1 INFO-VAX	Sat, 28 May 2005	Volume 2005 : Issue 295       Contents: Re: ABC problem with disk  Re: ABC problem with disk  Re: ABC problem with disk  Re: Byte Range Locking  (Bless)  Re: DEFRAG ovms 7.3-2  Re: DEFRAG ovms 7.3-2   RE: Encompass Board of Directors  Re: Encompass Board of Directors4 Re: HP Pulls Off a Respectable Second Fiscal Quarter? Re: HP to help governments with tracking users with NAtional ID ? Re: HP to help governments with tracking users with NAtional ID ? Re: HP to help governments with tracking users with NAtional ID ? Re: HP to help governments with tracking users with NAtional ID ? Re: HP to help governments with tracking users with NAtional ID ? Re: HP to help governments with tracking users with NAtional ID ? Re: HP to help governments with tracking users with NAtional ID ? Re: HP to help governments with tracking users with NAtional ID ? Re: HP to help governments with tracking users with NAtional ID ? Re: HP to help governments with tracking users with NAtional ID ? Re: HP to help governments with tracking users with NAtional ID ? Re: HP to help governments with tracking users with NAtional ID ? Re: HP to help governments with tracking users with NAtional ID  Re: Installing a printer Re: Installing a printer Re: Installing a printer# largest DEC collection  yada yada..  MultiHead issue on AS400 Re: New Cluster Interconnects  Re: New Cluster Interconnects  Re: New Cluster Interconnects ( Re: OpenVMS Advanced Technical boot camp- Re: QIOs in multi thread environment (Pascal) - Re: QIOs in multi thread environment (Pascal) % Re: VMS API (possible unix emulation) % Re: VMS API (possible unix emulation) % Re: VMS API (possible unix emulation)  Re: VMS TCPIP Services Re: VMS TCPIP Services  F ----------------------------------------------------------------------  % Date: Fri, 27 May 2005 21:10:08 -0500 2 From: David J Dachtera <djesys.nospam@comcast.net>" Subject: Re: ABC problem with disk+ Message-ID: <4297D2FF.ED7CC719@comcast.net>    Joe Sewell wrote:  >  > Paul Sture wrote:  > > Joe Sewell wrote:  > > H > > > The complete backup command is generated from DCL code, so I'm not? > > > precisely sure what it is. I believe it's something line:  > > >  > > > $ ABC INCREMENTAL K > > > $1$DKB100:/IGNORE=INTERLOCK/LOG=(something)/SUMMARY=(something) ! I'd 3 > > > have to figure out what the "something's" are  > > > E > > > The logs don't show anything except that it stops at that disk.  > > >  > > L > > Just a gut feeling response, but how does ABC actually execute the jobs? > > I > > Does ABC use spawned, detached or batch jobs? If so I'd be looking to # > > see if the quotas are adequate.  > G > It's running within a batch job, although the utility itself executes 3 > directly, I think.  I'll double-check the quotas.   # Check while the process is running:   " $ SHOW PROCESS/SUBPROC/ID=nnnnnnnn   --   David J Dachtera dba DJE Systems  http://www.djesys.com/  ) Unofficial OpenVMS Hobbyist Support Page: " http://www.djesys.com/vms/support/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/   " Unofficial OpenVMS-IA32 Home Page: http://www.djesys.com/vms/ia32/    Coming soon:& Unofficial OpenVMS Marketing Home Page   ------------------------------  % Date: Fri, 27 May 2005 21:15:10 -0500 2 From: David J Dachtera <djesys.nospam@comcast.net>" Subject: Re: ABC problem with disk+ Message-ID: <4297D42E.4ED904C0@comcast.net>    Joe Sewell wrote:  >  > David J Dachtera wrote:  > > Joe Sewell wrote:  > > > O > > > >Archive backup client, eh? Haven't heard of that since about 1998 or so. $ > > > >Not recommended then, either. > > > > O > > > >Use VMS BACKUP to backup VMS disks, or risk losing your data and/or your 8 > > > >ability to perform disaster recovery expediently. > > >  > > > Not an option. > > L > > Protecting your company's data by doing proper backups is not an option? > >  > > That's *GOTTA* be a first! > A > Oh, they do "proper backups" to the shared computer resources.    @ If it's not VMS BACKUP, it's not "proper" (or "supported"). Your company's data is at risk.  G ...but don't take my word for it. contact your OpenVSM support channels & and solicit their "official" position.   > The F > Alpha workstations used for development purposes, however, aren't asF > important, it seems; the standard is what we find in the labs, whereE > the system disk is routinely INITIALIZEd for a VMS upgrade.  (We're D > talking about hundreds of machines, and nobody has the time to run* > through the standard upgrade procedure.)  A The install and upgrade both take the same amount of time (unless - someone is judging by an out-dated standard).   > > > > This was installed by the system administrator (not me).J > > > The only reason I even noticed its existence on my machine is due toM > > > some system slowdown that I tracked down to this thing chewing up about ! > > > 5% CPU time in kernel mode.  > > D > > ...and the reason management continues to tolerate this is ... ? > H > ... because "management" doesn't know about it.  This is all under the? > auspices of IT, which is under a separate sector, which makes . > accountability more than a little difficult. > G > But enough of this.  ABC is what we're stuck with for now.  Help, not G > criticism of policies that are out of my hands, would be appreciated.   0 Your data is at risk. Period - end of statement.  H If this ABC thing is malfunctioning, it hardly qualifies as either safe,A proper or supported (by OpenVMS Engineering). If the vendor can't C support it either, well, does the word "pray" mean anything to you?    --   David J Dachtera dba DJE Systems  http://www.djesys.com/  ) Unofficial OpenVMS Hobbyist Support Page: " http://www.djesys.com/vms/support/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/   " Unofficial OpenVMS-IA32 Home Page: http://www.djesys.com/vms/ia32/    Coming soon:& Unofficial OpenVMS Marketing Home Page   ------------------------------  % Date: Fri, 27 May 2005 21:16:01 -0500 2 From: David J Dachtera <djesys.nospam@comcast.net>" Subject: Re: ABC problem with disk+ Message-ID: <4297D460.34A25E13@comcast.net>    William Webb wrote:  > I > On 27 May 2005 05:29:30 -0700, Joe Sewell <ultrajoe@spamcop.net> wrote:  > > David J Dachtera wrote:  > > > Joe Sewell wrote:  > > > > Q > > > > >Archive backup client, eh? Haven't heard of that since about 1998 or so. & > > > > >Not recommended then, either.	 > > > > > Q > > > > >Use VMS BACKUP to backup VMS disks, or risk losing your data and/or your : > > > > >ability to perform disaster recovery expediently. > > > >  > > > > Not an option. > > > N > > > Protecting your company's data by doing proper backups is not an option? > > >   > > > That's *GOTTA* be a first! > > G > > Oh, they do "proper backups" to the shared computer resources.  The H > > Alpha workstations used for development purposes, however, aren't asH > > important, it seems; the standard is what we find in the labs, whereG > > the system disk is routinely INITIALIZEd for a VMS upgrade.  (We're F > > talking about hundreds of machines, and nobody has the time to run, > > through the standard upgrade procedure.) > > @ > > > > This was installed by the system administrator (not me).L > > > > The only reason I even noticed its existence on my machine is due toO > > > > some system slowdown that I tracked down to this thing chewing up about # > > > > 5% CPU time in kernel mode.  > > > F > > > ...and the reason management continues to tolerate this is ... ? > > J > > ... because "management" doesn't know about it.  This is all under theA > > auspices of IT, which is under a separate sector, which makes 0 > > accountability more than a little difficult. > > I > > But enough of this.  ABC is what we're stuck with for now.  Help, not I > > criticism of policies that are out of my hands, would be appreciated.  > >  > >  > ! > This one might require courage:  > # > Suggest a disaster recovery test.  > E > Any good shop will have a post-mortem to evaluate what worked, what 4 > didn't and what can be done about them wot didn't. > = > If you don't think it'd be a career-limiting move, that is.   5 Depends how badly his boss handles self-embarassment.    --   David J Dachtera dba DJE Systems  http://www.djesys.com/  ) Unofficial OpenVMS Hobbyist Support Page: " http://www.djesys.com/vms/support/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/   " Unofficial OpenVMS-IA32 Home Page: http://www.djesys.com/vms/ia32/    Coming soon:& Unofficial OpenVMS Marketing Home Page   ------------------------------  # Date: Fri, 27 May 2005 22:03:26 GMT 1 From: Keith Parris <keithparris_NOSPAM@yahoo.com> ( Subject: Re: Byte Range Locking  (Bless)2 Message-ID: <OOMle.6178$lx5.4921@news.cpqcorp.net>   Richard Maher wrote:@ > can anyone confirm or deny that the only reason the Lock ValueN > Block was extended was to support UNIX satanic rituals on VMS? (Specifically > "Byte Range Locking")   G As Larry pointed out, byte range locking was first introduced long ago  A in support of Posix. I don't see any relationship between recent  F improvements in this area and the support for larger lock value block I size. Like longer DCL command lines and symbols, this was a feature long   requested by customers.    ------------------------------    Date: 27 May 2005 16:45:12 -0700 From: JimStrehlow@data911.com  Subject: Re: DEFRAG ovms 7.3-2C Message-ID: <1117237511.867719.206670@g47g2000cwa.googlegroups.com>   @ My manual approach to watching and monitoring fragmentation are:  * $!Look for contiguous available disk space $ MCR DFU  REPORT   disk:   +  Largest free extent              :  227644       0 $!See how badly fragmented a particular file is:' $ DEFINE/PROCESS  SYS$OUTPUT  TEMP.TEXT $ $ DUMP/HEADER/RECORD=END:0  filespec $ DEASSIGN/PROCESS SYS$OUTPUT ' $ SEARCH  TEMP.TEXT  "Dump of", "count" " $ DELETE/NOCONFIRM/LOG  TEMP.TEXT;       !To defragment a specific file: ( $ COPY/LOG  fileIn   fileOut /CONTIGUOUS    + $!If you can not copy/contiguous, then use: 4 $ MCR DFU  DEFRAGMENT /BESTTRY /STATISTICS  filename    ? Jim Strehlow, Data911 OpenVMS Systems Manager, Alameda, CA, USA    ------------------------------  % Date: Fri, 27 May 2005 21:20:42 -0500 2 From: David J Dachtera <djesys.nospam@comcast.net> Subject: Re: DEFRAG ovms 7.3-2+ Message-ID: <4297D57A.F491731E@comcast.net>    Nigel Barker wrote:  > R > On Thu, 26 May 2005 21:05:55 -0500, David J Dachtera <djesys.nospam@comcast.net> > wrote: > J > >Well, before you go doing that, first verify that disk fragmentation isF > >causing a performance problem. Otherwise, you'll be showing them "a# > >solution looking for a problem".  > M > The simplest way to decide whether disk fragmentation is actually causing a R > performance problem is to do a BACKUP/IMAGE of the disk in question then restoreD > that backup to the same disk. This will fully defragment the disk.  7 How does this help him analyze his present performance?   A He needs to know what he has before he defrags and after, and the B measured performance of both. Then, and only then, will he know if> fragmentation is an issue and worthy of being pursued further.  F The MONITOR utility is helpful for measuring such things as split I/OsC if he has that problem, as well as window turns, I/O queues and the 0 like. That's where I would start, if I were him.   --   David J Dachtera dba DJE Systems  http://www.djesys.com/  ) Unofficial OpenVMS Hobbyist Support Page: " http://www.djesys.com/vms/support/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/   " Unofficial OpenVMS-IA32 Home Page: http://www.djesys.com/vms/ia32/    Coming soon:& Unofficial OpenVMS Marketing Home Page   ------------------------------  % Date: Fri, 27 May 2005 13:39:47 -0500 / From: "Stuart, Ed" <Ed.Stuart@austinenergy.com> ) Subject: RE: Encompass Board of Directors T Message-ID: <DA4AD590CAF06845B671C398333A89C60A88D692@ohms.electric.ci.austin.tx.us>  K Then why is Interex doing their own event, while the HP Technology Forum is F being hosted by Encompass with input from ITUG and the OpenView group?? Granted it is an Encompass site, but browse around the info at: D <http://www.encompassus.org/events/annualconfintropage.html> and payL attention to the number of sessions and labs being offered at the Technology Forum vs. World.     EdE **Please apply a generous amount of all the usual disclaimers here.**      > -----Original Message-----: > From: Greg Cagle [mailto:gregc@removethisgregcagle.com] % > Sent: Friday, May 27, 2005 12:34 PM  > To: Info-VAX@Mvb.Saic.Com + > Subject: Re: Encompass Board of Directors  >  > Stuart, Ed wrote: D > > From what I have seen the membership and focus of Encompass and C > > different than that of Interex.  Encompass is composed of more  F > > technical front line folks while Interex appears to be more upper > > > level.  Encompass also promotes the exchange of ideas and  > information : > > at the membership level and encourages its members to 3 > present their solutions at the yearly conference.  > >  > > Ed > 3 > I don't think the membership profiles of the two  = > organizations are all that different IMHO. The majority of  ; > people who attend HP World are system administrators and  = > infrastructure managers, and the conference has always had  9 > strong participation from the membership at all levels.  > -- > Greg Cagle > gregc at gregcagle dot com >    ------------------------------  % Date: Fri, 27 May 2005 11:53:51 -0700 0 From: Greg Cagle <gregc@removethisgregcagle.com>) Subject: Re: Encompass Board of Directors 0 Message-ID: <119er64rteskr8a@corp.supernews.com>  B I don't think that's relevant to the discussion; the topic at handA is about the technical content and member participation levels of E the two conferences. There aren't a lot of specifics on the Encompass F site about the nature and level of the sessions, although the quantity is quite large.   C I've got experience with both organizations - all I'm trying to say > is that I don't think the nature of the organizations or theirC conferences is that dissimilar, and I encourage people to go to the 0 respective sites and draw their own conclusions.   - Greg   Stuart, Ed wrote: M > Then why is Interex doing their own event, while the HP Technology Forum is H > being hosted by Encompass with input from ITUG and the OpenView group?A > Granted it is an Encompass site, but browse around the info at: F > <http://www.encompassus.org/events/annualconfintropage.html> and payN > attention to the number of sessions and labs being offered at the Technology > Forum vs. World.   >  > EdG > **Please apply a generous amount of all the usual disclaimers here.**  >  >  >  >>-----Original Message-----: >>From: Greg Cagle [mailto:gregc@removethisgregcagle.com] % >>Sent: Friday, May 27, 2005 12:34 PM  >>To: Info-VAX@Mvb.Saic.Com + >>Subject: Re: Encompass Board of Directors  >> >>Stuart, Ed wrote:  >>C >>>From what I have seen the membership and focus of Encompass and  B >>>different than that of Interex.  Encompass is composed of more E >>>technical front line folks while Interex appears to be more upper  = >>>level.  Encompass also promotes the exchange of ideas and   >> >>information  >>9 >>>at the membership level and encourages its members to   >>3 >>present their solutions at the yearly conference.  >> >>>Ed  >>3 >>I don't think the membership profiles of the two  = >>organizations are all that different IMHO. The majority of  ; >>people who attend HP World are system administrators and  = >>infrastructure managers, and the conference has always had  9 >>strong participation from the membership at all levels.  >>-- >>Greg Cagle >>gregc at gregcagle dot com >>   --  
