1 INFO-VAX	Sat, 15 Apr 2006	Volume 2006 : Issue 209       Contents: ANN: HGFTP V3.2  Re: DSPP and OpenVMS media5 Re: Error codes from Distributed NetBeans for OpenVMS C Re: Getting device/directory info for an arbitrary user (F$GETUAI?) E Re: Is OpenVMS certified yet for 64-way Hard partitions (GS1280 M64). E Re: Is OpenVMS certified yet for 64-way Hard partitions (GS1280 M64). E Re: Is OpenVMS certified yet for 64-way Hard partitions (GS1280 M64). E RE: Is OpenVMS certified yet for 64-way Hard partitions (GS1280 M64). " Re: LWP::Simple crashes on VMSperl* Re: manually resetting file revision dates* Re: manually resetting file revision dates4 Re: set file empty (dfu set <file>/eblock=1/ebyte=0)  F ----------------------------------------------------------------------  % Date: Sat, 15 Apr 2006 12:21:57 -0500 - From: Hunter Goatley <goathunter@goatley.com>  Subject: ANN: HGFTP V3.2( Message-ID: <44412BB5.60409@goatley.com>  A HGFTP V3.2 is now available for download.  HGFTP is an FTP client E and server for OpenVMS VAX, OpenVMS Alpha, and OpenVMS IA64.  It runs : with MultiNet, TCPware, and TCP/IP Services on VAX, Alpha,7 and IA64, as well as CMU-IP on VAX.  It has a number of E features not supported by other VMS FTP clients and servers including > support for STRU O VMS (which TCP/IP Services lacks), multipleE anonymous FTP accounts, support for "ftp:" URLs, support for FTP site A aliases, and lots more.  It can be used instead of or in addition / to your existing stack's FTP client and server.   B This version adds support for the MDTM (MoDification TiMe) commandA in both the server and the client and the SET WINDOW_SIZE command = in the HGFTP client.  This version also upgrades the included A NETLIB component to V2.4 for Alpha and Itanium (NETLIB is forever F frozen at V2.3D for VAX; that version is included with HGFTP for VAX).  0 I hope to add ODS-5 support in the next release.   http://www.process.com/openvms/   5 ftp://ftp.process.com/vms-freeware/fileserv/hgftp.zip : http://vms.process.com/ftp/vms-freeware/fileserv/hgftp.zip   And on the other mirrors soon.   --     Hunter ------9 Hunter Goatley, Process Software, http://www.process.com/ B PreciseMail Anti-Spam Gateway for OpenVMS, Tru64, Solaris, & Linux9 goathunter@goatley.com     http://www.goatley.com/hunter/    ------------------------------  # Date: Sat, 15 Apr 2006 13:08:09 GMT " From:   VAXman-  @SendSpamHere.ORG# Subject: Re: DSPP and OpenVMS media 0 Message-ID: <00A543F4.940F168F@SendSpamHere.ORG>  _ In article <06041419522018_202002C3@antinode.org>, sms@antinode.org (Steven M. Schweda) writes:  >  > - >From: Hoff Hoffman <hoff-remove-this@hp.com>  >  >> Mark Schafer wrote: >>  M >> > If VMS Engineering provided o/s images, I would consider providing that  I >> > download also. (Hoff: call me if I'm missing out on an opportunity.)  >>  H >>    There have been informal engineering discussions around providing I >> block-based disk image copies of the distribution disks (eg: ISO, raw  K >> disk, or binary copies, depending on your preferred terminology), but I  K >> am presently aware of no particular demand for and accordingly no plans   >> to provide these images.  > E >   There may be no demand from the _paying_ customers, but I suspect H >that a lot of us parasites would be pleased to be able to suck down VMS5 >CD-ROM images.  You know, like the ones for Solaris:  > 2 >      http://www.sun.com/software/solaris/get.jsp > G >They seem to use Zip-compressed CD-ROM images (although in some cases, H >the compression doesn't buy much).  (The file names say "ISO", but it's >a lie.) > J >>    There are a couple of issues that would have to be resolved as part H >> of generating and providing this version of these images, of course. L >> (Nothing in technology is ever quite as easy as it looks (nor as easy as L >> the marketeers involved might lead you to believe; but I digress), and I I >> already know where a couple of the problem areas here will lurk.  