1 INFO-VAX	Thu, 07 Aug 2003	Volume 2003 : Issue 433       Contents:) Re: Computer Systems & Automation Manager  Re: Corrupted Indexed RMS file Re: Corrupted Indexed RMS file Re: Corrupted Indexed RMS file& Fast_Skip w/ SCSI tape drives question* Re: Gaining control of build dependency...* Re: Gaining control of build dependency... GBE as Cluster Interconnect  Re: GBE as Cluster Interconnect 0 How many errors are too many when making CD-R's?4 Re: How many errors are too many when making CD-R's?+ Re: Idiot (I know that's redundant) spammer + Re: Idiot (I know that's redundant) spammer + Re: Idiot (I know that's redundant) spammer + Re: Idiot (I know that's redundant) spammer ( Re: Licence manager and class scheduling( RE: Licence manager and class scheduling2 Re: Migrate obsolete VAX/VMS SW to OpenVMS Itanium Re: Netware is no VMS  Re: Netware is no VMS  Re: Netware is no VMS  Re: Netware is no VMS  Re: Netware is no VMS  Re: Netware is no VMS  Re: PDP-11 OS Release Dates  Re: PDP-11 OS Release Dates  Re: PDP-11 OS Release Dates  Re: PDP-11 OS Release Dates  Re: PDP-11 OS Release Dates 3 Re: pls help: install OS on DS10L without console ? 6 Re: Server consolidation & VMS - "Back to the Future"!3 Re: Sun tries to woo AlphaServer users away from HP 3 Re: Sun tries to woo AlphaServer users away from HP 
 WASD and RDB?  Re: WASD and RDB?  Re: WASD and RDB? J Re: [OT] Re stopped clocks being right twice a day (was Re: HP FUDBusting)J Re: [OT] Re stopped clocks being right twice a day (was Re: HP FUDBusting)  F ----------------------------------------------------------------------  % Date: Thu, 07 Aug 2003 05:15:09 +0800 , From: Paul Repacholi <prep@prep.synonet.com>2 Subject: Re: Computer Systems & Automation Manager- Message-ID: <87k79qcwte.fsf@prep.synonet.com>   = koehler@eisner.nospam.encompasserve.org (Bob Koehler) writes:   e > In article <1ca82fc6.0308050008.292ed469@posting.google.com>, dooleys@snowy.net.au (dooley) writes:  >> Phil D >> http://jobs.careerone.com.au/search/dsp_show_job.cfm?AD_ID=229537  ( >   So how is Port Kembla in the spring?  D Spring? Kembla does not have spring... Standard is rain/drizle/stormA wit coal dust. At least they are closing down the smelter, so the / copper content of everything should drop some.     The steel works is it?   --  < Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda. @                                              West Australia 6076* comp.os.vms,- The Older, Grumpier Slashdot. Raw, Cooked or Well-done, it's all half baked.F EPIC, The Architecture of the future, always has been, always will be.   ------------------------------  % Date: Wed, 06 Aug 2003 21:39:43 -0400 $ From: Hein <hein_cov@eps.zk.dec.com>' Subject: Re: Corrupted Indexed RMS file . Message-ID: <3F31ADDF.CF7EB388@eps.zk.dec.com>    You received good advice so far.  J You did not indicate clearly what kind of help you are looking for though.  H Are you still trying to recover data, or did you since restore a backup.2 Was the backup clean or did it have the loop also?  2 >         Creation Date:   25-JUL-2002 04:54:07.46A >         Revision Date:   23-JUL-2003 23:50:51.05, Number: 30188 2 >         Backup Date:     26-JUL-2003 13:45:23.23  B Hmmm.... revision date before creation date. Somewhat odd. A copy?I That revision count is high. This is after how many days? You may need to J teach your application to keep files open longer to make it easier for the system. I This is no excuse for corruption though, more a performance matter, and a  'chances' changer.  # >         File Monitoring: disabled   >         Global Buffer Count: 0  F Monitoring and global buffers are often goodness. Again, no excuse for corruption though     A > ***  VBN 1848100:  The bucket chains for key #0 contain a loop. 7 > Unrecoverable error encountered in structure of file.   @ I have some tools in the VMS FREEWARE inteh RMS_TOOLS directory.G They may help identifying buckets and key values around the corruption. F Of course this is only interesting if you are still trying to recorver
 more data.J The is also a tunign presentation there, which also give patching/recovery hints. Check it out.    G > We checked also disk for errors (analyze/disk/read) and error log for F > some errors that could be related to this file corruption. We didn't > find any problems.  
 Good move.  F > (VMS721_RMS-V0500) that mentions RMS loop during attempted duplicate9 > key insert for compressed keys - this might be related.    Not related.  H > Could this corruption be caused by application crash? This seems to beG > quite serious corruption to the file to be caused by mere application  > crash   I No. Not likely. RMS buffers are in exec mode, out of reach of application  code running amok.J The structural integrety is pretty safe, from appliacation misbehaving. Of course application couldI ask RMS to delete the wrong records, make teh wrong updates, but the file  structure shouldI be safe. Unless... the appliaction requested DEFERRED WRITE. In that case  a user processJ coudl be sitting on a critical buffer. Concurrent access, causing flushes,E tends to minimize potential damage, but I suppose there are scenarios * where an update for an entire bucket might not happen.   = > (fortunately it's test machine - but we would like to avoid % > similar problems in production...).   A Nothing you can do about it from application level, other then an  occasional explicit = call to $FLUSH if DFW is selected. Check hardware evern more.   J > Our system is OpenVMS version 7.2-1, machine is Alpha Server 2100 5/300.  / Probably with SWXCR (KZPSC) storage controller? E Verify scsi cable lengths are witing specs and terminators are in the  right place.  
 Good luck, Hein.    ------------------------------  $ Date: Wed, 6 Aug 2003 23:48:22 -0400  From: John Santos <JOHN@egh.com>' Subject: Re: Corrupted Indexed RMS file 5 Message-ID: <1030806234430.1951A-100000@Ives.egh.com>    On Wed, 6 Aug 2003, Hein wrote:   " > You received good advice so far. > L > You did not indicate clearly what kind of help you are looking for though. > J > Are you still trying to recover data, or did you since restore a backup.4 > Was the backup clean or did it have the loop also? > 4 > >         Creation Date:   25-JUL-2002 04:54:07.46C > >         Revision Date:   23-JUL-2003 23:50:51.05, Number: 30188 4 > >         Backup Date:     26-JUL-2003 13:45:23.23 > D > Hmmm.... revision date before creation date. Somewhat odd. A copy?  F Minor point: creation was a year ago.  Revision was about 2 weeks ago.< (I don't know how long ago the original message was posted.)  K > That revision count is high. This is after how many days? You may need to   * About a year, meaning about 80 revs a day.  L > teach your application to keep files open longer to make it easier for the	 > system. K > This is no excuse for corruption though, more a performance matter, and a  > 'chances' changer.   [...]    --   John Santos  Evans Griffiths & Hart, Inc. 781-861-0670 ext 539   ------------------------------  % Date: Thu, 07 Aug 2003 00:34:15 -0400 $ From: Hein <hein_cov@eps.zk.dec.com>' Subject: Re: Corrupted Indexed RMS file . Message-ID: <3F31D6C7.B335701A@eps.zk.dec.com>   John Santos wrote:  ! > On Wed, 6 Aug 2003, Hein wrote:  > 6 > > >         Creation Date:   25-JUL-2002 04:54:07.46E > > >         Revision Date:   23-JUL-2003 23:50:51.05, Number: 30188  > > F > > Hmmm.... revision date before creation date. Somewhat odd. A copy? > H > Minor point: creation was a year ago.  Revision was about 2 weeks ago.  9 Ooops... Don't know how I missed that. Indeed a year ago. M So now may comment would become: How come this file was not re-converted in a  year! C This is no excuse for corruption though, more a performance matter. H For your production system, I woudl hope you have a more regular convert schedule in mind.    :-).   Thanks for the correction. Hein.    ------------------------------  # Date: Thu, 07 Aug 2003 02:16:30 GMT + From: Jeff Cameron <JCam90502@jcameron.com> / Subject: Fast_Skip w/ SCSI tape drives question 2 Message-ID: <BB57048D.A728%JCam90502@jcameron.com>  L > This message is in MIME format. Since your mail reader does not understand< this format, some or all of this message may not be legible.   --B_3143042190_19843835 . Content-type: text/plain; charset="ISO-8859-1" Content-transfer-encoding: 8bit    To any who can help,  B In versions of VMS 7.2-2 and earlier the SCSI magnetic tape driverI SYS$MKDRIVER.EXE had code such that if an error or mount verification had D occurred on a tape device that supported fast skip, the driver wouldL permanently disable fast skip for that device, regardless of the setting setG by the $SET MAGTAPE/FAST_SKIP=option command. The only way to re-enable K fast_skip was to reboot the system. Now, Compaq/HP has provided a patch for # this problem which works just fine.   