1 INFO-VAX	Mon, 13 Nov 2006	Volume 2006 : Issue 624       Contents: Datacenter standarts(off topic)  Re: DECconnect adapters  Re: DECconnect adapters . Re: Easiest / Cheapest way to get into OpenVMS. Re: Easiest / Cheapest way to get into OpenVMS. Re: Easiest / Cheapest way to get into OpenVMS. Re: Easiest / Cheapest way to get into OpenVMS. Re: Easiest / Cheapest way to get into OpenVMS. Re: Easiest / Cheapest way to get into OpenVMS Re: Most "classic" VMS version Re: Novice's questions7 Re: Performance comparison Alpha ES40 vs Itanium rx3600 7 Re: Performance comparison Alpha ES40 vs Itanium rx3600 E Problem with sys$system:AXPVMS$PCSI_INSTALL_MIN.COM (VMS V8.3, Alpha)  Re: TIE/MSAI Re: TIE/MSAI Re: Where to get a cable? 6 Re: writing fixed size records with one shorter record6 Re: writing fixed size records with one shorter record6 Re: writing fixed size records with one shorter record6 Re: writing fixed size records with one shorter record6 Re: writing fixed size records with one shorter record  F ----------------------------------------------------------------------    Date: 12 Nov 2006 13:25:24 -0800 From: apogeusistemas@gmail.com( Subject: Datacenter standarts(off topic)B Message-ID: <1163366724.304055.49730@h54g2000cwb.googlegroups.com>  F Do you know where could I get informations about datacenter standarts, like? how storage data tapes, off site data storage, how carrier data  tapes,etc ?  Regards    ------------------------------  % Date: Sun, 12 Nov 2006 16:34:08 -0500 ' From: Stuart Fuller <stufuller@usa.net>   Subject: Re: DECconnect adapters/ Message-ID: <gjgk24-3vv.ln1@dadsys2.fuller.com>    JF Mezei wrote:    > Jack Patteeuw wrote: >>  G >> I'm looking for a "reasonably priced" supplier of H8575-E DECconnect  >> adapters. > / > What is a -E ? From decconnect to what type ?  > G > I have some -A connectors (decconnect on one side, and female DB25 on 
 > the other).   = The -E is to connect to the serial port of a Sun workstation.    --             Stu    ------------------------------  + Date: Mon, 13 Nov 2006 04:20:48 +0000 (UTC) , From: cdl@deeptow.ucsd.edu (Carl Lowenstein)  Subject: Re: DECconnect adapters* Message-ID: <ej8rr0$1j62$2@news1.ucsd.edu>  / In article <gjgk24-3vv.ln1@dadsys2.fuller.com>, ) Stuart Fuller  <stufuller@usa.net> wrote:  >JF Mezei wrote: >  >> Jack Patteeuw wrote:  >>> H >>> I'm looking for a "reasonably priced" supplier of H8575-E DECconnect
 >>> adapters.  >>  0 >> What is a -E ? From decconnect to what type ? >>  H >> I have some -A connectors (decconnect on one side, and female DB25 on >> the other). > > >The -E is to connect to the serial port of a Sun workstation.  I What I did when faced with that problem was take a D-connector pin puller J to an H8575-F (male for DCE) and swap the pins to fit the Sun workstation.F Hint -- Sun workstation is female DTE, unlike DEC workstation which isF male DTE.  Pins in the same place on the connector, but the other sex.       carl --  B     carl lowenstein         marine physical lab     u.c. san diegoB                                                  clowenst@ucsd.edu   ------------------------------    Date: 12 Nov 2006 14:38:55 -0800+ From: "Mike G." <michael.graffam@gmail.com> 7 Subject: Re: Easiest / Cheapest way to get into OpenVMS B Message-ID: <1163371135.099871.233930@e3g2000cwe.googlegroups.com>  C Thank you to everyone who has responded to this query. Nearly every B reply was quite helpful, and it is very nice to see an active, and, friendly user base to help me along on this.   Dan Foster wrote: i > In article <1163271707.958092.100090@i42g2000cwa.googlegroups.com>, Mike <mikegraffam@yahoo.com> wrote:  > > E > > What is the easiest / cheapest way to start playing with OpenVMS?  > ? > Given your specifications, it does indeed sound like SIMH and B > OpenVMS/VAX 7.3 would be the perfect fit for a 'portable' setup. ... I > There's a eBay auction going on now for a 433au in the SF Bay area, and B > an incredible price -- just under $100 for the buy-it-now price.F > Seriously a VERY good deal... you probably want to jump on this one!  G I did exactly that. For the price, its a steal -- and even if VMS isn't  for me, D its a nice machine for NetBSD. Although, I'm going to use SIMH on my laptopG too -- at least in the beginning. This way, I can backup my SIMH config  and F HD images, and fearlessly hack on VMS with impunity. If something goesB wrong, a simple 'cp' solves my troubles. And it will provide a few hours of fun" while on my bi-weekly train rides.  F > Here's a reasonable attempt at documenting VMS and UNIX equivalents: > B > http://www.physnet.uni-hamburg.de/physnet/vms-unix-commands.html  ( Interesting read. My initial impression:  B VMS commands are verbose. Great for learning, but perhaps tiresome after ; they are in rote memory? Its a good thing that I type well.   F The operators used for scripting look like FORTRAN. Very, very cool :)  E > Keep the questions coming! I'm pretty sure you'll do just fine. I'm D > quite comfortable with multiple platforms including UNIX/Linux andD > didn't find it to be difficult to learn VMS. Just took a while for > things to sink in.  F Thanks! I'm definately looking forward to having a go with this stuff. I had myF appetite whetted many years ago, while using the Delphi online service (didn't E even have SLIP/PPP then!). Delphi ran on VMS, and one fine day, after  reading G my email and playing a bit of a text adventure game (wasn't 'adventure'  .. but IE forget the name) a thundercrack gave me some magical line noise on my  1200 bpsF modem that sent me out of the familiar menu interface straight to DCL.
