1 INFO-VAX	Fri, 03 Nov 2000	Volume 2000 : Issue 614       Contents:F Re: %EDT-F-INTERERR, Internal software error; contact a Compaq support Re: Alpha equivalent Re: Alpha equivalent Re: Alpha equivalent Re: Alpha equivalent Re: Alpha equivalent Re: Alpha equivalent Alpha equivalent0 Re: date bug in pcal (Calendar printing program)% Re: DCPS 1.8 and LN32 Duplex Printing $ Re: Decnet over IP - where to start?0 Development (Was Re: Sun Bloodies Compaq's Nose)4 Re: Development (Was Re: Sun Bloodies Compaq's Nose)' Re: Downloading FDL files from FTP site  Dynamic strings using VAX COBOL  Re: eBay (guess what) again. Re: eBay (guess what) again. Re: eBay (guess what) again. Re: eBay (guess what) again. Re: GAWK for VMS?  Re: GAWK for VMS? " Re: How to boost DLT backup speed?" Re: How to boost DLT backup speed?# How to decluster the cluster system & Re: Is "errno" thread safe in OpenVMS?& Re: Is "errno" thread safe in OpenVMS?& Re: Is "errno" thread safe in OpenVMS?& Re: Is "errno" thread safe in OpenVMS? Locks  Re: OS Update Ramification Porting from Unix to OpenVMS Re: RECALL question/suggestion. . Research Systems Inc. dropping support for VMS2 Re: Research Systems Inc. dropping support for VMS SCP (Secure CoPy) for OpenVMS? Re: Sun "uptime" belly laugher Re: Sun "uptime" belly laugher Re: Sun Bloodies Compaq's Nose Re: Sun Bloodies Compaq's Nose Re: Sun Bloodies Compaq's Nose+ Re: Sun's haunting ecache problems detailed # the lights are out at Northernlight ' Re: the lights are out at Northernlight ' RE: VAX Mail question - corrupted files < Re: VAXstation, console terminal, reboot, power-up, license? Re: VMS databases for hobbyist? - Re: VMS font configuration with Reflections-X * Re: Windows NT to OpenVMS access questions* Re: Windows NT to OpenVMS access questions  F ----------------------------------------------------------------------   Date: 2 Nov 2000 16:37:32 PST T From: Fairfield@SLAC.Stanford.EDU (Ken Fairfield; SLAC: 650-926-2924; FAX: 926-3515)O Subject: Re: %EDT-F-INTERERR, Internal software error; contact a Compaq support 3 Message-ID: <vVRQ1VOmm8xo@mccdev.slac.stanford.edu>   H         Please forgive me for reordering David's remarks, but they'll be#     easier to respond to if I do...   g In article <3A00F514.37832A8B@earthlink.net>, "David J. Dachtera" <djesys.nospam@earthlink.net> writes:  [...about EDT vs. TPU...] $ > *PLENTY* of reasons to not use it! > H > Here's a sample (from my EDTINI.EDT file, comments added for clarity):  H > DEFINE KEY      FUNCTION 34 AS "CHGUSR."(Make select range upper case)H > DEFINE KEY GOLD FUNCTION 34 AS "CHGLSR."(Make select range lower case) > 8 > DEFINE KEY GOLD  > AS "(SHR)."          (Shift right.)7 > DEFINE KEY GOLD  < AS "(SHL)."          (Shift left.)   H         These four functions  are  already  present  in  the default EVEH     section.   Binding them to keys in an initialization file is as easyH     as  what  you've  shown  above,  even  if  the  syntax  is  slightly     different.  3 > DEFINE KEY GOLD  B AS "EXT CHANGE =?'Buffer: '.."   H         If my _guess_ as to  what  this  does  is correct (and there's aH     good  chance  it's  not...),  EVE already has this.   That  is,  EVEH     readily maintains multiple buffers  and  has  a  BUFFER  command  toH     switch between them (and will prompt if no buffer name is provided).  5 > DEFINE KEY GOLD  D AS "DSR."		(Delete select range, J >                                          all other buffers (incl. PASTE)= >                                          remain unchanged.)   H         Trivially present already in  EVE.   Press  SELECT a second time/     (or GOLD-SELECT for cases that require it).   = > DEFINE KEY GOLD  G AS "PASTE=?'Paste from which buffer? '." 9 > DEFINE KEY GOLD  K AS "CUTSR=?'Cut to which buffer? '."   H         Here I'll guess that the tools available (EDT vs.  EVE) _impose_H     a style of work.  I  can't  imagine  why  I'd need to specify "whichH     buffer" for a Cut or Paste operation.  One merely splits the screen,H     displays  one  buffer in one window and another buffer in the  otherH     window, cuts (SELECT/REMOVE or SELECT/STORE  TEXT)  in  one  window,H     positions  the  cursor  in the other window and pastes (INSERT HERE)$     there.  Have I missed something?   + > DEFINE KEY GOLD  A AS "EXT CHANGE =ALT.."   B         I simply don't know what this does and so can't comment...  A > DEFINE KEY GOLD  \ AS "DATE."           (Modelled on WPS-Plus.) : > DEFINE KEY GOLD  C AS "D+CCUNDC."	(Transpose characters)6 > DEFINE KEY GOLD  J AS "DEWWUNDW."	(Transpose words.)9 > DEFINE KEY GOLD  L AS "D+NL L UNDL."	(Transpose lines.)   H         These  are  _trivially_   implemented   as  TPU  functions.   ByH     "trivially",  I mean 5-10 lines each at most.  Of course, style  and8     comments would probably dictate doubling that count.  3 > DEFINE KEY GOLD  X AS "SEL -KS +C CUTSR=UNPASTE." J >                                         (An "UNPASTE" key - suffers fromD >                                          a long-standing EDT bug.)  H         This is probably  the  hardest  to  implement, _if_ I understandH     what  it  does.  If it works in any way like LSE's UNDO command,  itH     would be tough to do in EVE.  If the idea is to take the last  rangeH     inserted  (PASTE'd)  and  remove  (CUT)  it, that would not be hard,H     meaning I know  how  I'd  do  it,  but  it  would  be  a  bit of TPU     programming.  J > Consider the amount of code needed to replace these functions in TPU andJ > you will understand why EDT remains the editor of choice for many common
 > operations.   H         No, I disagree.  Not  much  code  at  all  for  13 out of the 14E     examples you give.  A bit of code for the 14th, but quite doable.   I > (I've been building my EDTINI.EDT file since 1982, and see no reason to % > change to TPU at the present time.)   H         I not try to evangelize for EVE.  I think editor preferences areH     very  personal  and  based   on   familiarity   and  emotions  (likeH     preferences  in  members  of the opposite gender).  There  are  someH     arguments to be made the presence or absence  of  some  function  orH     another  in  one or another editor, but rarely is a choice of editorH     based on rational evaluation of all  the alternatives.  [That may beC     too strong, but you get the point...it's like language wars...]   H         I happen to like EVE because  of its extendability, and over theH     years  I've added more "cute" little functions than you can shake  aH     stick at.  But I'd never try to convince a  colleague  that  my  EVEH     section  is  "better"  than EDT, or LSE, or Emacs, or anything else.H     In fact, a colleague of mine from  my graduate school days still useH     a  line-by-line editor (called Wylvax).  And another (junior to  me)H     graduate student wrote the longest thesis of any to come out of  ourH     experiment (about 450 pages) using two fingers on each hand in Teco!H     He  could  type  faster  and  more  accurately than anyone I've seen     before or since!!            -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: Thu, 02 Nov 2000 12:55:43 -0500 0 From: David Beatty <David.Beatty.NOSPAM@sas.com> Subject: Re: Alpha equivalent 2 Message-ID: <HqoBOnb8XyDy26pnnsthNc0A5llP@4ax.com>  % Unsure future of VMS????  See the URL B http://www.openvms.compaq.com/solutions/government/coe/index.html.? This agreement implies a 20 year commitment to VMS.  Definitely  not unsure ...   David R. Beatty   > On Thu, 02 Nov 2000 15:43:40 GMT, bawilhelm@my-deja.com wrote:  B >We will soon be implementing PeopleSoft HR, FR and Student at ourE >University.  We are a VMS shop and were going to move to UNIX on Sun H >for our database server, due to the unsure future of OpenVMS.  With theF >recent move on Oracle's part to increase the support level for VMS weH >are considering staying with VMS on Alpha.  We had a Sun machine spec'dH >for the project and I am wodering if someone could give me a comparableF >alpha configuration based on performance.  I have been looking at theH >specs of both architectures and have been having difficulty coorelating	 >the two.  >  >Sun Enterprise 4500 Server . >4 x 400MHZ UltraSPARC CPU's 8M external cache >4 Gbyte memory  >  >Thanks  >Brent Wilhelm >Salisbury State University  >System Admin/DBA  >bawilhelm@ssu. edu  >  > ' >Sent via Deja.com http://www.deja.com/  >Before you buy.   ------------------------------  # Date: Thu, 02 Nov 2000 19:00:50 GMT = From: system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-)  Subject: Re: Alpha equivalent 0 Message-ID: <009F285D.C54D145C@SendSpamHere.ORG>  _ In article <mdiM5.242$0f6.6088@news.get2net.dk>, "Anders" <ccc28376a@post.cybercity.dk> writes: K >Explain what on earth made caused your determination "unsure future".  Its < >future is as certain as your mobile phone working tomorrow.  G We here on the comp.os.vms newsgroup would like to answer your question G but we are afraid it would be a waste of time "due to the unsure future # of" the Salisbury Steak University.    --O VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COM              O city, n., 1. a place where trees are cut down and streets are named after them.    ------------------------------  $ Date: Thu, 2 Nov 2000 18:59:31 +0100, From: "Anders" <ccc28376a@post.cybercity.dk> Subject: Re: Alpha equivalent 0 Message-ID: <mdiM5.242$0f6.6088@news.get2net.dk>  J Explain what on earth made caused your determination "unsure future".  Its; future is as certain as your mobile phone working tomorrow.     L <bawilhelm@my-deja.com> wrote in message news:8ts23a$7ei$1@nnrp1.deja.com...C > We will soon be implementing PeopleSoft HR, FR and Student at our F > University.  We are a VMS shop and were going to move to UNIX on SunI > for our database server, due to the unsure future of OpenVMS.  With the G > recent move on Oracle's part to increase the support level for VMS we I > are considering staying with VMS on Alpha.  We had a Sun machine spec'd I > for the project and I am wodering if someone could give me a comparable G > alpha configuration based on performance.  I have been looking at the I > specs of both architectures and have been having difficulty coorelating 