And  H >> generating these images can involve more than a straight copy, in at  >> least one case.  :-)  > 3 >   I'm sure that the details would be interesting.  > I >>    But again, if this particular distribution mechanism is desired or  O >> desirable, then it is certainly something that should be formally requested.  > G >   Would adding "If you please, Sir" provide the kind of formality you  >seek?  J My only issue with this would be for the IA64 OpenVMS distro.  Presently, I this comes on a DVD.  What if one does not have a DVD burner to create an  OpenVMS IA64 DVD?  --  K VAXman- A Bored Certified VMS Kernel Mode Hacker   VAXman(at)TMESIS(dot)COM              5   "Well my son, life is like a beanstalk, isn't it?"     ------------------------------    Date: 15 Apr 2006 07:47:59 -0700( From: "Scott" <DocTrinsograce@gmail.com>> Subject: Re: Error codes from Distributed NetBeans for OpenVMSC Message-ID: <1145112479.455255.233900@j33g2000cwa.googlegroups.com>   E But... when you do that... the rest of us lurkers loose the advantage  of hearing the answer. :-(   ------------------------------    Date: 15 Apr 2006 08:11:04 -0700< From: "Hein RMS van den Heuvel" <heinvandenheuvel@gmail.com>L Subject: Re: Getting device/directory info for an arbitrary user (F$GETUAI?)A Message-ID: <1145113864.040168.3960@j33g2000cwa.googlegroups.com>    Agreed! C As long as there is no lexical F$GETUAI and the process has prives, 1 then when not just read SYSUAF directly from DCL? > Any tool will have to do just that, through SYS$GETUAI or not.F Elegant, too the poit, As simple as possible, but no simpler, and fast to boot   A The only tricky part, bypassing SYS$GETUAI could be if the fields  offsets were to change. A They haven't for the past decades, but I suppose it could happen.   A So just open sysuaf/SHARE=WRITE, READ/KEY the target, pick up the  fields. 
 Sample below.    Hein.   C $! uaf_defdevdir.com <username>    Hein van den Heuvel, April 2006. 6 $!libr/extr=$uafdef/out=uafdef.tmp sys$library:lib.mlb $!sea uafdef.tmp lgicmd  $!EQU    UAF$S_DEFDEV    32  $!EQU    UAF$T_DEFDEV    116 $!EQU    UAF$S_DEFDIR    64  $!EQU    UAF$T_DEFDIR    148? $define sysuaf sys$disk:[]sysuaf.dat  ! Local copy for testting ! $open /write/read/share=write uaf . 'f$parse("SYSUAF","SYS$SYSTEM:.DAT",,,"SYNTAX_ ONLY") $loop: $ read/end=done uaf rec 1 $ if p1.nes."" then read/end=done/key=&p1 uaf rec . $ defdev=f$extr(116+1,f$cvsi(116*8,8,rec),rec). $ defdir=f$extr(148+1,f$cvsi(148*8,8,rec),rec) $ username=f$extr(0,12,rec) 1 $ write sys$output username, ": ", defdev, defdir  $ if p1.eqs."" then goto loop  $done:
 $close uaf   ------------------------------    Date: 15 Apr 2006 01:38:17 -0700; From: "johnhreinhardt@yahoo.com" <johnhreinhardt@yahoo.com> N Subject: Re: Is OpenVMS certified yet for 64-way Hard partitions (GS1280 M64).B Message-ID: <1145090297.387090.88890@g10g2000cwb.googlegroups.com>   David J. Dachtera wrote: > Dave Froble wrote: > > L > > > But consider this: *if* the application is structured such that it canM > > > only run on one node at a time, it means that even if you are migrating C > > > from VMS-Alpha to VMS-IA64, you will configure the IA64 thing L > > > seperately, test it separately, and only once ready, you go "big bang"M > > > by extracting the databases from the Alpha system and loading them onto K > > > the IA64 system, hoping you get no new emergency patients during that  > > > period of downtime.  > > > > > I expect better from someone who is running a VMS cluster! > > I > > The storage is on a SAN!  It remains there!  VMS disks are VMS disks! J > > No different!  Now if it was being moved to AIX, then yes, a migrationD > > of some type for the data.  