiple-N processes (one per node) to advertise themselves as being available to receive: transactions from any process on any node of the cluster ?     SERVER COORDINATIONf  N If I have multiple server processes on multiple nodes on the cluster, is thereG some technology which allows these to talk to each other in a broadcast I fashion ? (eg: server 1 send a single message that is received by all theTH servers). And if a new server joins, then it too starts to receive thoseI "broadcast" messages. (and ideally, when one leaves, all others receive ad message of it leaving).       $ TRANSACTION CONSUMPTION/LOAD SHARING  F Ideally, I'd like to have the multiple servers intelligently share theI transactions. While one server has "reserved" some 30 transactions to onesK remote partner, the other server reserves the next 30 transactions (to some E other or perhaps same partner) and so on. This way, the throughput is O maximised, and if one node fails, then the servers on the other nodes continue.o  J However, how can one distribute transactions amongst those servers withoutH having a single server coordinate this (which represents single point of failure if that node fail?  L For instance, during slow period, if a single transaction is received, whichI server will get to pick that one up and deliver it as fast as possible ? n  N Are there any technologies which do this type of stuff (or papers that discuss this) ? ? ?t   ------------------------------  % Date: Wed, 16 Aug 2000 16:32:39 -0700f From: James.F.Duff@fhs.com4 Subject: Re: Cluster transaction processing paradigmC Message-ID: <OFEF353513.C23E8789-ON8825693D.007F1E53@HEALTHNET.COM>    JF Mezei writes: >[snip]  >rF >Question: what technologies exist right now that would allow multipleG >processes (one per node) to advertise themselves as being available to5 receive ; >transactions from any process on any node of the cluster ?-  D Reliable Transaction Router, a layered product available from the Q.K IIRC, you can set it up so multiple nodes in a network (not just a cluster) G can handle transactions.  Really good for mirrored transaction systems.s    I >If I have multiple server processes on multiple nodes on the cluster, isc therelH >some technology which allows these to talk to each other in a broadcastJ >fashion ? (eg: server 1 send a single message that is received by all theI >servers). And if a new server joins, then it too starts to receive thosehJ >"broadcast" messages. (and ideally, when one leaves, all others receive a >message of it leaving).   Same answer.  G >Ideally, I'd like to have the multiple servers intelligently share therJ >transactions. While one server has "reserved" some 30 transactions to oneG >remote partner, the other server reserves the next 30 transactions (tot someF >other or perhaps same partner) and so on. This way, the throughput isF >maximised, and if one node fails, then the servers on the other nodes	 continue.s >aK >However, how can one distribute transactions amongst those servers withoutoI >having a single server coordinate this (which represents single point ofa >failure if that node fail?l >eG >For instance, during slow period, if a single transaction is received,l whichfI >server will get to pick that one up and deliver it as fast as possible ?l >eG >Are there any technologies which do this type of stuff (or papers thath discuss  >this) ? ? ?  G Take a look at http://www.software.digital.com/rtr/htm_docs/grs-wp2.htm   J This paper describes the way the Australian Stock Exchange did distributedH transaction handling.  Having worked there, I know it works particularly well.i  E The only drawback with RTR is it is quite an obscure product.  When ImE was working at the ASX for example, the only support for RTR was fromeI a European arm of Digital (can't remember off hand which country it was)!    Regards, Jim. -- James.F.Duff@healthnet.com Pure personal opinion    ------------------------------  % Date: Wed, 16 Aug 2000 18:02:34 -07004! From: Shane.F.Smith@Healthnet.come4 Subject: Re: Cluster transaction processing paradigmC Message-ID: <OF47729E91.A3405FEF-ON8825693E.0004D862@HEALTHNET.COM>l  I If RTR doesn't look like a good match, you might like to take a look at a K company called TIBCO. Their core product is a multi-platform (IncludingVMS)a? UDP-based broadcast mechanism, which is mainly designed for the J one-broadcaster, many receivers model. They can do transaction processing,J and have add-ons for various queueing and fault tolerance methods that mayG be of interest. I'm not recommending it, I haven't used it enough to doiC that, but I'll happily share what I know of it offline if you like.c   Shanef          . James.F.Duff@fhs.com on 08/16/2000 04:32:39 PM   To:   Info-VAX@Mvb.Saic.Como cc:   5 Subject:  Re: Cluster transaction processing paradigmo           JF Mezei writes: >[snip]o > F >Question: what technologies exist right now that would allow multipleG >processes (one per node) to advertise themselves as being available tor receive ; >transactions from any process on any node of the cluster ?r  D Reliable Transaction Router, a layered product available from the Q.K IIRC, you can set it up so multiple nodes in a network (not just a cluster)DG can handle transactions.  Really good for mirrored transaction systems.t    I >If I have multiple server processes on multiple nodes on the cluster, ism there H >some technology which allows these to talk to each other in a broadcastJ >fashion ? (eg: server 1 send a single message that is received by all theI >servers). And if a new server joins, then it too starts to receive thoseyJ >"broadcast" messages. (and ideally, when one leaves, all others receive a >message of it leaving).   Same answer.  G >Ideally, I'd like to have the multiple servers intelligently share thetJ >transactions. While one server has "reserved" some 30 transactions to oneG >remote partner, the other server reserves the next 30 transactions (tos someF >other or perhaps same partner) and so on. This way, the throughput isF >maximised, and if one node fails, then the servers on the other nodes	 continue.i >TK >However, how can one distribute transactions amongst those servers withouttI >having a single server coordinate this (which represents single point ofi >failure if that node fail?  >iG >For instance, during slow period, if a single transaction is received,e whichoI >server will get to pick that one up and deliver it as fast as possible ?c >eG >Are there any technologies which do this type of stuff (or papers thatn discussr >this) ? ? ?  G Take a look at http://www.software.digital.com/rtr/htm_docs/grs-wp2.htmr  J This paper describes the way the Australian Stock Exchange did distributedH transaction handling.  Having worked there, I know it works particularly well.   E The only drawback with RTR is it is quite an obscure product.  When IiE was working at the ASX for example, the only support for RTR was fromhI a European arm of Digital (can't remember off hand which country it was)!e   Regards, Jim. -- James.F.Duff@healthnet.com Pure personal opinion    ------------------------------  % Date: Wed, 16 Aug 2000 19:54:02 -0500 ) From: "John E. Malmberg" <wb8tyw@qsl.net> ' Subject: Re: Copy from VMS to NT server . Message-ID: <spmdqf4hr5j74@corp.supernews.com>  ) <teroconnor@my-deja.com> wrote in messageF$  news:8ne632$nj5$1@nnrp1.deja.com... >o9 > > > We recently moved some fileshares from a VMS server 7 > > > (with Pathworks) to an NT box. We have some batch ; > > > jobs which copied files to VMS directories for use by 2 > > > clients mapping onto Pathworks shares. Now I5 > > > need to change these copy procedures on the VMS ; > > > system to direct the output files to the NT box. What : > > > is the best way to do this - is FTP the only option?8 > > > The NT server only runs TCPIP. Any examples of the > > > syntax to use? > >p  > > Recent VMS versions support: > > D > > $ COPY/NODE file ip-node"username password"::"path-and-filename" > >a > > example: > >nB > > $ COPY/FTP FOOBAR.DAT NTSRV"joe vmsuser":"C:\Store\Foobar.dat" > >l > > Arne > >e > Thanks Arne,& > I have tried this but get an error - >o5 > copy /ftp t.t  oilcom"userxxx pwxxx"::"c:\t.t" /logo, > %FTP-E-NETERR, I/O error on network device6 > -SYSTEM-F-REJECT, connect to network object rejected >n0 > Am I missing something on the NT server side ?  L There are several versions of FTP for Windows NT 4.0, all of them optionallyJ installed components.  There are also three management interfaces to them.  K I would recommend checking the ftp logs or the event logs on the Windows NTe. Server to see why the login is being rejected.  I The account that you use to ftp into the NT server must have the right to 6 log into it locally.  That is usually not the default.  H The default configuration for at least one of the FTP servers is to only allow anonymous logins.-  L And finally while you may realize that the password in NT is case sensitive,K in many cases you must qualify the username with the domain or server who's 5 SAM database it is in such as "DOMAIN\userxxx pwxxx".R   -Johnp wb8tyw@qsl.network   ------------------------------  % Date: Thu, 17 Aug 2000 11:31:01 +1000./ From: "Phil Howell" <howellp@snowyhydro.com.au>,' Subject: Re: Copy from VMS to NT servers. Message-ID: <2eHm5.234$Vb.5554@ozemail.com.au>   we currently use9 (however the syntax changes slightly on each ucx release)e* $       assign/nolog  'ftp_file' move_file@ $       ftp <servername> /username="xxxxxxx" /password="*******"
         ascii          cd <targetdir>         pwdo         case         put move_fileb         byea  7 You have to check that the ftp service is running on NTn  and that user permissions are ok4 (this is configured in IIS if you are running that)\ Phil) <teroconnor@my-deja.com> wrote in messagep# news:8ndpka$9nl$1@nnrp1.deja.com...tI > We recently moved some fileshares from a VMS server (with Pathworks) tos> > an NT box. We have some batch jobs which copied files to VMSE > directories for use by clients mapping onto Pathworks shares. Now IeF > need to change these copy procedures on the VMS system to direct theF > output files to the NT box. What is the best way to do this - is FTPE > the only option? The NT server only runs TCPIP. Any examples of thea > syntax to use?	 > Thanks,o > Ter  >r >n( > Sent via Deja.com http://www.deja.com/ > Before you buy.    ------------------------------  % Date: Wed, 16 Aug 2000 20:50:35 -0400p2 From: "Richard B. Gilbert" <DRAGON@compuserve.com>- Subject: Re: DECnet Plus copying time problemm6 Message-ID: <200008162050_MC2-AFEA-FEE@compuserve.com>  D         I've met these "gurus".  None yet has been able to give me aG reasonable explanation of *why*,  SQE should be off.  I just turn it ondJ anyway, or used to; damned near everything is 10-base-T or 100-base-T now=  " with few transceivers in evidence!  ' Message text written by "Brian Tillman"w, >>Try the following command on both systems: >rE >    $ mcr ncl show csma-cd station * Collision Detect Check Failures  > @ >It should be 0 for each interface.  If it's a large number, and increasing, J >your ethernet network is not configured correctly.  In my experience it'= si> >because SQE has not been enabled on the ethernet transceiver.  > Our network gurus insist that it be OFF on every transceiver.<   ------------------------------  % Date: Wed, 16 Aug 2000 18:03:30 +0000h- From: SysAdmin <djesys.nospam@fsi.net.mapson> & Subject: Re: Detached process question. Message-ID: <399AD772.164E2A6E@fsi.net.mapson>   Shraga Broyer wrote: >  > Hi._H > I am trying to run a DCL script as a detached process using RUN/DETACH > SYS$SYSTEM:LOGINOUT.EXE . E > I noticed that I can only specify GROUP or SYSTEM-wide logical name A > for the paths of the INPUT and OUTPUT files of the process (form> > example JOB-wide assignments such as  /INPUT=SYS$LOGIN:<someC > file>/OUTPUT=SYS$LOGIN:<some file>   won't work). I don't want toeH > create any new system or group-wide logical names. Can someone suggest > me a work around?r  D Well, what I did just now is this: I made a .COM file containing the following commands:S  