I But my question is this. In version 7.2-2 and earlier versions,one could L determine if fast skip for a tape device was disabled by doing the following :    $ANALYZE/SYSTEM  SDA>SHOW DEVICE <tape-device>  SDA>EXAMINE UCB+15C   L This would result in a hexadecimal long word. Looking at bit 12 of this wordL (identified by this mask %X'00001000'), if the bit is set, then fast_skip isG enabled, and if not set then fast_skip is permanently disabled. However J sometime between 7.2-2 and 7.3-1 this method of test seems to be no longerF valid, even though the native 7.3-1 SCSI magtape driver still disables* fast_skip on errors or mount verification.  L The patch kit dec-axpvms-vms731_fibre_scsi-v0300 does indeed fix the problem on VMS 7.3-1 (see : L http://ftp.support.compaq.com/patches/public/Readmes/vms/dec-axpvms-vms731_f ibre_scsi-v0300--4.README)  F Can anyone who has access to the source for the 7.3-1 SYS$MKDRIVER.EXEH provide a new test method to determine if fast skip has been disabled or not?   Thank you in advance.    Jeff Cameron     --B_3143042190_19843835 - Content-type: text/html; charset="ISO-8859-1" + Content-transfer-encoding: quoted-printable    <HTML> <HEAD>5 <TITLE>Fast_Skip w/ SCSI tape drives question</TITLE>  </HEAD>  <BODY>3 <FONT FACE=3D"Courier New">To any who can help,<BR>  <BR>L In versions of VMS 7.2-2 and earlier the SCSI magnetic tape driver SYS$MKDR=M IVER.EXE had code such that if an error or mount verification had occurred o= M n a tape device that supported fast skip, the driver would permanently disab= M le fast skip for that device, regardless of the setting set by the $SET MAGT= O APE/FAST_SKIP=3Doption command. The only way to re-enable fast_skip was to reb= M oot the system. Now, Compaq/HP has provided a patch for this problem which w=  orks just fine.<BR>  =A0<BR> N But my question is this. In version 7.2-2 and earlier versions,=A0one could d=M etermine if fast skip for a tape device was disabled by doing the following =  :<BR>  =A0<BR>  $ANALYZE/SYSTEM<BR> * SDA&gt;SHOW DEVICE &lt;tape-device&gt;<BR> SDA&gt;EXAMINE UCB+15C<BR> =A0<BR> L This would result in a hexadecimal long word. Looking at bit 12 of this wor=M d (identified by this mask %X'00001000'), if the bit is set, then fast_skip = M is enabled, and if not set then fast_skip is permanently disabled. However s= M ometime between 7.2-2 and 7.3-1 this method of test seems to be no longer va= M lid, even though the native 7.3-1 SCSI magtape driver still disables fast_sk= ( ip on errors or mount verification. <BR> =A0<BR> L The patch kit dec-axpvms-vms731_fibre_scsi-v0300 does indeed fix the proble=M m on VMS 7.3-1 (see : http://ftp.support.compaq.com/patches/public/Readmes/v= 4 ms/dec-axpvms-vms731_fibre_scsi-v0300--4.README)<BR> =A0<BR> L Can anyone who has access to the source for the 7.3-1 SYS$MKDRIVER.EXE prov=M ide a new test method to determine if fast skip has been disabled or not?<BR=  >  =A0<BR>  Thank you in advance.<BR>  <BR> Jeff Cameron<BR> </FONT>  </BODY>  </HTML>      --B_3143042190_19843835--    ------------------------------  * Date: Wed, 6 Aug 2003 18:02:18 +0000 (UTC)- From: lewis@spyder.mitre.org (Keith A. Lewis) 3 Subject: Re: Gaining control of build dependency... . Message-ID: <bgrfr9$rlt$2@newslocal.mitre.org>   brendan@scanoptics.com.au (Brendan Tregear) writes in article <4ea25e04.0308051640.153088e4@posting.google.com> dated 5 Aug 2003 17:40:08 -0700:E >The problem is this: I need a way to implement in MMS something such B >that my utility com file is run each time BEFORE the source filesB >dependency to the local include file is determined EVERY time the- >system is built. I have tried the following:  > - >my_source_file DEPENDS_ON local_include_file   H The source file doesn't depend on the include file, does it?  The objectG file depends on both the source file and the include file.  This may be  where your problem is.  2 >local_include_file DEPENDS_ON global_include_file >    RUN UTILITY COM FILE  > C >Now, every time the global include file is updated, my utility com G >file will run. This works fine, except there is now a build dependency 7 >that means the source file gets rebuilt every time the G >global_include_file changes, which is what i don't want! What I really  >want is this: > = >IF global_include_file is newer than local_include_file THEN  >   RUN UTILITY COM FILE >ENDIF  3 That's what the above MMS rule is supposed to do.     + --Keith Lewis              klewis$mitre.org > The above may not (yet) represent the opinions of my employer.   ------------------------------  $ Date: Wed, 6 Aug 2003 16:11:47 -0400  From: John Santos <JOHN@egh.com>3 Subject: Re: Gaining control of build dependency... 5 Message-ID: <1030806155546.1951A-100000@Ives.egh.com>   ! On 6 Aug 2003, Bob Koehler wrote:   s > In article <4ea25e04.0308051640.153088e4@posting.google.com>, brendan@scanoptics.com.au (Brendan Tregear) writes:  > ) > > Sorry for the long post... Any ideas?  > - >    Can't you do this with nested MMS files?  >  >    descrip.mms: 5 >   local_include_file DEPENDS_ON global_include_file  >      RUN UTILITY COM FILE # >      MMS/DESCRIPTION=descrip2.mms  >  >    descrip2.mms:0 >   my_source_file DEPENDS_ON local_include_file  @ I'm not an MMS expert by any means, but I think it boils down toG the question: Does MMS assume that if there is a chain of dependencies, B it needs to rebuild everything, or does it re-evaluate the need to rebuild at each step?   @ In other words, your utility that creates the local_include_file@ could build it in a scratch area, compare the new version to the@ existing version, and only replace the existing one with the newB one if they are materially different (where the utility determines  which differences are material.)  ? However, if MMS still insists on rebuilding the app because the A local_include_file must have changed since it is dependent on the ? global include file, then it will still rebuild everything.  On = the other hand, if MMS re-evaluates at each step, it will say ? "okay, local depends on global, global is newer, rebuild local. ? Next, app depends on local, local is older than app, no need to > rebuild."  You still need to run your utility each time, so we< are assuming this is much faster then actually compiling and linking.  > Another possibility, does MMS allow you to define a callout toE determine "older"?  Or is it just based on file creation/modification > dates?  If it allows you to define a method to determine which= is older, you could use your utility to create a test version @ of local_include_file from global_include_file, compare the test= version to the current version, and if identical, then return  the appropriate status to MMS.   HTH.     --   John Santos  Evans Griffiths & Hart, Inc. 781-861-0670 ext 539   ------------------------------   Date: 6 Aug 2003 22:07:02 +0200 2 From: "Lars Holmstrm" <lars.holmstrom@flysta.net>$ Subject: GBE as Cluster Interconnect( Message-ID: <3f315fe3$1@news.wineasy.se>  H Can any one share some experience about using GigaBit Ethernet/FC as the cluster interconnect. I In particular which GBE switches to use and which one to avoid. Also some I performance differences between GBE switches that may affect the Cluster.   J I would also like to learn something about recoverytime for a cluster withE larger separation (10-50 km) running GBE/FC as ISL. In particular for 1 SCS-failover to GBE when in a recovery situation.   @ Last I may replace FDDI in a cluster with 3 identical sites withL Gigaswitches and a fourth Quorum site (there is a particular reason for thisG setup). The quorum node has 3 FDDI adapters, one for each of the remote I sites. Does any one has some experience to share, first how to migrate to L GBE while keeping the services up, and second what to consider from a design perspective.   /Lars    ------------------------------   Date: 6 Aug 2003 19:47:23 -0700 1 From: keithparris_NOSPAM@yahoo.com (Keith Parris) ( Subject: Re: GBE as Cluster Interconnect< Message-ID: <cf15391e.0308061847.73725ff@posting.google.com>  b "Lars Holmstrm" <lars.holmstrom@flysta.net> wrote in message news:<3f315fe3$1@news.wineasy.se>...J > Can any one share some experience about using GigaBit Ethernet/FC as the > cluster interconnect.   D Lots of customers these days are moving from CI- (HSJ controller) orD SCSI-based (HSZ controller) storage to Fibre Channel storage, with a= SAN as the Storage Interconnect and Gigabit Ethernet (or Fast A Ethernet) as the Cluster Interconnect instead of CI (or DSSI), so  you're not alone.   A When I helped build E*Trade's 2nd disaster-tolerant cluster a few D years ago, we put in one rail of FDDI (GIGAswitch/FDDI) and one railE of GbE (Cisco 6509s) in parallel, as GbE was fairly new at the time.  ; A quick scan of my (woefully-imcomplete) directory of known D disaster-tolerant VMS cluster sites shows at least 10 VMS DT cluster6 sites known to be using GbE as a cluster interconnect.  B Note that Fibre Channel doesn't yet carry inter-node cluster (SCS)D traffic; it carries only SCSI-3 protocol for storage at this point. E So it can only be a Storage Interconnect at this point, not a Cluster F Interconnect (and so also can't do both at the same time as did CI and DSSI).  E Also, at this point (7.3-1) all interrupts for LAN adapters go to the @ Primary CPU in an SMP system, so if you're using CI now with SMP: systems with Fast_Path enabled, you could risk Primary CPU@ interrupt-state saturation if you switched over to using GigabitC Ethernet or another LAN as the cluster interconnect (or switched to F Memory Channel, for that matter, as it also lacks Fast_Path support). E Fast_Path support for PEDRIVER and LANs is slated for 7.3-2, which is  in Field Test now.  