 After playing C around for a bit, and having very limited success in navigating the  alien system, I F realized I was paying good money to Delphi and Tymnet to be confused.. and possiblyD risking my account with Delphi managment. So I hit the red button on
 the modem and F re-connected. So now, nearly 16 years later, curiousity has gotten the better of me -- . and I can actually afford hardware to run VMS.   Another fine day. :)  G Thanks again, everyone. I'm sure I'll be back soon with many questions.  Now I'm off to read the FAQ.    Mike   ------------------------------  % Date: Sun, 12 Nov 2006 17:21:02 -0600 % From: Dan Foster <usenet@evilphb.org> 7 Subject: Re: Easiest / Cheapest way to get into OpenVMS 5 Message-ID: <slrnelfb2u.5bi.usenet@zappy.catbert.org>   m In article <1163371135.099871.233930@e3g2000cwe.googlegroups.com>, Mike G. <michael.graffam@gmail.com> wrote:  > G >> Here's a reasonable attempt at documenting VMS and UNIX equivalents:  >>C >> http://www.physnet.uni-hamburg.de/physnet/vms-unix-commands.html  > * > Interesting read. My initial impression:  D > VMS commands are verbose. Great for learning, but perhaps tiresomeC > after they are in rote memory? Its a good thing that I type well.   A Nah, there's a special trick: the command names themselves can be : abbreviated to four letters (and in certain cases, three).   For instance, you can do:   	 $ ANA/SYS   
 instead of   $ ANALYZE /SYSTEM   G or $ ANALY /SYST or whatever combination. The parser only cares about a @ minimum number of unique characters that will help it identify a$ particular command and qualifier(s).   Another example:   $ DIR   
 instead of   $ DIRECTORY   E These, by the way, were one of the few three-letter exceptions to the ! minimum four-letter parsing rule.   C With your own DCL symbols (which in UNIX, we'd probably call 'shell H aliases'), you're allowed to have an one character minimum. So you could set this up in your LOGIN.COM:   $ c = "type /page nl:"  F That defines a symbol named 'c' which will execute TYPE /PAGE NL: whenD you type 'c'. So, really, the long command or qualifier names aren'tD usually a hindrance in actual practice. Only thing with defining DCLD symbols is to make sure it doesn't clash with an existing command orD even a future command that a system manager might eventually insert.  B There's something special about VMS commands that becomes apparentD after a while: there tend to be consistency with qualifiers (the CLIC command 'arguments'; they're known as qualifiers in the VMS world). > For instance, /OUTPUT and /LOG are two very common qualifiers.  C Why's that? Because VMS was designed to be orthogonal from the very F beginning... so you have a remarkable amount of consistency internallyB and externally. This helps with the user experience since it makesG things more predictable. This is also the same reason why a common rant A by the VMS greybeards concerns the haphazard design of UNIX's CLI 
 arguments. ;)   B In fact, VMS has a structured way of programmatically dealing withF qualifier parsing that's even cooler than getopt[s], and actually doesF make it much easier for programs to check for qualifiers and argumentsG as VMS will automagically do common housekeeping tasks like that on the  program's behalf.   A This means that programs will check for these things in a single, . published, well documented, consistent manner.  E I could elaborate more on the command line definition (CLD) stuff but  perhaps for another day.  H > The operators used for scripting look like FORTRAN. Very, very cool :)  G Speaking of which, FORTRAN was *very* popular in VMS's early days; much A of the example code in the earlier documentation used FORTRAN. In H today's documentation, new code examples tend to be C, but FORTRAN still% has a place in the OpenVMS community.   E VMS itself is pretty language-agnostic because DEC created VMS with a B common code calling convention. Basically, this allows you to callG functions in pretty much any other language under VMS. Call C code from E BASIC? Cake. Call BASIC from C? Cake. Call FORTRAN from C? Cake. Call ! MACRO-32 (assembly) from C? Cake.   E Stuff like that. You can't easily do that with UNIX even today, alas. F (Think of the headaches involved in doing C and C++ linkage, nevermind$ for other languages, on UNIX/Linux.)  F VMS itself was implemented in several languages (as the FAQ explains).H The main one is BLISS, but has code in other languages such as MACRO, C,( FORTRAN, and probably one or two others.  B > I had my appetite whetted many years ago, while using the Delphi > online service  @ Ah! Very nice. Didn't know that they ran their front-end on VMS.   -Dan   ------------------------------  % Date: Sun, 12 Nov 2006 19:45:09 -0500 3 From: "Richard B. Gilbert" <rgilbert88@comcast.net> 7 Subject: Re: Easiest / Cheapest way to get into OpenVMS : Message-ID: <WMSdnY_Dt8GOXcrYnZ2dnUVZ_sCdnZ2d@comcast.com>   Mike G. wrote:  E > Thank you to everyone who has responded to this query. Nearly every D > reply was quite helpful, and it is very nice to see an active, and. > friendly user base to help me along on this. >  > Dan Foster wrote:  > i >>In article <1163271707.958092.100090@i42g2000cwa.googlegroups.com>, Mike <mikegraffam@yahoo.com> wrote:  >>D >>>What is the easiest / cheapest way to start playing with OpenVMS? >>? >>Given your specifications, it does indeed sound like SIMH and B >>OpenVMS/VAX 7.3 would be the perfect fit for a 'portable' setup. >  > ...  > I >>There's a eBay auction going on now for a 433au in the SF Bay area, and B >>an incredible price -- just under $100 for the buy-it-now price.F >>Seriously a VERY good deal... you probably want to jump on this one! >  > I > I did exactly that. For the price, its a steal -- and even if VMS isn't 	 > for me, F > its a nice machine for NetBSD. Although, I'm going to use SIMH on my > laptopI > too -- at least in the beginning. This way, I can backup my SIMH config  > and H > HD images, and fearlessly hack on VMS with impunity. If something goesD > wrong, a simple 'cp' solves my troubles. And it will provide a few > hours of fun$ > while on my bi-weekly train rides. >  > F >>Here's a reasonable attempt at documenting VMS and UNIX equivalents: >>B >>http://www.physnet.uni-hamburg.de/physnet/vms-unix-commands.html >  > * > Interesting read. My initial impression: > D > VMS commands are verbose. Great for learning, but perhaps tiresome > after = > they are in rote memory? Its a good thing that I type well.  >   A Typing well is not as important on VMS as it is on Unix.  If you  C misspell something in DCL, you'll most likely get an error message.   F On Unix, any random string might be the name of an executable file on H your path!!!!  A typo as root could get very interesting and the system ) won't even ask "Do you REALLY mean that?"   H It's even relatively clear what a lot of the DCL commands do since they E tend to be English words bearing some relation to the function being  
 performed.  G Anyone who has ever had to type on a Model 33 teletype understands why  I Unix commands are terse; typing anything on one of those things was HARD  K WORK.  I kicked mine good-bye about thirty-two years ago and good riddance!    <snip>   ------------------------------  % Date: Sun, 12 Nov 2006 20:10:08 -0500 3 From: "Richard B. Gilbert" <rgilbert88@comcast.net> 7 Subject: Re: Easiest / Cheapest way to get into OpenVMS : Message-ID: <QYSdnS7NdZ1zWMrYnZ2dnUVZ_uCdnZ2d@comcast.com>   Dan Foster wrote:   o > In article <1163371135.099871.233930@e3g2000cwe.googlegroups.com>, Mike G. <michael.graffam@gmail.com> wrote:  > G >>>Here's a reasonable attempt at documenting VMS and UNIX equivalents:  >>> C >>>http://www.physnet.uni-hamburg.de/physnet/vms-unix-commands.html  >>* >>Interesting read. My initial impression: >  > D >>VMS commands are verbose. Great for learning, but perhaps tiresomeC >>after they are in rote memory? Its a good thing that I type well.  >  > C > Nah, there's a special trick: the command names themselves can be < > abbreviated to four letters (and in certain cases, three). >  > For instance, you can do:  >  > $ ANA/SYS  >  > instead of >  > $ ANALYZE /SYSTEM  > I > or $ ANALY /SYST or whatever combination. The parser only cares about a B > minimum number of unique characters that will help it identify a& > particular command and qualifier(s). >  > Another example: >  > $ DIR  >  > instead of > 
 > $ DIRECTORY  > G > These, by the way, were one of the few three-letter exceptions to the # > minimum four-letter parsing rule.  > E > With your own DCL symbols (which in UNIX, we'd probably call 'shell J > aliases'), you're allowed to have an one character minimum. So you could  > set this up in your LOGIN.COM: >  > $ c = "type /page nl:" > H > That defines a symbol named 'c' which will execute TYPE /PAGE NL: whenF > you type 'c'. So, really, the long command or qualifier names aren'tF > usually a hindrance in actual practice. Only thing with defining DCLF > symbols is to make sure it doesn't clash with an existing command orF > even a future command that a system manager might eventually insert. > D > There's something special about VMS commands that becomes apparentF > after a while: there tend to be consistency with qualifiers (the CLIE > command 'arguments'; they're known as qualifiers in the VMS world). @ > For instance, /OUTPUT and /LOG are two very common qualifiers. > E > Why's that? Because VMS was designed to be orthogonal from the very H > beginning... so you have a remarkable amount of consistency internallyD > and externally. This helps with the user experience since it makesI > things more predictable. This is also the same reason why a common rant C > by the VMS greybeards concerns the haphazard design of UNIX's CLI  > arguments. ;)   C I don't believe that the CLI has "arguments" as such.  