 SHLOG.COM:   $ SHOW LOGICAL $ EXIT    ...and ran it detached, like so:  < $ run/detach/inp=shlog.com/out=shlog.log sys$system:loginout  E When you look at the output, it may explain a lot. Then again, it mayV raise more questions.v  F If you need to run a long process, might you try SPAWN/NOWAIT instead?  H If you need this to survive you logging out, might you try running it in batch?  D If neither of those is suitable, then you'll need to make up for theH lack of the usual process and job level logical names in some other way.  E It's not that there's anything broken, it's just the way things work.v   You might try it this way:   $ RUN/DETACH -) 	/INPUT='F$TRNLNM("SYS$LOGIN")'some_file- + 	/OUTPUT='F$TRNLNM("SYS$LOGIN")'some_file -e 	SYS$SYSTEM:LOGINOUT.EXE  @ ...so the logicals get translated before the command is actually	 executed.d   David J. Dachterah   ------------------------------   Date: 16 Aug 2000 18:49:57 GMT2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)& Subject: Re: Detached process question6 Message-ID: <8nenol$b7q$3@mailint03.im.hou.compaq.com>  b In article <399ad08f.198595074@news.bmc.com>, Shraga_Broyer@bmc.nospam.com (Shraga Broyer) writes:G :I am trying to run a DCL script as a detached process using RUN/DETACHn :SYS$SYSTEM:LOGINOUT.EXE .  D :I noticed that I can only specify GROUP or SYSTEM-wide logical name@ :for the paths of the INPUT and OUTPUT files of the process (for= :example JOB-wide assignments such as  /INPUT=SYS$LOGIN:<somecB :file>/OUTPUT=SYS$LOGIN:<some file>   won't work). I don't want toG :create any new system or group-wide logical names. Can someone suggest  :me a work around?  J   Translate the specifications to physical references before passing them?  J   Only the input and output specifications are a factor here, as these are   needed early...o  J   If you want to use logical names, then the definitions must reside in a K   logical name table that is visible to the created process.  (Even if you oK   tried to get sneaky and define the logical names in the job table of the /J   created process, there's no reliable way to get LOGINOUT only as far as J   creating the process job table without also having it try to access the &   specified input and output devices.)  L   Also note that LOGINOUT only creates the SYS$LOGIN, SYS$LOGIN_DEVICE, and M   SYS$SCRATCH logical names if it is processing a login that is based on the iK   contents of a SYSUAF record -- without access to the SYSUAF record, this  +   information is not available to LOGINOUT.m  K   If you want to see these particular logical names created, please specifys7   the /AUTHORIZE qualifier on the RUN/DETACHED command.>  N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  # Date: Wed, 16 Aug 2000 18:58:04 GMTa= From: system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-)t& Subject: Re: Detached process question0 Message-ID: <009EEB1A.D807EAA2@SendSpamHere.ORG>  b In article <399ad08f.198595074@news.bmc.com>, Shraga_Broyer@bmc.nospam.com (Shraga Broyer) writes: >t >Hi.G >I am trying to run a DCL script as a detached process using RUN/DETACHt >SYS$SYSTEM:LOGINOUT.EXE .  D >I noticed that I can only specify GROUP or SYSTEM-wide logical name@ >for the paths of the INPUT and OUTPUT files of the process (for= >example JOB-wide assignments such as  /INPUT=SYS$LOGIN:<sometB >file>/OUTPUT=SYS$LOGIN:<some file>   won't work). I don't want toG >create any new system or group-wide logical names. Can someone suggests >me a work around? >e >TIA    F If you use RUN/DETACHED you're creating a new job and/or process.  TheF process and job tables of the process issuing the RUN command will notE be accessible by the process/job created.  You really don't have many G options if you want to refer to files, devices, etc. by a logical name,i@ you have to use a "shared" logical name table (group or system).   --O VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COMd   ------------------------------  % Date: Wed, 16 Aug 2000 14:56:36 -0400 ' From: David Beatty <dabeat@wnt.sas.com>n& Subject: Re: Detached process question2 Message-ID: <P+CaOVcqQeWG=R1VwP2sz6srH1oh@4ax.com>  B I've done this kind of thing before just fine.  Are you specifyingA a UIC other than the one you are initiating the RUN command from?   That might be part of the issue.  C If you mean the SYS$LOGIN directory for the account you want to runtC under (as opposed to the initiating account), you should be able toy% leave off SYS$LOGIN: on the filespec.d   You could also try  * /Input='F$TrnLnm("Sys$Login")'<input file>- /Output='F$TrnLnm("Sys$Login")'<output file>.r  8 Finally, there might be a file access/privilege problem.  & Try one of the above.  David R. Beatty  F On Wed, 16 Aug 2000 17:36:24 GMT, Shraga_Broyer@bmc.nospam.com (Shraga Broyer) wrote:   >a >Hi.G >I am trying to run a DCL script as a detached process using RUN/DETACH  >SYS$SYSTEM:LOGINOUT.EXE . oD >I noticed that I can only specify GROUP or SYSTEM-wide logical name@ >for the paths of the INPUT and OUTPUT files of the process (for= >example JOB-wide assignments such as  /INPUT=SYS$LOGIN:<someoB >file>/OUTPUT=SYS$LOGIN:<some file>   won't work). I don't want toG >create any new system or group-wide logical names. Can someone suggesta >me a work around? >  >TIA h >    ------------------------------    Date: 16 Aug 2000 19:07:33 -05009 From: Kilgallen@eisner.decus.org.nospam (Larry Kilgallen),& Subject: Re: Detached process question+ Message-ID: <8kwEDWtjsrj0@eisner.decus.org>c  b In article <399ad08f.198595074@news.bmc.com>, Shraga_Broyer@bmc.nospam.com (Shraga Broyer) writes:  H > I am trying to run a DCL script as a detached process using RUN/DETACH > SYS$SYSTEM:LOGINOUT.EXE . E > I noticed that I can only specify GROUP or SYSTEM-wide logical nameeA > for the paths of the INPUT and OUTPUT files of the process (for > > example JOB-wide assignments such as  /INPUT=SYS$LOGIN:<someC > file>/OUTPUT=SYS$LOGIN:<some file>   won't work). I don't want touH > create any new system or group-wide logical names. Can someone suggest > me a work around?I  9 	$ RUN/DETACH/INPUT='f$trnlnm("MY_JOB_LOGICAL_NAME")' ...l   ------------------------------  % Date: Wed, 16 Aug 2000 21:29:26 -0500e7 From: "David J. Dachtera" <djesys.nospam@earthlink.net>m3 Subject: DJE Systems DECUS Presentations on the Web - Message-ID: <399B4E06.9623AC1D@earthlink.net>v   Folks,  D The word came the other day that I won't be speaking at CETS2000. OfG course, that also means no Affordable BOF, unless someone else heads ite
 up. *SIGH*  E So, I've put my San Diego presentations on the web. You can find themr via a link at this URL:i   http://www.djesys.com/vms/#tipsh   The presentations are:  ! o Introduction to DCL Programmingo   o Intermediate DCL Programming  / o Zip and Unzip for OpenVMS - Secrets Revealed!e  H I cleaned up some of the errors we found during the presentations at SanH Diego. Please help me find the rest. Your notes and (polite) correctionsG are welcome and invited. The presentations are available as HTML, .PDF,i .PPT and .PS (uncompressed).  F I may still make an OpenVMS Freeware CD-ROM available for sale (cheap,G circa. $5 US) on the web. I've found an e-store site that I can use andsD I'm currently preparing The DJEstore for a debut sometime later this season.   G Stay tuned. If I can find providers for the VMS trinkets that have beenlB requested in another thread, and/or if I can get permission to useH Compaq trademark symbols and such (OpenVMS "licensed" merchandise), I'll+ consider adding them to the store, as well.l   -- h David J. Dachterah dba DJE Systemst http://www.djesys.com/  : Unofficial Affordable OpenVMS Home Page and Message Board: http://www.djesys.com/vms/soho/o  H This *IS* comp.os.vms. So, a certain bias in postings is to be expected.  @ Feel free to exercise your rights of free speech and expression.  F However, attacks against individual posters, or groups of posters, are strongly discouraged.o   ------------------------------  # Date: Thu, 17 Aug 2000 03:59:24 GMT42 From: "Zane H. Healy" <healyzh@shell1.aracnet.com>7 Subject: Re: DJE Systems DECUS Presentations on the Web 4 Message-ID: <wqJm5.91$M62.34299@typhoon.aracnet.com>  E In comp.os.vms David J. Dachtera <djesys.nospam@earthlink.net> wrote:sI > Stay tuned. If I can find providers for the VMS trinkets that have been.D > requested in another thread, and/or if I can get permission to useJ > Compaq trademark symbols and such (OpenVMS "licensed" merchandise), I'll- > consider adding them to the store, as well.   K Hey...  That'd be cool!  Especially baseball caps and t-shirts :^)  Now you . just need to convince Compaq to let you do it!   			Zane    ------------------------------  % Date: Wed, 16 Aug 2000 11:59:47 -0400g* From: David A Froble <davef@tsoft-inc.com>" Subject: Re: dose vms run on intel- Message-ID: <399ABA73.74115996@tsoft-inc.com>p  
 Jeremy wrote:  > C > dose vms run on intel or is it still an alpha vax thing and if itiJ > dosen't run on intel where can a person get used inexpensive and working: > hardware? This Gnome thing makes me want to change os's. > 
 > ThankYou >  > Jeremy > ( > Sent via Deja.com http://www.deja.com/ > Before you buy.l   EBAY - and be careful.   Dave   -- t4 David Froble                       Tel: 724-529-04504 Dave Froble Enterprises, Inc.      Fax: 724-529-0596> DFE Ultralights, Inc.              E-Mail: davef@tsoft-inc.com6 T-Soft, Inc.  170 Grimplin Road  Vanderbilt, PA  15486   ------------------------------  # Date: Thu, 17 Aug 2000 01:11:12 GMTe2 From: "Zane H. Healy" <healyzh@shell1.aracnet.com>" Subject: Re: dose vms run on intel4 Message-ID: <QYGm5.86$M62.31597@typhoon.aracnet.com>  3 Hoff Hoffman <hoffman@xdelta.zko.dec.nospam> wrote:hN >   I'd avoid systems with a Q-bus, UNIBUS, BI or similar, at least until you N >   are rather more familiar with OpenVMS and with VAX hardware, and I'd also   J Most definitly, while these are fun systems what Hoff doesn't mention hereK is that getting disks for them, especially the Q-Bus systems can be a royalnL pain.  Most of the older Q-Bus systems that are available use MFM HD's whichJ by now are very unreliable.  You don't want to know what a SCSI controllerK for one of these systems goes for.  If they aren't using MFM disks they are,$ still probably not using SCSI disks.  L If you don't like playing with hardware you don't want any of these systems!  K >   avoid VAX systems with 7xx, 6xxx, 7xxx, 8xxx, 9xxx or 10xxx series for aH >   home use due to the age or the physical size and power requirements.  I I don't know...  The VAX 7000's I've seen don't look that bad :^)  You'lltG also probably want to avoid VAX 4000's (note, I did not say VAXstation, K VAXstation 4000's are very nice).  Of course personally I'd love a VAX 4000u to replace my MicroVAX 3 :^)  N >   With the Alpha series systems, most any (used) AlphaStation series system J >   would be a good choice, and most any low-end AlphaServer.  The larger L >   AlphaServer systems (AlphaServer 2100 and larger) will likely be larger ' >   (and more expensive) than you want.u  I For a hobbyist system one thing to pay close attention to is how much theRJ RAM for the system costs.  You're likely to find that you want to add moreK RAM to the system after you've used it for a while.  For what it would havejH taken to add 128MB of RAM to my AlphaStation 500/333 I was able to get aJ PWS433au (which I then had to come up with the $$$'s to add 256MB to)!  IfG you buy an Alpha I'd personally recommend getting one with 128MB of RAMtH minimum.  While it's possible to run OpenVMS on a 64MB system I've foundJ that even 96MB doesn't cut it, with 112MB being what I personally consider to be the acceptable minimum.    			Zanet   ------------------------------  % Date: Wed, 16 Aug 2000 21:13:54 -0500x) From: "John E. Malmberg" <wb8tyw@qsl.net>p" Subject: Re: dose vms run on intel/ Message-ID: <spmig8dar5j145@corp.supernews.com>a  4 There is a VAX emulator that runs on Intel hardware.   http://www.charon-vax.com/  < People have reported to have been able to run OpenVMS on it.    A "Zane H. Healy" <healyzh@shell1.aracnet.company> wrote in message/. news:QYGm5.86$M62.31597@typhoon.aracnet.com...	 >  You'llEI > also probably want to avoid VAX 4000's (note, I did not say VAXstation,gH > VAXstation 4000's are very nice).  