1 INFO-VAX	Thu, 22 Jun 2000	Volume 2000 : Issue 347       Contents: RE: Affordable Debate Summary 2 Re: Affordable Debate Summary (was: VAX on Intel?)2 Re: Affordable Debate Summary (was: VAX on Intel?)2 Re: Affordable Debate Summary (was: VAX on Intel?)2 Re: Affordable Debate Summary (was: VAX on Intel?) Re: Amazon?  Re: Amazon?  RE: Amazon?   Re: Besides Pathworks for Mac...# Cisco Has DECNet Ping - VMS Doesn't ' Re: Cisco Has DECNet Ping - VMS Doesn't ' RE: Cisco Has DECNet Ping - VMS Doesn't 0 Re: Clusterwide logical suggestion/documentation0 RE: Clusterwide logical suggestion/documentation0 RE: Clusterwide logical suggestion/documentation Re: Common Alpha Executable " Re: DECNET-Plus, DTSS, NTP and UTC" Re: DECNET-Plus, DTSS, NTP and UTC" Re: DECNET-Plus, DTSS, NTP and UTC" Re: DECNET-Plus, DTSS, NTP and UTC Equivalent of LAT ports in UCX" Re: Equivalent of LAT ports in UCX1 Re: Error In VMS SPD For AlphaServer 2100 5/375 ?  Finding a Hardware Address Re: Finding a Hardware Address Re: Fun VMS Facts? Re: Hotmail - Re: How do I calculate CONNECT TIME from DCL?  Internet & scripting$ Re: Mitnick (was Re: Fun VMS Facts?)$ Re: Mitnick (was Re: Fun VMS Facts?)$ Re: Mitnick (was Re: Fun VMS Facts?) Re: New to VMS. RE: OpenVMS clusters vs other systems clusters. RE: OpenVMS clusters vs other systems clusters. RE: OpenVMS clusters vs other systems clusters) Re: OpenVMS UCX/FTP and Internet Explorer ) Re: OpenVMS UCX/FTP and Internet Explorer ) Re: OpenVMS UCX/FTP and Internet Explorer 5 Re: Proxy problem: Why does node:: work but 0:: fail? ' Re: Questions about memory (DS20, ES40)  Re: QUEUE ENTRY NUMBERS... Re: QUEUE ENTRY NUMBERS... Re: Real Time VMS?? Re: Storage Works / Snapshots / Maybe it's time to skip OpenVMS  Re: User mode logicals re:  User mode logicals * Re: VAX & Alpha exist but where is OpenVMS re: VAX on Intel?  Re: VAX on Intel?  Re: VAX on Intel?  Re: VAX on Intel?  Re: VAX on Intel?  Re: VAX on Intel?  Re: Verify of Backups $ Re: VMS Evangelism:  Modest Proposal( VMS systems manager/programmer available Win2K on Alpha resumed?? Re: Win2K on Alpha resumed??  F ----------------------------------------------------------------------  % Date: Thu, 22 Jun 2000 13:31:11 -0400 . From: Hank Vander Waal <hvanderw@novagate.com>& Subject: RE: Affordable Debate Summary8 Message-ID: <001201bfdc6f$a913c930$939433cf@mscmain.com>  " I could not agree more with Rich!!  L I have been installing small systems for the last 16 years using MVII, 3100s and now small Alpha C boxes.  The largest customer I have has 80 users.  None of them are L interested in wether Baan or any other company has or has not supported VMS.J They want a solution that does not cost an arm and a leg and does not BSODI every week or month.  None of them have an MIS person on staff and really H like that fact that they don't need one since the "box in the back room" never stops!G I also have had customers that started with 3 users on a MVII that have K grown to 45 users on an "old" Alpha 2100 system and adding another 16 users K in the next 2 months.  Not big by the standards mentioned in this group but K there are LOTS of them out there and could be lots more if we could get the I INITIAL price down!  Good little companies do not stay little!!  And they J don't like to switch computer systems. So if we can get them on VMS at theG low end we will keep them for a long time and they will continue to pay 3 support agreements and buy new hardware & software!      Hank Vander Waal CIS 
 Hudsonvile MI     E The company I work for has been putting VMS systems into small-medium G businesses for 20 years.  Everything from small consumer loan companies H with 4 users to decent size manufacturing companies with multiple plantsF and up to a few hundred simultaneous users.  Very few clusters, mostlyD single MV3100 systems, largest systems sold were a VAX 4000-705a and? a dual AS1000a-4/500 SCSI cluster.  All running custom software @ developed in house.  Until last year most used VT terminals (nowB peecees running Telnet and an emulator, and I'm gleefully tracking1 relative failure/cost/replacement rates on that!)   C We can put in a DS10 single system cheaper now than practically any G MV3100 cost back then, save for the lack of ADL licenses for the users. E We had embedded reseller paks for the needed onsite development tools B (BASIC mostly) and now have ASAP/CSA for in-house, at least on the> Alpha systems (and we do maintain a _lot_ of VAX sites still).  C Three of our customers over time have either grown into or combined D into quite large companies.  Two of them are still VMS sites (thoughD much larger now than we support) and have done big business with DEQB on sales and support, the third moved to the merge-partners systemD (Novell-based, and has moved twice since to NT and now HPUX).  I forD one believe in the absolute necessity for the Q to be targetting the  small commercial market for VMS.  D Since we don't rely on packaged applications, the lack of SAP, BAAN,B Peoplesoft, etc is not an issue in our market.  However we are now@ competing with vendors who can drop a prepackaged NT box and 'x'E cheap client clones for less than the cost of the DS10 and VMS alone, F not counting our software, setup, etc.  That solution can work, though@ usually is far less customized to the customer's specific needs,> but the difference in price speaks quite loudly in their ears.> We're still busy, and still have new systems going in, but the? resistance is growing, and the _bulk_ of it is the price of the A system.  Places that were happy to spend $25K on an MV3100 system @ 10 years ago look in the pc mags and see complete NT servers for= under $2000 and the $7000-ish DS10 with VMS and user licenses > throws them completely.  Cluster?  Even the ones who could use6 it decide that at that price they can live without it.  B I believe that the Q really needs to produce a lower cost of entryA system.  A DS10 or equivalent, 128MB RAM, hard drive, tape drive, = VMS, and 'x' users (5 or 10) for under $5000 (and under $3500 < would be great, since thats the round cost of a good quality> name brand NT server with some apps in the papers here).  Some> time ago someone made a recommendation of a 'cluster in a box'= for under $10K {I think that was the price, but I may be mis- > remembering}.  That one would _absolutely_ make it possible to? put a cluster in some of those medium size places that actually A do need the capabilities a cluster provides, but are so ingrained ; with NT server pricing that they won't consider the current  product and pricing.  C If the Q needs to reduce the automatic support period to meet those : price targets, fine.  The up-front hardware/OS cost is the# biggest hurdle we seem to be facing < these days, and tradeoffs on warranty periods would not be aE problem.  After all there's always the support contract afterwards...      Rich Jordan  rjordan@mcs.net    ------------------------------  % Date: Thu, 22 Jun 2000 10:07:50 +0100 . From: Roger@natron.demon.co.uk (Roger Barnett); Subject: Re: Affordable Debate Summary (was: VAX on Intel?) - Message-ID: <880084008wnr@natron.demon.co.uk>   = Perhaps Compaq should consider providing VMS profiles at the   lower end ?    For example:; a. development VMS, supporting lower cost licences for the  D     various compilers, development versions of layered products, etcA     but: limited in terms of numbers of users, will not run lower (     cost production scale databases, etc  H [ could use this to push the ASAP or whatever it is called now program ]  > b. application VMS, supporting lower cost licences for various     non-development productsB     but: limited in terms of numbers of users, will not run lower      cost compilers etc  ( c./d.  as above but for a 2-node cluster    7 As far as I can see this would only need changes to LMF ( ( and to the sales model...       uhoh )    
 Roger Barnett    ------------------------------  % Date: Thu, 22 Jun 2000 03:40:01 -0700 * From: "Nikita V. Belenki" <kit@nospam.net>; Subject: Re: Affordable Debate Summary (was: VAX on Intel?) 4 Message-ID: <%7m45.1$0x.557@nuq-read.news.verio.net>  B "David J. Dachtera" <djesys.nospam@earthlink.net> wrote in message' news:39517C3B.BE5246A9@earthlink.net...   D > > Not at all.  What I've been trying to drum into you is that it's
 reasonableG > > at this point in time for Compaq to devote minimal resources to its I > > *smallest and least lucrative* VMS customers:  they've been abandoned  for a J > > decade or more anyway, and right now Compaq has more important fish to fry L > > (and several other systems better positioned to sell in the very low end > > anyway).F > That is, you're trying to drum Compaq's position into us, instead ofH > being sensitive to the needs of the customer (the cornerstone of every/ > successful, profitable, long-lived business).   K Which is exactly what Compaq does. The customer wants Windows, the customer C gets Windows. So if Compaq wants to support its "smallest and least - lucrative" customers, it knows where to look.   J > > 2.  Compaq can't exploit it, because there's little evidence that (for VMS)7 > > any worthwhile demand in that market segment exists I > high price = low (or no) demand (Business 101, 1st semester, Chapter 2:  > The Law of Supply and Demand)    Or: < low demand = high price (which is exactly the same equation)  H There is virtually no demand for VMS in the low-end, high-volume market,A regardless of its price. Because there is no low-end, high-volume J *solutions* using VMS, and also because there is no "low-end, high-volume"L people familiar with VMS. And even if you drop the price of VMS to zero, youK will change nothing. Guess why? Because zero-price non-proprietary Linux is L already there, and the "winning" features of VMS are just irrelevant on this market.   C Or could you present just one VMS-based *complete solution* that is K "high-volume-ready", can sustain the competition against Linux and will not 2 be a threat for the current VMS-related cash flow?   Kit. kit # kits.net   ------------------------------  # Date: Thu, 22 Jun 2000 14:21:57 GMT 3 From: "Gord Coulman" <nospam_gcoulman@ccinet.ab.ca> ; Subject: Re: Affordable Debate Summary (was: VAX on Intel?) ; Message-ID: <9ip45.12660$vT6.1607773@news1.telusplanet.net>   K What IT people are short of these days is time.  Wouldn't it be great to be H able to buy, for example, a pre-configured database server that will runD reliably for months or years with little setup or intervention?  TheL advantage that NT has at the low end stems mainly from perceived ease of useD and software that can be installed with one "double click" (in salesK literature, anyway).  Perhaps OpenVMS could make some inroads in this area, 9 given competitive pricing and a well-integrated solution.    Gord Coulman  3 Nikita V. Belenki <kit@nospam.net> wrote in message . news:%7m45.1$0x.557@nuq-read.news.verio.net...D > "David J. Dachtera" <djesys.nospam@earthlink.net> wrote in message) > news:39517C3B.BE5246A9@earthlink.net...  > F > > > Not at all.  What I've been trying to drum into you is that it's > reasonableI > > > at this point in time for Compaq to devote minimal resources to its K > > > *smallest and least lucrative* VMS customers:  they've been abandoned  > for a L > > > decade or more anyway, and right now Compaq has more important fish to > fry J > > > (and several other systems better positioned to sell in the very low end  > > > anyway).H > > That is, you're trying to drum Compaq's position into us, instead ofJ > > being sensitive to the needs of the customer (the cornerstone of every1 > > successful, profitable, long-lived business).  > D > Which is exactly what Compaq does. The customer wants Windows, the customerE > gets Windows. So if Compaq wants to support its "smallest and least / > lucrative" customers, it knows where to look.  > L > > > 2.  Compaq can't exploit it, because there's little evidence that (for > VMS)9 > > > any worthwhile demand in that market segment exists K > > high price = low (or no) demand (Business 101, 1st semester, Chapter 2: ! > > The Law of Supply and Demand)  >  > Or: > > low demand = high price (which is exactly the same equation) > J > There is virtually no demand for VMS in the low-end, high-volume market,C > regardless of its price. Because there is no low-end, high-volume L > *solutions* using VMS, and also because there is no "low-end, high-volume"J > people familiar with VMS. And even if you drop the price of VMS to zero, you J > will change nothing. Guess why? Because zero-price non-proprietary Linux isI > already there, and the "winning" features of VMS are just irrelevant on  this	 > market.  > E > Or could you present just one VMS-based *complete solution* that is I > "high-volume-ready", can sustain the competition against Linux and will  not 4 > be a threat for the current VMS-related cash flow? >  > Kit. > kit # kits.net >  >  >    ------------------------------    Date: 22 Jun 2000 10:21:59 -0500+ From: rjordan@Mars.mcs.net (Richard Jordan) ; Subject: Re: Affordable Debate Summary (was: VAX on Intel?) ' Message-ID: <8itaun$3em$1@Mars.mcs.net>   E The company I work for has been putting VMS systems into small-medium G businesses for 20 years.  Everything from small consumer loan companies H with 4 users to decent size manufacturing companies with multiple plantsF and up to a few hundred simultaneous users.  Very few clusters, mostlyD single MV3100 systems, largest systems sold were a VAX 4000-705a and? a dual AS1000a-4/500 SCSI cluster.  All running custom software @ developed in house.  Until last year most used VT terminals (nowB peecees running Telnet and an emulator, and I'm gleefully tracking1 relative failure/cost/replacement rates on that!)   C We can put in a DS10 single system cheaper now than practically any G MV3100 cost back then, save for the lack of ADL licenses for the users. E We had embedded reseller paks for the needed onsite development tools B (BASIC mostly) and now have ASAP/CSA for in-house, at least on the> Alpha systems (and we do maintain a _lot_ of VAX sites still).  C Three of our customers over time have either grown into or combined D into quite large companies.  Two of them are still VMS sites (thoughD much larger now than we support) and have done big business with DEQB on sales and support, the third moved to the merge-partners systemD (Novell-based, and has moved twice since to NT and now HPUX).  I forE one believe in the absolute necessity for the Q to be targetting the    small commercial market for VMS.  D Since we don't rely on packaged applications, the lack of SAP, BAAN,B Peoplesoft, etc is not an issue in our market.  However we are now@ competing with vendors who can drop a prepackaged NT box and 'x'E cheap client clones for less than the cost of the DS10 and VMS alone, F not counting our software, setup, etc.  That solution can work, though@ usually is far less customized to the customer's specific needs,@ but the difference in price speaks quite loudly in their ears.  > We're still busy, and still have new systems going in, but the? resistance is growing, and the _bulk_ of it is the price of the A system.  Places that were happy to spend $25K on an MV3100 system A 10 years ago look in the pc mags and see complete NT servers for  = under $2000 and the $7000-ish DS10 with VMS and user licenses > throws them completely.  Cluster?  Even the ones who could use6 it decide that at that price they can live without it.  B I believe that the Q really needs to produce a lower cost of entryA system.  A DS10 or equivalent, 128MB RAM, hard drive, tape drive, = VMS, and 'x' users (5 or 10) for under $5000 (and under $3500 < would be great, since thats the round cost of a good quality> name brand NT server with some apps in the papers here).  Some> time ago someone made a recommendation of a 'cluster in a box'= for under $10K {I think that was the price, but I may be mis- > remembering}.  