Possibly the AIX software isn't even: > > compatable with the VMS data structures within Oracle. > H > The tablespace files are Stream_LF, but the record lengths are well inH > excess of anything RMS could handle without a lot of fancy code on top > of it - like Oracle! > G > Seriously, the tablespace files themselves are probably not portable, " > but I could be wrong about that. >   C It depends on the version of Oracle.  Oracle has a feature that was D introduced in V8i called "transportable tablespaces" which basicallyG allow you to move a tablespace (consisting of some number of datafiles) G from one database to another as a kind of high(er) speed export/import. B  Until 10g you could only transport between databases of identicalG architecture - i.e. Alpha VMS to Alpha VMS or Sparc64 to Sparc64.  With E 10g you can "easily" transport between any architecture which has the E same endianess and with a run through a conversion utility inside the @ RMAN tool you can change the endianess of the data and transportE between disimilar architectures.  (See Metalink article # 243304.1 if ? you have access)  Now what this means is that you can migrate a B database to another architecture a lot easier then the traditionalF export/import but this does NOT enable simultaneous access to the sameE data by different architectures.   So you could clone your production C database to another architecture as a dry run but both could not be G live at the same time.  The Oracle RAC option which does allow multiple ? systems to access the same datafiles simultaneously is only for G (Oracle) clusters of identical architecture systems.  So you cannot add F an IA64 system to your Alpha cluster and slide it into your Oracle RACB database as some sort of rolling upgrade.  There will have to be a= downtime period for the conversion of the data either through = transportable tablespaces or the classic export/import cycle.    > > I'd expect that J > > an Oracle data type is an Oracle data type, but it could be worse than	 > > that.  >    It sometimes is.  G > I could see some possible issues re: floating point formats, if those 3 > are in use at all, not mention big/little endian.  >   F Yep. Endian is the big thing. Though 10g gets around that (see above). Also 32 vs. 64 bit systems.   C > > But in what's being discussed, adding an itanic to the cluster,  > > there is no data migration.  >   G Sorry, yes there is.  Even the transportable tablespace feature doesn't  get you around that.  D > Should not be, just as in VAX -> Alpha migrations, with noteworthy- > exceptions like datatype/endianness issues.  >   G Since both the VAX and Alpha are little endian I think it was the 32/64 G bit difference that did in the VAX-->Alpha datafile transfer.  The only D option was the export/import routine.  (HP VAX OpenVMS CertificationD Matrix, Article #52574.1) While not exactly "frought with danger andF mistakes" there are always those little gotcha's that pop up.  Ask any@ seasoned Oracle DBA.  Though things are much better now.   GoingF between Oracle versions is always interesting the first few times.  V6@ to V7 was fun when the integrity (no relation)  constraints thatD existed in V6 but weren't enforced were suddenly enforced in V7.  Oh& yes, very interesting days those were.  L > > > In such a scenario, moving from VMS to AIX isn't that different exceptN > > > that a few system managers need to learn AIX (but vast majority of users > > > won't see a difference)  > > K > > Bullshit!  What it it requires a dump from one database and a load into 7 > > another?  That is fraught with danger and mistakes.  > G > How so? I was under the impression that a database export would yield E > files that could be imported into, for example, a UN*X-based Oracle J > database, regardless of the source machine(s). Data and schemas exported@ > should be entirely portable, or am I totally off-base on that?  E Nope, you're right.  Oracle's Export tool creates a dump file that is B readable across all supported O/S by their Import tool.  