C Verell Boaen of VMS Engineering has done tests of different cluster B interconnects and their latency, bandwidth, and host CPU overhead,? with the results presented most recently at HP-ETS in St. Louis A (there's an older version from DFW Days on the http://vmsone.com/ E website).  It would be worthwhile to get a copy of his presentation.  C (I can send you one via e-mail; just remove _NOSPAM from my address F above when you request it.)  Verell will be presenting this at the VMSC Advanced Technical Bootcamp in November also, if you can arrange to  attend that.  @ There are two Gigabit Ethernet host adapters out now, the older,C slower (for lock requests), but proven DEGPA, and the newer, faster + (for lock requests), but less-mature DEGXA.   E Do you do MSCP-serving across FDDI today? and have NISCS_MAX_PKTSZ at B 4474 to take advantage of large packets?  If so, look carefully atE Jumbo Packet support on GbE, and be careful, as some vendors say they C support Jumbo Packets if they support ANYTHING over 1498 bytes, not . necessarily the full size the standard allows.  D And if you are doing MSCP-serving now, consider that the addition ofD inter-site Fibre Channel links would allow direct access to storage,F relegating MSCP-serving to a backup path in the event of a FC failure.  D But if you can only have a limited number of inter-site links due toE cost (like two between a pair of sites), it may be better to have two D dual-purpose SCS links than to have one SCS link and one FC link andD either be unable to continue use of both of those sites (if the soleD SCS link fails) or to have to fall back to MSCP-serving over the SCS! link (if the sole FC link fails).   K > In particular which GBE switches to use and which one to avoid. Also some K > performance differences between GBE switches that may affect the Cluster.   A My preference is the switches from Digital Network Products Group 9 (http://dnpg.com/).  If you've had good results with your D GIGAswitch/FDDI boxes, consider dnpg's Gigabit Ethernet boxes, built  with the same design philosophy.  D I worry about switch OS's that are designed with IP as their primary@ purpose in life, rather than bridging -- with IP, dropping small= multicast packets can actually be helpful in resource-starved E conditions (like ARP broadcast storms).  PEDRIVER's Hello packets are D small multicast packets, and if a lot of them get dropped, nodes can) leave the cluster with CLUEXIT bugchecks.   B Newer Ciscos seem to work fine, although I would avoid the low-endF (e.g. Catalyst 4000 series) boxes (which have collision domains across@ multiple ports) if you can and stick to the ones with a separateB collision domain per port (e.g. 6500 series), if you want to avoid congestion.   F One way to measure a switch's performance as a cluster interconnect isB to time actual lock requests through that switch.  You can use theF LOCKTIME tool originally written by Roy G. Davis, author of VAXclusterB Principles -- see http://encompasserve.org/~parris/locktime*.*  ByB enabling and disabling different LAN adapters using SCACP, you canC measure the performance of each interconnect separately.  Test your C existing infrastructure first, and don't accept new technology that E can't do better than that, especially under heavy load.  I found some A pretty good results for GbE on ES45s recently -- 200 microseconds = latency through a Cisco switch, or 140 microseconds through a ? cross-over cable, which is very close to the 120 microseconds I  measured for Memory Channel 2.  L > I would also like to learn something about recoverytime for a cluster withG > larger separation (10-50 km) running GBE/FC as ISL. In particular for 3 > SCS-failover to GBE when in a recovery situation.   D As noted above, FC cannot serve as a backup SCS link yet (but that's$ in the Roadmap as a future feature).  B So I'm confused about the question, and I'm thinking that the onlyF failover possible today would be if the Fibre Channel link failed, andF the storage traffic failed over to MSCP (which can happen provided youE are running OpenVMS version 7.3-1 and above, which has the ability to 3 fail over between direct FC and MSCP-served paths).   > If you have a single extended LAN between sites, Spanning TreeD reconfiguration times can be an issue that requires running with theA RECNXINTERVAL parameter set to a value higher than the default 20 D seconds.  With multiple, completely separate LANs, you could use the default value.  ? Or by "recovery" did you mean to talk about shadow full-copy or @ full-merge times?  With Fibre Channel storage, you will get fullB merges instead of the mini-merges you may be used to with CI-based@ (HSJ) storage (until the successful completion of the host-basedE mini-merge project which is underway in VMS Engineering at present).  = I did a presentation on shadow copies and merges at HP-EUW in 1 Amsterdam -- see http://www2.openvms.org/kparris/   B > Last I may replace FDDI in a cluster with 3 identical sites withN > Gigaswitches and a fourth Quorum site (there is a particular reason for thisI > setup). The quorum node has 3 FDDI adapters, one for each of the remote K > sites. Does any one has some experience to share, first how to migrate to N > GBE while keeping the services up, and second what to consider from a design > perspective.  D This cluster sounds intriguing, and I'd love to learn more about it.  D Since PEDRIVER can handle multiple LAN paths, one common approach inF migrating between interconnects is to leave the existing LAN in place,F and install the new interconnect as a second LAN, with LAN adapters ofD both types in the systems during the transition.  The most risk-freeA approach here would be to initially lower the priority of the new A interconnect so the cluster continues to use the old interconnect 9 while you evaluate the operation and stability of the new F interconnect.  Next, turn the new interconnect on in parallel with theB old, at the same priority.  Later, lower the priority of the olderF interconnect.  Finally, remove the old interconnect.  Allow sufficient; testing time at each stage before you progress to the next.   B If a particular node doesn't have enough PCI slots to double-up onE adapters (say, perhaps your quorum node here), see if you can add one @ adapter and transition one link at a time to the new technology, following the strategy above.    ------------------------------   Date: 6 Aug 2003 15:14:41 -0700 . From: spamsink2001@yahoo.com (Alan E. Feldman)9 Subject: How many errors are too many when making CD-R's? = Message-ID: <b096a4ee.0308061414.4951f26d@posting.google.com>    Hello,  B I have been asked to archive data to CDs. I used to have access toB CD-Rs that came in standard CD jewel cases. I used to get anywhereD from zero to only a few errors with the individually packaged CD-R's* that each come with its own CD jewel case.  B Now I've been given "stack" CD-R's and I've been getting many moreE errors. I recently got 195 recoverable errors on a single disc. These , errors are produced during my "verify pass".  C My Method: I prepare an LD logical disk. I copy the logical disk to @ the CD-R on a CD-writer using CD_RECORD. Since the disk is never@ MOUNT-ed, I get, and expect, no errors during this part. Then, I1 unload the disc and load it into a CD-ROM device:   A Disk $1$DKA100: (NODEX), device type RRD43, is online, allocated, 
 deallocate on C     dismount, mounted, software write-locked, file-oriented device, 
 shareable,     error logging is enabled.   F MOUNT it, and do a BACKUP/IMAGE/COMPARE between that and the LD drive.@ BACKUP finds no errors, but the error count on the RRD43 was 195F errors on a recent disc. They appear to be all recoverable errors, butD is it so many that some time in the future, due to scratches or what! not, I may lose some of the data?   5 Would I be justified in asking for the more expensive 0 individually-packaged CD-R's that I used to use?  / Thanks. A typical error message follows my sig.    Alan E. Feldman      /  ******************************* ENTRY     836.  ******************************* F  ERROR SEQUENCE 20677.                           LOGGED ON:        SID 12000003F  DATE/TIME  5-AUG-2003 21:59:54.86                            SYS_TYPE 04010001   SYSTEM UPTIME: 45 DAYS 06:15:10F  SCS NODE: NODEX                                               VAX/VMS V6.2   8  DEVICE ERROR  KA47  CPU FW REV# 3.  CONSOLE FW REV# 0.1   D  RRD43 SUB-SYSTEM, UNIT _NODEX$DKA100:, CURRENT LABEL "FT_CORP_DSK2"           HW REVISION     34383031 9                                        HW REVISION = 1084         ERROR TYPE            05 C                                        EXTENDED SENSE DATA RECEIVED         SCSI ID               01 3                                        SCSI ID = 1.         SCSI LUN              00 4                                        SCSI LUN = 0.        SCSI SUBLUN           00 7                                        SCSI SUBLUN = 0.         PORT STATUS     00000001 ?                                        %SYSTEM-S-NORMAL, NORMAL 
