1 INFO-VAX	Fri, 06 Sep 2002	Volume 2002 : Issue 491       Contents: Re: "inview" Article Re: "inview" Article Re: "inview" Article Re: "inview" Article Re: "inview" Article Re: "inview" Article Re: "inview" Article Re: After Memory Upgrade Re: After Memory Upgrade	 Re: am/pm 3 Beware of disappearing St. Louis hotel reservations ' Burning an image with an unknown format + Re: Burning an image with an unknown format > Re: Can one still vest Vax executables onto Alpha OpenVMS V7.3 Cancelling writes to a log file  Re: Compression Utility? Re: Compression Utility? Convert Blocks to MB Re: Convert Blocks to MB Re: Convert Blocks to MB Re: Convert Blocks to MB Re: Convert Blocks to MB% Re: Copy of OpenVMS hobbyist CD in UK % Re: Copy of OpenVMS hobbyist CD in UK % Re: Copy of OpenVMS hobbyist CD in UK % Re: Copy of OpenVMS hobbyist CD in UK % Re: Copy of OpenVMS hobbyist CD in UK = Curly's wisdom on how to handle HP's Enterprise Systems Group A Re: Curly's wisdom on how to handle HP's Enterprise Systems Group A Re: Curly's wisdom on how to handle HP's Enterprise Systems Group A Re: Curly's wisdom on how to handle HP's Enterprise Systems Group  Re: definition of SYS$SYSROOT   Disk size limit on Alpha-500au ?$ Re: Disk size limit on Alpha-500au ? Re: Help installing Pine Re: Help installing Pine+ Re: IBM did it, Sun did it, why Compaq not?  Re: Itanic2 performance  Re: Itanic2 performance  Re: Itanic2 performance  Re: Itanic2 performance # ja foste enganado no teu trabalho ?  Re: MATCHC Instruction8 Re: Microsoft reeling from hack attacks, VMS just yawns!8 Re: Microsoft reeling from hack attacks, VMS just yawns!9 Re: netscape/modzilla/HP/open-source: system requirements  Re: OpenSSL woes Post answers in forum people Request for Tools guide P Response to NIST's take on noon and midnight (was Re: Is the HP/Compaq merger re9 Re: Secret New Products (was: Is the HP/Compaq merger...) ' Sending Emails with OpenVms 7.2 + QUEUE + Re: Sending Emails with OpenVms 7.2 + QUEUE + Re: Sending Emails with OpenVms 7.2 + QUEUE , Re: SET PROTECTION (SET SECURITY/PROTECTION), Re: SET PROTECTION (SET SECURITY/PROTECTION)A Re: Sick of the trolls, whiners, and non-customers in this forum? A Re: Sick of the trolls, whiners, and non-customers in this forum? A Re: Sick of the trolls, whiners, and non-customers in this forum? A Re: Sick of the trolls, whiners, and non-customers in this forum? A Re: Sick of the trolls, whiners, and non-customers in this forum?  Re: SOAP Toolkit The Gauntlet is Cast!  Re: The Gauntlet is Cast!  Re: The Gauntlet is Cast! % Re: VAX / Network / Hardware question  VAXstation 3100 Model 76 SPX  Re: VAXstation 3100 Model 76 SPX  Re: VAXstation 3100 Model 76 SPX  Re: VAXstation 3100 Model 76 SPX& Re: VaxStation 4000/60 need hard drive& RE: VaxStation 4000/60 need hard drive& Re: VaxStation 4000/60 need hard drive& Re: VaxStation 4000/60 need hard drive Re: VMS Printing To CA/Dispatch ( Re: Weird, but you get that - marketing.F What features/capabilities would you like to see available on OpenVMS?J Re: What features/capabilities would you like to see available on OpenVMS?J Re: What features/capabilities would you like to see available on OpenVMS?J Re: What features/capabilities would you like to see available on OpenVMS?J Re: What features/capabilities would you like to see available on OpenVMS?P Re: What features/capabilities would you like to see available on OpenVMS? OpenVP Re: What features/capabilities would you like to see available on OpenVMS? OpenVP Re: What features/capabilities would you like to see available on OpenVMS? OpenVP Re: What features/capabilities would you like to see available on OpenVMS? OpenVP Re: What features/capabilities would you like to see available on OpenVMS? OpenVP Re: What features/capabilities would you like to see available on OpenVMS? OpenVA Re: Who are the largest VMS customers. Top ten? (Was: VMS and HP) A Re: Who are the largest VMS customers. Top ten? (Was: VMS and HP)  XP1000 667 selling !!!) YEA!  I have a working Dec 3000 300 alpha   F ----------------------------------------------------------------------  % Date: Thu, 05 Sep 2002 14:10:17 -0400 - From: JF Mezei <jfmezei.spamnot@videotron.ca>  Subject: Re: "inview" Article , Message-ID: <3D779E08.65FED326@videotron.ca>  ( Andrew Harrison SUNUK Consultancy wrote:+ > Solaris Internet Mail Server was replaced + > by Netscape's IMAP server as the standard  > IMAP server that we supplied.   K Did Sun purchase the SIMS/PMDF product before or after the association with 
 Netscape ?L Seems like Sun would have wasted much money when it purchased Innosoft. What* has happened to what used to be Innosoft ?    6 > The Netscape server won out though some bits of SIMS1 > were incorporated into it. iPlanet is as I said  > simply a re-branding.     I If it is just rebranding, how come my ISP would have gone through so many N troubles and email outages lasting over a full day, followed by other problemsN and errors when they moved from Netscape to iPlanet ? I have so far heard onlyJ tidbits, including some about incompetence, but also underestimates of how* massive the database conversions would be.  P It would seem to me that this was far more than a simple software upgrade to me.   ------------------------------  # Date: Thu, 05 Sep 2002 20:41:55 GMT * From: "Bill Todd" <billtodd@metrocast.net> Subject: Re: "inview" Article C Message-ID: <%jPd9.444825$m91.17180118@bin5.nnrp.aus1.giganews.com>   8 "David Webb" <david20@alpha2.mdx.ac.uk> wrote in message% news:al7g7o$t8c$1@aquila.mdx.ac.uk... 7 > In article <3D767E57.C14BF7C4@videotron.ca>, JF Mezei & <jfmezei.spamnot@videotron.ca> writes: > >Keith Parris wrote:J > >> I can take (and I have often seen this happen) an application writtenK > >> with no knowledge of clusters, put it onto a VMS Cluster platform, and F > >> I can immediately reap the benefits of the VMS Cluster platform's > > L > >That assumes that the application was written to allow multiple instances of< > >itself to co-exist. (for instance, file IO, logging etc). > L > In general it doesn't. For instance for logging, if the application reallyD > won't let you name the logfile and insists on opening a new one on startup,J > then VMS' file versioning comes to the rescue.  The first application toH > startup opens a logfile. The second one running on a different cluster nodeC > opens a logfile with the same name but one higher version number. : > Both then continue to write to their own opened logfile.  L I suggest that your response is incorrect on at least one count.  While yourJ example of transparent creation of multiple log files *might* work (thoughK would not if the application was designed to append to an existing log file E rather than create a new one), 'in general' if the application wasn't J written to allow multiple instances of itself then multiple such instancesC will trip over themselves in updating any common files (such as the L hypothetical aforementioned appended-to log file):  either they'll open themK prohibiting sharing (the safe way to do it) thus preventing other instances J from using them, or they'll open them without locking at all (uncontrolledJ sharing) and screw them up internally with overlapping concurrent updates,G or they'll open them appropriately shared (perhaps just from habit) but H won't *use* the interlocked access in a manner that keeps their contentsJ consistent (because they aren't aware that other concurrent updates may be occurring).   D Furthermore, you were responding to a statement that in order for anJ application 'written with no knowledge of clusters' to 'reap the benefits'C of VMS clustering the application must be written to allow multiple K instances of itself.  If you wish to contest that assertion, please provide L examples of how a single-instance application can 'reap the benefits' of VMSK clustering:  there may certainly be some (e.g., if VMS provides an external L system mechanism that can be set up to monitor an application and restart itI on another node if its host node fails, as some other clustering products 8 do), but they're not so obvious as to go without saying.   - bill   ------------------------------  % Date: Thu, 05 Sep 2002 22:52:10 -0400 - From: JF Mezei <jfmezei.spamnot@videotron.ca>  Subject: Re: "inview" Article , Message-ID: <3D781859.B1183D6C@videotron.ca>   David Froble wrote: N > really don't understand the claims of some about automatic re-start on other$ > cluster nodes, should a node fail.  I You'll note that Tandem doesn't have this capability "out of the box". An N application must be "tandem aware" in order to allow  a backup process to take# over transparently mid-transaction. K The application must "chat" with its backup process to feed it the data and  checkpoints etc.  K In a way, a Galaxy with shared memory might allow this type of architecture M where a backup process would "follow" what the master is doing and be able to J continue if needed. But that is also a waste of resources since the backup' process would do no nothing productive.    ------------------------------  % Date: Thu, 05 Sep 2002 21:03:36 -0400 ( From: David Froble <davef@tsoft-inc.com> Subject: Re: "inview" Article , Message-ID: <3D77FEE8.3060604@tsoft-inc.com>   Bill Todd wrote:  : > "David Webb" <david20@alpha2.mdx.ac.uk> wrote in message' > news:al7g7o$t8c$1@aquila.mdx.ac.uk...  > 7 >>In article <3D767E57.C14BF7C4@videotron.ca>, JF Mezei  >>( > <jfmezei.spamnot@videotron.ca> writes: >  >>>Keith Parris wrote: >>> I >>>>I can take (and I have often seen this happen) an application written J >>>>with no knowledge of clusters, put it onto a VMS Cluster platform, andE >>>>I can immediately reap the benefits of the VMS Cluster platform's  >>>>L >>>That assumes that the application was written to allow multiple instances >>>  > of > < >>>itself to co-exist. (for instance, file IO, logging etc). >>> L >>In general it doesn't. For instance for logging, if the application reallyD >>won't let you name the logfile and insists on opening a new one on >>
 > startup, > J >>then VMS' file versioning comes to the rescue.  The first application toH >>startup opens a logfile. The second one running on a different cluster >> > node > C >>opens a logfile with the same name but one higher version number. : >>Both then continue to write to their own opened logfile. >> > N > I suggest that your response is incorrect on at least one count.  While yourL > example of transparent creation of multiple log files *might* work (thoughM > would not if the application was designed to append to an existing log file G > rather than create a new one), 'in general' if the application wasn't L > written to allow multiple instances of itself then multiple such instancesE > will trip over themselves in updating any common files (such as the N > hypothetical aforementioned appended-to log file):  either they'll open themM > prohibiting sharing (the safe way to do it) thus preventing other instances L > from using them, or they'll open them without locking at all (uncontrolledL > sharing) and screw them up internally with overlapping concurrent updates,I > or they'll open them appropriately shared (perhaps just from habit) but J > won't *use* the interlocked access in a manner that keeps their contentsL > consistent (because they aren't aware that other concurrent updates may be
 > occurring).  > F > Furthermore, you were responding to a statement that in order for anL > application 'written with no knowledge of clusters' to 'reap the benefits'E > of VMS clustering the application must be written to allow multiple M > instances of itself.  If you wish to contest that assertion, please provide N > examples of how a single-instance application can 'reap the benefits' of VMSM > clustering:  there may certainly be some (e.g., if VMS provides an external N > system mechanism that can be set up to monitor an application and restart itK > on another node if its host node fails, as some other clustering products : > do), but they're not so obvious as to go without saying. >  > - bill    K I think that a more reasonable statement to make is that if an application  Q supports multiple users/instances of it's functionality on a single system, then  Q it can benefit from the scaling offered by clustering multiple systems.  Such an  P application would already support shared access to files through use of the DLM M or comparable locking tools, so for applications using the DLM, sharing in a  P cluster is rather transparent, with respect to being on a cluster.  If there is L a different locking tool, it WOULD have to be cluster aware, unless it used  cluster aware features of VMS.      N I have to note that while scaling of resources is a benefit of VMS clusters, IL really don't understand the claims of some about automatic re-start on other" cluster nodes, should a node fail.    K Say that I have an interactive application that's accepting user input and  P storing the data until a point is reached where a transaction can be written to L disk.  Sometimes the amount of data can be substantial.  Should the node on N which the application go down, the data stored in memory is irrevocably lost. N Yes, the application design could be such that every step gets saved on disk, Q and re-start capability is designed in.  But if not, what's in memory is, as far   as I know, lost.  P Even worse, a bad application design that does partial database updates, and no P re-start capability is built into the application.  When that is interrupted in H the middle, the database is corrupted with a partial transaction.  Yes, P application design issues, but nothing about the surviability of a cluster to a R single node going down addresses such problems.  Clusters are good, but not magic.   Dave   ------------------------------  # Date: Fri, 06 Sep 2002 02:59:10 GMT * From: "Bill Todd" <billtodd@metrocast.net> Subject: Re: "inview" Article A Message-ID: <2SUd9.196191$On.7825328@bin3.nnrp.aus1.giganews.com>   5 "David Froble" <davef@tsoft-inc.com> wrote in message & news:3D77FEE8.3060604@tsoft-inc.com...   ...   L > I think that a more reasonable statement to make is that if an applicationE > supports multiple users/instances of it's functionality on a single  system, thenI > it can benefit from the scaling offered by clustering multiple systems.   6 I believe that's what JF was saying, and it's correct.   ...   L > Say that I have an interactive application that's accepting user input andF > storing the data until a point is reached where a transaction can be
 written toJ > disk.  Sometimes the amount of data can be substantial.  Should the node onI > which the application go down, the data stored in memory is irrevocably  lost. I > Yes, the application design could be such that every step gets saved on  disk, K > and re-start capability is designed in.  But if not, what's in memory is,  as far > as I know, lost.  K Even if you saved every step, you could still always lose the last step the  user had performed.   L The way you make an application non-stop from the user input viewpoint is toH run it on a separate system from the one the user is sitting at.  If theF user's desktop fails, s/he doesn't expect the keyboard to fail over toH something else with no data loss.  But if the user's desktop is running,C it's certainly possible to ensure that nothing behind it fails in a K user-visible manner - the easiest way being to have intelligent software on I the desktop that knows how to fail over to a different back-end server if E the host it's talking with suddenly croaks and resubmit the last user F transaction if necessary (yes, the fail-over host also has to be smartF enough to ascertain whether that transaction needs to be redone or was already captured).   > J > Even worse, a bad application design that does partial database updates, and no4 > re-start capability is built into the application.  L If the application can't survive restart after a power outage, then it's notH clear that it's one worth considering in this discussion.  But if it canL survive such a restart, then it can survive fail-over to another instance in exactly the same manner.     When that is interrupted in I > the middle, the database is corrupted with a partial transaction.  Yes,uL > application design issues, but nothing about the surviability of a cluster to aI > single node going down addresses such problems.  Clusters are good, but 
 not magic.  J No, but they can come close if appropriate supporting services exist.  ForL example, as I mentioned earlier, if the cluster includes a facility that canK be charged with watching a specific application instance on a specific nodetJ and told to restart that application if it dies (and do so on another nodeK if the original host node dies), then a single-instance application with no J cluster awareness whatsoever can run close to non-stop (modulo the time itE takes to restart itself, and assuming that it does not fail by, e.g.,yI entering an internal infinite loop which the supervising facility doesn't  recognize as a failure).   - bill   ------------------------------   Date: 5 Sep 2002 20:51:19 -0700l1 From: keithparris_NOSPAM@yahoo.com (Keith Parris)m Subject: Re: "inview" Articles= Message-ID: <cf15391e.0209051951.36201d9c@posting.google.com>r  u "Bill Todd" <billtodd@metrocast.net> wrote in message news:<%jPd9.444825$m91.17180118@bin5.nnrp.aus1.giganews.com>... ( > 'in general' if the application wasn'tL > written to allow multiple instances of itself then multiple such instancesE > will trip over themselves in updating any common files (such as thecN > hypothetical aforementioned appended-to log file):  either they'll open themM > prohibiting sharing (the safe way to do it) thus preventing other instanceseL > from using them, or they'll open them without locking at all (uncontrolledL > sharing) and screw them up internally with overlapping concurrent updates,I > or they'll open them appropriately shared (perhaps just from habit) but J > won't *use* the interlocked access in a manner that keeps their contentsL > consistent (because they aren't aware that other concurrent updates may be
 > occurring).m  F It appears there's some confusion here between basic multi-user access* considerations and cluster considerations.  C@ As I recall, the majority of applications written for VMS in theE pre-cluster timeframe did allow for multiple instances of themselves,f> otherwise they would have been pretty useless in a timesharingD environment where there were multiple simultaneous users of the same application program.  B With respect to multi-user access and contention: most programmers? simply took advantage of the default behavior of the high-leveloE languages on VMS: for example, if you wrote an interactive program inl= Fortran and read a record in an indexed file, that record wasnE automatically locked until you either wrote it back with an update or0B read a different record.  So multi-user data protection was there.  A That same behavior was preserved in a clustered environment whereSF different users ran the same program on different nodes, thanks to the: cluster-wide file system and the distributed lock manager.  F > Furthermore, you were responding to a statement that in order for anL > application 'written with no knowledge of clusters' to 'reap the benefits'E > of VMS clustering the application must be written to allow multiplebM > instances of itself.  If you wish to contest that assertion, please provideoN > examples of how a single-instance application can 'reap the benefits' of VMS* > clustering:  there may certainly be some  C First of all, I never said every application written in pre-clusteriB days could run on multiple nodes in a cluster -- certain features,C such as mailboxes, common event flag clusters, and global sections, C were never "cluster-ified".  But in my experience, most applicationn programs could.h  B But even if we accept your contention that most applications couldE only run on a single node even in a clustered environment, they still @ benefit from the clustered environment, because if that one nodeF fails, they can be restarted on another single node and continue.  AndA in a disaster-tolerant cluster, their data is preserved despite a0E disaster and they can be moved to run again, albeit on a single node,-D at the surviving site.  So they obtain benefits from clustering, and? from disaster-tolerant clustering, by virtue of running in that  environment.  M > clustering:  there may certainly be some (e.g., if VMS provides an externaloN > system mechanism that can be set up to monitor an application and restart itK > on another node if its host node fails, as some other clustering productso > do)f  : You mean like an application server running in batch using) $SUBMIT/QUEUE=some_generic_queue/RESTART?a  B Or connection to a LAT service that is offered by multiple clusterB nodes, and the user performs the "failover" by logging back in, at, which point he's directed to a working node?  C A big advantage of VMS clusters over most Unix clusters is that youaA most often don't NEED to monitor an application and restart it on A another node because it is already running on ALL of the nodes ato once.a. ----------------------------------------------. Keith Parris | parris at encompasserve dot org   ------------------------------  # Date: Fri, 06 Sep 2002 05:13:52 GMTu* From: "Bill Todd" <billtodd@metrocast.net> Subject: Re: "inview" Article A Message-ID: <kQWd9.198576$On.7934297@bin3.nnrp.aus1.giganews.com>   > "Keith Parris" <keithparris_NOSPAM@yahoo.com> wrote in message7 news:cf15391e.0209051951.36201d9c@posting.google.com...n7 > "Bill Todd" <billtodd@metrocast.net> wrote in message2? news:<%jPd9.444825$m91.17180118@bin5.nnrp.aus1.giganews.com>....* > > 'in general' if the application wasn'tD > > written to allow multiple instances of itself then multiple such	 instances G > > will trip over themselves in updating any common files (such as thenK > > hypothetical aforementioned appended-to log file):  either they'll open  themE > > prohibiting sharing (the safe way to do it) thus preventing other 	 instancess@ > > from using them, or they'll open them without locking at all