That one would _absolutely_ make it possible to? put a cluster in some of those medium size places that actually A do need the capabilities a cluster provides, but are so ingrained < with NT server pricing that they won't consider the current  product and pricing.    C If the Q needs to reduce the automatic support period to meet thosed; price targets, fine.  The up-front hardware/OS cost is the  # biggest hurdle we seem to be facingm= these days, and tradeoffs on warranty periods would not be a AE problem.  After all there's always the support contract afterwards...m     Rich Jordann rjordan@mcs.net       ------------------------------    Date: 22 Jun 2000 07:54:43 -05009 From: Kilgallen@eisner.decus.org.nospam (Larry Kilgallen)t Subject: Re: Amazon?+ Message-ID: <ZT$B3dyCbNiK@eisner.decus.org>i  \ In article <39513257.8B3F507A@videotron.ca>, JF Mezei <jfmezei.spamnot@videotron.ca> writes:* > Andrew Harrison SUNUK Consultancy wrote:H >> Amazon kicked the Alphaservers that were hosting most of Amazons backL >> end systems out in favour of HP V series machines last year in a quid pro9 >> quo deal with HP (they got to sell HP kit via Amazon).r > N > The Compaq news relese now only talked about web servers. So Compaq was ableN > to put a positive spin even if it lost out big with Amazon. My guess is thatO > once Amazon called Compaq's bluff and signed with HP with all the fanfare and)N > notes that Amazon was dropping Compaq, Compaq may have finally sharpened itsM > pencils and found a way to get Amazon to keep some Compaq gear in its shop.e  ; Any shop that is highly dependent on computers and has lotst> would be extremely foolish to switch brands entirely away fromB something upon which they have depended.  Remember the differences between various brands of Unix.W   ------------------------------  % Date: Thu, 22 Jun 2000 14:38:43 +0100 B From: Andrew Harrison SUNUK Consultancy <andrew.nospam@uk.sun.com> Subject: Re: Amazon?* Message-ID: <395216E3.D6929A15@uk.sun.com>   "Main, Kerry" wrote:  	 > Andrew,  >eN > Like most fud, you are making statements that are far from the truth. Please > re-read the release. >AL > Do you honestly believe that a company would post a news release like thisB > (with a quote from the Customer IT manager) if it were not true?N > <http://www1.compaq.com/pressrelease/0,1494,wp%7E14583_2%21ob%7E31415_1_1,00 > .html> >-  I No Kerry I am sure it is true, its just a really pointless press release.h EveryoneH knows that Amazon is chucking Alpha boxes out in favour of HP's and haveH allready chucked a whole load of systems out when the origional V-series went in.  G Why both send out a press release, the writing is on the wall at Amazona and this simply looks bad.   >u2 > Geeezzz... Andrew, your fud has had better days. >gH > Yep, Compaq made a bid for some NEW servers and it lost. That happens. >a  * Nope thats not what happened. Sorry Kerry.  K Amazon has for some time wanted to get into supplying electronics equipmentS, computers and their components over the WEB.  O Compaq didn't lose a bid for new Servers because the bid for Servers was linked   O to Amazons desire to sell in this case HP computers, printers etc through their   N WEB channel. Both of the HP/Amazon deals have been quite publically a tradeoffM of Amazon using HP servers for the bulk of their infrastructure in return fore HP& kit that Amazon can sell over the WEB.  J And nor is this FUD, the HP V-Series bought last year did replace existingF AlphaServers running Tru64, the new deal will see HP systems replacing@ AlphaServer and Sun's in the mid tier of Amazons infrastructure.  O You should not take this as a criticism of Tru64 or AlphaServers its not, sinces  E the HP servers were chosen over Compaq and Sun servers because Amazone' want to distribute HP desktops etc etc.   M If you work for Compaqs PC division then you can take it as a criticism since K it is clear that Amazon either did not contemplate selling Compaq equipmentpG via their WEB sites or they considered that HP was a better brand of PCVI which justified the risks of moving their backend infrastructure from onea platform to another.  E And of course Amazon will at least in the short term need to buy more  AlphaserversO because their requirements are continually growing and the HP switchover is not    going to happen over night.a     >iN > However, given that Amazon has a huge amount of Alpha's in place (with plansI > to buy more as indicated in the above), I am sure you realize that when K > deals for new servers are made, no company makes immediate plans to throwsM > the old ones out the window. The new ones will be slowly phased in over theeK > lifetime of the agreement and the onus is now on the new vendor to ensure F > their stuff performs much better than the Customers current servers. >q  L Quite but given that Amazon have said that they will be replacing 90% of theM AlphaServers with HP's this large installed base will be getting smaller once " the switchover gets in full swing.  I But don't fool yourself that the GS boxes performing faster will make anyn difference,tI if performance had been a real buying criteria then Amazon would not havee bought" the HP V-Series in the first case.   Regardso Andrew HarrisonA Enterprise IT Architect    ------------------------------  % Date: Thu, 22 Jun 2000 11:07:45 -0400w# From: John Vottero <John@MVPSI.com>' Subject: RE: Amazon?D Message-ID: <C15945A9D9EFCF11BA8B08002BBF1CCC0CD72A@berry.mvpsi.com>  K Don't be so naive Andrew!  Isn't it obvious that HP made a deal?  "Buy sometD of our V-Series and let us do a big press release and we'll give you preferred pricing."n  L The Compaq press release says "Amazon currently runs all of its Web servers,C and the majority of its installed base, on Compaq's midrange servermL platform".  They run some small application on a V-Series so they can sell aJ million printers.  Certainly, if they were dumping the Alpha platform they> wouldn't buy millions of dollars worth of them now would they?   > -----Original Message-----) > From: Andrew Harrison SUNUK Consultancy # > [mailto:andrew.nospam@uk.sun.com]e' > Sent: Thursday, June 22, 2000 9:39 AMf > To: Info-VAX@Mvb.Saic.Comu > Subject: Re: Amazon? >  >  > "Main, Kerry" wrote: >  > > Andrew,) > >u? > > Like most fud, you are making statements that are far from n > the truth. Pleasei > > re-read the release. > >o= > > Do you honestly believe that a company would post a news o > release like thistD > > (with a quote from the Customer IT manager) if it were not true? > > @ > <http://www1.compaq.com/pressrelease/0,1494,wp%7E14583_2%21ob% > 7E31415_1_1,00
 > > .html> > >u > = > No Kerry I am sure it is true, its just a really pointless o > press release.
 > Everyone= > knows that Amazon is chucking Alpha boxes out in favour of r > HP's and have.8 > allready chucked a whole load of systems out when the  > origional V-series
 > went in. > @ > Why both send out a press release, the writing is on the wall  > at Amazons > and this simply looks bad. >  > >d4 > > Geeezzz... Andrew, your fud has had better days. > >p= > > Yep, Compaq made a bid for some NEW servers and it lost.   > That happens.- > >- > , > Nope thats not what happened. Sorry Kerry. > 8 > Amazon has for some time wanted to get into supplying  > electronics equipmenta. > computers and their components over the WEB. > ? > Compaq didn't lose a bid for new Servers because the bid for V > Servers was linked > @ > to Amazons desire to sell in this case HP computers, printers  > etc through their  > ; > WEB channel. Both of the HP/Amazon deals have been quite e > publically a tradeoffe3 > of Amazon using HP servers for the bulk of their t > infrastructure in return for > HP( > kit that Amazon can sell over the WEB. > < > And nor is this FUD, the HP V-Series bought last year did  > replace existingH > AlphaServers running Tru64, the new deal will see HP systems replacingB > AlphaServer and Sun's in the mid tier of Amazons infrastructure. > 6 > You should not take this as a criticism of Tru64 or  > AlphaServers its not, sinceM > G > the HP servers were chosen over Compaq and Sun servers because Amazon-) > want to distribute HP desktops etc etc.0 > @ > If you work for Compaqs PC division then you can take it as a  > criticism sincee= > it is clear that Amazon either did not contemplate selling @ > Compaq equipment> > via their WEB sites or they considered that HP was a better 
 > brand of PCs4 > which justified the risks of moving their backend  > infrastructure from one  > platform to another. > G > And of course Amazon will at least in the short term need to buy morec > Alphaservers@ > because their requirements are continually growing and the HP  > switchover is nota >  > going to happen over night.m >  >  > >i? > > However, given that Amazon has a huge amount of Alpha's in i > place (with plans : > > to buy more as indicated in the above), I am sure you  > realize that whent? > > deals for new servers are made, no company makes immediate ( > plans to throw= > > the old ones out the window. The new ones will be slowly c > phased in over the= > > lifetime of the agreement and the onus is now on the new s > vendor to ensureH > > their stuff performs much better than the Customers current servers. > >o > : > Quite but given that Amazon have said that they will be  > replacing 90% of the; > AlphaServers with HP's this large installed base will be   > getting smaller once$ > the switchover gets in full swing. > > > But don't fool yourself that the GS boxes performing faster  > will make anyi
 > difference,a= > if performance had been a real buying criteria then Amazon - > would not have > bought$ > the HP V-Series in the first case. > 	 > Regardsi > Andrew Harrisono > Enterprise IT Architecta >  >  >  >    ------------------------------    Date: 22 Jun 2000 08:08:35 -05009 From: Kilgallen@eisner.decus.org.nospam (Larry Kilgallen)w) Subject: Re: Besides Pathworks for Mac...x+ Message-ID: <ngs8Kxx9b5gi@eisner.decus.org>e  S In article <39517300.7D0721CF@mediaone.net>, Ed Wilts <ewilts@mediaone.net> writes:e > Christoph Gartmann wrote:  >> gN >> I remember a problem with certain types of Alpha processors though. Perhaps- >> someone can comment a bit more about that?  > J > The problem was that the EV6 check utility (whose name I can't remember)C > failed on the executables.  It also fails on a whole bunch of VMS  > executables.  G There are known bad coding practices that can trigger "false positives"1G with that tool, having to do with not separating code from data psects.i   ------------------------------  % Date: Thu, 22 Jun 2000 10:46:33 +0100g/ From: Nigel Arnot <sysmgr@maxwell.ph.kcl.ac.uk>m, Subject: Cisco Has DECNet Ping - VMS Doesn't6 Message-ID: <009EBFBF.988B47F3.9@maxwell.ph.kcl.ac.uk>   > ? > Cisco routers capable of routing DECNet and LAT have a DECNet = > ping that allows the user to specify the size and number ofn > packets to be transmitted. > A > Is it possible to add this functionality to VMS, to be availble A > to both DECnet Phase IV and DECNet-PLUS systems, perhaps as an s > enhancement of DTSEND ?u > 6 > --Jerry Leslie     (my opinions are strictly my own) >   $ All together now: "OH YES IT DOES!!"  + NCP LOOP NODE node-id COUNT n LENGTH nn    o   or at lower levels,e   NCP LOOP CIRCUIT circuit-ida NCP LOOP LINE line-id   $ NCP HELP LOOP  for lots more info.       	Yours,l
 		Nigel Arnott- 		NRA@MAXWELL.PH.KCL.AC.UK                   s  7 		"In the beginning there was nothing, which exploded."    ------------------------------  # Date: Thu, 22 Jun 2000 15:06:29 GMT / From: "Richard L. Dyson" <rick-dyson@uiowa.edu> 0 Subject: Re: Cisco Has DECNet Ping - VMS Doesn't) Message-ID: <3951E525.147DE396@uiowa.edu>i   Jerry Leslie wrote: J > NCP's loop node doesn't seem to provide any quantitative results, unless4 > I'm doing something wrong, a distinct possibility: > ; >   <SCCN02> mcr ncp loop node sccvx6 count 100 length 3000, >   <SCCN02>  = 	Can anyone suggest what the DECnet V equivalent would be?  I F some of them beasts and since they mostly have worked without problem,G I have not needed to learn NCL. :)  NCP is sort-of still around, but it)> only does a few things anymore and that is not one of them. :(   $ MCR NCP Loop Node BlahB %NCP-W-SYSMGT, System-specific management function not supported     Thanks!d rick -- rH Richard L. Dyson                                    rick-dyson@uiowa.eduH  _   _      _____                http://www-pi.physics.uiowa.edu/~dyson/H | | | |    |_   _|   Systems Analyst                     O: 319/335-1879H | | | | of   | |     The University of Iowa            FAX: 319/335-17536 | \_/ |     _| |_    Department of Physics & Astronomy-  \___/     |_____|   Iowa City, IA 52242-1479.   ------------------------------  % Date: Thu, 22 Jun 2000 13:12:55 -0400a0 From: arturo saavedra <arturo.saavedra@wcom.com>0 Subject: RE: Cisco Has DECNet Ping - VMS Doesn't4 Message-ID: <000301bfdc6d$1b3ea6e0$14b324a6@CJ4733A>   you can try something like n  > $ mcr ncl loop loopback application name local:.node, count 10   that should do it.    A Remember that the mirror application ( object under phase iv ) is- enabled as well.  a     -----Original Message-----H From: News Administrator [mailto:news@sysadm.physics.uiowa.edu]On Behalf Of Richard L. Dysono& Sent: Thursday, June 22, 2000 11:06 AM To: Info-VAX@Mvb.Saic.Com 0 Subject: Re: Cisco Has DECNet Ping - VMS Doesn't     Jerry Leslie wrote:2J > NCP's loop node doesn't seem to provide any quantitative results, unless4 > I'm doing something wrong, a distinct possibility: > ; >   <SCCN02> mcr ncp loop node sccvx6 count 100 length 3000  >   <SCCN02>  = 	Can anyone suggest what the DECnet V equivalent would be?  I F some of them beasts and since they mostly have worked without problem,G I have not needed to learn NCL. :)  NCP is sort-of still around, but it > only does a few things anymore and that is not one of them. :(   $ MCR NCP Loop Node BlahB %NCP-W-SYSMGT, System-specific management function not supported     Thanks!n rick -- cH Richard L. Dyson                                    rick-dyson@uiowa.eduH  _   _      _____                http://www-pi.physics.uiowa.edu/~dyson/H | | | |    |_   _|   Systems Analyst                     O: 319/335-1879H | | | | of   | |     The University of Iowa            FAX: 319/335-17536 | \_/ |     _| |_    Department of Physics & Astronomy-  \___/     |_____|   Iowa City, IA 52242-1479t   ------------------------------    Date: 22 Jun 2000 08:18:41 -05009 From: Kilgallen@eisner.decus.org.nospam (Larry Kilgallen)r9 Subject: Re: Clusterwide logical suggestion/documentatione+ Message-ID: <OuY1wfHkIBeW@eisner.decus.org>t  \ In article <3951A5DB.EF197AE7@videotron.ca>, JF Mezei <jfmezei.spamnot@videotron.ca> writes:F > The VMS 7.2 on-line help on creating logical names (HELP SYS $CRELNMM > arguments) stipulates that logical names are restricted to 32 characters innM > the process and system directory tables are limited to 31 characters, whilea> > others are limited to 255. No mention of clusterwide tables. > L > The VMS Cluster manual makes no mention of the length of the logical name.O > http://WWW.OPENVMS.DIGITAL.COM:8000/72final/4477/4477pro_007.html#index_x_357n > B > Also, that page provides confusing examples of clustwide tables:  D Your post seems to be a commentary on specific documentation errors.  @ Your decision to tell us all about it would seem more reasonable? if you were to _start_ with an explanation that you had in factrB mailed the documentation comments to the designated Internet email@ address included in the front of every VMS document.  _Then_ youA could explain your technical points (leaving out the specifics ofh' which document erred in which fashion).c   ------------------------------  % Date: Thu, 22 Jun 2000 09:21:38 -0300i1 From: "Boyle, Darren" <boyledj@bankofbermuda.com>(9 Subject: RE: Clusterwide logical suggestion/documentationtK Message-ID: <9F664D538536D411BD3200508B6FF01A0247F8@bdant027.bda.bobda.com>   I Surely the best approach would be DEFINE/CLUSTER....No ?.  We have group,e! job, process etc, why not cluster  - Darren   > ----------4 > From: 	JF Mezei[SMTP:jfmezei.spamnot@videotron.ca]( > Sent: 	Thursday, June 22, 2000 2:36 AM > To: 	Info-VAX@Mvb.Saic.Com8 > Subject: 	Clusterwide logical suggestion/documentation > F > The VMS 7.2 on-line help on creating logical names (HELP SYS $CRELNMJ > arguments) stipulates that logical names are restricted to 32 characters > inG > the process and system directory tables are limited to 31 characters,g > whilee> > others are limited to 255. No mention of clusterwide tables. > L > The VMS Cluster manual makes no mention of the length of the logical name.L > http://WWW.OPENVMS.DIGITAL.COM:8000/72final/4477/4477pro_007.html#index_x_ > 357- > B > Also, that page provides confusing examples of clustwide tables: > 	 > <quote>lI > The following example shows how to create a clusterwide logical name ini > theaF > default clusterwide logical name table, LNM$CLUSTER_TABLE, using the > DEFINE > command: r > C >  $ DEFINE/TABLE=LNM$CLUSTER_TABLE logical-name equivalence-stringh