 SUCCESSFUL2                                         COMPLETION        SCSI CMD        8B7D1108                             0040 +                                        READ         SCSI STATUS           02 6                                        CHECK CONDITION     EXTENDED SENSE DATA           EXTENDED SENSE  000100F0                         0AC87D11                         00000000                         00000018                             0000 6                                        RECOVERED ERRORE                                        RECOV DATA WITH ERR CORRECTION  APPLD         UCB$B_ERTCNT          B9 =                                        185. RETRIES REMAINING         UCB$B_ERTMAX          01 ;                                        1. RETRIES ALLOWABLE         ORB$L_OWNER     00010004 :                                        OWNER UIC [001,004]        UCB$L_CHAR      1ECD4008 ;                                        DIRECTORY STRUCTURED 4                                        FILE ORIENTED/                                        SHARABLE 0                                        AVAILABLE.                                        MOUNTED4                                        ERROR LOGGING0                                        ALLOCATED<                                        SOFTWARE "WRITE LOCK"7                                        CAPABLE OF INPUT 8                                        CAPABLE OF OUTPUT4                                        RANDOM ACCESS      F  V A X / V M S        SYSTEM ERROR REPORT         COMPILED  6-AUG-2003 15:15:37E                                                                       	 PAGE 390.            UCB$W_STS           0000         UCB$L_OPCNT     00022710 >                                        141072. QIO'S THIS UNIT        UCB$W_ERRCNT        010E <                                        270. ERRORS THIS UNIT        IRP$W_BCNT          8000 C                                        TRANSFER SIZE 32768. BYTE(S)         IRP$W_BOFF          0110 <                                        272. BYTE PAGE OFFSET        IRP$L_PID       00C00018 6                                        REQUESTOR "PID"        IRP$Q_IOSB      00000001 C                        9C010000        IOSB, 0. BYTE(S) TRANSFERRED    ------------------------------  $ Date: Wed, 6 Aug 2003 19:45:20 -0400  From: John Santos <JOHN@egh.com>= Subject: Re: How many errors are too many when making CD-R's? 5 Message-ID: <1030806193646.1951B-100000@Ives.egh.com>   % On 6 Aug 2003, Alan E. Feldman wrote:    > Hello, > D > I have been asked to archive data to CDs. I used to have access toD > CD-Rs that came in standard CD jewel cases. I used to get anywhereF > from zero to only a few errors with the individually packaged CD-R's, > that each come with its own CD jewel case. > D > Now I've been given "stack" CD-R's and I've been getting many moreG > errors. I recently got 195 recoverable errors on a single disc. These . > errors are produced during my "verify pass". > E > My Method: I prepare an LD logical disk. I copy the logical disk to B > the CD-R on a CD-writer using CD_RECORD. Since the disk is neverB > MOUNT-ed, I get, and expect, no errors during this part. Then, I3 > unload the disc and load it into a CD-ROM device:   F CD-R's tend to be labeled with a maximum write speed.  Most CD-writingE programs can be adjusted to write at various speeds (don't know about F CD_RECORD, though, as I've never used it.)  It may be that the cheaperB CD-R's that come in a stack have a slower max write speed than the8 more expensive ones that come in individual jewel cases.  2 Bottom line, try a lower write speed, if possible.  D As far as long-term viability is concerned, my guess is you can readB the disks okay now because the drive is applying the ECC code, butE if it gets a scratch, and thus multiple bit errors in the same block,  ECC may not suffice any more.   B If the data on the disk is backup save-sets, and you used /CRC and? a reasonable /GROUP size, BACKUP may still be able to read them ; when most other things can't, but if you lose the directory @ structure or INDEXF.SYS, you may be out of luck.  It sounds like@ you haven't reached that point yet, but after a few scratches...   --   John Santos  Evans Griffiths & Hart, Inc. 781-861-0670 ext 539   ------------------------------  % Date: Wed, 06 Aug 2003 12:13:43 -0700 0 From: Mark Berryman <Mark.Berryman@Mvb.Saic.Com>4 Subject: Re: Idiot (I know that's redundant) spammer' Message-ID: <3f30f0f8$1@cpns1.saic.com>    Bradford J. Hamilton wrote: \ > In article <3f2fed1e$1@cpns1.saic.com>, Mark Berryman <Mark.Berryman@Mvb.Saic.Com> writes: > !Bradford J. Hamilton wrote: > !snip!T > !> (Of course, not every MGM/PHB *really* feels this way - I know of an owner of aQ > !> small business who thinks that Word is a POS, and laments the loss of WP for  > !> VMS.)	:-) > !tL > !Beg pardon?  Why would he lament a loss that doesn't exist?  WP is still J > !available for VMS with both character-cell and X-Window support.  I've = > !got a recently purchased copy running my system right now.w > !a > 
 > Hi Mark, > O > After my original post, I Google'd some links which led me to believe that WPtN > was still available - it may be, but AFAIK, not from Corel.  Is it a currentM > version of WP, or is it WP 5.1?  What kind of support is available for it? c9 > Please post a link or two that I might be able to view.s  F It is available from Legacy2000 (http://www.legacy-2000.com) and both H 5.1 and 7.1 are shipped with the same order.  It is a supported product C but, since I haven't needed support, I have no knowledge as to the - quality of support.   
 Mark Berrymane   ------------------------------   Date: 6 Aug 2003 16:01 CDT' From: carl@gerg.tamu.edu (Carl Perkins) 4 Subject: Re: Idiot (I know that's redundant) spammer, Message-ID: <6AUG200316014486@gerg.tamu.edu>  5 "David J. Dachtera" <djesys.nospam@fsi.net> writes... C }My question is this: in WhineBloze Exploder, when a directory with G }sub.'s is highlighted, is there a keyboard shortcut to expand the listg3 }of subs without having to mouse and click the "+"?R }  }--  }David J. Dachtera  F Up and down arrows navigate up and down the list in its current state.   Right arrow expands a "+".  B Left arrow contracts the sub elements of the current item, if any, to create a "+".   --- Carl   ------------------------------   Date: 6 Aug 2003 16:44 CDT' From: carl@gerg.tamu.edu (Carl Perkins)i4 Subject: Re: Idiot (I know that's redundant) spammer, Message-ID: <6AUG200316442814@gerg.tamu.edu>  5 "David J. Dachtera" <djesys.nospam@fsi.net> writes...  }Peter Weaver wrote: }> c }> David J. Dachtera wrote:o }> >...?8 }> > My question is this: in WhineBloze Exploder, when a }> directory withe; }> > sub.'s is highlighted, is there a keyboard shortcut tod }> expand the list7 }> > of subs without having to mouse and click the "+"?v }> i7 }> The Right Arrow does that, here are a few more keys;  }> R }> Press                 To A }> END                   Display the bottom of the active window. > }> HOME                  Display the top of the active window.9 }> NUM LOCK+ASTERISK     Display all subfolders under thet }> selected folder.- }>   on numeric keypad (*)= }> NUM LOCK+PLUS SIGN    Display the contents of the selectedi
 }> folder. }>   on numeric keypad (+)6 }> NUM LOCK+MINUS SIGN   Collapse the selected folder. }>   on numeric keypad (-); }> LEFT ARROW            Collapse current selection if it's.% }> expanded, or select parent folder.d: }> RIGHT ARROW           Display current selection if it's( }> collapsed, or select first subfolder. } ( }Excellent!! *MUCH* faster than mousing! } B }....and this is documented ...where? I'd like to spread the word. }  }--  }David J. Dachtera  A Somewhere in the help is a list of a bunch of keyboard shortcuts.3B I think these, or at least some of them, are on it. I have no ideaF where to find the list. I have seen it several times, but it is almostB impossible to find on purpose. As I recall it is a sub page off ofE some accessibility related page, and it is not easy to find (or maybe B not possible to find) via the index or even the search function ofC the help. This is on Windows NT V4.0 - they may have made it easier 0 to find since then (but I wouldn't count on it).  C The list also seems to be included in 3rd party books on how to useiG the thing. It's probably posted all over the place on the web as well - $ there is one on the MS website here:  > http://support.microsoft.com/support/kb/articles/q126/4/49.asp  @ (That's a simplified URL, actually - the one it ends up using is strangely convoluted.)  ? This set is under "Windows Explorer Tree Control" on that page.r  I The page states that it is for the various versions from 95 to Millenium,pG but most (or possibly all) of them work with the NT through XP set too..  