 (uncontrolled E > > sharing) and screw them up internally with overlapping concurrent  updates,K > > or they'll open them appropriately shared (perhaps just from habit) butlL > > won't *use* the interlocked access in a manner that keeps their contentsK > > consistent (because they aren't aware that other concurrent updates may. be > > occurring).n >nH > It appears there's some confusion here between basic multi-user access, > considerations and cluster considerations.   I don't believe so.e   >oB > As I recall, the majority of applications written for VMS in theG > pre-cluster timeframe did allow for multiple instances of themselves, @ > otherwise they would have been pretty useless in a timesharingF > environment where there were multiple simultaneous users of the same > application program.  G However, in most cases they allowed multiple *independent* instances ofiI themselves in multiple disjoint environments (e.g., user 'home' directoryhE sub-trees, though read-only portions of the application may have beenoB shared).  Which is effectively the same as being a single-instance. application in the context of this discussion.  L Or are you referring to multi-user (rather than multi-instance) applicationsK where a single process served multiple users?  They're also single-instanceyK applications in this context, though they obviously don't translate as well L to the cluster idiom because of the difficulty such a single instance has inF grabbing user input from and producing local output to any node in the cluster.   > D > With respect to multi-user access and contention: most programmersA > simply took advantage of the default behavior of the high-leveltG > languages on VMS: for example, if you wrote an interactive program in ? > Fortran and read a record in an indexed file, that record wasgG > automatically locked until you either wrote it back with an update oreD > read a different record.  So multi-user data protection was there.  H Irrelevant:  that may guarantee the internal structural integrity of theL file, but it does nothing to guarantee the *logical* integrity of the file'sI contents.  Only the application can do that, because only the application-K knows what defines that logical consistency - and if the application wasn'teG written under the assumption that multiple instances of itself would be3I updating shared files concurrently, the logical consistency of such filesr) will likely be trashed under such access.-   > C > That same behavior was preserved in a clustered environment wherenH > different users ran the same program on different nodes, thanks to the< > cluster-wide file system and the distributed lock manager.  I And that's why applications written to allow multiple instances to updatecB shared data consistently are precisely those applications that areK best-positioned to leverage the strengths of clustering transparently.  ButaK they don't get that way (save possibly by chance) just by using RMS and itseG transparent locking:  they have to have it designed into their internals functioning.   >nH > > Furthermore, you were responding to a statement that in order for anD > > application 'written with no knowledge of clusters' to 'reap the	 benefits' G > > of VMS clustering the application must be written to allow multiple:G > > instances of itself.  If you wish to contest that assertion, please7 provide@L > > examples of how a single-instance application can 'reap the benefits' of VMS , > > clustering:  there may certainly be some > E > First of all, I never said every application written in pre-clustersD > days could run on multiple nodes in a cluster -- certain features,E > such as mailboxes, common event flag clusters, and global sections,bE > were never "cluster-ified".  But in my experience, most applicationa > programs could.t  E All the above observations appear to apply to applications written tooK cooperate either with other instances of themselves or at least with *some*cH other processes on the system.  So not only did no one accuse David (forF that is whom I was responding to, not you) or, for that matter, you ofI saying anything resembling what you deny saying above, but neither is thenL ability of such applications to run in a cluster relevant to this particular sub-thread of discussion.m   >eD > But even if we accept your contention that most applications could< > only run on a single node even in a clustered environment,   Which of course I never said.n    they still B > benefit from the clustered environment, because if that one nodeC > fails, they can be restarted on another single node and continue.   H Just as they could be restarted on their own, rebooted node.  Neither isL transparent (your choice of characterization), nor occurs "without having toK modify the application code and without having to write failover scripts touE be invoked on a failure" (your words, though the batch-queue /RESTARTrJ approach, if it does this, might not be considered equivalent to writing a fail-over script, I suppose).u  J And looking back at that post I see a pre-echo of some of what I respondedH to (the second time, now) above:  "I can run it on multiple nodes in theI cluster at once, not just one node at a time, because the file system andoL RMS do the right thing."  I didn't respond to those words originally becauseL I interpreted them to mean "... because the file system and RMS do the right9 thing in a cluster just as they do on a single node for amI multi-instance-aware application" rather than "...because the file systemeL and RMS make it unnecessary for an application to be multi-instance-aware inE a cluster just as they do on a single node":  the former assertion isa correct, but the latter is not.x     And C > in a disaster-tolerant cluster, their data is preserved despite alG > disaster and they can be moved to run again, albeit on a single node, F > at the surviving site.  So they obtain benefits from clustering, andA > from disaster-tolerant clustering, by virtue of running in thatl > environment.  L That's arguably correct, but your original statement (even leaving aside theJ "without having to modify the application code and without having to writeE failover scripts to be invoked on a failure" part) was that you could D "*immediately* [emphasis added] reap the benefits of the VMS ClusterG platform's higher availability and scalability", not that such benefitscI would accrue some time down the road after a failure.  Combined with yourpJ comment about running on multiple nodes, that caused JF to reply that thisI was only true for multi-instance-aware applications:  he was correct, andt. that's what I've been pointing out ever since.   >sF > > clustering:  there may certainly be some (e.g., if VMS provides an externalE > > system mechanism that can be set up to monitor an application ando
 restart itD > > on another node if its host node fails, as some other clustering products > > do)r >i< > You mean like an application server running in batch using+ > $SUBMIT/QUEUE=some_generic_queue/RESTART?t  J If that does what I think it may (track the executing instance and restartJ it - possibly somewhere else - if it or its node fails), it would qualify.   >iD > Or connection to a LAT service that is offered by multiple clusterD > nodes, and the user performs the "failover" by logging back in, at. > which point he's directed to a working node?  E Nope:  that reconnects the user to the cluster, but doesn't reconnectdG him/her to what s/he was doing unless the server-side element either isoJ multi-instance already or something else brings a new replacement instance up on the new server node.   >pE > A big advantage of VMS clusters over most Unix clusters is that yourC > most often don't NEED to monitor an application and restart it on C > another node because it is already running on ALL of the nodes ate > once.   I Nope:  multi-instance applications are supported on Unix clusters just asRL they are on VMS clusters, though those that on VMS might elect to coordinateG through the DLM typically do so through use (as I observed before, somesH might call it 'misuse', but it does work) of the file system in the Unix cluster environment.   - bill   ------------------------------  # Date: Fri, 06 Sep 2002 01:23:00 GMTt! From: rob.buxton@wcc.spam.govt.nzc! Subject: Re: After Memory Upgrade % Message-ID: <3d7802e9.167229212@news>e  @ IIRC the SYSGEN Parameter PHYSICALPAGES used to have be manually. adjusted for some of the older versions of VMS  , Could this also be the cause of the problem?  B Maybe there's a Set Physicalpages in Modparams that has been there since earlier days.f      A On 5 Sep 2002 10:56:16 -0700, keithparris_NOSPAM@yahoo.com (Keitho Parris) wrote:  N >Jim <jim@jim.com> wrote in message news:<Xns927F74B7FFA90jimwork@10.1.0.4>...N >> Jumping from screen to screen takes a long time, there seems to be alot of N >> ENQ/DEQ (ie. file locks?) and the command $monitor cluster/interval=1 does - >> NOT update every 1 second like it used to.s >hE >Given that an AUTOGEN was done, you might want to see what parameteroA >changes AUTOGEN did (comparing SYS$SYSTEM:SETPARAMS.DAT files ofeF >different dates with $DIFFERENCES/PARALLEL might be the easiest way). > G >You may have to just start from scratch the your performance analysis,dF >using the techniques in the performance manual in the documentation. F >It is conceivable that the performance problem has nothing to do with; >the memory upgrade, other than being coincidental in time.k > D >If you don't already have DECamds / Availability Manager installed,: >you might find that tool helpful, and it's free with VMS. > D >A generic technique I've found very useful is to identify a processG >that seems slow, and then try to identify what it's waiting on most ofs> >the time.  look at it under SDA ($ANALYZE/SYSTEM) using a DCLC >procedure that loops, looking at SDA> SHOW PROCESS/CHANNELS output E >looking for "Busy" channels, which can indicate I/O bottlenecks, andaC >SDA> SHOW PROCESS/LOCKS looking for lock requests that are queued,s$ >which can indicate lock contention. >nD >With regard to the ENQ/DEQs, you didn't say what version of VMS youA >were running, but if you're running 7.2-2 or higher, the new SDA G >extension LCK (under $ANALYZE/SYSTEM, do LCK at the SDA> prompt to seesG >LCK's online help) can be used to show what is going on in the area ofcD >locking.  SDA> LCK ACTIVE will show the most-active lock trees on aG >given node (and what file they correspond to).  LCK can also trace allaF >lock requests, if it comes down to that.  If you're not running 7.2-2@ >or above, or would like a cluster-wide locking activity summaryD >instead of the per-node info that LCK gives, use the LOCK_ACTV tool/ >from the VMS Freeware CD under [KP_LOCKTOOLS].t > G >With regard to the slowness observed in MONITOR CLUSTER, you might see E >if a network problem might be contributing to the cluster slowness. oE >There, it would be helpful to know if you are using DECnet or TCP/IPiE >for MONITOR CLUSTER, and whether or not your cluster uses the LAN asf >the cluster interconnect./ >-----------------------------------------------/ >Keith Parris | parris at encompasserve dot orgm   ------------------------------   Date: 5 Sep 2002 22:46 CDT' From: carl@gerg.tamu.edu (Carl Perkins)9! Subject: Re: After Memory Upgrader, Message-ID: <5SEP200222462861@gerg.tamu.edu>  5 keithparris_NOSPAM@yahoo.com (Keith Parris) writes...1N }Jim <jim@jim.com> wrote in message news:<Xns927F74B7FFA90jimwork@10.1.0.4>...N }> Jumping from screen to screen takes a long time, there seems to be alot of N }> ENQ/DEQ (ie. file locks?) and the command $monitor cluster/interval=1 does - }> NOT update every 1 second like it used to.h } E }Given that an AUTOGEN was done, you might want to see what parametertA }changes AUTOGEN did (comparing SYS$SYSTEM:SETPARAMS.DAT files ofaF }different dates with $DIFFERENCES/PARALLEL might be the easiest way). } G }You may have to just start from scratch the your performance analysis,jF }using the techniques in the performance manual in the documentation. F }It is conceivable that the performance problem has nothing to do with; }the memory upgrade, other than being coincidental in time.t/ }Keith Parris | parris at encompasserve dot orge  & Coincidences can be quite irritating.   C Is it possible that someone did the leap in logic "now we have moreMD memory, so I can adjust the working set quota for this account to beF REALLY HUGE"? Amongst other things, a proccess that used to be limitedC by pagefaulting a lot could now be limited only by CPU speed (or ita? could be eating up all that shiny new memory, and more, leaving + everybody else even worse off than before).a  B If the autogen was done using feedback data, you have to ask "what feedback data?"   @ Specifically, it wouldn't happen to have been feedback data that= was collected before the new memory was added, would it? Thist would clearly be inappropriate.o  @ A complete list of everything that was done after the memory wasC added (or shortly before, in anticipation of it) could be helpfull,rA but is not likely to be available. A single typo in MODPARAMS.DAT C could mess things up pretty bad (adding a digit to, or removing one . from, a system parameter's value for example).  H Review MODPARAMS.DAT and reconsider any MAX_* settings and any "FOO = x"C hardcoded settings. You may benefit from increasing these now. Also J consider the values used for various MIN_* settings as it may be benficialF to increase those now too. Then autogen with feedback data accumulated since the memory upgrade.i  I It is also a good time to review working set quota settings for accounts.e7 You may se a benefit from increasing some of these too.u   --- Carl   ------------------------------  # Date: Thu, 05 Sep 2002 18:27:07 GMT + From: Jeff Campbell <jcampbell@ins-msi.com>e Subject: Re: am/pm* Message-ID: <3D7796A1.FBFD965@ins-msi.com>   Here's one'u  <    <http://www.universal-radio.com/catalog/clocks/4437.html>  
 Jeff Campbellr n8wxs@arrl.net   Brian Tillman wrote: > G > >Well, I'm 100% sure that there have *never* been produced any analogtH > >(the normal round type) clock/watch that have a 24h dial, even though1 > >there are many countries that don't use am/pm.a > M > Then you are 100% wrong.  I have seen them.  There are two types.  One, theaN > more common, has 1-12 in large numerals and 13-00 in small numerals radiallyI > inward.  The other (and, as I say, I have seen one, has 01 to 12 on theiN > right half f the dial (with 12 in the bottom-most position) and 13-00 on theL > left side (with 00 at the top).  The hour hand moves once around the clock > in 24 hours. > H > >Some wrist watches have both 1-12 and 13-24 printed on the dial, and,F > >if not, you probably don't need a calculator to do the "conversion"
 > >anyway. > K > You contradict yourself.  Above you say you're "100% sure" that none haveuC > ever been produced and here you way that they have.  Which is it?g > --C > Brian Tillman                   Internet: tillman_brian at si.comgC > Smiths Aerospace                          tillman at swdev.si.comn? > 3290 Patterson Ave. SE, MS      Addresses modified to prevente> > Grand Rapids, MI 49512-1991     SPAM.  Replace "at" with "@": >        This opinion doesn't represent that of my company   ------------------------------   Date: 5 Sep 2002 16:42:53 -0600 - From: Kilgallen@SpamCop.net (Larry Kilgallen)B< Subject: Beware of disappearing St. Louis hotel reservations3 Message-ID: <6GDl3x$5EM1r@eisner.encompasserve.org>a  E            <<< EISNER::DRA1:[NOTES$LIBRARY]SYMPOSIA_PUBLIC.NOTE;1 >>> G          -< Symposia_Public / HP-ETS 2002 / October 6-11 / St. Louis >-BP ================================================================================P Note 592.0     Beware of disappearing St. Louis hotel reservations    No repliesP EISNER::KILGALLEN "Larry Kilgallen, LJK Software"    39 lines   5-SEP-2002 16:41P --------------------------------------------------------------------------------P >    <<< Note 579.11 by EISNER::KILLEEN "Jeff Killeen CETS Business Manager" >>> > Q > > Then I have to ask it.  Will there be a good ol' "pen and paper" registrationu > > process? > : > Business decision - nope - telephone only alternative...  / A funny thing happened on my way to St Louis...g  G Since I picked up my airline tickets today, I was reminded to call backaE to 888-219-7186 in order to get the confirmation number for the hotelaD reservation I made last Friday.  At the time I made the reservation,D they said they could not give a confirmation number then but I could! get it by calling back this week.e  C Much to my surprise, they had lost all traces of my reservation andaI even my registration for the Symposium.  Eventually there was a three-waya phone call involving:n  ' 	800-516-6582 vendor (identity unknown) $ 	888-219-7186 vendor (Merits Travel)0 	Victim trying to go to the DECUS Symposium (me)  2 There seem to have been two failure of the system:  8 	800- vendor failed to notify 888- vendor of a telephone 		registration.s  7 	888- vendor totally lost what they told me last Fridaye  		was a valid hotel reservation.  F The next step is that the supervisor of the 800- vendor representativeE I talked to is supposed to give back-channel notification to the 888-mF vendor so that when I call back tomorrow they will at least agree thatE I am eligible to get a hotel reservation, if not that they claimed to  have gotten me one.   C Certainly when they claim to have accomplished this, I will also bep verifying it with the hotel. -- tN ==============================================================================I The Boulder Pledge: "Under no circumstances will I ever purchase anythinggJ      offered to me as the result of an unsolicited email message. Nor willI      I forward chain letters, petitions, mass mailings, or virus warningsdH      to large numbers of others. This is my contribution to the survival      of the online community."N ==============================================================================   ------------------------------   Date: 5 Sep 2002 16:03:31 -0700d, From: srp336@getcoactive.com (Steve Pfister)0 Subject: Burning an image with an unknown format= Message-ID: <45126e60.0209051503.2a33d981@posting.google.com>o  B I've got some sort of OpenVMS 7.2 image file that someone was niceE enough to supply me with a few months ago. I was able to successfullyhE use it to install a copy on an emulator. Then, I bought a Microvax onrD eBay and wanted to reinstall the copy of OpenVMS on it. I don't haveA the person's name anymore, and I don't think he ever told me whatcF created it (assuming he knew in the first place...). Try as I might, IA couldn't get anything to recognize it (I'm running w2k). I put itnE aside because the new hobbyist kits would be available soon (that waseE around May, and the kits were supposed to be out in July). Here it isaE September now, and the Microvax is gathering dust in my basement. I'd E like to get something out of the image...I don't want to have to sell  the machine.  A Is there any way I can figure out how to burn this image to a CD?a   Thanks!s   ------------------------------  % Date: Thu, 05 Sep 2002 19:35:01 -0400 * From: "Stanley F. Quayle" <stan@stanq.com>4 Subject: Re: Burning an image with an unknown format. Message-ID: <3D77B1E5.25300.27FC444@localhost>  , On 5 Sep 2002 at 16:03, Steve Pfister wrote:C > Is there any way I can figure out how to burn this image to a CD?   B I do this all the time.  You need a SCSI CD burner, and a copy of  cdrecord for VMS:o  1    ftp://v36.chemie.uni-konstanz.de/cdrecord_vms/s  B Alternatively, you could boot your PC to Linux and burn the image  file, also using cdrecord.  
 --Stan Quaylel Quayle Consulting Inc.  