 > </quote> >  > further along:	 > <quote>sC > For clusterwide definitions in a common SYSTARTUP_VMS.COM, Compaq  > recommends@ > that you use a conditional definition, such as the following:  > 1 >  $ IF F$TRNLNM("CLUSTER_APPS") .EQS. "" THEN - a6 >  _$ DEFINE/TABLE=LNM$SYSCLUSTER/EXEC CLUSTER_APPS -  >  _$ 1$DKA500:[COMMON_APPS]  
 > </quote> >  > L > The second example, which uses LNM$SYSCLUSTER works in the expected way byF > making logical names "visible" on all nodes with a SHOW LOG command,	 > whereas @ > in the first examble (LNM$CLUSTER_TABLE) requires that you use5 > /TABLE=LNM$CLUSTER_TABLE with the SHOW LOG command.t >  > J > Also, by using two different logical name tables in the examples in that > pageC > without expolaining the differences between each can lead to usern > confusion. >     F **********************************************************************C This message and any files transmitted with it are confidential andAJ may be privileged and/or subject to the provisions of privacy legislation.M They are intended solely for the use of the individual or entity to whom theytL are addressed. If the reader of this message is not the intended recipient, B please notify the sender immediately and then delete this message.I You are notified that reliance on, disclosure of, distribution or copyings of this message is prohibited.   Bank of Bermuda F **********************************************************************   ------------------------------  % Date: Thu, 22 Jun 2000 11:21:36 -0400n# From: John Vottero <John@MVPSI.com>u9 Subject: RE: Clusterwide logical suggestion/documentationsD Message-ID: <C15945A9D9EFCF11BA8B08002BBF1CCC0CD72B@berry.mvpsi.com>  J I think that /JOB, /SYSTEM etc. are there for backwards compatibility.  If1 they had it to do over, it would just be /TABLE=.b   > -----Original Message-----8 > From: Boyle, Darren [mailto:boyledj@bankofbermuda.com]' > Sent: Thursday, June 22, 2000 8:22 AMn > To: Info-VAX@Mvb.Saic.Comc; > Subject: RE: Clusterwide logical suggestion/documentationE >  > @ > Surely the best approach would be DEFINE/CLUSTER....No ?.  We 
 > have group, # > job, process etc, why not clustere
 > - Darren >  > > ----------6 > > From: 	JF Mezei[SMTP:jfmezei.spamnot@videotron.ca]* > > Sent: 	Thursday, June 22, 2000 2:36 AM > > To: 	Info-VAX@Mvb.Saic.Com: > > Subject: 	Clusterwide logical suggestion/documentation > > H > > The VMS 7.2 on-line help on creating logical names (HELP SYS $CRELNM? > > arguments) stipulates that logical names are restricted to w > 32 characterss > > in> > > the process and system directory tables are limited to 31 
 > characters,u	 > > while @ > > others are limited to 255. No mention of clusterwide tables. > > = > > The VMS Cluster manual makes no mention of the length of   > the logical name.  > > @ > http://WWW.OPENVMS.DIGITAL.COM:8000/72final/4477/4477pro_007.h > tml#index_x_ > > 357c > > D > > Also, that page provides confusing examples of clustwide tables: > >  > > <quote> < > > The following example shows how to create a clusterwide  > logical name ini > > thevH > > default clusterwide logical name table, LNM$CLUSTER_TABLE, using the
 > > DEFINE
 > > command: t > > E > >  $ DEFINE/TABLE=LNM$CLUSTER_TABLE logical-name equivalence-strings > > </quote> > >  > > further along: > > <quote> E > > For clusterwide definitions in a common SYSTARTUP_VMS.COM, Compaqh > > recommendsB > > that you use a conditional definition, such as the following:  > > 3 > >  $ IF F$TRNLNM("CLUSTER_APPS") .EQS. "" THEN - e8 > >  _$ DEFINE/TABLE=LNM$SYSCLUSTER/EXEC CLUSTER_APPS -  > >  _$ 1$DKA500:[COMMON_APPS] T > > </quote> > >  > > ? > > The second example, which uses LNM$SYSCLUSTER works in the   > expected way byoH > > making logical names "visible" on all nodes with a SHOW LOG command, > > whereasqB > > in the first examble (LNM$CLUSTER_TABLE) requires that you use7 > > /TABLE=LNM$CLUSTER_TABLE with the SHOW LOG command.  > >  > > < > > Also, by using two different logical name tables in the  > examples in that > > pageE > > without expolaining the differences between each can lead to usern > > confusion. > >  >  > H > **********************************************************************E > This message and any files transmitted with it are confidential and-@ > may be privileged and/or subject to the provisions of privacy  > legislation.< > They are intended solely for the use of the individual or  > entity to whom theyC: > are addressed. If the reader of this message is not the  > intended recipient, D > please notify the sender immediately and then delete this message.4 > You are notified that reliance on, disclosure of,  > distribution or copyingo  > of this message is prohibited. >  > Bank of BermudayH > ********************************************************************** >  >    ------------------------------    Date: 22 Jun 2000 08:12:50 -05009 From: Kilgallen@eisner.decus.org.nospam (Larry Kilgallen)l$ Subject: Re: Common Alpha Executable+ Message-ID: <MOrNXwdx3Ixw@eisner.decus.org>g  a In article <8is42b$bra@gap.cco.caltech.edu>, mathog@seqaxp.bio.caltech.edu (David Mathog) writes:r  J > This would be one heck of a lot better for sales than Affinity ever was!  ? Mainly because the Linux (and even Unix) proper behavior is noti8 subject to change at the whim of some other corporation.  F I would suggest you organize a session for the Fall US DECUS SymposiumF in Los Angeles (your neighborhood) and also a lot of Advance Publicity to draw participation.   ------------------------------  % Date: Wed, 21 Jun 2000 22:46:43 -0700r& From: Ryan Moore <rmoore@qualcomm.com>+ Subject: Re: DECNET-Plus, DTSS, NTP and UTCrI Message-ID: <Pine.GSO.4.10.10006212236200.28995-100000@zinc.qualcomm.com>   G On 21 Jun 2000, Ken Fairfield; SLAC: 650-926-2924; FAX: 926-3515 wrote:  >     QUESTIONS  >     ---------aF >       o Would I, for instance,  be  better  off  to make this node aF >         DTSS  Server,  and if so, how do I get it to use  its  local+ >         system clock as corrected by NTP?   C Given that you have no other DECnet-Plus nodes on your network, I'deG recommend you forget about DTSS all together and just use NTP.  I'm not-J sure what would happen if you configured your machine as a DTSS server andG ran NTP at the same time.  Maybe DTSS and NTP would "fight" over how too! set the clock.  Not a good thing.h  6 To make sure you're not starting DTSS, just rename theH NET$DTSS_SERVER_STARTUP.NCL file to something else.  And add the commandH "exit" to the first line of the NET$DTSS_CLERK_STARTUP.NCL file.  Then IG don't think DTSS will start at all.  Then you configure NTP in Multinetf and you're happy.n  E Or, if you want to run a DTSS server, you could try to acquire a DTSSPF "time provider" that gets the time via NTP.  I don't have a pointer toJ such software, but someone else may know where to get something like that.? But if you do this, you don't want to configure NTP on the box.r  E A big difference between DTSS and NTP come when daylight savings timedE changes.  I believe NTP changes the clock in one step.  But DTSS willcI "slew" the time until the server gets back in sync.  Although there isn'taE a time "jump", it means that timestamps around the time of the change % don't reflect any reality whatsoever.t   -Ryan    ------------------------------  % Date: Thu, 22 Jun 2000 09:02:56 -0500i1 From: "Dave Gudewicz" <david.gudewicz@abbott.com>t+ Subject: Re: DECNET-Plus, DTSS, NTP and UTC 8 Message-ID: <8it63q$87d$1@fizban.fizban.pprd.abbott.com>  H I could be wrong but I thought Multinet did the twice a year time changeL automatically.  UCX and TCPIP services still need to be coaxed into changingJ the time.  If you have access to DSN there are several articles related to this issue.   G From my ever failing memory, you need to pay attention to the followingh	 logicals:m  !     sys$timezone_daylight_savingsl     sys$timezone_differentialt     sys$timezone_names     sys$timezone_rulet  H As for DTSS/NTP, we went through the same confusion (and still do) about2 what to do.  We now use NTP and it seems to be OK.  I Rumor has it that VMS 7.3 will "finally" fix the time issue, but that's an ways off.  Next year I'm told.   Dave...r  0 Ken Fairfield; SLAC: 650-926-2924; FAX: 926-3515. <Fairfield@SLAC.Stanford.EDU> wrote in message- news:M0Ze4F51Pb3r@mccdev.slac.stanford.edu...iJ >         I need to bring  up  a  bunch  of  old issues here: DECNET-Plus,J >     DTSS,  NTP,  but  also  UTC and the  various  SYS$TIMEZONE*  logical >     names... >  >     BACKGROUND >     ----------J >         I have a DS10 running  VMS  7.2-1  and DECNET-Plus with MultinetJ >     4.2A  installed.  Consider this a stand-alone system for the  momentJ >     even though I'll be adding it to my cluster soon.  The rest  of  theJ >     cluster  still  runs  DECNET  IV.   My  main  motivation for runningJ >     DECNET-Plus on this node is  that  this gives me DECNET-over-IP to aJ >     few  other  clusters/systems  at our site,  whereas  the  networkingJ >     people pulled the plug on DECNET routing in the site's routers  lastJ >     fall.   I.e.,  from  my cluster nodes (DECNET IV), I can't get thereJ >     from here (where "there" is  any  node  outside our local subnet via >     DECNET). >eJ >         Note that we use NTP  pretty  much universally on-site and there. >     are no DTSS servers configured anywhere. >cJ >         I spent probably two hours  at  deja.com this morning looking upJ >     old  discussions relating to DTSS (and the Too Few Servers  DetectedJ >     messages), how to block the messages, how to disable the DTSS Clerk,J >     how to set up your node as a DTSS Server, etc.  I'm still left  withJ >     basic  questions about how _best_ to configure DTSS _and_ use NTP to# >     synchronize the system clock.3 >m >     ISSUES >     ------J >         There are various and sundry  interactions between DTSS, NTP andJ >     UTC,  and the various system *TIME* logical names.  In particular, IJ >     found  that  if  I  did  a   DISABLE   DTSS   &   DELETE   DTSS   inJ >     NET$DTSS_CLERK_STARTUP.NCL,   and   followed   that   by   executingJ >     SYS$MANAGER:UTC$TIME_SETUP.COM,   that   on   reboot,   the  variousJ >     SYS$TIMEZONE*  logical names were wrong (that is, set to PST  valuesJ >     rather than PDT values).  I got similar results  if  I  renamed  theJ >     DTSS NCL file "out of the way", but in that case, I found that thereJ >     was  a  DTSS$SERVER  process  running  where before I had (or didn't) >     have) a DTSS$CLERK process running.p >eJ >         Finally, I put the  _default_ NET$DTSS_CLERK_STARTUP.NCL back inJ >     place,  execute  UTC$TIME_SETUP and rebooted, and _now_ the  logicalJ >     names are set correctly _and_ NTP adjusts the time correctly!  I  doJ >     get  the  Too Few Servers Detected OPCOM messages from DTSS, which IJ >     duly suppressed with  MC  NCL  BLOCK  EVENT DISPATCHER ... commands.G >     Nevertheless, I wonder if this is the correct, or best, solution?0 >  >     QUESTIONSg >     ---------eF >       o Would I, for instance,  be  better  off  to make this node aF >         DTSS  Server,  and if so, how do I get it to use  its  local+ >         system clock as corrected by NTP?e >hF >       o Given that there are no  DTSS  Servers  on the network, if IF >         continue to run as a DTSS Clerk, will DTSS change the systemF >         time  in  spite  of NTP?  [I  currently  suppress  automaticF >         spring and fall time changes  due  to  requirements  of  ourF >         production  node...I'd  be very distressed if DTSS "thought"F >         it was smarter than me, or NTP,  and did these changes on myE >         behalf.] What is required to set up a DTSS Server this way?s >AF >       o Could I simply wait  until  late  in the boot process, afterF >         DTSS  has  set up the SYSTEM logical names, to DISABLE  DTSSF >         and DELETE DTSS, or  would  I  still  have  problems  duringF >         spring  and  fall  time  changes?  Note that I currently runF >         SYS$MANAGER:UTC$CONFIGURE_TDF.COM  on   all   cluster  nodesF >         (DECNET IV) when I do the spring and fall time changes to beF >         sure  that SYS$TIMEZONE_DIFFERENTIAL and that "special  cellF >         in memory", EXE$GQ_TDF, get changed correctly  (as  well  as9 >         doing the appropriate stuff in MULTINET, etc.).  >r >y+ >         Thanks for any and all responses.i >o >             -Ken > --/ >  Kenneth H. Fairfield            |  Internet:e Fairfield@SLC.Slac.Stanford.Edum< >  SLAC, 2575 Sand Hill Rd, MS 46  |  Voice:    650-926-2924< >  Menlo Park, CA  94025           |  FAX:      650-926-3515L >  ------------------------------------------------------------------------- ----D >  These opinions are mine, not SLAC's, Stanford's, nor the DOE's...   ------------------------------   Date: 22 Jun 2000 14:54:26 GMT2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)+ Subject: Re: DECNET-Plus, DTSS, NTP and UTCo6 Message-ID: <8it9b2$306$2@mailint03.im.hou.compaq.com>   In article <M0Ze4F51Pb3r@mccdev.slac.stanford.edu>, Fairfield@SLAC.Stanford.EDU (Ken Fairfield; SLAC: 650-926-2924; FAX: 926-3515) writes:I :        I need to bring  up  a  bunch  of  old issues here: DECNET-Plus,yI :    DTSS,  NTP,  but  also  UTC and the  various  SYS$TIMEZONE*  logical8
 :    names...  ..     If it were me:  J   Shut off DTSS, and use NTP for synchronization.  While it is possible toK   get DTSS to query NTP for the time, the example code that permitted this -I   to be set up was found buggy and also assumed an old NTP, and was thus nJ   removed from the DTSS kit a while back.  (I'd have prefered to have seen(   it rewritten, but that didn't happen.)  M   Always please use SYS$MANAGER:UTC$TIME_SETUP.COM for timezone stuff -- not eM   the SYS$MANAGER:UTC$CONFIGURE_TDF.COM or SYS$MANAGER:UTC$TIMEZONE_SETUP.COMg