H You can gets lists for specific MS products, including operating systemsG such as Windows 2000 or XP but (strangely) not plain old NT, from here:t  D http://www.microsoft.com/enable/products/keyboard/keyboardsearch.asp  E The on-line help for Windows is sometimes/frequently not helpful, but-D there is a huge volume of info on the web at microsoft.com and other places, too.   --- Carl   ------------------------------   Date: 06 Aug 2003 21:52:09 GMT( From: ka2doug@cs.commoc.sc (DL Phillips)4 Subject: Re: Idiot (I know that's redundant) spammer> Message-ID: <20030806175209.03007.00001186@mb-m01.news.cs.com>   >r( >Excellent!! *MUCH* faster than mousing! > A >...and this is documented ...where? I'd like to spread the word.   < Umm, Start-> Help -> Keyboard Shortcuts and pick your topic.  K This has turned into an elightening discussion. Seems to be a lot of people O here with a very vocal dislike of winblowz but very little knowlege on which to K base such dislike. I dislike windows but I have enough experience with thisd crap to know why.u   ------------------------------  $ Date: Wed, 6 Aug 2003 13:46:55 -0400+ From: "Syltrem" <syltremzulu@videotron.com>c1 Subject: Re: Licence manager and class schedulingi4 Message-ID: <wabYa.2966$Gf3.13423@tor-nn1.netcom.ca>   Very good idea.   H But it's so hard to change the way people do things. And also that wouldB mean for a software vendor to spend time and money to enable classD scheduling etc, and the result would be that they will probably sell, licences for cheaper. Not worth it for them.   Money drives the world.-  J It would only work if Digital (pardon me, HP) would do it in the 1st placeF to start a new trend. But do they have any software licensed like this' (other than OpenVMS itself of course) ?c   -- Syltrem    OpenVMS 7.3-1 + Oracle 8.1.7.4H http://pages.infinit.net/syltrem (OpenVMS related web site, en franais)% ---zulu is not in my email address--- F "Jim Brankin" <jbrankin@ntlworld.com> a crit dans le message de news:2 863f19d6.0308060844.24e70849@posting.google.com...C > I have a suggestion for an enhancement to VMS. How about coupling 2 > class based scheduling with the licence manager. >rB > The object of the exercise would be to licence software to use aG > certain percentage of the resources of a machine. It seems to me thatiG > this would encourage server consolidation and that this would benefitd > VMS. >oG > The cost of software licences frequently varies with the CPU power ofhH > the machine, sometimes with the amount of memory and almost never with$ > the capacity of the I/O subsystem. >pE > It seems to me that this works against VMS, which can walk and chew B > gum at the same time. You can run more than one application on aH > single machine whereas in the Microsoft and Linux worlds that is rare.D > But if you run two apps on the one machine you pay licence fees asD > though each application had the machine to itself. This encourages6 > folk to use the one machine / one application model. > E > Now, if you could buy a licence which said that an application will F > use only half the CPU power of the machine then the problem would beE > solved. It should be possible for the licence manager and the class0H > based sheduler to co-operate to provide this. Just add the schecdulingG >  class and the percentage to the information specified in the licenceRH > and have the licence checking routine check that the user id is in the? > right scheduling class and that the percentage of the machines( > available to that class is acceptable. >s >  > What do you think? >t > Jimm >  >e >rL > -------------------------------------------------------------------------- --0 >                          Pure Personal Opinion >c9 >                        brankin at nildram dot co dot ukeL > -------------------------------------------------------------------------- --   ------------------------------  $ Date: Wed, 6 Aug 2003 22:05:16 -0400' From: "Main, Kerry" <kerry.main@hp.com>e1 Subject: RE: Licence manager and class scheduling R Message-ID: <FD827B33AB0D9C4E92EACEEFEE2BA2FB0C7A6D@tayexc19.americas.cpqcorp.net>   Jim,  : Re: licensing, class scheduler and server consolidation ..  G The savings in a server consolidation project are typically in reducing7H the number of one app one servers and moving these apps to fewer serversE with bigger cpu's. The trade-off is that now you have to be concernedvG with one application (or group of processes) taking over more resourcess: than their share of the system and impacting other groups.  H Hence, to enforce various service levels in place, a class scheduler (orE workload manager) is used to restrict specific workloads so that theys@ never take over more than x% of the cpu resources on the system.  G This is also where Galaxy can assist as well. If two applications do noaB play together well (for whatever reason), then you can put them onC separate OS partitions and let the Galaxy software dynamically moves@ cpu's back and forth between the 2+ OS instances as the workload? changes. You can also set Galaxy up for managing resources with,H something like "move 3 cpu's from the mid tier App Server after 21:00 onG InstanceA to the database /batch server on instanceB to assist in dailysD crunching, but then move the 3 cpu's back at 06:00 to assist the App. Server deal with the daily interactive loads."  G I suspect what you are hinting at is actually something a bit differenteE ie. resource based charging where detailed audit reports show exactlyaG how much CPU time each process utilized and the end user/business groupRA are charged accordingly. Its kind of like your hydro utility i.e.YG monthly billing based on historical estimate, but adjusted periodicallye" to deal with changes in workloads.   :-)n   Regardsf  
 Kerry Main Senior Consultant3 Hewlett-Packard (Canada) Co.! Consulting & Integration Servicesn Voice: 613-592-4660- Fax   : 613-591-4477 Email: kerryDOTmain@hpDOTcom-     (remove the DOT's and replace with "."'s)D OpenVMS DCL - the original .COM  =20h   >=20 > -----Original Message-----5 > From: Jim Brankin [mailto:jbrankin@ntlworld.com]=20S > Sent: August 6, 2003 12:45 PMa > To: Info-VAX@Mvb.Saic.Comc >=20= > I have a suggestion for an enhancement to VMS. How about=20 ; > coupling class based scheduling with the licence manager.e >=20? > The object of the exercise would be to licence software to=20i? > use a certain percentage of the resources of a machine. It=20 ? > seems to me that this would encourage server consolidation=20w" > and that this would benefit VMS. >=20A > The cost of software licences frequently varies with the CPU=20nB > power of the machine, sometimes with the amount of memory and=206 > almost never with the capacity of the I/O subsystem. >=20? > It seems to me that this works against VMS, which can walk=201= > and chew gum at the same time. You can run more than one=20 A > application on a single machine whereas in the Microsoft and=200 > Linux worlds that is rare.? > But if you run two apps on the one machine you pay licence=20e? > fees as though each application had the machine to itself.=20mF > This encourages folk to use the one machine / one application model. >=207 > Now, if you could buy a licence which said that an=20o@ > application will use only half the CPU power of the machine=20@ > then the problem would be solved. It should be possible for=208 > the licence manager and the class based sheduler to=20@ > co-operate to provide this. Just add the schecduling  class=20; > and the percentage to the information specified in the=20$A > licence and have the licence checking routine check that the=20 : > user id is in the right scheduling class and that the=20B > percentage of the machine available to that class is acceptable. >=20 >=20 > What do you think? >=20 > Jim=20 >=20 >=20 >=20@ > -------------------------------------------------------------- > --------------0 >                          Pure Personal Opinion >=209 >                        brankin at nildram dot co dot ukn@ > -------------------------------------------------------------- > -------------- >=20   ------------------------------  $ Date: Wed, 6 Aug 2003 19:46:00 -0400) From: "David @ Island" <david-@-hpaq.net>s; Subject: Re: Migrate obsolete VAX/VMS SW to OpenVMS Itaniumo/ Message-ID: <vj34o9bvcc5k76@news.supernews.com>   
 OK Good pointR  H And if we did, we owuld have to charge double the prices for the product   ;0(E  < "David J. Dachtera" <djesys.nospam@fsi.net> wrote in message! news:3F2B1D0E.8C3606FF@fsi.net...e > "David @ Island" wrote:e > >c  > > But is that really the case? > >b? > > Remember we have Alpha systems at below Intel boxen prices.a > > Loads of them in factIA > > Still don't hear the phone ringing any more than usual thougha >,A > ...and you've placed big, visible, people-grabbing ads in whichn > publications?e >  > --   > David J. Dachtera  > dba DJE Systemst > http://www.djesys.com/ >e* > Unofficial Affordable OpenVMS Home Page:! > http://www.djesys.com/vms/soho/l   ------------------------------  % Date: Wed, 06 Aug 2003 13:30:58 -0500O1 From: "David J. Dachtera" <djesys.nospam@fsi.net>  Subject: Re: Netware is no VMS' Message-ID: <3F314962.FA7AC1BC@fsi.net>b   "John Gemignani, Jr." wrote: > > > "David J. Dachtera" <djesys.nospam@fsi.net> wrote in message# > news:3F312E62.B2D691E7@fsi.net... 