 > the two. >  > Sun Enterprise 4500 Server/ > 4 x 400MHZ UltraSPARC CPU's 8M external cache  > 4 Gbyte memory >  > Thanks > Brent Wilhelm  > Salisbury State University > System Admin/DBA > bawilhelm@ssu. edu >  > ( > Sent via Deja.com http://www.deja.com/ > Before you buy.    ------------------------------   Date: 2 Nov 2000 10:15:31 -0700 1 From: nothome@spammers.are.scum (Malcolm Dunnett)  Subject: Re: Alpha equivalent , Message-ID: <X+lBu2ix23oC@malvm1.mala.bc.ca>  1 In article <mdiM5.242$0f6.6088@news.get2net.dk>,  1    "Anders" <ccc28376a@post.cybercity.dk> writes: L > Explain what on earth made caused your determination "unsure future".  Its= > future is as certain as your mobile phone working tomorrow.  > D     You'd think Oracle taking a year to get a version of 8i releasedH for VMS would cause a bit of concern. Sure they've promised to do better> in the future, but only time will tell if they really mean it.  B     Is Oracle boosting their commitment to VMS because they reallyB believe in the product or is Compaq bankrolling the effort? If theA latter it could end up being just like the Win2K on Alpha fiasco, A ie as soon as Compaq gets tired of throwing money down a hole the ) entire project folds in a matter of days.    ------------------------------  % Date: Thu, 02 Nov 2000 15:32:44 -0500 0 From: David Beatty <David.Beatty.NOSPAM@sas.com> Subject: Re: Alpha equivalent 2 Message-ID: <C84BOmNGaEu6Mt0vRk7hcj7NkcZc@4ax.com>  ; Considering that 80% to 90% of the databases running on VMS < are owned by Oracle (either Oracle or Rdb), I would think it: would be in Compaq's best interest to ensure that boosting5 Oracle on VMS is in the best interest of Oracle Corp.   > Unfortunately, logic and reality don't always go hand-in-hand!  < On a side note, I find it just as sad that my company does a7 shoddy job of promoting SAS on VMS, especially with the 9 64-bit and DecThreads enhancements on the way in the next  major release.  Go figure ...    David R. Beatty   @ On 2 Nov 2000 10:15:31 -0700, nothome@spammers.are.scum (Malcolm Dunnett) wrote:   2 >In article <mdiM5.242$0f6.6088@news.get2net.dk>, 2 >   "Anders" <ccc28376a@post.cybercity.dk> writes:M >> Explain what on earth made caused your determination "unsure future".  Its > >> future is as certain as your mobile phone working tomorrow. >>  E >    You'd think Oracle taking a year to get a version of 8i released I >for VMS would cause a bit of concern. Sure they've promised to do better ? >in the future, but only time will tell if they really mean it.  > C >    Is Oracle boosting their commitment to VMS because they really C >believe in the product or is Compaq bankrolling the effort? If the B >latter it could end up being just like the Win2K on Alpha fiasco,B >ie as soon as Compaq gets tired of throwing money down a hole the* >entire project folds in a matter of days.   ------------------------------  % Date: Thu, 02 Nov 2000 15:34:51 -0500 0 From: David Beatty <David.Beatty.NOSPAM@sas.com> Subject: Re: Alpha equivalent 2 Message-ID: <pc8BOlqTRTSX=INOeEwhdCY2gI44@4ax.com>  : I did, admittedly.  However, the response by Goradz Kikelj got the job done.   1 On Thu, 2 Nov 2000 14:18:26 -0500, "Peter Weaver"  <peter.weaver@stelco.ca> wrote:   C >"Brian Schenkenberger, VAXman-" <system@SendSpamHere.ORG> wrote in 3 >message news:009F285D.C54D145C@SendSpamHere.ORG... < >> In article <mdiM5.242$0f6.6088@news.get2net.dk>, "Anders"& ><ccc28376a@post.cybercity.dk> writes:@ >> >Explain what on earth made caused your determination "unsure >future".  Its? >> >future is as certain as your mobile phone working tomorrow.  >>A >> We here on the comp.os.vms newsgroup would like to answer your 	 >question C >> but we are afraid it would be a waste of time "due to the unsure  >future & >> of" the Salisbury Steak University. > G >If anyone knows the answer (I do not) we should be helping. Because of G >the wording in the first line I think most people missed the part that  >said; > E >| ...                                                           With  >theE >| recent move on Oracle's part to increase the support level for VMS  >we 2 >| are considering staying with VMS on Alpha.  ... >  >  >  >    ------------------------------  # Date: Thu, 02 Nov 2000 15:43:40 GMT  From: bawilhelm@my-deja.com  Subject: Alpha equivalent ) Message-ID: <8ts23a$7ei$1@nnrp1.deja.com>   A We will soon be implementing PeopleSoft HR, FR and Student at our D University.  We are a VMS shop and were going to move to UNIX on SunG for our database server, due to the unsure future of OpenVMS.  With the E recent move on Oracle's part to increase the support level for VMS we G are considering staying with VMS on Alpha.  We had a Sun machine spec'd G for the project and I am wodering if someone could give me a comparable E alpha configuration based on performance.  I have been looking at the G specs of both architectures and have been having difficulty coorelating  the two.   Sun Enterprise 4500 Server- 4 x 400MHZ UltraSPARC CPU's 8M external cache  4 Gbyte memory   Thanks
 Brent Wilhelm  Salisbury State University System Admin/DBA bawilhelm@ssu. edu    & Sent via Deja.com http://www.deja.com/ Before you buy.    ------------------------------  " Date: Thu, 2 Nov 2000 22:30:06 GMT/ From: "Richard L. Dyson" <rick-dyson@uiowa.edu> 9 Subject: Re: date bug in pcal (Calendar printing program) ) Message-ID: <3A01968E.177E3D5D@uiowa.edu>    David Mathog wrote:  > M > The PCAL calendar printing program has a bug that shifts all the dates over K > by 1 day for 2001 and later.  A version that works correctly in and after  > 2001 is now up at: > 8 >   http://seqaxp.bio.caltech.edu/pub/SOFTWARE/PSCAL.ZIP  ? FWIW:  The current version of PCAL (v4.7) for OpenVMS and otherp	 platformsl can be found at:  7 	ftp://www-pi.physics.uiowa.edu/~dyson/pcal-4_7-vms.zipe  G I don't believe it has any of the problems David found with a very muchbF older version.  It also has tons more options then this older release.: It is also much more VMS-friendly and compiles for OpenVMS out-of-the-box.e  G > If you're using this program you should replace it before users start   > printing bogus 2001 calendars.   Regards, Rick -- nH Richard L. Dyson                                    rick-dyson@uiowa.eduH  _   _      _____                http://www-pi.physics.uiowa.edu/~dyson/H | | | |    |_   _|   Systems Analyst                     O: 319/335-1879H | | | | of   | |     The University of Iowa            FAX: 319/335-17536 | \_/ |     _| |_    Department of Physics & Astronomy-  \___/     |_____|   Iowa City, IA 52242-1479M   ------------------------------  $ Date: Fri, 3 Nov 2000 05:12:22 +0800' From: "Maverick" <seawyk@letterbox.com>E. Subject: Re: DCPS 1.8 and LN32 Duplex Printing0 Message-ID: <8tsl8j$tcr$1@newsie.singa.pore.net>  > Thks for the reply, I'll test it out and revert on the results   rgds   Maverick  ? "David Mathog" <mathog@seqaxp.bio.caltech.edu> wrote in messageo& news:8tn9k3$o8r@gap.cco.caltech.edu...J > In article <paul.r.anderson-EF2AC8.14372531102000@news.compaq.com>, Paul- Anderson <paul.r.anderson@compaq.com> writes:h> > >In article <8tn34r$coo$1@newsie.singa.pore.net>, "Maverick"  > ><seawyk@letterbox.com> wrote: > >lI > >> However, print jobs that I have submitted with the duplex option hastH > >> also returned a message that says that duplex option not supported. > >  > >Please post >pI > It works on my system, so here are the values for the positive control:- >- > >-# > >   - PRINT command you are using3 >S > These both work@ >CB >   PRINTD == "PRINT/QUEUE=CPS_ANSI/NOFLAG/PARAMETERS=(SIDES=TWO)" >   PRINTMD ==L "PRINT/QUEUE=CPS_ANSI/NOFLAG/PARAMETERS=(LAYUP="(MARGINS=72,72,72,72;NOBORDE RS)",SIDES=TWO)" >a6 > >   - Result of SHOW QUEUE /FULL of LN32 print queue >g >   $ sho que/full/all cps_ansi " >   Generic printer queue CPS_ANSII >     /GENERIC=(CPS) /OWNER=[SYSMGR,SYSTEM] /PROTECTION=(S:M,O:D,G:R,W:S)y /SCHEDULE=(NOSIZE) >   $ sho que/full/all cpsH >   Printer queue CPS, idle, on SEQAXP::DCPS$, mounted form DCPS$DEFAULT (stock=DEFAULT)D) >     <LN32 (with duplexor) in 158 Braun>EC >     /BASE_PRIORITY=4 /DEFAULT=(FORM=DCPS$DEFAULT (stock=DEFAULT)) - /NOENABLE_GENERIC /LIBRARY=DCPS_LIB Lowercase 0 >     /OWNER=[SYSMGR,SYSTEM] /PROCESSOR=DCPS$SMB0 /PROTECTION=(S:M,O:D,G:R,W:S) /SCHEDULE=(NOSIZE) >LF > >   - Error message returned and whether it's printed on the printer  > >     or on the OpenVMS system >u' > Can't help with that one - no errors.y >nD > On the hardware side the original poster should make sure that theH > cord from the duplexor is firmly plugged into the appropriate locationE > on the LN32.  That's around the back of the machine and if somebodyaF > pulled it out when working on the machine (or never plugged it in inB > the first place) that would not be evident from the front of the
 > machine. > A > By the way, this is the best printer I've ever used.  It's just C > amazing how fast it can print (even in duplex mode) and the printRF > quality is excellent. We've had only one paper jam in the year we'veG > owned it.  The only other problem of note was a glitch involving DCPS F > and the Windows drivers (which caused some VMS print jobs to requeueB > and print again) and that was corrected by upgrading the printerC > firmware.  We upgraded DCPS too but that turned out not to be the.A > problem.  It's got real Adobe Postscript 3 in it so none of the A > postscript emulation problems arise. Genicom really should do asF > better job marketing these things - they are a lot less work to dealG > with than the HPs most everybody else is using and the cost is prettyrF > close (if you shop around).   Ours is only the duplexor and the baseH > unit which has enough capacity for us.  With the optional paper feeder> > under it the LN32 can print a small forest's worth of paper. >(