 Greg Cagle gregc at gregcagle dot com   ------------------------------  % Date: Fri, 27 May 2005 15:08:49 -0400 ( From: Bill Todd <billtodd@metrocast.net>= Subject: Re: HP Pulls Off a Respectable Second Fiscal Quarter = Message-ID: <PLadnXX6d-nf7QrfRVn-ow@metrocastcablevision.com>    Dave Froble wrote: > Bill Todd wrote: >  >> Fuck off, Fred. >  > J > I can only take that as an indicator that there isn't a valid rebuttal, = > and therefore an admission that Fred scores one point.  :-)   G Hmmm.  I guess that would make your interpretation just as incompetent  
 as Fred's.  . The morons seem to be out in force these days.   - bill   ------------------------------   Date: 27 May 2005 18:45:26 GMT$ From: "Doc." <doc@openvms-rocks.com>H Subject: Re: HP to help governments with tracking users with NAtional ID7 Message-ID: <Xns9663D33D4108Adcovmsrox@212.100.160.126>   % %NEWS-I-NEWMSG, William Webb wrote in / news:8660a3a105052709582e5c5cc7@mail.gmail.com    H > On 27 May 2005 11:20:42 -0500, Larry Kilgallen <Kilgallen@spamcop.net>	 > wrote:  G >> In article <8660a3a1050527044246cad97d@mail.gmail.com>, William Webb  >> <wil   > liam.w.webb@gmail.com> writes: >>  A >> > They'll have to come up with a different acronym for it, for  >> > multiple r 
 > easo> > ns:  >> >G >> > They already used the -nis suffix for a networking product back in  >> > the days of d|i|g|i|t|a|l >>  B >> It is so sad that you young people do not remember LinkWorks or1 >> DECset (either incarnation of either name) :-)  >>   >  > G > I'm old enough to have punched cards for computer class and have even  > seen core memory in use.  G I'm not that old, I remember using blank punch-cards as a note-passing  6 mechanism that was replaced by the infamous post-it.    J The funny thing was that the local paper for 3M HQ tried to make out that J the post-it predated email.  They did admit that most people couldn't see H the point to post-it notes until some evangelization was gone through.  I Seems to me the same was the case with email, but as it spread to a less  9 technical community the lessons about security were lost.      Doc. --  G OpenVMS:     Eight out of ten hackers prefer *other* operating systems. G http://www.openvms-rocks.com    Deathrow Public-Access OpenVMS Cluster.    ------------------------------   Date: 27 May 2005 19:23:00 GMT( From: bill@cs.uofs.edu (Bill Gunshannon)H Subject: Re: HP to help governments with tracking users with NAtional ID+ Message-ID: <3fpacjF8vpsgU1@individual.net>   7 In article <Xns9663D33D4108Adcovmsrox@212.100.160.126>, ' 	"Doc." <doc@openvms-rocks.com> writes: ' > %NEWS-I-NEWMSG, William Webb wrote in 1 > news:8660a3a105052709582e5c5cc7@mail.gmail.com   > I >> On 27 May 2005 11:20:42 -0500, Larry Kilgallen <Kilgallen@spamcop.net> 
 >> wrote: H >>> In article <8660a3a1050527044246cad97d@mail.gmail.com>, William Webb	 >>> <wil  ! >> liam.w.webb@gmail.com> writes:  >>> B >>> > They'll have to come up with a different acronym for it, for >>> > multiple r   >> easo> > ns: >>> > H >>> > They already used the -nis suffix for a networking product back in >>> > the days of d|i|g|i|t|a|l  >>> C >>> It is so sad that you young people do not remember LinkWorks or 2 >>> DECset (either incarnation of either name) :-) >>>  >>   >>  H >> I'm old enough to have punched cards for computer class and have even >> seen core memory in use.  > I > I'm not that old, I remember using blank punch-cards as a note-passing  8 > mechanism that was replaced by the infamous post-it.   > L > The funny thing was that the local paper for 3M HQ tried to make out that L > the post-it predated email.  They did admit that most people couldn't see J > the point to post-it notes until some evangelization was gone through.  / > Seems to me the same was the case with email,   G Are you saying email needed evangelization to succeed?  Some of us have G been doing Email since before there was an INTERNET and everyone I knew G in those days immediately saw the utility in the system compared to the 
 alternatives.   K >                                               but as it spread to a less  ; > technical community the lessons about security were lost.   H The day NFS let the INTERNET go was a dark day indeed.  While the amountH of available information has grown exponentially the actual value of the< net in general is going down almost as fast (if not faster!)   bill   --  J Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>       ------------------------------   Date: 27 May 2005 19:37:28 GMT$ From: "Doc." <doc@openvms-rocks.com>H Subject: Re: HP to help governments with tracking users with NAtional ID7 Message-ID: <Xns9663DC1879DC6dcovmsrox@212.100.160.126>   ( %NEWS-I-NEWMSG, Bill Gunshannon wrote in# news:3fpacjF8vpsgU1@individual.net    9 > In article <Xns9663D33D4108Adcovmsrox@212.100.160.126>, - >      "Doc." <doc@openvms-rocks.com> writes:   < >> I'm not that old, I remember using blank punch-cards as aF >> note-passing mechanism that was replaced by the infamous post-it.   >>  G >> The funny thing was that the local paper for 3M HQ tried to make out D >> that the post-it predated email.  They did admit that most peopleH >> couldn't see the point to post-it notes until some evangelization was? >> gone through.  Seems to me the same was the case with email,  > D > Are you saying email needed evangelization to succeed?  Some of usG > have been doing Email since before there was an INTERNET and everyone @ > I knew in those days immediately saw the utility in the system > compared to the alternatives.   H No.  Email didn't need any evangelists.  The computer delivered it just 4 fine without someone turning up on the doorstep. :-)  I I was referring to the adoption of it by the non-technical community and  H their perception that post-it notes predated email.  In the process (of F them adopting it) all the lessons that were learned about things like D executable content were thrown out the window to cater to a certain  segment of the population.   <snip>  C > The day NFS let the INTERNET go was a dark day indeed.  While the D > amount of available information has grown exponentially the actualB > value of the net in general is going down almost as fast (if not > faster!)    A My experience probably doesn't go back far enough, but I get the  J impression there's always been a supply of pranksters and the like on the G net.  Its opening up allowed in the Snake Oil salesmen and their ilk.   H Now governments are panicking about their presence and some of the good ? attributes of the net could be damaged by heavy-handed actions.      Doc. --  G OpenVMS:     Eight out of ten hackers prefer *other* operating systems. G http://www.openvms-rocks.com    Deathrow Public-Access OpenVMS Cluster.    ------------------------------  % Date: Fri, 27 May 2005 21:48:01 +0200 & From: Paul Sture <paul.sture@decus.ch>H Subject: Re: HP to help governments with tracking users with NAtional ID+ Message-ID: <3fpbqqF8vc5tU1@individual.net>    William Webb wrote:   O > On 27 May 2005 11:20:42 -0500, Larry Kilgallen <Kilgallen@spamcop.net> wrote:  > i >>In article <8660a3a1050527044246cad97d@mail.gmail.com>, William Webb <william.w.webb@gmail.com> writes:  >> >>O >>>They'll have to come up with a different acronym for it, for multiple reaso=  >>>ns: >>> I >>>They already used the -nis suffix for a networking product back in the  >>>days of d|i|g|i|t|a|l >>H >>It is so sad that you young people do not remember LinkWorks or DECset) >>(either incarnation of either name) :-)  >> >  >  > G > I'm old enough to have punched cards for computer class and have even  > seen core memory in use. >  > : ^ )   8 Hah! I'm old enough to have used punched cards at work.