Of course personally I'd love a VAX 4000 > to replace my MicroVAX 3 :^)  L The VAX 4000-10x series are nice, and have a SCSI bus.  Still pricey though.H If you can find a VAX 4000-50, especially with all the possible mountingE kits, that is also a nice system.  It fits in a Q-BUS chassis, but ist2 basically self contained and has a SCSI bus on it. >nK > For a hobbyist system one thing to pay close attention to is how much the L > RAM for the system costs.  You're likely to find that you want to add moreH > RAM to the system after you've used it for a while.  For what it would haveJ > taken to add 128MB of RAM to my AlphaStation 500/333 I was able to get aL > PWS433au (which I then had to come up with the $$$'s to add 256MB to)!  IfI > you buy an Alpha I'd personally recommend getting one with 128MB of RAM1J > minimum.  While it's possible to run OpenVMS on a 64MB system I've foundL > that even 96MB doesn't cut it, with 112MB being what I personally consider > to be the acceptable minimum.h  L Those RAM numbers are for ALPHA of course.  A VAX system will be quite happyK with much less RAM.  While more is better, 16MB is good for text only, 32MBs( and more will really help with graphics.   -John  wb8tyw@qsl.network   ------------------------------  # Date: Thu, 17 Aug 2000 03:56:32 GMTn2 From: "Zane H. Healy" <healyzh@shell1.aracnet.com>" Subject: Re: dose vms run on intel4 Message-ID: <QnJm5.90$M62.34299@typhoon.aracnet.com>  ( John E. Malmberg <wb8tyw@qsl.net> wrote:6 > There is a VAX emulator that runs on Intel hardware.   > http://www.charon-vax.com/  > > People have reported to have been able to run OpenVMS on it.  J It'll also run under Virtual PC on a Macintosh.  I managed to get part wayL towards getting OpenVMS V7.2 loaded in on it, when I ran out of time to messE with it.  One things for sure Charon will seriously bog down Windows!   D The big problem with it is that the hobbyist version doesn't includeL networking, so a VS4000/VLC is actually more useful!  Well a VLC is probably" also faster for the time being :^)  K Wow, I'd not looked at their web pages in a while, real improvement there. c  I OUCH, I just read up on their $490 full evaluation package.  For the low,2K low price of $490 you can get a evaluation kit that includes the networkingNK code turned on, and is limited to a 6 week period.  Since when do companiesu" have to pay to evaluate something!   GACK, I just found the prices:7 	The CHARON-VAX MicroVAX emulator for Windows ~$5900.00v2 	CHARON-VAXstation Lite for Windows	     ~$1900.00  J The Lite version has everything I want, but when you toss in the cost of aI laptop....  Since this would be for Hobbyist use, I don't see me spendingoL that kind of money.  Guess I'll have to look into what can be done with disk images and a real VMS system.t   			Zanee   ------------------------------   Date: 16 Aug 2000 23:21:04 GMT- From: "Phil Tregoning" <ptregoni@esoc.esa.de>  Subject: Re: File Headersv= Message-ID: <01c007d6$f263dac0$4b53b083@ptregoni.esoc.esa.de>0  8 SysAdmin <djesys.nospam@fsi.net.mapson> wrote in article% <399AB61F.FE3E47F7@fsi.net.mapson>...s
 > Data wrote:D > > E > > I actually meant how close to using up all the file headers on an 2 > > initialized disk ( # of files not disk space ) >    [SNIP]   > H > The volume may have been INITIALIZEd with /HEADERS=xxxxx in which caseJ > the INDEXF.SYS file will have been pre-extended forstalling the onset ofE > maximum INDEXF.SYS fragmentation (INDEXF.SYS can only have a singleGG > header). So, in some cases, you can tell by comparing the EOF size oftH > INDEXF.SYS to the ALQ size. If the EOF size is less than the ALQ size,I > INDEXF.SYS has not been extended since it was created; otherwise, you'dCH > need to DUMP the header of INDEXF.SYS to see how many extents the file > has.   FWIW  5   $ DUMP/HEADER/BLOCK=COUNT=0 DISK:[000000]INDEXF.SYSc   ...s"   Map area offset:             100"   Access control area offset:  255   ...o"   Map area words in use:        14   ...s  > When the "map area words in use" gets close to "Access control@ area offset" - "Map area offset" is when trouble is approaching.  9 Normally there are no ACEs on INDEXF.SYS, so this is whenl1 the "map area words in use" gets to be about 155.   J > To make a long story short, predicting the onset of "index file full" is > rather a dicey proposition.u >    Phil T   ------------------------------  % Date: Wed, 16 Aug 2000 20:27:54 -0400N2 From: "Richard B. Gilbert" <DRAGON@compuserve.com> Subject: Re: File HeadersS7 Message-ID: <200008162028_MC2-B005-B063@compuserve.com>e  J         As others have written, there's no really easy way.  Each file on=  J the disk requires a minimum of one header.  A badly fragmented file might=  " require a hundred or more headers.  F         VMS will, if allowed to do so, allocate a 1,000,000 block fileJ three or four blocks at a time.  This can burn up an incredible number of=  
 file headers.a  H         There are tools for coping with fragmentation.  The simplest andH cheapest is to allocate files close to their final sizes, and to specifyE reasonable extent sizes.  This takes care of the 1,000,000 block file ) allocated three or four blocks at a time.p  J         Next come the defragmentation packages.  There's a freeware packa= geJ that I haven't worked with for about ten years and I forget its name.  It=  J lacked polish but got the job done.  Perfectdisk from Raxco, I have worke= dpB with more recently.  It too gets the job done.  There's one calledD Diskeeper from Executive Software but: WARNING, their sales staff isJ extremely pushy and if you once let them have your phone number, the call= ssJ won't stop for years.  They also have peculiar ideas about honoring their=  @ contracts with people whose lines of business they disapprove ofJ (principally the makers of Ritalin).  There's Digital File Optimizer (DFO= )OE from Compaq (or maybe it's called "CFO" this week).   There are a fewaJ others.  Most of these packages have Analysis tools.  I don't recall if o= neH of them counts free file headers or not.  If not, you might write one; IJ believe there's a bitmap somewhere in the filestructure that keeps track = ofJ allocated and free file headers.  Writing the software to count the bits = is$ left as an exercise for the student.    E         If all else fails, make an image backup of your disk and thenf restore it.T   Message text written by "Data"B >I actually meant how close to using up all the file headers on an. initialized disk ( # of files not disk space )    B "David J. Dachtera" <djesys.nospam@earthlink.net> wrote in message' news:3999FBD4.53B340D8@earthlink.net...0
 > Data wrote:a > >m9 > > Is there a way to tell how close to full a disk is ??  >j > SHOW DEVICE/FULL ddcu  >s > -- > David J. Dachtera. > dba DJE Systemsr > http://www.djesys.com/ >f< > Unofficial Affordable OpenVMS Home Page and Message Board:" > http://www.djesys.com/vms/soho/<   ------------------------------  % Date: Thu, 17 Aug 2000 13:18:32 +0010E% From: paddy.o'brien@zzz.tg.nsw.gov.aua Subject: Re: File HeadersT5 Message-ID: <01JT2TI73R1U0032XQ@tgmail.tg.nsw.gov.au>U  M From the various answers here, possibly another plug for DFU might be called $ for.  O DFU's REPORT shows a lot of useful info about number of files, etc., on a disk.    Regards, Paddy   Paddy O'Brien, Transmission Development,s
 TransGrid, PO Box A1000, Sydney South,  NSW 2000, Australia'   Tel:   +61 2 9284-3063 Fax:   +61 2 9284-3050& Email: paddy.o'brien@zzz.tg.nsw.gov.au  M Either "\'" or "\s" (to escape the apostrophe) seems to work for most people,o; but that little whizz-bang apostrophe gives me little spam.s   ------------------------------  % Date: Wed, 16 Aug 2000 12:05:39 -0400g* From: David A Froble <davef@tsoft-inc.com>- Subject: Re: FTP to/from Alpha boxes - solvedE- Message-ID: <399ABBD3.C23064B5@tsoft-inc.com>    Andy Proctor wrote:0 >  > All,H > A month or so ago I posted a request for help with certain problems onN > PWS433au units (mainly) running VMS7.1-1h2 and UCX 4.2. The problem was thatM > ftp sessions could be opened and initiated from the PWS, but when trying toe9 > ftp onto the PWS, the request was rejected out of hand.aM > The problem came down to something simple in the end, (don't they all?) thetL > file protection for SYS$MANAGER:SYLOGIN.COM did not have W:RE access. OnceN > this was set, all works perfectly. Whether this would be an "approved fix" I > don't know, but it works! E > Thanks to all who have contributed to the problem, in the group andb6 > directly, especially Jesper Naur and David Dachtera.= > I hope this "fix" may help someone in the same predicament.a  P I wouldn't call it a 'fix', since this is the protection this file should alwaysN have, to allow access whenever any user logs in.  I'm guessing you never had aN non-prived user account on this system, or never used it.  It's a good idea toK always have at least one user account on any VMS system that has no specialiI privileges, a UIC outside the 'system' range, etc.  An attempt to use theeK account whenever there are apparent 'system' problems can be quite helpful.l   Dave   -- e4 David Froble                       Tel: 724-529-04504 Dave Froble Enterprises, Inc.      Fax: 724-529-0596> DFE Ultralights, Inc.              E-Mail: davef@tsoft-inc.com6 T-Soft, Inc.  170 Grimplin Road  Vanderbilt, PA  15486   ------------------------------   Date: 16 Aug 2000 18:40:22 GMT2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)* Subject: Re: Help with pagefile full error6 Message-ID: <8nen6m$b7q$2@mailint03.im.hou.compaq.com>  y In article <20000816163820.23163.qmail@web1702.mail.yahoo.com>, "Jeffrey L. Cowgill Jr." <cowgillbones@yahoo.com> writes:TN :I am having a problems with my Dec Alpha running OpenVMS 7.1.  I am getting aO :pagetable full message and then insufficient memory errors that will not allow' :me to log in as any user.  E   On no specific error message(s) -- please remember to include thoseuH   when posting -- I am going to guess that your primary pagefile is too H   small, or that you need one or more secondary pagefiles, or that your K   secondary pagefile(s) are either too small or are otherwise inaccessable.C  ; :Has anyone had this problem and do you know how to fix it?r  F   If you cannot boot far enough to log in normally, then boot OpenVMS    Alpha conversationally:e      >>> b -fl 0,1     SYSBOOT> SET STARTUP_P1 "MIN"     SYSBOOT> SET WRITESYSPARAMS 0     and then invoke:  +    $ DIRECTORY/SIZE SYS$SYSTEM:PAGEFILE.SYS     $ RUN SYS$SYSTEM:SYSGENB    SYSGEN> CREATE SYS$SYSTEM:PAGEFILE.SYS/SIZE=larger-than-current     You will then want to reboot.m  ;   I use this explicit sequence rather than the more normal:       $ @SYS$UPDATE:SWAPFILES  B   as a minimal bootstrap may not install enough to permit the use B   of SWAPFILES.COM.  If you can perform a full bootstrap, you will   want to use SWAPFILES.COM.  = :I already tried adding a new pagefile but this did not help..  A   Then maybe the new PAGEFILE.SYS file was itself not big enough?h  ? :The system was working until a power outage brought it down.  aB :The hardware checks out ok, so I am thinking I have corrupt file.  E   You likely do not have a corrupt pagefile, as the contents -- with  D   one salient exception -- are reinitialized when you reboot.  It isG   possible to have corrupt blocks located underneath the pagefile, but sH   the reported symptoms are not the usual footprint for that corruption.G   (To resolve that problem, you will typically want to create a new andiG   different pagefile (rather than enlarging the current file), reboot, oH   and then delete the old pagefile containing the bad blocks -- OpenVMS I   and the disk controllers will then "get rid" of the bad blocks for you.eG   Adding controller or host-based shadowing (RAID) can help here, too.)r  H   You *may* have a dump sitting in the pagefile and occupying space, do -   you have the system parameter SAVEDUMP set?a  I   Do you use secondary files?  (These are usually connected via DCL addedsG   into the command procedure SYPAGSWPFILES.COM.)  You might have lost acG   secondary disk, and without the secondary files your system lacks theyG   room for the amount of paging required.  (I'd also look at the amounteE   of physical memory configured vs the amount required, but that's a p   generic recommendation.)  N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  % Date: Wed, 16 Aug 2000 20:50:38 -0400-2 From: "Richard B. Gilbert" <DRAGON@compuserve.com>& Subject: Help with pagefile full error6 Message-ID: <200008162050_MC2-AFEA-FF1@compuserve.com>  J         You are likely to get more useful help faster if you will give th= es; full and exact text of the error message you are receiving.   >         You may have corrupted you AUTOGEN.PAR file.  If so, a conversational boot: SYSBOOT> USE DEFAULT SYSBOOT> CONTINUE'  G         Will give you a set of parameters that will boot and run almostoD anything.   You can then run AUTOGEN to get a set of parameters more- appropriate to your particular configuration.0      0 Message text written by "Jeffrey L. Cowgill Jr."	 >Hello, =g    J I am having a problems with my Dec Alpha running OpenVMS 7.1.  I am getti= ng anH pagetable full message and then insufficient memory errors that will not allow:J me to log in as any user.  I have to boot to sysgen level to get in at al= l.  J Has anyone had this problem and do you know how to fix it?  I already tri= edJ adding a new pagefile but this did not help.  The system was working unti= lb a J power outage brought it down.  The hardware checks out ok, so I am thinki= ng IM have corrupt file.  Any ideas?<a   ------------------------------  # Date: Wed, 16 Aug 2000 22:39:01 GMTt! From: Beyonder <beyonder@vrx.net> ; Subject: Re: Here we go again - WTB/T/etc source listing CDs' Message-ID: <399B1800.1F675DBB@vrx.net>i   "D.Webb" wrote:w  B > Given that so many other operating systems are going open-sourceL > eg Solaris - see http://www.sun.com/software/solaris/source/overview.html.J > Isn't it time for Compaq to review the price of the VMS source listings.  9 long overdue, unfortunately won't help me though. sigh...v  > > $2000 plus a $790/yr update cost is much much too expensive.   You're not kidding!t   B.   ------------------------------   Date: 16 Aug 2000 15:57:16 PDTT From: Fairfield@SLAC.Stanford.EDU (Ken Fairfield; SLAC: 650-926-2924; FAX: 926-3515); Subject: Re: Here we go again - WTB/T/etc source listing CD03 Message-ID: <+sGW7jEn+3A7@mccdev.slac.stanford.edu>V  1 In article <8ne79k$2un$1@aquila.news.mdx.ac.uk>, e.     	david20@alpha2.mdx.ac.uk (D.Webb) writes:  B > Given that so many other operating systems are going open-sourceL > eg Solaris - see http://www.sun.com/software/solaris/source/overview.html.  :         Yes, and 60 billion flys can't be wrong: eat %^&*!  J > Isn't it time for Compaq to review the price of the VMS source listings. > > > $2000 plus a $790/yr update cost is much much too expensive.  H         Too expensive for  whom/what?   There  are  licensing  issues inH     addition to (the minimum cost of) creating and distributing the CDs.H     I'd  think  the cost barrier would aid keeping the listings  in  theH     hands of "serious" people, as opposed to the kiddie  generation  whoH     doesn't  even  understand  basic copyright issues!  Certainly $2K isH     not a barrier to anyone _needing_  the  listings in order to do codeD     development that requires internals knowlege _and_ the listings.  H         Just because a hobbiest/high schooler  can't afford the listings-     doesn't mean they should be given away...R           -Ken -- tM  Kenneth H. Fairfield            |  Internet: Fairfield@SLC.Slac.Stanford.Edu :  SLAC, 2575 Sand Hill Rd, MS 46  |  Voice:    650-926-2924:  Menlo Park, CA  94025           |  FAX:      650-926-3515N  -----------------------------------------------------------------------------B  These opinions are mine, not SLAC's, Stanford's, nor the DOE's...   ------------------------------  # Date: Thu, 17 Aug 2000 03:21:06 GMTi! From: Beyonder <beyonder@vrx.net>e; Subject: Re: Here we go again - WTB/T/etc source listing CD:' Message-ID: <399B5A13.A4FF1D80@vrx.net>n  9 "Ken Fairfield; SLAC: 650-926-2924; FAX: 926-3515" wrote:A  J >         Too expensive for  whom/what?   There  are  licensing  issues inJ >     addition to (the minimum cost of) creating and distributing the CDs.J >     I'd  think  the cost barrier would aid keeping the listings  in  theJ >     hands of "serious" people, as opposed to the kiddie  generation  whoJ >     doesn't  even  understand  basic copyright issues!  Certainly $2K isJ >     not a barrier to anyone _needing_  the  listings in order to do codeF >     development that requires internals knowlege _and_ the listings.  / hah! give me a break. $2k is way too expensive. N I want to learn, to read and understand. I was even going to do a paper on it,. but not with people with attitudes like yours!  J >         Just because a hobbiest/high schooler  can't afford the listings/ >     doesn't mean they should be given away...   R I'm in university you dope! I guess you don't remember what that's like. No money,P and no way to get this stuff unless someone is very nice, even just a short-term loan.y  + <begin SEVERE rant! --(no flames please)-->o  N Seriously what we have here is a situation of the "haves" and the "have nots".N What we have here is a bunch of people who paid some money (probably not theirN own, probably the company paid for it and it's not a personal expense), and weJ have a bunch of very immature people saying "nyah nyah, we have the sourceP listings and you don't". Or more seriously: "we have the listings, and you neverR will". The sort of uptight, snobbish, stuck up attitude that seems to infest a lot? of the older-older school generation (that would be pre-gen-x).r  Q Where it doesn't matter what someone wants or needs, it doesn't matter the reasonWR or cause, it'll never be given out, not because you can't give it out, not becauseM you can't lend it, just because you don't want to, it's yours and you have tooR protect your precious listings, even though you've probably got dozens of them. It doesn't matter.   R It's an issue of control and inflated self-importance and inflated self-worth. YouQ think you're better than others are. You enjoy seeing people suffer. Someone askseO for a kind favor, and your response is "tough, forget it", when would it really < hurt you in any way to help out? No, it certainly would not.  R I mean, just a loan for god-sakes. I'd even pay for courier charges to me and alsoO the return charges. I'd GIVE someone a fiche listing set, AND I'd be willing tosR PAY up to $200 US just to BORROW them for a short time. For god-sakes, what do you people want, blood ? jeez...  J I thought these types of attitudes died a long time ago. I guess greed andA hypcrisy are still alive and well today as they have always been.   O But nobody cares, I'm not sure why I'm even surprised. I thought there was evenRN one kind-hearted soul here who would help, but you've all proven that kindness doesn't exist.  C Thank you for re-affirming my belief that the human race is doomed.t  
 <end rant>  R I apologize for this but I can't take these attitudes any longer. If someone wouldH be kind enough to grant my request, I'd be more than eternally greatful.   B.   ------------------------------  % Date: Wed, 16 Aug 2000 16:08:29 -0500 + From: "Cliff Squidlow" <csquid@hotmail.com>e= Subject: Re: How to backup files since the last image backup.(- Message-ID: <6lDm5.15$bG.16431@nnrp3.sbc.net>S  I You may want to consider the /INTERCHANGE switch - it's good for avoiding : backing up all the *.dir's if you don't need ACL, etc info   ------------------------------   Date: 16 Aug 2000 23:37:57 GMT- From: "Phil Tregoning" <ptregoni@esoc.esa.de>o1 Subject: Re: How to fake a device full situation. = Message-ID: <01c007d9$4deab920$4b53b083@ptregoni.esoc.esa.de>   8 SysAdmin <djesys.nospam@fsi.net.mapson> wrote in article% <399AA7E0.2258EE3A@fsi.net.mapson>...s > Phil Tregoning wrote:h	 > [snip]  0 > > I think LDDRIVER on the freeware disk should7 > > allow this - in fact you can also use it for fakinga$ > > mount verification and the like. > @ > Now THERE'S something I'd be interested in hearing more about! >  > Can you post some specifics? >  > David J. Dachterao >   8 Have a look on the freeware (V4.1) disk. I haven't tried9 this myself, but the example in the read me includes this1 command:  > $ LD WATCH LDA1 5/FILE=LDA1:[000000]JUNK.DAT/ACTION=ERROR=%X2C  : Any access to block 5 of JUNK.DAT results in "SYS-F-ABORT" (i.e. status code %x2c).  4 $ DUMP LDA1:[000000]JUNK.DAT/BLOCK=(START=5,COUNT=1)6 %DUMP-E-READERR, error reading LDA1:[000000]JUNK.DAT;1 -SYSTEM-F-ABORT, abort  
 Cool, huh?  7 It also mentions an example MOUNTVER.COM for putting an # LD device into mount verification.     Phil T   ------------------------------  % Date: Wed, 16 Aug 2000 19:44:35 -0500 7 From: "David J. Dachtera" <djesys.nospam@earthlink.net>o1 Subject: Re: How to fake a device full situation.a- Message-ID: <399B3573.F4A5DB20@earthlink.net>s   Phil Tregoning wrote:s > : > SysAdmin <djesys.nospam@fsi.net.mapson> wrote in article' > <399AA7E0.2258EE3A@fsi.net.mapson>...i > > Phil Tregoning wrote:h
 > > [snip]2 > > > I think LDDRIVER on the freeware disk should9 > > > allow this - in fact you can also use it for faking & > > > mount verification and the like. > >oB > > Now THERE'S something I'd be interested in hearing more about! > >o  > > Can you post some specifics? > >d > > David J. Dachterai > >f > : > Have a look on the freeware (V4.1) disk. I haven't tried; > this myself, but the example in the read me includes this'
 > command: > @ > $ LD WATCH LDA1 5/FILE=LDA1:[000000]JUNK.DAT/ACTION=ERROR=%X2C > < > Any access to block 5 of JUNK.DAT results in "SYS-F-ABORT" > (i.e. status code %x2c). > 6 > $ DUMP LDA1:[000000]JUNK.DAT/BLOCK=(START=5,COUNT=1)8 > %DUMP-E-READERR, error reading LDA1:[000000]JUNK.DAT;1 > -SYSTEM-F-ABORT, abort >  > Cool, huh? > 9 > It also mentions an example MOUNTVER.COM for putting ane$ > LD device into mount verification.  2 Great stuff! I'll look into it! Thanx for the tip!   --   David J. Dachtera  dba DJE Systems  http://www.djesys.com/  : Unofficial Affordable OpenVMS Home Page and Message Board: http://www.djesys.com/vms/soho/a  H This *IS* comp.os.vms. So, a certain bias in postings is to be expected.  @ Feel free to exercise your rights of free speech and expression.  F However, attacks against individual posters, or groups of posters, are strongly discouraged.h   ------------------------------  % Date: Wed, 16 Aug 2000 17:38:00 -050087 From: "David J. Dachtera" <djesys.nospam@earthlink.net>w, Subject: Re: HP LJ8000 won't print correctly- Message-ID: <399B17C8.F7F77FDA@earthlink.net>.   smz@my-deja.com wrote: >  > Hello, > + > Newbie question regarding HP8000 printer.nH > I can't seem to get this printer to format the page correctly (appearsF > to not recognize a carriage return).  Prints accross the first line,H > then continues off the sheet. Only prints the first page, instead of aF > series of pages with one line.  I'm guessing a forms issue currently > set at NOINIT. > > > Currently, set up identically as HP 5si which are formatting+ > successfully (except for the extra page).e   O.k Usual requests here:  ( How are you sending data to the printer?  
 If TCP/IP: TCP/IP stack & Version?n   If other, specify...   We'll go from there.   -- 0 David J. Dachtera8 dba DJE Systemsd http://www.djesys.com/  : Unofficial Affordable OpenVMS Home Page and Message Board: http://www.djesys.com/vms/soho/t  H This *IS* comp.os.vms. So, a certain bias in postings is to be expected.  @ Feel free to exercise your rights of free speech and expression.  F However, attacks against individual posters, or groups of posters, are strongly discouraged.D   ------------------------------  % Date: Wed, 16 Aug 2000 16:49:42 -0400r5 From: "Sue Skonetski" <susan.skonetski@compaq.nospam>m+ Subject: If Airlines were operating systems16 Message-ID: <8neun5$ckj$1@mailint03.im.hou.compaq.com>  H You may have seen this before since its been around for awhile,  but its worth a repeat.s   Suee   OS Air -D Passengers walk out onto the runway, grab hold of the plane, push itI until it gets into the air, hop on then jump off when it hits the ground.a TheyD grab the plane again, push it back into the air, hop on jump off ...  