The various  A programs take arguments passed in the command line.  The lack of  C orthogonality is due to the fact that each program parses it's own  C command line and the syntax can be different for each program (and  @ frequently is).  A few years ago a function called getopts() or H something close to that was introduced; it parses command lines and can G be told which options are legitimate.  It makes things a little better  B but "-a" can still mean something different to every program that  accepts it.    ------------------------------  % Date: Sun, 12 Nov 2006 19:46:22 -0600 % From: Dan Foster <usenet@evilphb.org> 7 Subject: Re: Easiest / Cheapest way to get into OpenVMS 5 Message-ID: <slrnelfjje.5bi.usenet@zappy.catbert.org>   m In article <QYSdnS7NdZ1zWMrYnZ2dnUVZ_uCdnZ2d@comcast.com>, Richard B. Gilbert <rgilbert88@comcast.net> wrote:  > D > I don't believe that the CLI has "arguments" as such.  The variousB > programs take arguments passed in the command line.  The lack ofD > orthogonality is due to the fact that each program parses it's ownD > command line and the syntax can be different for each program (andA > frequently is).  A few years ago a function called getopts() or E > something close to that was introduced; it parses command lines and E > can be told which options are legitimate.  It makes things a little E > better but "-a" can still mean something different to every program  > that accepts it.  4 Yeah, and then you also have the issue of (in UNIX):   	-a    	vs.   	--a[something]    	vs.   	a  & GNU tar under UNIX does, for instance:  3 	'-c' and 'c' to mean create, as well as '--create'   G Not all UNIX utilities will have the same alphabetical letter for -X vs  X, too.     But on VMS, it is nearly always:  ! 	<utility> /qualifier[=parameter]   G There are exceptions, such as UNIX utils ported to VMS that retains the + UNIX style argument handling, for instance.   C A common question is 'what are the CLI arguments for this utility?'    UNIX, maybe:   	% util --help
 	% util -h
 	% util -H 	% util help- 	% util -xhelp   (e.g. with the star utility)  	% man util   ? No real consistency. One has to fumble around, trying different  variations.    VMS is generally:    	$ HELP <utility>   D ...and the help is laid out in a common and consistent format, e.g.:   <utility name>   	<brief description>   	Format    	<usage format>   ! Additional information available:    	Parameter	Qualifiers ' 	<every individual qualifier goes here> 	 	Examples   H (And you can then ask it to give detailed info on individual qualifiers,D which includes the data types for input and outputs. Very handy when. you're programming in a variety of languages.)  G There are also certain qualifiers that will behave the same amongst all G VMS utilities and OS tools, too. In a sense, this is the VMS equivalent C of a GUI's copy/cut/paste menu functions in the sense that they are 1 universally consistent regardless of application.   E All told, UNIX's haphazard CLI utility argument handling is much less H straightforward than the more consistent VMS usage that is generally the+ same amongst all applications and OS tools.   F For UNIX, it's just the historical product of a fragmented development> process by many different outside sources and significant codeG divergence without any formal guidelines enforced until after the horse C had left the barn. Even to this day, applications are still left on F their own to parse things though presence of getopt[s] does help some.  A This is a frank look at things from someone whom administers both H platforms and has done so for a fair number of years. No bias here. UNIX< has its high points, but VMS usually has the cleaner design.  A That's owning to its (VMS) single-vendor professional experienced D engineers that remains for a long time throughout product lifecyclesF spanning many years and overall development history for a product thatF was designed for the future from literally day one in 1975. One gets aF real sense of this by reading various VMS 1.x product documentation orG source code and see how well they really did plan ahead for the future.    -Dan   ------------------------------  % Date: Mon, 13 Nov 2006 00:03:27 -0500 / From: "William Webb" <william.w.webb@gmail.com> 7 Subject: Re: Easiest / Cheapest way to get into OpenVMS I Message-ID: <8660a3a10611122103j5c344235idc3050f74f3c5e78@mail.gmail.com>   ( ------=_Part_2811_27369825.1163394207024; Content-Type: text/plain; charset=ISO-8859-1; format=flowed  Content-Transfer-Encoding: 7bit  Content-Disposition: inline   ' Get an account on the Deathrow Cluster.    