*  G DECset I know, but I don't remember LinkWorks (though the SPD contains  = references to such things as TeamWorks, which I do remember).   E * Coincidentally, I just hit a litter bin from the other side of the  I room with a piece of crumpled paper today, and was asked if I had played  G basketball when younger. My answer was "No. My real training came from   using punched cards".    ------------------------------  % Date: Fri, 27 May 2005 16:22:18 -0400 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> H Subject: Re: HP to help governments with tracking users with NAtional ID, Message-ID: <4297817A.B06F19AE@teksavvy.com>   William Webb wrote:   F > I've got some of those pixellated cat "AlphaPowered" graphics if youE > (or anybody else) is interested.  I'd like to see someone turn them  > into X bitmaps.     5 If it is big enough, then by all means send it to me.    ------------------------------  % Date: Fri, 27 May 2005 16:51:21 -0400 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> H Subject: Re: HP to help governments with tracking users with NAtional ID, Message-ID: <42978847.9AE33D9F@teksavvy.com>   Bill Gunshannon wrote:I > Are you saying email needed evangelization to succeed?  Some of us have I > been doing Email since before there was an INTERNET and everyone I knew I > in those days immediately saw the utility in the system compared to the  > alternatives.   H Lets not forget ALLIN1 which was the largest installed-seat email system1 in the world and which startd in the early 1980s.   E However, communications with the outside world were very basic before C the internet became widely available, often through clunky X25 line  editor interfaces.  E I managed a group of bankers for about 15 years with email. Despite a E clunky interface, the service was a success because they had specific ? information to exchange. So there was enough traffic to make it : worthwile for the secretaries to check the email each day.  E Anopther group I helped get started wasn't so lucky. There were a vew C visionaries who saw the real need for information exchange. But the E members of that organisations didn't really have anything specific to H send, and with little traffic, there was veryt little incentive to checkG the emails  regularly, and because members didn't check regularly, high C value information didn't have confidence level of being received in * time, so members still used other methods.    G What made EMAIL really work is when corporations hoooked their internal F email systems to the internet so that staff could exchange emails withF anyone else, and more importantly, have emails delivered to their deskG instead of them having to use a clunky interface to access and external 
 email system.    ------------------------------    Date: 27 May 2005 16:01:52 -0500; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler) H Subject: Re: HP to help governments with tracking users with NAtional ID3 Message-ID: <OeluntS4NXl4@eisner.encompasserve.org>   G > I'm old enough to have punched cards for computer class and have even  > seen core memory in use.  G    You don't have to be that old.  We're still using it.  It's very rad     hard.   ------------------------------  % Date: Sat, 28 May 2005 07:00:42 +0930 * From: Mark Daniel <mark.daniel@vsm.com.au>H Subject: Re: HP to help governments with tracking users with NAtional ID- Message-ID: <429791c3@duster.adelaide.on.net>    Bob Koehler wrote:G >>I'm old enough to have punched cards for computer class and have even  >>seen core memory in use. >  > I >    You don't have to be that old.  We're still using it.  It's very rad  >    hard.    7 I hope you are using an equally hard computing machine.   ; The Hollerith electro-mechanical Tabulator springs to mind!   F +--------------------------------------------------------------------+E   Mark Daniel                         http://wasd.vsm.com.au/adelaide F   mailto:Mark.Daniel@wasd.vsm.com.au (Mark.Daniel@dsto.defence.gov.au);   A pox on the houses of all SPAMers.  Make that two poxes. F +--------------------------------------------------------------------+   ------------------------------   Date: 27 May 2005 22:00:56 GMT( From: bill@cs.uofs.edu (Bill Gunshannon)H Subject: Re: HP to help governments with tracking users with NAtional ID+ Message-ID: <3fpjkoF90q5gU1@individual.net>   , In article <42978847.9AE33D9F@teksavvy.com>,0 	JF Mezei <jfmezei.spamnot@teksavvy.com> writes: > Bill Gunshannon wrote:J >> Are you saying email needed evangelization to succeed?  Some of us haveJ >> been doing Email since before there was an INTERNET and everyone I knewJ >> in those days immediately saw the utility in the system compared to the >> alternatives. > J > Lets not forget ALLIN1 which was the largest installed-seat email system3 > in the world and which startd in the early 1980s.  > G > However, communications with the outside world were very basic before E > the internet became widely available, often through clunky X25 line  > editor interfaces.  C UUCP based Email and USENET were doing just fine before ALLIN1 and, D yes, before X25 too.  It's pretty hard to do no as I don't know thatA anyone bothered to archive that newsgroup, but a look at the UUCP > maps makes for interesting reading.  The world was pretty wellC connected even before there was an ARPANet, much less the NSFNet or  what we kow as the INTERNET.   > G > I managed a group of bankers for about 15 years with email. Despite a G > clunky interface, the service was a success because they had specific A > information to exchange. So there was enough traffic to make it < > worthwile for the secretaries to check the email each day.  A Back in the early days of USENET there was a world full of people D with information to exchange.  People who became familiar with EmailA and USENET in college went on to expand the network ever further. D For example, there is a Car Dealership in Towanda, PA (find that oneC a map!) that paid for nightly calls to Cornell in order to have his  small Unix system "on the net".    > G > Anopther group I helped get started wasn't so lucky. There were a vew E > visionaries who saw the real need for information exchange. But the G > members of that organisations didn't really have anything specific to J > send, and with little traffic, there was veryt little incentive to checkI > the emails  regularly, and because members didn't check regularly, high E > value information didn't have confidence level of being received in , > time, so members still used other methods.  B Not the fault of the medium.  I have people here at the university@ who are still like that. (We have someone in a position of power> over our IT infrastructure who was instrumental in eliminating@ USENET news from the University cause he didn't see any value in it.)   >  > I > What made EMAIL really work is when corporations hoooked their internal H > email systems to the internet so that staff could exchange emails withH > anyone else, and more importantly, have emails delivered to their deskI > instead of them having to use a clunky interface to access and external  > email system.   B What you call "clunky interfaces" are more than suitable for EmailF and are still in use by many people, including people who's experienceE doesn't go back to the days when that was all there was.  No one ever 9 got infected with an email borne virus using Pine or Elm.    bill   --  J Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>       ------------------------------   Date: 27 May 2005 22:05:41 GMT( From: bill@cs.uofs.edu (Bill Gunshannon)H Subject: Re: HP to help governments with tracking users with NAtional ID+ Message-ID: <3fpjtlF90q5gU2@individual.net>   3 In article <OeluntS4NXl4@eisner.encompasserve.org>, > 	koehler@eisner.nospam.encompasserve.org (Bob Koehler) writes: > H >> I'm old enough to have punched cards for computer class and have even >> seen core memory in use.  > I >    You don't have to be that old.  We're still using it.  It's very rad 