 > Regards, >  > David Mathog > mathog@seqaxp.bio.caltech.eduK@ > Manager, sequence analysis facility, biology division, Caltech   ------------------------------   Date: 2 Nov 2000 15:30:09 PSTDT From: Fairfield@SLAC.Stanford.EDU (Ken Fairfield; SLAC: 650-926-2924; FAX: 926-3515)- Subject: Re: Decnet over IP - where to start?D3 Message-ID: <YVtn4QkoMiIu@mccdev.slac.stanford.edu>   * In article <8tmr2q$tde$1@nnrp1.deja.com>, 6     	Mike Price <mike.price@littlewoods.co.uk> writes:H > We have some boxes using decnet phase 5 on VMS7.1-2. They also use UCXI > 4.2. Our comms people don't like us using DECnet throught their routersYC > so I was wondering about using decnet over IP to keep them quiet.   H         You don't say whether you're currently using OSI under DECnet V.H     There are three different transports  possible  in Phase V: OSI, NSPH     (the  old  Phase  IV  method)  and  TCP/IP.   Which  you  use  is  a1     configuration matter on your Phase V systems.l  H         Depending on your requirements  (in  Europe, you _may_ very wellH     want/need the OSI transport), you can limit yourself to any one, any!     two, or all three transports.t  H         If your systems have only  a  single network adapter, and if UCXH     is  already  setup  to  provide DNS, then you really  only  need  to     execute,  *         $ @SYS$MANAGER:NET$CONFIGURE BASIC  7     on your systems.  One of the first prompts will be,f  I * Enter the directory services to use on the system [LOCAL,DECDNS,DOMAIN]   H     If you want OSI  and  TCP/IP,  respond  with DECDNS,DOMAIN .  If youH     want  to use Phase IV address and TCP/IP, respond with LOCAL,DOMAIN.H     I believe you can specify DOMAIN by itself if you don't want to  useH     OSI or NSP, bu I have a vague recollection that you may be forced toH     add  one  of  the  two  DECnet transports as well.  The installation     manual notes that,  D        "If you choose to enter  more  than one directory service forD         your  system,  enter  them in the order  of  priority.   TheD         ordering  of  this  list  is  _very_  important:  the  firstD         directory  srvice  entered  in  this  list is considered the:         primary directory service to use on the system..."  H         As an example, in addition to  the "main" AlphaServers and VAXesH     in  my  cluster, which are running DECnet IV under VMS 7.2-1/7.2,  IH     have two small Alphas, each with their own system disk, on which I'mH     running DECnet-Plus.  On these two  system,  I  have  the  directoryH     services configured as LOCAL,DOMAIN.  This allows me to connent withH     NSP  (LOCAL  database) to the other VMS system on the LAVC, but withH     TCP/IP (DOMAIN)  to  other  VMS  clusters  on-site  that are runningH     DECnet-Plus.  The main issue here is that DECnet is _not_ routed on-H     site.  Therefore, a SET HOST NODEX (where NODEX is _not_ in my localH     network) will _immediately_ fail and the DEnet-Plus will immediatelyH     retry  with DOMAIN (and connect!).  If I were to have configured theH     directories in the other order, DOMAIN,LOCAL, the DECnet-Plus  nodesH     would  attempt to connect to the Phase IV nodes via DOMAIN and wouldH     wait something like 60 to  90  seconds  (don't recall exactly which)H     before timing-out the TCP/IP connection attempt and retrying on NSP.  H         This is all to say that TCP/IP DECnet transport works very well,H     but you need to consider the capabilities of the target nodes you'reH     trying to reach and the protocols that the network will pass between	     them.D           -Ken -- eM  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, 02 Nov 2000 19:32:41 +0000   From: steven.reece@quintiles.com9 Subject: Development (Was Re: Sun Bloodies Compaq's Nose)oH Message-ID: <OF11B3954B.CFBE35FA-ON8025698B.006A4B95@qedi.quintiles.com>  I I've just sat through two days of presentations on OpenVMS, the future ofiJ OpenVMS and the DII COE work which Compaq have done, are doing and will be  doing for DII COE Certification.  # My conclusion from the experience ?l   I'M IMPRESSED!   I'M VERY IMPRESSED IN FACT!!  I Compaq have done a lot of great work here and it is all for the future ofoG OpenVMS.  The renaissance is real, the future is real and the future is5I OpenVMS.  The commitment to OpenVMS comes right from the top.  If nothingjG else, you don't spend the kind of cash that Compaq have spent on recentRJ work on something that has only a few months to live.  OpenVMS is here and is here to stay.  G For those customers that want the best hardware platform but don't wantuI OpenVMS then try Tru-64 or Linux.  There's no need to go to Wintel if you  don't want to.  H Real cash and real work is going into pulling software vendors back, notB least of which must be the DII COE work since it provides a common environment.  D As for Sun bloodying Compaq's nose - it's only because they can't do better.t   ------------------------------  $ Date: Thu, 2 Nov 2000 14:07:42 -06001 From: "Dave Gudewicz" <david.gudewicz@abbott.com>-= Subject: Re: Development (Was Re: Sun Bloodies Compaq's Nose)r8 Message-ID: <8tshb7$9c5$1@fizban.fizban.pprd.abbott.com>  L This was good to hear, thanks you for your report, and similar to what I and# others felt at CETS2000 last month.a  H We were told that on VMS's top 10 list of things to do, DII COE took the5 first 9 slots and there was nothing in the 10th.  ;-)a   Dave...2  - <steven.reece@quintiles.com> wrote in messagelB news:OF11B3954B.CFBE35FA-ON8025698B.006A4B95@qedi.quintiles.com... > K > I've just sat through two days of presentations on OpenVMS, the future ofeL > OpenVMS and the DII COE work which Compaq have done, are doing and will be" > doing for DII COE Certification. > % > My conclusion from the experience ?  >h > I'M IMPRESSED! >i > I'M VERY IMPRESSED IN FACT!! > K > Compaq have done a lot of great work here and it is all for the future ofeI > OpenVMS.  The renaissance is real, the future is real and the future iseK > OpenVMS.  The commitment to OpenVMS comes right from the top.  If nothingiI > else, you don't spend the kind of cash that Compaq have spent on recent L > work on something that has only a few months to live.  OpenVMS is here and > is here to stay. >@I > For those customers that want the best hardware platform but don't wantyK > OpenVMS then try Tru-64 or Linux.  There's no need to go to Wintel if youm > don't want to. >eJ > Real cash and real work is going into pulling software vendors back, notD > least of which must be the DII COE work since it provides a common > environment. >eF > As for Sun bloodying Compaq's nose - it's only because they can't do	 > better.r >t >i   ------------------------------  % Date: Thu, 02 Nov 2000 20:31:54 -0500o2 From: norm lastovica <norman.lastovica@oracle.com>0 Subject: Re: Downloading FDL files from FTP site) Message-ID: <3A02158A.928486A@oracle.com>o  * absolutely.  I've had very good luck using) ZIP "-9Vvj" with the ZIP off the freeware - CD.  Preserves VMS file attributes perfectly.u   Guard Hewitt wrote:  > L > We have run into problems with customers downloading FDL files off our FTPF > site using a browser. The only way we've found to download the filesK > correctly is using the GET/FDL command from VMS. Is there a way to zip or > > package these files so they make it off the FTP site intact?   -- 2> norman lastovica / oracle rdb engineering / usa / 610.696.4685   ------------------------------  # Date: Fri, 03 Nov 2000 01:56:11 GMTa From: deja_foo@my-deja.com( Subject: Dynamic strings using VAX COBOL) Message-ID: <8tt5vp$8hp$1@nnrp1.deja.com>)  - I've been checking out creating and accessinge* dynamic strings using LIB$SGET1_DD and the- LIB$SCOPY / STR$COPY family of routines. I'veo0 seen samples in a few languages but haven't seen" a COBOL (VAX or Alpha) sample yet.  0 Anyone have a COBOL sample showing how to create- a dynamic string, load it with the value of ab. working-storage field, and maybe even copy the+ value of the dynamic string back to anotheri working-storage field?    & Sent via Deja.com http://www.deja.com/ Before you buy.n   ------------------------------  $ Date: Thu, 2 Nov 2000 23:38:17 -0500  From: John Santos <JOHN@egh.com>% Subject: Re: eBay (guess what) again.d5 Message-ID: <1001102233333.2093B-100000@Ives.egh.com>f  + On Thu, 2 Nov 2000, Terry C. Shannon wrote:g   > > > "Jordan Henderson" <jordan@lisa.gemair.com> wrote in message& > news:8ts3ap$vdh$1@lisa.gemair.com.... > > In article <3A016143.F178C010@uk.sun.com>,6 > > andrew harrison  <andrew.nospam@uk.sun.com> wrote: >  > <snip> > > A > > >I have not bothered detailing the depths of your FUD attemptl? > > >with respect to Sun and Java but as I keep saying you needI > > >to get more mirror time.  > > >u > >k? > > I'm glad you brought up Java again.  Because your post here > > > where you said that the Java available for OpenVMS is Java@ > > "yes and no" is a classic example of FUD, which I countered.A > > There's little doubt that for the best, most up-to-date, Javao> > > almost everyone will tell you that Sun is the place to go.< > > Using this fact against competitors is a pure, classical > > FUD technique. > I > One wonders if Java ultimately will bite Solaris in the backside. Sun'spF > claim to fame in the Unix space (and rightfully so) is the humongousM > boatload of apps available on Solaris. Tru64 of course pales in comparison,G4 > and apps availability hath cost the Q many a sale. > J > OK, what's gonna happen if and when ISVs start writing their key apps inK > Java? If Java is write-once, run-anywhere, and Compaq, et al, have decenta= > Java implementations, apps availability many no longer be ah > platform-dependent issue.   A No problem.  Sun has an easy work-around.  Every 6-8 months, theywC release a new version of Java (with an obfuscated version numberingk? scheme to make it hard to tell what's going on) that is neitherd@ upward nor downward compatible with the previous version.  Since> it takes 6-8 months for all the other vendors to catch up, Sun@ always has the latest and greatest Java implementation and every@ one else is second best.  The lack of compatibility ensures that= you can't just recompile your Java sources on the new release ? (God forbid you copy over your .class files), but have to check0& everything out, forcing further delay.   --   John Santosa Evans Griffiths & Hart, Inc. 781-861-0670 ext 539   ------------------------------  % Date: Thu, 02 Nov 2000 15:42:09 +0000 0 From: andrew harrison <andrew.nospam@uk.sun.com>% Subject: Re: eBay (guess what) again.r* Message-ID: <3A018B51.9B95654F@uk.sun.com>   Warren Spencer wrote:O > H > andrew.nospam@uk.sun.com (andrew harrison) wrote in <3A016143.F178C010 > @uk.sun.com>:h > G > >My statement that Digital/Compaq were unable to show any performanceeA > >benefits of using a 64bit OS and VLM for DBMS's measured usingnI > >standard benchmarks such as TPC/SAP/PeoplSoft etc is TRUE, you may not F > >like it but you won't get any support from anyone at Compaq to backA > >you up because there is no support to available. Well with theh1 > >exception of Kerry but you want support right.y > 2 > SAP/Peoplesoft is hardly a "standard benchmark". >   8 They are standard applications benchmarks, audited using8 standard run conditions and providing a standard results3 reporting mechanism. The spec of the benchmarks is :