 ----------C Stanley F. Quayle, P.E. N8SQ  +1 614-868-1363  Fax: +1 614 868-1671-1 8572 North Spring Ct. NW, Pickerington, OH  43147n= Preferred address:  stan@stanq.com       http://www.stanq.comf   ------------------------------  $ Date: Thu, 5 Sep 2002 21:04:15 -04002 From: "warren sander" <sander.ma.ultranet@rcn.com>G Subject: Re: Can one still vest Vax executables onto Alpha OpenVMS V7.3 + Message-ID: <al8v2k$avr$1@bob.news.rcn.net>s  G try http://www.openvms.compaq.com/openvms/products/omsva/omsva.html forn0 OpenVMS migration software for vax to alpha v1.2 it's new, updated decmigrate.e   -warrenb   --B ------------------------------------------------------------------9 Warren Sander                           OpenVMS Marketing B Hewlett-Packard Company         Work:  warren.sander@remove.hp.comG 200 Forest Street MR01-3/J1     Personal: sander@remove.ma.ultranet.comc3 Marlboro, MA 01752                   (508) 467-4875 5    My opinions are my own and I only speak for myselfe,          Read http://www.openvms.compaq.com/B ------------------------------------------------------------------< "David J. Dachtera" <djesys.nospam@fsi.net> wrote in message! news:3D76C443.87D37E6B@fsi.net...o > Hoff Hoffman wrote:e > >i7 > > In article <3D6DB653.48AB3722@telocity.com>, Kolotht <koloth@telocity.com> writes:o! > > > or earlier?  Why the limit?E > >rK > >   Please acquire and read DECmigrate documentation for specific details K > >   and for the current release and the product support status.   Thanks!cL > >   The DECmigrate materials and information is available via a URL listedG > >   in the OpenVMS FAQ; in the OpenVMS Frequently Avoided Quotations.  >a  > Quotations from Chairman Hoff? >r > -- > David J. Dachterao > dba DJE Systemsr > http://www.djesys.com/ >e* > Unofficial Affordable OpenVMS Home Page:! > http://www.djesys.com/vms/soho/u   ------------------------------  % Date: Thu, 05 Sep 2002 14:54:32 -0400i- From: JF Mezei <jfmezei.spamnot@videotron.ca>n( Subject: Cancelling writes to a log file, Message-ID: <3D77A864.E572FE1A@videotron.ca>  & Is it possible to do the following : ?  K as a transaction is being processed, a whole bunch of logging is written tooH file. If the transaction is succesful, then that logging is erased and a9 simpler one line message is written to indicate success. e  C This way, if the transaction fails (or programs crashes), plenty ofdK information is available in the logfile. But if everything works fine, then I that redundant information is removed to remove clutter from the logfile.e  K Could one use "ftell" at the start of a transaction, write a whole bunch ofaE stuff, and if succesful, then "fseek" to go back to the start of that*J transaction's log and then write the one line "success" message and signal7 that this is the new end of file until the next write ?n   ------------------------------   Date: 5 Sep 2002 14:35:38 -0700e' From: geff@excite.com (geff@excite.com)l! Subject: Re: Compression Utility?y= Message-ID: <2119a17f.0209051335.33e6e689@posting.google.com>:  A Thanks a million.  I was able to point to www.process.com for the.D info-zip files and was able to successfully link them on my system. 0 Very useful resource.  That for the good advice.   geff     David Gray <GrayD@turpinltd.com> wrote in message news:<79DE14DAA1A8D3119B6E00805FEB980BAE4276@zakary.turpin-distribution.com>... M > How about Zip & Unzip from http://www.info-zip.org/pub/infozip/Zip.html#VMSU >  > -----Original Message-----3 > From: FalconFan [mailto:falconfan@falconfan.com]   > Sent: 05 September 2002 05:34l > To: Info-VAX@Mvb.Saic.Comt > Subject: Compression Utility?  > A > I'm in need of a VMS compression utility.  I need to be able totE > compress a large binary file then ftp it from one server to anothertE > server then uncompress.  I don't have a lot of experience in VMS incF > this area.  Anyone care to recommend a good compression utility (and* > where to find it) for a beginner to use? >  > Thanks for any info :) > FalconFann   ------------------------------  # Date: Fri, 06 Sep 2002 02:22:32 GMTp1 From: "David J. Dachtera" <djesys.nospam@fsi.net>c! Subject: Re: Compression Utility?d' Message-ID: <3D7816BC.CF7EB869@fsi.net>e   "geff@excite.com" wrote: > G > I've tried info-zip and can't get it to work.  Any other suggestions?l   Yes. See these URLs:4 http://www.djesys.com/vms/support/zipunzip/index.htm0 http://www.djesys.com/vms/hobbyist/zipunzip.html http://www.djesys.com/zip.html  http://www.djesys.com/unzip.html  ? If you're still having problems after reviewing these, write mefG privately - how to demung the reply-to should be obvious - and I'll try- to help.  F VMS ZIP and UNZIP are very widely used and can be very helpful. You'll= do well to master any challenges you may be having with them.3   -- . David J. Dachtera  dba DJE Systemsa http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/l   ------------------------------   Date: 5 Sep 2002 14:41:39 -0700c' From: geff@excite.com (geff@excite.com)  Subject: Convert Blocks to MB5= Message-ID: <2119a17f.0209051341.237f1101@posting.google.com>   D I checked the faq and it doesn't mention how to do this.  Is there aB simple way to convert the block figures I get when I do a "dir" to# megabytes which is the way I think?    Thanks for any info, geff   ------------------------------  % Date: Thu, 05 Sep 2002 18:22:11 -0400o- From: JF Mezei <jfmezei.spamnot@videotron.ca>N! Subject: Re: Convert Blocks to MBa, Message-ID: <3D77D902.CAB99ADA@videotron.ca>   "geff@excite.com" wrote: > F > I checked the faq and it doesn't mention how to do this.  Is there aD > simple way to convert the block figures I get when I do a "dir" to% > megabytes which is the way I think?e   1 block is 512 bytes.e 2 blocks is 1 k 1 2*1024 blocks is 1 meg (one "tousand" kilobytes).   E I.e, divide number of blocks by 2048 and you get number of megabytes. = Multiply number of blocks by 512 and you get number of bytes.a  O Note that this is an approximation since the last block may not be fully used. A  2 DIR/ALL provides blocks used and blocks allocated.  V And the disk cluster factor may result in a file containg 10 bytes taking up 3 blocks.   ------------------------------  % Date: Thu, 05 Sep 2002 20:40:51 -0400a2 From: rdeininger@mindspring.com (Robert Deininger)! Subject: Re: Convert Blocks to MBhK Message-ID: <rdeininger-0509022040510001@1cust113.tnt2.nashua.nh.da.uu.net>l  5 In article <3D77D902.CAB99ADA@videotron.ca>, JF Mezeic% <jfmezei.spamnot@videotron.ca> wrote:/   > 3 >DIR/ALL provides blocks used and blocks allocated.,   Correction:r   DIR/SIZE=ALL  ( In this case, the "SIZE" does matter....   ------------------------------   Date: 5 Sep 2002 23:00 CDT' From: carl@gerg.tamu.edu (Carl Perkins) ! Subject: Re: Convert Blocks to MB , Message-ID: <5SEP200223002856@gerg.tamu.edu>  + geff@excite.com (geff@excite.com) writes...oE }I checked the faq and it doesn't mention how to do this.  Is there ahC }simple way to convert the block figures I get when I do a "dir" tow$ }megabytes which is the way I think? }  }Thanks for any info,L }geffs   Since 1 block = 512 bytes...  C Divide blocks by 2048 to get binary megabytes (1048576 bytes each).w   or  H Divide blocks by 1953.125 to get decimal megabytes (1000000 bytes each).   or  D Devide blocks by 2000 to get a quick approximation that is a smidgenD less than half way between these two. (Much easier to do in you head than the other two.)   --- Carl   ------------------------------  % Date: Fri, 06 Sep 2002 00:15:45 -0400-) From: FalconFan <falconfan@falconfan.com> ! Subject: Re: Convert Blocks to MB-8 Message-ID: <suagnukih9ticp1o1lh5ab2i1jn7f6rhrn@4ax.com>  ; Hey - thanks for your help.  I think I understand it now :)g   Geff  F On 5 Sep 2002 14:41:39 -0700, geff@excite.com (geff@excite.com) wrote:  E >I checked the faq and it doesn't mention how to do this.  Is there a0C >simple way to convert the block figures I get when I do a "dir" toi$ >megabytes which is the way I think? >Q >Thanks for any info,t >geffa   ------------------------------  # Date: Fri, 06 Sep 2002 02:01:34 GMTt1 From: "David J. Dachtera" <djesys.nospam@fsi.net>u. Subject: Re: Copy of OpenVMS hobbyist CD in UK' Message-ID: <3D7811D5.80F6DCF1@fsi.net>D   issinoho wrote:e > D > Technically yes, but many software packages which offer an imagingE > feature, use the term 'ISO image' liberally, in this case WinImage.u  G My advice would be to write to the software company and point out theirsH error. Classifying dism images, regardless of format, as "ISO images" isB a *GROSS* error and demonstrates one's lack of sophistication, not+ mention casting doubt on one's credibility.e  D > Interestingly, it took me a little while to find a PC package thatH > would image an ODS disk properly, most packages fail to even recogniseB > that a disk is in the drive! Does anyone else have any favourite > methods of achieving this?   Yes. I use VMS. See these URLs:e' http://www.djesys.com/vms/cdrom.html#pd $ http://www.djesys.com/vms/cdrom.html  G You can use the technique described for physical disks for image largereF than a CD-ROM, of course. They just take longer to transfer around. ToC write it back to a physical disk, just reverse the procedure shown:   1 1. MOUNT the target disk /FOREIGN, write enabled.w% 2. COPY the image file to the device.o 3. DISMOUNT the target disk.  A Note, however, that the source and target disk geometries must beo compatible.   E To eliminate this limit, use VMS to make a /IMAGE BACKUP of the disk.nE Then, you can restore the resulting saveset to any disk big enough toiC contain the contents of the saveset. To transer the saveset around,eE especially if non-VMS systems will participate in the transfer at anyOB point, use ZIP for VMS to pack the saveset into a platform neutral6 archive. Remember to use the /VMS (or "-V") qualifier.  $ > I'm also aware of MKIMAGE from theF > Softresint people which seems to mimic the *NIX dd command, although6 > they have now removed this from their download area.   Have a look at:i http://www.gearcd.com/  H I'm currently using an older version of Gear for Windows. It works quite$ nicely for duping ODS CDs onto CD-R.   -- 0 David J. Dachterar dba DJE Systems  http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/e   ------------------------------  # Date: Fri, 06 Sep 2002 02:03:43 GMTf1 From: "David J. Dachtera" <djesys.nospam@fsi.net> . Subject: Re: Copy of OpenVMS hobbyist CD in UK' Message-ID: <3D781259.DC3D5084@fsi.net>m   issinoho wrote:k > 
 > Good grief!2E > Fine, but 'ISO image' has been taken into the public domain in mucha  1 How many "wrongs" does it take to make a "right"?T  C > the same way as 'Hoover' or 'Xerox' has. Pedantry is fine if it'slH > educational, but let's not start trying to score points of one another9 > - we can't all guarantee 'correct usage' on all topics.e  C Well, you can do what everyone else does and get what everyone elsed3 gets: mediocrity. Excellence requires a finer edge.u   -- c David J. Dachtera. dba DJE Systemss http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/l   ------------------------------  # Date: Fri, 06 Sep 2002 02:05:35 GMT.1 From: "David J. Dachtera" <djesys.nospam@fsi.net> . Subject: Re: Copy of OpenVMS hobbyist CD in UK' Message-ID: <3D7812C9.99678240@fsi.net>a   Larry Kilgallen wrote: > f > In article <d0141774.0209050103.6242d98d@posting.google.com>, issinoho@slayme.com (issinoho) writes: > > Good grief! G > > Fine, but 'ISO image' has been taken into the public domain in much E > > the same way as 'Hoover' or 'Xerox' has. Pedantry is fine if it'shJ > > educational, but let's not start trying to score points of one another; > > - we can't all guarantee 'correct usage' on all topics.u > H > We are (or some of us are) trying to communicate technical informationK > accurately.  If you want to be fuzzy, use a fuzzy term like "disc image".A > H > Or should we ask people to pose their comments in terms of their "VAX" > even when they mean "Alpha" ?=  F The "generic" reference "VAX" means everything from the machine itselfG (VAX or Alpha) to the operating system and the applications and layeredo products, taken collectively...y   ...among the unsophisticated.i   -- - David J. Dachtera- dba DJE Systems- http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/J   ------------------------------  # Date: Fri, 06 Sep 2002 02:12:25 GMTi1 From: "David J. Dachtera" <djesys.nospam@fsi.net>u. Subject: Re: Copy of OpenVMS hobbyist CD in UK' Message-ID: <3D781461.9F2A6E41@fsi.net>    Larry Kilgallen wrote: > ] > In article <3D76D029.EDBD2B0E@fsi.net>, "David J. Dachtera" <djesys.nospam@fsi.net> writes:2 > > Tim Shoppa wrote:  > P > >> And to be even more pedantic, ISO-9660 is known in the source code listings# > >> as "ODS-3" in a few places :-)y > >dL > > ...which makes it even more amazing that ODS-3 is not supported natively% > > by the file primitives or by RMS.  > E > So far as I can see it is supported by RMS and the file primitives,iE > but only for read.  The way RMS finds the protection of an ISO-96605! > file, for instance, is via QIO.  > C > Writing is considerably more difficult, since the ISO-9660 formatyC > does not lend itself to on-the-fly modifications used by RMS when C > writing.  Even contemplating a software program to allow multipleiE > simultaneous writers to modify an ISO-9660 volume is mind-boggling.   7 I've always tended to view that as a practical case foreH reverse-engineering. If we knew whether and/or where RMS expects to findF the appropriate metadata, it would be possible to write an MKISOFS_VMSE variant that could place the necessary bits where RMS expects to finds them.   G The resultant FS would be RO (read-only), of course, but that's because = of the design limitations of ISO-9660. AFAIK: files *MUST* bewH contiguous, cluster size *MUST* be 4 (not real sure about that one), andC directory entries must appear in ascending ASCII sequence. I may bee. quite wrong on all of those counts, of course.   --   David J. Dachteras dba DJE Systemss http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/a   ------------------------------   Date: 5 Sep 2002 22:57:13 -0600o- From: Kilgallen@SpamCop.net (Larry Kilgallen)=. Subject: Re: Copy of OpenVMS hobbyist CD in UK3 Message-ID: <c9dm+nAPnNVB@eisner.encompasserve.org>   [ In article <3D781461.9F2A6E41@fsi.net>, "David J. Dachtera" <djesys.nospam@fsi.net> writes:g > Larry Kilgallen wrote: >> g^ >> In article <3D76D029.EDBD2B0E@fsi.net>, "David J. Dachtera" <djesys.nospam@fsi.net> writes: >> > Tim Shoppa wrote: >> eQ >> >> And to be even more pedantic, ISO-9660 is known in the source code listings $ >> >> as "ODS-3" in a few places :-) >> >M >> > ...which makes it even more amazing that ODS-3 is not supported natively'& >> > by the file primitives or by RMS. >> tF >> So far as I can see it is supported by RMS and the file primitives,F >> but only for read.  The way RMS finds the protection of an ISO-9660" >> file, for instance, is via QIO. >> hD >> Writing is considerably more difficult, since the ISO-9660 formatD >> does not lend itself to on-the-fly modifications used by RMS whenD >> writing.  Even contemplating a software program to allow multipleF >> simultaneous writers to modify an ISO-9660 volume is mind-boggling. > 9 > I've always tended to view that as a practical case fortJ > reverse-engineering. If we knew whether and/or where RMS expects to findH > the appropriate metadata, it would be possible to write an MKISOFS_VMSG > variant that could place the necessary bits where RMS expects to find. > them.   D Ok, I think we are on the same wavelength.  I had thought you wantedA the actual writing of the CDROM to be done through the normal RMS C primitives for writing an ODS-2 disks - e.g., squeeze this new filerC into the middle of and existing directory while other processes arenB actively reading it.  The Path Table structure in ISO-9660 sort of@ guarantees such an update cannot be done safely in a manner thatE would always produce a valid ISO-9660 disc (image) should the machine8' crash during the writing of the update.r  I > The resultant FS would be RO (read-only), of course, but that's because.? > of the design limitations of ISO-9660. AFAIK: files *MUST* be1J > contiguous, cluster size *MUST* be 4 (not real sure about that one), andE > directory entries must appear in ascending ASCII sequence. I may be-0 > quite wrong on all of those counts, of course.  > There are three "levels" of ISO-9660 interchange.  At level 3,@ files do not have to be contiguous.  This is crucial if you need? to write a file of more than 600 MB to a CDROM volume set.  The-= maximum number of ISO-9660 volumes allowed in a volume set is D 65535, but that likely will exhaust the MTBF of your CD-R burner :-)  D ISO-9660 allows it's equivalent of "cluster size" for files to be asB low as 512 bytes, while that for directories must be at least 2048@ bytes, and the two can differ on a single disc.  But that is the@ standard, and in practice ISO-9660 discs are always written with? both numbers being 2048.  While a primary goal might be just toa> satisfy VMS needs, it would be good if simple files (e.g., not@ indexed or relative files) could also be read by other operatingB systems.  Since existing discs use a 2048 byte logical block size,A operating systems don't have much practice with anything else, soiD there is not much sense in tempting fate by writing with a different logical block size.s  ? You point out an interesting nuance about directory ordering --m> unlike VMS the "." and ";" characters are considered according! to the normal collating sequence.e   ------------------------------  # Date: Fri, 06 Sep 2002 01:54:33 GMT * From: "Bill Todd" <billtodd@metrocast.net>F Subject: Curly's wisdom on how to handle HP's Enterprise Systems GroupC Message-ID: <tVTd9.448714$m91.17412495@bin5.nnrp.aus1.giganews.com>-  L http://investor.cnet.com/investor/brokeragecenter/newsitem-broker/0-9910-108 2-20371578-0.html?tag=ltnc  E "The key to turning the Enterprise Systems Group (ESG) around is cost H cutting. He plans to take out $150-200 million in the next two quarters;K some 70% of the planned 15,000 headcount reduction could be in ESG. He saiduA that ESG should be able to make money with no revenue growth. Anye& improvement can be quickly leveraged."  % The news just keeps getting better...r   ------------------------------  $ Date: Thu, 5 Sep 2002 22:50:06 -0400* From: "r_mclean" <bobmclean@bellsouth.net>J Subject: Re: Curly's wisdom on how to handle HP's Enterprise Systems GroupA Message-ID: <zvUd9.15518$%P6.6376714@e3500-atl1.usenetserver.com>t  K Well, about 7,000 have been "workforce reduced" as of today, with about 200oL out of ESG, so the 70% number (which is 10,500) is impossable to hit. And heH is right ESG should be able to make money at the current level of sales.    