It's evenE mostly backwards compatible so that the 10g Import can read back to a F V6 export.  The opposite is not true however ( not surprising though). :-)    > I > > > Also, remember that in any such move, it will also likely involve a 4 > > > change of version, so lots of testing involve. > > I > > For the itanic?  If Cerner certifies it, then most of that is already 	 > > done.  > ; > Brash assumption. That's why my partner and I have a job.  > D > > To AIX?  Massive amounts of testing data to insure that data was > > properly migrated. > > N > > > If the application is nice in a cluster and the database formats are theL > > > same and you can just take the raw database files and move them to theL > > > VMS-IA64 machine, then yeah, it would be much simpler to move from VMS. > > > to VMS. Somehow, I doubt it is the case. > > P > > Why would you doubt this?  It's VMS we're talking about, not some cheesy OS. >   3 But it's an Oracle database not an RMS file...  ;-)    ------------------------------   Date: 15 Apr 2006 12:54:50 GMT( From: bill@cs.uofs.edu (Bill Gunshannon)N Subject: Re: Is OpenVMS certified yet for 64-way Hard partitions (GS1280 M64).+ Message-ID: <4ac8oqFsd3jeU1@individual.net>   9 In article <wdednV7Yfe3dCN3ZnZ2dnUVZ_tednZ2d@libcom.com>, * 	Dave Froble <davef@tsoft-inc.com> writes: > I > Actually, I don't know Oracle, nor what it would actually take.  If we  C > haven't scared off most people, perhaps someone could speculate.  I > Without knowing exactly what Cerner is doing, it's hard to be precise.  J > But in going from VMS to AIX I'd expect there to be significant issues.   D I would expect little difficulty migrating at the application level.@ T real work is at the application developer level and Cerner has already done that.  G Here's a simple test.  Both PHP an MySQL run on VMS and Unix.  How hard G is it to take a pure PHP application that accesses a MySQL database and F move it from one system to the other?  How hard to just move the data-F base?  I suspect that from the standpoint of the end user, moving fromG VMS to AIX is no more difficult than moving from VMS-Alpha to VMS-IA64.    bill     --  J Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>       ------------------------------  % Date: Sat, 15 Apr 2006 09:15:06 -0400 ) From: "Ken Robinson" <kenrbnsn@gmail.com> N Subject: Re: Is OpenVMS certified yet for 64-way Hard partitions (GS1280 M64).H Message-ID: <7dd80f60604150615m26c297b3k71c3dea24c748df5@mail.gmail.com>  L On 15 Apr 2006 12:54:50 GMT, Bill Gunshannon <bill@cs.uofs.edu> wrote (in p= art): I > Here's a simple test.  Both PHP an MySQL run on VMS and Unix.  How hard I > is it to take a pure PHP application that accesses a MySQL database and H > move it from one system to the other?  How hard to just move the data- > base?   = This is not hard to do at all, as long as the versions of the D webserver (usually Apache), PHP, and MySQL match or are pretty closeE (or you know that you've scripted for the lowest common denominator).   . I have done all of the following combinations:A Written & Tested on Windows, moved to and run successfully on VMS B Written & Tested on Windows, moved to and run successfully on UNIX> Written & Tested on Unix, moved to and run successfully on VMSB Written & Tested on Unix, moved to and run successfully on Windows> Written & Tested on VMS, moved to and run successfully on UNIXA Written & Tested on VMS, moved to and run successfully on Windows   G That's for pure PHP/MySQL that doesn't use anything specific to the OS.    Ken    ------------------------------  % Date: Sat, 15 Apr 2006 10:22:31 -0400 ' From: "Main, Kerry" <Kerry.Main@hp.com> N Subject: RE: Is OpenVMS certified yet for 64-way Hard partitions (GS1280 M64).T Message-ID: <FA60F2C4B72A584DBFC6091F6A2B86840125B0E6@tayexc19.americas.cpqcorp.net>   > -----Original Message-----J > From: David J. Dachtera [mailto:djesys.nospam@NeOaSrPtAhMlNiOnWk.net]=20 > Sent: April 14, 2006 11:51 PM  > To: Info-VAX@Mvb.Saic.Com : > Subject: Re: Is OpenVMS certified yet for 64-way Hard=20 > partitions (GS1280 M64). >=20   [snip..]   >=20F > ...except that due to gov't certification requirements, and the timeD > required to acieve them, healthcare tends to be the furthest back. >=20  H So, the issue is not that Cerner is behind on OS versions, but rather isH handcuffed because of Govt regulations? That simply tells me that Cerner= is doing what it needs to do in terms of ensuring compliance.   > > > > > > Consider that the last sale dates for Alpha are just > > > around the corner , > > > > > and Cerner has yet to certify I64.	 > > > > >  > >=20E > > Again, do you think Cerner might need a DB being available on the ? > > platform before it certs its application or should it go=20  > ahead without  > > it?  >=20 > Say, "middleware". >=20  D Perhaps I do not understand the environment as well as you (I do notD work day-to-day in Cerner environment), but my understanding is thatD most of the Cerner servers are either big AIX or big OpenVMS serversE (with Citrix front ends as I recall), are you saying that they should E add new servers to the mix to run the middleware independently of the  back end servers?   @ [remember that network IO adds exponential delays to the overallE solution as compared to direct IO's and IO intensive app's can take a G major overall performance hit when slicing into many different "network  layers".  6 Remember the Govt certification effort on all of this.  D Again, you obviously work in this environment, so help me understand3 what you are positioning when you say "middleware".   C In general terms, I start to get nervous twitches when people start > talking about breaking up a host based, extremely intensive IOD application into multiple network layers where every network IO adds2 considerable latency into the overall solution.=20  F This is reason why SAP has been promoting return to middleware runningB on the same server as the DB server for IO intensive consolidation% initiatives. IT layers consolidation.    > > > > A > > > > I wonder if the availability of Oracle 10G on IA64 had=20  > any impact on 
 > > > > that?  > > > = > > > Arguable, but unlikely. Cerner's disenchantment with=20  > VMS/HP is fast# > > > becoming the stuff of legend.  > > > A > > > > And before anyone jumps on Oracle, keep in mind that they  > > > also had to 7 > > > > get the 10G Alpha release out the door as well.  > > > > B > > > > Here is a question for you - has Cerner cert'ed Oracle 10G > > > yet on anyA > > > > platform? What about RAC? [Perhaps they have, but I am=20  > interested to  > > > > confirm] > > > = > > > 9i-RAC, yes. We're puuting that in soon. 10g, doubtful.  > > >   A Not surprising, most Customers are still running Oracle 9i or 8i.    > >=20> > > So, if putting in Oracle 9i RAC, and assuming it scales=20 > even reasonably H > > well, is this not a better solution to your problem that putting all# > > your egg's in one large basket?  >=203 > No - see the prior discussion of licensing costs.  >=20  4 Ah, so the issue is not technical, but a cost issue?  D Lets not mix the arguments - is it a technical or cost issue you are concerned with?   A Whether you use 2 x 32CPU servers or 1 64CPU server, since Oracle B charges per CPU, the base Oracle costs should be the same. What isE different is that Oracle RAC adds something like 25-50% of additional D per cpu license cost, but then you also have the ability to shutdownF servers with zero impact on application availability, so there is someB additional benefits over a hot-cold (active-passive) server model.  	 [snip...]   A > > > > As others have stated, the 64CPU cert on OpenVMS Alpha=20  > can likely be G > > > > done, but it will require some firm business justifications and ' > > > > commitment from real Customers.  > > > < > > > How 'bout, "staying in business"? Is that considered a > > > "justification"? > > >  > > > -- > >=20A > > Absolutely. However, doing expensive testing when you have=20  > no CustomersA > > or ISV's willing to commit to a purchase is a great way to=20 