 available.  @ Vendors running the benchmark have to use the same applications,: same queries, same middleware, connected clients etc etc.   > With SAP Sd R3 or R4 for example they get to choose how their > configure the HW and which DBMS to use though most people use  Oracle but thats about it.  @ In many ways they are more standard than TPC-C for example where@ almost everything is up for grabs, DBMS, Middleware, programming6 language, shared everything or shared nothing etc etc.   Regardsn Andrew Harrisone Enterprise IT Architects   ------------------------------   Date: 2 Nov 2000 11:04:41 -0500t/ From: jordan@lisa.gemair.com (Jordan Henderson)s% Subject: Re: eBay (guess what) again.n* Message-ID: <8ts3ap$vdh$1@lisa.gemair.com>  * In article <3A016143.F178C010@uk.sun.com>,2 andrew harrison  <andrew.nospam@uk.sun.com> wrote: >Jordan Henderson wrote: >> n- >> In article <39FFFA48.DC111710@uk.sun.com>, 6 >>   andrew harrison <andrew.nospam@uk.sun.com> wrote: >> > >> > [snip]p >> >G >> > The capabilities of the HW platform is also important, VLM without C >> > a reasonable amount of available memory isn't that usefull and E >> > that was always the issue with the 8400/GS140. This was probablytE >> > the reason why Digital/Compaq were never able to demonstrate the-
 >> > benefits-H >> > of 64bit/VLM on any standard DBMS benchmarks that could be comparedB >> > with other vendors results. In fact there is no evidence thatD >> > Compaq tried to make any use of VLM for any of these benchmarksE >> > and if they did there the results obtained were not competitive." >> > >> qI >> Can anyone doubt that Andrew is on a FUD campaign against Compaq here?. >>   >Jordane >a? >Lets examine the statement you are complaining about shall we s? >and try to get to the bottom of your FACT vs FUD dillema shallo >we. >eE >My statement that Digital/Compaq were unable to show any performance @ >benefits of using a 64bit OS and VLM for DBMS's measured using H >standard benchmarks such as TPC/SAP/PeoplSoft etc is TRUE, you may not D >like it but you won't get any support from anyone at Compaq to back@ >you up because there is no support to available. Well with the / >exception of Kerry but you want support right.  > G >So this apparently is FUD, its TRUE, can it be FUD if its TRUE. Now ofM? >course I will be happy to be accused of FUD if you are saying ,@ >that FUD is simply a statement that is entirely TRUE but which   >is simply uncomfortable to you. >   A How do you know if it's true if there's "no evidence" as you say? B It's just conjecture on your part.  It may be true, it may not be.A Regardless of whether it's true or not, which you can't prove oneu@ way or the other, it's definitely FUD.  There may be a number of> reasons that Compaq/Digital didn't produce benchmarks that you? aren't aware of.  To gather from this that there's no advantageA< to VLM on Compaq OS's from "no evidence" is just conjecture.  > As I pointed out, but you apparently don't understand (or are = taking a pose of ignorance so as to forward your agenda), FUD'= and fact are orthogonal to one another.  FUD typically has a  A grain of fact, taken out of context, used in isolation from othereA facts, presented with some spin to bring about Fear, Uncertainty l and Doubt in the reader.  A I redirect your attention to the history of the term FUD.  It was @ invented by Gary Amdahl, in description to what IBM was doing to= his customers.  It just so happened that much of what IBM wasl; saying was true, you couldn't expect as good support on IBMtB operating systems when using Amdahl machines, the future of AmdahlA as a vendor was not as secure as IBM, etc. etc.  The problem withu? this kind of FUD is that it's designed to create an atmosphere e< where the statements are even more true than they once were.  < Today, Microsoft does exactly this to it's competitors, and : apparently, so does Sun.  You say above that you are happy9 to be accused of FUD.  Since you appear to represent Sun  , here, that's the conclusion that I can draw.  > >I have not bothered detailing the depths of your FUD attempt < >with respect to Sun and Java but as I keep saying you need  >to get more mirror time.e >t  ; I'm glad you brought up Java again.  Because your post hereo; where you said that the Java available for OpenVMS is Java i> "yes and no" is a classic example of FUD, which I countered.  > There's little doubt that for the best, most up-to-date, Java : almost everyone will tell you that Sun is the place to go.8 Using this fact against competitors is a pure, classical: FUD technique.  Some of what you said may be true, but the> intent is to drive customers from the OpenVMS/Java combination  and thus make it even more true.  ; So, are we to assume that Sun is interested in FUDding it's- Java partners? 1  < I also pointed out that my posts are not FUD at all, because> I'm not trying to scare Sun customers.  If I was interested in? that I would post elsewhere.  Another point that you are eithero< too dense to understand or are taking a pose to forward your agenda.e  > As I don't actually believe you to be an idiot, I suspect that> you have an agenda to FUD Compaq and OpenVMS.  If you convince> me that you don't have such an agenda, I may have to reexamine# my beliefs about your intelligence.e     >Regards >Andrew Harrison >Enterprise IT Architect   -Jordan Henderson  jordan@greenapple.comV   ------------------------------   Date: 2 Nov 2000 11:09:04 -0500a/ From: jordan@lisa.gemair.com (Jordan Henderson)n% Subject: Re: eBay (guess what) again.u* Message-ID: <8ts3j0$vm4$1@lisa.gemair.com>  * In article <3A01880E.69E2616F@oracle.com>,4 norm lastovica  <norman.lastovica@oracle.com> wrote: >andrew harrison wrote:i >	.n >	.k >	.tG >> My statement that Digital/Compaq were unable to show any performance A >> benefits of using a 64bit OS and VLM for DBMS's measured usingEI >> standard benchmarks such as TPC/SAP/PeoplSoft etc is TRUE, you may noteF >> like it but you won't get any support from anyone at Compaq to back3 >> you up because there is no support to available.N >O* >	Certainly, at the time, the world record1 >TPM/C number on the 8400 with Rdb was using VLM  0 >on VMS (well more than 10gb of the database in 2 >cache).  This was back around 1995.  You may have >been misinformed in the past.  A Oh, don't bother with facts that support Compaq/Digital.  They'lln@ have no affect on Mr. Harrison, at all.  He's only interested in= presenting facts, in isolation, that detract from Compaq and  ? OpenVMS.  Much like you'd expect of someone with a charter from  marketing would behave._   -Jordan Hendersonr jordan@greenapple.coms   ------------------------------  % Date: Thu, 02 Nov 2000 20:41:05 +0100 ( From: Bernd Eckstein <B.Eckstein@cli.de> Subject: Re: GAWK for VMS?& Message-ID: <3A01C351.55C09298@cli.de>   johnson@umtc.de wrote: > I > Can anybody tell me how to go about getting awk for a VMS machine which  > doesn't have a C compiler? > - http://www.montagar.com/freeware/GAWK-2_15_6/    --  E Outside of a dog, a book is man's best friend. Inside of a dog, it issE too dark to read.                                      (Groucho Marx)   ( Mit freundlichen Gruessen / Best regardsC B.Eckstein, CLI GmbH - mailto:B.Eckstein@cli.de - http://www.cli.detC Matthiashofstr. 28, D-52064 Aachen - Fon: +49 241 47051-0, Fax: -89i   ------------------------------  % Date: Thu, 02 Nov 2000 22:02:14 -0600w7 From: "David J. Dachtera" <djesys.nospam@earthlink.net>d Subject: Re: GAWK for VMS?- Message-ID: <3A0238C6.2F74331A@earthlink.net>d   johnson@umtc.de wrote: >  > Bernd wrote: > >C > >johnson@umtc.de wrote:  > >>L > >> Can anybody tell me how to go about getting awk for a VMS machine which > >> doesn't have a C compiler?  > >>0 > >http://www.montagar.com/freeware/GAWK-2_15_6/ >  > $ r gawk.exe0 > %DCL-W-ACTIMAGE, error activating image MTHRTLL > -CLI-E-IMGNAME, image file UMTC$DKA0:[SYS0.SYSCOMMON.][SYSLIB]UVMTHRTL.EXE< > -SYSTEM-F-SHRIDMISMAT, ident mismatch with shareable image > $c   Unfortunate, yes.h  F Hunter has been successful distributing ZIP, UNZIP and other things inH object form for both VAX and Alpha (LINK on the target system at install= time). Maybe that should be a standard practice for freeware.n -- r David J. Dachtera  dba DJE Systems5 http://www.djesys.com/  : Unofficial Affordable OpenVMS Home Page and Message Board: http://www.djesys.com/vms/soho/o  F This *IS* an OpenVMS-related newsgroup. 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.B   ------------------------------  " Date: Thu, 2 Nov 2000 21:28:36 GMT/ From: "Richard L. Dyson" <rick-dyson@uiowa.edu> + Subject: Re: How to boost DLT backup speed?d) Message-ID: <3A018824.3F26BE07@uiowa.edu>    Alan Greig wrote:t? > A TK89 can sustain over 30GB/hr under VMS. Really it can. Try!B > comparing your settings with those below. Especially the account > quotas > C > >       What basic Mount/Backup commands with qualifiers for high5F > >performance do others use for DLT tape drives off AlphaServer 1200, > >doing > >full /Image backups?d > = > For a fast backup to TK89 (change verify below if required)h > . > $ BK :== BACKUP/IMAGE/IGNORE=(INTER,LABEL) -& >   /BLOCK=65535/LAB=BACKUP/noVERIFY - >   /media=compaction -r  / 	My only change here is from 32256 up to 65535.   A > Also the account that run BACKUP should have something like the  > following: > ; > Maxjobs:         0  Fillm:       600  Bytlm:       128000 ; > Maxacctjobs:     0  Shrfillm:      0  Pbytlm:           0 ; > Maxdetach:       0  BIOlm:       300  JTquota:       4096y; > Prclm:          10  DIOlm:      4096  WSdef:       200000-; > Prio:            4  ASTlm:      4096  WSquo:       200000e; > Queprio:         0  TQElm:        40  WSextent:   1000000(; > CPU:        (none)  Enqlm:     32767  Pgflquo:    2000000$  < 	My backup account had lower settings than these.  