   procedures.(  I   V7.3 has integrated time management, and I'm working with several otheroJ   folks to get the OpenVMS FAQ and the OpenVMS V7.3 documentation updated.H   Currently C and IP use the base OpenVMS time management.  Starting in F   V7.3, DTSS also uses base OpenVMS for time management, and I've alsoF   asked for an "ignore" logical to disable DTSS startup when required.G   (Rather than the current scheme that requires it to start and then togH   shut down.)  Support for (optionally enabled) automatic daylight time @   switchovers also (finally) finds its way into OpenVMS in V7.3.  N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  % Date: Thu, 22 Jun 2000 18:33:24 +0200l. From: "Marc Van Dyck" <marc.vandyck@skynet.be>+ Subject: Re: DECNET-Plus, DTSS, NTP and UTC * Message-ID: <8itevf$u52$1@news0.skynet.be>   Hoff,   = Being a VMS guy from the good old days, I'm doing all my timeb synchronization with DTSS,I and with a program that I have written to feed the time from an external,3 radio-controlled clock to DTSS.  J Does that mean that when 7.3 arrives, I will have to rework that program ?K it uses the DTSS provider interface as described in the documentation. Will- that interface remain valid ?z  E Also, later on, my external clock will disappear and the time will be  provided to me by NTP.L Will I be able to extract the time from NTP and feed it into DTSS, and how ?   -- Marc  (Marc.Vandyck@skynet.be)L "Hoff Hoffman" <hoffman@xdelta.zko.dec.nospam> a crit dans le message news:+ 8it9b2$306$2@mailint03.im.hou.compaq.com...m >rK >   V7.3 has integrated time management, and I'm working with several othertL >   folks to get the OpenVMS FAQ and the OpenVMS V7.3 documentation updated.I >   Currently C and IP use the base OpenVMS time management.  Starting inCH >   V7.3, DTSS also uses base OpenVMS for time management, and I've alsoH >   asked for an "ignore" logical to disable DTSS startup when required.I >   (Rather than the current scheme that requires it to start and then tohI >   shut down.)  Support for (optionally enabled) automatic daylight time B >   switchovers also (finally) finds its way into OpenVMS in V7.3. >-, >  --------------------------- pure personal# opinion ---------------------------e1 >    Hoff (Stephen) Hoffman   OpenVMS Engineeringm hoffman#xdelta.zko.dec.com >u   ------------------------------  % Date: Thu, 22 Jun 2000 17:19:55 +0200"4 From: "Keith Martin" <keith.martin@uk.origin-it.com>' Subject: Equivalent of LAT ports in UCX Y Message-ID: <0E16861EE7BCD111BE9400805FE6841F10A25788@c1s5x001.cor.srvfarm.origin-it.com>E  I Does anybody know if there is any equivalent to the LAT port in UCX under  VMS and how they are created?r   ------------------------------   Date: 22 Jun 2000 15:47:58 GMT2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)+ Subject: Re: Equivalent of LAT ports in UCXt6 Message-ID: <8itcfe$4dn$1@mailint03.im.hou.compaq.com>   In article <0E16861EE7BCD111BE9400805FE6841F10A25788@c1s5x001.cor.srvfarm.origin-it.com>, "Keith Martin" <keith.martin@uk.origin-it.com> writes:J :Does anybody know if there is any equivalent to the LAT port in UCX under :VMS and how they are created?  #   The OpenVMS FAQ section entitled:e  ?     MGMT36. How can I set up reverse telnet (like reverse LAT)?o     might be of interest here.  #   The OpenVMS FAQ is available via:   "     http://www.openvms.compaq.com/  N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------   Date: 22 Jun 2000 14:57:42 GMT2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman): Subject: Re: Error In VMS SPD For AlphaServer 2100 5/375 ?6 Message-ID: <8it9h6$306$3@mailint03.im.hou.compaq.com>  R In article <8is5sr$gp$1@joe.rice.edu>, leslie@clio.rice.edu (Jerry Leslie) writes:4 :The OpenVMS SPD "Alpha Systems Supported" section:: ..B :mentions that "AlphaServer 2100 (All chip speeds)" are supported,D :but "Hoff", in message ID 8hh1a7$5v1$2@mailint03.im.hou.compaq.com, :states: :iC :  "There are examples where OpenVMS Alpha itself does not support A4 :   specific platforms (eg: AlphaServer 2100 5/375)"  G   I'll pass this (apparent error) along to the SPD maintainer.  Thanks!h  N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  # Date: Thu, 22 Jun 2000 16:50:25 GMT ) From: sfm1115@bjcmail.carenet.org (Shawn)e# Subject: Finding a Hardware Addressn1 Message-ID: <395242ee.268633865@news.starnet.net>g  B How can I find out what the Hardware Address is for an AlphaServer running OpenVMS 7.1o  A For Example when I am in ncp and trying to issue the connect nodebC <System Name> it responds telling me I need the hardware address ofm" the box I am trying to connect to.  : How do I go to that box and retrieve the hardware address?     Thanks   Shawn    ------------------------------   Date: 22 Jun 2000 17:43:45 GMT2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)' Subject: Re: Finding a Hardware Addressg6 Message-ID: <8itj8h$6s3$1@mailint03.im.hou.compaq.com>  ] In article <395242ee.268633865@news.starnet.net>, sfm1115@bjcmail.carenet.org (Shawn) writes:lC :How can I find out what the Hardware Address is for an AlphaServerE :running OpenVMS 7.1  C   You might want to take a look at the OpenVMS FAQ section entitledrD   "MISC7.  How to determine the network hardware address?".  The FAQD   itself is available via a link at http://www.openvms.compaq.com/,    among other places.e  N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  # Date: Thu, 22 Jun 2000 15:47:52 GMT * From: Art Rice <arice.NOcSPAM@ue.itug.org> Subject: Re: Fun VMS Facts?w8 Message-ID: <i1d4lskfj2rcjukm976mvokb6adcvhrpf4@4ax.com>  . On Mon, 19 Jun 2000 11:31:31 -0400, Tim Shoppa! <shoppa@trailing-edge.com> wrote:q   >Peter Weaver wrote: >> ED >> I took a look at http://www.kevinmitnick.com/sentest.html since IF >> thought a printout of the actual statement posted on the wall would< >> look good. The closest thing I can find is Mitnik saying; >> l
 >> ========== @ >> I have 20 years experience circumventing information security >> measures, and canA >> report that I have successfully compromised all systems that Ic >> targeted for   >> unauthorized access save one.
 >> ==========o >>  G >> I could not find any place where he said what that one system was. InG >> could not find any mentioned of VMS, Digital or DEC in the document.hI >> If anyone finds a quote where he says that the one system was VMS theni >> let us know.e >eB >I'm sure he's broken into VMS systems - Kevin Mitnick is a masterE >of "human factors" weaknesses, and calling up the VMS system manager C >and saying that you're a user who has lost his password works just*? >as well there as it does anywhere else.  And 12+ years ago the D >"FIELD SERVICE" account and password tricks were in place in a huge# >fraction of the sites running VMS.:  D I actually heard about one place where the FIELD SERVICE account hadC been expired purposely and someone set the new password as SERVICE.e     -- a
 Art Rice   **t# Special Data Processing Corporation & --------------------------------------' All opinions expressed are mine and do n% not reflect the views of my employer.    ------------------------------  % Date: Mon, 19 Jun 2000 17:58:16 +0200a= From: Arne =?iso-8859-1?Q?Vajh=F8j?= <arne.vajhoej@gtech.com>( Subject: Re: Hotmail) Message-ID: <394E4318.C8255626@gtech.com>a   John Vottero wrote:-9 > "Arne Vajhj" <arne.vajhoej@gtech.com> wrote in message0% > news:394BA97A.FFA2C827@gtech.com....: > > The danish computerworld online told, that hotmail was4 > > unavailable for a long period and that they lost > > email for 330000 users.0 > > ) > > Does anyone have details about that ?5 > >eK > > AFAIK then the Hotmail backend is on Solaris ! Maybe we have a new Ebay  > > ????  D > It probably happened when they tried to switch from Solaris to NT!   This was just weeks ago.  @ I thougth it was years ago MS tried the Solaris-NT move (without	 success).-   Have they tried again ?    Arne   ------------------------------   Date: 22 Jun 2000 08:04 CST ' From: carl@gerg.tamu.edu (Carl Perkins)g6 Subject: Re: How do I calculate CONNECT TIME from DCL?- Message-ID: <22JUN200008043730@gerg.tamu.edu>r  5 kaplow_r@eisner.decus.org.mars (Bob Kaplow) writes...uM }> Here's a version that should work with pretty much any valid VMS date-timetJ }> for the login and current time, up to a login time on 17-nov-1858 and a: }> curent time on 31-dec-9999 (I think it should, anyway). } K }It's not quite so easy. I tried this and found some processes on my systemy7 }with more uptime than the system uptime! A neat trick.D
 }	Bob Kaplow	v   Oops.t  , I knew I should have tested some more cases.  ; As usual, the best way to debug your program is to post it.h   --- Carl   ------------------------------    Date: 22 Jun 2000 06:45:57 -0700 From: locate@locate-now.comS Subject: Internet & scriptingvO Message-ID: <67EFFCF1.69DA5DC9.5f80b4b24baa3cb43895ee5eabf00388@locate-now.com>c   --=200006220639=) Content-Type: text/plain;charset=US-ASCII   ; For you programmers or those who are interested in computereI langage. I found while surfing the internet a source that explains, html,-) pearl, asp, python, smil,javascrip, etc. 2  0 Check At: http://www.locate-now.com/adcomp1.htm ! click on "internet and scripting": --=200006220639=--   ------------------------------    Date: 22 Jun 2000 07:49:19 -05009 From: Kilgallen@eisner.decus.org.nospam (Larry Kilgallen)c- Subject: Re: Mitnick (was Re: Fun VMS Facts?)C+ Message-ID: <ysoU3N$5VIjW@eisner.decus.org>n  ^ In article <sl26h0k1e7f56@corp.supernews.com>, "Peter Weaver" <peter.weaver@stelco.ca> writes:  A > better yet, have Regis ask a contestent "According to Mitnick'snF > testimony in front of the U.S. Congress, what was the only operatingE > system he could not break into? ... Is that your final answer?...")D > would be great for VMS.e  @ Would it ?  What is your current programming language after that1 television game show asked a question about Ada ?i   ------------------------------  % Date: Thu, 22 Jun 2000 11:34:29 -0400l- From: "Peter Weaver" <peter.weaver@stelco.ca>2- Subject: Re: Mitnick (was Re: Fun VMS Facts?)U/ Message-ID: <sl4ch4l3e7f154@corp.supernews.com>.  $ Larry Kilgallen wrote in message ...> >In article <sl26h0k1e7f56@corp.supernews.com>, "Peter Weaver"  <peter.weaver@stelco.ca> writes: >oB >> better yet, have Regis ask a contestent "According to Mitnick's= >> testimony in front of the U.S. Congress, what was the onlyd	 operatingsF >> system he could not break into? ... Is that your final answer?...") >> would be great for VMS. >tA >Would it ?  What is your current programming language after that 2 >television game show asked a question about Ada ?  F OK, you got me, I am one of the three people in North America who haveC never seen the show so I have never heard the question about Ada...,     -- Peter Weaver   ------------------------------    Date: 22 Jun 2000 13:32:07 -05009 From: Kilgallen@eisner.decus.org.nospam (Larry Kilgallen)C- Subject: Re: Mitnick (was Re: Fun VMS Facts?)e+ Message-ID: <WAmTkhDsgNnK@eisner.decus.org>u  _ In article <sl4ch4l3e7f154@corp.supernews.com>, "Peter Weaver" <peter.weaver@stelco.ca> writes:t& > Larry Kilgallen wrote in message ...? >>In article <sl26h0k1e7f56@corp.supernews.com>, "Peter Weaver"c" > <peter.weaver@stelco.ca> writes: >>C >>> better yet, have Regis ask a contestent "According to Mitnick's > >>> testimony in front of the U.S. Congress, what was the only > operatingoG >>> system he could not break into? ... Is that your final answer?...")  >>> would be great for VMS.e >>B >>Would it ?  What is your current programming language after that3 >>television game show asked a question about Ada ?  > H > OK, you got me, I am one of the three people in North America who haveE > never seen the show so I have never heard the question about Ada...t  F Well I am another of those three people, but some weeks back I startedF reading incomprehensible references on Ada mailing lists that had someI antecedent discussion or event I had obviously missed.  Actually, though,Y I didn't miss it much.   ------------------------------  % Date: Mon, 19 Jun 2000 17:59:00 +0200 = From: Arne =?iso-8859-1?Q?Vajh=F8j?= <arne.vajhoej@gtech.com>  Subject: Re: New to VMSt( Message-ID: <394E4344.982371E@gtech.com>   Vikas Desai wrote:M >     I am trying to write a program which will do some actions automaticallyiI > for a user when he receives a mail in his inbox. As soon as the mail isLK > received I want to process the mail and perform some actions depending onl > the contents of the mail.o > L >     My problem is how do I trap the event of receiving a mail and how do I > manipulate the mail box. >  > OS = OpenVMS V7.1 on Alpha.r   Look for the DELIVER utility.    Arne   ------------------------------  % Date: Thu, 22 Jun 2000 07:30:59 -0400e+ From: "Main, Kerry" <Kerry.Main@compaq.com>-7 Subject: RE: OpenVMS clusters vs other systems clustersfJ Message-ID: <910612C07BCAD1119AF40000F86AF0D80528444E@kaoexc4.kao.dec.com>   Bill,t   [just catching up on emails ..]   E >>> But another way of looking at the situation is that even with its D current lead VMS isn't setting the world on fire (in terms of marketK penetration) compared to its competition, and since the competition appearspL to be moving a good deal faster than VMS is this situation seems likely only to get worse.>>   L Well, I'll will be the first to say that I agree that certain things done inH the past could have been done much better from a marketing perspective.   L However, does that mean the future is doomed as well? Is there no way a past  mistake(s) can not be corrected?  K OpenVMS is a $4B a year business. 17% of OpenVMS business last year was newBC business. Andrew will not want to believe this, but thats reality. e  K Of course, more can be done and more is indeed what is happening behind ther( scenes. This work will soon be evident.   L We now have Customers that are requesting NT and UNIX TO OpenVMS porting andK migration assistance. As a multi-OS provider, Compaq can assist in whatevert- OS Customers want migration assistance with. Q  C Check out these pointers and then tell me that the situation is noti improving. M  G <http://www.openvms.digital.com/gsseries/index.html> (look at recent GS  Series testimonials)H <http://www.openvms.digital.com/gsseries/quotes.html> (including Oracle)D <http://www.compaq.com/alphaserver/gs/quotes/etrade.html> (one of my favourites)o  ? <http://www.openvms.digital.com/e-postcard1/> (ensure sound on)-- <http://www.openvms.digital.com/e-postcard2/>-L Yes, the last two are marketing sound bites, but is this not what folks haveK been asking for? Are these not the types of things that one can use to kickGI off a presentation on OpenVMS with? [Click on the title below the picture-3 and save the .exe file to run directly on your PC.]a   Recent partner site updates:; <http://www.cognos.com/adtpci/> Web Applications on OpenVMSL  <http://www.cognos.com/openvms/>  C <http://www.e-dbms.com/analysts/2000/benchmark.html> Cache database % performance press release on OpenVMS.lG <http://www.legato.com/News/pr00022index.html> (Legato planning OpenVMSo release)  K Anyway, I know the response will likely be something to the effect "we haveoL heard it before..", but I guess my response to that would be "stay tuned .."   :-)    Regards,  
 Kerry Main Senior Consultant,