 > hasten that 
 > > point. >=20B > ...unless your competition is already doing it which case you=20 > either get& > with the program or GTFO of the way. >=20< > > Most vendors (not just HP) are well beyond the techie=20 > concept of "build D > > it and they will come" attitudes - that went out many years ago. >=20? > ...along with the platforms that were sold that way during=20  > their heyday. B > When the business/sales model changed and put the cart before=20 > the horse,A > the market evaporated, as we have seen - and continue to see=20  > - with VMS > and Alpha. >=20  E Ok, lets play devil's advocate for a moment. Lets say HP goes through E all of the effort to certify 64CPU's with no commitments from anyone. C Just ignores Cerner, Oracle and Govt's formal overall solution cert  requirements.=20  G HP eats the cert cost with the techie vision of "build it and they will 	 come ..".   E Then, continuing the what-if scenario, Cerner comes out and says "the D RAC model licensing costs have been ironed out and due to the higher@ availability model, we recommend our Customers not buy the 64CPUH active-passive model, but rather adopt the higher availability 2 x 32CPU server active-active model".  4 So, should HP just say "oh well .. And eat the cost?  G Lets be clear - no vendor is going to make a large certification effort G without some level of commitment from its target Customers or Partners. , That is simply not a good business practice.   Regards   
 Kerry Main Senior Consultant  HP Services Canada Voice: 613-592-4660  Fax: 613-591-4477  kerryDOTmainAThpDOTcom (remove the DOT's and AT)=20  4 OpenVMS - the secure, multi-site OS that just works.   ------------------------------  % Date: Sat, 15 Apr 2006 01:14:36 -0500 6 From: "Craig A. Berry" <craigberry@mac.com.spamfooler>+ Subject: Re: LWP::Simple crashes on VMSperl @ Message-ID: <craigberry-51B164.01143615042006@free.teranews.com>  @ In article <craigberry-0F69E8.07410312042006@free.teranews.com>,8  "Craig A. Berry" <craigberry@mac.com.spamfooler> wrote:  E > In article <1144705012.190392.170990@u72g2000cwu.googlegroups.com>, / >  "juna" <ggl.20.jjpon@spamgourmet.com> wrote:  > D > > It seems that my last post didn't go through. It went like this: > > 
 > > Craig, > > D > > The crash goes away when I deassign the logical PERL_ENV_TABLES.= > > I had it defined to LNM$PROCESS because the readme.vms in  > > DBD-Oracle1.16 told me to.E > > After a little more reading on the subject, I can keep DBD-Oracle ) > > happy and prevent the crash if I do a > > > define PERL_ENV_TABLES crtl_env,clisym_global,lnm$process. > > D > > Can you (or any other VMSperl expert) tell me if my fix is okay?* > > or am I going to break something else? > >  > > Regards, > >  > > Juna > > B > > PS. My apologies to the creator(s) of LWP::Simple. Your module+ > > wasn't the problem, just the messenger.  > F > Thanks for narrowing it down to the definition of PERL_ENV_TABLES.  E > I've now been able to reproduce the bug in the current development  0 > release of Perl by simply doing the following: > & > $ define perl_env_tables lnm$process+ > $ perl -e "print join qq/\n/, keys %ENV;" J > Fatal VMS error (status=100052) at D0:[CRAIG.perl]vms.c;1, line 1071 at  > -e line 1.- > %RMS-F-SYN, file specification syntax error  > J > So this should be fairly straightforward to fix.  In the meantime, your B > workaround sounds as good as any.  There is of course no way to H > guarantee that any particular definition of PERL_ENV_TABLES will work / > for every script or module you may encounter.   H I've just committed a patch for this issue, which turned out to just be F the removal of a wayward exclamation mark.  