WWWebb  B On 11 Nov 2006 11:01:48 -0800, Mike <mikegraffam@yahoo.com> wrote: >  > Hello everyone,  > H > What is the easiest / cheapest way to start playing with OpenVMS? I'veG > had a simply glorious 15-year run with UNIX-like systems now, and can F > say very happily, that I've never run Windows (ok, I dual booted DOS0 > and Minix on a 286, but that doesn't count :). > G > But I've always been curious about VMS, and my purposes at this point D > are completely for learning. I don't have any specific applicationI > right now, and probably wont: at present, I maintain no desktop systems  > of my own. > G > I had a little Alpha-based DEC Multia that I ran Linux on a number of I > years ago, before it gave up the ghost. I imagine that putting together I > another Alpha is the best choice? What would the minimum system be? I'm I > OK with a text-only interface; but playing with X on VMS would probably  > be neat too. > H > Also; what about emulator solutions in the short term? I travel a lot,E > and my primary machine is a laptop (Linux/x86-64). It would be very I > useful for me to have an emulated VMS. I understand that SIMH will boot > > OpenVMS w/ TCP/IP support. Has anyone here done this? Is itsG > performance suitable for learning (navigating the system by text, and  > compiling C code, to start). > H > Finally: what about manuals or books? A system command reference wouldE > be quite helpful. Bear in mind, I don't even know the VMS analog to 
 > 'ls'. :) > I > Any VMS Elmers that can offer some advice would be greatly appreciated.  >  >      --   Ajilon Consulting  Site resident at Quest Diagnostics = first.x.last@f$edit(contents of previous line,"COLLAPSE").com   ( ------=_Part_2811_27369825.1163394207024+ Content-Type: text/html; charset=ISO-8859-1  Content-Transfer-Encoding: 7bit  Content-Disposition: inline    Get an account on the Deathrow Cluster.<br><br>WWWebb<br><br><div><span class="gmail_quote">On 11 Nov 2006 11:01:48 -0800, <b class="gmail_sendername">Mike</b> &lt;<a href="mailto:mikegraffam@yahoo.com">mikegraffam@yahoo.com </a>&gt; wrote:</span><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">Hello everyone,<br><br>What is the easiest / cheapest way to start playing with OpenVMS? I've<br>had a simply glorious 15-year run with UNIX-like systems now, and can<br>say very happily, that I've never run Windows (ok, I dual booted DOS<br>and Minix on a 286, but that doesn't count :).<br><br>But I've always been curious about VMS, and my purposes at this point <br>are completely for learning. I don't have any specific application<br>right now, and probably wont: at present, I maintain no desktop systems<br>of my own.<br><br>I had a little Alpha-based DEC Multia that I ran Linux on a number of <br>years ago, before it gave up the ghost. I imagine that putting together<br>another Alpha is the best choice? What would the minimum system be? I'm<br>OK with a text-only interface; but playing with X on VMS would probably  <br>be neat too.<br><br>Also; what about emulator solutions in the short term? I travel a lot,<br>and my primary machine is a laptop (Linux/x86-64). It would be very<br>useful for me to have an emulated VMS. I understand that SIMH will boot <br>OpenVMS w/ TCP/IP support. Has anyone here done this? Is its<br>performance suitable for learning (navigating the system by text, and<br>compiling C code, to start).<br><br>Finally: what about manuals or books? A system command reference would  <br>be quite helpful. Bear in mind, I don't even know the VMS analog to<br>'ls'. :)<br><br>Any VMS Elmers that can offer some advice would be greatly appreciated.<br><br></blockquote></div><br><br clear="all"><br>-- <br>Ajilon Consulting u <br>Site resident at <br>Quest Diagnostics<br>first.x.last@f$edit(contents of previous line,&quot;COLLAPSE&quot;).com   * ------=_Part_2811_27369825.1163394207024--   ------------------------------    Date: 12 Nov 2006 18:25:01 -0800+ From: "madcrow" <madcrow.maxwell@gmail.com> ' Subject: Re: Most "classic" VMS version B Message-ID: <1163384701.711946.47220@i42g2000cwa.googlegroups.com>  G Hmm, so it looks like the consensus seems to be split between 5.5-2 and D "just use the newest version you can get". That means I'll be tryingG 5.5-2 for now. Eventually, I'll get a get the most recent VAX version I C can and compare... Might actually make for an interesting write-up, C although I guess DEC (or by that point Compaq, I guess) already did  something like that...   ------------------------------    Date: 12 Nov 2006 22:07:40 +01006 From: peter@langstoeger.at (Peter 'EPLAN' LANGSTOEGER) Subject: Re: Novice's questions , Message-ID: <45579b2c$1@news.langstoeger.at>  z In article <890539d90611110004y3467693eh598c47239c93a5f1@mail.gmail.com>, "Carl Friedberg" <frida.fried@gmail.com> writes:M >TWlrZSwKCk9uIDEwIE5vdiAyMDA2IDIzOjQwOjI4IC0wODAwLCBNWk4gPE1pa2VabW5AZ21haWwu M >Y29tPiB3cm90ZToKPgo+ICIiIlN0ZXBoZW4gSG9mZm1hbiDQydPBzCjBKToKPiAiIiIKPgo+ID4g M >ICAgSWYgdGhpcyBpcyB0aGUgdHlwaWNhbCBob2JieWlzdCBsaWNlbnNlLCBleHRyYWN0IHRoZSBj  >...   What language is this?  N Do I really need to MX_DECODE this garbage by hand? Please use text only here.   --   Peter "EPLAN" LANGSTOEGER % Network and OpenVMS system specialist  E-mail  peter@langstoeger.atF A-1030 VIENNA  AUSTRIA              I'm not a pessimist, I'm a realist   ------------------------------  % Date: Mon, 13 Nov 2006 00:57:38 +0800  From: prep@prep.synonet.com @ Subject: Re: Performance comparison Alpha ES40 vs Itanium rx36000 Message-ID: <87d57s4n7h.fsf@k9.prep.synonet.com>  , "Syltrem" <syltremzulu@videotron.ca> writes:  C > That's ok, we'll keep the Alpha around for some (long) time after  > installing the Integrity.   C Don't, go find some ES45s or 47s and split the AS load onto another ' machine. Save you a lot of pain `soon'.    --  < Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda. @                                              West Australia 6076* comp.os.vms,- The Older, Grumpier Slashdot. Raw, Cooked or Well-done, it's all half baked.F EPIC, The Architecture of the future, always has been, always will be.   ------------------------------  % Date: Sun, 12 Nov 2006 21:14:09 -0500 ( From: Bill Todd <billtodd@metrocast.net>@ Subject: Re: Performance comparison Alpha ES40 vs Itanium rx3600G Message-ID: <lNadnTFdfPVvScrYnZ2dnUVZ_vadnZ2d@metrocastcablevision.com>    George Cook wrote:   ...   H > If I recall correctly, even the simplest IA64 instruction is 128 bits.  C I think you recall incorrectly:  doesn't Itanic place three 41-bit  $ instructions in each 128-bit bundle?  G > EPIC is/was a poorly thought out concept.  I still cannot believe the A > number of people who actually fell for all the hype or that any = > competent CPU designer actually thought it was a good idea.   H Well, you ought to keep in mind the fact that the basic Itanic approach A was designed in the late '80s (ISTR that the HP project began in  G December, 1988).  At that time, and in fact until about 1995 (when the  > serious difficulties in Itanic began to become evident to the H engineering groups), there was no real proof that RISC would succeed in G its quest for real benefits from out-of-order execution, nor that IA32  J would succeed in becoming competitive (let alone a leader) in performance.  I EPIC was an alternative (though also risky) approach to what at the time  D appeared to be an otherwise-intractable problem, and it's still not H clear that a *good* design and implementation would be anything like as E much of a dog as Itanic has been.  Unfortunately, since the ex-Alpha  G team got booted off Tukwila (which may be the end of the Itanic line),   we may never even get a hint.   H Of course, the blatantly unrealistic hype that has surrounded Itanic in . *this* decade has no justification whatsoever.   - bill   ------------------------------  % Date: Sun, 12 Nov 2006 23:14:11 -0500 - From: bradhamilton <bradhamilton@comcast.net> N Subject: Problem with sys$system:AXPVMS$PCSI_INSTALL_MIN.COM (VMS V8.3, Alpha)* Message-ID: <4557F113.8070703@comcast.net>   Has anyone seen this problem?   C Installed minimal boot on a non-system disk.  Installation went as  I expected (I've done this a number of times on V7.2X/7.3X systems without   issue).   C When I went to boot from the "alternate" boot disk, I received the   following at the console:    >>>>b -fl e,0 dkb600 > % > System will be reset prior to boot.  >  > B > ff.fe.fd.fc.fb.fa.f9.f8.f7.f6.f5.CPU 0 speed is 2.31 ns (433MHz) > # > ef.ee.ed.ec.eb.f4.ea.e9.e8.e7.e6. 4 > ewa0: link up : Negotiated  100BaseTX: full duplex > e5.e4.$ > Digital Personal WorkStation 433au > CPU 0 booting  > % > (boot dkb600.6.0.1009.0 -flags e,0) 4 > block 0 of dkb600.6.0.1009.0 is a valid boot block, > reading 1226 blocks from dkb600.6.0.1009.0 > bootstrap code read in5 > base = 1d8000, image_start = 0, image_bytes = 99400  > initializing HWRPB at 2000# > initializing page table at 1ca000  > initializing machine state% > setting affinity to the primary CPU  > jumping to bootstrap code 1 > %APB-I-FILENOTLOC, Unable to locate SYSBOOT.EXE H > %APB-I-LOADFAIL, Failed to load secondary bootstrap, status = 00000910 >  > halted CPU 0 >  > halt code = 5  > HALT instruction executed  > PC = 20004048          > warning -- HWRPB is invalid.  E Since I am a hobbyist, I can't report this through "formal" channels.    ------------------------------    Date: 12 Nov 2006 21:56:25 +01006 From: peter@langstoeger.at (Peter 'EPLAN' LANGSTOEGER) Subject: Re: TIE/MSAI , Message-ID: <45579889$1@news.langstoeger.at>  S In article <00A5E796.D066F448@SendSpamHere.ORG>, VAXman-  @SendSpamHere.ORG writes: G >Does anybody know of if there a place where I can download viable kits  >for these?   ! They are only downloadable by HP. D Problem is, that some of these kit are not packaged there. Complain!  L This is one reason (the other is the original dates of the .PCSI* kit files)F why I always want packaged (BACKUP, DCX or most preferred ZIPEXE) kitsF and at least the VMS ECOs (along with many kits for years now) are now ZIPEXE.    --   Peter "EPLAN" LANGSTOEGER % Network and OpenVMS system specialist  E-mail  peter@langstoeger.atF A-1030 VIENNA  AUSTRIA              I'm not a pessimist, I'm a realist   ------------------------------    Date: 12 Nov 2006 21:58:21 +01006 From: peter@langstoeger.at (Peter 'EPLAN' LANGSTOEGER) Subject: Re: TIE/MSAI , Message-ID: <455798fd$1@news.langstoeger.at>   In article <1163102423.292335.181900@b28g2000cwb.googlegroups.com>, "johnhreinhardt@yahoo.com" <johnhreinhardt@yahoo.com> writes: F >Have you tried checking the RMS record format to verify that they areH >correct? All of the PCSI$COMPRESSED files I have are seq, fixed 512byte >records  3 Indeed. And the .PCSI files are seq, fixed 8192byte    --   Peter "EPLAN" LANGSTOEGER % Network and OpenVMS system specialist  E-mail  peter@langstoeger.atF A-1030 VIENNA  AUSTRIA              I'm not a pessimist, I'm a realist   ------------------------------  + Date: Mon, 13 Nov 2006 04:26:24 +0000 (UTC) , From: cdl@deeptow.ucsd.edu (Carl Lowenstein)" Subject: Re: Where to get a cable?* Message-ID: <ej8s5g$1j62$3@news1.ucsd.edu>  8 In article <lbw2.2h4ufy@no-mx.forums.yourdomain.com.au>,9 lbw2  <lbw2.2h4ufy@no-mx.forums.yourdomain.com.au> wrote:  > % >The Grue - James T. Sprinkle Wrote:  H >> Where is a good place to get a resonably priced mmj <-> db9f cable? I >> just H >> got a MicroVAX 3100 and it is just sitting there waiting for me to do >> something with it. : )  >>   >> The Grue  >*MMJ to DB-9 Female* G >This hooks a PC's serial port to a Digital MMJ device (such as a VT220 @ >terminal).  Digital made an adapter called a H8571-J with these
 >pinouts: ( >-MMJ                      DB-9 Female- B >1  Orange                1  Connect to pins 6 and 8 *and* White / >orange # >2  Blue                  2  Green  " >3  White / blue          3  Blue $ >4  White / green         4  Orange > >5  Green                 5  White / blue *and* White / green B >6  White / orange        6  Connect to pins 1 and 8 *and* White / >orange  >7  No connection ) >8  Connect to pins 1 and 6 *and* White /  >orange  >9  No connection  >  >Use them all the time.   . There is something peculiar about this table. D The column labeled MMJ has 9 entries, the MMJ connector has 6 pins. L The column labeled DB-9 has 6 entries, the connector called DB-9 has 9 pins.  - Maybe I just don't understand how to read it.        carl --  B     carl lowenstein         marine physical lab     u.c. san diegoB                                                  clowenst@ucsd.edu   ------------------------------  % Date: Mon, 13 Nov 2006 00:19:06 +0100 9 From: Albrecht Schlosser <ajs567XNewsX-XNewsX@tiscali.de> ? Subject: Re: writing fixed size records with one shorter record $ Message-ID: <ej8a4e$1l2$1@online.de>   Hein RMS van den Heuvel wrote: > Albrecht Schlosser wrote: ! >> Hein RMS van den Heuvel wrote:  >>> Albrecht Schlosser wrote: = >>>  -->  FAT$V_RTYPE=FAT$C_UNDEFINED     FAB$B_RFM=FAB$C_UDF R >> I understand "FAB$B_RFM=FAB$C_UDF", but what does "FAT$V_RTYPE=FAT$C_UNDEFINED"; >> mean? (guess: the same, but what language/tool/context?)  > I > Interesting you picked up on that! It was a deliberate cut & paste from  > the SET FILE /ATTR help text.   	 Ah, I see   4 > The RFM=UDF referds to DCL dealing with those bits  % that was obvious, I didn't mention it   6 > The FAB stuf is the RMS language as you surmised and   right   I > The FAT stuff refers to the ACP QIO interface to those same attributes.   C I didn't know that, never had to do with the ACP QIO interface. But  surely you did ;-)    H >  > I think that I will give that a try. Just found a piece of my own C > code that Q >> writes a (real fixed length) file with sys$put. Setting the correct attributes  >> should not be a big deal, > 4 > Correct. Coudl be done while initializing the FAB. > ? >> but I do hope that the resulting file would then be  usable.  >  > That will be the big test.I > Some standard tools ('search if I recall correctly') don't care to deal  > with UDF files. F > You can always SET FILE/ATT=(RFM=FIX,LRL=512,MRS=512) afterwards :-)  D I thought of that, too. But how would it be possible from within my F program that writes the file? What would I have to do then? Close the F file, reopen it, set the FAB fields and close it again? Would that do + the trick, or would there be more involved?   I Or would I set the attributes to RFM=UDF before sys$create, then sys$put  A all the records, and just reset to RFM=FIX just before sys$close?   E I can see that XAB$L_EBK and XAB$W_FFB are XAB fields, but if I write E undefined format records, then I think that I wouldn't need to modify H these fields. The only fields I think that would have to be modified are!   FAB$W_MRS and FAB$B_RFM, right?   G Maybe also XAB$W_LRL, if I wrote records longer than 512 Bytes, but not " if I write no more than 512 Bytes?  2 Thanks for your help, I really do appreciate that!   