 Compaq Canadad Professional ServicesP Voice : 613-592-4660 FAX   : 819-772-7036 Email : kerry.main@compaq.come       -----Original Message-----, From: Bill Todd [mailto:billtodd@foo.mv.com]& Sent: Wednesday, June 21, 2000 1:01 AM To: Info-VAX@Mvb.Saic.Com 7 Subject: Re: OpenVMS clusters vs other systems clustersZ      4 Main, Kerry <Kerry.Main@compaq.com> wrote in messageD news:910612C07BCAD1119AF40000F86AF0D805284436@kaoexc4.kao.dec.com... > JF,F >aA > While Tru64 now has some of the capabilities that is in OpenVMSe clustering,n* > the OpenVMS world is not standing still. >SK > This is the same argument that folks have said about other OS's (like NT)gK > that will "soon" have the same features as OpenVMS. When they catch up ine asJ > few years to where OpenVMS is today, they find that OpenVMS has moved on up > a number of notches already.  F As usual, you're putting as good a face on things as is possible.  ButJ another way of looking at the situation is that even with its current leadD VMS isn't setting the world on fire (in terms of market penetration)K compared to its competition, and since the competition appears to be moving3F a good deal faster than VMS is this situation seems likely only to get worse.  I Some of the blame can be placed on Compaq's (and DEC's before it) abysmal K (non-)marketing, so improvements in that area should temporarily give VMS azF bounce or at least halt its downward slide (yes, its numbers have beenG relatively constant over time, but in a growing market that constitutestJ losing ground).  But over anything but the short run, VMS is going to haveD to develop functionally at the same rate as its competition to avoid returning to decline.    >oJ > So, if you are a mission critical Customer that needs high availability, doK > you implement a departmental OS today because in a few years, you feel it,I > will have the same features as your current OS? Or, do you realize thatd whenI > the dept OS reachs where you are today, the current mission critical OS. will  > be 2 or 3 levels higher again?  K VMS has no lock on high availability any more - well, it never did, but nowOL its competition includes 'standard' Unix and even Windows environments.  TheJ eagerness with which customers are accepting such solutions makes it clearL that VMS's advantages, real though they may be, are not decisive compared toG other considerations (in other words, the competition offers *adequate*e6 availability plus many attractions that VMS does not).  D Compaq can certainly derive some incremental revenue and profit fromJ allowing VMS out of the closet.  But it won't get anything like the marketJ penetration it *could* get without a good deal of development work as well5 (IMO a good deal *more* than its road maps indicate).o   - bill   >oL > It is a constantly sliding upward scale. As each new release of a OS comesF > out, it is usually better, more improved than the last. W2K has moreF > improved features and is more stable than NT4. Absolutely. For thoseE > Customers that have implemented NT solutions, this is a good thing.k > K > However, OpenVMS V7.2-1 is also more improved, and has more features thaneL > OpenVMS V7.1. OpenVMS V7.3 will raise the RASS (reliability, availability,) > scalability, security) bar even higher.  >fJ > With respect to Tru64 now having some of the OpenVMS cluster features, IJ > view this as a strength of having multiple OS offerings. Tru64 gets some ofG > the good stuff from OpenVMS. OpenVMS gets some of the good stuff fromu Tru64.> > Same goes with respect to sharing with the Tandem NSK folks. >tH > I would be surprised if the IBM folks did not look at this in the sameH > manner. Sharing of technologies between multiple OS's is a good thing. >eG > For those vendors that offer only one OS, enhancements, ideas and newl stuffa* > is usually coming from the same sources. >-J > >>> I do not know if that is really true or not. If it is true, it meansH > that any vendor can develop clustering technology that rivals VMS in a very > short time.<<< >-G > Not. No matter how one slices it, clustering is really tough stuff tonL > design, implement and support - especially if you plan to put into mission > critical environments. >@
 > Regards, >u > Kerry Main > Senior Consultant, > Compaq Canada, > Professional Servicesr > Voice : 613-592-4660 > FAX   : 819-772-7036 > Email : kerry.main@compaq.comn >n >h >" > -----Original Message-----6 > From: JF Mezei [mailto:jfmezei.spamnot@videotron.ca]( > Sent: Saturday, June 17, 2000 10:09 PM > To: Info-VAX@Mvb.Saic.Coms9 > Subject: Re: OpenVMS clusters vs other systems clustersc >a >r > Larry Kilgallen wrote:E > > I doubt that you will find Compaq funding such a study for public.F > > release.  The only "good" result possible from a Compaq standpointE > > is one that measured Tru64 and VMS as absolutely identical, since ; > > otherwise one departement or another would be offended.  >aF > The Montreal Compaq office has said a few times that Tru64's currentL > clustering capabilities now include ALL the functionality that was present > in+ > openVMS (past tense is theirs, not mine).a >aK > I do not know if that is really true or not. If it is true, it means thata > anypJ > vendor can develop clustering technology that rivals VMS in a very short > time.h >a   ------------------------------  # Date: Thu, 22 Jun 2000 16:44:49 GMTt= From: system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-) 7 Subject: RE: OpenVMS clusters vs other systems clusters70 Message-ID: <009EBFD0.1DD7E428@SendSpamHere.ORG>  x In article <910612C07BCAD1119AF40000F86AF0D80528444E@kaoexc4.kao.dec.com>, "Main, Kerry" <Kerry.Main@compaq.com> writes:@ ><http://www.openvms.digital.com/e-postcard1/> (ensure sound on)  5 ... and how do I do that with Netscape 3.03 on VMS???     . ><http://www.openvms.digital.com/e-postcard2/>M >Yes, the last two are marketing sound bites, but is this not what folks haveSL >been asking for? Are these not the types of things that one can use to kickJ >off a presentation on OpenVMS with? [Click on the title below the picture4 >and save the .exe file to run directly on your PC.]  C Will Compaq provide a PeeCee for these or persist in neglecting andiB alienating the VMS users that they currently possess???  If CompaqC will provide me with the PeeCee (sorry, it would have to be runningSC Linux as I will have no demon seed spawning under my roof), I mightj. be able to see what these links are all about.  C Sorry, Kerry, but if they were truly interested in getting the wordqD out, these links (IMHO) would be less PeeCee specific.  It just doesD *NOT* demonstrate that they have sincere intentions for VMS.  PeeCee" or not to be, that is their quest.    L >Anyway, I know the response will likely be something to the effect "we haveM >heard it before..", but I guess my response to that would be "stay tuned .."a   To what channel???    1 BTW, I was great to finally meet you back in May.> --O VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COMt   ------------------------------  + Date: Thu, 22 Jun 2000 17:06:02 +0000 (   ) 3 From: Christopher Smith <chriss@Mufasa.pubserv.com>>7 Subject: RE: OpenVMS clusters vs other systems clustersmJ Message-ID: <Pine.LNX.4.05.10006221703390.28849-100000@Mufasa.pubserv.com>  9 On Thu, 22 Jun 2000, Brian Schenkenberger, VAXman- wrote:   E > Sorry, Kerry, but if they were truly interested in getting the wordoF > out, these links (IMHO) would be less PeeCee specific.  It just doesF > *NOT* demonstrate that they have sincere intentions for VMS.  PeeCee$ > or not to be, that is their quest.  G Well, it could be argued that nearly all the stupid people in the worldEJ (which are a majority) have peesees, so they're certainly getting the wordG out.  Maybe they just didn't think it nessecary to preach to the choir.r   Regards,   Chris-    O ===============================================================================i@ "My two cents"			(http://rootworks.com/twocentsworth.cgi?128562)= Christopher Smith(chriss@pubserv.com)			Prgramer^W ProgrammerF Prime Synergy of Champaign, IL.u% -------------------------------------1I "Where a calculator on the ENIAC is equipped with 18,000 vacuum tubes andeH weighs 30 tons, computers in the future may have only 1,000 vacuum tubes; and weigh only 1.5 tons." -- Popular Mechanics, March 1949  O -------------------------------------------------------------------------------    ------------------------------    Date: 22 Jun 2000 09:30:22 -0500, From: koehler@eisner.decus.org (Bob Koehler)2 Subject: Re: OpenVMS UCX/FTP and Internet Explorer+ Message-ID: <JUNFY36jhvOE@eisner.decus.org>   ` In article <IUe45.15559$Pz6.104147@news1.mts.net>, "Randy Jung" <rpjung@mb.sympatico.ca> writes: > L > As far as compliance with the RFC goes I took a quick look at RFC 0959 andJ > it looked to me there was no specification for the output of a directoryG > command. Indeed, I didn't even find any reference to "DIRECTORY" as a8J > command and the "LIST" and "NAME LIST" commands did not specify anythingI > about what the output looked like. Perhaps whose an expert on FTP couldaN > point out where these commands are defined or where I missed the definition. >   H    They're not.  The RFCs are incredibly sloppy.  I've been working withJ    the RFC959 (FTP).  In passive mode when you open a data connection you F    get (or send) a "the host and port address this server is listeningF    on".  If you've read the RFC carfully you may be able to figure out<    this information is embedded somewhere in a 227 response.  E    Notice no details are given as to the layout of the information;  aF    one might assume that the data is at least layed out in the manner I    specified by example for a DATA PORT command (non-passive connection) 0    "PORT h1,h2,h3,h4,p1,p2".  D    Generally you do get a 227 response ending in "PORT" and a comma G    separated list of numbers over the control socket which include the lD    IP addresss of the other system and the port to connect to.  But -    sometimes you get parentheses around them.n  G    The RFC doesn't say you could have parentheses, but some FTP servers-F    do it.  IIRC it doesn't say the data will always come at the end ofF    the 227 response, but it seems like it does.  If you're writing an E    FTP client you have to test against just about every platform you 0B    can get your hands on and live with what they really did, whileH    trying not to assume anything (i.e. parse for "PORT" because the data,    might not be at the end of the response).    a   ------------------------------    Date: 22 Jun 2000 09:40:16 -0500, From: koehler@eisner.decus.org (Bob Koehler)2 Subject: Re: OpenVMS UCX/FTP and Internet Explorer+ Message-ID: <p4iT7eSQqlQ9@eisner.decus.org>w  \ In article <39518217.98212106@videotron.ca>, JF Mezei <jfmezei.spamnot@videotron.ca> writes:  O > Sites should just tell users to switch browsers because the fault is in theira2 > browser not complying to the internet standards.  E    I go just a little easier on them.  I post my .com files on my web H    server and note on the page that IE users will have to select save to    disk.  G    If they use IE and can't read, or don't actually know how to use IE,>#    they probably don't need my DCL.e   ------------------------------   Date: 22 Jun 2000 14:00:49 GMT' From: david20@alpha2.mdx.ac.uk (D.Webb)>2 Subject: Re: OpenVMS UCX/FTP and Internet Explorer0 Message-ID: <8it66h$pj9$1@aquila.news.mdx.ac.uk>  e In article <01JQVVQS6T2Q003G3C@Eisner.DECUS.org>, Jeff Schreiber <SCHREIBER@Eisner.DECUS.org> writes:m, >Randy Jung <rpjung@mb.symaptico.ca> writes: >>L >>As far as compliance with the RFC goes I took a quick look at RFC 0959 andJ >>it looked to me there was no specification for the output of a directoryG >>command. Indeed, I didn't even find any reference to "DIRECTORY" as aaJ >>command and the "LIST" and "NAME LIST" commands did not specify anythingI >>about what the output looked like. Perhaps whose an expert on FTP couldtN >>point out where these commands are defined or where I missed the definition. >> >mK >    The original 959 authors cut a few corners, and there are a few placesaI >    where they skimped on the standard.  One easy example can be seen ino >    reference to pathname:  >  >      pathnameo >>F >         Pathname is defined to be the character string which must beG >         input to a file system by a user in order to identify a file.rG >         Pathname normally contains device and/or directory names, and G >         file name specification.  FTP does not yet specify a standardcE >         pathname convention.  Each user must follow the file namingoC >         conventions of the file systems involved in the transfer.  >kE >    Instead of designing a standard filenaming format, they opted toaJ >    make it the user's responsibility to know the file naming conventions" >    on each side of the transfer. >-F >    An interesting side effect of the "user's responsibility" is thatI >    something like "ls -la" to a unix system will try to get a directory:K >    list of files named "-la", however unix systems saw that and sent back H >    the output of an "ls -la".  So we had customers asking for "ls -la"K >    support in VMS, because unix did it, therefore we were violating RFCs.i > I >    Anyway... onto the point at hand.  The important section is where itl" >    talks about the LIST command: >  >         LIST (LIST)I >aI >            This command causes a list to be sent from the server to the I >            passive DTP.  If the pathname specifies a directory or other G >            group of files, the server should transfer a list of fileseE >            in the specified directory.  If the pathname specifies auH >            file then the server should send current information on theI >            file.  A null argument implies the user's current working or C >            default directory.  The data transfer is over the datawE >            connection in type ASCII or type EBCDIC.  (The user must D >            ensure that the TYPE is appropriately ASCII or EBCDIC).H >            Since the information on a file may vary widely from systemI >            to system, this information may be hard to use automatically C >            in a program, but may be quite useful to a human user.n >uL >    That last sentence is the "out" for the authors.  "this information mayL >    be hard to use automatically in a program".  Since Internet Explorer isK >    going to find the filenames, and make hyperlinks out of them, they areB6 >    automatically using the information in a program. > J >    So in general, it's Internet Explorer who is at fault, as the RFC has? >    no standard for the formatting of the LIST command output.w >v  M Internet Explorer should instead do what I presume Netscape does and use the i
 NLST command.l  K The RFC does in fact give a format for the results returned by this commandi8 (as shown by the part of the RFC I have marked with *s).          NAME LIST (NLST)   K      This command causes a directory listing to be sent from server to user K      site. The pathname should specify a directory or other system-specificoK      file group descriptor; a null argument implies the current directory. r  E *     The server will return a stream of names of files and no other o *     information. nD *     The data will be transferred in ASCII or EBCDIC type over the M *     data connection as valid pathname strings separated by <CRLF> or <NL>. ?= *     (Again the user must ensure that the TYPE is correct.) f    I      This command is intended to return information that can be used by aaL      program to further process the files automatically. For example, in the1      implementation of a "multiple get" function.o  p  M Hence Internet Explorer should use this command to get it's list of filenamesb@ which can then be displayed as icons and be setup as hyperlinks.  P If it really wished to add in the other information provided by the LIST commandL then it should firstly use the NLST command to get the list of filenames and7 then use the LIST command to get the other information.a  N It could then match up the filenames (and create hyper-links out of them ) andA display the other information returned by the LIST command as is.   O I doubt if this is worth doing (the matching would always run the risk of beingrJ caught out by strange choices of filename and additional information being returned eg fictional example  G LIST returns   filetype (A for ASCII, B for binary) filename  filesize e   filename happens to be A  O Which A do you make a hyperlink out of since you don't know the format returned. by the LIST command ?    )     M However bottom line is Internet Explorer should be using the NLST command notoL the LIST command since this is the command which should be used by programs  according to the RFC.k      
 David Webb VMs and Unix team leader CCSS Middlesex University   ------------------------------   Date: 21 Jun 2000 22:29:42 GMT* From: bleau@umtof.umd.edu (Lawrence Bleau)> Subject: Re: Proxy problem: Why does node:: work but 0:: fail?) Message-ID: <8irfkm$snm$1@hecate.umd.edu>l  Z In article <8ip591$qgh$1@pyrite.mv.net>, morris@iridium.mv.net (Skipper W. Morris) writes:B >So if you have clusters, the problem might be that you don't have? >a proxy for the other possible nodename in the proxy database.t  5 No, this is a standalone system, no cluster involved..  B >Easiest thing is do a "set host 0" vs "set host foo" and when you= >login look at where the system thinks the connect came from.t  3 It says I came from 6903::SYSTEM when I SET HOST 0.n= It says I come from ULEIS::SYSTEM when I do a SET HOST ULEIS.r< 6903 is the correct numeric equivalent of my Decnet address.  G This works fine.  The problem is using FAL with a numeric address.  Thef intruder database shows:  A Intrusion       Type       Count        Expiration         SourcefF    NETWORK      SUSPECT       1   21-JUN-2000 18:33:29.39  6903::BLEAU     Lawrence Bleau University of Maryland" Physics Dept., Space Physics Group 301-405-6223 bleau@umtof.umd.edu    ------------------------------  % Date: Thu, 22 Jun 2000 09:19:01 -0500a1 From: "Dave Gudewicz" <david.gudewicz@abbott.com> 0 Subject: Re: Questions about memory (DS20, ES40)8 Message-ID: <8it71v$8df$1@fizban.fizban.pprd.abbott.com>  - Ray T. <lists@aik.tec.sc.us> wrote in messagee' news:394EA579.170B27B5@aik.tec.sc.us...  > Rudolf Wingert wrote:a > >i