 >    hard.   : So which do they have on ISS?  An IBM 24 or a Univac 1610?   bill   --  J Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>       ------------------------------    Date: 27 May 2005 18:16:48 -0500- From: Kilgallen@SpamCop.net (Larry Kilgallen) H Subject: Re: HP to help governments with tracking users with NAtional ID3 Message-ID: <S9TAmrJMdVTE@eisner.encompasserve.org>   T In article <3fpbqqF8vc5tU1@individual.net>, Paul Sture <paul.sture@decus.ch> writes:  I > DECset I know, but I don't remember LinkWorks (though the SPD contains  ? > references to such things as TeamWorks, which I do remember).o  . So it is the _second_ LinkWorks that you know.   Which DECset do you know ?   ------------------------------  % Date: Fri, 27 May 2005 20:22:22 -0400m- From: JF Mezei <jfmezei.spamnot@teksavvy.com> H Subject: Re: HP to help governments with tracking users with NAtional ID, Message-ID: <4297B9AA.3A5BF7DD@teksavvy.com>   Bill Gunshannon wrote:J > >> I'm old enough to have punched cards for computer class and have even > >> seen core memory in use.l  < > So which do they have on ISS?  An IBM 24 or a Univac 1610?    F They'd need a specially made for NASA card reader for the ISS. With noH gravity, the weight on top of the stack of cards wouldn't be pushing theG cards down to be sucked one by one by the card reader. And upon exitingvC the machine, the cards would fly all over the place because gravitym0 wouldn't pull them down into the output tray :-)  G So NASA undoubtedly spend hundreds of billions designing a punched cardgC reader that workd in 0g.  (Similar to the urban myth stories of thei space pen).t   ------------------------------  % Date: Fri, 27 May 2005 21:16:42 -0400d- From: William Webb <william.w.webb@gmail.com> H Subject: Re: HP to help governments with tracking users with NAtional ID7 Message-ID: <8660a3a105052718167e63cda7@mail.gmail.com>i  : On 5/27/05, JF Mezei <jfmezei.spamnot@teksavvy.com> wrote: > Bill Gunshannon wrote:L > > >> I'm old enough to have punched cards for computer class and have eve= n  > > >> seen core memory in use.  >=20> > > So which do they have on ISS?  An IBM 24 or a Univac 1610? >=20 >=20H > They'd need a specially made for NASA card reader for the ISS. With noJ > gravity, the weight on top of the stack of cards wouldn't be pushing theI > cards down to be sucked one by one by the card reader. And upon exitingeE > the machine, the cards would fly all over the place because gravity 2 > wouldn't pull them down into the output tray :-) >=20I > So NASA undoubtedly spend hundreds of billions designing a punched cardeE > reader that workd in 0g.  (Similar to the urban myth stories of the/
 > space pen)./ >=20& What about *punching* cards in Zero G.  = You'd end up with the penultimate "hanging (floating?) chads"s   WWWebb   --=20 C NOTE: This email address is only used for noncommerical VMS-relateds correspondence.eC All unsolicited commercial email will be deemed to be a request fora8 services pursuant to the terms and conditions located at# http://bellsouthpwp.net/w/e/webbww/r   ------------------------------  % Date: Fri, 27 May 2005 16:34:55 -0400y- From: JF Mezei <jfmezei.spamnot@teksavvy.com>h! Subject: Re: Installing a printer + Message-ID: <4297846E.BD5151D@teksavvy.com>    Paul Anderson wrote:B > I suspect since the great majority of customers purchasing an HPH > printer aren't going to be running it on OpenVMS, and not all of those: > people would run DCPS, it would not make sense to do so.    E It would be a great marketing ploy to suddently include VMS "drivers"sH for all HP printers being shipped. It would send a strong signal that HPL has taken ownership of VMS and that VMS is still alive and widely available.   ------------------------------  # Date: Sat, 28 May 2005 00:01:53 GMT9  From: John Santos <john@egh.com>! Subject: Re: Installing a printerh* Message-ID: <RxOle.5399$3u3.3402@trnddc07>  
 MJS wrote: > Dave Froble wrote: > 
 >> MJS wrote:a >>$ >>> Peter 'EPLAN' LANGSTOEGER wrote: >>>dC >>>> In article <42941b7d$0$347$cc9e4d1f@news.dial.pipex.com>, MJS ,+ >>>> <maniac449@nospam.hotmail.com> writes:  >>>>J >>>>> We'd like to use the parallel port 'LRA0:' but also have the option I >>>>> of Ethernet.  As LRA0: is displayed when I 'SH DEV' I take it that o3 >>>>> the driver for the port is already installed?h >>>> >>>> >>>> >>>> >>>>H >>>> Yup. The driver for the parallel port is part of OpenVMS itself andE >>>> is loaded automatically if the hardware is found (at boot time).t >>>>L >>>> But, you can't use DCPS (as was my previous suggestion) on the parallelB >>>> port. DCPS requires bidirectional communication (with the PS  >>>> interpreter)aI >>>> and doesn't like parallel ports even when bidirectional (don't know w
 >>>> why).L >>>> And DCPS has a lot of goodies I personally can not imagine to ever miss >>>> again. So reconsider LAN. >>>> >>>eH >>> I've been wondering about that, read the DCPS System Manager manual H >>> today but it only referred to Network or Serial printing. Can't say / >>> I've ever heard of a serial printer though?a >> >> >>! >> That makes me feel really old.  >  > I > I'm no spring chicken myself, just more used to PC systems (and Amigas sJ > from a long time ago) & using Parallel, USB or Ethernet; never heard of > > serial printing.  Oh what a sheltered upbringing I've had :) > H > Trying to setup a printer on VMS has been a real eye-opener, I wish I  > could say in a good sense. >  >>J >>> Looks as though my only real option is to acquire DCPS from somewhere H >>> an use the second Ethernet port on the AlphaServer for the printer, - >>> best I start reading up on UCX as well :(r >> >> >>D >> No.  You don't run a direct point to point connection when using C >> ethernet.  Just hook the printer up as another node on whatever c& >> existing ethernet you already have. >> > I > I've use direct point to point Ethernet before (for a home PC network) tG > by using a cross-over cable instead of a patch cable, are you saying  8 > that this won't work for AlphaServer to Printer setup?  A It should work fine, it's just more complicated to set up (makingu> sure your IP routing is right, etc.) and seems like a waste of; an ethernet port.  If you are all ready on a LAN, then justt@ add another drop for the printer and access it over the existing? LAN.  Then other systems can use the printer too.  (You can setr; up the VMS system to share the printer, by acting as an LPDs; server or by defining a LAT service on it or any of severals: other ways, but again it's more work to set up, and no one? can access the printer in the extremely unlikely event :-) thatp the Alpha is down.)   = Save your extra ethernet port for clustering or LAN failover.   K > Apologies for the late reply but a certain football match in Turkey last  I > night had my complete attention, 'You'll Never Walk Alone' (sorry that U > crept our ;) ).k >  > Thanks to allr > MJSn     --   John Santosi Evans Griffiths & Hart, Inc. 781-861-0670 ext 539   ------------------------------  % Date: Fri, 27 May 2005 23:35:08 -0400b' From: Dave Froble <davef@tsoft-inc.com> ! Subject: Re: Installing a printeri0 Message-ID: <119fplts7rnei21@corp.supernews.com>   Paul Anderson wrote:? > In article <42962ef8$0$388$cc9e4d1f@news.dial.pipex.com>, MJS ' > <maniac449@nospam.hotmail.com> wrote:m >  > A >>If DCPS is the 'best' way to use a printer on VMS, why isn't itr; >>included on the driver disk as standard, particularly for  >>'Bulk/Commercial' printers?  >  > G > Are you wondering why DCPS isn't on the CD you get with the printer?  D > During our Genicom days, we actually did that, especially when theG > printer required a new version of DCPS.  