 > > [snip]K > > Huh? Aren't coffee stains darker than beer stains? Coffee with sugar is-B > > probably at least as sticky as beer when it dries, I s'pose... > D >     But if you come home late with coffee stains your wife doesn't& >     ask where you've been, does she? > = >     "What is that?  A COFFEE STAIN!!!!  You've been to thatsD >     boutique again!  That's it!  I'm going home to mother, and I'm >     taking the PC!"l > E >     (Good ridance ... to the PC, that is ... you deserve every blue  >     screen that you get.)m   *LOL*    Uxcellent, dude!   -- u David J. Dachtera  dba DJE Systemse http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/h   ------------------------------  $ Date: Wed, 6 Aug 2003 23:05:19 +0100( From: "John Travell" <john@jomatech.com> Subject: Re: Netware is no VMS9 Message-ID: <bgrube$po1nb$1@ID-120847.news.uni-berlin.de>   3 "jlsue" <jefflsxxxz@sbcglobal.net> wrote in messaget2 news:ueb2jvs5e4upk428af4h2orv3bu1j3vvfg@4ax.com...1 > I know this take us a little off-topic, but....e >(I > What new features do you believe are needed above what's already in theiF > roadmap?  For me, at this point, I don't see any real missing killerJ > feature that's both feasible (i.e., not something like process migration' > between cluster nodes), and valuable.  >   L I do not know how true is is, but I once heard a war story about someone whoJ installed the same swapfile on two nodes in a cluster. Allegedly a process< was swapped out from one machine and swapped into the other.( According to the tale, it nearly worked.H The gotcha was that the I/O device owned by that process on the original: system was owned by a different process on the new system.  G It is by no means all we would need, but a first step would be a singlee clusterwide I/O database.iD Every known device would need to flag whether it was a node-specificJ non-migratable device, or one that could be moved  to, or re-created on, a! different machine in the cluster.eK Next we would need a properly managed global swapfile. Like the quorum file K this would need node specific write zones as well as controlled access to a5 common area.K There are bound to be many other problems that would need to be sorted, butt I think it could be done..  K All we have left is the combined willpower of the customer base to persuadeyJ the the purse holders that this needs to be done, and to fix engineering's: priority queue so that it happens soon. (Dream on boy! :-)     -- John Travell" Independent VMS crashdump analyst. john- at - jomatech - dot - comt +44-(0)23-92552229 http://www.jomatech.com/           ---m& Outgoing mail is certified Virus Free.: Checked by AVG anti-virus system (http://www.grisoft.com).A Version: 6.0.506 / Virus Database: 303 - Release Date: 01/08/2003    ------------------------------  $ Date: Wed, 6 Aug 2003 21:29:32 -0400* From: "Bill Todd" <billtodd@metrocast.net> Subject: Re: Netware is no VMS2 Message-ID: <UoSdnciTA-TCMayiXTWJjg@metrocast.net>  3 "jlsue" <jefflsxxxz@sbcglobal.net> wrote in messagee2 news:ueb2jvs5e4upk428af4h2orv3bu1j3vvfg@4ax.com...   ...h  I > What new features do you believe are needed above what's already in ther
 > roadmap?  J How about that new file system that VMS developers have been talking about for 3 - 4 years now?  G How about some of those explicitly uncommitted 'future investigations'?1G Real 'fork' is another item VMS developers have been talking about (andrH actually semi-promising, though with no hard date) for several years nowE (seeing it called 'uncommitted' under those circumstances seems a bit ? alarming) - and Unix98 compliance, plus support for a real UnixnA 'personality' that Unix people as well as applications could feel L comfortable working in, is hardly unimportant to wider acceptance of VMS (ifF that's something HP has any interest in; IBM, it should be noted, *is*I interested in such support, and has done an excellent job of adding Linuxa@ personalities even to its decidedly non-Unixy i-series systems).  L But there should be people around here in a far better position than I am to% make up a list of desirable features.o   - bill   ------------------------------   Date: 6 Aug 2003 22:48:07 -0500i+ From: young_r@encompasserve.org (Rob Young)n Subject: Re: Netware is no VMS3 Message-ID: <KByM6ZZQjhd2@eisner.encompasserve.org>   _ In article <UoSdnciTA-TCMayiXTWJjg@metrocast.net>, "Bill Todd" <billtodd@metrocast.net> writes:  > 5 > "jlsue" <jefflsxxxz@sbcglobal.net> wrote in messageo4 > news:ueb2jvs5e4upk428af4h2orv3bu1j3vvfg@4ax.com... >  > ...  > J >> What new features do you believe are needed above what's already in the >> roadmap?t > L > How about that new file system that VMS developers have been talking about > for 3 - 4 years now? >    	Back to February:  | http://groups.google.com/groups?selm=rdeininger-2202030718590001%40user-uinj4lr.dialup.mindspring.com&oe=UTF-8&output=gplain  > In article <G8ednVIlc4atTsujXTWcpQ@metrocast.net>, "Bill Todd" <billtodd@metrocast.net> wrote:   M >> I doubt enhancements to TPU will get higher priority than, say, a new filel
 >> system. >nD >And since plans for a new file system seem to have pretty much beenK >languishing for over three years now, with no obvious progress (not even a M >mention on the roadmap through early 2005), well, draw your own conclusions.e  H What makes you think the file system has been languishing?  Just because nothing has shipped?  _ http://groups.google.com/groups?selm=v5kru8khbin516%40corp.supernews.com&oe=UTF-8&output=gplainl   re: new file systemi  L At our last LUG meeting here in Chicago, we heard that a new VMS file systemE is being worked on.  There were lessons learned the last time around.w   -- Dave...l   ----    N > But there should be people around here in a far better position than I am to' > make up a list of desirable features.r  @ 	Why not?  You know enough about OSes.  Many of your ideas would 	apply to VMS, wouldn't they?s  
 			       Rob    ------------------------------  % Date: Thu, 07 Aug 2003 07:39:16 +0200  From: Dirk Munk <munk@home.nl> Subject: Re: Netware is no VMS2 Message-ID: <bgsotd$6ol$1@news3.tilbu1.nb.home.nl>   Bill Todd wrote:5 > "jlsue" <jefflsxxxz@sbcglobal.net> wrote in messagen4 > news:ueb2jvs5e4upk428af4h2orv3bu1j3vvfg@4ax.com... >  > ..., >  > I >>What new features do you believe are needed above what's already in they
 >>roadmap? >  > L > How about that new file system that VMS developers have been talking about > for 3 - 4 years now? > I > How about some of those explicitly uncommitted 'future investigations'?dI > Real 'fork' is another item VMS developers have been talking about (and J > actually semi-promising, though with no hard date) for several years nowG > (seeing it called 'uncommitted' under those circumstances seems a bit A > alarming) - and Unix98 compliance, plus support for a real UnixBC > 'personality' that Unix people as well as applications could feelc > comfortable working in,   2 AFAIK the Unix stuff is what we get with VMS 7.3-2  4 is hardly unimportant to wider acceptance of VMS (ifH > that's something HP has any interest in; IBM, it should be noted, *is*K > interested in such support, and has done an excellent job of adding LinuxlB > personalities even to its decidedly non-Unixy i-series systems). > N > But there should be people around here in a far better position than I am to' > make up a list of desirable features.l >  > - bill >  >  >    ------------------------------   Date: 6 Aug 2003 23:31:58 -0500T+ From: young_r@encompasserve.org (Rob Young)e Subject: Re: Netware is no VMS3 Message-ID: <lGeqfr8i1gsm@eisner.encompasserve.org>o  d In article <bgrube$po1nb$1@ID-120847.news.uni-berlin.de>, "John Travell" <john@jomatech.com> writes:5 > "jlsue" <jefflsxxxz@sbcglobal.net> wrote in messageo4 > news:ueb2jvs5e4upk428af4h2orv3bu1j3vvfg@4ax.com...2 >> I know this take us a little off-topic, but.... >>J >> What new features do you believe are needed above what's already in theG >> roadmap?  For me, at this point, I don't see any real missing killersK >> feature that's both feasible (i.e., not something like process migrationr( >> between cluster nodes), and valuable. >> >   ? 	Maybe in a few days a few good ideas surface.  I guess you are ? 	saying all the good ideas for OS improvement have already beenn 	invented?  Of course not.  % 	I keep returning to this one.  TRAM:t  [ http://groups.google.com/groups?selm=VA.00000201.25384c70%40sture.ch&oe=UTF-8&output=gplaine  ? 	But that is dead or not worth focusing on.  Here is one betterm 	than TRAM that will be coming:r  5 http://www.internetnews.com/infra/article.php/2219411o   June 10, 2003 - IBM, Infineon Develop Better Magnetic Memory ) By Michael Singer     H What if you could turn computers on and off as quickly as a light switchN without having to wait for software to "boot up"? Sound like science fiction?   O Researchers at IBM (Quote, Company Info) and German-based Infineon TechnologiesgM AG (Quote, Company Info) don't think so. The two companies Monday highlightedeM their advances on Magnetic (or Magnetoresistive) Random Access Memory (MRAM). H The nonvolatile memory technology uses magnetic, rather than electronic,E charges to store bits of data. The companies say they've advanced thegN technology so much it could replace current computer memory as early as 2005.   < 	To me this one is a "not if"  but a "when" proposition.  It 	will be coming.    F 	The missing killer VMS feature is true WRITEBEHIND (with confidence) > 	at a filesystem level.  With MRAM, the writes are not lost on9 	power outtage.  So as Fred mentioned what the impetus or ! 	