 Mac Airways - F The cashiers, flight attendants and pilots all look the same, talk the same and act the same.  A When you ask them questions about the flight, they reply that you F don't want to know, don't need to know, and would you please return to your seat and watch the movie.   Windows Airlines -D The terminal is neat and clean, the attendants courteous, the pilots capable.  J The fleet of Lear Jets the carrier operates is immense. Your jet takes offL without a hitch, pushes above the clouds and at 20,000 feet explodes without
 a warning.   OS/2 Skyways -E The terminal is almost empty - only a few prospective passengers millc about.F The announcer says that a flight has just departed, although no planes appear to be on the runway.   D Airline personnel apologize profusely to customers in hushed voices,? pointing from time to time to the sleek, powerful jets outside.fC They tell each passenger how great flight will be on these new jets6B and how much safer it will be than Windows Airlines, but they willE have to wait a little longer for the technicians to finish the flight  systems.  # Maybe until mid-1996. Maybe longer.$   Fly Windows NT -F Passengers carry their seats out onto the tarmac and place them in the@ outline of a plane. They sit down, flap their arms, and make jet( swooshing sounds as if they were flying.   Unix Express -E Passengers bring a piece of the airplane and a box of tools with themyK to the airport. They gather on the tarmac, arguing about what kind of planeeG they want to build.  The passengers split into groups and build several I different aircraft, but give them all the same name. Only some passengersn reach > their destinations, but all of them believe they have arrived.   VMS AirhF You get on the plane via a smooth-running escalator following a swift, efficient check-inB Passengers can choose a size of plane to suit their journey but anD intercom is an optional extra  Your seat is comfortable but spartan.F A command-line option brings a touch of decadence.  Your stewardess is6 Sharon Stone: Getting on but still well worth the rideB There are numerous refreshments to choose from with a mind-blowingL number of options, all of which are deliverable.  The plane has an extra setI of wings so it can support NT air without effort. All passengers can takeP	 advantageLH of an extra seat for failover.  All flight controls are mirrored and the systemsFL are 'fly by command' rather than 'fly by pushing button.  If an engine fails another J seamlessly takes it's place.  If the 'plane starts to labour through heavy loadingnH another engine is seamlessly powered up.  If the journey is extended the	 'plane isoJ upgraded in flight.  If the 'plane does happen to crash there's no problem because theu3 passengers also exist on another one 150 miles awayl   ------------------------------  % Date: Wed, 16 Aug 2000 16:17:18 -0400w- From: JF Mezei <jfmezei.spamnot@videotron.ca>u, Subject: Re: Ken Olson's "snake oil" comment, Message-ID: <399AF6CD.867F9E08@videotron.ca>   David A Froble wrote:tN > I had always knew that there was a shift in about 1980, give or take severalP > years, where DEC stopped listening to the customers, and began in many ways to, > tell the customer what was best for them.   L Didn't this coincide with a growth period when DEC began to hire ex-IBMers ?L Whenever I heard news of DEC hiring an ex guy from IBM, I always though "bigL mistake, the last think DEC needs is to become like IBM. If it wants to beat5 IBM, it must avoid the pitfals that IBM was stuck in.a   ------------------------------  % Date: Wed, 16 Aug 2000 20:50:37 -0400e2 From: "Richard B. Gilbert" <DRAGON@compuserve.com>  Subject: Re: LN03 Printer Issues6 Message-ID: <200008162050_MC2-AFEA-FEF@compuserve.com>  9 Message text written by INTERNET:gartmann@immunbio.mpg.deoD >In article <200008142336_MC2-AFAB-7EC1@compuserve.com>, "Richard B.( Gilbert" <DRAGON@compuserve.com> writes:: >        Thirteen years ago, the LN03 was a great printer! >-F >        Most people have retired these dinosaurs.  They are slow, lowJ >resolution, and expensive to maintain.  There is little reason, other th= an, >nostalgia, for anyone to keep them running.  > I still have one because it is able to print on cardboard ;-)<  9         I always used a "Magic Marker" (tm) for that! ;-)o   ------------------------------  % Date: Thu, 17 Aug 2000 03:51:58 +0930e/ From: Mark Daniel <mark.daniel@wasd.vsm.com.au>e Subject: Re: Mail / Message-ID: <399ADBC6.185430ED@wasd.vsm.com.au>R   HorseNuts wrote:  B (our sympathies, must make sitting for long periods uncomfortable)   ------------------------------  % Date: Wed, 16 Aug 2000 16:29:28 -0400g- From: JF Mezei <jfmezei.spamnot@videotron.ca>r" Subject: Re: Mass adding of users?, Message-ID: <399AF9A6.EB87759A@videotron.ca>  G I think that each site has its own specific needs, so there is no "one"w' solution to adding users automatically.a  N However, a good command file could be written faily easily to read a data fileM into various fields (student name, number, password, account, expiration dategH etc) and then do the UAF account creation, create the directory (set itsX ownership and version limit) and add a disk quota entry (and then loop to next student).  G Do you have established naming procedures ? Does your site already haveoL established procedures to come up with a unique UIC for each account ? ThereL are a lot of small details to look at (which is why there isn't a definitive% mass-user-creation program available.    ------------------------------  % Date: Wed, 16 Aug 2000 16:42:36 -0400t- From: JF Mezei <jfmezei.spamnot@videotron.ca>o" Subject: Re: Mass adding of users?, Message-ID: <399AFCB9.6533A27C@videotron.ca>   Another [obligatory] comment.w  E If you have ALL-IN-1 (or office server) on your machines, it has nicelM scripting capabilities and tools to help in creation of users. (for instance,0M finding first free UIC inside a group). You can use these to inspire yourselft in creating your own routines.  L (the ALL-IN-1 routines go much firther than creating the username/directory,2 they also create a whole bunch of ALL-IN-1 stuff).  E However, the direct access to the UAF via scripts is very useful (for+J instance, to generate lists of expired usernames which you can then delete etc).   Y There are similar tools availble outside of ALL-IN-1 as well. But they are "unsupported".i   ------------------------------  % Date: Wed, 16 Aug 2000 16:50:19 -0500t7 From: "David J. Dachtera" <djesys.nospam@earthlink.net> " Subject: Re: Mass adding of users?- Message-ID: <399B0C9B.93F8F292@earthlink.net>p   JF Mezei wrote:r >  > Another [obligatory] comment./ > G > If you have ALL-IN-1 (or office server) on your machines, it has nice/O > scripting capabilities and tools to help in creation of users. (for instance,rO > finding first free UIC inside a group). You can use these to inspire yourselfi  > in creating your own routines. > N > (the ALL-IN-1 routines go much firther than creating the username/directory,4 > they also create a whole bunch of ALL-IN-1 stuff). > G > However, the direct access to the UAF via scripts is very useful (for L > instance, to generate lists of expired usernames which you can then delete > etc).u > [ > There are similar tools availble outside of ALL-IN-1 as well. But they are "unsupported".y  F I suppose Dan Sugalski may have tuned out due to the static of recent.  B Still, I was just wondering whether VMS Perl has any UAF interface) modules akin to All-in-1's UAF "dataset".0   --   David J. Dachtera5 dba DJE Systems- http://www.djesys.com/  : Unofficial Affordable OpenVMS Home Page and Message Board: http://www.djesys.com/vms/soho/d   ------------------------------  # Date: Wed, 16 Aug 2000 20:02:43 GMTu' From: Gary Robinson <spgfr@my-deja.com>o  Subject: Newbie Decnet+ Question) Message-ID: <8nes0k$j7t$1@nnrp1.deja.com>   B I've recently inherited a Compaq Alpha system (VMS 7.1)and need toC change the node name and ip address. There are no manuals, and I've/B been flailing around with net$configure and ucx to no avail. Could someone please help?   TIA,  
 Gary Robinson.    & Sent via Deja.com http://www.deja.com/ Before you buy.s   ------------------------------   Date: 16 Aug 2000 23:18:22 GMT) From: leslie@clio.rice.edu (Jerry Leslie) $ Subject: Re: Newbie Decnet+ Question' Message-ID: <8nf7fu$9nb$1@joe.rice.edu>a  ( Gary Robinson (spgfr@my-deja.com) wrote:D : I've recently inherited a Compaq Alpha system (VMS 7.1)and need toE : change the node name and ip address. There are no manuals, and I've D : been flailing around with net$configure and ucx to no avail. Could : someone please help?   : TIA,   : Gary Robinson   > That procedure is documented in the OpenVMS FAQ, available at:  8   http://www.openvms.digital.com/wizard/openvms_faq.html   Look for the section titled:  =   MGMT9.  How do I change the node name of an OpenVMS System?e  - You might find this section of value as well:s  <   DOC1.   Where can I find online copies of OpenVMS manuals?  4 --Jerry Leslie     (my opinions are strictly my own)   ------------------------------    Date: 16 Aug 2000 18:46:09 -0500* From: kuhrt@eisner.decus.org (Marty Kuhrt) Subject: NFS serving and ypd+ Message-ID: <rIgBi8b7h21h@eisner.decus.org>   > We have a couple of Alphaservers and HSZ40s collecting dust in? our computer room at the moment.  Our Solaris geeks are lookinge= to implement a new disk farm and have been looking at NetApp,i Auspex, HSGs, etc.    C I said, "Hey, I could NFS serve disks from a VMScluster to you!".  w  : They scratched their pointy little heads and said, "Hmm...> Scalability, fault tolerance, clustering...  (They've heard my: "VMS is better" bit before, so they know some of the basic@ qualities)  Sounds like it might be a good idea.  Can you use yp for UID/GID syncing?"  c   I said, "I'll look into it"t  < So, is there a yp client for VMS?  Has anyone done something similar?  Gotchas?  Pitfalls?i  = I'm going to be setting up the Alphaservers and the HSZs as ae: test cluster, probably running V7.2-1 and Multinet 4.2A.  @ Performance isn't an issue at this point, just proof of concept.   ------------------------------  % Date: Wed, 16 Aug 2000 19:11:45 -0400l- From: Niranjan Rajaghatta <hemanir@yahoo.com>e' Subject: Re: openhack challenge and VMS ) Message-ID: <399B1FB0.63DFBA8A@yahoo.com>o  C Guys. I get the point. What Iam saying is, has COMPAQ/DEC done it ?t1 Have they atleast sponsored something like this ? 3 Have they atleast made it well known ? I think not.sJ Iam not talking about third party initiative here (although I did not know about them before).oH When a first party(COMPAQ/DEC) does a thing like this, it gets a totally different dimension.3 That is the dimension that VMS has to venture into.u   -RSN   Niranjan Rajaghatta wrote:  ; > Recently, pcweek, rigged up a web site 'openhack.com' andt > challenged/invited> > hackers to hack the site. It was rigged up using unix boxes.( > Solaris security bugs were discovered.4 > All this is not relevant to what I want to convey. > Read more atD > http://www.zdnet.com/eweek/stories/general/0,11011,2600258,00.html >1G > Since VMS is purported to be a much more secure system, and there are: > lessF > known exploits on VMS boxes, why should not compaq do something like > this ?D > Iam sure VMS will come out on top and it will be a great marketing > posture/strategyE > for compaq. Something like 'hackvax.com' should be a catchy phrase.-I > Beat the right drums about VMS's invulnerability and chanting the right"	 > phrasese3 > should give a boost to the VMS web server market.s >a6 > Anybody at compaq listening and open to such ideas ? >o > TIA, > -RSN7 > PS : If somebody does implement this, I want royalty.3   ------------------------------  % Date: Wed, 16 Aug 2000 10:09:17 -0400 / From: Bradford Hamilton <Brad.Hamilton@fmr.com> ) Subject: Re: OpenVMS Patches Mailing Listt' Message-ID: <399AA08D.3A3ED4BD@fmr.com>t  G I just un-subscribed and re-subscribed a user, because the user did notuH receive this morning's patches.  I'll keep an eye out to see if that did
