1 INFO-VAX	Mon, 05 Mar 2001	Volume 2001 : Issue 127       Contents:$ Automate Startup of DECvoice App.'s? Re: dec 3000 300 crashing  Re: dec 3000 300 crashing ' Re: Drop C/C++ better sooner than later  Further APAC details9 Re: How to convert an indexed file to a sequential file?? 4 Re: How to licence boot strapping a hobbyist system? Re: Just Two Days Left to Vote1 RE: Low Cost Alpha Price Drop - Now $850 Complete 1 RE: Low Cost Alpha Price Drop - Now $850 Complete # Re: OpenVMS and cluster reliability  Oracle/OVMS7.1 Memory tuning?  Piping / file filters  Re: Piping / file filters   F ----------------------------------------------------------------------  % Date: Sun, 04 Mar 2001 21:39:50 -0600 7 From: "David J. Dachtera" <djesys.nospam@earthlink.net> - Subject: Automate Startup of DECvoice App.'s? - Message-ID: <3AA30A86.AD8BA0EA@earthlink.net>   G Does anyone know how to automate the startup of a DECvoice application?   A Naturally, I inherited no DECvoice documentation when I got a new E full-time gig last fall. So, I'm hoping someoneone out there has some  useful expertise.   B The current procedure requires the manual use of the menu-orientedF VAE$AM program. I'm hoping someone knows what this program does behindF the scenes and whether the proce.'s and programs it runs "for you" can; be automated and run during the system startup ro in batch.   / Anyone have anything they'd care to pass along?    --   David J. Dachtera  dba DJE Systems  http://www.djesys.com/  : Unofficial Affordable OpenVMS Home Page and Message Board: http://www.djesys.com/vms/soho/   F This *IS* an OpenVMS-related newsgroup. So, a certain bias in postings is to be expected.  @ Feel free to exercise your rights of free speech and expression.  F However, attacks against individual posters, or groups of posters, are strongly discouraged.    ------------------------------  % Date: Sun, 04 Mar 2001 17:26:12 -0500 2 From: rdeininger@mindspring.com (Robert Deininger)" Subject: Re: dec 3000 300 crashingL Message-ID: <rdeininger-0403011726120001@user-2ivea72.dialup.mindspring.com>  ; In article <3AA212BD.D3E644A3@infopuls.com>, Christof Brass  <brass@infopuls.com> wrote:   I > This is a late follow up to the firmware memory test problem. I finally F had to shut down my system and could perform these tests. Here are the results: >  > 1.Test > 5 > T-STS-MEM - LLSC Test-Addr 01000000 FWD Wr 00000000 2 > ? T-ERR-MEM - stl_c bcache hit at Addr: 013fc154) > T-STS-MEM - Uncorrected Error count = 1 ; > ?-T-ERR-MEM - FAILED status = 20 test hit addr = 013fc154  > ?? 801    MEM 0x0020 >  >     84    FAIL >  > 2.Test > 5 > T-STS-MEM - LLSC Test-Addr 01800000 FWD Wr 00000000 > > ? T-ERR-MEM - stlc bcache miss with victim at Addr: 019f2374) > T-STS-MEM - Uncorrected Error count = 1 ; > ?-T-ERR-MEM - FAILED status = 20 test hit addr = 019f2374  > ?? 801    MEM 0x0020 >   	 Christof,   D    I found an obscure note in the firmware update instructions.  The7 memory test on the DEC 3000 300 will fail unless you do    >>> SET MOP OFF   I before the memory test.  I confirmed this behavior on two of our systems.   N I believe the operation of the MOP software interferes with the LLSC function.   --   Robert Deininger rdeininger@mindspring.com    ------------------------------  % Date: Mon, 05 Mar 2001 04:07:11 +0000 ) From: Christof Brass <brass@infopuls.com> " Subject: Re: dec 3000 300 crashing, Message-ID: <3AA310EF.ECFE0523@infopuls.com>   Robert Deininger wrote:  > = > In article <3AA212BD.D3E644A3@infopuls.com>, Christof Brass  > <brass@infopuls.com> wrote:  > K > > This is a late follow up to the firmware memory test problem. I finally H > had to shut down my system and could perform these tests. Here are the
 > results: > > 
 > > 1.Test > > 7 > > T-STS-MEM - LLSC Test-Addr 01000000 FWD Wr 00000000 4 > > ? T-ERR-MEM - stl_c bcache hit at Addr: 013fc154+ > > T-STS-MEM - Uncorrected Error count = 1 = > > ?-T-ERR-MEM - FAILED status = 20 test hit addr = 013fc154  > > ?? 801    MEM 0x0020 > >  > >     84    FAIL > > 