 > > Hello, > > D > > I do have some questions about the memory for the DS20 and ES40.E > > I did see, that there is DIMM with 512 MB for the DS10. Can I usetG > > four of them per bank on the DS20? Also there are 1024MB DIMMs out.s) > > Are they also usable within the DS20?iK > > Does anybody know the MHz of DS20 memory? The DS10 memory (SX-MS310-xy)oH > > is desribed as 100MHz and the same for XP900 as 87MHz. What's right?I > > If I am right, then the memory for the ES40 are a bankmodul with fouraJ > > DIMMs. Can I buy a cheap one (e.g. 128MB for an ES40 AlphaStation) andH > > exchange the DIMMs with higher sized cheap standard industry memory? > >a" > > TIA and regards Rudolf Wingert >a > Hi Rudolf, > G > I just know the DS10.  It has four slots yes, but when you buy memory E > for it from Compaq, you get a pair.  So the max memory is 2GB whichaE > you buy as a qty of two 1GB parts which come as two pairs of memory H > cards.  Yes you can use an industry standard memory IF it is the rightI > type and specs.  BUT to be safe, it would be best to stick to companiesOJ > that are approved for maintenance by Compaq and make memory specificallyB > marked for the computer you have, such as DataRam and Camintonn.H > Kingston also marks memory as made for DS-10, etc.  Note that the part2 > numbers are different between DS-10 and DS-20's. >r > Ray T.  B I believe there was a recent announcement that Compaq is no longerL supporting maintenance for DataRam and Camintonn.  It started in Nov. 99 and# ended May 00.  Dunno about Kinston.d   Dave...l   ------------------------------   Date: 22 Jun 2000 07:23:22 CDT= From: wayne@tachysoft.xxx.044962.killspam.0138 (Wayne Sewell)s# Subject: Re: QUEUE ENTRY NUMBERS...e. Message-ID: <ZTDSQUHpgF0K@tachxxsoftxxconsult>  W In article <sl20o1bve7f159@news.supernews.com>, "John Vottero" <John@mvpsi.com> writes: > > "Mark D. Jilson" <jilly@clarityconnect.com> wrote in message. > news:3950CAF7.54D97D4A@clarityconnect.com...J >> What problems are these numbers creating??  Entry numbers do not have aH >> documented format so any software that thinks it knows the format andI >> relies on this is faulty and needs to be fixed. One should treat entryFJ >> numbers as an opaque longword value and the sequencing of the job entryC >> number is undocumented, and is subject to change without notice.t >> > N > People want to type/remember 3 or 4 digit entry numbers, not 7 digits.  It'sK > the same whine that was heard when clusters first showed up and PIDs whend( > from a few hex digits to 8 hex digits.  M That's why I like decterms.  I can just cut and paste the sumbitches from shoe	 sys.  :-)e     --  O ===============================================================================iM Wayne Sewell, Tachyon Software Consulting  (281)812-0738  wayne@tachysoft.xxxn: http://www.tachysoft.xxx/www/tachyon.html and wayne.html  K change .xxx to .com in addresses above, assuming you are not a spambot  :-)rO ===============================================================================sO Otter, on dining with Bluto:"It's perfectly safe if you keep your arms and legs, 			away from his mouth."   ------------------------------   Date: 22 Jun 2000 14:42:19 GMT2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)# Subject: Re: QUEUE ENTRY NUMBERS...n6 Message-ID: <8it8kb$306$1@mailint03.im.hou.compaq.com>  C   I will repeat certain very important phrases in this posting, ando(   particularly the following key phrase:  7     The queue entry number is an opaque longword value.h    w In article <A5DB1F2A104EEACF.393F5AFF9D8A3C40.A3B58F97120E5C8F@lp.airnews.net>, Chris Scheers <asi@airmail.net> writes:b :Terry Marosites wrote:tO :>    But if that is the case how come I don't see any jobs over 4 digits on myeO :> VAX (V7.1). I am migrating to alpha (v7.1 also) will I see it then?   I haveuM :> been on Alpha's for years and have not noted any entry on 4 digits before, 3 :> is there a setting that control the wrap number?e  *   Again, no, there is no control for this.  5   The queue entry number is an opaque longword value.   E   The entry number should be treated just like the process  PID -- it D   is an opaque longword value.    It is not a small integer, and (inE   retrospect) it was probably a mistake to ever make it look like it r!   might be a small integer value.e   ..F :The last time I looked, the queue manager allocated a table for entryH :numbers 1-9999.  When a new job is created, entry numbers are allocatedI :sequentially from this table, skipping any numbers that are currently inhF :use.  After entry number 9999, the search restarts at entry number 1.  A   Correct, for this part of the current scheme.  This may change.b  5   The queue entry number is an opaque longword value.r  H :This works fine unless all of the entry numbers are used, i.e., entriesI :1-9999 all exist.  When this happens, the entry number table is extendedr% :adding new numbers, say 10000-19999.O  
   Correct.  5   The queue entry number is an opaque longword value.1  F :Even if you clear out all the jobs, once the table has been extended,F :you will continue to see the higher numbers until the job database is :recreated.e  
   Correct.  5   The queue entry number is an opaque longword value.s  H :AFAIK, the only way to see the higher numbers is to have 10,000 or more :entries simultaneously in use.e  F   Incorrect.  Configurations with multiple queue managers running can G   (and do) add some bits (currently) to the middle of the entry number.   5   The queue entry number is an opaque longword value.t  I   The current entry number assignment format -- subject to change without=7   notice -- looks something like this (decimal) format:=           X-QM-NNNNw  A           NNNN : 0 to 9999 : The queue manager entry number rangeeC           QM   : 00 to 04  : queue manager number (The QMan ID - 1)fD           X    : 0 to 9    : overflow, to extend the 0 to 9999 range  5   The queue entry number is an opaque longword value.   J   This scheme has been in place since the "new" queue manager was releasedJ   back in V5.5, though -- like the process PID -- this entry number value E   can and should always be treated as an opaque longword value.  This K   scheme can be changed in the future of course, as the format and contentshI   and algorithm involved in the assignment of the queue entry numbers is eJ   considered to be undocumented, and the value itself is considered to be    an opaque longword value.n  )   Oh, did I remember to mention :-) that:i  7     The queue entry number is an opaque longword value.   N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  # Date: Thu, 22 Jun 2000 16:15:06 GMTo* From: Art Rice <arice.NOcSPAM@ue.itug.org> Subject: Re: Real Time VMS? 8 Message-ID: <dge4lsoo9o33hsip75ekurmu4mpp70d272@4ax.com>  @ On Thu, 15 Jun 2000 20:32:37 +0100, Mark <Mark@spam-me-baby.com> wrote:   >Hi,G >    I was wondering if anyone know of any Hard Real Time extentions tomI >VMS? I currently work on a RT Linux (And extention to the standard LinuxpF >Kernel which allows true real time performance) system at home, whichI >got me thinking if such a thing was avaliable for VMS. The sort of thinghC >I'd be looking is well with in the micro second range for periodicy >tasks, interrupts etc.a > : >I'd be grateful for any info anyone's got on the subject. >n >Cheerss >i >Mark   > Seems to me that when Rockwell Switching Systems (now RockwellF Electronic Commerce) came out with the Galaxy ACD, that it was running> on the PDP11.  PBX and ACD switches are something that I wouldE consider Real-Time (or pretty darn close to it.)  Currently, they usedF proprietary hardware of their own design running OS9.  That machine is called Spectrum.   -- n
 Art Rice   **d# Special Data Processing Corporationr& --------------------------------------' All opinions expressed are mine and do e% not reflect the views of my employer.s   ------------------------------    Date: 22 Jun 2000 11:24:48 -0500- From: young_r@eisner.decus.org (Robert Young)dH Subject: Re: Storage Works / Snapshots / Maybe it's time to skip OpenVMS+ Message-ID: <zy42kBjaMDJJ@eisner.decus.org>a  r In article <3E5EC94AC27E4386.D26E5BDAB54A4480.5A1B3C3225E6F830@lp.airnews.net>, kuff@tessco.com (Hal Kuff) writes: > K >    yes sadly we are running one application using RMS files..... Guess weaI > could do Journaling but can't get a straight answer from Compaq on thatt	 > either.2 > L >    How would you script this, if no supportable tool is out there?  RumorsI > on HXTerm have abounded for ages.....    Would love to hear about thesea
 > options....  >  >   D         I wouldn't know how to go about adding Journaling capability=         to RMS.  A quick Wizard search reveals it hasn't beent8         a topic (yet).  From the VMS FAQ , we find this:  + http://www.faqs.org/faqs/dec-faq/vms/part3/   B FILE3.  How do I make sure that my data is safely written to disk?  E If your application must absolutely guarantee that data is available,lK no matter what, there's really no substitute for RMS Journalling.  However, J you can achieve a good degree of data integrity by issuing a SYS$FLUSH RMSD call at appropriate times (if you're using RMS, that is.)  If you'reM using a high-level language's I/O system, check that language's documentationnF to see if you can access the RMS control blocks for the open file.  InC C you can use fflush followed by fsync.  Note that fsync, which wasnK undocumented for VAX C but is documented for DEC C, takes a file descriptorn as an argument, not a *FILE.           ---            The method is simply:e  ,                 1) Quiesce (FREEZE) Database"                 2) Break shadowset$                 3) UNFREEZE Database#                 4) Backup shadowseti'                 5) Re-add shadow membert  K         Journals allow databse to be restored fully by applying to previouspG         good backup, as you are probably aware.  Journals are analogous>         to Oracle's redo logs.  F         All of 1-5 is DCL code for me with appropriate database calls.  G         You may be able to get a clean copy if you could somehow freezeiL         your database.  The problem is you may also lose a whole day's worthK         of work/entries... not good.  I would persue Journalling vigorously L         and work with your developers to add a hook to quiesce the databases6         if that functionality doesn't currently exist.  D         My HSDSA comment was directed at controller based mirrors orF         0+1.  What I am working with is Volume Shadowing and since theH         price of that license has dropped, I would persue that more thanD         the mirror based solution.  Storage is cheap these days, forF         high availability some of what I am doing involves a shadowsetG         where the master is a RAID1 (controller based mirrorset) membereJ         and the secondary shadowset member is a single drive.  This allowsD         me to be covered during those brief periods where the shadowE         member is broken out for backups.  A hardware failure in that 8         timeframe would be troublesome to say the least.  G         If you are running a business, you can't afford to skimp unless (         you are under draconian budgets.  H         Very recently (less than 1 month ago) power was lost at a clientH         site.  Power lost in a UPS backed up DataCenter... one of 2 PDUsH         tripped unexpectedly.  Without proper layout and redundant powerH         supplies plugged into opposing PDUs, I would have had a disaster@         to recover from.  I am pressing very hard to get 2 smallG         UPSes to plug my 115 volt based storage into.  Silly, isn't it?rG         But with UPS, I can set CACHE_UPS TRUE and battery status won'tCE         effect me.  If I had made the assumption that the datacenter yA         is on UPS, let's just set CACHE_UPS to TRUE AND not awareuB         I had lost a controller battery (slim but without periodicA         checks of the batteries .. could happen) I would have hadtJ         database corruption.  CACHE_POLICY and settings are a whole other          issue...  #                                 Robl   >  >  > K > In article <cVxVcTJMSKlp@eisner.decus.org>, young_r@eisner.decus.org (Robn > Young) wrote:  > 
 >> In articlelF > <05E9483E465FF40C.DBE5A72396AEDFFB.0693B6D86E224ECB@lp.airnews.net>,$ > kuff@tessco.com (Hal Kuff) writes: >> > p >> > vJ >> >    What we're looking for is a snapahot via command files or API thatJ >> > allows us to only take users offline for 5 mins.... cloning a stripedK >> > mirrorset would defeat the purpose....  Adding mirrors and taking themtL >> > away to convert them to units for mounting would be clumsy and probably >> > unscriptablem >> >   >> eK >>         It would be scriptable using HSDSA but I believe that is for HSJi" >>         use only (pretty sure). >> gE >>         I went back and read your original post .. there is littlea >>         detail here.. >> o, >>         Going to make a couple guesses.   >>  > >>                 1)  You are using older database technologyE >>                         (old but good) that allows FREEZING of the % >>                         databases.P >> cI >>                 2)  You're not using Oracle as you can perform backupsnK >>                         online with that and there would be no downtime.e >> eM >>                 3)  Since you are spending now or shortly, you are lookingnL >>                         at an ESA1000 (or follow-on) using Fibre Channel. >> bH >>         Yes, this can be done and in the timeframe you are looking atM >>         via DCL if the above are true and you have a reasonable number of t >>         volumes.    >> tE >>         The FREEZE mentioned in 1) causes writes to be flushed if iD >>         it is what I think it is... drop us more info out here or >>         drop me an email. >> h& >>                                 Rob   ------------------------------  % Date: Thu, 22 Jun 2000 08:30:31 -0400y, From: Howard S Shubs <hshubs@mindspring.com> Subject: Re: User mode logicalsn> Message-ID: <hshubs-98AB4E.08303122062000@news.mindspring.com>  6 In article <3951A8FF.905C39EF@videotron.ca>, JF Mezei % <jfmezei.spamnot@videotron.ca> wrote:h  I >Is there a way to create a cluster-wide logical which lives only during . >theK >time the application is executing ? (eg: logicals go away once application E >exits). (The program uses $CRELNM to create a bunch of cluster-wide a	 >logicalso  F Which application should they go away after?  A process logical knows G when -the- application terminates.  A job (or group?  Hm, something to -( try today) or greater logical might not.   -- l Howard S Shubs, the Denim Adept)   ------------------------------  % Date: Thu, 22 Jun 2000 14:15:38 +0100D/ From: Nigel Arnot <sysmgr@maxwell.ph.kcl.ac.uk>i  Subject: re:  User mode logicals7 Message-ID: <009EBFDC.CDAFDB25.18@maxwell.ph.kcl.ac.uk>v   > O > I was under the impression that user mode logicals died after an image exits.o > N > But it seems that this applies only to process logicals and not to system or > cluster logicals.g  K This is also how it's documented (in "OpenVMS Programming Concepts Manual",sM chapter on logical names). The discussion of image rundown is a subsection of C that on process logicals, implying it does not apply to the others.U > M > Is there a way to create a cluster-wide logical which lives only during thefL > time the application is executing ? (eg: logicals go away once applicationO > exits). (The program uses $CRELNM to create a bunch of cluster-wide logicals.  >   E The obvious one would be to SYS$DCLEXH an exit-handler to tidy up ther= names. If you worry about them being left behind after $STOP,sH you could use a lock-file opened for exclusive access when you start andF never closed. If a subsequent invocation *can* open the file, the onlyM explanation is that the creator process has died, and you use open-success assA a cue to re-set the logical names. Or you could stow its PID and tG JPI$_IMAGECOUNT into one of the logical names, and check it's alive and.# not iamge-exited using SYS$GETJPIW.   C I've seen a few situations where folks were using logical names forSL interprocess communications where a small shared ISAM file was an altogetherD neater solution. Not saying it applies here, but it's easy to forget+ how powerful and easy that solution can be.h   	Yours, 