I believeC I got them from suggestions in the various release notes and Backup2C sections of the manuals long ago.  I'll make some increases and runt tonight to see what happens.  G > Also could you post  a copy of the basic Accounting information for acD > job which logs in, run your backup procedure then logs out. Here's2 > what I see on an ES40 backing up well over 100GB > 9 >  SYSTEM       job terminated at  2-NOV-2000 08:21:01.28  >  >   Accounting information:iO >   Buffered I/O count:             734336      Peak working set size:   203616 O >   Direct I/O count:              4480978      Peak virtual size:       372576PO >   Page faults:                     16052      Mounted volumes:              2oQ >   Charged CPU time:        0 01:14:40.53      Elapsed time:       0 04:21:01.28u  F Here is my accounting from last night, with the old settings.  I have E made some changes to the process quotas and the blocksize and should   have a new result tomorrow.o  8   BACKUP       job terminated at  2-NOV-2000 05:44:32.92     Accounting information: E   Buffered I/O count:          666420         Peak working set size: n 19120pE   Direct I/O count:           1472846         Peak page file size:   m 60208'5   Page faults:                   3366         Mounted  volumes:            2sA   Charged CPU time:           0 00:26:25.84   Elapsed time:     0i 03:44:32.89   @ 	These results are from an AlphaServer 2100 OpenVMS v6.2 with anG old KZPAA controller and a TZ87 drive backing up about 24.816x10^6 blks-E (or ~11.83 GB).  I am also trying to tune an AlphaServer 1200/OpenVMSm/ v7.1-1H2 with a TK89 on a FWD SCSI controller. g   Regards, Rick -- eH Richard L. Dyson                                    rick-dyson@uiowa.eduH  _   _      _____                http://www-pi.physics.uiowa.edu/~dyson/H | | | |    |_   _|   Systems Analyst                     O: 319/335-1879H | | | | of   | |     The University of Iowa            FAX: 319/335-17536 | \_/ |     _| |_    Department of Physics & Astronomy-  \___/     |_____|   Iowa City, IA 52242-1479a   ------------------------------  # Date: Thu, 02 Nov 2000 14:51:46 GMTe% From: agreig@my-deja.com (Alan Greig)o+ Subject: Re: How to boost DLT backup speed? 2 Message-ID: <3a0175c1.2070863053@news.newsguy.com>  E On Thu, 02 Nov 2000 12:01:44 +0000, steven.reece@quintiles.com wrote::   >mK >But you've got the other issue with DLTs of "Can I keep them supplied witheL >enough data to keep writing continuously"?  This is allied to "my tape saysL >it can store 20GB but I'm only getting 15GB before it requests another" andI >"my backup took 3 hours before data compaction was applied but now takesh >6.5 hours".L >If the DLT doesn't get data quick enough to carry on writing it will, afterK >a specific period, stop, rewind to the end of data, and then start writingkK >again when it has something to write.  This leads to tape scrubbing and tooL >excessively long backup times as the tape scans backwards and forwards past >the tape heads.   Hi Steven, o  ? [Made if back from DECUS ok as you can see. How was day 2 whichDE unfortunately I couldn't stay for? Did Steve Hofman put on a suit forrD the defence types? Anyway most of the info below is really for those* following the thread rather than yourself]  E In my case I am backing up from dual HSZ80s with striped mirrorsets (r< most volumes three stripes dual mirror). On a decent (ie notD disastrously fragmented) filesystem I can easily read at a sustainedF 30MB/sec so keeping the 10MB/sec tape going isn't a problem with large> contiguous file. Additionally I have WSQUO and WSDEF at 200000@ pagelets so there's always 10secs worth of data (at 10MB/sec) inB BACKUP's buffers. This gives me enough headroom to keep the drivesF streaming even with directories containing thosuands of small files onE a fragmented volume. You need to stick CHANNLCNT way up (I have it at A 2047) as well as FILLM and the other process quotas as well whichr@ otherwise prevent BACKUP from performing well with lots of small files.  C All in all, I can usually keep the drives streaming even when doingeA online backups from user volumes in the middle of the day. That'ss, should I actually want to do this of course.  C Even with the help of Legato support our NT team can only get about F one third of the performance out of a TK89 under NT as I can under VMSD when backing up tens of thousands of small files - such as a typicalD network fileserver. Our PC background Compaq Account Manager told usE there must be a fault with the TK89 as it can't possibly be faster onr& VMS than on NT. I tried to explain....  B In case anyone points out the obvious - yes I know we could streamF under NT if we were willing to use ludicrous amounts of mirrorsets andD striping but it isn't really cost effective. Guess that's why NetApp boxes are becoming so popular.       --
 Alan Greig   ------------------------------  # Date: Fri, 03 Nov 2000 03:51:11 GMT * From: iwan tamimi <iwantam@singnet.com.sg>, Subject: How to decluster the cluster system) Message-ID: <8ttcnd$dt7$1@nnrp1.deja.com>    Hi All,o  C Currently we have Clustered OpenVMS system throuhh CI cosnsist of 2eA AS8400 and AS8200. Because the applications has been migratied to6= TRU64, we have to decomission the 2 systems from the cluster.o  G How the right way to do it? Make it 1 member cluster or just remove theu cluster? How to do it?   Thank you and Regards,   Iwan T.B    & Sent via Deja.com http://www.deja.com/ Before you buy.s   ------------------------------  % Date: Thu, 02 Nov 2000 19:15:04 -0000>- From: wspencer@ap.nospam.org (Warren Spencer)e/ Subject: Re: Is "errno" thread safe in OpenVMS?u/ Message-ID: <t03f9oec6rgb87@news.supernews.com>n  + R.Brodie@rl.ac.uk (Richard Brodie) wrote ino" <8trk5a$1bnm@newton.cc.rl.ac.uk>:    >b2 >"Mark" <mark@NOSPAMtechop.co.uk> wrote in message* >news:3a014a03.7286577@news.force9.net...  >oE >> We are porting a large application from UNIX to VMS.  The existingrB >> code makes heavy use of the "errno" variable.  Is this variable, >> thread safe when used with POSIX threads? >tD >To be pedantic, it's not a variable at all: it's a macro defining aF >function returning a thread local status. So yes, it should be thread >safe.   >   F It's my understanding that errno is set by rtl functions when a error F occurs.  In a multi-threaded process, an rtl call by one thread could J affect the results of a different thread inspecting that value - assuming I the timing conditions were "just right" (thread 1 makes a successful rtl dH call, thread 2 makes a unsuccessful rtl call [errno gets set], thread 1 J inspects errno and sees the error from thread 2).  I've never known errno  to be thread safe.  C But a far more important question is:  Why are you porting in this -J direction?  btw, good choice on the target os!  What does the application  do?e   ws   -- c3 << What if there were no hypothetical questions? >>$   ------------------------------  % Date: Thu, 02 Nov 2000 11:29:59 -0800w! From: Shane.F.Smith@Healthnet.comm/ Subject: Re: Is "errno" thread safe in OpenVMS?rD Message-ID: <OFC415DCAC.9FA7AC44-ON8825698B.006A2173@foundation.com>  J Yes. I don't have the page reference, but it is documented as thread safe.E I ran into the same thing myself a few months back and had to hit thetH manuals quite hard to answer the question. (I had to port a little commsH package to Weeniedoze, and found that errno is NOT thread safe there, so< being paranoid I had to check I hadn't blown it on VMS too).   $ SET MODE/RANT=ONK On Weeniedoze, you have to use a Weeniedoze-specific function call to get atH thread safe "last error" value. They suggest that if you want a reliableH errno, you #define errno to have the preprocessor substitute it with theF function call. Errno is useless in that context anyway, so why doesn'tJ errno automatically call the function when running threaded? Just one moreK way Micro$haft tries to make sure code is hard to port to another platform.tF And don't get me started on nonstandard berkely sockets, or malloc not reporting error conditions.... $ SET MODE/RANT=OFF    Shanet          8 mark@NOSPAMtechop.co.uk (Mark) on 11/02/2000 03:05:00 AM   To:   Info-VAX@Mvb.Saic.Coml cc:e  , Subject:  Is "errno" thread safe in OpenVMS?     Hi,s  B We are porting a large application from UNIX to VMS.  The existing? code makes heavy use of the "errno" variable.  Is this variablet) thread safe when used with POSIX threads?s   TIAt
 Mark Williamsb   ------------------------------  % Date: Thu, 02 Nov 2000 16:00:39 -0600 / From: Chris Scheers <chris@applied-synergy.com> / Subject: Re: Is "errno" thread safe in OpenVMS?mO Message-ID: <B30CCF6575F8BED5.E011D16523043452.635B4481AC8A6C1A@lp.airnews.net>y   Mark wrote:j >  > Hi,u > D > We are porting a large application from UNIX to VMS.  The existingA > code makes heavy use of the "errno" variable.  Is this variabled+ > thread safe when used with POSIX threads?s  F Generally, yes.  Take a look at ERRNO.H and see how errno is defined. < If it is defined by a routine that gets its address (usually. xxx$get_errno_addr), it should be thread safe.  > If it is defined as an extern variable, it is not thread safe.  D It depends on the version of the compiler/includes/libraries you are using.  G -----------------------------------------------------------------------t$ Chris Scheers, Applied Synergy, Inc.  C Voice: 817-237-3360            Internet: chris@applied-synergy.com l   Fax: 817-237-3074a   ------------------------------  % Date: Thu, 02 Nov 2000 22:59:32 -0500 " From: Dan Sugalski <dan@sidhe.org>/ Subject: Re: Is "errno" thread safe in OpenVMS?j; Message-ID: <5.0.0.25.0.20001102225901.02f1e7e0@24.8.96.48>y  & At 11:05 AM 11/2/00 +0000, Mark wrote: >Hi, >nC >We are porting a large application from UNIX to VMS.  The existingo@ >code makes heavy use of the "errno" variable.  Is this variable* >thread safe when used with POSIX threads?  