 the trick!   Brad   Peter Weaver wrote:v > D > Richard L. Dyson wrote in message <39993CA9.5D088DFA@uiowa.edu>... > > ... D > > However, speaking of the devil, I just today got a patch listingC > >(ALPMANA02_071) for the first time in a LONG time (I sware it isc > longerH > >than last Feb 2000 like Peter reported...)  But, time flies when your > >having fun... :)0 > >]A > > Now, is this just a hiccup or is the patch list back on-line?s > >We'll have to watch and see.i >  > > ...v > G > Two people here received this patch notice today, two others did not, C > but our service rep tells us that all four of us are on the list.e >  > -- > $! Peter Weaveru: > $ input = "0756475627E277561667562704374756C636F6E23616"	 > $ i = 0. > $loop:. > $ my_address2[i*4,4]=%X'F$EXTRACT(i,1,input)
 > $ i = i + 1s, > $ if i .lt. f$length(input) then goto loop3 > $ write sys$output "My address is ''my_address2'"i   ------------------------------  % Date: Wed, 16 Aug 2000 13:36:33 -0700 / From: Terry Marosites <TMarosites@unitedad.com>r Subject: RE: OpenVMSwareM Message-ID: <1137A4A23A51D311B2D600105A1D5213026FDEDD@seantexch.unitedad.com>a  C If you take some of the Compaq surveys you can get Compac t-shirts.i7 But they are not as nice as the old VMS paraphernalia. VJ I love my 2 VMS baseball caps, both just Say VMS , one has a Cheesier Cat E and the other the shark. got them at the Col. Springs office while onKG vacation near by. and yes I have a "I love my VAX" bumper sticker on mynK truck, My field rep in 1980 ordered a bunch after getting some new systems.TD I don't think they do that anymore :( maybe they just don't have the. attitude that some could still be prod of VMS    Terry      -----Original Message-----# From: hoffman@xdelta.zko.dec.nospamd& [mailto:hoffman@xdelta.zko.dec.nospam]& Sent: Tuesday, August 15, 2000 1:42 PM To: Info-VAX@Mvb.Saic.Comt Subject: Re: OpenVMSware      J In article <p4wl5.99934$A%3.1321059@news1.rdc2.pa.home.com>, "Joseph Gill" <gilljb@home.com> writes:/I :Does anyone know if Compaq has an OpenVMSware web site where you can buypC :cool OpenVMS shirts, hats, mugs, etc simular to Sun's Sunware sites :(http://www.sun.com/sunware/)?i  F   The OpenVMS-specific materials are generally available (usually for     free) at trade shows and such.  F   I know of no website where this OpenVMS merchandise can be ordered, +   though that is an interesting suggestion.w  2  --------------------------- pure personal opinion ---------------------------.L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com      5 *****************************************************     5 *****************************************************c4 Any views or opinions are solely those of the author) and do not necessarily represent those of  United Advertising Media.n5 *****************************************************I4 The information transmitted is intended only for the1 person or entity to which it is addressed and mayt3 contain confidential and/or privileged material. If 3 you are not the intended recipient of this message,a. please do not read, copy, use or disclose this3 communication and notify the sender immediately. Ity0 should be noted that any review, retransmission,2 dissemination or other use of, or taking action in- reliance upon, this information by persons orm- entities other than the intended recipient isc prohibited.f5 *****************************************************o **   ------------------------------  % Date: Wed, 16 Aug 2000 09:12:19 +1000t+ From: "Stephen Bishop" <sbishop@mincom.com>- Subject: Perl on VMS+ Message-ID: <8ncikf$3k0$1@sol.mincom.oz.au>e  	 Hi there!s  I I'm new to VMS and Perl, and was hoping someone could tell me what is theg/ usual install directory for Perl under OpenVMS?    I know Unix looks like:f #!/usr/local/bin/Perl5  2 But have no idea what this usually looks like VMS.  J Also, I don't really know how to represent directory structures in my Perl code under VMS.o  7 Any help anyone can offer, will be greatly appreciated.h   Thanks in advance! Cheers!i Stevei   ------------------------------  % Date: Wed, 16 Aug 2000 20:00:39 -0500t7 From: "David J. Dachtera" <djesys.nospam@earthlink.net>a Subject: Re: Perl on VMS- Message-ID: <399B3937.F0238226@earthlink.net>t   Stephen Bishop wrote:n >  > Hi there!  > K > I'm new to VMS and Perl, and was hoping someone could tell me what is thei1 > usual install directory for Perl under OpenVMS?d >  > I know Unix looks like:a > #!/usr/local/bin/Perl5 > 4 > But have no idea what this usually looks like VMS. > L > Also, I don't really know how to represent directory structures in my Perl > code under VMS.a > 9 > Any help anyone can offer, will be greatly appreciated.s >  > Thanks in advance!	 > Cheers!a > Steve   6 Have a look at http://www.sidhe.org/vmsperl/index.html  : There's a link there for install instructions and the FAQ.   Might help, y'never know...n   -- o David J. Dachterah dba DJE Systems  http://www.djesys.com/  : Unofficial Affordable OpenVMS Home Page and Message Board: http://www.djesys.com/vms/soho/n  H This *IS* comp.os.vms. So, a certain bias in postings is to be expected.  @ Feel free to exercise your rights of free speech and expression.  F However, attacks against individual posters, or groups of posters, are strongly discouraged.n   ------------------------------  + Date: Wed, 16 Aug 2000 23:12:49 -0400 (EDT) " From: Dan Sugalski <dan@sidhe.org> Subject: Re: Perl on VMSF Message-ID: <Pine.LNX.4.21.0008162308550.2570-100000@tuatha.sidhe.org>  * On Wed, 16 Aug 2000, Stephen Bishop wrote:   > Hi there!n > K > I'm new to VMS and Perl, and was hoping someone could tell me what is thef1 > usual install directory for Perl under OpenVMS?l  G Perl lives under the PERL_ROOT logical, generally speaking. That can bey pointing anywhere you like.o   > I know Unix looks like:t > #!/usr/local/bin/Perl5 > 4 > But have no idea what this usually looks like VMS.  eE Perl's usually in as either a symbol or command in DCLTABLES, so thisY tends to look like:l  	   #! perl.    Nice 'n simple.o  L > Also, I don't really know how to represent directory structures in my Perl > code under VMS.h  G Use the native VMS dir structure, or a unixy equivalent. If you want to ' open FOO:[BAR.BAZ]XYZZY.TXT, then just:e  '    open FILE, 'FOO:[BAR.BAZ]XYZZY.TXT;'s   or, if you'd rather:  &   open FILE, '/FOO/BAR/BAZ/XYZZY.TXT';  I Either will work. Filenames are case-insensitive on VMS, though functionsdJ that hand them to you (like readdir) tend to hand them back in lower-case.  F Do be careful with dollar signs, as VMS is fond of them, but perl does> variable interpolation inside double-quoted strings with them.   					dan   ------------------------------  % Date: Wed, 16 Aug 2000 22:37:40 +0200I+ From: "M. van Eck" <m.van.eck02@freeler.nl>p Subject: Posix support8 Message-ID: <8neqqd$r6v$1@rubidium.news.lion-access.net>  J I have heard rumours that Posix support still exists for VMS. Could anyoneJ tell me if that is true? If Yes, who is supporting it and is a version for VMS 7.2 available?  
 M. van Eck   ------------------------------   Date: 16 Aug 2000 23:20:33 GMT) From: leslie@clio.rice.edu (Jerry Leslie)o Subject: Re: Posix support' Message-ID: <8nf7k1$9nb$2@joe.rice.edu>,  * M. van Eck (m.van.eck02@freeler.nl) wrote:L : I have heard rumours that Posix support still exists for VMS. Could anyoneL : tell me if that is true? If Yes, who is supporting it and is a version for : VMS 7.2 available?   See the OpenVMS FAQ:  8   http://www.openvms.digital.com/wizard/openvms_faq.html    specifically the section titled:  @   SOFT2.  Where can I find the UNIX <whatever> tool for OpenVMS?  4 --Jerry Leslie     (my opinions are strictly my own)   ------------------------------  % Date: Thu, 17 Aug 2000 04:12:59 +09300/ From: Mark Daniel <mark.daniel@wasd.vsm.com.au> 7 Subject: Re: Problem setting up YAHMAIL with OSU serverr/ Message-ID: <399AE0B3.3ACF6777@wasd.vsm.com.au>,  P I have updated a new minor release with the source code fix, and put that on the
 download page      http://wasd.vsm.com.au/wasd/  T so that others won't have this problem (though why it's taken so long to be reportedS is a bit unclear, I believe there are a number of PMDF installations around).  This T doesn't affect non-PMDF builds so there no need to go from 1.3.1 to 1.3.2 otherwise.  S Apologies, it looks like I failed to transfer an updated source file back to my own N development system (which lacks PMDF) after compiling and fixing this issue on@ another system an admin lets me use for the PMDF side of things.  8 >       status = PMDFwriteHeader (&PmdfNqCtx, &PmdfHdr);0 > .............................................^A > %CC-E-UNDECLARED, In this statement, "PmdfHdr" is not declared.g5 > at line number 699 in file SW$:[YAHMAIL]SENDMSG.C;1c > ...M  E +-------------------------------------------------------------------+hD  Mark Daniel            Opinions my own ... and on loan from others.E  mailto:Mark.Daniel@wasd.vsm.com.au (Mark.Daniel@dsto.defence.gov.au)eE +-------------------------------------------------------------------+    ------------------------------  % Date: Wed, 16 Aug 2000 01:43:21 -0700o5 From: Richard  <maher_rjNOmaSPAM@hotmail.com.invalid>sF Subject: Re: Programmatic access to byte count rx/tx on TCP/IP socket.9 Message-ID: <1f1200dc.27933669@usw-ex0108-062.remarq.com>u   Hi,o  4 The on-line documentation is brilliant! Well done to whoever set it up.  < http://www.openvms.digital.com:8000/72final/6529/6529profile _010.html#tcppmch05_14  + Look at $qio io$_sensemode (ioctl) fionreada  7 I'm not sure what counts you want but try $getdviw with19 devbufsiz, opcnt,recsize,refcnt,transcnt at least some of@< them are active and have useful information for BG: devices.   Regards Richard Maher.  1 PS. I thought you said your dog does not bite :-)h     * Sent from AltaVista http://www.altavista.com Where you can also find related Web Pages, Images, Audios, Videos, News, and Shopping.  Smart is Beautifulr   ------------------------------  % Date: Wed, 16 Aug 2000 03:20:10 -0700 5 From: Richard  <maher_rjNOmaSPAM@hotmail.com.invalid>rF Subject: Re: Programmatic access to byte count rx/tx on TCP/IP socket.9 Message-ID: <033fd065.40d506ec@usw-ex0108-062.remarq.com>    Hi,   ; I thought that I'd better point out that the usefullness orp; obtainability of device information *may* vary depending on  your TCP/IP supplier.o  ; If anyone is interested or is in a position to pass this on-9 to TCPWARE people then here are three differences between1 UCX and TCPWARE.  1 1) TCPWARE requires OPER privilege to io$_setmodej tcp_probe_idle. UCX does not.   8 2) TCPWARE accepts, but appears to ignore, socket option4 ucx$c_share. You get an error when you try to attach another channel.  : 3) io$_acpcontrol inetacp$c_trans is the dog's bollocks on: ucx and will gleefully translate a binary internet address9 into an ascii nodename! TCPWARE just does what it says ina: the documentation (hard to blame it for that) that is , it returns ascii 1.2.3.4.  7 What is the maximum buffer size that can be returned bye8 inetacp_func$c_gethostbyaddr? The examples I've seen use9 256 or 512 bytes but there is nothing in the manuals that : tell you how much memory to allocate of provide a function; to tell you. I'd prefer inet$acp$c_trans to work on all VMS ; versions of TCP/IP but if I have to change I'd like to knowh* what the C socket RTL (or such like) does.   Regards Richard Maher.     * Sent from AltaVista http://www.altavista.com Where you can also find related Web Pages, Images, Audios, Videos, News, and Shopping.  