5 "Bill Todd" <billtodd@metrocast.net> wrote in message.= news:tVTd9.448714$m91.17412495@bin5.nnrp.aus1.giganews.com...s >-L http://investor.cnet.com/investor/brokeragecenter/newsitem-broker/0-9910-108 > 2-20371578-0.html?tag=ltnc >.G > "The key to turning the Enterprise Systems Group (ESG) around is cost J > cutting. He plans to take out $150-200 million in the next two quarters;H > some 70% of the planned 15,000 headcount reduction could be in ESG. He saidC > that ESG should be able to make money with no revenue growth. Anyi( > improvement can be quickly leveraged." >p' > The news just keeps getting better...- >  >r >    ------------------------------  # Date: Fri, 06 Sep 2002 03:24:57 GMT5* From: "Bill Todd" <billtodd@metrocast.net>J Subject: Re: Curly's wisdom on how to handle HP's Enterprise Systems GroupC Message-ID: <deVd9.466110$2p2.18453746@bin4.nnrp.aus1.giganews.com>   5 "r_mclean" <bobmclean@bellsouth.net> wrote in messaget; news:zvUd9.15518$%P6.6376714@e3500-atl1.usenetserver.com...rI > Well, about 7,000 have been "workforce reduced" as of today, with aboute 200rG > out of ESG, so the 70% number (which is 10,500) is impossable to hit.-  E Well, yuh - I noticed that as well.  But my comment was about Curly'shK expressed attitude, not his competence (which we already understand all tooD well).   - bill   ------------------------------  % Date: Fri, 06 Sep 2002 00:35:06 -0400'- From: JF Mezei <jfmezei.spamnot@videotron.ca>eJ Subject: Re: Curly's wisdom on how to handle HP's Enterprise Systems Group, Message-ID: <3D783076.D4AD4EBE@videotron.ca>   Bill Todd wrote:G > "The key to turning the Enterprise Systems Group (ESG) around is costpJ > cutting. He plans to take out $150-200 million in the next two quarters;M > some 70% of the planned 15,000 headcount reduction could be in ESG. He saidw? > that ESG should be able to make money with no revenue growth.     S That would mean some 10500 employees losing their jobs in enterprise systems group.t  N How many  wintel replated employees are in ESG ? (eg: proliant and HP's wintel server products ?).-  J If they eliminate only the HP wintel server stuff, how many employees does that represent ?  H Considering that the "true" enterprise products are all undergoing IntelM funded transitions to that IA64 thing, there wouldn't be much in cost savingst by cutting them.  H However, Capellas opened his big inept mouth again. To him "enterprise =@ wintel", but to customers "entreprise = everything BUT wintel".   N he should have been more specific. For instance, he could have said: "Some 70%H of employee reductions will come through the rationalisation of industryI standard servers in the ESG". That would have sent a clear message to the:I "real" enterprise customers that their product was safe for now. CapellasaT wording added uncertaintly to HP's true intentions with the real enterprise systems.   ------------------------------  # Date: Fri, 06 Sep 2002 02:28:34 GMT01 From: "David J. Dachtera" <djesys.nospam@fsi.net>.& Subject: Re: definition of SYS$SYSROOT' Message-ID: <3D78182D.B64AFDF8@fsi.net>m   Phillip Helbig wrote:y > F > At various times, there have been discussions about (system) logicalD > names which are search lists, in particular why SYS$SYSROOT is NOTC > defined as SYS$SPECIFIC,SYS$COMMON, but rather as <translation ofc > SYS$SPECIFIC>,SYS$COMMON.b > 0 > A google search turns up WAY too many threads. > I > I think Alan Feldman(?) recently posted a summary of this, but I didn't)+ > find it with a search (admittedly quick).- > J > Can someone point me to a concise summary which has no wrong information > and includes historical info?a > H > On a (perhaps) related note: since the logicals which use specific andE > common stuff are defined as search lists anyway, why not just leaven? > VMS$COMMON in the search list, i.e. why have the [.SYSCOMMON]-H > subdirectory be a link to the VMS$COMMON directory?  Does this perhapsI > go back to a time when SET FILE/ENTER existed but search lists did not?g  F Possibly, but I encountered a system where I'm currently working whereF the VMS$COMMON directory was MISSING! ...and the system booted and ran
 just fine!   -- f David J. Dachtera  dba DJE Systemsm http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/.   ------------------------------   Date: 6 Sep 2002 05:24:16 GMTc7 From: yehavi@vms.huji.ac.il (Yehavi Bourvine (58-4279))g) Subject: Disk size limit on Alpha-500au ?e$ Message-ID: <2002Sep6.052416@hujicc>   Hello,K   I had a broken 9GB SCSI disk on Alpha 500au. I tried placing a 36GB driveoI instead of it; the VMS recognised it but placed it "offline". Only when I K placed another 9GB disk there it came ok. I tried upgrading the firmware ton0 version 6.0 (the latest one) but it didn't help.)   The VMS version is V7.1-1H2.  Any idea?a  <                                            Thanks, __Yehavi:   ------------------------------  % Date: Fri, 06 Sep 2002 07:38:21 +0200e From: Dirk Munk <munk@home.nl>- Subject: Re: Disk size limit on Alpha-500au ?h1 Message-ID: <J5Xd9.1571$H6.275715@zwoll1.home.nl>g    Yehavi Bourvine (58-4279) wrote: > Hello,M >   I had a broken 9GB SCSI disk on Alpha 500au. I tried placing a 36GB drive-K > instead of it; the VMS recognised it but placed it "offline". Only when IeM > placed another 9GB disk there it came ok. I tried upgrading the firmware toD2 > version 6.0 (the latest one) but it didn't help.+ >   The VMS version is V7.1-1H2.  Any idea?a > > >                                            Thanks, __Yehavi:  N You don't mention what kind/make 36 GB disk you're trying to use. Most likely K the SCSI driver isn't happy with the responses it gets from the drive, and t places the drive in "offline".  N There is no practical limit to the disk size. In other words there is a limit / but there are no disks that size for sale :-) .m  L With "firmware V6.0" you most likely mean that you used the firmware CD 6.0. When you issue:l  / 	$ write sys$output f$getsyi("console_version")D  " you should get V7.2-1 as response.   ------------------------------  $ Date: Tue, 3 Sep 2002 19:58:05 -0400, From: "Francisco Ortega" <fortega@iblues.cc>! Subject: Re: Help installing PinesA Message-ID: <ZJQd9.13824$%P6.5930602@e3500-atl1.usenetserver.com>n   What about this type of errora  C     u_char          sin6_len;           /* length of this struct */t ....^e< %CC-E-MISSINGTYPE, Missing type specifier or type qualifier.    + What steps do you recommend for me to take.i     thanks a lot  - "Hans Vlems" <hvlems@iae.nl> wrote in messagee0 news:al2ucd$1lubm2$1@ID-143435.news.dfncis.de...G > Since that "no tcpip stack" related message was apparently just that,nE > informational rather than en error, I wouldn't pay attention to it.iG > However the error message that said: " Extraneous parameter names ares> > ignored. at line number 295 in module RESOLV of text libraryI > SYS$COMMON:[SYSLIB]DECC$RTLDEF.TLB;2" is somewhat more of a problem. AnrF > indication that the compiler was not installed correctly, or not the7 > expected options were chosen during the installation.m >c/ > <david20@alpha2.mdx.ac.uk> schreef in berichte' > news:al22r3$5iq$1@aquila.mdx.ac.uk...aG > > In article <al0hl1$1lu4hv$1@ID-143435.news.dfncis.de>, "Hans Vlems"r > <hvlems@iae.nl> writes:tI > > >Compilation warnings are repeated by the LINKER and the LIBRARIAN tor > signalK > > >that something may have escaped the programmer's attention. Could very: > wellH > > >be that the program was written in C and that the compiler had some
 > warningsC > > >that had no effect on code quality (that seems possible in C).uB > > >So all the object files are stored in an object library (.OLB
 filetype): > amL > > >convenient way to store all those compiler output files without getting ai% > > >directory with too many entries.a > > >n > > I note the line :- > >iK > > >> %VMSBUILD-I-CCLIENT, Building C-client library with DEC C and NO TCPl > > >support > >tJ > > Could it be that there is a qualifier which you need to provide to theG > > build specifying which TCPIP stack (UCX, TCPWARE, MULTINET) you aree
 > building	 > > with.v > >eJ > > I know a number of products which you have to build with commands like > >c > > @MAKEVMS  XXX YYY  UCXTCP/ > >dJ > > where XXX, YYY and UCXTCP are various qualifiers used in the build and6 > > UCXTCP tells the build to use the UCX TCPIP stack.K > > These are usually used by ifdefs in the code to pull in various includen > files. > >R > [snip] >e >S   ------------------------------  $ Date: Wed, 4 Sep 2002 15:41:28 -0400, From: "Francisco Ortega" <fortega@iblues.cc>! Subject: Re: Help installing Pine A Message-ID: <_JQd9.13827$%P6.5930602@e3500-atl1.usenetserver.com>   5 What steps do you recommend to have this thing going.m   thanks    - "Hans Vlems" <hvlems@iae.nl> wrote in messages0 news:al2ucd$1lubm2$1@ID-143435.news.dfncis.de...G > Since that "no tcpip stack" related message was apparently just that,oE > informational rather than en error, I wouldn't pay attention to it.oG > However the error message that said: " Extraneous parameter names arey> > ignored. at line number 295 in module RESOLV of text libraryI > SYS$COMMON:[SYSLIB]DECC$RTLDEF.TLB;2" is somewhat more of a problem. An F > indication that the compiler was not installed correctly, or not the7 > expected options were chosen during the installation.t >a/ > <david20@alpha2.mdx.ac.uk> schreef in berichti' > news:al22r3$5iq$1@aquila.mdx.ac.uk...iG > > In article <al0hl1$1lu4hv$1@ID-143435.news.dfncis.de>, "Hans Vlems"  > <hvlems@iae.nl> writes:MI > > >Compilation warnings are repeated by the LINKER and the LIBRARIAN to- > signalK > > >that something may have escaped the programmer's attention. Could veryf > wellH > > >be that the program was written in C and that the compiler had some
 > warningsC > > >that had no effect on code quality (that seems possible in C).sB > > >So all the object files are stored in an object library (.OLB
 filetype): > a L > > >convenient way to store all those compiler output files without getting an% > > >directory with too many entries.y > > >  > > I note the line :- > >rK > > >> %VMSBUILD-I-CCLIENT, Building C-client library with DEC C and NO TCP  > > >support > >fJ > > Could it be that there is a qualifier which you need to provide to theG > > build specifying which TCPIP stack (UCX, TCPWARE, MULTINET) you are 
 > building	 > > with.  > > J > > I know a number of products which you have to build with commands like > >n > > @MAKEVMS  XXX YYY  UCXTCPt > >uJ > > where XXX, YYY and UCXTCP are various qualifiers used in the build and6 > > UCXTCP tells the build to use the UCX TCPIP stack.K > > These are usually used by ifdefs in the code to pull in various includec > files. > >s > [snip] >. >    ------------------------------  # Date: Thu, 05 Sep 2002 22:27:42 GMTa# From: "John Smith" <a@nonymous.com>I4 Subject: Re: IBM did it, Sun did it, why Compaq not?J Message-ID: <yTQd9.177186$8aG1.41891@news01.bloor.is.net.cable.rogers.com>  3 "Rudolf Wingert" <win@fom.fgan.de> wrote in messagea0 news:200209050622.IAA22311@sinet1.fom.fgan.de... >sF > No. I did think about the Veritas file system. The newest example isC > Corel. Why did Compaq not pay for WordPerfect as a standard underrF > OpenVMS? Why did they not pay for porting other Corel office productE > and Corel Draw to OpenVMS? I am happy that there a few ISVs (inclu-oG > ding IBM), which do know the power of OpenVMS and do have application- > for OpenVMS.  L Wordperfect put their eggs into the Windows basket. That meant coding to theL Win API. In order to reap any performance out of a windowed app, it is in myL experience necessary to use the native API. Now one could use something likeI Bristol Technologies (if they are still around and still writing code foriF VMS) to provide a 'translation' layer, but that only adds to the cost,L complexity, and lack of perceived user-friendliness due to the slower window response times.w  J Don't forget that at the time, Palmer et. al. were saying that Windows and Unix were the future.e   ------------------------------  # Date: Thu, 05 Sep 2002 22:28:51 GMTi0 From: Paul Winalski <prune@ZAnkh-Morpork.mv.com>  Subject: Re: Itanic2 performance8 Message-ID: <lcmfnu0upamtoidn3084nid49v4mvgg8qk@4ax.com>  , On Wed, 04 Sep 2002 13:52:44 -0400, JF Mezei% <jfmezei.spamnot@videotron.ca> wrote:s  N >Is it really fair then to claim that IA64 is a 64 bit architecture if all its2 >performance metrics are measured in 32 bit mode ?  > Perhaps not, but it's within the rules of the Spec benchmarks.  K >Considering that the 8086 will continue to power 32 bit servers at a loweryO >price and respectable performance, wouldn't it be fair to think that those who K >can live within the limits of 32 bits would stay on the traditional winteld
 >servers ?  > That certainly is one of the risks to Itanium's success in the! marketplace, as it was for Alpha.   M >And if performance of the IA64 in 64 bit mode sucks, it won't really do wellhI >in the market it was designed for: those users who need the 64 bit mode.s  C Itanium integer Spec performance is lower with LP64, but it's stilldB quite respectable.  And there are important applications out there? where the benefits of the 64-bit address space far outweigh theeA increase in data structure size caused by 64-bit pointers.  Largea1 database and transaction processing, for example.   
 ---------- Remove 'Z' to reply by email.a   ------------------------------   Date: 6 Sep 2002 02:01:41 GMTt& From: Rick Jones <foo@bar.baz.invalid>  Subject: Re: Itanic2 performance* Message-ID: <al92a5$p79$2@web1.cup.hp.com>  1 Paul Winalski <prune@zankh-morpork.mv.com> wrote:nD > The "HP-UX compiler technology" you are talking about is the IntelE > compiler.  The Itanium 2 benchmarks in question were produced using-F > Intel's compilers running on HP-UX.  The plan last time I looked was2 > for VMS to be "stuck with" this same technology.  F Intel's compilers run under Linux (and I suppose Windows). They do notE run under HP-UX. Similarly, the HP-UX compilers only run under HP-UX.TE So, a SPECcpu2000 result on a system running HP-UX used HP compilers.eF A SPECcpu2000 result on a system running Linux would not have used the
 HP compilers.s  C This can be confirmed for those results published by looking at thee# disclosures at http://www.spec.org/d  
 rick jones -- tH Wisdom Teeth are impacted, people are affected by the effects of events.F these opinions are mine, all mine; HP might not want them anyway... :)A feel free to post, OR email to raj in cup.hp.com  but NOT BOTH....   ------------------------------   Date: 6 Sep 2002 01:55:51 GMTn& From: Rick Jones <foo@bar.baz.invalid>  Subject: Re: Itanic2 performance* Message-ID: <al91v7$p79$1@web1.cup.hp.com>  ) Bill Todd <billtodd@metrocast.net> wrote: B > And then there SPECweb99_SSL, where McKinley turned in even more= > stellar numbers.  While you'd expect a benchmark containingaC > encryption to let Itanic2 shine (at least if encryption played an,B > important part in the overall results), there are some questions > that pop up there as well.  > > While I don't know how important disk performance is to thisF > benchmark, the results (with a single exception) suggest that it may? > dominate it.  The two top performers in the 4-processor class5E > (Itanic2 and PA-RISC) use the most (15) and fastest (15Krpm) disks,(B > while the relatively poorly-performing Alpha and Sun entries useF > fewer (9) and slower (10Krpm for Sun, and AFAICT 7200 rpm for Alpha)F > disks (the test results give only a part number for the Alpha disks,D > and the only Google reference I could find that seemed to shed anyF > light on their speed was a Tru64 cluster upgrade comparison documentE > where the same part was used and the only 9 GB disks mentioned were@ > 7200 rpm ones).j  B Do not attribute to need what may be more accurately attributed to( accident of what was around at the time.  E There are two sources of possible disc I/O for a SPECweb99* benchmarkc? - access to the working set, and logging. For a given number ofwF conforming connections, SPECweb99 and SPECweb99_SSL have the same size& working-set and the same logging rate.  I Consider that four-CPU rx5670 SPECweb99_SSL figure of 1520 SPECweb99_SSL:9  O http://www.spec.org/osg/web99ssl/results/res2002q3/web99ssl-20020708-00009.htmli  A In the "FDR" you will see the fileset size as being ~4983 MB. ThemC system RAM was also shown to be 24GB, which was more than enough toSF have the entire working-set in RAM, even if each of the Zeus processesE had its own private copy of each URL. There was virtually no disc I/OcD for the working set. (No, I did not run that benchmark, but have runC many SPECweb99 benchmarks, and am in close contact with the currentd) crop of folks running the UX benchmarks).r  D Also, note in the "SUT Notes" that one 10K RPM disc was used for theC working set, it was the logs that were placed on the HP va7100 diskc) array (the 15 15K PRM discs you mention).   ? So, lets consider the logs; perhaps that was generating lots ofnA I/O. For that, let's go back to a SPECweb99 result with a similarw> number of conforming connections. For that, let's go back to a; SPECweb99 result for a two-CPU, 550 MHz HP 9000 Model A500:c  I http://www.spec.org/osg/web99/results/res2000q4/web99-20001030-00071.html   F It achieved 2109 SPECweb99, and had a working set of ~6872 MB. It alsoB used the binary log format, and would have been generating logs atC roughly 2109/1520 or 1.39X that of the rx5670 SPECweb99_SSL result.-  9 The A500 configuration used two, 10K rpm disc mechanisms.0  > From this one can safely assume that a bevy of 15 15K RPM discD mechanisms are not necessary, though certainly sufficient :) for the" I/O requirements of the benchmark.  E SPEC reporting rules require that what was on the system be reported,(? whether it was "necessary" to achieve the performance result orlF not. In this instance, the SPECweb99_SSL systems happened to have discA arrays connected to them, and it was convenient to use them (whenoE running lots of SPECweb99* runs, the logs can start to add-up, havingpD lots of space for the logs is convenient) even though it does result( in people drawing incorrect conclusions.  D To get a better/more accurate idea of whether or not a configurationB might have needed a given number of discs, you need to look at theC reported fileset size, and compare that with the RAM in the system,iE how the URLs were being cached - were they being cached in malloc()ed F memory, mmap private, mmap shared, or in the filesystem buffer cache -8 and the number of discrete web server processes running.  