Albrecht   ------------------------------    Date: 12 Nov 2006 16:26:37 -0800< From: "Hein RMS van den Heuvel" <heinvandenheuvel@gmail.com>? Subject: Re: writing fixed size records with one shorter record C Message-ID: <1163377597.358820.184540@f16g2000cwb.googlegroups.com>    Albrecht Schlosser wrote:   J > Or would I set the attributes to RFM=UDF before sys$create, then sys$putC > all the records, and just reset to RFM=FIX just before sys$close?   G Yes! ... but you have to say 'pretty please' before that through a call G to the  (undocumented) SYS$MODIFY. That is a whole alternative solution E as it would allow you to change the record size as needed. The LINKER E uses this to be able to write fixed-length-record image data followed * by variable length debug records as neeed. See sample usage below.   A I don't like Bob's suggestion of 'stream file'. It mangles binary $ files.  Is that what tiff files are?= RMS Stream files will drop leading binary zero bytes on $PUT. < And could those TIFF files not have too many, or too few LFs  G Jack could well be right on the formatted/unformated io suggestion, but F i have not played with fortran IO in a while. Some experiments needed.   Program follows. Enjoy! Hein.    #include <stdio> #include <rms> #include <rmedef>  main (int argc, char *argv[])  {  struct FAB	fab;  int		s;   < fab = cc$rms_fab;			/* Make this a real FAB	(bid and bln) */ fab.fab$b_fac = FAB$M_PUT; fab.fab$l_fna = argv[1];" fab.fab$b_fns = STRLEN( argv[1] ); s = sys$open(&fab);      if (!(s & 1)) return s;  fab.fab$l_ctx = RME$C_SETRFM;  fab.fab$v_esc = 1;8 if (argc > 2) {  sscanf(argv[2], "%d", &fab.fab$w_mrs );!     } else fab.fab$w_mrs = 32256;  fab.fab$b_rfm = FAB$C_FIX; s=sys$modify(&fab);      if (!(s & 1)) return s;  return sys$close(&fab);  }              > G > I can see that XAB$L_EBK and XAB$W_FFB are XAB fields, but if I write G > undefined format records, then I think that I wouldn't need to modify J > these fields. The only fields I think that would have to be modified are# >   FAB$W_MRS and FAB$B_RFM, right?  > I > Maybe also XAB$W_LRL, if I wrote records longer than 512 Bytes, but not $ > if I write no more than 512 Bytes? > 4 > Thanks for your help, I really do appreciate that! > 
 > Albrecht   ------------------------------    Date: 12 Nov 2006 16:50:23 -0800< From: "Hein RMS van den Heuvel" <heinvandenheuvel@gmail.com>? Subject: Re: writing fixed size records with one shorter record C Message-ID: <1163379023.217305.165660@h54g2000cwb.googlegroups.com>    JF Mezei wrote:   > Hein RMS van den Heuvel wrote:K > > Yes! ... but you have to say 'pretty please' before that through a call & > > to the  (undocumented) SYS$MODIFY. >  > D > After all those years, how come SYS$MODIFY is still undocumented ?  E Because some of its usages would make it all too easy to shoot onself  in the foot.B The (possible) added value of adding documenting the call does notD weigh up against the potential cost fielding support calls for this.  G It probably also only has minimal error handling as it only is supposed E to have know and knoledgeable users. Documenting it would also mean a + full code review and regression test suite.    Just my best guess,  Hein.    ------------------------------  + Date: Mon, 13 Nov 2006 02:25:31 +0000 (UTC) 7 From: moroney@world.std.spaamtrap.com (Michael Moroney) ? Subject: Re: writing fixed size records with one shorter record ( Message-ID: <ej8l2r$luj$1@pcls4.std.com>  ; Albrecht Schlosser <ajs567XNewsX-XNewsX@tiscali.de> writes:    >Hein RMS van den Heuvel wrote: G >> You can always SET FILE/ATT=(RFM=FIX,LRL=512,MRS=512) afterwards :-)   E >I thought of that, too. But how would it be possible from within my  G >program that writes the file? What would I have to do then? Close the  G >file, reopen it, set the FAB fields and close it again? Would that do  , >the trick, or would there be more involved?  I Note that if you do that SET FILE on a file with a partially filled final F block and read it in later, a read of the final block will return 512 I bytes, ignoring the end-of-file offset.  This means you'll get zeroes (or C junk) following the end of the data of the partial block.  I'm sure J that a fixed length record file with the EOF byte not at the end of a fullF record is totally unsupported, and the behavior I saw may change from  version to version of VMS.   ------------------------------  % Date: Sun, 12 Nov 2006 21:48:14 -0500 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> ? Subject: Re: writing fixed size records with one shorter record 7 Message-ID: <22e5$4557dcbb$cef8887a$13558@TEKSAVVY.COM>    Hein RMS van den Heuvel wrote:E >> After all those years, how come SYS$MODIFY is still undocumented ?  > G > Because some of its usages would make it all too easy to shoot onself  > in the foot.  + Is it more dangerous than SET FILE/ATTRIB ?    ------------------------------   End of INFO-VAX 2006.624 ************************