catalyst for Galaxy's birth was:   g http://groups.google.com/groups?selm=366D4B85.94E0DC5%40star.enet.dec_nospam.com&oe=UTF-8&output=gplainu  >   Congratulations, you've just discovered much of the rationalB   behind Galaxy (too late, I'm on 4 of the patent applications ;-)  A   Galaxy - actually that's a marketing name, Adaptive PartitionedsC   Multi-Processing (APMP) is the product just about to ship in it'si:   initial form from Compaq on the OpenVMS Alpha TurboLaserC   (AlphaServer 8x00).  It is actually targeted downstream at futureDF   highly scaleable systems that will eventually have hundreds of CPUs,G   a terabyte of memory, and potentially hundreds of I/O busses.  Worse,eB   the hardware designers are designing these big-ass systems using5   various forms of non-uniform memory and I/O access.r   ---a  9 	It isn't unreasonable to project MRAM as the much needed > 	catalyst to push a VMS filesystem re-write in a VERY excitingB 	direction.  Even if a future Itanium server had a portion of MRAMB 	so that the HP/UX folks and NSK, and VMS folks could target their: 	OS improvements towards.  In other words, a collaborative: 	effort.  One can imagine IBM knows what MRAM can and will: 	do for them.  Certainly don't want the competition to get: 	a leg up, do you?  Here would be a chance to be proactive; 	on the filesystem re-write, targetting future hardware and > 	sell the MRAM concept company-wide for corporate buy-in, etc.  9 	Imagine the ideas for managing writes and IO if all youro. 	writes to disk could occur in a lazy fashion.   				Robs  B Men with walkie-talkie                  I'm home again to you babeC Men with flashlights waving             You know it makes me wonder G Up upon the tower                       Sittin' in the quiet slipstreamc> The clock reads daylight savin'         Rollin' in the thunder  -                                 -- Neil Youngr                              n   ------------------------------  % Date: Wed, 06 Aug 2003 11:51:19 -0700e* From: Alan Frisbie <Abuse@Flying-Disk.com>$ Subject: Re: PDP-11 OS Release Dates. Message-ID: <3F314E27.3010108@Flying-Disk.com>   jmfbahciv@aol.com wrote:- > In article <3F2FEB55.C73F6F0@yk.rim.or.jp>,e, >    Ishikawa <ishikawa@yk.rim.or.jp> wrote:  ; >> Since the memory was so tight, saving code space thusly l: >> by having a single READ/WRITE routine that is used both7 >> for input/output was considered a win. (This was note& >> thread safe if I recall correctly.)8 >> I heard that this code was done by Bill Gates himself7 >> and that he was very proud of the code trick and theD >> memory saving.    > GoodfuckingGRIEF!e  > > Bill Gates thinking would have improved if somebody had done > a lobotomy.  n  ? It is worst than that.   There are places in Altair BASIC where @ they saved space by jumping into the second byte of a three-byte? instruction!   Yes, it's true.   I'm not kidding.   In the latei: 1970's, a group of us disassembled Altair BASIC (easy) and> proceeded to reverse-engineer and comment it (hard).   It took< me a while to understand that section of code, after which I felt, shall we say, "unclean".   Alan   ------------------------------  * Date: Wed, 6 Aug 2003 18:50:11 +0000 (UTC). From: dfevans@bcr10.uwaterloo.ca (David Evans)$ Subject: Re: PDP-11 OS Release Dates/ Message-ID: <bgril3$fnv$1@tabloid.uwaterloo.ca>c  . In article <3F314E27.3010108@Flying-Disk.com>,, Alan Frisbie  <Abuse@Flying-Disk.com> wrote: >d@ >It is worst than that.   There are places in Altair BASIC whereA >they saved space by jumping into the second byte of a three-byte 2 >instruction!   Yes, it's true.   I'm not kidding.  F   This sort of thing is still touted as "real, skilled programming" by? some die-hard eight-bitters (what a good unintentional pun that D was...).  I'm sure the practice would receive saged nods of approval? in, say, comp.sys.cbm.  Every now and then they go on about how @ current software is bloated and crappy because programmers don't( know these kinds of tools of efficiency.   --  M David Evans          (NeXTMail/MIME OK)             dfevans@bbcr.uwaterloo.ca M Ph.D. Candidate, Computer/Synth Junkie     http://bbcr.uwaterloo.ca/~dfevans/ M University of Waterloo         "Default is the value selected by the composer M Ontario, Canada           overridden by your command." - Roland TR-707 Manual    ------------------------------  % Date: Wed, 06 Aug 2003 13:22:14 -06000% From: Dan O'Reilly <dano@process.com> $ Subject: Re: PDP-11 OS Release DatesA Message-ID: <5.2.0.9.2.20030806131806.02172300@raptor.psccos.com>   ( At 12:50 PM 8/6/2003, David Evans wrote:/ >In article <3F314E27.3010108@Flying-Disk.com>, - >Alan Frisbie  <Abuse@Flying-Disk.com> wrote:  > > B > >It is worst than that.   There are places in Altair BASIC whereC > >they saved space by jumping into the second byte of a three-byteF4 > >instruction!   Yes, it's true.   I'm not kidding. > H >   This sort of thing is still touted as "real, skilled programming" by@ >some die-hard eight-bitters (what a good unintentional pun thatE >was...).  I'm sure the practice would receive saged nods of approvalP@ >in, say, comp.sys.cbm.  Every now and then they go on about howA >current software is bloated and crappy because programmers don't0) >know these kinds of tools of efficiency.1  J Errrr, I can't say I would disagree with them.  I've maintained for a longG time it took a lot more skill to produce large, complicated systems on, I say, RSX than it does on VMS simply because you have to be concerned with F a much more restricted address space (and I've done so on both).  WhenJ VMS first came out, I had a friend tell me of VMS: "it's great - if I needJ to allocate a buffer to enter a name in, I allocate a megabyte".  Granted,G that's on the ridiculous side, but it's also somewhat illustrative of a  mindset.  H On the other hand, there's something to be said about not having to beatJ yourself up against a brick wall to get something done.  Self-flagellation: just ain't that fun, if you're forced to do it...<grin>...   ------J +-------------------------------+----------------------------------------+J | Dan O'Reilly                  |  "There are 10 types of people in this |J | Principal Engineer            |   world: those who understand binary   |J | Process Software              |   and those who don't."                |J | http://www.process.com        |                                        |J +-------------------------------+----------------------------------------+   ------------------------------  $ Date: Wed, 6 Aug 2003 15:29:18 -0400* From: "Bill Todd" <billtodd@metrocast.net>$ Subject: Re: PDP-11 OS Release Dates2 Message-ID: <INCdnVcZmYa9yqyiXTWJkw@metrocast.net>  ; "David Evans" <dfevans@bcr10.uwaterloo.ca> wrote in message ) news:bgril3$fnv$1@tabloid.uwaterloo.ca...Y0 > In article <3F314E27.3010108@Flying-Disk.com>,. > Alan Frisbie  <Abuse@Flying-Disk.com> wrote: > >nB > >It is worst than that.   There are places in Altair BASIC whereC > >they saved space by jumping into the second byte of a three-bytec4 > >instruction!   Yes, it's true.   I'm not kidding. >,H >   This sort of thing is still touted as "real, skilled programming" byA > some die-hard eight-bitters (what a good unintentional pun thateF > was...).  I'm sure the practice would receive saged nods of approvalA > in, say, comp.sys.cbm.  Every now and then they go on about howeB > current software is bloated and crappy because programmers don't* > know these kinds of tools of efficiency.  L Which is just about as idiotic as condemning the practice (and skill) in the4 context of the time and system in which it was used.  K Times change, constraints often differ radically across different platformsO4 even in the *same* time, and so do 'best practices'.   - bill   ------------------------------  % Date: Wed, 06 Aug 2003 21:23:22 -0000a2 From: m_thompson-NOSPAM@ids.net (Michael Thompson)$ Subject: Re: PDP-11 OS Release Dates/ Message-ID: <vj2seacu0nhc51@corp.supernews.com>g  J In article <INCdnVcZmYa9yqyiXTWJkw@metrocast.net>, billtodd@metrocast.net  says...e >s >e< >"David Evans" <dfevans@bcr10.uwaterloo.ca> wrote in message* >news:bgril3$fnv$1@tabloid.uwaterloo.ca...1 >> In article <3F314E27.3010108@Flying-Disk.com>,t/ >> Alan Frisbie  <Abuse@Flying-Disk.com> wrote:e >> >C >> >It is worst than that.   There are places in Altair BASIC whereiD >> >they saved space by jumping into the second byte of a three-byte5 >> >instruction!   Yes, it's true.   I'm not kidding.  >>I >>   This sort of thing is still touted as "real, skilled programming" byyB >> some die-hard eight-bitters (what a good unintentional pun thatG >> was...).  I'm sure the practice would receive saged nods of approval B >> in, say, comp.sys.cbm.  Every now and then they go on about howC >> current software is bloated and crappy because programmers don't-+ >> know these kinds of tools of efficiency.- >-M >Which is just about as idiotic as condemning the practice (and skill) in theM5 >context of the time and system in which it was used.n >sL >Times change, constraints often differ radically across different platforms5 >even in the *same* time, and so do 'best practices'.y >  >- billr  K I disassembled some 8080 code a couple of decades ago. It took me a little cN while to figure out that it was self modifying and changed the destination of ) jumps. That probably saved a byte or two.O   ------------------------------   Date: 6 Aug 2003 18:27:44 -0700o' From: st-wong@alumni.cuhk.net (ST Wong)'< Subject: Re: pls help: install OS on DS10L without console ?= Message-ID: <28073c51.0308061727.327ad51a@posting.google.com>e  9 It works now!  Thanks to all who responded to my posting.    Regards, /STi   ------------------------------  % Date: Wed, 06 Aug 2003 13:48:32 -0500l1 From: "David J. Dachtera" <djesys.nospam@fsi.net> ? Subject: Re: Server consolidation & VMS - "Back to the Future"! ' Message-ID: <3F314D80.F8178B01@fsi.net>p   "John Gemignani, Jr." wrote: > > > "David J. Dachtera" <djesys.nospam@fsi.net> wrote in message# > news:3F312AE1.F2B8E567@fsi.net...r