 rick jones --  H Wisdom Teeth are impacted, people are affected by the effects of events.F these opinions are mine, all mine; HP might not want them anyway... :)A feel free to post, OR email to raj in cup.hp.com  but NOT BOTH...    ------------------------------  # Date: Fri, 06 Sep 2002 03:21:50 GMT * From: "Bill Todd" <billtodd@metrocast.net>  Subject: Re: Itanic2 performanceC Message-ID: <hbVd9.449729$m91.17499438@bin5.nnrp.aus1.giganews.com>-  3 "Rick Jones" <foo@bar.baz.invalid> wrote in messageM$ news:al91v7$p79$1@web1.cup.hp.com...   ...f  D > Do not attribute to need what may be more accurately attributed to* > accident of what was around at the time.  C I didn't necessarily attribute it to need (which is why I posed thegK observation as a question rather than an assertion) - I just noted that the J performance results seemed to correlate quite closely with the performance of the disk subsystems used.  J But your response was exactly what I hoped to get (well, I suppose I wouldH have been happier if it had been more favorable to Alpha, but what I wasH primarily looking for was someone with some actual knowledge of what wasJ going on in the benchmark).  And despite a bit of discussion in comp.arch,? considerably more here, and a veritable blizzard of activity atoK realworldtech.com, yours is the first contribution to shed any actual lightM on the matter.   >,G > There are two sources of possible disc I/O for a SPECweb99* benchmarkMA > - access to the working set, and logging. For a given number of/H > conforming connections, SPECweb99 and SPECweb99_SSL have the same size( > working-set and the same logging rate. > K > Consider that four-CPU rx5670 SPECweb99_SSL figure of 1520 SPECweb99_SSL:! >  >fL http://www.spec.org/osg/web99ssl/results/res2002q3/web99ssl-20020708-00009.h tmla > C > In the "FDR" you will see the fileset size as being ~4983 MB. The-E > system RAM was also shown to be 24GB, which was more than enough toj% > have the entire working-set in RAM,u  G That part I noticed, though without knowing exactly what 'fileset size'c meant I couldn't be certain.  #  even if each of the Zeus processespG > had its own private copy of each URL. There was virtually no disc I/O"F > for the working set. (No, I did not run that benchmark, but have runE > many SPECweb99 benchmarks, and am in close contact with the currentt+ > crop of folks running the UX benchmarks).d >aF > Also, note in the "SUT Notes" that one 10K RPM disc was used for theE > working set, it was the logs that were placed on the HP va7100 diskr+ > array (the 15 15K PRM discs you mention).s >iA > So, lets consider the logs; perhaps that was generating lots of1C > I/O. For that, let's go back to a SPECweb99 result with a similar @ > number of conforming connections. For that, let's go back to a= > SPECweb99 result for a two-CPU, 550 MHz HP 9000 Model A500:" >nK > http://www.spec.org/osg/web99/results/res2000q4/web99-20001030-00071.html  >mH > It achieved 2109 SPECweb99, and had a working set of ~6872 MB. It alsoD > used the binary log format, and would have been generating logs atE > roughly 2109/1520 or 1.39X that of the rx5670 SPECweb99_SSL result.B >s; > The A500 configuration used two, 10K rpm disc mechanisms.a  I Indeed, and I'll take your (implied) word for it that the logging load offE SPECweb99 is at least essentially identical to that of SPECweb99_SSL.t  L However, if the two benchmarks are pretty much identical save for the use ofK SSL, the second set of results above seems to make it absolutely clear thaniC the benchmark is dramatically dominated by the encryption activity:u@ otherwise, how could a pair of 550 MHz PA8600s running SPECweb99L significantly out-perform 4 top-end McKinleys running SPECweb99_SSL?  If so,G that also explains why the latter could blitz four 833 MHz Alphas in an6C ES40:  the one thing McKinley (and even Merced) really shines at is   FP/encryption-style performance.  J So thanks - I'll enter this insight into the discussion over at rwt.com asE well, and it should at least put that portion of the subject to rest.a   - bill   ------------------------------  * Date: Thu, 5 Sep 2002 23:46:04 +0000 (UTC) From: ouuktk@dwdddaae.comd, Subject: ja foste enganado no teu trabalho ?* Message-ID: <1031269120.387266@proxynews1>  < ############     http://burlasleo1.cjb.net     #############  9 - ACT TO STOP CRIMINALS AND ABUSE OF POWER IN ENTERPRISESa= - HAJA PARA PARAR COM CRIMINOSOS E ABUSO DE PODER EM EMPRESASt  < ############     http://burlasleo1.cjb.net     ############# jiptvd   ------------------------------   Date: 5 Sep 2002 16:20:14 -0700t, From: colive@technologEase.com (Chris Olive) Subject: Re: MATCHC Instruction = Message-ID: <b10654c6.0209051520.7efc4b6f@posting.google.com>i  Y briggs@encompasserve.org wrote in message news:<ffk8NQb3w5bT@eisner.encompasserve.org>...e2 > colive@technologEase.com (Chris Olive) wrote in 5 >  <b10654c6.0209041450.6e6f10f6@posting.google.com>:r > >"Stanley Reynolds" <nospam_stanley_reynolds@yahoo.com> wrote in message news:<PNpd9.437$2L.74228@e3500-atl2.usenetserver.com>...n? >  [Nested attribution lost.  I think it was Chris Olive again]oG > >> > It wouldn't be terribly hard to view your object "string" in 65keG > >> > blocks and iterate those blocks of your "string" over the MATCHCnJ > >> > instruction <x> number of times for a string of <x> blocks, esp. ifG > >> > your match string fits within one block.  You could do it with aiC > >> > custom .MACRO more than likely, and just replace your MATCHCdI > >> > instruction instances with your custom .MACRO (call it BLKMTCHC orcH > >> > something like that.)  R0 holds the address at the end of the 65kI > >> > block (or any block size you choose up to 65k). If you don't get amE > >> > match on MATCHC, increment your starting object pointer (usingMM > >> > register displacement for example) by R0 on no match, and iterate overfD > >> > MATCHC again until you get a match or run out of blocks (or aJ > >> > sub-block as in the case of the last block short of an even block.)I > >> > Could conceivably be done with a custom .MACRO I think... (If your-L > >> > match string can be larger than 65k, then this would perhaps be a bit > >> > of a different story.)l > >> > >  [...]L > >> Would this work in the case of a string that crossed a block boundary ? >  dH > >Yes.  Say you had a string that was 262,420 bytes long.  262420/65535G > >= 4 blocks plus 280 bytes left over or 280 bytes which would cross asH > >logical 65k block boundary.  Your first 4 iterations over MATCHC thenF > >would specify 65535 for the source length (3rd operand).  Your lastH > >iteration would specify the 280 bytes left over as the source length. > J > No.  It would not work.  You need to increment by blocksize-searchsize+1 > I > For instance, with a 65535 byte block size, a 262420 byte search stringn* > and a 100 byte search object, you would: > ' > MATCHC ( 100, objadr, 65535, srcadr )n- > MATCHC ( 100, objadr, 65535, srcadr+65436 )a. > MATCHC ( 100, objadr, 65535, srcadr+130872 ). > MATCHC ( 100, objadr, 65535, srcadr+196308 ), > MATCHC ( 100, objadr, 676, srcadr+261744 ) > B > Assuming that I haven't made any arithmetic or fencepost errors. >   F A very good observation.  Thanks.  This is right.  And the incrementerD would not be conveniently in R0 then, as you also assert.  Increment logical blocks as indicated.  E The concept however still remains and I still maintain it can be doned  with a replacement .MACRO... 8-)   Regards, Chrish -----t Chris Oliveo colive(at)technologEase(dot)comi   ------------------------------  # Date: Thu, 05 Sep 2002 18:18:29 GMTl* From: "WarlockD" <warlockd@drakesmith.net>A Subject: Re: Microsoft reeling from hack attacks, VMS just yawns!M? Message-ID: <VdNd9.165748$_91.212970@rwcrnsc51.ops.asp.att.net>   F Funny, I wonder if it has to do with office, if it has to do with mail programs and such?    5 "Bob Ceculski" <bob@instantwhip.com> wrote in messageU7 news:d7791aa1.0209050442.391fa60c@posting.google.com...w# Microsoft reeling from hack attackst  . Patch what you can, worry about what you can't  0 By Paul Hales: Thursday 05 September 2002, 13:08  > MICROSOFT REPORTS that hackers are having a field day with itsC software. It seems the harder the company tries to shore up all itseB products, the more the hackers, crackers and whacky-bacciers enjoy' worming their way through the defences.-F Last week the company posted an advisory here warning of an "increasedA level of hacking activity" that it had been tracking. The hackinggD attempts show similar symptoms and "behaviors", the company said andB affect a whole slew of Microsoft operating systems from Windows 98 onwards.  D Microsoft admits to being clueless as to how to stop the attacks andC warns of a "spike" of activity involving back-door trojans slippingo0 errant files onto PCs connected to the Internet.  F The company has also released details of a Certificate Validation Flaw- that could enable identity spoofing, it says.u  E The alert carries the maximum severity rating, Critical and affects ao? similarly long list of Microsoft operating system products. The B company has released patches for Windows 2000 and Windows XP whichE administrators should install immediately, it says. You'll find those  here.i  > The company is still working on patches for its other productsC including Windows 98, the INQ preferred Win 98 SE, Windows Me and a ? bunch of Mac products including versions of Office for the Mac.   2 The voles are beavering away on those as we write.   ------------------------------  * Date: Thu, 5 Sep 2002 22:44:15 +0000 (UTC)+ From: david20@alpha2.mdx.ac.uk (David Webb)uA Subject: Re: Microsoft reeling from hack attacks, VMS just yawns!:+ Message-ID: <al8mnv$c3d$1@aquila.mdx.ac.uk>M  l In article <VdNd9.165748$_91.212970@rwcrnsc51.ops.asp.att.net>, "WarlockD" <warlockd@drakesmith.net> writes:G >Funny, I wonder if it has to do with office, if it has to do with mail> >programs and such?a >   / And it's probably about to get much much worse.r  O It appears that shatter attacks (see http://security.tombom.co.uk/shatter.html)t# are even worse than first appeared.n2 see http://www.net-security.org/article.php?id=166  O For those who aren't familar with this. Shatter allows a user to escalate theiriO privileges if any privileged program has opened a window on the user's desktop.s. The problem is with the Windows messaging API.  N Microsoft's response was that no privileged program should be doing this sinceJ the desktop was a security boundary. However no one seems to have told theI application writers (and from the article above Microsoft break their owneN rules). It also now looks as though you may not need the privileged program to* have even opened a window on your desktop.      
 David Webb VMS and Unix team leader CCSS Middlesex University         >e6 >"Bob Ceculski" <bob@instantwhip.com> wrote in message8 >news:d7791aa1.0209050442.391fa60c@posting.google.com...$ >Microsoft reeling from hack attacks >e/ >Patch what you can, worry about what you can't- >-1 >By Paul Hales: Thursday 05 September 2002, 13:08i >J? >MICROSOFT REPORTS that hackers are having a field day with itssD >software. It seems the harder the company tries to shore up all itsC >products, the more the hackers, crackers and whacky-bacciers enjoyk( >worming their way through the defences.G >Last week the company posted an advisory here warning of an "increased B >level of hacking activity" that it had been tracking. The hackingE >attempts show similar symptoms and "behaviors", the company said and C >affect a whole slew of Microsoft operating systems from Windows 98 	 >onwards.l > E >Microsoft admits to being clueless as to how to stop the attacks and-D >warns of a "spike" of activity involving back-door trojans slipping1 >errant files onto PCs connected to the Internet.n > G >The company has also released details of a Certificate Validation Flawh. >that could enable identity spoofing, it says. >eF >The alert carries the maximum severity rating, Critical and affects a@ >similarly long list of Microsoft operating system products. TheC >company has released patches for Windows 2000 and Windows XP which-F >administrators should install immediately, it says. You'll find those >here. >o? >The company is still working on patches for its other products D >including Windows 98, the INQ preferred Win 98 SE, Windows Me and a@ >bunch of Mac products including versions of Office for the Mac. > 3 >The voles are beavering away on those as we write.a >c >a   ------------------------------  # Date: Fri, 06 Sep 2002 01:47:39 GMTa1 From: "David J. Dachtera" <djesys.nospam@fsi.net> B Subject: Re: netscape/modzilla/HP/open-source: system requirements' Message-ID: <3D780E93.94736B3A@fsi.net>    Colin Blake wrote: >  > JF Mezei wrote:  > # > >Won't CSWB be renamed to HPSWB ?e > > L > >If it applies only to Digital products (aka: VMS), why not brtand it DSWB > >o > >Or better yet, VMS-Mozilla ?s > > 3 > Maybe its time for a "name that product" game :-)e  E ...and I know the perfect prize: free commercial licenses for VMS andoF layered products for the lifetime of the winner and his/her first heirA (that is, it can be willed to one generation after the winner, not. further). *NOT* limited to any hardware class.   -- u David J. Dachterao dba DJE Systemsh http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/t   ------------------------------  % Date: Fri, 06 Sep 2002 05:16:51 +0200n2 From: martin@radiogaga.harz.de (Martin Vorlaender) Subject: Re: OpenSSL woese; Message-ID: <3d781e23.524144494f47414741@radiogaga.harz.de>e  3 Martin Vorlaender (martin@radiogaga.harz.de) wrote:C5 > Martin Vorlaender (martin@radiogaga.harz.de) wrote:lC > > I'm trying to integrate SSL with ht://Dig to enable it to index  > > https URLs.L ... 7 > The final point in the code where the error occurs iss1 > MD_RAND\ssleay_rand_bytes\%LINE 7540 (listing):  > C >         RANDerr(RAND_F_SSLEAY_RAND_BYTES,RAND_R_PRNG_NOT_SEEDED); D >         ERR_add_error_data(1, "You need to read the OpenSSL FAQ, "= >                 "http://www.openssl.org/support/faq.html");y  D The FAQ suggested that the application should call RAND_load_file().) I put that in, and it started to work... n  E When I've tested it some more, and sorted out the legal implications,r I'll announce the new release.  7 A big thank you goes to Rick Barry for ongoing support.d   cu,    Martin -- oG So long, and thanks        | Martin Vorlaender  |  VMS & WNT programmer 4 for all the books...       | work: mv@pdv-systeme.deG In Memoriam Douglas Adams  |   http://www.pdv-systeme.de/users/martinv/g;             1952-2001      | home: martin@radiogaga.harz.der   ------------------------------  # Date: Thu, 05 Sep 2002 23:13:05 GMTo, From: "Kenneth Farmer" <kfarmer@openvms.org>% Subject: Post answers in forum people/< Message-ID: <5yRd9.3913$ip3.436237@twister.southeast.rr.com>  J If you had legitamate suggestion please post in forum.  That's where he'llI be checking for responses.  Not sure if he visits the newsgroups.  So far I not one suggestion has been listed in the forum.  I guess that's good forh% OpenVMS, bad for potential ISV's.  :)s  2 http://www.openvms.org/phorum/read.php?f=6&i=4&t=4   kf   --   Kenneth Farmer http://www.Tru64.org http://www.OpenVMS.org http://www.LinuxHPC.orgi      7 "Kenneth Farmer" <kfarmer@openvms.org> wrote in messaget6 news:n3Od9.1537$jF4.362242@twister.southeast.rr.com...7 > From the General forum on OpenVMS.org by Jim Johnson:p4 > http://www.openvms.org/phorum/read.php?f=6&i=4&t=4 >c > ---------------------- > E > I'm interested in finding out what sort of features or capabilities  OpenVMSpJ > customers might find interesting. I have a personal reason for this -- IJ > have a small company that might be interested in building infrastructure orH > integration products for OpenVMS, and I'd like to get a handle on what sortst* > of products there would be a demand for. >hL > However, I'm also interested generally. I don't hear a lot of conversationG > in comp.os.vms or in here that I can see could translate into producteF > opportunities. Is the market truly dead, or have we all just stopped asking?k >nK > So, for right now, I'd like us to brainstorm a bit about what we'd reallyeG > like to see to make OpenVMS better in our environments -- it could be 8 > additional VMS-specific features, portability aids, or= > interoperation/integration capabilities. Or something else.  > L > So, to get a discussion going, below is a list of things that have occured: > to me in the past that I posted in comp.os.vms in march. >.H > I'd say "what do you want?", only that's got some bad connotations for those D > that followed a certain TV series... Instead, let's get some ideas flowing. >  > Jim. >u > TP Application Building Toolso >hF > - A clusterwide, transaction-capable, persistant named property bag. >iH > This would provide a simple to use container for name/attribute pairs,K > similar to clusterwide logical names, with the addition that the contentssG > would be automatically hardened to disk, and the access could be dones underx > transaction control. >wI > This would allow clusterwide applications to cheaply share data betweenAC > instances of themselves, and use transactions to maintain logicalu > consistency. >t > - Transaction capable IPCr > L > $ICC offers a high performance local and cluster communications mechanism.K > This would layer on top of that and provide a fully automatic transactiontL > branch management feature. Thus, a client-server application would be ableI > to just issue a 'send' request while inside a transaction and know thatk theeK > actions at the remote end would automatically be processed as part of then > transaction. >e > View it as ACMS-extra-lite.n >h > - TIP interoperability >oJ > Provide an API that gateways between DECdtm and TIP. Include the support toG > actually connect and interopate with a Microsoft MTS based front end.0 >h >m > Java Interoperability: >l# > - ACMS to J2EE migration toolkit.  > F > This would process TDL, CDD, and the other associated files to buildK > equivalent JSP, EJB, and servlet structures. As well as supporting accessi toG > Java-only logic, this would either use an available infrastructure toeE > connect back to an existing ACMS task, or incorporate one to do so.U > L > This would let an ACMS customer either interoperate with J2EE, or migrate,4 > over time, to a fully J2EE compliatnt environment. >a > - Other to J2EE migrationh >uH > This is just a placeholder to point out that the other TP monitor-typeK > things that Digital shipped could also be migrated in a similar manner --  > DECintact, MessageQ, etc.t >H > -------------------------- >5J > Let him know what features you would like to see available.  Respond at:4 > http://www.openvms.org/phorum/read.php?f=6&i=4&t=4 >  >P >e > Kenu >w > -- >i > Kenneth Farmer > http://www.Tru64.org > http://www.OpenVMS.org > http://www.LinuxHPC.orga >o >U >5 >H   ------------------------------   Date: 5 Sep 2002 14:17:16 -0600e+ From: young_r@encompasserve.org (Rob Young)a  Subject: Request for Tools guide3 Message-ID: <453$YIB1lfpO@eisner.encompasserve.org>o  9 	Here is a request.  It is sort of a request for overview1/ 	and installation guides for all OpenVMS tools.   = 	I would love to see a document describing all the managementhG 	tools for OpenVMS.  Even if it is a midnight project.  Essentially it g 	would:o  # 		1)  Describe what the tool is ford 		2)  Cost a 		3)  Suggested uses9 		4)  If obsoleted, explain replacement and missing/addeds 			features of new tool(s).s3 		5)  How to set it up with all major TCP/IP stacksy  > 	I just got email about a new rev level of HP Insight Manager.  F 	I spent a while getting that going (SNMP is a pain folks, my opinion), 	with a few DSNlink exchanges to go with it.  A 	My brain is getting soft sorting out tools.  OpenVMS management lF 	station (not using for legit reasons don't believe it works on bootupI 	saving/starting print queue configs with Multinet NTY symbiont), WEBES, .< 	Availability Manager.  