Smart is BeautifulD   ------------------------------  % Date: Thu, 17 Aug 2000 08:10:28 +0930 / From: Mark Daniel <Mark.Daniel@wasd.vsm.com.au>dF Subject: Re: Programmatic access to byte count rx/tx on TCP/IP socket.. Message-ID: <399B185C.351050C@wasd.vsm.com.au>   Richard wrote:  - > Look at $qio io$_sensemode (ioctl) fionreade  a My understanding of FIONREAD is it returns the count of incoming data, buffered but still unread.n  9 > I'm not sure what counts you want but try $getdviw withv; > devbufsiz, opcnt,recsize,refcnt,transcnt at least some ofi> > them are active and have useful information for BG: devices.   It's for parent-child socket sharing.  I'd simply like the parent to be able to independently keep track of the data transfered by the child (I was hoping  for a simple $GETDVI on the BG).  = None of the generic DVI items seem to provide what I'm after.b  3 > PS. I thought you said your dog does not bite :-)i  > Unless you live next door to me I do not know what this means.  ! Thanks for your response Richard.u   ------------------------------  % Date: Thu, 17 Aug 2000 08:01:18 +0930i/ From: Mark Daniel <Mark.Daniel@wasd.vsm.com.au>aF Subject: Re: Programmatic access to byte count rx/tx on TCP/IP socket./ Message-ID: <399B1636.8CFFA68B@wasd.vsm.com.au>/  I It's for parent-child socket sharing.  I'd simply like the parent to keep * track of the data transfered by the child.   "Mark D. Jilson" wrote:v  G > What will you be doing with the counts or is this just an exercise inl$ > obtaining obscure information? ;*) >  > Mark Daniel wrote: > >-K > > Is it possible to access via $QIO the sent and received counts of a BG:r > > device socket. > > K > > Furthermore (with due deference to Hoff's current pendantry :^) if this B > > is possible then is it practical?  And again, if so provide anK > > explanation and possibly a code example if it's not straightforward (ornI > > include where the interface documentation may be found :^)   If it isaL > > possible, is it known what versions of the BG: drive (UCX, DTCPIP, etc.)C > > it is supported for (and if so please include this info. also)?t > >nC > > Hmmm, I have a feeling I've left something out.  Oh well.  TIA.t >  > --H > Jilly   - Working from Home in the Chemung River Valley - Lockwood, NYL >         - jilly@clarityconnect.com                      - Brett Bodine fanL >         - Mark.Jilson@Compaq.com                        - since 1975 or so5 >         - http://www.jilly.baka.com               -n   ------------------------------  % Date: Wed, 16 Aug 2000 16:36:08 -0400 - From: JF Mezei <jfmezei.spamnot@videotron.ca>t Subject: Re: RD51 vs. RD53, Message-ID: <399AFB35.C257DC2F@videotron.ca>  L As someone else noted, if the RD53 fails to boot/spin up, it does show up asJ an RD51 drive. I had such a beast once. It would try/start to spin up, butN then get tired and stop. The electronics probably worked but the motor didn't.  H The RQDX3E module (and also I beleiv on the RQDX3 controller) has optionJ switches to determine if a unit is going to be a disk RDxx or a diskquetteL drive RXxx. But if VMS sees an RD51, I would *think* that your config is OK.   ------------------------------   Date: 16 Aug 2000 18:24:15 GMT2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)& Subject: Re: Reconfigure Network Cards6 Message-ID: <8nem8f$b7q$1@mailint03.im.hou.compaq.com>  [ In article <8nee4c$1sh$1@bw107zhb.bluewin.ch>, "Markus Eymann" <eymannm@bluewin.ch> writes:i< :How could I delete my Network Devices and create them new??  B   Beware: one-line questions are often rather difficult for folks >   to answer, particularly given the potential for ambiguity...      I will assume this is OpenVMS.     The usual questions apply:  ?     What is the particular problem you are attempting to solve?      Which OpenVMS version?       Which platform?  w      Which network protocol(s)?        Which network controller(s)?E     Is this specific to any particular applications?  (If so, which?)s  * :please send answers to eymannm@bluewin.ch     Ask here, get an answer here.   N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  % Date: Wed, 16 Aug 2000 15:14:48 -05001+ From: "Cliff Squidlow" <csquid@hotmail.com>.! Subject: Re: Samba (was PC-NFS ?)e, Message-ID: <MyCm5.11$bG.9320@nnrp3.sbc.net>  : "Phil Howell" <howellp@snowyhydro.com.au> wrote in message- news:rLmm5.90458$N4.2172868@ozemail.com.au...h >m6 > Cliff Squidlow <csquid@hotmail.com> wrote in message [SNIP]9 > samba is an alternative means of exporting file systemss [SNIP]G I have some limited experience with Samba when I set up a Linux server,kE actually I have very limited Linux/Unix/etc. experience period.  If IfJ remember correctly, it sets up a LANMAN type share on the server, which isE probably what I'm looking for.  Is there a VMS version of this around.@ somewhere?  Security would not be an issue on a home LAN server.   ------------------------------  % Date: Thu, 17 Aug 2000 11:12:33 +1000e/ From: "Phil Howell" <howellp@snowyhydro.com.au>t! Subject: Re: Samba (was PC-NFS ?)e. Message-ID: <KYGm5.224$Vb.5063@ozemail.com.au>   I go to  http://samba.anu.edu.au/samba/ which re-directs me to4 http://www.ifn.ing.tu-bs.de/ifn/sonst/samba-vms.html Phil4 Cliff Squidlow <csquid@hotmail.com> wrote in message& news:MyCm5.11$bG.9320@nnrp3.sbc.net... > < > "Phil Howell" <howellp@snowyhydro.com.au> wrote in message/ > news:rLmm5.90458$N4.2172868@ozemail.com.au...1 > >38 > > Cliff Squidlow <csquid@hotmail.com> wrote in message > [SNIP]; > > samba is an alternative means of exporting file systems1 > [SNIP]I > I have some limited experience with Samba when I set up a Linux server,gG > actually I have very limited Linux/Unix/etc. experience period.  If IiL > remember correctly, it sets up a LANMAN type share on the server, which isG > probably what I'm looking for.  Is there a VMS version of this aroundKB > somewhere?  Security would not be an issue on a home LAN server. >m >    ------------------------------   Date: 16 Aug 2000 20:08:33 GMT2 From: mathog@seqaxp.bio.caltech.edu (David Mathog) Subject: socket paths???, Message-ID: <8nesc1$bdk@gap.cco.caltech.edu>  K I'm having a look at the Postgresql code - and it's turning out to be much aH easier to work on than any of the other "portable" databases I've seen.   K However, in one location it tries to include <sys/un.h> (absent on OpenVMS)e@ and does this with some of the values from it in another header:  @ /* Configure the UNIX socket address for the well known port. */   #if defined(SUN_LEN)! #define UNIXSOCK_PATH(sun,port) \aN         (sprintf((sun).sun_path, "/tmp/.s.PGSQL.%d", (port)), SUN_LEN(&(sun))) #else ! #define UNIXSOCK_PATH(sun,port) \r?         (sprintf((sun).sun_path, "/tmp/.s.PGSQL.%d", (port)), \ H          strlen((sun).sun_path)+ offsetof(struct sockaddr_un, sun_path)) #endif  G What the heck is all this "sun" stuff and how do you work around it on s	 OpenVMS? n   Thanks,-   David Mathog mathog@seqaxp.bio.caltech.edu ? Manager, sequence analysis facility, biology division, Caltech *   ------------------------------  % Date: Wed, 16 Aug 2000 15:48:42 -0600*% From: Dan O'Reilly <dano@process.com>  Subject: Re: socket paths???C Message-ID: <4.3.2.7.2.20000816154623.00ba8818@pop.clsp.uswest.net>r  D Those are "UNIX domain sockets".  They're accessed as if they were aE directory path in UNIX, and they provide many of the same things thattE a file would, such as protection via ACL's, that sort of thing.  VerydE handy, unless you want to make them work on VMS, in which case you'relF generally screwed.  About all you can do is convert the offending code* into using mailboxes or something similar.  * At 02:08 PM 8/16/2000, David Mathog wrote:K >I'm having a look at the Postgresql code - and it's turning out to be muchrH >easier to work on than any of the other "portable" databases I've seen. >*L >However, in one location it tries to include <sys/un.h> (absent on OpenVMS)A >and does this with some of the values from it in another header:B >oA >/* Configure the UNIX socket address for the well known port. */$ >$ >#if defined(SUN_LEN) " >#define UNIXSOCK_PATH(sun,port) \@ >         (sprintf((sun).sun_path, "/tmp/.s.PGSQL.%d", (port)),  > SUN_LEN(&(sun))) >#else" >#define UNIXSOCK_PATH(sun,port) \A >         (sprintf((sun).sun_path, "/tmp/.s.PGSQL.%d", (port)), \sJ >          strlen((sun).sun_path)+ offsetof(struct sockaddr_un, sun_path)) >#endife > G >What the heck is all this "sun" stuff and how do you work around it ond	 >OpenVMS?s >  >Thanks, >e
 >David Mathog  >mathog@seqaxp.bio.caltech.edu? >Manager, sequence analysis facility, biology division, Caltecha   ------I +-------------------------------+---------------------------------------+3I | Dan O'Reilly                  |                                       | I | Principal Engineer            |  "Those are my principles. If you     |sI | Process Software              |   don't like them I have others."     |eI | http://www.process.com        |                    -- Groucho Marx    |sI +-------------------------------+---------------------------------------+c   ------------------------------  # Date: Thu, 17 Aug 2000 01:15:48 GMTl% From: jlsue <jlsuexxxz@dialupnet.com>s- Subject: Re: TCPIP - upgrade from 4.2 to 5.0ao8 Message-ID: <51fmps8jki5644817dj7sj6hmon48ubeub@4ax.com>  . On Tue, 15 Aug 2000 06:50:30 -0700, Mike Price5 <mike.priceNOmiSPAM@littlewoods.co.uk.invalid> wrote:r  - >Thanks to those who have responded by e-mailn1 >To answer my own question using Compaqs responseo >:8 >the symbols have changed from UCX..... to TCPIP..... to >'draw a line under UCX' >i@ >This means that we need to change all out TCPIP programs to get >them to re-link >oA >I don't really think that adhears to the  normal concepts of VMSt >upgrades!!o  @ Well, this really isn't a VMS upgrade.... it's a layered productF upgrade.  Anyway, yes, these changes have happened in the past in bothB VMS and layered products.  Usually a read of the release notes andC installation guide will provide enough info to properly prepare fort the upgrade.      1 Not speaking for anyone, certainly not DEC/Compaqn- (get rid of the xxxx in my address to e-mail)i   ------------------------------   Date: 16 Aug 2000 23:47:21 GMT- From: "Phil Tregoning" <ptregoni@esoc.esa.de>i Subject: Re: UCX BUFFERS ?= Message-ID: <01c007da$9ee44980$4b53b083@ptregoni.esoc.esa.de>a  : Terry Marosites <TMarosites@unitedad.com> wrote in articleD <1137A4A23A51D311B2D600105A1D5213026FDED5@seantexch.unitedad.com>... > Hello All, > ( > Tracing a hesitation problem I did a: " > $ UCX SHOW COMMUNICATION/MEMORY K > to see if I  have run out of buffers( see below for display).  I see thatsC > the free count for small_static buffers stays mostly at 0 and theYH > Large_static buffers stay is the single digit seldom to zero. How do I9 > increase the number of buffers for the static buffers?   >    Unsurprisingly, using:  + $ UCX SET COMMUNICATION/SMALL_BUFFERS=MIN:nh  5 But in your example there had been no waits or drops, 4 which in my experience usually means the buffers are OK.s   Phil T   ------------------------------  % Date: Thu, 17 Aug 2000 14:46:45 +1000h/ From: Ebeling Ross <Ross.Ebeling@tg.nsw.gov.au>-+ Subject: VAXStation 4000 Model 60 internals H Message-ID: <9E277D5010F0D311A8CB00508B7152B41DFF1C@exho1.tg.nsw.gov.au>  
 Colleagues   I am trying to establish a "hobbyist" VAX system to keep my hand in at home. I have the opportunity to bid on a VAXstation 4000 model 60 in an upcoming auction. I thought I knew something about older PCs, but VAXen are totally new to me!-   Can anyone tell me what form of CD-ROM is used in the VAXstations? Is it a "common or garden" ISE (as in standard PC), or is it a 50-pin SCSI?  [ Can I use any 50-pin SCSI HDD, or do I need a particular low-level format or ROM for VAXen?    Is "VAXen" the plural of VAX?S  h I suppose the most significant question is - "is there a VAX hardware FAQ, and where can I get a copy?"    Any help would be appreciated.  