But a good percentage of our 4 > customers were going to use the printer with DCPS. > B > I suspect since the great majority of customers purchasing an HPH > printer aren't going to be running it on OpenVMS, and not all of those: > people would run DCPS, it would not make sense to do so. > A > On the other hand, the entire DCPS kit and documentation can ber > compressed to less then 7 MB.y >  > Paul >   : I'd guess that most of such CDs have lots of unused space.   -- k4 David Froble                       Tel: 724-529-04504 Dave Froble Enterprises, Inc.      Fax: 724-529-0596> DFE Ultralights, Inc.              E-Mail: davef@tsoft-inc.com 170 Grimplin Roade Vanderbilt, PA  15486o   ------------------------------  # Date: Sat, 28 May 2005 04:59:49 GMTa From: Opcom <eccm@swbell.net>d, Subject: largest DEC collection  yada yada..< Message-ID: <9VSle.2993$3D6.2684@newssvr12.news.prodigy.com>   Friends of the DECWorld,   It's me, Opcom again..  @ I am bound to announce that the eldest and holiest parts of the B 'collection' now have new homes, right down to the last 8" floppy / and DECtape. No sacred items have been defiled./  @ I am especially pleased that the inoperable VAX 11/780 has been @ taken to be used to re-case a working '780 belonging to another > enthusiast that in the past had fallen off a liftgate and got  rather smashed up.  B There are some remains to be had, some for free, some not (weight A inversely proportional to price), you must pick up or send truck.e  @ Approximately 7 VAX 6000 machines. 208V Y 3 phase. some 6500's, > some 6400's with BI bus. Some SDI, DSSI, and CI. Includes two < pristine 6400 types with tall SF type storage which used to A reside at Sony wafer FAB. Maybe some RA92's, have to check. Lots e$ of extra parts, boards, cables, etc.  B DEC Pro 380 integrated desk model. Working with o/s, network, and  color CRT on elevator platform.a   DEC Rainbowe   Vaxmatee  ; About an 8' high stack of vaxstations and vaxstation-sized uB storage expansion cabinets. Several graphics head CRT's for same,  and cables, keyboards, etc.   5 About a 4' stack of "Lunchbox" sized vaxstation gear.   	 KIMS box.1  A And for the outright junque collectors and apocalyptic hoarders, pA a few near perfect IBM/Compaq XT/PC's, Compaq luggable PC's, one cB mint 286, and quite a nice collection of XT laptops and "beehive" @ or ye olde style terminals. 5.25" floppy disks and drives, etc.  ad nauseum.i  - I could be persuaded to ship the small stuff.P  > If interested, please decode the secret e-mail address at the ) bottom of the message and drop me a note.    FOB Dallas Texas  9 Maybe I'll post some pics soon, have to get around to it.e  
 Thank you,   "remember the Wiz"   -- S
 rest begards,c Patrick Jankowiako	 VMS 4evere  +    r e c y c l e r AT s w b e l l DOT n e t    ------------------------------    Date: 27 May 2005 15:15:30 -0700' From: "Graham" <grahambayley@gmail.com>-! Subject: MultiHead issue on AS400 B Message-ID: <1117232130.237342.14570@f14g2000cwb.googlegroups.com>   Hi all,   G I have three AS400's all triple head.  I recently upgraded to LCD's ando  E noticed only on one machine that head 1 of (0,1,2) on machine 1 is atnD a different resolution to the others. Screen 0 @ 1280x1024, Screen 1! @ 960x1024, Screen 2 @ 1280x1024.   6 We did some checking and the normal files are in order; DECW$DEVICE_CONFIG_GY.COM and DECW$PRIVATE_SERVER_SETUP.COM   < The issue seems to be not a DECW issue as on a cold boot theB res of the screen 1 remains at 960x1024.  I looked at the graphics/ cards.  All cards are the same jumper settings.e  ? Anyone any ideas how to get screen 1 to 1280x1024 ... please ??e   Graham   ------------------------------  # Date: Fri, 27 May 2005 21:56:05 GMTd1 From: Keith Parris <keithparris_NOSPAM@yahoo.com> & Subject: Re: New Cluster Interconnects2 Message-ID: <VHMle.6177$2B5.1386@news.cpqcorp.net>   Richard Maher wrote:K > 1) I saw 10 Gigabit NIC support scheduled for VMS 8.3 (Depending on whichhL > slide you look at it says "Integrity Servers Only"). Now 10x what a lot ofI > people are using for a cluster-interconnect at the moment sounds prettytG > s__t-hot to me! Especially if you're moving big lock-trees around the/N > cluster. Given that this functionality is less than a year away, surely someI > performance figures or at least anecdotal evidence should be available?w  G One way to estimate the impact is to see how Gigabit Ethernet compared aI with Fast Ethernet. Bandwidth went up by close to 10x; latency went from 0E about 240 microseconds for a round-trip lock conversion request with mI Fast Ethernet to about 200 microseconds on Gigabit Ethernet, as measured DG on a typical Alpha box of the recent past (Wildfire or ES40). So don't  ! expect a 10x latency improvement.m  M > I mean, if I was an Rdb engineer that had used p___-poor DLM performance asnN > the rationale for sticking all of my R&D eggs in the stand-alone single-nodeK > basket, then I'd be interested in what's happening with this. Right? "ButfM > it's not the bandwidth, it's the latency that gets ya." Well that brings me* > to the next slide. . .  F Rdb supports Row Cache in Galaxy Shared Memory between multiple nodes.  J > 2) Next Generation Low-Latency Interconnects Post 8.3 (Integrity Servers > Only)  > A > Am I the only person getting their jollies out of this or what?B    Nope. I also find this exciting.  E Potential candidate technologies one would naturally want to look at  ( could include Infiniband and RDMA/iWARP.  I Infiniband promises low latency, but hasn't really taken off much in the xF industry yet, and hardware is expensive. Some initial proponents have I subsequently backed out (like Intel). A lot of people are waiting to see g how this turns out.   D RDMA/iWARP looks to have broad potential industry support, and will B quite possibly be built into commodity Ethernet adapters. Latency G wouldn't be quite as low, but price would be low and price/performance r
 very good.  E Should be interesting. In any case, I'm know VMS Engineering has its  F finger on the pulse of the technologies available in the marketplace, C and will provide a quality solution with the best interests of the - customers in mind.  M > Will there be a special limit on the distances between nodes for this stuffaM > to work? (Like memory channel) Can you have a Disaster Tolerant Low Latencyr
 > Cluster?  H Infiniband has some fairly low distance limitations, unless you include 0 an Infiniband router, and that's for IP traffic.  G For anything Ethernet-based I don't expect distance limitations. Other sA than how far you can drive light over fiber, there's no inherent ,I distance limit in Gigabit Ethernet today, for example. But of course the rD longer your inter-site distance, the more delay there is due to the F speed of light over the distance, which mitigates against low latency. ---gI Looks likely I'll be doing a hands-on workshop at HP Technology Forum on  E Long-Distance OpenVMS Clusters. We'll explore some of the impacts of nB long distances on performance in that workshop, for folks who are  interested.    ------------------------------  % Date: Sat, 28 May 2005 02:47:03 +0200 3 From: "Dr. Dweeb" <NOSPAM_5msg0h202@sneakemail.com>v& Subject: Re: New Cluster Interconnects= Message-ID: <4297bf85$0$78287$157c6196@dreader1.cybercity.dk>    Dr. Dweeb wrote: > Keith Parris wrote:  >> Richard Maher wrote:nG >>> 1) I saw 10 Gigabit NIC support scheduled for VMS 8.3 (Depending oniF >>> which slide you look at it says "Integrity Servers Only"). Now 10xD >>> what a lot of people are using for a cluster-interconnect at theH >>> moment sounds pretty s__t-hot to me! Especially if you're moving bigH >>> lock-trees around the cluster. Given that this functionality is lessA >>> than a year away, surely some performance figures or at leastm+ >>> anecdotal evidence should be available?e >>@ >> One way to estimate the impact is to see how Gigabit EthernetB >> compared with Fast Ethernet. Bandwidth went up by close to 10x;A >> latency went from about 240 microseconds for a round-trip lockiE >> conversion request with Fast Ethernet to about 200 microseconds onbE >> Gigabit Ethernet, as measured on a typical Alpha box of the recentHF >> past (Wildfire or ES40). So don't expect a 10x latency improvement. >>@ >>> I mean, if I was an Rdb engineer that had used p___-poor DLMG >>> performance as the rationale for sticking all of my R&D eggs in the>D >>> stand-alone single-node basket, then I'd be interested in what'sE >>> happening with this. Right? "But it's not the bandwidth, it's the2E >>> latency that gets ya." Well that brings me to the next slide. . .u >>B >> Rdb supports Row Cache in Galaxy Shared Memory between multiple	 >> nodes.dD >>> 2) Next Generation Low-Latency Interconnects Post 8.3 (Integrity >>> Servers Only)> >>> C >>> Am I the only person getting their jollies out of this or what?  >># >> Nope. I also find this exciting.  >>G >> Potential candidate technologies one would naturally want to look ata+ >> could include Infiniband and RDMA/iWARP.  >>G >> Infiniband promises low latency, but hasn't really taken off much in-G >> the industry yet, and hardware is expensive. Some initial proponents"A >> have subsequently backed out (like Intel). A lot of people aret% >> waiting to see how this turns out.- >>F >> RDMA/iWARP looks to have broad potential industry support, and willD >> quite possibly be built into commodity Ethernet adapters. Latency7 >> wouldn't be quite as low, but price would be low and@ >> price/performance very good.r >>G >> Should be interesting. In any case, I'm know VMS Engineering has its H >> finger on the pulse of the technologies available in the marketplace,E >> and will provide a quality solution with the best interests of the  >> customers in mind.t >>D >>> Will there be a special limit on the distances between nodes forE >>> this stuff to work? (Like memory channel) Can you have a Disasterk! >>> Tolerant Low Latency Cluster?t >>B >> Infiniband has some fairly low distance limitations, unless you; >> include an Infiniband router, and that's for IP traffic.i >>C >> For anything Ethernet-based I don't expect distance limitations.p@ >> Other than how far you can drive light over fiber, there's noF >> inherent distance limit in Gigabit Ethernet today, for example. ButF >> of course the longer your inter-site distance, the more delay thereB >> is due to the speed of light over the distance, which mitigates >> against low latency. ---h >aC > Indeed - it is the physical absolute barrier for comminucation ofoB > data or energy over distance, something to do with instantaneousG > causal events not being possible and all that (if I recall my physicsp > of relativity correctly) :-) >w > Dweeb. >s  L But perhaps not http://homepage.sunrise.ch/homepage/schatzer/space-time.html   Dweeb.  H >> Looks likely I'll be doing a hands-on workshop at HP Technology ForumG >> on Long-Distance OpenVMS Clusters. We'll explore some of the impactsuG >> of long distances on performance in that workshop, for folks who areo >> interested. r   ------------------------------  % Date: Sat, 28 May 2005 02:37:50 +0200a3 From: "Dr. Dweeb" <NOSPAM_5msg0h202@sneakemail.com>y& Subject: Re: New Cluster Interconnects= Message-ID: <4297bd5d$0$78279$157c6196@dreader1.cybercity.dk>5   Keith Parris wrote:i > Richard Maher wrote:F >> 1) I saw 10 Gigabit NIC support scheduled for VMS 8.3 (Depending onE >> which slide you look at it says "Integrity Servers Only"). Now 10xeC >> what a lot of people are using for a cluster-interconnect at thewG >> moment sounds pretty s__t-hot to me! Especially if you're moving big G >> lock-trees around the cluster. Given that this functionality is lessi@ >> than a year away, surely some performance figures or at least* >> anecdotal evidence should be available? >sH > One way to estimate the impact is to see how Gigabit Ethernet comparedE > with Fast Ethernet. Bandwidth went up by close to 10x; latency wentsF > from about 240 microseconds for a round-trip lock conversion requestF > with Fast Ethernet to about 200 microseconds on Gigabit Ethernet, asA > measured on a typical Alpha box of the recent past (Wildfire oro3 > ES40). So don't expect a 10x latency improvement.o >.? >> I mean, if I was an Rdb engineer that had used p___-poor DLMaF >> performance as the rationale for sticking all of my R&D eggs in theC >> stand-alone single-node basket, then I'd be interested in what'stD >> happening with this. Right? "But it's not the bandwidth, it's theD >> latency that gets ya." Well that brings me to the next slide. . . >iH > Rdb supports Row Cache in Galaxy Shared Memory between multiple nodes. >oC >> 2) Next Generation Low-Latency Interconnects Post 8.3 (Integrity  >> Servers Only) >>B >> Am I the only person getting their jollies out of this or what? >g" > Nope. I also find this exciting. > F > Potential candidate technologies one would naturally want to look at* > could include Infiniband and RDMA/iWARP. >tF > Infiniband promises low latency, but hasn't really taken off much inF > the industry yet, and hardware is expensive. Some initial proponents@ > have subsequently backed out (like Intel). A lot of people are$ > waiting to see how this turns out. >.E > RDMA/iWARP looks to have broad potential industry support, and willoC > quite possibly be built into commodity Ethernet adapters. LatencyNH > wouldn't be quite as low, but price would be low and price/performance > very good. >eF > Should be interesting. In any case, I'm know VMS Engineering has itsG > finger on the pulse of the technologies available in the marketplace,cD > and will provide a quality solution with the best interests of the > customers in mind. >TC >> Will there be a special limit on the distances between nodes for D >> this stuff to work? (Like memory channel) Can you have a Disaster  >> Tolerant Low Latency Cluster? >nA > Infiniband has some fairly low distance limitations, unless youh: > include an Infiniband router, and that's for IP traffic. >eH > For anything Ethernet-based I don't expect distance limitations. OtherB > than how far you can drive light over fiber, there's no inherentF > distance limit in Gigabit Ethernet today, for example. But of courseE > the longer your inter-site distance, the more delay there is due totC > the speed of light over the distance, which mitigates against low  > latency. ---  J Indeed - it is the physical absolute barrier for comminucation of data or K energy over distance, something to do with instantaneous causal events not eM being possible and all that (if I recall my physics of relativity correctly) t :-)n   Dweeb.  G > Looks likely I'll be doing a hands-on workshop at HP Technology Forum F > on Long-Distance OpenVMS Clusters. We'll explore some of the impactsF > of long distances on performance in that workshop, for folks who are > interested.    ------------------------------    Date: 27 May 2005 19:25:50 -0700! From: susan_skonetski@hotmail.comt1 Subject: Re: OpenVMS Advanced Technical boot camppB Message-ID: <1117247150.448116.92600@z14g2000cwz.googlegroups.com>   We are now at 197.   Suei   ------------------------------  % Date: Fri, 27 May 2005 16:31:52 -0400g- From: JF Mezei <jfmezei.spamnot@teksavvy.com> 6 Subject: Re: QIOs in multi thread environment (Pascal), Message-ID: <429783B7.CD11A6CD@teksavvy.com>   Erich Weber wrote:K > I've supposed, that if one thread waiting in a $QIOW for input, the other-' > threads starting working immediately.0  G I remember a discussion of this a year or two ago here. You may wish to  try google to retrieve it.  