None of this is real hard but it is D 	increasingly difficult figuring out why or if I need the tool, and  	getting them going.  E 	I think a one-stop doc that is a very nice overview and walk through4D 	with typical setups for all tools would be a good thing.  Realizing? 	some installations are more involved and maybe you could refer#> 	to installation guides that reside elsewhere (how to find the 	installation guides).  D 	At the very least, an overview guide for all tools if nothing else.   				Rob   EB 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 slipstreamM> The clock reads daylight savin'         Rollin' in the thunder  -                                 -- Neil Youngo   ------------------------------   Date: 5 Sep 2002 17:06:48 -0700i. From: SPAMSINK2001@YAHOO.COM (Alan E. Feldman)Y Subject: Response to NIST's take on noon and midnight (was Re: Is the HP/Compaq merger re_= Message-ID: <343f30ae.0209051606.5384c97c@posting.google.com>3   "Webb, William W Raleigh, NC" <wwebb1@email.usps.gov> wrote in message news:<BA52530E3149734A9BAABDBBFA808E49028BEA60@rlghncst964.usps.gov>... > Sorry, I rounded.5 > What I meant to say was: >  > AM 00:00:01 => 11:59:59  > PM 12:00:01 => 23:59:59s    E Do you mean to say that times like 00:00:00.50 are neither am nor pm?u (!)c    . > Here's what the US govt says on the issue of > noon and midnight: > 7 > http://www.boulder.nist.gov/timefreq/general/misc.htmi [...]i+ > Are noon and midnight 12 a.m. or 12 p.m.?  > N > This is a tricky question. The answer is that the terms 12 a.m. and 12 p.m. # > are wrong and should not be used.o > K > To illustrate this, consider that "a.m" and "p.m." are abbreviations for sI > "ante meridiem" and "post meridiem." They mean "before noon" and "aftery	 > noon," aI > respectively. Noon is neither before or after noon; it is simply noon. tC > Therefore, neither the "a.m." nor "p.m." designation is correct. hN > On the other hand, midnight is both 12 hours before noon and 12 hours after M > noon. Therefore, either 12 a.m. or 12 p.m. could work as a designation for  @ > midnight, but both would be ambiguous as to the date intended. [...]W  E They can say what they want, but in this case they are wrong. The sun A almost never crosses the meridian at 1200 standard time, so their4F argument falls apart. I just wrote to them again about this. You quote them and I question them.b  B By the old definitions, they are right, if you apply them to local; solar time. But they don't apply to standard time. The new, E unfortunately not yet "officially" recognized definitions, work great E for am/pm digital clocks showing standard time. The new definition ishA logical, mathematically self-consistent, sensible, useful, and deb facto. What more do you want?   A Remember, it used to be "official" that all heavenly bodies movedc around the Earth.1  A So my question is, why do we have to be forever shackled to thesecF obsolete, Old World definitions when the new ones are clearly superiorA for modern life? Actually, we don't, fortunately, because the newa> definitions are what all am/pm digital clocks use, ebmedded or
 otherwise.  B Why can't the NIST site at least acknowledge what is already fact?  They can mention both versions!:  F Traditionally, ante meridiem blah blah blah.... But with the advent of/ modern life, digital clocks, blah blah blah....a   Would that be so terrible?   Dislcaimer: JMO  Alan E. Feldmani spamsink2001 at yahoo dot com    ------------------------------  # Date: Thu, 05 Sep 2002 22:57:11 GMTr# From: "John Smith" <a@nonymous.com>eB Subject: Re: Secret New Products (was: Is the HP/Compaq merger...)I Message-ID: <bjRd9.105574$GK2.47253@news02.bloor.is.net.cable.rogers.com>n  5 "WarlockD" <warlockd@drakesmith.net> wrote in messagew) news:ryMc9.271661$983.530159@rwcrnsc53...lK > Heh, next thing you know, they will introduce the "Alpha Vax", 'Now sucks00 > more volume than any of its Vax predecessors!"   It sucks faster.   ------------------------------  $ Date: Thu, 5 Sep 2002 19:00:08 -0400, From: "Francisco Ortega" <fortega@iblues.cc>0 Subject: Sending Emails with OpenVms 7.2 + QUEUEA Message-ID: <rfRd9.13862$%P6.5942362@e3500-atl1.usenetserver.com>c    Hi,D         I originally wrote about pine, which I have not been able toL configure yet. Since I figure that it may take me longer for the compilationJ errors, I decide to try the regular mail. I went ahead and put an internetL address on the TO:. Well , the system complain about not having SMTP. I wentI and configure SMTP using the sys$manager:tcpip$config  for the client andaK also (I Don't think I need it ) I configure the POP3 for the server (server # options). ( JUST ENABLE THOSE TWO).i  H Well, then when I try to send the email, it complaint about not having a queue on the systema  5 so I created the que using start/queue/manager/queue. I That work but I don't remember how to create the sys$batch and sys$print.i5 I went ahead and try to send another email using mail L Now it's telling me that there is no such queue, but I don't know what queueH is need it . maybe sys$smtp or tcp$smtp , I'm just guessing, but I don'tE remember how to create it.  Does someone has an email client softwareiJ running or even pine arealdy compile that can be send to me. Really, All IJ want is to have an email client to receive emails in my alpha. Have a news  client, it would be even better.  = After I get that running, maybe I can have an email server :)-    . Thanks a lot for all your help before and now.   Francisco R. Ortega2   ------------------------------  $ Date: Fri, 6 Sep 2002 12:36:40 +10009 From: forMsytAhm@optusShom.com.aKu (Mark(un-MASK)Forsyth)a4 Subject: Re: Sending Emails with OpenVms 7.2 + QUEUE7 Message-ID: <slrnang55n.6l.forMsytAhm@plague.bogus.com>.  U On Thu, 5 Sep 2002 19:00:08 -0400, Francisco Ortega <fortega@iblues.cc> gushed forth:p   [lossy compression]   b I just downloaded pine for VMS from http://www.agh.cc.kcl.ac.uk/files/vms/pine-vms/ , unzipped it,O configured it and used it. Took about 20 minutes using the prebuil executables./    
 Running :-   DEC 3000 Model 300@   DIGITAL TCP/IP Services for OpenVMS Alpha Version V4.2 - ECO 12   on a DEC 3000 Model 300 running OpenVMS V7.1-2    C No problems. (Yeah I know, it's old but it works well for me..:-) )      -- w     Ooroo 	 Mark F...r  $ Another Optus Cable Traffic Monitor.3 http://www.members.optushome.com.au/forsythm/traff/    ------------------------------  % Date: Fri, 06 Sep 2002 06:04:58 +0200r2 From: martin@radiogaga.harz.de (Martin Vorlaender)4 Subject: Re: Sending Emails with OpenVms 7.2 + QUEUE; Message-ID: <3d78296a.524144494f47414741@radiogaga.harz.de>y  + Francisco Ortega (fortega@iblues.cc) wrote:a  G > I went and configure SMTP using the sys$manager:tcpip$config  for theaI > client and also (I Don't think I need it ) I configure the POP3 for theo4 > server (server options). ( JUST ENABLE THOSE TWO).  B I think I remember that enabling SMTP doesn't start it right away.  rJ > Well, then when I try to send the email, it complaint about not having a > queue on the systems  ? SMTP is probably not started - it creates the queue on startup.i8 You may want to try "$ @SYS$MANAGER:TCPIP$SMTP_STARTUP".  i7 > so I created the que using start/queue/manager/queue.s  F That command (if it really works) starts the queue manager. It doesn'tE have to do anything with creating new queues. Besides that, you don'tt create SMTP queues by hand.b  0 Read the Compaq TCP/IP Services documentation at< http://www.openvms.compaq.com/doc/ on how to configure SMTP.   cu,M   Martin --  A                      | Martin Vorlaender  |  VMS & WNT programmers. Microsoft's answer   | work: mv@pdv-systeme.deA to OpenVMS is        |   http://www.pdv-systeme.de/users/martinv/e5 Windows NT 10.0.     | home: martin@radiogaga.harz.dei   ------------------------------   Date: 5 Sep 2002 16:05:31 -0700e, From: colive@technologEase.com (Chris Olive)5 Subject: Re: SET PROTECTION (SET SECURITY/PROTECTION) = Message-ID: <b10654c6.0209051505.7138380e@posting.google.com>o  | Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com> wrote in message news:<01KM5BGDDTN09QUS7H@sysdev.deutsche-boerse.com>...H > > > While I'm on the topic, why isn't there F$PROTECTION to return the# > > > protection mask on an object?  > > E > > Does F$FILE_ATTRIBUTES (filename,"PRO") not the things you want ?  > ) > Yes, of course, forgot about that.  :-(e > J > Pursuing the analogy with F$PRIVILEGE, is there a need for a lexical to * > change the (protection) file attributes?  E About 10 years ago, I wrote a command procedure called CHMOD.COM thatsD worked roughly the way Unix 'chmod' does/did.  It only worked in theF 'chmod nnn' format where 'nnn' is an octal number like '755'.  (Not inD 'chmod o+rx' style.)  My "VMS-ation" of chmod was to view the "RWED"F as a bit mask where R=8, W=4, E=2, and D=1.  The command procedure wasF simple: define all possible (16) combinations (masks) as strings, thenF take a hex bit-mask string as the first parameter and equate it to the2 string/mask, and viola! I had a protection string.   Rough example:  . $! Define bit masks (16 possible combinations) $ mask_0 :=T $ mask_1 := :D $ mask_2 := :E $ mask_3 := :ED. $ mask_4 := :W $ :y $ :h $ mask_15 := :RWED $ :  $ :f $ bitmask = p1 $ :n $ :o $ world = "", $ world_mask = %X'f$extract( 3, 1, bitmask )  $ world = "W" + mask_'world_mask $ :0 $ ::  F and the same for all the other protections, SYSTEM, OWNER, and GROUP. F You get the idea.  So 'chmod FF22 filename.txt' would build the stringD "(S:RWED, O:RWED, G:E, W:E)" on the fly, and I passed it to SET PROTC and I was done.  (The "_" character was an "overlook" character, so A 'chmod __22 filename.txt' would just build "(G:E, W:E)" and leaveL% SYSTEM and GROUP alone for instance.)h  E You could combine that logic with say a '+' on the front of your mask)> and -- you'd have to check each files EXISTING protection with> F$FILE_ATTR( filename, "PRO" ) first -- apply that mask to theC existing protection by ORing it.  If you break your protection mask/C passed as P1 down into real numbers, ORing becomes a piece of cake,NB which is why I even mentioned all this.  If you have to manipulateF strings, well... yuck.  But if you can convert your RWED's to numbers,E all the better: Get the mask you want to add, convert it to a number,i? OR with your existing file protection mask (which you of course|E converted to a number using the same logic), and viola: you have your A new mask (which you can convert back to a string rather easily asw
 shown above.)e  A You are welcome to the code for CHMOD.COM for a jump start if you.E want.  I thought the automatic hex conversion in DCL (ie. num = %Xnn):C worked as a great facilitator, and allowed me to view RWED as a bitsC mask for a fairly straight-forward solution.  So maybe you can rollr; your own solution in much the same way?  It would be worthy  freeware....   Chrisw -----' Chris Oliven colive(at)technologEase(dot)comv   ------------------------------   Date: 5 Sep 2002 16:11:52 -0700 . From: SPAMSINK2001@YAHOO.COM (Alan E. Feldman)5 Subject: Re: SET PROTECTION (SET SECURITY/PROTECTION)p= Message-ID: <343f30ae.0209051511.5e2e5922@posting.google.com>/  | Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com> wrote in message news:<01KM59B9G7XU9QUS7H@sysdev.deutsche-boerse.com>...J > Suppose I have a large amount of files [...]*.* to which I want to make G > sure that for, say, World, Read privilege is granted.  However, some eJ > will already have, say, W:RE or W:E while others will have just W:.  Is K > there a simple way to ADD the R priv?  Just specifying W:R will override  E > what is already there, which is not what I want.  (I think this is lE > something like chmod o+r in unix.)  Is this REALLY an area where a e= > feature one really needs is present in unix and not in VMS?h > E > While I'm on the topic, why isn't there F$PROTECTION to return the a > protection mask on an object?     T Well, R is equivalent to RE (if you can read it, you can run it), so why not just do    # $ SET FILE [...]*.* /PROT=W=R   ! ?t    E (or SET SECURITY/PROTECTION or whatever it is in your version of VMS),  A That will save you from writing the otherwise necessary DCL code.e     Disclaimer: JMHO Alan E. Feldmanm spamsink2001 at yahoo dot comn   ------------------------------  # Date: Thu, 05 Sep 2002 22:53:47 GMTi# From: "John Smith" <a@nonymous.com> J Subject: Re: Sick of the trolls, whiners, and non-customers in this forum?J Message-ID: <%fRd9.177587$8aG1.53752@news01.bloor.is.net.cable.rogers.com>  < "Terry C. Shannon" <terryshannon@attbi.com> wrote in message9 news:21xd9.153574$_91.209973@rwcrnsc51.ops.asp.att.net...  >'I > Umm. JF, for the umteenth time, are yoe even a DEC/CPQ/CUSTOMER? When'so the 7 > last time  you attended a DECUS? Bought a broduct????p     Terry,  L What difference does it make if he's a current customer or not? Many of JF's? points are just as valid today as they were when he was a majore customer/user.  G Does he HAVE to be a current customer or does he have to be a not-quitexE Armani Analyst (perhaps Brooks Brothers guy at best) in order for hisn, observations and opinions to be listened to?  K JF is in a position like so many of those who lurk/contribute here - former L customers who would dearly LOVE to be working with VMS again professionally.E But most of these people cannot take the professional risk with theirrJ careers to step forward and champion VMS within their respective companiesG without seeing significant evidence that HP is truly positively seriouswG about the VMS business. And for the others who are current users of VMSnF professionally, many (not all) feel like they are walking on quicksandL within their organizations with respect to the longevity of VMS within their shops.  L If HP realized that there are tens of thousands of JF's worldwide willing toH take that risk so long as HP stood up at the plate and made the attempt,J then maybe HP would do something truly in the public eye to market VMS. AsI unbelieveable as it may seem to the rest of the world, despite the rapingJJ the VMS customer base has been subjected to by Digital/Compaq, and perhapsC HP, there is still an enormous amount of goodwill available for the I OPERATING SYSTEM - maybe not the owners though, unless they take positiveu steps towards VMS.  G Just to make the playing field level, let's turn your question that youEI asked JF around - what was the last thing you bought that was VMS-relateduL and gave revenue to VMS's owner (whomever that was)? For me, it was a coupleF of low-end Alphas about a year ago, which we have subsequently sold to
 customers.  G In 500 words or less, what's HP going to do that's really eye-openinglyyG positive for VMS in the next 100 days? Market/advertise it widely, killoI Itanic and revive Alpha, announce a $200 million OpenVMS porting fund forrE ISV's? Or will it be to just wander around from exisiting customer toP0 existing customer, saying "Please don't leave."?  K I'm not picking on you or JF. One doesn't have to be a baseball player or abL person who has bought a ticket to a game in order to make observations aboutJ the situation leading up to the near-strike last week, or any other aspect of the game.   ------------------------------  % Date: Thu, 05 Sep 2002 19:58:48 -0400 - From: JF Mezei <jfmezei.spamnot@videotron.ca>VJ Subject: Re: Sick of the trolls, whiners, and non-customers in this forum?, Message-ID: <3D77EFB2.B0A67805@videotron.ca>   John Smith wrote: D > If HP realized that there are tens of thousands of JF's worldwide   H Please, don't bring apocalyptic scenarios into the discussion. The above; scenario would be enough to force many to commit hara-kiri..  M If I had been born a few generations later, I could threathen to clone myself H tens of thousands of times unless countries gave me access to their gold reserves :-) :-) :-) :-)   ------------------------------   Date: 5 Sep 2002 01:18:12 GMTr& From: peter@abbnm.com (Peter da Silva)J Subject: Re: Sick of the trolls, whiners, and non-customers in this forum?- Message-ID: <al6bck$kon@web.eng.baileynm.com>l  8 In article <magbnus4bvg9ta2gtnqinfedp78nhpud9s@4ax.com>,' Alan Greig  <a.greig@virgin.net> wrote: H > Had first Digital , then Compaq and perhaps HP helped me every step ofG > the way to push VMS but it failed in the marketplace, I would feel nooG > animosity But instead I've spent the last decade fighting them tryingd' > to push me towards Unix then Windows.o  C Anyone else wear their "Affinity" freebies to get a rise out of oldtF DEC hands? You can always tell the people who watched the company melt: down, they scowl at your "Affinity - VMS/NT" polo shirt...   --  O I've seen things you people can't imagine. Chimneysweeps on fire over the roofsqO of London. I've watched kite-strings glitter in the sun at Hyde Park Gate.  All L these things will be lost in time, like chalk-paintings in the rain.   `-_-'K Time for your nap.  | Peter da Silva | Har du kramat din varg, idag?    'U`E   ------------------------------   Date: 5 Sep 2002 17:53:02 GMTw& From: peter@abbnm.com (Peter da Silva)J Subject: Re: Sick of the trolls, whiners, and non-customers in this forum?- Message-ID: <al85lu$3i4@web.eng.baileynm.com>n  ? In article <21xd9.153574$_91.209973@rwcrnsc51.ops.asp.att.net>,.0 Terry C. Shannon <terryshannon@attbi.com> wrote:B > Umm. JF, for the umteenth time, are yoe even a DEC/CPQ/CUSTOMER?  ) What difference could that possibly make?0   -- .O I've seen things you people can't imagine. Chimneysweeps on fire over the roofs O of London. I've watched kite-strings glitter in the sun at Hyde Park Gate.  All.L these things will be lost in time, like chalk-paintings in the rain.   `-_-'K Time for your nap.  | Peter da Silva | Har du kramat din varg, idag?    'U`    ------------------------------   Date: 5 Sep 2002 18:03:33 GMTa& From: peter@abbnm.com (Peter da Silva)J Subject: Re: Sick of the trolls, whiners, and non-customers in this forum?- Message-ID: <al869l$4a2@web.eng.baileynm.com>o  > In article <L2xd9.131360$kp.763782@rwcrnsc52.ops.asp.att.net>,0 Terry C. Shannon <terryshannon@attbi.com> wrote:N > > Terry, neither you nor Bill have enough credibility left to be spending it > > this profligately.  @ > Shimatta! I have made a mistakle. Your refund is in the mail..  L You're sending me some credibility back? Cool... what were you doing muckingF around with my credibility in the first place? You been flogging otherH people's credibility? Speculating in hot credibility on the spot market?  M I tried grey-market credibility once, but it was badly cut with spin control,o1 so I'll be glad to get the real thing back again.o   --  O I've seen things you people can't imagine. Chimneysweeps on fire over the roofseO of London. I've watched kite-strings glitter in the sun at Hyde Park Gate.  AlliL these things will be lost in time, like chalk-paintings in the rain.   `-_-'K Time for your nap.  | Peter da Silva | Har du kramat din varg, idag?    'U`t   ------------------------------   Date: 5 Sep 2002 17:46:26 -0700r+ From: tonymcg27@yahoo.com.au (Tony McGrath)b Subject: Re: SOAP Toolkitc= Message-ID: <521ebd9f.0209051646.6e181ab3@posting.google.com>   R "Jakob Erber" <erberj@yahoo.de> wrote in message news:<3d7709b3$1@news.post.ch>... > Hello, > . > I tried to get the SOAP toolkit for VMS from > C > http://www.openvms.compaq.com/openvms/products/ips/soap/soap.html? > M > I registered and was promised, that a download url will be sent to me soon,. > but nothing happens. > $ > Did somebody actually get the KIT? > 	 > regardsl >  > Jakobt   G'day Jakob,F Yeah the same thing happened to me last week when I wanted to download. the USP kit. The promised email never arrived.E I just guessed what the required URL would be, based on the structurec= of the similar download URL for the CSWS kit. Worked a treat. E I don't want to get in trouble by announcing it here, I'll send you ao private email. Cheers from Oz,o   Tony -- Tony McGrath OpenVMS Support Groupn Toll Transport, IT Dept.