The patch is visible here:  < http://public.activestate.com/cgi-bin/perlbrowse?patch=27808  $ That should make it into Perl 5.8.9.E *** Posted via a free Usenet account from http://www.teranews.com ***    ------------------------------    Date: 15 Apr 2006 09:12:03 -0700< From: "Hein RMS van den Heuvel" <heinvandenheuvel@gmail.com>3 Subject: Re: manually resetting file revision dates C Message-ID: <1145117523.953470.210230@z34g2000cwc.googlegroups.com>   E That example is not Alpha compliant, and uses the more efficient, but E much more tricky to understand QIO call to modify file attributes. It @ also fails to test the return status on the time conversion... a  relatively likely troouble spot.  D Here is a variant, based on that example, which just uses RMS calls.E It works for me. Like the original it expects the filename to operate ( on point to by a logical name "filename"   Enjoy... Hein.    $ type rdt.mar         .psect data,wrt,noexe          $fabdef          $xabrdtdef  F infab:  $fab    fnm=<filename>,-                ;Filename is a logicalB                 fac=<get,put>,-                 ;File access modes                 xab=xabrdt   xabrdt: $xabrdt    time_buff:      .blkb    80  time_buff_desc: .long    80 #                 .address  time_buff   @ prompt: .ascid  /revision time-MONTH MUST BE IN CAPS-dd-MMM-yyyy
 hh:mm:ss /           .psect  code,nowrt,exe         .entry  begin,^M<> ;  ; open the file  ;          $open   fab=infab          blbc    r0, done ; , ; move desired date into revision date field ;          PUSHAL          PROMPT&         PUSHAL          TIME_BUFF_DESC*         CALLS           #2,G^LIB$GET_INPUT  1         $BINTIM_S       TIMBUF = TIME_BUFF_DESC,- 2                         TIMADR = xabrdt+xab$l_rdt0         blbc r0, done  ; 7 ;  close the file, modify revision date on the deaccess  ;          $close  fab=infab 8 done:   ret             ; Return with final status in r0         .end begin   ------------------------------    Date: 15 Apr 2006 10:24:05 -0700< From: "Hein RMS van den Heuvel" <heinvandenheuvel@gmail.com>3 Subject: Re: manually resetting file revision dates C Message-ID: <1145121845.761824.178620@z34g2000cwc.googlegroups.com>   D I already had RDT examples in Fortran, C and Pascal, but for grins I just made one in Cobol. 9 It is a litte rough in that it hardcode's RMS structures. G Notably a manually layout of an XABRDT instead of some library include, 2 and a hard coded 36 for FAB$L_XAB (from $FABDEF in SYS$LIBRARY:STARLET.MLB)E And, for the example, both filename and new data/time are hard coded. & Still, it should show the principle...   Hein     IDENTIFICATION DIVISION. PROGRAM-ID. rdt_test.  ENVIRONMENT DIVISION.  INPUT-OUTPUT SECTION. 
 FILE-CONTROL.      SELECT TEST_FILE)         ASSIGN               TO TEST_NAME ,         FILE STATUS          IS FILE_STATUS.   DATA DIVISION.
 FILE SECTION. 
 FD TEST_FILE.  01 TEST_RECORD.      03  SOME-DATA PIC X(80).   WORKING-STORAGE SECTION.  5 01  FILE_STATUS                 PIC X(02) VALUE "  ". ( 01  FAB_PT                      POINTER.( 01  XABRDT_PT                   POINTER.( 01  FAB_XAB                     POINTER.   01 XABRDT_REC.2   05 XABRDT_BLN_COD     PIC S9(9) COMP VALUE 5150.!   05 NXT        USAGE IS POINTER.    05 RVN        PIC 9(9) COMP.   05 RDT        PIC 9(18) COMP.   : 01 NEW_TIME     PIC X(23) VALUE "01-APR-2001 01:23:45.67".   PROCEDURE DIVISION.    MY_MAIN SECTION. MAIN.        OPEN I-O  TEST_FILE.     IF FILE_STATUS = "00" THEN0        CALL "DCOB$RMS_CURRENT_FAB" GIVING FAB_PT     ELSEC        DISPLAY "ERROR: TEST FILE OPEN ERROR, STATUS: ", FILE_STATUS         STOP RUN      END-IF.   E     CALL "SYS$BINTIM" USING BY DESCRIPTOR NEW_TIME, BY REFERENCE RDT.   $     ADD 36 TO FAB_PT GIVING FAB_XAB.-     SET XABRDT_PT TO REFERENCE OF XABRDT_REC. &     CALL "OTS$MOVE3" USING BY VALUE 4,3                             BY REFERENCE XABRDT_PT, -                             BY VALUE FAB_XAB.        CLOSE TEST_FILE.