I Errno is guaranteed to be thread-safe when used with POSIX threads. It's  , part of the standard. (Which is rather nice)   					Dan  I --------------------------------------"it's like this"------------------- 2 Dan Sugalski                          even samurai? dan@sidhe.org                         have teddy bears and event;                                       teddy bears get drunka   ------------------------------  % Date: Thu, 02 Nov 2000 19:54:47 -0800'% From: "E. Higa" <higasan@pacbell.net>u Subject: Locks+ Message-ID: <3A023707.DBEC5C66@pacbell.net>r  B I'm looking for somewhere that will give good detailed informationF regarding locks. running through monitor cluster the locks on one nodeA it the 15,000 mark at times.  i've pointed to a few programs i'vesG noticed but without a good explaination unfortnately no one will listeno. to me... any help or info would be appreciated   ------------------------------   Date: 2 Nov 2000 16:55:11 PSTmT From: Fairfield@SLAC.Stanford.EDU (Ken Fairfield; SLAC: 650-926-2924; FAX: 926-3515)# Subject: Re: OS Update Ramificationo3 Message-ID: <ESFDKL37aUUP@mccdev.slac.stanford.edu>   H In article <8tt0qc$4ae$1@nnrp1.deja.com>, cjdouglass@my-deja.com writes: [...].E > Is the 6.2 B saveset tape bootable? this would allow me to create akG > fresh installation, free from any unwanted or unrecognized files. ...o  H         Actually, sort-of.  :-) Do an /IMAGE restore of the B saveset toH     the target disk (which means  you  destroy  any  data you had on theH     target),  and  when  finished, copy the remaining  savesets  to  theH     [000000] directory on that disk.  Then boot the to  root  E  on  theH     target  disk  and  you'll  immediately  start  the  VMS installation     procedure.           -Ken -- nM  Kenneth H. Fairfield            |  Internet: Fairfield@SLC.Slac.Stanford.Edua:  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: Fri, 03 Nov 2000 04:12:33 GMTo5 From: njc@CC47532-A.ewndsr1.nj.home.com (Neil Cherry)n% Subject: Porting from Unix to OpenVMSq> Message-ID: <slrn904eqm.4dh.njc@CC47532-A.ewndsr1.nj.home.com>  @ I have some Unix daemons that I want to port over to Open VMS (aF MicroVAX BA123 w/KA-630 processor). Any pointers? The code is in C andB I will need to access the serial ports (I don't have any yet but I8 have access to terminal servers that support LAT & IP).    Thanks   -- /H Linux Home Automation           Neil Cherry             ncherry@home.netC http://members.home.net/ncherry                         (Text only)aB http://meltingpot.fortunecity.com/lightsey/52           (Graphics)0 http://linuxha.sourceforge.net/				(SourceForge)   ------------------------------  # Date: Thu, 02 Nov 2000 14:53:55 GMT0% From: agreig@my-deja.com (Alan Greig) ( Subject: Re: RECALL question/suggestion.2 Message-ID: <3a017fdb.2073448531@news.newsguy.com>  B On Thu, 02 Nov 2000 12:24:23 GMT, richard_maher@my-deja.com wrote:   >Hi, >wG >I hardly know PIPE at all and have been following recent examples withmD >great interest, but in your example is it a case of the sub-process/ >having it's own "empty" command recall buffer?  >kG >Sorry, I don't have the faintest idea of when a sub-proc is created oru1 >what || vs && does. Is it in the DCL dictionary?m   $ HELP PIPE PARAMETERi     --
 Alan Greig   ------------------------------  % Date: Thu, 02 Nov 2000 13:28:24 -0700 1 From: Vance Haemmerle <vance@toyvax.Tucson.AZ.US>o7 Subject: Research Systems Inc. dropping support for VMS.3 Message-ID: <3A016BF8.56B821D8@toyvax.Tucson.AZ.US>e  < Just received my IDL 5.4 cover letter that says that version? 5.4 is the last version to support OpenVMS.  That's certainly a>? kick in the pants to scientific computing on VMS.  RSI was justm8 bought by Kodak, maybe that had something to do with it.   -- Vance Haemmerlef vance@toyvax.Tucson.AZ.US    ------------------------------  % Date: Thu, 02 Nov 2000 22:30:05 -0800G From: prymmer/kahn <@best.com>; Subject: Re: Research Systems Inc. dropping support for VMSe( Message-ID: <3A025B6C.A062C6B7@best.com>  H I am almost positive that I've seen posts to mailing lists from folks at KodaktH in Rochester NY that were running VMS.  I suspect that the RSI folks are0 having a hard time finding VMS savvy developers.  H There may be a few "open Source" projects that might have an easier timeD being ported to VMS.  NumPy the numerical Python package and PDL theB Perl Data Language come to mind.  I am not sure if either was ever> eplicitly ported to VMS, but either one could be - at least inD principle.  Should you decide to undertake such a task just rememberH that you'd be doing what the folks who left RSI ultimately refused to do  (if that is of any consolation).   PP   Vance Haemmerle wrote:  > > Just received my IDL 5.4 cover letter that says that versionA > 5.4 is the last version to support OpenVMS.  That's certainly agA > kick in the pants to scientific computing on VMS.  RSI was just9: > bought by Kodak, maybe that had something to do with it. >p > -- > Vance Haemmerlet > vance@toyvax.Tucson.AZ.USc   ------------------------------  % Date: Thu, 02 Nov 2000 15:32:08 -0700e* From: Terry Aardema <taardema@nrcan.gc.ca>' Subject: SCP (Secure CoPy) for OpenVMS? ' Message-ID: <3A01EB68.2967@nrcan.gc.ca>   C Does anyone know of a SCP program for OpenVMS (specifically OpenVMSKH 7.2-1 Alpha, but don't be shy!)? This would be to allow Windows (bleck!)> clients to Secure-ly CoPy files to-and-fro via the SSH server.  D I know SCP is a UNIXism, but the alternative of FTPing through a SSHD client is almost to painful to bear! The user training involved, the/ problems about PASV and firewalls ... <shudder>0  H Of course, alternatives to SCP will also be cheerfully accepted (as long; as we can encrypt the username/password, I'll be happy) ...   
 advTHANKSance   
 Terry Aardema    ------------------------------  % Date: Thu, 02 Nov 2000 11:43:57 -0800 ! From: Shane.F.Smith@Healthnet.com0' Subject: Re: Sun "uptime" belly laugher7D Message-ID: <OFB34FF3AB.B7F4FB4E-ON8825698B.006B50BC@foundation.com>   Andrew, you wrote:= >I can only think that Kerry was confused and didn't actually2= >mean to suggest that logging people out every night improvesn
 >security.  J You're not thinking that through, Andrew. I have personally seen a cleanerI wandering around during her coffee break tapping peoples keyboards to see G if anyone had left themselves logged in and unsecured. (The next day we J told the cleaning contractors we never wanted to see her again.) EverybodyJ gets careless occasionally, so unless there's an automatic solution you're' going to be vulnerable sooner or later.u  J Granted, automatically logging out idle processes isn't the only solution,F but it is often a valid one. Come to think of it, I can't think of anyJ company I've worked at or with that didn't use it for at least some groups	 of users.i   Shaneu          D andrew harrison <andrew.nospam@uk.sun.com> on 11/02/2000 04:13:17 AM   To:   Info-VAX@Mvb.Saic.Comz cc:   ( Subject:  Re: Sun "uptime" belly laugher     Robert Deininger wrote:t >-- > In article <OmmmvxNi3fak@eisner.decus.org>,n: Kilgallen@eisner.decus.org.nospam (Larry Kilgallen) wrote: >s > >z > > >> This is essential forE > > >> accountability.  Otherwise you can't really tell who did what.f > > > I > > > Ummm.  I don't follow you.  Unless you only allow 1 user at a time,P how doH > > > enforced logouts improve accountability?  Is this kind of like the networkf/ > > > security goons who just unplug the wires?  > > G > > VMS has lots of auditing capability, but it is all tied back to who3K > > logged in.  If an Rdb record gets modified, or a VMS file gets deleted, 2 > > you can have a guaranteed record of who did it _PROVIDED_PEOPLE_LOG_OUT_. > K > Of course I lock the screen when I go home.  If someone gets my password,  heJ > can unlock the screen and pretend to be me.  But he could just as easilyE > log in a fresh session and pretend to be me, if he has my password.t >fG > I wonder how many of the folks Kerry would force to log out each day,y haveI > written their passwords in an obvious place on/in their desks.  Forcing-J > people to log out doesn't even scratch the surface of real security.  (I+ > realize I'm preaching to the choir here.)s >L< I can only think that Kerry was confused and didn't actually< mean to suggest that logging people out every night improves	 security.w  ? Logging people out every night and then changing their passworde@ of course would improve system security but it might not be that< popular with the users or the admins every morning when they tried to log in.  @ You can also log people out and then disable their account untilC office hours the next day. A security package called BOKS supportedo@ this functionality but even this can be very unpopular with both users and admins.t  < We use a combination of smart cards and userID/password this< defeats the person who finds out what my password is because they also need the card.   Regardss Andrew Harrisons Enterprise IT Architecta   ------------------------------  % Date: Thu, 02 Nov 2000 23:32:14 -0500c6 From: dunstan <dunstanSpamIsBestWhenFried@provide.net>' Subject: Re: Sun "uptime" belly laughero* Message-ID: <3A023FCE.85D1D0F@provide.net>   John Santos wrote:   >eI > All my customers ("Real World") feel the same way.  But they have usersfI > who are not programmers and who don't know the SYSTEM password, and who F > work at desks or in cubicals, not in the machine room or in lockableH > offices.  Sessions left logged in overnight are both a security and anJ > accountability risk.  Leaving a terminal or workstation un-attended even# > while going to lunch is an issue.a >1   As an interesting aside:  N Several years ago my boss had me write an Xwindows program that would forciblyR engage the screen saver after N minutes of inactivity, inactivity being defined asQ mouse not moving and nothing being typed in at the keyboard (the former was easy,tR the latter was downright ugly).  All of us with the system password tended to rushQ off and fight fires on occasion,  and we sometimes (inadvertently) left a session Q around logged in as SYSTEM.  Heck, even our own accounts had slightly higher thans normal privileges.  P Anyway, you could set the thing to kick in after N minutes during prime time, orT after M minutes during non-prime time (you could decide what time-of-day constitutedR prime time, too).  It handled multi-head displays, two-password accounts, etc.  It@ also ran as SYSTEM so an unprivileged user couldn't kill it off.  