$ Laverton North, VIC, Australia, 3006) tony_mcgrath (at) toll (dot) com (dot) aua   ------------------------------  # Date: Fri, 06 Sep 2002 02:52:29 GMT-1 From: "David J. Dachtera" <djesys.nospam@fsi.net>r Subject: The Gauntlet is Cast!' Message-ID: <3D781DC5.143EF9CF@fsi.net>O  F Got the following from InfoWorld today. Anyone interested in casting aH similar challenge at VMS Management's feet should see my text below this	 info. ...l  & SUN YET TO ANSWER CHALLENGE FROM USERS  , Posted September 04, 2002 05:08 Pacific Time  7 SUN MICROSYSTEMS HAS yet to respond to a challenge from 4 a group of its users made earlier this week, calling4 for an public debate about the future of the Solaris! operating systems on Intel chips.f  3 John Groenveld, associate research engineer at Penno5 State University, shelled out $9,292 to place an openU5 letter in the Tuesday edition of the San Jose Mercury - News that challenged Scott McNealy, chairman,u0 president and chief executive officer at Sun, to4 discuss the future of Solaris on the Intel platform.4 Groenveld is one of many users who have battled with/ Sun over the last nine months to make Solaris 9 5 available on a wide range of Intel-based hardware andd0 to gain support from Sun for the product. Sun in1 January halted development on Solaris 9 for Intelb3 chips as a cost-cutting measure. Last month it madet4 Solaris on Intel available as a product, but only on its Sun LX50 server.     For the full story:vM http://www.infoworld.com/articles/hn/xml/02/09/04/020904hndebate.xml?0905thamb   (The URL may have wrapped.)h  D Anyone interested in casting a similar challenge at VMS Management'sC feet should e-mail me privately. How to demung the reply-to address-E should be obvious, but it would help me keep e-mail separate if folksAH would write to me at djesys at dls dot net or djbenedict at aol dot com./ My snail-mail address can be found at this URL:.  " http://www.djesys.com/contact.html  H If 10,000 people will each commit to sending $1 U.S. (send no money now,D I'm just trying to gauge interest), we could buy ad space in a major0 newspaper. We'd need to decide where, of course.  D I'll be contacting publications in an effort to assess the financialD scale of this undertaking. I'll report back to the group when I have some numbers...e   -- p David J. Dachtera  dba DJE Systemsn http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/l   ------------------------------   Date: 5 Sep 2002 22:53:28 -0600a+ From: young_r@encompasserve.org (Rob Young)s" Subject: Re: The Gauntlet is Cast!3 Message-ID: <rCmE28372xE2@eisner.encompasserve.org>0  [ In article <3D781DC5.143EF9CF@fsi.net>, "David J. Dachtera" <djesys.nospam@fsi.net> writes:    > ( > SUN YET TO ANSWER CHALLENGE FROM USERS > 3 > January halted development on Solaris 9 for Intela5 > chips as a cost-cutting measure. Last month it madeS6 > Solaris on Intel available as a product, but only on > its Sun LX50 server. >   ; 	They have to put up artificial barriers.  The LX50 is a 1Uo= 	server.  If for example they supported a 4-way or 8-way Delli: 	box, some wise-acre would come along and run Oracle on it< 	and show it outperforms Oracle on UltraSparc.  Oh... that's$ 	right, Sun doesn't do tpmC anymore.  @ 	But for the same reason, Linux runs on LX50.  Linux and Solaris9 	on Intel is for edge computing, remember Zander's words?t  1 http://www.eweek.com/article2/0,3959,30668,00.aspa  O We're hot about Linux, we think it's a great low-end play in the edge computingiM market, and the product we are bringing out this summer will catapult us into 
 the lead.      	Catapult?  Okay.o  ; 	Don't expect Scotty to budge.  Anymore than you would havenA 	expected Digital's senior management to make a radical change in ? 	direction in 1987 or 1988 when revenues and profits were still:3 	growing.  Shoot, guess Sun's revs aren't growing. w  A 	Anyhow, point in this ... Sun can't embrace Intel, can't embrace@/ 	Linux.  The good ship Sun is left to flounder:-    M "As the weak IT spending picture weighs on the sector, Sun is also challenged)N by several company-specific issues," wrote J.P. Morgan analyst Bill Shope. "WeE believe that Sun continues to grapple with a core market slowdown andd increasing competition."  H In the low-end, Sun is facing a threat from the Microsoft-Linux industryN standard server vendors, such as Dell, he told clients. The threat is startingN to creep up into the midrange Unix server market, where pricing competition isO fierce. In the high end, Sun faces off with IBM and Hewlett-Packard (HPQ: news,u chart, profile).  N At least eight Wall Street brokerages cut their earnings and revenue estimatesJ for the computer hardware and services company, not only for Sun's current0 fiscal first quarter but also for 2003 and 2004.    C 	You know.. instead of Sun being the Novell of the 2000's.... thinkt> 	of Sun as the Apple of the 2000's.  Apple is still alive, but@ 	largely irrelevant.  One-trick pony Sun is headed there as they< 	refuse to adopt the industry standard.  Isn't Apple so cute8 	running on those PowerPC chips?   Ooohhhhh ... Aaahhhh.  @ 	That San Jose Mercury article could have been an Apple focused B 	article 12 years ago urging Steve Jobs to get over to Intel from 
 	Motorola.   				Rob    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 slipstreamx> The clock reads daylight savin'         Rollin' in the thunder  -                                 -- Neil Youngs   ------------------------------  % Date: Thu, 05 Sep 2002 23:02:14 -0500t7 From: "Craig A. Berry" <craigberry@nospam.telocity.com> " Subject: Re: The Gauntlet is Cast!G Message-ID: <craigberry-2D5ADC.23021305092002@news.directvinternet.com>c  ' In article <3D781DC5.143EF9CF@fsi.net>,G3  "David J. Dachtera" <djesys.nospam@fsi.net> wrote:a  F > Anyone interested in casting a similar challenge at VMS Management's# > feet should e-mail me privately. c  B Dave, in your enthusiasm, you seem to have left out a few salient C points, such as challenge to do what?  I had several snide guesses bD prepared (the most logical one based on your post being to have the H OpenVMS group support Solaris on Intel!) but as much fun as the guesses H might be it would better to have you tell us what you want to challenge  them to do.b   ------------------------------  $ Date: Thu, 5 Sep 2002 17:01:50 -0400; From: "Brian Tillman" <tillman_brian@notnoone.notnohow.com>c. Subject: Re: VAX / Network / Hardware question$ Message-ID: <3d77c69b$1@news.si.com>  D >I've not dealt with a WS150, but I have used DR250's.  Since no oneB >else has responded, I'll chip in.  You can connect to a DR250 and@ >execute Decnet-IV style management commands with NCP.  You use: > 
 >$ MCR NCP8 >NCP> set executive node DECRTR user foo password xxxxxx  = That should be EXECUTOR, not EXECUTIVE.  See HELP SET in NCP.y -- nA Brian Tillman                   Internet: tillman_brian at si.com A Smiths Aerospace                          tillman at swdev.si.comE= 3290 Patterson Ave. SE, MS      Addresses modified to prevent0< Grand Rapids, MI 49512-1991     SPAM.  Replace "at" with "@"8        This opinion doesn't represent that of my company   ------------------------------  # Date: Thu, 05 Sep 2002 19:25:59 GMTl# From: NetBoot <netboot@netboot.com>o% Subject: VAXstation 3100 Model 76 SPXB5 Message-ID: <MPG.17e1790e3b02836a98970f@news.ptd.net>   F Can anyone tell me about this system and compared to other VaxStation  systems?   ------------------------------  # Date: Thu, 05 Sep 2002 19:41:49 GMT,4 From: Tim Llewellyn <tim.llewellyn@blueyonder.co.uk>) Subject: Re: VAXstation 3100 Model 76 SPXJ0 Message-ID: <3D77B1F3.D9A9D160@blueyonder.co.uk>  D Better than all the other VAXStation 3100 models, the SPX controllerJ is optimized for X rather than VWS, which is good. Not as good/fast as theG VAXStation 4000/60, though some of those only have a mono frame buffer.c   Regards,       NetBoot wrote: > G > Can anyone tell me about this system and compared to other VaxStationR
 > systems?   -- i tim.llewellyn@blueyonder.co.uk s  H * PLEASE NOTE tim.llewellyn@cableinet.co.uk address is NO LONGER VALID *   ------------------------------  # Date: Thu, 05 Sep 2002 19:51:52 GMT # From: NetBoot <netboot@netboot.com>o) Subject: Re: VAXstation 3100 Model 76 SPXm5 Message-ID: <MPG.17e17f1ca209436b989711@news.ptd.net>-  1 In article <3D77B1F3.D9A9D160@blueyonder.co.uk>, r& tim.llewellyn@blueyonder.co.uk says... > F > Better than all the other VAXStation 3100 models, the SPX controllerL > is optimized for X rather than VWS, which is good. Not as good/fast as theI > VAXStation 4000/60, though some of those only have a mono frame buffer.$ > 
 > Regards, >  >  >  > NetBoot wrote: > > I > > Can anyone tell me about this system and compared to other VaxStation  > > systems? >  >     G hmm, Ok, so what OS can I put on it?  I'm also getting the 4000/60 too.e   ------------------------------  # Date: Thu, 05 Sep 2002 20:17:47 GMTd4 From: Tim Llewellyn <tim.llewellyn@blueyonder.co.uk>) Subject: Re: VAXstation 3100 Model 76 SPX 0 Message-ID: <3D77BA61.C14F7489@blueyonder.co.uk>   NetBoot wrote: > 2 > In article <3D77B1F3.D9A9D160@blueyonder.co.uk>,( > tim.llewellyn@blueyonder.co.uk says... > >tH > > Better than all the other VAXStation 3100 models, the SPX controllerN > > is optimized for X rather than VWS, which is good. Not as good/fast as theK > > VAXStation 4000/60, though some of those only have a mono frame buffer.w > >n > > Regards, > >i > >i > >p > > NetBoot wrote: > > >eK > > > Can anyone tell me about this system and compared to other VaxStation  > > > systems? > >  > >s > I > hmm, Ok, so what OS can I put on it?  I'm also getting the 4000/60 too.r   Lucky person :-)  K VMS and presumably Ultrix/VAX will work. Maybe some of the free unices too.rE I think the SPX required VMS 5.5-2HW2 or something like that, but I'dl go for 7.2 or 7.3.   Regards,   -- g tim.llewellyn@blueyonder.co.uk m  H * PLEASE NOTE tim.llewellyn@cableinet.co.uk address is NO LONGER VALID *   ------------------------------  # Date: Thu, 05 Sep 2002 18:08:39 GMTd4 From: Tim Llewellyn <tim.llewellyn@blueyonder.co.uk>/ Subject: Re: VaxStation 4000/60 need hard driver0 Message-ID: <3D779C1E.E1F9FD40@blueyonder.co.uk>   NetBoot wrote: > E > In article <al7vjm$17ee$1@half.spin.it>, balzano-spam-avoid-@iol.it 	 > says.... > >u4 > > "NetBoot" <netboot@netboot.com> wrote in message3 > > news:MPG.17e146b1de232a2a98970c@news.ptd.net...   > > > What hard drive do I need? > >  > > scsi > >e# > > > Are these drives proprietary?F > >, > > no > >p > > L. > >d > >a > >u >  > Is there a size limit?  E Something dim in the mists of my memory suggests there is a thermallyoE imposed limit on the type of disk. I certainly remember having RZ26's?? in that box, maybe you can put an rz28 (2gig), rz29 is too hot.u  . I guess if you're willing to plug and play ...   HTH. l   -- t tim.llewellyn@blueyonder.co.uk n  H * PLEASE NOTE tim.llewellyn@cableinet.co.uk address is NO LONGER VALID *   ------------------------------  $ Date: Thu, 5 Sep 2002 11:24:50 -0700# From: "Tom Linden" <tom@kednos.com>i/ Subject: RE: VaxStation 4000/60 need hard drive 9 Message-ID: <CIEJLCMNHNNDLLOOGNJIKEOOFKAA.tom@kednos.com>t  B I have a couple of 7200 rpm IBM drives (4.5 and 18GB) in a 4000/909 which is the identical enclosure, and have had no problemt   >-----Original Message-----t< >From: Tim Llewellyn [mailto:tim.llewellyn@blueyonder.co.uk], >Sent: Thursday, September 05, 2002 11:09 AM >To: Info-VAX@Mvb.Saic.Com0 >Subject: Re: VaxStation 4000/60 need hard drive >  >  >a >  >NetBoot wrote:l >>  F >> In article <al7vjm$17ee$1@half.spin.it>, balzano-spam-avoid-@iol.it
 >> says... >> >5 >> > "NetBoot" <netboot@netboot.com> wrote in messaget4 >> > news:MPG.17e146b1de232a2a98970c@news.ptd.net...! >> > > What hard drive do I need?m >> >	 >> > scsi  >> >$ >> > > Are these drives proprietary? >> > >> > no- >> > >> > L.  >> > >> > >> > >>   >> Is there a size limit?r > F >Something dim in the mists of my memory suggests there is a thermallyF >imposed limit on the type of disk. I certainly remember having RZ26's@ >in that box, maybe you can put an rz28 (2gig), rz29 is too hot. >v/ >I guess if you're willing to plug and play ...m >  >HTH.  >  >--   >tim.llewellyn@blueyonder.co.uk  >sI >* PLEASE NOTE tim.llewellyn@cableinet.co.uk address is NO LONGER VALID *n >a >---' >Incoming mail is certified Virus Free.n; >Checked by AVG anti-virus system (http://www.grisoft.com).e@ >Version: 6.0.381 / Virus Database: 214 - Release Date: 8/2/2002 >u ---a& Outgoing mail is certified Virus Free.: Checked by AVG anti-virus system (http://www.grisoft.com).? Version: 6.0.381 / Virus Database: 214 - Release Date: 8/2/2002i   ------------------------------  # Date: Thu, 05 Sep 2002 19:48:10 GMTg* From: "Jack Peacock" <peacock@simconv.com>/ Subject: Re: VaxStation 4000/60 need hard drivenC Message-ID: <_xOd9.321882$Aw4.13362877@bin2.nnrp.aus1.giganews.com>r  0 "NetBoot" <netboot@netboot.com> wrote in message/ news:MPG.17e146b1de232a2a98970c@news.ptd.net...w; > What hard drive do I need?  Are these drives proprietary?f >eL I'd recommend you avoid Quantum Fireball drives.  I've had trouble with themE before, something on the command tag queueing I believe.  I do have apL 3100/10 running on a Micropolis 2210 1GB drive now but I've had trouble withI Micropolis on Alpha machines so you might want to avoid those if you have  any old ones around.    Jack Peacocke   ------------------------------  # Date: Thu, 05 Sep 2002 22:17:50 GMTf# From: "John Smith" <a@nonymous.com>n/ Subject: Re: VaxStation 4000/60 need hard drivehK Message-ID: <iKQd9.177038$8aG1.155961@news01.bloor.is.net.cable.rogers.com>i  0 "NetBoot" <netboot@netboot.com> wrote in message/ news:MPG.17e146b1de232a2a98970c@news.ptd.net...-; > What hard drive do I need?  Are these drives proprietary?o  K I seem to recollect that there was some firmware required to make a non-DEC L drive able to run as a boot disk. But that may be VMS version-specific - the" 4000's were released on VMS 5.5-x.  K I used Toshiba 2+ Gb drives way back when in the VaxStation 4000 series andiL Toshiba provided me with a different firmware chip to install on the drives.K That was back in the days of them having a disk products division office in.I Marlborough or thereabouts - a guy named Charlie DiChiara, I think. Mightc- still have some correspondence kicking aroundl   ------------------------------  # Date: Fri, 06 Sep 2002 02:16:14 GMTe1 From: "David J. Dachtera" <djesys.nospam@fsi.net> ( Subject: Re: VMS Printing To CA/Dispatch' Message-ID: <3D781549.75F91973@fsi.net>t   Jan-Erik Sderholm wrote:  > ? > SNA/RJE can't print directly to a VMS queue, you have to scane; > for files and submit them for printing in some batch job.r  4 I believe it can print to a spooled device, however.  > > SNA/PrE can (as the name says) print directly to VMS queues,= > but there is a limit on the number of queues at 127 or 128.   G I seem to recall having well over 200 PrE streams running through a -CTr? gateway. We had to increase CHANNELCNT to permit it, of course.    -- c David J. Dachterab dba DJE Systems1 http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/p   ------------------------------  % Date: Fri, 06 Sep 2002 09:50:44 +1000 1 From: Paddy O'Brien <paddy.o'brien@tg.nsw.gov.au>i1 Subject: Re: Weird, but you get that - marketing.i, Message-ID: <3D77EDD4.9000306@tg.nsw.gov.au>   > ? > Wow! Backwards apostrophe disease! First case I've ever seen.  >  >  > [...]     H Hadn`t you heard that we`ve almost run out of ' (damn, I`ve just wasted 	 another).    Regards, Paddy O`Brien      G ***********************************************************************b  C "This electronic message and any attachments may contain privilegedo> and confidential information intended only for the use of the B addressees named above.  If you are not the intended recipient of C this email, please delete the message and any attachment and adviseeB the sender.  You are hereby notified that any use, dissemination, 7 distribution, reproduction of this email is prohibited.0  A If you have received the email in error, please notify TransGrid iA immediately.  Any views expressed in this email are those of the 9= individual sender except where the sender expressly and with iC authority states them to be the views of TransGrid.  TransGrid useso> virus scanning software but excludes any liability for viruses contained in any attachment."e  G ***********************************************************************a   ------------------------------  # Date: Thu, 05 Sep 2002 19:15:31 GMTo, From: "Kenneth Farmer" <kfarmer@openvms.org>O Subject: What features/capabilities would you like to see available on OpenVMS?g< Message-ID: <n3Od9.1537$jF4.362242@twister.southeast.rr.com>  5 From the General forum on OpenVMS.org by Jim Johnson:e2 http://www.openvms.org/phorum/read.php?f=6&i=4&t=4   ----------------------  K I'm interested in finding out what sort of features or capabilities OpenVMS.H customers might find interesting. I have a personal reason for this -- IK have a small company that might be interested in building infrastructure orwL integration products for OpenVMS, and I'd like to get a handle on what sorts( of products there would be a demand for.  J However, I'm also interested generally. I don't hear a lot of conversationE in comp.os.vms or in here that I can see could translate into producttL opportunities. Is the market truly dead, or have we all just stopped asking?  I So, for right now, I'd like us to brainstorm a bit about what we'd reallylE like to see to make OpenVMS better in our environments -- it could beg6 additional VMS-specific features, portability aids, or; interoperation/integration capabilities. Or something else.d  J So, to get a discussion going, below is a list of things that have occured8 to me in the past that I posted in comp.os.vms in march.  L I'd say "what do you want?", only that's got some bad connotations for thoseK that followed a certain TV series... Instead, let's get some ideas flowing.e   Jim.   TP Application Building Tools   D - A clusterwide, transaction-capable, persistant named property bag.  F This would provide a simple to use container for name/attribute pairs,I similar to clusterwide logical names, with the addition that the contentsyK would be automatically hardened to disk, and the access could be done under/ transaction control.  G This would allow clusterwide applications to cheaply share data betweenpA instances of themselves, and use transactions to maintain logicalg consistency.   - Transaction capable IPCg  J $ICC offers a high performance local and cluster communications mechanism.I This would layer on top of that and provide a fully automatic transaction J branch management feature. Thus, a client-server application would be ableK to just issue a 'send' request while inside a transaction and know that thetI actions at the remote end would automatically be processed as part of ther transaction.   View it as ACMS-extra-lite.a   - TIP interoperability  K Provide an API that gateways between DECdtm and TIP. Include the support toaE actually connect and interopate with a Microsoft MTS based front end.      Java Interoperability:  ! - ACMS to J2EE migration toolkit.o  D This would process TDL, CDD, and the other associated files to buildL equivalent JSP, EJB, and servlet structures. As well as supporting access toE Java-only logic, this would either use an available infrastructure tonC connect back to an existing ACMS task, or incorporate one to do so.p  J This would let an ACMS customer either interoperate with J2EE, or migrate,2 over time, to a fully J2EE compliatnt environment.   - Other to J2EE migration   F This is just a placeholder to point out that the other TP monitor-typeI things that Digital shipped could also be migrated in a similar manner --o DECintact, MessageQ, etc.n   --------------------------  H Let him know what features you would like to see available.  Respond at:2 http://www.openvms.org/phorum/read.php?f=6&i=4&t=4       Keny   --   Kenneth Farmer http://www.Tru64.org http://www.OpenVMS.org http://www.LinuxHPC.orgc   ------------------------------  % Date: Thu, 05 Sep 2002 19:49:14 -0000n- From: wspencer@ap.nospam.org (Warren Spencer)oS Subject: Re: What features/capabilities would you like to see available on OpenVMS?s5 Message-ID: <9280A9BF9warrenspencer1977@216.168.3.30>   - kfarmer@openvms.org (Kenneth Farmer) wrote in 2 <n3Od9.1537$jF4.362242@twister.southeast.rr.com>:   6 >From the General forum on OpenVMS.org by Jim Johnson:3 >http://www.openvms.org/phorum/read.php?f=6&i=4&t=4r >n >----------------------p >rD >I'm interested in finding out what sort of features or capabilitiesG >OpenVMS customers might find interesting. I have a personal reason foriD >this -- I have a small company that might be interested in buildingH >infrastructure or integration products for OpenVMS, and I'd like to get@ >a handle on what sorts of products there would be a demand for. >n> >However, I'm also interested generally. I don't hear a lot ofF >conversation in comp.os.vms or in here that I can see could translateE >into product opportunities. Is the market truly dead, or have we allo >just stopped asking?  > C >So, for right now, I'd like us to brainstorm a bit about what we'deD >really like to see to make OpenVMS better in our environments -- it@ >could be additional VMS-specific features, portability aids, or< >interoperation/integration capabilities. Or something else. >tC >So, to get a discussion going, below is a list of things that have-A >occured to me in the past that I posted in comp.os.vms in march.m >CG >I'd say "what do you want?", only that's got some bad connotations foreI >those that followed a certain TV series... Instead, let's get some ideasn