 > > [snip]G > > Yes, it will be costly, especially as HP will not be assisting witheH > > marketing costs for our VMS ventures. Still, I believe the long-termL > > rewards will be there. VMS has staying power, despite decades of neglect > > and "unmarketing". > M >     But keep in mind that the subsequent rewards fall mostly into the handsUN >     of those who produce (and market, and sell), and associated system salesJ >     with VMS from HP (IMHO) won't entice HP into bringing about the true >     VMS renaissance.  H ...which is why I believe it's up to us, the folks out here on the frontG lines, to make it happen. We're smaller and have less resources, that's E true; however, at the same time, we're lighter and more agile and canoF more readily adapt to our customers' needs than a lumbering, lethargic? behemoth like hp or "OpenVMS, a wholly-owned subsidiary of hp".-  	 >     So:  > 5 >         IF (VMS runs on mix-and-match hardware) ANDlJ >             (the "other creative ones" in the field make it do much more- >                     than off-the-shelf VMS)  >         THENG >             I'd like to believe that VMS may gain in popularity againh6 >             despite what the owner is doing with it;  B ...or not doing, as the case may be (and certainly appears to be).   > >tL > > I believe the time is now, more than it has been in a while, to take theL > > ball and run with it. Do what HP refuses to do: Bring the VMS message toK > > the streets and remind the world what reliability, security and restfulD > > nights are all about.t > J >     I am finally getting around to doing some of the stuff that I alwaysG >     wanted to do but never did because I was too busy doing stuff ford >     someone else.   F I think we all know where that's at: too busy earning a living to make
 any money.  < >     P.S. I'm going to take my Vaxes out of moth balls now.  H Hhmmm... interesting thought: would napthalene have a negative effect on any plastic/rubber/etc. parts?   -- e David J. Dachterap dba DJE Systems> http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/    ------------------------------  % Date: Thu, 07 Aug 2003 05:20:24 +0800 , From: Paul Repacholi <prep@prep.synonet.com>< Subject: Re: Sun tries to woo AlphaServer users away from HP- Message-ID: <87fzkecwkn.fsf@prep.synonet.com>o  ; "Fred Kleinsorge" <my-last-name@stardotzko.dec.com> writes:n   > SDL has been ported to C.r  : Well, that is good news. Will we be seeing the new version on the freeware disk?g   -- f< Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda.h@                                              West Australia 6076* comp.os.vms,- The Older, Grumpier Slashdot. Raw, Cooked or Well-done, it's all half baked.F EPIC, The Architecture of the future, always has been, always will be.   ------------------------------  # Date: Thu, 07 Aug 2003 01:51:22 GMT / From: "Ken Farmer" <kfarmer@NOSPAM.OpenVMS.org>c< Subject: Re: Sun tries to woo AlphaServer users away from HP? Message-ID: <ugiYa.326973$nr.13403188@twister.southeast.rr.com>)  6 "JF Mezei" <jfmezei.nospam@istop.com> wrote in message# news:3F295EA3.892CA085@istop.com...  > Rob Young wrote:D > > "You're going to see Linux and Windows absolutely eviscerate the
 midrange Unixh > > market," said Capellas.  > >lI > >         Capellas was/is the source and we beat this about quite a bit  > >         1.2 years ago. >pJ > Sorry, Winkler was the original source. Curly was merely continuing what hisb( > handlers such as Winkler had told him.    4 CRN: Capellas: Linux, Windows Will 'Eviscerate' Unix7 http://www.tru64.org/stories.php?story=02/05/03/2615381r     -- Kenneth Farmer  <><     EnterpriseUnix.org  |  Tru64.org OpenVMS.org  |  dcl.OpenVMS.orgm$ EnterpriseLinux.org  |  LinuxHPC.org   ------------------------------   Date: 6 Aug 2003 22:09:46 GMT2 From: healyzh@aracnet.com6 Subject: WASD and RDB?, Message-ID: <bgruba015s9@enews1.newsguy.com>  L I'm looking for examples of interfacing WASD and RDB.  Basically I'm wantingJ to setup a CGI to extract and format data from RDB.  Needless to say, this' is going to be a learning experience...    		Zane   ------------------------------  # Date: Wed, 06 Aug 2003 23:09:00 GMTgL From: winston@SSRL.SLAC.STANFORD.EDU ("Alan Winston - SSRL Admin Cmptg Mgr") Subject: Re: WASD and RDB?6 Message-ID: <00A23FBD.82C9DF73@SSRL.SLAC.STANFORD.EDU>  H In article <bgruba015s9@enews1.newsguy.com>, healyzh@aracnet.com writes:  M >I'm looking for examples of interfacing WASD and RDB.  Basically I'm wanting@K >to setup a CGI to extract and format data from RDB.  Needless to say, thisp( >is going to be a learning experience...  N In general, you need some kind of glue to interface web servers and databases.  L WASD should be able support Rdbweb, the Oracle-supplied web agent which letsK you write routines as stored SQL procedures.  (Haven't tried it or seen it; O Rdbweb certainly works under OSU and Apache.  Rdbweb limits you to one databasea9 at a time, and I find writing SQL code for CGIs tedious.)e  K There are Rdb plugins for Perl and Python, both of which work (well) under s WASD.t  L I think you could use Tomcat (Java Server Pages, from the Apache project) orJ JBOSS with WASD; it doesn't seem like there's a lot of intimate connectionL between the JSP processor and the web server.  You can use either the OracleE thin JDBC client with the OCI-type SQL Services or the newly-releasede Rdb-specific thin JDBC client.  J (In general, if you're using a glue language, you care about the webserverJ supporting the language, and the language supporting the database, but theJ webserver itself doesn't have to do much for the database, or vice-versa.)   -- Alano --  O ===============================================================================m0  Alan Winston --- WINSTON@SSRL.SLAC.STANFORD.EDUM  Disclaimer: I speak only for myself, not SLAC or SSRL   Phone:  650/926-3056aM  Paper mail to: SSRL -- SLAC BIN 99, 2575 Sand Hill Rd, Menlo Park CA   94025aO ===============================================================================s   ------------------------------   Date: 7 Aug 2003 04:38:49 GMTy From: healyzh@aracnet.come Subject: Re: WASD and RDB?, Message-ID: <bgsl4p014bm@enews2.newsguy.com>  M "Alan Winston - SSRL Admin Cmptg Mgr" <winston@ssrl.slac.stanford.edu> wrote:hJ > In article <bgruba015s9@enews1.newsguy.com>, healyzh@aracnet.com writes:  N >>I'm looking for examples of interfacing WASD and RDB.  Basically I'm wantingL >>to setup a CGI to extract and format data from RDB.  Needless to say, this) >>is going to be a learning experience...g  P > In general, you need some kind of glue to interface web servers and databases.  I I'm thinking either perl or DCL.  The odds are I'll use a little of both,e but primarily perl.n  N > WASD should be able support Rdbweb, the Oracle-supplied web agent which letsM > you write routines as stored SQL procedures.  (Haven't tried it or seen it;-Q > Rdbweb certainly works under OSU and Apache.  Rdbweb limits you to one database ; > at a time, and I find writing SQL code for CGIs tedious.)-  H Unfortunatly I don't think I have the option of playing with Rdbweb, I'mE using a copy of Rdb download from the OTN for learning purposes.  I'dmK already figured on the "fun" of writing SQL code for the CGI's.  It doesn'ts2 look as bad as some things I've had to do in Perl.  M > There are Rdb plugins for Perl and Python, both of which work (well) under l > WASD.r  / I'll have to try and find Rdb modules for Perl.s  I Python and Java are out, as I don't speak either, and don't care to learnr% them while trying to re-learn SQL :^)s   			Zanee   ------------------------------   Date: 6 Aug 2003 13:20:21 -0700e* From: dfreybur@yahoo.com (Doug Freyburger)S Subject: Re: [OT] Re stopped clocks being right twice a day (was Re: HP FUDBusting)e= Message-ID: <7960d3ee.0308061220.34a842c0@posting.google.com>l   Alan E. Feldman wrote: > Bill Todd wrote: >cH > > Obscurity may help some, but even in an absolute sense VMS is almostL > > certainly more secure than most other systems out there, including Unix. >i* > Almost? What do you mean, "almost"?  :-|  H VMS was designed with security more in mind than Unix, so it started outG more secure.  VMS has fewer units, so it is a less common target, so itiI is less vulnerable.  Unix has a more active support/development staff, sor< Unix security bugs are more likely to be fixed more quickly.   ------------------------------   Date: 6 Aug 2003 15:35:27 -0500n; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler)3S Subject: Re: [OT] Re stopped clocks being right twice a day (was Re: HP FUDBusting)"3 Message-ID: <jj46udVNGlnD@eisner.encompasserve.org>e  j In article <7960d3ee.0308061220.34a842c0@posting.google.com>, dfreybur@yahoo.com (Doug Freyburger) writes: > Alan E. Feldman wrote: >> Bill Todd wrote:n >>I >> > Obscurity may help some, but even in an absolute sense VMS is almost M >> > certainly more secure than most other systems out there, including Unix.a >>+ >> Almost? What do you mean, "almost"?  :-|  > J > VMS was designed with security more in mind than Unix, so it started outI > more secure.  VMS has fewer units, so it is a less common target, so ityK > is less vulnerable.  Unix has a more active support/development staff, so > > Unix security bugs are more likely to be fixed more quickly.  ?    That is the most incompetent bull I've heard in a long time.u   ------------------------------   End of INFO-VAX 2003.433 ************************