 End of aside.h   ------------------------------   Date: 2 Nov 2000 10:11:02 -0700 1 From: nothome@spammers.are.scum (Malcolm Dunnett)i' Subject: Re: Sun Bloodies Compaq's NoseT, Message-ID: <oU+J6PndbXsU@malvm1.mala.bc.ca>  ; In article <szhM5.8433$Ho3.76126@typhoon.ne.mediaone.net>, G:     "Terry C. Shannon" <terryshannon@mediaone.net> writes:   > 8 > <fabio_compaq@ep-bc.petrobras.com.br> wrote in messageH > news:OF4BF37E05.A6348C18-ON8325698B.0058611F@ep-bc.petrobras.com.br...6 >> Well people... what can I say about  Sun x Compaq ? >>K >> I am OpenVMS addicted but the company I am working for  will convert the 
 >> largestG >> database of the company from the OpenVMS servers (about 23) to a fewn1 >> Enterprise 10000 running SAP, in January/2002.  >>L >> So ? Will be 450.000 - 23 Compaq Customers or 10.000.000 - 30.000 users ? >> >> What do you think ? > M > I think the Compaq sales rep in the account in question ought to be invited/N > to a "Come To Jesus" meeting to explain to senior Compaq management just how) > and why Compaq ceded an account to Sun.r >   L     I don't see VMS listed as a supported platform for SAP, so sticking withH VMS wouldn't appear to be an option. Perhaps instead of scapegoating theN sales rep someone should ask those higher up in Compaq management what they're doing to remedy that situation.a  K     Personally, I would applaud a customer who is forced to abandon VMS duepK to lack of software availability choosing a non-Compaq platform. The battler$ cry of any good VMS bigot should be:  '      "Give me VMS or give me Sun!!" :-)t  G    VMS's future is much safer if Compaq management firmly believes thatn2 every lost VMS customer is a lost Compaq customer.   ------------------------------   Date: 2 Nov 2000 20:19:08 +0100e* From: eplan@kapsch.net (Peter LANGSTOEGER)' Subject: Re: Sun Bloodies Compaq's Nosey( Message-ID: <3a01be2c@news.kapsch.co.at>  q In article <szhM5.8433$Ho3.76126@typhoon.ne.mediaone.net>, "Terry C. Shannon" <terryshannon@mediaone.net> writes:s >l7 ><fabio_compaq@ep-bc.petrobras.com.br> wrote in messagenG >news:OF4BF37E05.A6348C18-ON8325698B.0058611F@ep-bc.petrobras.com.br... 6 >> Well people... what can I say about  Sun x Compaq ? >>K >> I am OpenVMS addicted but the company I am working for  will convert theo
 >> largestG >> database of the company from the OpenVMS servers (about 23) to a fewf1 >> Enterprise 10000 running SAP, in January/2002.- >>L >> So ? Will be 450.000 - 23 Compaq Customers or 10.000.000 - 30.000 users ? >> >> What do you think ? >RL >I think the Compaq sales rep in the account in question ought to be invitedM >to a "Come To Jesus" meeting to explain to senior Compaq management just howi( >and why Compaq ceded an account to Sun.  3 And so must about 60% of DEQ's sales people, too...h   -- u< Peter "EPLAN" LANGSTOEGER           Tel.    +43 1 81111-2651; Network and OpenVMS system manager  Fax.    +43 1 81111-888r< <<< KAPSCH AG  Wagenseilgasse 1     E-mail  eplan@kapsch.netH A-1121 VIENNA  AUSTRIA              "I'm not a pessimist, I'm a realist"   ------------------------------  % Date: Thu, 02 Nov 2000 19:27:40 -0000e- From: wspencer@ap.nospam.org (Warren Spencer)n' Subject: Re: Sun Bloodies Compaq's Noses/ Message-ID: <t03g1c9avphv28@news.supernews.com>i  , fabio_compaq@ep-bc.petrobras.com.br wrote inB <OF4BF37E05.A6348C18-ON8325698B.0058611F@ep-bc.petrobras.com.br>:   4 >Well people... what can I say about  Sun x Compaq ? > I >I am OpenVMS addicted but the company I am working for  will convert the  >largestE >database of the company from the OpenVMS servers (about 23) to a fewh/ >Enterprise 10000 running SAP, in January/2002.  >0H >So ? Will be 450.000 - 23 Compaq Customers or 10.000.000 - 30.000 users >? " >r >What do you think ? > 	 >Regards,a >p >FC.  H I think you'd better buy large cosmic-ray umbrellas for your E10000's.  F Apparently they're required to prevent cache-related operating system K crashes.  There's been talk in this newsgroup and the trade press recently aJ about this so-called design oversight (no ecc), and Sun's attempted cover-L up (non-disclosure agreement with E*Trade).  Is your management prepared to I answer for spending big bucks on broken hardware?  Are they aware of the o> existing examples of failure in the industry of this hardware?   ws   ------------------------------  " Date: Thu, 2 Nov 2000 20:37:03 GMT- From: "Richard D. Piccard" <piccard@ohio.edu>P4 Subject: Re: Sun's haunting ecache problems detailed' Message-ID: <3A01D06E.6416163@ohio.edu>    Terry Kennedy wrote:  A > Brian Schenkenberger, VAXman- <system@sendspamhere.org> writes:eO > > But parity is usually only good for single bit errors.  Multiple bit errorsdN > > may go undetected and thus, you might be feeding the system with bad data. >-I >   Sure, but you have to draw the line somewhere. If you get a multi-biteH > error in cache with ECC, it's uncorrectable, and if is data that can'tG > be re-fetched from main memory (for example, if it was configured for-G > write-back and the writeback hasn't happened yet) you're dead anyway.@ >hL >   There is a whole lot of statistical modeling that goes on to decide whatN > level of error detection/correction needs to be provided, based on the specs5 > and presumed error rates of the various components.U            ^^^^^^^^^^^^^^^^^^d  N There is the problem:  SUN was simply too optimistic in their basic error rate estimation.   #                                 RDPh     >f >h6 >         Terry Kennedy             http://www.tmk.com7 >         terry@tmk.com             Jersey City, NJ USA-   --B ==================================================================B Dick Piccard                           Academic Technology ManagerB piccard@ohio.edu                                 Computer ServicesB http://oak.cats.ohiou.edu/~piccard/                Ohio University   ------------------------------  # Date: Thu, 02 Nov 2000 20:49:37 GMTu= From: system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-) , Subject: the lights are out at Northernlight0 Message-ID: <009F286C.F7B3FC00@SendSpamHere.ORG>  L I just fired up Netscape V3.03 on the old Alphastation to do a search.  I'veL been using Northernlight since they so proudly hail the use of OpenVMS Alpha6 on their main page.  Sadly, I find myself staring at:      vvv---  grammar???H ------------------------------------------------------------------------= //the writes the channel changer for northernlight.com pages.A   function write_chan_chgr() {  I document.write("<table border=0 cellpadding=0 cellspacing=0 width=85%>");nX document.write("<tr><td colspan=4><img src=t.gif border=0 width=1 height=1></td></tr>");h document.write("<tr><td colspan=3 bgcolor=#ffffff><img src=t.gif border=0 width=1 height=1></td></tr>");j document.write("<tr><td bgcolor=#ffffff width=1 height=1><img src=t.gif border=0 width=1 height=1></td>");' document.write("<td bgcolor=#ffffff>");rZ document.write("<table border=0 cellpadding=0 cellspacing=0 width=100% BGCOLOR=#ffffff>");^ document.write("<tr><td bgcolor=#ffffff><img src=t.gif border=0 width=1 height=6></td></tr>"); document.write("<tr>");d document.write("<TD align=center valign=middle BGCOLOR=#ffffff><a href=docs/gen_help_where.html><font face=arial,helvetica size=2>Help</font></a></td>");  document.write("</tr>"); document.write("<TR>");2 document.write("<TD align=center valign=middle BGCOLOR=#ffffff><a href=docs/member_accounts.html><font face=arial,helvetica size=2>Accounts</font></a></td>"); document.write("</tr>"); document.write("<TR>");. document.write("<TD align=center valign=middle BGCOLOR=#ffffff><a href=docs/about_company_mission.html><font face=arial,helvetica size=2>About</font></a></td>");  document.write("</tr>");_ document.write("<tr><td bgcolor=#ffffff><img src=t.gif border=0 width=1 height=10></td></tr>");e' document.write("<TR BGCOLOR=#ffffff>");  document.write("<TD align=center valign=middle BGCOLOR=#ffffff><a href=/cgi-bin/cl_cliplist.pl><font face=arial,helvetica size=2>Alerts</font></a></td>"); document.write("</tr>");' document.write("<TR BGCOLOR=#ffffff>");  document.write("<TD align=center valign=middle BGCOLOR=#ffffff><a href=http://finance.northernlight.com/portfolio.asp><font face=arial,helvetica size=2>Portfolio</font></a><BR></td>"); document.write("</tr>");^ document.write("<tr><td bgcolor=#ffffff><img src=t.gif border=0 width=1 height=6></td></tr>"); document.write("</table>");e document.write("</td>");f document.write("<td bgcolor=#ffffff width=1 height=1><img src=t.gif border=0 width=1 height=1></td>"); document.write("</tr>");h document.write("<tr><td colspan=3 bgcolor=#ffffff><img src=t.gif border=0 width=1 height=1></td></tr>"); document.write("</table>");e   }tH ------------------------------------------------------------------------  + So, I disable JavaScript and I try again.  o  L I then decided to leave a message about this issue to NorthernLight at theirL link /docs/gen_help_comments.html.  On this page is a form selection for the2 operating system in use.  Here are the selections:     Windows 95   Windows 98   Windows NT   Windows 2000