 		Nigel Arnotk- 		NRA@MAXWELL.PH.KCL.AC.UK                   -  7 		"In the beginning there was nothing, which exploded."0   ------------------------------  # Date: Thu, 22 Jun 2000 16:08:59 GMT * From: Art Rice <arice.NOcSPAM@ue.itug.org>3 Subject: Re: VAX & Alpha exist but where is OpenVMS)8 Message-ID: <p9e4lss3hqvf7gb7d9caom926n7irqr2u5@4ax.com>  4 On Thu, 15 Jun 2000 16:44:51 -0400, "Ebinger . Eric" <EEbinger@drc.com> wrote:   3 >Microsoft Word spelling and grammar for the stringh >y >vms >m, >complains about capitalization and suggests >VMS >Vas >CMS >GMS >JMS > < >HDJKLHJKLDFHKLJDFHL results in "Not in Dictionary:" with no8 >suggestions.  It appears that VMS is in the dictionary. >   F Tried it for NSK and no problem.  Tried it with "nsk" and it suggested "ask."     -- e
 Art Rice   **a# Special Data Processing Corporation & --------------------------------------' All opinions expressed are mine and do  % not reflect the views of my employer.y   ------------------------------  % Date: Thu, 22 Jun 2000 10:40:43 +0100e/ From: Nigel Arnot <sysmgr@maxwell.ph.kcl.ac.uk>h Subject: re: VAX on Intel?6 Message-ID: <009EBFBE.C7E54F18.4@maxwell.ph.kcl.ac.uk>  , Andrew, surely you can do better than this?!   Andrew Harrison writes:t > 5 > What are the reasons why Linux has been sucessfull.  >  > 1.    Its small:/ > 2.    Its fast, particularly on small systems.! > 3.    Its more reliable then NTQ' > 4.    Its relatively simple to manageQ > 5.    Its free or nearly freen > J > But Linux is only architected to support the 1-2 CPU sweet spot occupiedF > by NT, it doesn't have  sophisticated SMP support, clustering direct > filesystems,N > commercial quality volume managers etc etc which the commercial UNIX's have. > 3 > So what happens when all this get added to Linux.o >  > 1.    It is no longer smallm  G We have VMS to prove that software bloat is not inevitable. And much of L Linux is modular-loadable, so although some growth in the size of the kernelO will doubtless occur, there's no reason to think that it has to be be dramatic.   J Care to let us know how much bloat there's been between the original Sparc o/s and the latest Solaris?   F > 2.    Its no longer as fast as it was, particularly on small systems  H Again, VMS is a counter-example. In fact, it's only Microsoft who manageE to make the rate of increase of software inefficiency exceed the rate=L of increase of processor performance! Although it's nice that Linux is stillJ highly usable on a "scrap" 80386, I doubt if there will be many complaints7 if future versions need a 486 or even a 100MHz Pentium.o   Again, how are Sun doing?u  ' > 3.    Its still more reliable than NT=) > 4.    Its no longer as simple to manage=  G Yet again, VMS proves this is not inevitable. And if the fancy bits areeJ loadable and you don't load them, you don't need to manage them. HopefullyO even if they have to load, they'll be happily ignoreable with default settings.n Like with VMS.  % > 5.    Its still free or nearly freen > G > Since this is also a description of most commercial UNIX's the choiceeB > of Linux or a commercial UNIX OS becomes one that is going to be> > based much more on issues like applications availability and > support, OS support etc.  K Commercial unix nearly free?! I suppose it depends where you draw the line.hL Fancy filesystems and clusterability (in the wide sense) still cost a prettyF packet. Linux and the open-source movement have the potential to make G alternatives to these commercial money-spinners free. Hence Microsoft'shM attempts to obfusticate and patent their protocols, and the "embrace, extend,o? extinguish" approach to open standards. (I don't know enough togB say whether they are uniquely evil or merely the worst offenders).  M Methinks Sun are suffering corporate nightmares about 200-foot-tall penguins!e   	Yours,I
 		Nigel Arnota- 		NRA@MAXWELL.PH.KCL.AC.UK                   n  7 		"In the beginning there was nothing, which exploded."e   ------------------------------  % Date: Thu, 22 Jun 2000 11:14:36 +0100lB From: Andrew Harrison SUNUK Consultancy <andrew.nospam@uk.sun.com> Subject: Re: VAX on Intel?* Message-ID: <3951E70C.AC22DDF9@uk.sun.com>   JF Mezei wrote:t  * > Andrew Harrison SUNUK Consultancy wrote:5 > > So what happens when all this get added to Linux.s > >  > > 1.    It is no longer smallfH > > 2.    Its no longer as fast as it was, particularly on small systems > ...rI > > Since this is also a description of most commercial UNIX's the choicenD > > of Linux or a commercial UNIX OS becomes one that is going to be@ > > based much more on issues like applications availability and > > support, OS support etc. > N > But if Linux is still "free" (especially on large corporation with many manyJ > seats (or requirements to handle manu IP connections), and large supportJ > organisations provide as good a job of supporting it, do the proprietaryM > Unixes of the world really still have such an edge worth paying the extra ?   O Solaris is essentially free, Sun charges 75 dollars for the OS distribution for  x86 machinesD with up to 8 CPU's. Once you have the distribution you can use it on as many systems as you want.  G This is much much less than many of the Linux vendors CD distributions.o  N Red-Hat for example charge 2,500 dollars for their Enterprise Edition of Linux and=M 179 dollars for their Professional Edition and Enteprise Edition does not getw close G to matching the capabilities of Solaris 8 for large enterprise systems.   K Granted Sun charges more for systems with more than 8 CPU's but since LinuxtL does not operate in the greater than 8 CPU market and since there hardly anyG commodity Intel systems with more than 8 CPU's this doe no seem to be a  harsh restriction.  G So perhaps you should be asking, is it worth paying extra for Linux ???n     regards  Andrew Harrisona Enterprise IT Architecta   ------------------------------    Date: 22 Jun 2000 08:06:02 -05009 From: Kilgallen@eisner.decus.org.nospam (Larry Kilgallen)d Subject: Re: VAX on Intel?+ Message-ID: <sIKBeSwbBrno@eisner.decus.org>d  \ In article <395136F8.83A2832A@videotron.ca>, JF Mezei <jfmezei.spamnot@videotron.ca> writes: > Bill Todd wrote:C >> that *if* Compaq is willing at least to try to revive VMS in theeM >> marketplace, they will want to start in areas where results can be readily-N >> measured - both in volume and in profits.  If they do that and the response/ >> is ho-hum, then they won't have risked much , > P > Yes, this is your guess, and probably Compaq's beancounters' thinking as well.  ? And that thinking must prevail over experimental ventures under.0 the current Compaq position on the stock market.  > As a big VMS fan, I would be _extremely_mad_ if people on this= newsgroup caused Compaq to make VMS so unprofitable that theyl> later dropped it. But I am not really thinking that is likely.   ------------------------------  % Date: Thu, 22 Jun 2000 13:25:41 +0100 / From: Nigel Arnot <sysmgr@maxwell.ph.kcl.ac.uk>h Subject: Re: VAX on Intel?7 Message-ID: <009EBFD5.D34CF1A0.20@maxwell.ph.kcl.ac.uk>y   > X-Gateway-Source-Info: USENETd$ > Sender: infovax-request@ulcc.ac.uk >  >  Bill Todd writes:r  < > Nigel Arnot <sysmgr@maxwell.ph.kcl.ac.uk> wrote in message3 > news:009EBF08.049ED189.22@maxwell.ph.kcl.ac.uk...t >  > > L > > The bigger issue is the VMS and VMS layered products software prices. AsN > > long as Compaq keep the single band structure for everything from the ES40J > > downwards, then the price of VMS on low-end platforms is going to look? > > expensive -- the lower you go, the more expensive it looks.  > J > This might be the most telling argument to take to Compaq:  if indeed itK > believes in tiered pricing models, then adding a tier in recognition that-N > entry-level sales are a special case is entirely compatible with its current > approach.d   Agree. > 
 >  And theM > > production cost of copies of software is very close to zero: whatever youn5 > > sell it for is pretty much all positive cashflow.B > >8I > > So, to first order Compaq could afford to sell whatever minumum Alpha I > > platform they havew running VMS for no more than the cost of the sameaG > > platform running Linux. Why, therefore, do they charge (much) more?a > ! > 1.  Software development costs.s  O I'd say that all the parts of VMS most relevant to an entry-level customer were K developed years ago (1984-90 timescale, apart from the Alpha porting). Most < of the new stuff is aimed more toward the big-iron hardware. >  > 2.  Software support costs.r  D If the customer wants it, let the customer pay for it. Don't make it compulsory.  >  > 3.  Software cost of sales.c   Don't understand this one. > % > 4.  Relative value of the software.t  K This is a key point. We often complain that Compaq advertising is preachinglI to the choir. I'd say that at the entry level, they are also marketing toHI the choir. Someone who doesn't have a VMS background is not going to takeFG Compaq's word if they say "it's four times better so we'll charge four oI times as much". It's got to sell against the others, price for price, andeE then win friends. The payback is later when those friends scale up tofE a bigger VMS system, or tell their friends to buy VMS and sales startc growing.   > >tL > > The argument for so doing is second-order: probably that they don't want5 > > to cannibalise their profitable high-end systems.d > I > 5.  Yes, that too, but as you note there are ways to avoid much of that)
 > problem. >  >  They may worry thatL > > customers might replace an ES40, or maybe something bigger, by a clusterI > > of small cheap VMS systems. So they keep VMS expensive and VMSclustera) > > still more so so that doesn't happen.a > > M > > IMO, what they miss in this analysis is that a new customer for VMS needs M > > to start somewhere, and it's far, far easier to start small than to start N > > big. Two of VMS's key andvantages are cluster-ability and scalability. ButF > > the proportionally high software costs for the lowest VMS platformG > > serve to place both of these advantages out of reach of the low-end M > > customers who'd maybe LIKE a non-stop (cluster) system, but can't justify M > > the cost. So the present policy is locking out what may be a large sourcewN > > of NEW customers -- the ones who can't contribute much to Compaq's profitsN > > with their first small purchase, but who are most likely to come back someI > > time later for a bigger system if their needs grow and they like whats	 > they'veeM > > got. Salesmen in just about any industry know the value of "a foot in thesE > > door" It's something that you can't price for a bean-counter, andm > something ( > > that Compaq seem to ignore with VMS. > N > That's not entirely fair:  VMS is one hell of a system to be able to buy forF > $1200.  Admittedly it lacks some things you might want, like TCP/IP,N > software RAID (at least shadowing), and defragmention, that are bundled withL > common low-end server-level systems nowadays, and perhaps isn't discountedK > when purchased with hardware as other systems are, both of which would be $ > reasonable to point out to Compaq.  I You know that. I know that. Someone who has never had a VMS system beforee@ doesn't, and isn't going to believe everything Compaq tells him. > K > In other words, there are reasonable suggestions one might make to CompaqcN > (more bundled software, discounts when purchased with the hardware) based onE > common industry practices that would reduce the cost of VMS withoutaK > explicitly raising the issue of more direct competition with other CompaqmK > offerings (Windows and Linux) that Compaq may quite understandably see asAN > addressing a somewhat different market than it (at least currently) wants to > target with VMS. >  > >hK > > Any mention of service cost in this is a red herring. Customers wantingrL > > service from Compaq will pay for it as an option. Other customers may beK > > quite happy with nothing apart from access to patches vis the Internet,eK > > and if they are forced to pay for support that they don't need it's yet 	 > another  > > reason that VMS loses out. > >9I > > So my suggestion would be that the combined cost of a software bundlepN > > including VMS, Cluster, Volshad, DFO, TCPIP, and a restricted-user license > for>K > > a programming language of choice should be set at a smallish percentage>* > > of the price of the hardware platform. > C > What other platforms bundle language support with a low-end OS?  t  L You can get C and Fortran for free on any Unix (GCC/G++/G77) and even on NT F (Cygwin/GCC etc) (there are a few strings attached to Cygwin, but it's cheap even if you buy it..  F I don't think there's an Alpha/VMS port of the GNU compilers. If thereD were, bundling them would be an OK alternative. (MAybe Compaq should" fund that port? - throwaway idea).   > If youN > can't make the case that this is standard practice, why should Compaq be theN > exception?  Unless the license you're suggesting above is for non-commercialF > use only, don't get greedy:  as long as Compaq's language pricing isF > reasonably comparable to other vendors', be happy - because Compaq's* > compilers are often noticeably superior.  K I said single-user to imply it was for non-intensive in-house use. DefiningEG noncommercial is tricky, but if it said that object code developed withnG this compiler license must either be distributed with sources for free,hI or not distributed outside the customer's sites at all, I'd say that was  
 fairly OK.   > I > This applies even more to bundling cluster support.  Compaq has alreadyiI > halved its cluster license price, and created a $1500 workstation-classeN > license option as well.  Who else bundles clustering support?  (Don't answerK > that Linux does:  its support isn't anything like VMS's, Tru64's, or evensJ > Sun's.  And don't say "But Linux *will* be offering real clustering RealC > Soon Now":  Compaq can cross that bridge if and when it's built.)5  J No: that's too late. That's when the competition gains an unbeatable priceJ advantage. Compaq should be using VMScluster to attack NOW, while they areL still ahead of the pack. And don't forget, in order for a cluster license toG be anything other than a piece of paper, you have to buy another set of  VMS hardware.   I The Workstation-class cluster license specifically denies you the abilityhE to construct a fault-tolerant cluster. It basically gets you what NFSo9 gets you with Unix, except you have to pay $1500 for it.     > I > If you want just the limited clustering that Win2K supports, you pay an5L > extra $2500 or more per node to get Advanced Server instead of just ServerI > (and clustering appears to constitute most of the functional difference L > between the two).  If you want clustering on Tru64, last I knew it cost anK > extra $3K/node.  I don't think Sun gives away its cluster option, either.oK > So be happy if you can get Compaq to reduce the per-node VMS full cluster"N > license price to something similar:  you'll be getting a bargain compared to > the alternatives.   I And why NOT make VMS a bargain? I'm not suggesting anything that involves H an immediate loss: you'll make profit on every unit of hardware shipped,I the manufacturing cost of software is basically nil so any money recievedaM for it is positive cashflow, and if the result is that VMS once again becomescI hot news and an industry trend-setter instead of "legacy", with a rapidlytJ growing number of entry-level sales, it's a handsome win. Should this not I work, the loss is small (since low-end VMS isn't a big seller anyway, theu2 main loss would be from cannibalisation effects).    > " >  I don't buy the cannibalisationL > > argument -- in most cases someone buying a bigger VMS platform needs the > IOL > > bandwidth it provides, which a small cluster can't -- but if the low-endI > > bundled cluster license was restricted to no more than three nodes, IS> > > don't think it would hurt much. (You need three to build a > disaster-tolerantn