 advthanksancev   Ross Ebeling   ross.ebeling@tg.nsw.gov.au      F **********************************************************************I "This electronic message and any attachments may be confidential.  If you J are not the intended recipient of this message could you please delete theH message and any attachments and advise the sender.  TransGrid uses virusI scanning software but excludes any liability for viruses contained in anyo attachment".  G This email may contain privileged and confidential information intendedeC only for the use of the addressees named above.  If you are not thewG intended recipient of this email, you are hereby notified that any use,aJ dissemination, distribution, reproduction of this email is prohibited.  If> you have received this email in error, please notify TransGrid@ immediately.  Any views expressed in this email are those of theF individual sender except where the sender expressly and with authority* states them to be the views of TransGrid".  F **********************************************************************   ------------------------------  + Date: Thu, 17 Aug 2000 05:40:06 +0000 (   ).3 From: Christopher Smith <chriss@Mufasa.pubserv.com> / Subject: Re: VAXStation 4000 Model 60 internals I Message-ID: <Pine.LNX.4.05.10008170522530.3282-100000@Mufasa.pubserv.com>-  ( On Thu, 17 Aug 2000, Ebeling Ross wrote:   > I am trying to establish a "hobbyist" VAX system to keep my hand in at home. I have the opportunity to bid on a VAXstation 4000 model 60 in an upcoming auction. I thought I knew something about older PCs, but VAXen are totally new to me!-  J You may find that you like them.  I certainly do.  At any rate, a hobbyistI system is a very good idea as far as I'm concerned.  The VAXStation 4000s  are very nice VAXen, too.s  I One word of caution:  VAXen are not PCs the way most people use the term. E If you've used worstation class machines before, you'll know what I'm.I talking about.  If not, let me say, for instance, that these are machines#J which are actually designed to be kept running (though, you don't have to,I of course).  They are also very particular about how they're shut down --nI PCs didn't, until recently, care so much.  Aside from those caveats, theyoH have all sorts of bennifits over the PC machines of today, for instance,G their ability to use a serial connection as the console monitor, or theWJ debugger, and incredibly nice set of tools (for testing things, formatting' disks, etc) that's built in to the ROM.-   > Can anyone tell me what form of CD-ROM is used in the VAXstations? Is it a "common or garden" ISE (as in standard PC), or is it a 50-pin SCSI?  I ISA is common these days?  The only things I've ever seen it used in wereu1 intel platform stuff, and a couple macintoshes ;)t  E All kidding aside, not only is it a SCSI drive, but it must also be an@ drive which can have a 512 byte block size (if memory serves...)  E Don't be too worried about finding one of these, though, they're verylI common, and, in fact are the very same drives which will boot up sun, hp,rH and sgi workstations (these machines also have this restriction) -- justF to name a few.  In short, if you can't find one that says it will workG with a VAX, you can just use one that says it will work with a sun(sgi,t" hp...), and you'll likely be fine.  ] > Can I use any 50-pin SCSI HDD, or do I need a particular low-level format or ROM for VAXen?W   Maybe, no, and no.. ;)  E There is no special ROM.  You can format the drive yourself, and I've I never noticed a drive that wouldn't work in a VAXStation 4000 -- but some H of them would just have to be tried.  In general, the answer is probablyF yes, get any SCSI hard disk.  Specifically, I've had success with someF older quantum and seagate disks, so those are the only ones I can give$ info about from personal experience.   > Is "VAXen" the plural of VAX?    Yes.  j > I suppose the most significant question is - "is there a VAX hardware FAQ, and where can I get a copy?"   H That's a tough one.  There's the comp.os.vms FAQ -- I'd start with that.H I don't have a current pointer, but do a search on dejanews through thisE group.  You'll probably find it -- and there was at least one pointer. posted to the FAQ just today.c  H There's also a comp.sys.dec -- you might check to see if they have a FAQ too.   HTH, and regards,m   Chris   O ===============================================================================i@ "My two cents"			(http://rootworks.com/twocentsworth.cgi?128562)= Christopher Smith(chriss@pubserv.com)			Prgramer^W Programmerr Prime Synergy of Champaign, IL.c% -------------------------------------eI "Where a calculator on the ENIAC is equipped with 18,000 vacuum tubes andsH weighs 30 tons, computers in the future may have only 1,000 vacuum tubes; and weigh only 1.5 tons." -- Popular Mechanics, March 1949 dO -------------------------------------------------------------------------------    ------------------------------  % Date: Wed, 16 Aug 2000 15:57:56 -0400n+ From: Clayton Kroh <ckroh@MAIL.CLARION.EDU>m Subject: VMS Mail probleme+ Message-ID: <399AF243.69BEFD87@clarion.edu>   G In preparation for a PMDF upgrade, I recently installed the OpenVMS ECOrH patch ALPMAIL03_071 on my Alpha system and needed to back it out becauseH it prevented our WebMail application from being able to delete messages.  E To back out the ALPMAIL03 patch I was told that I could simply rename G the five new images that were part of the patch. Here is what I did andMG it seemed to work okay because now we can delete messages from webMail.t   cd AXPDISK:[VMS$COMMON.SYSEXE]$ rename mail.exe   mail.exe_alpmail03, rename mail_old.exe   mail_old.exe_alpmail032 rename mail_server.exe   mail_server.exe_alpmail03 cd AXPDISK:[VMS$COMMON.SYSLIB]* rename mailshr.exe   mailshr.exe_alpmail03, rename mailshrp.exe   mailshrp.exe_alpmail03  H My problem now is that when users try to send mail via VMS mail and theyA use the "@" in the "To" field (ie. sending outside of the Clarion : University domain) we get this error listed at the bottom.  D POP3 and even Pine mail seems to work fine and we only see the error from VMS mail.  - Any ideas as to where I should start looking?i   Thanks in advance.     EXAMPLE:   $ mail  
 MAIL> send
 To:     ckroh 
 Subj:   test1 H Enter your message below. Press CTRL/Z when complete, or CTRL/C to quit:  
 this works  Exiti< New mail on node AXP3 from AXP3::CKROH        "CLAYTON KROH"  
 MAIL> send To:     ckroh@clarion.edu 0 %MAIL-E-ERRACTRNS, error activating transport IN> %RMS-E-IRC, illegal record encountered; VBN or record number =
 2156233140   MAIL>      -- Clayton Kroh Manager, Enterprise Services Clarion University of PA ckroh@clarion.edu5 (814) 393-1673   ------------------------------  % Date: Wed, 16 Aug 2000 11:03:58 -0700D0 From: Mark Berryman <Mark.Berryman@Mvb.Saic.Com>  Subject: Re: VMS Vs any other OS, Message-ID: <399A751E.694EEDF7@Mvb.Saic.Com>  > To all those participating in this thread - a gentle reminder:  D This forum is for the discussion of VMS and topics directly relatingG thereto.  Discussions of one's perception of the intellectual merits of G another are not related to VMS in any way that I can discern and, thus,rA are not appropriate for this forum.  I would ask to you move this0' discussion to a more appropriate venue.d   With thanks,  
 Mark Berryman: Mark.Berryman@Mvb.Saic.Com Info-VAX administrator   ------------------------------  # Date: Wed, 16 Aug 2000 18:33:48 GMTl= From: system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-)c  Subject: Re: VMS Vs any other OS0 Message-ID: <009EEB17.7423FF9B@SendSpamHere.ORG>  R In article <8neh7g$44p$1@pyrite.mv.net>, "Bill Todd" <billtodd@foo.mv.com> writes: >o1 >David P. Murphy <dpm@myths.com> wrote in message ) >news:spldh91nr5j78@news.supernews.com...i) >> Bill Todd <billtodd@foo.mv.com> wrote:  >>* >> > David P. Murphy <dpm@myths.com> wrote >>I >> >> how reminiscent of carl.os.vms:  one person deciding when "opinion"r >> >> becomes "pollution". >n >....c >r@ >> You've got a hair-trigger there, Bill.  Any particular reason? >> you're trying to insult me?  You've already claimed that you-E >> descend to the level of the person with whom you're conversing ---f2 >> but I've been holding a clean, reasonable chat. >lK >As with the quote I included above?  My impression is that comparison with:B >Lydick would constitute a pretty direct insult, but since I neverK >encountered his offerings that's only an inference.  And I won't bother toPH >recapitulate from your other posts considerably more direct aspersions.  G Then you're a newbie here, eh?  The long-time reader here will rememberrG the general atmosphere of this newsgroup when it was carl.os.vms and, I F am fairly certain I can comment for the majority that, we would prefer4 that comp.os.vms not erode to that state ever again.  F That said, your mannerisms are very much akin to that of the late CJL.F Come in, offer some useful and/or pertinent information, and then pro-F ceed to spew vehement remarks and slurs which the great many here will! expurgate with a kill file entry.   F Please tone it down.  This thread is acheiving a noise-to-signal ratioA the likes of which have not been seen here in a very great while.a   --O VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COMm   ------------------------------    Date: 16 Aug 2000 15:03:04 -0400/ From: jordan@lisa.gemair.com (Jordan Henderson)w  Subject: Re: VMS Vs any other OS* Message-ID: <8neoh8$2b4$1@lisa.gemair.com>  , In article <399A751E.694EEDF7@Mvb.Saic.Com>,2 Mark Berryman  <Mark.Berryman@Mvb.Saic.Com> wrote:? >To all those participating in this thread - a gentle reminder:r > E >This forum is for the discussion of VMS and topics directly relatinglH >thereto.  Discussions of one's perception of the intellectual merits ofH >another are not related to VMS in any way that I can discern and, thus,B >are not appropriate for this forum.  I would ask to you move this( >discussion to a more appropriate venue. >w  @ Wait a minute!  Are gentle reminders appropriate for this forum?  
 >With thanks,w >d >Mark Berryman >Mark.Berryman@Mvb.Saic.Com  >Info-VAX administrator-   -Jordan HendersonA jordan@greenapple.comr   ------------------------------  % Date: Wed, 16 Aug 2000 13:34:55 -0700 ! From: Shane.F.Smith@Healthnet.comt  Subject: Re: VMS Vs any other OSC Message-ID: <OF007CD65B.A84B232B-ON8825693D.00710463@HEALTHNET.COM>k  B No, physical violence is required to stop some of these people....   Shaneo          C jordan@lisa.gemair.com (Jordan Henderson) on 08/16/2000 12:03:04 PMs   To:   Info-VAX@Mvb.Saic.Como cc:h  ! Subject:  Re: VMS Vs any other OSt    , In article <399A751E.694EEDF7@Mvb.Saic.Com>,2 Mark Berryman  <Mark.Berryman@Mvb.Saic.Com> wrote:? >To all those participating in this thread - a gentle reminder:( >uE >This forum is for the discussion of VMS and topics directly relating)H >thereto.  Discussions of one's perception of the intellectual merits ofH >another are not related to VMS in any way that I can discern and, thus,B >are not appropriate for this forum.  I would ask to you move this( >discussion to a more appropriate venue. >a  @ Wait a minute!  Are gentle reminders appropriate for this forum?  
 >With thanks,- >- >Mark Berryman >Mark.Berryman@Mvb.Saic.Com0 >Info-VAX administratora   -Jordan Hendersone jordan@greenapple.coms   ------------------------------    Date: 16 Aug 2000 22:03:54 -0500* From: young_r@eisner.decus.org (Rob Young)  Subject: Re: VMS Vs any other OS+ Message-ID: <Anbr+5Vm8aRO@eisner.decus.org>o  p In article <009EEB17.7423FF9B@SendSpamHere.ORG>, system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-) writes:T > In article <8neh7g$44p$1@pyrite.mv.net>, "Bill Todd" <billtodd@foo.mv.com> writes:   > H > That said, your mannerisms are very much akin to that of the late CJL.H > Come in, offer some useful and/or pertinent information, and then pro-H > ceed to spew vehement remarks and slurs which the great many here will# > expurgate with a kill file entry.b >   = 	One interesting thing about Carl though.  In between callingn: 	the poster "crap for brains" or worse, he would post nice= 	working examples of code.  Got me started or leaning towards  	understanding several things.   				Robp   ------------------------------  % Date: Wed, 16 Aug 2000 23:25:17 +0200   From: Paul Sture <paul@sture.ch>  Subject: Re: VMS Vs any other OS+ Message-ID: <VA.000000a1.03e06b7f@sture.ch>H  9 In article <8neh7g$44p$1@pyrite.mv.net>, Bill Todd wrote:Q) > From: "Bill Todd" <billtodd@foo.mv.com>. > Newsgroups: comp.os.vmss" > Subject: Re: VMS Vs any other OS' > Date: Wed, 16 Aug 2000 13:00:33 -0400a >  > O Now then Mr Todd. I do have a complaint about you. When you climbed onto c.o.v eN (what - 9 months or so ago?), you reminded me of myself a few years ago, when O I was deprived of VMS at work, but still from memory, could come out with bits -/ of sound advice, or positive comments at least.   O Nowadays I wonder about your motives. I have seen you slag various people off,  O sometimes in discussions I wish to join in, often it dissaudes me from joining  H in. As various folks here know, I am not afraid to join in to technical N discussions - if I am wrong, someone will (normally) gently correct me, and I - will in return apologize if I misled someone.-  N I can trawl back over the last few months to find your comments, but I do not ( wish to do so. However, I must say this:  N My memory of many of your arguments goes like "...some technical argument"... E "but since that is not my current field, I cannot say for certain..."o  N After your exchanges of the last few days (nay, weeks, months...), I must ask H "What do you do for a living now?" OS details etc... Come on, even some A frustrations, and we may be able to suggest directions for you...e  & "What _is_ your current expertise???"   M Finally, "Why do you take such joy in what I can only call emotional insults nK to folks who will normally bend over backwards to swap technical info with .
 others here?"r  M You are getting close to being added to my killfile, although I would rather hD not do that, because when on form I _do_ feel you have something to  contribute.g ___w
 Paul Sture Switzerlando   ------------------------------  % Date: Wed, 16 Aug 2000 16:44:02 -0400n% From: "Brian Tillman" <tillman_brian>x Subject: Re: X windows$ Message-ID: <399afcc3$1@news.si.com>  J >There is this X windows program called SmartX that I use to connect to myJ >VMS Server. Now when I configure my SmartX session (which is actually a XK >windows client), there is a place where I give it my username,password andcL >the command to run when it starts up. This command is run when a successfulK >connection is made to the server. This is done I believe so that the usersAJ >can configure their own commands at startup (maybe run a script that will do2 >some tasks and then startup a X windows session).  K First, let's get the terminology correct.  In an X environment, your windoweK display device is the SERVER and the program what wants to put up a window,r* running on your VMS system, is the CLIENT.  > One command you should be able to use is, on your server side,  5 $ @sys$system:pcx$server 4,0,0,TCPIP, xxx.xxx.xxx.xxxu8 sys$system:DECW$TERMINAL.EXE -display xxx.xxx.xxx.xxx:0&  4 where xxx.xxx.xxx.xxx is your X server's IP address. --A Brian Tillman                   Internet: tillman_brian at si.comtA Smiths Industries, Inc.                   tillman at swdev.si.com8= 3290 Patterson Ave. SE, MS      Addresses modified to prevent < Grand Rapids, MI 49512-1991     SPAM.  Replace "at" with "@"8        This opinion doesn't represent that of my company   ------------------------------   End of INFO-VAX 2000.458 ************************