   Windows 3.1    Macintosh    UNIX  K Notice anything missing?  What does this say about NorthernLight?  They use K VMS but they don't expect anybody else to?  Seems there's a rather dim bulbp in that NorthernLight.  :(   --O VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COMa             O city, n., 1. a place where trees are cut down and streets are named after them.    ------------------------------  % Date: Thu, 02 Nov 2000 16:04:53 -0500 # From: Jim Agnew <agnew@hsc.vcu.edu>c0 Subject: Re: the lights are out at Northernlight+ Message-ID: <3A01D6F5.ECEF6D56@hsc.vcu.edu>>  K oops.. posted before i noticed the grist of what Brian was talking about.. o  & "Brian Schenkenberger, VAXman-" wrote: > rather dim bulbt > in that NorthernLight.  :( >  > --Q > VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COMh > Q > city, n., 1. a place where trees are cut down and streets are named after them.e   ------------------------------  $ Date: Thu, 2 Nov 2000 22:57:04 -0500  From: John Santos <JOHN@egh.com>0 Subject: RE: VAX Mail question - corrupted files5 Message-ID: <1001102225227.2093A-100000@Ives.egh.com>d  + On Thu, 2 Nov 2000, John Macallister wrote:y  M > >I don't know how the files are being corrupted.  That's what I'm trying to  > find > >out.  > L > What I was seeking from you was an example file segment showing one of theL > file's original and corrupted segments. Just one or two records would have/ > been enough to help direct any investigation.t > M > However, after seeing the details about sending a 38Kblock file I tried theoG > same here using nodes over which I have full control. I sent the mail2N > A->B->C where the distances A-B and B-C are 5000 miles and A and C are local > to each other. > M > I found that I couldn't, even after a few attempts, get the DECnet relayingp > to complete i.e. >  >  MAIL  from A to  B::C::user > F > would hang at some random point. I tried the same with SMTP relayingJ > involving the same nodes and that worked without any problem first time. > K > I've often transferred Gigabytes of data in one operation without troubleiN > using DECnet over the same links A-B amd B-C using both copy and BACKUP to aH > remote saveset directly node to node and not via relaying. The problem% > appears to be with the relay step. e > N > What happened with the DECnet mail relaying is that a lost packet caused theM > DECnet transfer to hang. My guess is that the precise DECnet behaviour willuN > vary with the DECnet version and so some versions may hang while others willL > hiccup and continue the transfer perhaps losing some data or even fail and< > restart, which I'd guess is what's happening in your case. > G > If you have software support I'd recommend that you report it. As thetL > problem is reproducible it should be straighforward to determine the cause > and obtain a fix (?).o >  > I hope this helps. >  > John  A Relaying was not an essential ingredient of the original problem.c  B The only reason for trying A:: -> B:: -> A:: was that the originalA poster had no access to system B:: so couldn't set up any kind of'B test environment there.  However, he did receive reports that mail? he sent from A:: to B:: was received by users on B:: corrupted.   > I think this problem of (apparently) dropped packets causing a= transfer to hang is a separate issue, and needs to be pursued  separately.e   --   John Santos  Evans Griffiths & Hart, Inc. 781-861-0670 ext 539   ------------------------------  % Date: Thu, 02 Nov 2000 15:57:17 -0600 / From: Chris Scheers <chris@applied-synergy.com>wE Subject: Re: VAXstation, console terminal, reboot, power-up, license?iO Message-ID: <4309AAAA347A0388.24E9773D7B4D774D.26375208AC287893@lp.airnews.net>l   "Zane H. Healy" wrote: > = > In comp.sys.dec Phillip Helbig <helbig@astro.rug.nl> wrote:mI > > I'm puzzled by the following behaviour.  First, when I switch off the K > > console terminal and switch it back on, I'm back at the >>> and have towJ > > tell it to boot.  Second, doing the same while a program (such as SHOWL > > CLUSTER, MONITOR etc) shows the power-up sequence (memory test etc) when% > > the terminal is switched back on.f > J > > First, is this behaviour expected?  Second, if so, can someone explain > > it?L > G > I've got the same thing with a VAXstation 3100 I've got.  This is themN > expected behaviour.  However, if you want to just access it via the net, andM > are only using the VT320 to boot it you can set it up to automatically bootoM > off of the hard drive, then don't connect a terminal to it.  It should thenyM > boot up and let you access it via the network.  This is what I've done withc
 > my 3100. > N > What I'm wondering is does this happen with the VAXstation 4000's, I know itI > doens't with a DEC 3000/300LX or a AlphaStation 200 4/233 (thankfully).d  B VAXstation 4000-90s also show the "break to console" behavior when turning off the terminal.   F I have the console port on my VAXstations wired to a DECserver 200.  IG can then connect to the console when I need to and, since the DECservereA is left on, I don't have a problem with the machines going to theo console.  H Most of the time, I just turn the machines on and let them boot.  I only& use the console if there is a problem.  G -----------------------------------------------------------------------s$ Chris Scheers, Applied Synergy, Inc.  C Voice: 817-237-3360            Internet: chris@applied-synergy.com s   Fax: 817-237-3074    ------------------------------  # Date: Fri, 03 Nov 2000 03:42:48 GMTp2 From: "Zane H. Healy" <healyzh@shell1.aracnet.com>( Subject: Re: VMS databases for hobbyist?7 Message-ID: <YuqM5.7643$Qz2.230011@typhoon.aracnet.com>   - alphaman-nix-spam@hsv.sungardtrust.com wrote:-D > Yes.  Check out Cache from Intersys -- you can get a free downloadG > along with a license (currently good through the end of the year) at:h  0 >  http://www.e-dbms.com/beta/distributions.html  E > It runs on a variety of platforms, but OpenVMS is its first love...o  D > (I've not tried it yet myself, but it does look promising, and its- > legacy in the OpenVMS world is well known.)v  J Well, based on this info I snagged it.  You've got to convert the save setI back into a valid file as it's been munged, in fact it got transferred aswI text.  I've gotten it installed and running, now I'm trying to figure out  what to do with it.t   			Zanet   ------------------------------  % Date: Thu, 02 Nov 2000 22:23:18 -0800o From: prymmer/kahn <@best.com>6 Subject: Re: VMS font configuration with Reflections-X( Message-ID: <3A0259D5.8F3E77AD@best.com>   Start with:       HELP FONT  N Where FONT is the DCL verb that invokes the program known as bdf2pcf on Unix XN implementations (or bdf2snf if you're on a VAX).  The fonts for VMS DECwindowsK are in a place like SYS$SYSROOT:[SYSFONTS...] or somesuch.  The HELP moduleuJ for FONT will tell you where they are located.  You can pick up BDF sourceM files for most X fonts from various X windows places, including lots of linuxo and BSD ftp sites.   PP   Jim Jennis wrote:    > Hi VMS Colleagues, >eJ > We run (at least) one each of just about every X-capable OS known to manE > here (Linux, Solaris, Tru64, HP-UX, OpenVMS), and I'm trying to getd> > Reflections-X set up to work transparently with all of them. > J > I have just gotten a brand new DS-10 with OpenVMS 7.2-1 and CDE (DWMOTIFM > 1.2-5) installed. Logging into the console via CDE everything seems to workbL > correctly, however when I use Reflections-X to telnet in or to start a CDE7 > login session, every time I start a Decterm I get....y >oG > cant find font -*-terminal-Bold-R-Normal-18-140-100-100-c-0-ISO8859-*k > M > Sounds like either my Reflections-X or my VMS fonts or logicals pointing to  > them are not set up properly.E >rI > I'd appreciate any tips on getting DW fonts properly configured for use3 > with Reflections-X.. >. > TIA! >0 > Jim?: > --------------------------------------------------------9 > FSC - Building Better Information Technology Solutions-a9 >       from the Production Floor to the Customer's Door.m: > --------------------------------------------------------7 > Jim Jennis, Technical Director for Commercial Systemsr  > Fuentez Systems Concepts, Inc. > 1 Discovery Place, Suite 2 > Martinsburg, WV. 25401 > USA  >e% > Phone: +001 (304) 263-0163 ext. 235p > Fax:   +001 (304) 263-0702& > Email: jjennis@discovery.fuentez.com >        jhjennis@shentel.netu( > WEB: http://www.discovery.fuentez.com/   ------------------------------  % Date: Thu, 02 Nov 2000 19:07:03 -0000b- From: wspencer@ap.nospam.org (Warren Spencer) 3 Subject: Re: Windows NT to OpenVMS access questionsf/ Message-ID: <t03eqnjcmb2ff5@news.supernews.com>e  & casinoop2@aol.com (CasinoOp2) wrote in0 <20001101184030.27040.00000163@ng-cs1.aol.com>:   H >I am trying to get some Windows NT PCs to talk to an Alpha server. FromF >the PC I can login to VMS using Reflection and can send/receive filesI >using Reflection FTP Neighborhood but I can't map a drive to the server.oA >Also when the PC boots, I get a message saying it can't find thelF >Pathworks licenses. When I click on Network Neighborhood, then EntireG >Network, then NetWare or Compatible Network, I get the message "UnablepF >to browse  the network. The network path was not found." I don't knowI >what path is being talked about or how to even find out. Can anyone sheds >some light on this for me?   
 -- snip --  I I've noticed that you need to wait 5 or 10 minutes after rebooting an NT >K box before it becomes aware of other nodes on the network.  This is due to >J a Microsoft networking component called the "Browser" which contains some A funky algorithm for keeping a list of known nodes on the network.p   ws   -- r3 << What if there were no hypothetical questions? >>n   ------------------------------   Date: 02 Nov 2000 19:26:54 GMT# From: casinoop2@aol.com (CasinoOp2)E3 Subject: Re: Windows NT to OpenVMS access questionsn: Message-ID: <20001102142654.12753.00000498@ng-fi1.aol.com>  ' >Do you have the appropriate licenses ?n  E Yes, the licenses in question are loaded (Show License PW* lists two,s" PWLMXXXCA7.02 and PWXXWINAT07.01).  < >Did you configure the Pathworks licensing agent on the pc ?  N I installed Pathworks and as best I can recall all went ok. But I don't recall3 a separate configuring step for the license server.e  @ >Do you have warnings or error in the pc's pathworks event log ?  O There are many entries in the Pathwork logs. The ones with a red X just say "NoqJ license servers are available to satify request". Double-clicking does not$ bring up anything that really helps.  J I don't know if this helps or not, but when I run the Pathworks ConnectionK Manager, I get an error "No TAPI device configured.". What is a TAPI device. anyhow?   N I have looked through the book OpenVMS and Windows NT Integration for Dummies,! but I can't locate any help here.m  
 Clark CalkinsE Schafer Corp ccalkins@schafercorp.com   ------------------------------   End of INFO-VAX 2000.614 ************************