 >flowing.  >a >Jim.< >8 >TP Application Building Tools >.E >- A clusterwide, transaction-capable, persistant named property bag.s >gG >This would provide a simple to use container for name/attribute pairs,lA >similar to clusterwide logical names, with the addition that theaG >contents would be automatically hardened to disk, and the access coulds# >be done under transaction control.s >oH >This would allow clusterwide applications to cheaply share data betweenB >instances of themselves, and use transactions to maintain logical
 >consistency.o >a >- Transaction capable IPC > @ >$ICC offers a high performance local and cluster communicationsI >mechanism. This would layer on top of that and provide a fully automaticcI >transaction branch management feature. Thus, a client-server applicationoH >would be able to just issue a 'send' request while inside a transactionC >and know that the actions at the remote end would automatically be & >processed as part of the transaction. >s >View it as ACMS-extra-lite. >N >- TIP interoperabilityo >DI >Provide an API that gateways between DECdtm and TIP. Include the supportSI >to actually connect and interopate with a Microsoft MTS based front end.a >r >  >Java Interoperability:a > " >- ACMS to J2EE migration toolkit. >vE >This would process TDL, CDD, and the other associated files to build0C >equivalent JSP, EJB, and servlet structures. As well as supporting > >access to Java-only logic, this would either use an availableH >infrastructure to connect back to an existing ACMS task, or incorporate >one to do so. t >mB >This would let an ACMS customer either interoperate with J2EE, or< >migrate, over time, to a fully J2EE compliatnt environment. >y >- Other to J2EE migration >tG >This is just a placeholder to point out that the other TP monitor-type0G >things that Digital shipped could also be migrated in a similar manneru >-- DECintact, MessageQ, etc.l >  >--------------------------9 >1I >Let him know what features you would like to see available.  Respond at:h3 >http://www.openvms.org/phorum/read.php?f=6&i=4&t=4  >c >Ken >. >t >Kenneth Farmert >http://www.Tru64.org  >http://www.OpenVMS.org  >http://www.LinuxHPC.org   Kenneth,  B Here's a couple of things that frustrate me in the last 12 months:  R * An SMTP API that can do MIME encoding right (instead of those interactive tools  like MIME and MPACKl  J * Programatic access to ODBC-compliant databases hosted on other platforms   * An XML parser callable from C1    J Nothing very big here, but there's huge inter-operability gains to be had.   ws   -- m   Warren Spencer' Senior Software Engineer (not a writer)  The Associated Press  < ** Time flies like an arrow.  Fruit flies like a bananna. **   ------------------------------   Date: 5 Sep 2002 15:25:15 -0600 - From: frey@encompasserve.org (Sharon Guthrie)SS Subject: Re: What features/capabilities would you like to see available on OpenVMS?o3 Message-ID: <QFfpR2Ouu0Is@eisner.encompasserve.org>   k In article <n3Od9.1537$jF4.362242@twister.southeast.rr.com>, "Kenneth Farmer" <kfarmer@openvms.org> writes:y  L > However, I'm also interested generally. I don't hear a lot of conversationG > in comp.os.vms or in here that I can see could translate into product2N > opportunities. Is the market truly dead, or have we all just stopped asking? > < 	Heck no, it's just that VMS is damned near perfect already!  N (Ok, now all the nitpickers are going to chime in and discredit me.  heheheh  E Go ahead and make suggestions for improvements, I'm only being a VMS   cheerleader here.)   Sharon   ------------------------------   Date: 5 Sep 2002 15:33:35 -0600h- From: Kilgallen@SpamCop.net (Larry Kilgallen)aS Subject: Re: What features/capabilities would you like to see available on OpenVMS?,3 Message-ID: <NtdMeS7tsPgL@eisner.encompasserve.org>   e In article <9280A9BF9warrenspencer1977@216.168.3.30>, wspencer@ap.nospam.org (Warren Spencer) writes:   D > Here's a couple of things that frustrate me in the last 12 months: > T > * An SMTP API that can do MIME encoding right (instead of those interactive tools  > like MIME and MPACK  > L > * Programatic access to ODBC-compliant databases hosted on other platforms > ! > * An XML parser callable from Cd  A The Apache Group has one called Xerces, which various people haveNB ported to VMS.  I thought I read that VMS V7.3-1 comes with a portB of Xerces on one of the ancillary disks (not part of the base OS).   ------------------------------  % Date: Thu, 05 Sep 2002 22:55:51 +020072 From: martin@radiogaga.harz.de (Martin Vorlaender)S Subject: Re: What features/capabilities would you like to see available on OpenVMS?.; Message-ID: <3d77c4d7.524144494f47414741@radiogaga.harz.de>   . Warren Spencer (wspencer@ap.nospam.org) wrote:. > kfarmer@openvms.org (Kenneth Farmer) wrote: 8 > >From the General forum on OpenVMS.org by Jim Johnson:5 > >http://www.openvms.org/phorum/read.php?f=6&i=4&t=4s > >i > >----------------------e > >sF > >I'm interested in finding out what sort of features or capabilities2 > >OpenVMS customers might find interesting. [...] >o! > * An XML parser callable from C,  : I didn't have any major troubles to build expat[1] on VMS.   cu,    Martin  * [1] http://sourceforge.net/projects/expat/ -- EG So long, and thanks        | Martin Vorlaender  |  VMS & WNT programmer-4 for all the books...       | work: mv@pdv-systeme.deK In Memoriam Douglas Adams  |       http://www.pdv-systeme.de/users/martinv/0;             1952-2001      | home: martin@radiogaga.harz.deA   ------------------------------  # Date: Thu, 05 Sep 2002 20:20:03 GMTB4 From: Tim Llewellyn <tim.llewellyn@blueyonder.co.uk>Y Subject: Re: What features/capabilities would you like to see available on OpenVMS? OpenVe0 Message-ID: <3D77BAEA.B41861B2@blueyonder.co.uk>   JF Mezei wrote:  >   f > N > However, perhaps one idea would be the ability to "archive" a process. StoreN > all symbols and process logicals in a file, along with current directory, pfL > key definitions, command recall buffer and other process attributes. Then,N > these could be restored later on. (for instance, you need to log off and logH > back in again to use a new DCLTABLES.EXE, but want to maintain all theJ > environment you had before, or you want to maintain your environment and3 > restore it after a VMSINSTAL which zaps it clean.f > 2 > something such as PROCESS/SAVE=ALL myproject.dat9 > and                       PROCESS/RESTORE myproject.dati > J > where /SAVE and /RESTORE would have values such as ALL, LOGICAL, SYMBOL,  > RECALL, CURRENTDIR, KEYS etc).  7 Surely this is all achievable with a DCL script or two?    -- P tim.llewellyn@blueyonder.co.uk t  H * PLEASE NOTE tim.llewellyn@cableinet.co.uk address is NO LONGER VALID *   ------------------------------  % Date: Thu, 05 Sep 2002 16:20:40 -0400 - From: JF Mezei <jfmezei.spamnot@videotron.ca>sY Subject: Re: What features/capabilities would you like to see available on OpenVMS? OpenVu, Message-ID: <3D77BC8E.C2068EAF@videotron.ca>   Kenneth Farmer wrote:2K > So, for right now, I'd like us to brainstorm a bit about what we'd really09 > like to see to make OpenVMS better in our environments c  I Conversion tools to allow VMS users to read/use documents coming from the9I outside (eg: proprietary microsoft file formats). In other words, retake,t0 revive and rejuvenate the CDA converter library.  H > This would provide a simple to use container for name/attribute pairs,K > similar to clusterwide logical names, with the addition that the contentstM > would be automatically hardened to disk, and the access could be done undero > transaction control.  N It is called an indexed file with a record having 2 fields: a name field and aN value field. ALL-IN-1 has had that for decades (the PST, personal symbol tableI file). In A1 terms, any variable that begins with "$" is treated as a PST2I variable and is automatically fetched from and stored in the PST and thuso survives image activations.e  L However, perhaps one idea would be the ability to "archive" a process. StoreL all symbols and process logicals in a file, along with current directory, pfJ key definitions, command recall buffer and other process attributes. Then,L these could be restored later on. (for instance, you need to log off and logF back in again to use a new DCLTABLES.EXE, but want to maintain all theH environment you had before, or you want to maintain your environment and1 restore it after a VMSINSTAL which zaps it clean.n  0 something such as PROCESS/SAVE=ALL myproject.dat% and			  PROCESS/RESTORE myproject.datb  H where /SAVE and /RESTORE would have values such as ALL, LOGICAL, SYMBOL, RECALL, CURRENTDIR, KEYS etc).   ------------------------------  % Date: Thu, 05 Sep 2002 17:22:00 -0400a- From: JF Mezei <jfmezei.spamnot@videotron.ca>eY Subject: Re: What features/capabilities would you like to see available on OpenVMS? OpenVc, Message-ID: <3D77CAEB.67575070@videotron.ca>   Sharon Guthrie wrote:nE >         Heck no, it's just that VMS is damned near perfect already!t  M Are you kidding ? Not only does VMS still require users to press return aftereJ having typed in their username and another return for the password, but it< still  lacks automatic field completion on those two inputs.  L I should be able to just type "j" and then VMS automatically completes it toI "jfmezei" for the username, and I should be able to type "i" as the firstoK letter of the password and VMS should be able to automatically complete thebL password field, provided that sufficient letters have been typed to uniquelyN identify the username. And there should be no need to press return at the end.  N If you're going to compete against Microsoft, you might as well compete at theN same level: put all sorts of fancy features that look great for marketing evenI though from a security perspective they might be a disaster. Think of thesJ claims of productivity improvements that could be made with such a featureL since hundreds of thousands of users would save a few seconds every day when they login.*  K This feature alone should be sufficient to raise productivity levels in ther6 USA to boost its economy and take it out of recession.   :-) :-) :-) :-) :-) :-)i   ------------------------------  % Date: Thu, 05 Sep 2002 22:34:18 +0200o9 From: Jan-Erik =?iso-8859-1?Q?S=F6derholm?= <aaa@aaa.com> Y Subject: Re: What features/capabilities would you like to see available on OpenVMS? OpenVd' Message-ID: <3D77BFCA.D69AD051@aaa.com>n   Warren Spencer wrote:t > 
 > Kenneth, > D > Here's a couple of things that frustrate me in the last 12 months: > S > * An SMTP API that can do MIME encoding right (instead of those interactive toolss > like MIME and MPACKh  F Yes, the MIME tool is built to be interactive, but MPACK/MUNPACK works
 just great& in batch, if that was what you wanted.   > L > * Programatic access to ODBC-compliant databases hosted on other platforms  8 Well, there is realy no ODBC-compliant databases at all.A But there is a lot of databases that have ODBC-drivers that worksr6 with the ODBC API in the different Windows plattforms.  C ODBC is just an (standard) interface between the appication and the9# specific database network protocol..  > ODBC is *not* a network protocol. All ODBC drivers for all the< major databases uses the specific network protocol from each? vendor. Rdb uses SQL/Services, Oracle uses SQL*net, Sybase usese OpenClient and so on.M  D And the database *server* doesn't realy know that the access is done  through something called "ODBC".  B There was at a time a "ODBC gateway for Rdb" that you could use to> connect to other databases using ODBC, but this product used a; Windows box as the "ODBC-gateway" to reach the ODBC driver.e  9 What we need is support for the different vendors networks protocol API's on VMS.  A I think "Attunity Connect" (or whatever the name is) does provideu9 some sort of standard API usable for different databases.h     Best Regards Jan-Erik Sderholm.a   ------------------------------  % Date: Thu, 05 Sep 2002 15:56:49 -0500n7 From: "Craig A. Berry" <craigberry@nospam.telocity.com>tY Subject: Re: What features/capabilities would you like to see available on OpenVMS? OpenVt% Message-ID: <3d77c512$1_3@nopics.sjc>    Warren Spencer wrote:   D > Here's a couple of things that frustrate me in the last 12 months: > T > * An SMTP API that can do MIME encoding right (instead of those interactive tools  > like MIME and MPACKe  G There are a number of existing ways to skin this cat. Probably the Perl ! module MIME::Lite is the easiest.   L > * Programatic access to ODBC-compliant databases hosted on other platforms  > Are all of these existing products unsuitable for some reason?  # Attunity: <http://www.attunity.com>l  < EasySoft: <http://www.easysoft.com/products/2002/main.phtml>  I OpenRDA: <http://www.atinet.com/Products/OpenRDA/ODBC_UNIX/ODBC_Unix.asp>o    ! > * An XML parser callable from C   5 expat 1.95.4 or later has VMS support out of the box:   ( <http://sourceforge.net/projects/expat/>   ------------------------------  # Date: Fri, 06 Sep 2002 02:34:43 GMTs1 From: "David J. Dachtera" <djesys.nospam@fsi.net>uY Subject: Re: What features/capabilities would you like to see available on OpenVMS? OpenVu' Message-ID: <3D78199D.8CFCBDE5@fsi.net>t   Kenneth Farmer wrote:o > 7 > From the General forum on OpenVMS.org by Jim Johnson:i4 > http://www.openvms.org/phorum/read.php?f=6&i=4&t=4 >  > ---------------------- > M > I'm interested in finding out what sort of features or capabilities OpenVMSuJ > customers might find interesting. I have a personal reason for this -- IM > have a small company that might be interested in building infrastructure oroN > integration products for OpenVMS, and I'd like to get a handle on what sorts* > of products there would be a demand for. > L > However, I'm also interested generally. I don't hear a lot of conversationG > in comp.os.vms or in here that I can see could translate into productnN > opportunities. Is the market truly dead, or have we all just stopped asking?  H Try a Google search of comp.os.vms for "wish list". I posted a number ofG items a year or so ago, mostly pertaining to DCL programming and usage.r  H Other posters have already listed some *VERY* good items in this thread.   -- a David J. Dachtera, dba DJE Systems  http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/a   ------------------------------  % Date: Fri, 06 Sep 2002 08:59:00 +1000 . From: Burnie M <burniem.NOSPAM@ozemail.com.au>J Subject: Re: Who are the largest VMS customers. Top ten? (Was: VMS and HP)8 Message-ID: <adofnu8i0rthni0o0uetvl2opinn6hrlpi@4ax.com>  E On 5 Sep 2002 05:01:26 -0700, bobmarlow@postmaster.co.uk (Bob Marlow)l wrote:  ~ >"C.W.Holeman II" <cwhii5@Julian5Locals.com> wrote in message news:<aFBd9.8528$LI2.565066@newsread2.prod.itd.earthlink.net>... >> JF Mezei wrote: >> .K >> > HP may not care about what the many outside the circle of the few verybO >> > large and important customers think. HP may be focused only on keeping the-= >> > remaining VMS large customers who generate the profits. o >> t. >> Who are the largest VMS customers. Top ten?    $ >No.of Vaxes or money? In any case:-E >BP and Shell were very heavy users, though I don't know the figures.iB >BAE Systems (Uk) would be big, I think, as they took what was GEC >Other oil/defence companies?t    
 Barclays Bank    ------------------------------  % Date: Fri, 06 Sep 2002 07:54:13 +0200t From: Dirk Munk <munk@home.nl>J Subject: Re: Who are the largest VMS customers. Top ten? (Was: VMS and HP)1 Message-ID: <BkXd9.1625$H6.277948@zwoll1.home.nl>    C.W.Holeman II wrote:  > JF Mezei wrote:w >  > H >>HP may not care about what the many outside the circle of the few veryL >>large and important customers think. HP may be focused only on keeping the: >>remaining VMS large customers who generate the profits.  >  > - > Who are the largest VMS customers. Top ten?i > N All kind of companies and government institutions HP/Compaq is not allowed to J mention. That is part of the tragedy. These customers rely heavely on the ? reliability and quality of VMS, and we are not allowed to know.T   ------------------------------  $ Date: Thu, 5 Sep 2002 20:50:22 -04001 From: "Island (hpaq.net)" <dbturner@islandco.com>  Subject: XP1000 667 selling !!!a/ Message-ID: <unfv0oddi1bf68@news.supernews.com>p   www.hpaq.net     -- Island Computers US Corp.m 2700 Gregory Streeto Savannah GA 31404h Toll Free: 1-877 636 4332e International: 001 912 447 6622e  Facsimile:      001 912 201 0096 dbturner@hpaq.nets www.hpaq.net   ------------------------------  # Date: Thu, 05 Sep 2002 22:14:47 GMTb* From: "WarlockD" <warlockd@drakesmith.net>2 Subject: YEA!  I have a working Dec 3000 300 alpha) Message-ID: <rHQd9.9661$Jo.873@rwcrnsc53>s  J It went into the terminal unlike the the 400 I had.  Nice 150mhz processor too!  L Anyone know where the startup sound came from?  It sounds like a grandfather clock.   ------------------------------   End of INFO-VAX 2002.491 ************************de a transactionC >and know that the actions at the remote end would automatically be & >processed as part of the transaction. >s >View it as ACMS-extra-lite. >N >- TIP interoperabilityo >DI >Provide an API that gateways between DECdtm and TIP. Include the supportSI >to actually connect and interopate with a Microsoft MTS based front end.a >r >  >Java Interoperability:a > " >- ACMS to J2EE migration toolkit. >vE >This would process TDL, CDD                                                                                                                                                                                                                                                                                                     	    
            
                                                                                 !    "    #    $    %    &    '    (    )    *    +    ,    -    .    /    0    1    2    3    4    5    6    7    8    9    :    ;    <    =    >    ?    @    A    B    C    D    E    F    G    H    I    J    K    L    M    N    O    P    Q    R    S    T    U    V    W    X    Y    Z    [    \    ]    ^    _    `    a    b    c    d    e    f    g    h    i    j    k    l    m    n    o    p    q    r    s    t    u    v    w    x    y    z    {    |    }    ~                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                        