     STOP RUN.    ------------------------------    Date: 15 Apr 2006 07:41:07 -0700< From: "Hein RMS van den Heuvel" <heinvandenheuvel@gmail.com>= Subject: Re: set file empty (dfu set <file>/eblock=1/ebyte=0) B Message-ID: <1145112067.934434.40400@e56g2000cwe.googlegroups.com>   Dave Froble wrote:I > Why would you want to truncate an existing file and re-write it instead 4 > of just deleting the file and creating a new file?  ) You'd loose the file-id, and performance.   G Both methods will of course have to find space and write out an updated  file header.@ However, file creation also involves finding a free file header,G marking it as used and finding an empty space in the directory possibly ( shuffling up the directory to make room,1 The preceding delete adds all of that in reverse.   6 Yes, caches will help, but why go through the motions?  B > It cannot be to re-use the existing allocated storage, since the$ > truncate must release the storage.  D Right, but if oly the EOF is changed, such as now suggested even theF allocated blocks will be re-used. btw.. More easily so than the deleteG case this allows you to 'undo' a recent recent eof change and may offer - a peak into past data for debugging purposes.    Hein.    ------------------------------   End of INFO-VAX 2006.209 ************************                                                                                                                                                                                                                                                                                                                                                                                                                ⊣$Qr+Z)r-!\
<[.j
sș%mFBc?۲52f)Z=EvY#?6m^6k@e*_F!A.uI8
f rb=֣
j(j,ZC3d]5tv/^i{BY$MfNB򾡖w	/t<\k:bqn6
.-9 ={Oe$1!j@*" F8dcEcp<Av-h=Eȝi\2􄨈Gӷ,*
Į7^|r_HnX
3uJDeՇy3+S4	l=9t;S	rl+`!OX9&Qb1n8_X*]fsSV6iOz- l3E(E2:aGU3:!ÊӚ32x?g19Wk@Qnyr〩SHߖ_.f
%W۴<8 4ȗt|C0?=-`ɅRwţ噯GU<ˠ'%ՔL+ȢCnjfizШr6 {<o<y8i6ƑGvaJ>shiRq
0!|>GSţ֦1hb#] RX>}"$s0PzѲ	ՖáR[_e9zBәxIjH(Q~gYV~9/ b)7Ef=pX;UKз
ܣ q,dg8g1-7-51*ƻ TZ˥&I]Y&=|w6[Q;:7?Pa(&a,3}CLs/6J-rL<9r"qZ`w0	hЫFaI;*j89m
dhKSt
-9QcYNl<AY
M\PȑXp\s$!2#g

2g[f6b2EfG^d7oq1W8<8in K5MmBeuJgdg:}_7ip)ۍoCd{0 `wصwݡ %ҫJ5&RRCq%'CCEgxm92?$[e8EWñ[@jP'x9C#19Cy|((l7n3RH&y)qDρQВG 2|G ҃:n!ܾ)F^L$:P	 &]\_<"~1MPV!!QL"X\oPݍg.d$|h{@}`N^f=Ρ	1	eTTߢ1#T <<C*[&?O IӖ 'My?Τ蔲GO+qϞT?dWE@_ "SP!x"^#h;*8iE5ɛoa\&lP$
/ijAvZ
]@oA33\
_5B[ӨYE.8**	u"*٣ *clEnW8o
bG@x`ch$ۜm4QHvJj\baLskdߵ׆8Kwm0$t,DX0Il?]-Ӱ( cL43;3czjtWfWήkk2b;3#'"j?hAX
KhRm6`/0@?$Q_&L"dq#3gw_TfF{{pkGh/h9^.k\U,
n.Y?݊.ңzWO@@rpT̽%<6!m\/dΧd";	Rl^]5j%:-~ɍXe^JepEEW躒|Svԝ3'Q/O`nEbCd[.ʟoEa$ldj,,rQc-գŬeE]:;x
ti
TZ;]%^1Io# 0)	# z9:2,WE<
A4GNۊ1t."Tç;m:H5YVSoPnCR[[+,[?VVO([j9"*kX?)cXh(m
A:O.PciZ)pµY͊9.JhMA<`w&LuQu'r˧{wlYPŖh<h6HVȯt}YqCzWPXj-yCWNbKE4OYX@PϛS[ZeQ*CZ-w:/'oEj0?tWu+e&A l
*uЖܤ zn)=;tN*%3@
<F.nKrfsj$}q7j!5$}OeoVX*\ڳLsuy2bz6N<
jqO{C7WJ	C46M	CWWЁ~b )Z~V	TїUhEU;zH#Ϛ&oYiJ-ܗ\Vx楬oӇu*/Jaܒ#fYBD|Xk8 oW,l5R^ÇZ:sgA/+DL~NB#-XABBYiV8x{U*̛Tv/{|X.7i9:kk~d=Ղ+<N."xӥqv}T"n`00׫hU.rѨT`,[~O{=~:4
"-1^+%2sZ??Vif