G I think it had something to do with QIOW putting the *process* to LEF.    5 Here is what the DEC/Posix Threads Manual has to say:,     ## B.12.6 Blocking System ServiceswM In OpenVMS Alpha Version 7.0 and later, with few exceptions a blocking systemeO service call is thread synchronousthat is, only the calling thread is blocked. G The exceptions are services that do not block in user mode and servicese that set. common event flags. (See also Section B.12.8.)K When a thread calls a system service that must block, the OpenVMS scheduleroL makes an upcall to allow the Threads Library to schedule another user threadD to execute. Therefore, only the calling thread is blocked, all other threads areeJ unaffected, and the process continues running. When the service completes,J the thread is awakened by means of another upcall, and the Threads Library; schedules it to run again at the threads next opportunity.aM This applies to all W forms of system services, for example, $QIOW, $END_mL TRANSW, and $GETJPIW. Additionally, this applies to the following event flag' services: $WAITFR, $WFLAND, and $WFLOR.o ##  H So it seems that if you don't use event flags in the QIOW, you should beG OK. Also Google the group (or the FAQ ?) to find out how to NOT specify @ an event flag. Seems that using "0" still usess some event flag.   ------------------------------  % Date: Sat, 28 May 2005 05:42:35 +020082 From: martin@radiogaga.harz.de (Martin Vorlaender)6 Subject: Re: QIOs in multi thread environment (Pascal); Message-ID: <4297e8ab.524144494f47414741@radiogaga.harz.de>e  . JF Mezei <jfmezei.spamnot@teksavvy.com> wrote:J > So it seems that if you don't use event flags in the QIOW, you should beI > OK. Also Google the group (or the FAQ ?) to find out how to NOT specifyfA > an event flag. Seems that using "0" still uses some event flag.i  E There's EFN$C_ENF "Event No Flag" (with a value of 127) - at least inv the C header file EFNDEF.H.t   cu,r   Martin -- aA                            | Martin Vorlaender  |  OpenVMS rules!g4  UNIX is user friendly.    | work: mv@pdv-systeme.deG  It's just selective about |   http://www.pdv-systeme.de/users/martinv/r;  who its friends are.      | home: martin@radiogaga.harz.dew   ------------------------------    Date: 27 May 2005 14:54:28 -0700 From: anuj.goyal@gmail.com. Subject: Re: VMS API (possible unix emulation)B Message-ID: <1117230868.266144.36010@g47g2000cwa.googlegroups.com>  
 Mike and Ken,5  G thanks for the links they really helped out, now I want to test out therE C++ compiler on a OpenVMS system.  None of the 'free' systems seem toeG have the C++ compilers installed on them, I especially want to test outa 8.2 on Integrity (ia64).   see section 2.11  - http://h71000.www7.hp.com/faq/vmsfaq_002.htmlI    I am trying to port the STLport:+ (http://www.stlport.org/doc/platforms.html)e   ------------------------------    Date: 27 May 2005 17:42:40 -0700 From: anuj.goyal@gmail.com. Subject: Re: VMS API (possible unix emulation)C Message-ID: <1117240960.238041.314690@g49g2000cwa.googlegroups.com>d  A I was mistaken, the deathrow cluster has a recent version of C++.n   DAHMER $ CXX /VERSIONo* Compaq C++ V6.3-020 for OpenVMS Alpha V7.2   ------------------------------  % Date: Fri, 27 May 2005 21:09:02 -0400h+ From: Ken Robinson <kenrbnsn1@patmedia.net>i. Subject: Re: VMS API (possible unix emulation)A Message-ID: <6.2.3.0.2.20050527210754.090c3d68@mail.patmedia.net>   2 At 08:42 PM 5/27/2005, anuj.goyal@gmail.com wrote:B >I was mistaken, the deathrow cluster has a recent version of C++. >- >DAHMER $ CXX /VERSION+ >Compaq C++ V6.3-020 for OpenVMS Alpha V7.2n  0 This is the information on the testdrive system:   $ sho lic/ch1 OpenVMS I64/LMF Charge Information for node TD183OF This is an HP rx1620  (1.60GHz/3.0MB), with 2 CPUs active, 2 socket(s)3 Type: PPL,   Units Required: 2  (I64 Per Processor)r	 $ sho net   6 Product:  TCP/IP        Node:  td183.testdrive.hp.com  Address(es):  192.233.54.183 $ 	 $ cxx/ver-$ HP C++ V7.1-011 on OpenVMS IA64 V8.2 $ sho sys/noprocI OpenVMS V8.2  on node TD183  27-MAY-2005 21:05:51.80  Uptime  11 10:31:11    ------------------------------   Date: 27 May 2005 19:03:17 GMT$ From: "Doc." <doc@openvms-rocks.com> Subject: Re: VMS TCPIP Services 7 Message-ID: <Xns9663D646FA8A1dcovmsrox@212.100.160.126>e  2 %NEWS-I-NEWMSG, Peter 'EPLAN' LANGSTOEGER wrote in$ news:42975a0f$1@news.langstoeger.at   D > In article <3forqgF8ufffU1@individual.net>, bill@cs.uofs.edu (Bill > Gunshannon) writes: D >>OK, I know this has to be something simple but I have followed allE >>the instructions I have been able to find in the documentataion andnD >>it still doesn't work.  I need to increase INTSTKPAGES in order to >>run TCPIP. > ( > What system ? INTSTKPAGES is VAX only.  I Bill has been talking for the past couple of weeks about getting VAX kit -; up and running and collaborating with some of his PDP toys.   E I'm not knowledgeable enough to provide a definitive answer, but I'd i guess at needing to useh   MIN_INTSTKPAGES=xH  > to get the parameter to always meet your minimum requirements.     Doc. -- 0G OpenVMS:     Eight out of ten hackers prefer *other* operating systems.JG http://www.openvms-rocks.com    Deathrow Public-Access OpenVMS Cluster.    ------------------------------  % Date: Fri, 27 May 2005 16:43:12 -0400 - From: JF Mezei <jfmezei.spamnot@teksavvy.com>N Subject: Re: VMS TCPIP Services , Message-ID: <4297865E.5770AE3D@teksavvy.com>  
 "Doc." wrote:oJ > Bill has been talking for the past couple of weeks about getting VAX kit= > up and running and collaborating with some of his PDP toys.t  * From a James Bond Movie:  Boys with toys !   > guess at needing to use  >  > MIN_INTSTKPAGES=x     E Just for the record, on both my machines, I have min_intstkpages = 16i6 and when all is said and done, AUTOGEN sets it to 16.   E Autogen on VAX is no longer such an "automatic" thing and doesn't setn0 some parameters correctly (in particular SPTREQ)   ------------------------------   End of INFO-VAX 2005.295 ************************as pretty wellC connected even before there was an ARPANet, much less the NSFNet or  what we kow as the INTERNET.   > G > I managed a group of bankers for about 15 years with email. Despite a G > clunky interface, the service was a success because they had specific A > information to exchange. So there was enough traffic to make it < > worthwile for the secretaries to check the email each day.  A Back in the early days of USENET there was a world full of pFqΉHgD	ޮ넭+iR7'vHP@y)p7ttc7"­4:RdWzxV	G7F&౑ZuGL0Li*n`ҽU>-KP44nD#f)ɈюT"oN{0ެH}Y&[W%u͉z-B#ol7yT*4YRh}0%QdR(RWe8rfH^[0V:UgClmoR8#)3!AH)p%+16 p  
9Y3򘪯oޚn)Z'dW=vK>99wg
ӣ9P7ODa^>P<.]9ՙ`^SU-H3O/
F	!	"H>
I[FRN>_WF 0gG _)on=OlsO@~R.s-n<kuck	-i?.BW#KtwW(-6IW(/%'̗8_B9c]\͂iؕƦ{5eGWtxp
l&:#Xaic+uWA_xzqd+ `E@kNLRC<-q +f8 )T0|MsHYU9WVCl2K_j?Z&^x|*rJXXR\Cdŀ)a>\Q!/96N@6e%鼴D	aFNNdW	q1&	p.<ea3g>!,K8h:1*:}Vkhc.
PI1%+68Q+"p&R(=Jl)f!=◱ΐ$+Oz/x#u{ *]@VSP XrEzdsSn;c>o``zVUDLYKMf[4@oʖWva]_ҘG?];RKpT6DMe[
0[ewKlo6:焣N$L]6tW9qhr^T(۪gx*.	W1Zu1v@KL`ރIR08z?kg&ܮd8;ƋV
_x]b⋈sʎ53E4?rDƮ`I,,ך0d Ax݈"t#`jCr^lf"W+rs˯S nƕ)p	bU
#EP4Z!	PJ07HH;A
r&麸/*Wc#Zd>kͶ

!/3D
)~)\fmKnT(u~	\b^d9c-*
jFGdۅt#Z%B N_X}05L룜!<1(QuA;leFJ,M>gG(-dx3P1GH`:b{$;d,#Oůg=<X:S6?-`ΒXDHཚlBn`GWMBq-: 	jPFop[B!I榫a ͚X# ȟ*iA#}CΓ!l"mާ["HUwt9ޜY{V
4E62`cٵ۶rX1r\~aHT^)j,|ur.V^:r}檿
gapG/ާcd~hdF
DTITucwF\ itS%q<wyJ=*BWߋ<\	~<^Q՗`֢88`eG&vG7Ȯ5EHYFO0h4k)d0̙Nrbh9d̌1DPg4F |ɞ9
1jAJLTyS	"B ˃a*WxÉ!NkALfћ~/?<s!
a*φb[And.*(0zY(PCǫJ(0j9%c%FQ$ersZO(3jq\q\
ꐌz># |+	"0hM+}f$m60d4ͬ<>j@뉍-G{,YVg%(om>zrGHB8٘l;hZ
Ar+2qE2y"oVHpXV%5BC@=<;9A]۽MCPi3xh _{2k`	RT/DIW(YdN62	A1Kٲ
#>Y2XOnMnrK	'.
v7HdE]+:aNR!1)m蚂Oz'I l_pF
nx2etlҮJ)3J<@:>殮1JefXkb"co$_aNJ*	:``eK P 	Ae;!M_z+Ë!vNjC8.!oGBF$Jn1<4D tUv܌51)ÁYy̔-aɇgH TD`M]F8Q[ҰFyf鎓"c$ٽ
9\K^n+^2j zc	rsC0_1ZN('ir3u)I7Rz Vъy<