 > > cluster).u > L > This assertion keeps coming up.  I suppose if you want a disaster-tolerantE > cluster that provides *completely* uninterrupted service, it may behL > technically true, but I'd assert that no entry-level customer has any real > need for this. > H > A two-node cluster that mirrors the data at both (separated) nodes canL > continue to run if the secondary node (the one with zero votes) fails, andM > can be recovered in minutes by reconfiguring the secondary node stand-aloneeL > if the primary node fails, can it not?  If so, wouldn't this be sufficient& > for any truly entry-level situation?  F true, but why errect a barrier that stops them from purchasing a thirdK set of VMS hardware if they like the thought of being disaster- rather thanCG fault-tolerant? And although the two-node setup can be reconfigured in wI seconds in the primary fails, it requires manual intervention, and eithereI an operator presence at both locations or some careful planning involving L modem-operated consoles, and hacker-proofing thereof, and ... The three-nodeE solution is capable of looking after everything that its safeguarding G against all by itself,  and so faults can be ignored until tomorrow. OraE even until next week. It might therefore be much preferred by a small.D organisation that doesn't want to have someone technically competentD around or on call 24x7. The third node reduces a crisis to a routine problem.   > B >  The system should also come with VMS-Perl and a webserver (bothL > > open-source, but Compaq should offer full support for these to those who
 > > want it).  > >oL > > And once you've got the pricing attractive from the very bottom upwards, > useeN > > VMScluster as the unifying concept and the one to base VMS advertising on.M > > Is there any other system where you can grow from a couple of entry-leveleH > > systems to a huge "mainframe" without ever having had an application	 > outage? N > > And isn't that *exactly* what a start-up e-business wants: easy, virtuallyM > > unlimited scalability and no visible shutdowns or breakdowns to annoy then > > customers? > H > This is certainly one pillar of the advertising campaign Compaq shouldI > create.  But if it's really a telling advantage, then customers will bebE > willing to pay something for it, as long as the cost is reasonable.   F Indeed ... once the telling is being done by lots of industry sources,F not just Compaq. To get there, they have to get a fair number of smallC VMS clusters into use. And once those customers start extolling itseD virtues, isn't it better to keep the lower price and increase volumeI (and thence the probability of more lugrative upgrades to bigger systems)lD than to choke off a possible renaissance by increasing prices again?   > N > The problem with low-end VMS is not primarily price:  reducing it would helpM > some *after* other issues are addressed, but without addressing those othergD > issues you might well *give* it away and still have few customers.  L Quite possibly. Giving low-end VMS away would be a more aggressive move thanF the fixed-percentage bundle I was proposing, but not necessarily a bad8 one. It would get the maximum possible number of low-endE VMS installations onto customer sites, and to first order there wouldaC still be profit in it (on the hardware and support-contract sales).tD Of course, against that there is a loss from cannibalisation of someI larger VMS sites, so one has to balance risk (of loss now) against reward  (more VMS sales in the future).    > F > The problem with low-end VMS is largely the same as the problem withJ > mid-range VMS:  it just doesn't fit all that well into currently-popularG > existing environments, so the barrier to entry in any low-end (hence,sL > usually 'commodity') situation where its unique strengths aren't perceivedB > as critical is huge.  Make VMS a platform approachable by users,H > developers,and support personnel as something familiar, which includesL > integration with their current environments in a manner that allows VMS toN > add value to them (without incurring significant down-sides in the process),L > and that barrier largely disappears - and allows VMS's strengths to shine.C > Only at that point does it make sense to consider radical low-endrH > penetration strategies such as parity pricing and IA hardware support.  G I'd invert that: push it NOW on the strengths that it has NOW (of which J I see VMScluster and the easy fault- and disaster-tolerance that gives youB as the greatest, and one of the easiest to advertise, and one thatG non-technical managers of e-businesses might be receptive to). Anything I involving changes to VMS will take time - at least a year, unless alreadyo@ in progress - by which time free LinuxClusters may be a reality.   > >eM > > VMS should be competing hard in the *server* market. At the low end, it's, > not. > G > Absolutely.  But, again, pricing alone won't help much as long as VMS M > servers lag in both performance and features compared to Windows (and oftenrG > Unix) servers (and remain 'foreign-looking' to both users and supporte	 > staff).y  J Sometimes this will be true. But a low-end low-cost VMScluster (much more L so than  a standalone VMS) offers something that the rest can't or don't do,I right now. So why not cut prices NOW, before someone else forces Compaq'syE hand. Why not for once be seen as an industry leader, not a reluctantsD follower?  (The obvious answer might be that Compaq is scared of theF wrath of Bill ... but now is the best possible time to risk that, withI Microsoft pinned under the legal spotlights for quite some time to come).w   > I > > Sometime, maybe soon, this window of opportunity will close, when the M > > competition finally catches up to where VMScluster was more than a decadet > ago.I > > If VMS is still seen as "legacy" then, it won't get any more chances.m > L > I agree that the window may well close, and it will almost certainly closeL > soon if VMS both fails to take major steps toward real integration (in theL > areas mentioned above) *and* fails to take significant steps to *maintain*G > its technical leadership.  And while addressing only one of those twoiE > problems might be sufficient for VMS to retain its current industryeL > standing, I suspect both will have to be addressed if it is to have a real
 > revival.   I won't disagree there.v   	Yours,>
 		Nigel Arnotl- 		NRA@MAXWELL.PH.KCL.AC.UK                      7 		"In the beginning there was nothing, which exploded."l   ------------------------------    Date: 22 Jun 2000 11:00:03 -05009 From: Kilgallen@eisner.decus.org.nospam (Larry Kilgallen)p Subject: Re: VAX on Intel?+ Message-ID: <shV1xmHnnFE1@eisner.decus.org>s  i In article <009EBFD5.D34CF1A0.20@maxwell.ph.kcl.ac.uk>, Nigel Arnot <sysmgr@maxwell.ph.kcl.ac.uk> writes:e  H > I don't think there's an Alpha/VMS port of the GNU compilers. If thereF > were, bundling them would be an OK alternative. (MAybe Compaq should$ > fund that port? - throwaway idea).  @ DEC funded the port of GNAT (the GNU Ada compiler) to Alpha VMS.@ As part of that deal, the result had to support certain features= used in the older DEC Ada (Ada 83) product for calling systemk< services, etc.  The result is that GNAT on all platforms now  has those new (to GNU) features.   ------------------------------  % Date: Thu, 22 Jun 2000 17:35:15 +0200 = From: Arne =?iso-8859-1?Q?Vajh=F8j?= <arne.vajhoej@gtech.com>k Subject: Re: VAX on Intel?) Message-ID: <39523233.5AB3DA85@gtech.com>a   Nigel Arnot wrote:# > > 1.  Software development costs.  > Q > I'd say that all the parts of VMS most relevant to an entry-level customer were M > developed years ago (1984-90 timescale, apart from the Alpha porting). Mostl> > of the new stuff is aimed more toward the big-iron hardware.  G Who pays for the software engineers and QA engineers working on VMS nowo ?k  9 Just testing of whether old features in a new VMS versionl: works with a zillion combinations of hardware and software is a huge task.    > > 3.  Software cost of sales.s >  > Don't understand this one.  G We always point fingers at Compaq because Compaq sales people forget toz; mention VMS and because Compaq does not put out massive VMSv: adds. But if we want them to do those things, then we must9 also realize that there are only one group to pay for it:  us the VMS customers.    Arne   ------------------------------  % Date: Thu, 22 Jun 2000 21:27:36 +0930r% From: Jeremy Begg <jeremy@vsm.com.au>v Subject: Re: Verify of Backups* Message-ID: <3951FF30.7136CEA4@vsm.com.au>  H For a novel way to think about your backup strategies, visit "The Tao of- Backup" at http://www.rocksoft.com/taobackup/            Jeremy Beggi  ;   +-------------------------------------------------------+c;   |            VSM Software Services Pty. Ltd.            | ;   |                 http://www.vsm.com.au/                |e;   |       "OpenVMS Systems Management & Programming"      | ;   |-------------------------------------------------------|i;   | P.O.Box 801, Unley,     |  E-Mail:  jeremy@vsm.com.au |e;   | South Australia 5061    |   Phone:  +61 8 83592155    |t;   |-------------------------|  Mobile:  0414 422 947      |f;   |  A.C.N. 068 409 156     |     FAX:  +61 8 82231777    |e;   +-------------------------------------------------------+i   ------------------------------  % Date: Thu, 22 Jun 2000 01:12:04 -0700p1 From: Vance Haemmerle <vance@toyvax.Tucson.AZ.US> - Subject: Re: VMS Evangelism:  Modest Proposal 3 Message-ID: <395167E4.2B96E715@toyvax.Tucson.AZ.US>h   Terry C. Shannon wrote:c > H > Ya know, those of us who support OpenVMS would probably get further byN > purchasing one (or maybe ten) shares of CPQ stock than by ranting on Usenet.N > As a stockholder, you are entitled to attend annual meetings and such. These/ > make great venues for getting yourself heard.e > M > What's more, the Q might place more weight on a letter or suggestion from a J > stockholder than a letter from a non-stockholder (who's to know that you > only have 10 shares?). > 	 > cheers,  > 	 > terry st= > Who used to have 9 shares, but now has a hell of a lot more   I   I recently sold my 1100 shares.  I still believe in the products but noeL longer in the management of the company.  I've found better places to put my money.   -- Vance Haemmerle  vance@toyvax.Tucson.AZ.USI   ------------------------------  % Date: Thu, 22 Jun 2000 11:23:50 +0100e1 From: "Rob Murphy" <rob@stonecoldsolutions.co.uk>c1 Subject: VMS systems manager/programmer availabler# Message-ID: <8ispk7$j40$1@gxsn.com>i  4 Is anyone out there looking for a decent VMS systems; manager/programmer/engineer with fifteen years experience ?   D I have worked for a multitude of companies, including Reuters, GlaxoL Wellcome, Coopers & Lybrand and Ford. I have been working with VMS (both VAXJ and Alpha) since 1986, and aside from doing systems & network management I. also program in Pascal, C, Fortran and others.  J I also have lots of Unix & NT experience, as well as VB and Delphi coding.  D I am based in Essex, in England, but I am more than happy to travel.  I If anyone out there has a position open, please drop me an email and I'lle
 forward a CV.    Thanks,a  
 Rob Murphy rob@stonecoldsolutions.co.uk# http://www.stonecoldsolutions.co.ukm   ------------------------------  % Date: Thu, 22 Jun 2000 18:00:44 +0100   From: steven.reece@quintiles.com! Subject: Win2K on Alpha resumed??e> Message-ID: <80256906.005D90A7.00@qedilc01.qedi.quintiles.com>  L There's an article posted on http://www.aceshardwaare.com/ commenting on the* resumption of service on Windows on Alpha.2 No idea if it's true or not, it could just be FUD.   Steve.   ------------------------------   Date: 22 Jun 2000 17:44:42 GMT2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)% Subject: Re: Win2K on Alpha resumed??x6 Message-ID: <8itjaa$6s3$2@mailint03.im.hou.compaq.com>  a In article <80256906.005D90A7.00@qedilc01.qedi.quintiles.com>, steven.reece@quintiles.com writes: M :There's an article posted on http://www.aceshardwaare.com/ commenting on the + :resumption of service on Windows on Alpha.p3 :No idea if it's true or not, it could just be FUD.t     False.  N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------   End of INFO-VAX 2000.347 ************************including VMS, Cluster, Volshad, DFO, TCPIP, and a restricted-user license > for>K > > a programming language of choice should be set at a smallish percentage>* > > of the price of the hardware platform. > C > What other platforms bundle language support with a low-end OS?  t  L You can get C and Fortran for free on any Unix (GCC/G++/G77) and even onpleted.  <<< CWD /lt95b/antivirus9 >>> 250 Connected to /disk$misc/decus/lt95b/antivirus/.A <<< TYPE A >>> 200 Type A ok. <<< PORT 195,113,19,90,84,11955 >>> 200 Port 84.119 at Host 195.113.19.90 accepted.d
 <<< LIST >>> 150 List started. >>> 226 Transfer completed.c <<< CWD /lt95b/antivirus/pcT< >>> 250 Connected to /disk$misc/decus/lt95b/antivirus/pc/. <<< TYPE A >>> 200 Type A ok. <<< PORT 195,113,19,90,85,117<5 >>> 200 Port 85.117 at Host 195.113.19.90 accepted.m
 <<< LIST >>> 150 List started.  >>> 226 Transfer completed.s <<< CWD /lt95b/linuxaxp 8 >>> 250 Connected to /disk$misc/decus/lt95b/linuxaxp/. <<< TYPE A >>> 200 Type A ok. <<< PORT 195,113,19,90,88,474 >>> 200 Port 88.47 at Host 195.113.19.90 accepted.
 <<< LIST >>> 150 List started.v >>> 226 Transfer completed.o <<< CWD /lt95b/netsrcn6 >>> 250 Connected to /disk$misc/decus/lt95b/netsrc/. <<< TYPE A >>> 200 Type A ok. <<< PORT 195,113,19,90,88,25195 >>> 200 Port 88.251 at Host 195.113.19.90 accepted.>
 <<< LIST >>> 150 List started.  >>> 226 Transfer completed.n <<< CWD /lt95b/plot_routine/< >>> 250 Connected to /disk$misc/decus/lt95b/plot_routine/. <<< TYPE A >>> 200 Type A ok. <<< PORT 195,113,19,90,91,215.5 >>> 200 Port 91.215 at Host 195.113.19.90 accepted.f
 <<< LIST >>> 150 List started.r >>> 226 Transfer completed.k <<< CWD /lt95b/texmf5 >>> 250 Connected to /disk$misc/decus/lt95b/texmf/.1 <<< TYPE A >>> 200 Type A ok. <<< PORT 195,113,19,90,97,193d5 >>> 200 Port 97.193 at Host 195.113.19.90 accepted.n
 <<< LIST >>> 150 List started.o >>> 226 Transfer completed.  <<< CWD /lt95b/tktcl5 >>> 250 Connected to /disk$misc/decus/lt95b/tktcl/. <<< TYPE A >>> 200 Type A ok. <<< PORT 195,113,19,90,101,24.5 >>> 200 Port 101.24 at Host 195.113.19.90 accepted. 
 <<< LIST >>> 150 List started.  >>> 226 Transfer completed.n <<< CWD /lt95b/tools5 >>> 250 Connected to /disk$misc/decus/lt95b/tools/. <<< TYPE A >>> 200 Type A ok.  <<< PORT 195,113,19,90,103,1216 >>> 200 Port 103.121 at Host 195.113.19.90 accepted.
 <<< LIST >>> 150 List started.  >>> 226 Transfer completed.n
 <<< CWD /mnm, >>> 250 Connected to /disk$misc/decus/mn/. <<< TYPE A >>> 200 Type A ok. <<< PORT 195,113,19,90,105,8825 >>> 200 Port 105.88 at Host 195.113.19.90 accepted.5
 <<< LIST >>> 150 List started.c >>> 226 Transfer completed.o <<< CWD /mn/netc0 >>> 250 Connected to /disk$misc/decus/mn/net/. <<< TYPE A >>> 200 Type A ok. <<< PORT 195,113,19,90,109,16 5 >>> 200 Port 109.16 at Host 195.113.19.90 accepted.T
 <<< LIST >>> 150 List started.  >>> 226 Transfer completed.  <<< CWD /mn/net/cr3 >>> 250 Connected to /disk$misc/decus/mn/net/cr/.  <<< TYPE A >>> 200 Type A ok.  <<< PORT 195,113,19,90,112,2186 >>> 200 Port 112.218 at Host 195.113.19.90 accepted.
 <<< LIST >>> 150 List started.l >>> 226 Transfer completed.> <<< CWD /mn/perl