 > > 2.Test > > 7 > > T-STS-MEM - LLSC Test-Addr 01800000 FWD Wr 00000000 @ > > ? T-ERR-MEM - stlc bcache miss with victim at Addr: 019f2374+ > > T-STS-MEM - Uncorrected Error count = 1 = > > ?-T-ERR-MEM - FAILED status = 20 test hit addr = 019f2374  > > ?? 801    MEM 0x0020 > >  >  > Christof,  > F >    I found an obscure note in the firmware update instructions.  The9 > memory test on the DEC 3000 300 will fail unless you do  >  > >>> SET MOP OFF  > K > before the memory test.  I confirmed this behavior on two of our systems.  > P > I believe the operation of the MOP software interferes with the LLSC function. >  > -- > Robert Deininger > rdeininger@mindspring.com   < Sorry, I forgot that although I read it after you posted it.   Sorry for this accumulation of tautologies (of course I can only forget what I knew and read something after it has been posted ;-) ).   Before I posted my results I searched the Google Usenet archive but there were only two messages about this topic and my news reader rendered all the messages of this thread as unavailable.    I'll repeat the test with SET MOP OFF if I can shutdown the system (too many windows normally). I don't understand why the test failed the first two times and then succeeded.Robert Deininger wrote:    ------------------------------  % Date: Mon, 05 Mar 2001 09:07:43 +0010 % From: paddy.o'brien@zzz.tg.nsw.gov.au 0 Subject: Re: Drop C/C++ better sooner than later5 Message-ID: <01K0TZ1AYLLE00AJ83@tgmail.tg.nsw.gov.au>    Brian,  7 >Thanks Paddy but this doesn't answer the Why.  Why, I   >want to know, does : >the "debugging on VAX for this very reason" need to be?     [snips]   0 >routine in M64 is the way to go.  What I don't  >understand is the debugging	 >comment.   7 That was Dave Weatherall's comment.  I've never looked  8 at the machine code in a debug session, my mistakes are  probably too obvious :-(  8 Since we use F90/F95 code, we can only develop on Alpha 5 now, so for reasons previously mentioned, I have not  7 bothered with looking at the machine for quite a while.    Regards, Paddy   Paddy O'Brien, Transmission Development, 
 TransGrid, PO Box A1000, Sydney South,  NSW 2000, Australia    Tel:   +61 2 9284-3063 Fax:   +61 2 9284-3050& Email: paddy.o'brien@zzz.tg.nsw.gov.au  8 Either "\'" or "\s" (to escape the apostrophe) seems to  work for most people, 6 but that little whizz-bang apostrophe gives me little  spam.    ------------------------------   Date: 4 Mar 2001 15:15:37 -0500 + From: young_r@encompasserve.org (Rob Young)  Subject: Further APAC details 3 Message-ID: <wdjrNaAaL5Yu@eisner.encompasserve.org>   D Came upon the following found in a recent SKC.  Thought others might like a gander.   Rob       I Compaq Australia and the Australian Partnership for Advanced Computing on M February 14 reported that Compaq will provide Alpha supercomputing technology N to the APAC National Facility in Canberra, Australia. The agreement was signedH between Compaq and The Australian National University which acts as hostJ institution for APAC. Under the three year alliance, Compaq will supply anG AlphaServer SC system including more than 450 Alpha CPUs, together with H system and development software. Compaq will also work with APAC and itsN partners to provide expertise and solutions for Australian research, educationG and commercial firms. The initial system, consisting of top-of-the-line M AlphaServer ES4x quadprocessor nodes, will be operational in April 2001. When I completed in October, APACs AlphaServer SC will rank in the top 60 of the C worlds most powerful computing systems while serving as the largest ? supercomputer available to Australian researchers and industry.   J A Cachegate Comeuppance? For sure, mate! The February 14 APAC announcementL serves as a bona fide St. Valentines Day Massacre for the cosmic-ray-plaguedK StarFire enterprise server. As previously reported in SKC, Sun Microsystems L was the incumbent in the APAC account, but Le Roi Soleils four-pack of UltraO Enterprise 10000 StarFire servers failed to cut the mustard or pass its initial K acceptance  test so APAC renewed its Hunt For High Performance by reopening N bids last October. Compaq HPTC VP Bill Blake and his team forthwith supplantedJ the StarFires with a Better Answer: We chose to partner with Compaq on theG basis of it being the company in Australia with the most cost-effective M solution to the computing capacity we need, explained APAC executive director  Professor John O'Callaghan.    ------------------------------  $ Date: Mon, 5 Mar 2001 13:19:14 +1100/ From: "Phil Howell" <phowell@snowyhydro.com.au> B Subject: Re: How to convert an indexed file to a sequential file??1 Message-ID: <uGCo6.2931$o4.128201@ozemail.com.au>   D Larry Kilgallen <Kilgallen@eisner.decus.org.nospam> wrote in message- news:aHNh2px7SK6i@eisner.encompasserve.org... = > In article <t9tv57rtj1fla5@corp.supernews.com>, "Bill Ames"  <billames@accunet.net> writes:E > > Im trying to convert an indexed file to a sequential file so that  records F > > can be merged faster.  I would like to use a DCL command for this. > 9 > Is this a contest ?  Is there a prize for most-arcane ?  > , No because the TECO method would always win.   ------------------------------  % Date: Mon, 05 Mar 2001 04:29:07 +0000 ) From: Christof Brass <brass@infopuls.com> = Subject: Re: How to licence boot strapping a hobbyist system? , Message-ID: <3AA31613.AC7AC33E@infopuls.com>   Many Thanks!  CSA is on my list. The hobbyist system has now the DW-MOTIF licence PAK but decided to change its behaviour. After booting the basic things the monitor goes off (black screen) when the DECwindows login prompt should be displayed. If I log in remotely I SHOW SYSTEM shows a DTLOGIN and a DTGREET process. If I blindly put in SYSTEM/<password> these two processes disappear but then a DTSESSION and a DTWM is visible (unfortunately only with SHOW SYSTEM not on the screen). The first time after licensing OPENVMS-ALPHA the DECwindows login prompt was displayed nicely. Then I switched off the machine and interrupted the boot process just before the DECwindows prompt would have appeared, entered the licence PAKs in character mode and rebooted. 5 Has anybody an idea why the screen switched to black?    ------------------------------    Date: 04 Mar 2001 16:45:15 -08003 From: Eric Smith <eric-no-spam-for-me@brouhaha.com> ' Subject: Re: Just Two Days Left to Vote 0 Message-ID: <qhvgppavs4.fsf@ruckus.brouhaha.com>  ; Kilgallen@eisner.decus.org.nospam (Larry Kilgallen) writes: F > 	"JavaScript is the only way to provide the proper user experience."  D Ah!  So they consider the "proper user experience" to be one that is? only available to those who are (A) not security-conscious, and " (B) run certain selected browsers.  C I've run surveys on web sites without JavaScript, and I didn't have ? any problem whatsoever in providing a "proper user experience".   H I think what they really mean is a "cool-looking whizzy user interface", which I have no use for.   ------------------------------  % Date: Sun, 04 Mar 2001 16:33:38 -0600 + From: "Main, Kerry" <Kerry.Main@compaq.com> : Subject: RE: Low Cost Alpha Price Drop - Now $850 CompleteR Message-ID: <BE56C50EA024184DAF48F0B9A47F5CF44C886C@kaoexc01.americas.cpqcorp.net>   Hunter,   D Just remembered that a few weeks ago, I removed my Netscape Java andJ Javascript capability. Just turned it back on and indeed the site is a bit
 messed up.  I Simple test - Netscape - edit - preferences - advanced .. leave "Java" on # and click OFF "Enable Javascript".    E Now, reload Netscape, reaccess site and see if it is still messed up.    Regards,  
 Kerry Main Senior Consultant  Compaq Canada Inc. Professional Services  Voice: 613-592-4660  Fax  :  819-772-7036 Email: Kerry.Main@Compaq.com     -----Original Message-----4 From: Hunter Goatley [mailto:goathunter@goatley.com] Sent: March 4, 2001 1:36 PM  To: Main, Kerry 3 Cc: 'goathunter@goatley.com'; Info-VAX@Mvb.Saic.Com : Subject: RE: Low Cost Alpha Price Drop - Now $850 Complete    , "Main, Kerry" <Kerry.Main@Compaq.com> wrote: > I >It would appeat the site has been fixed as I just accessed with Netscape & >V4.76 and Opera 5.0 with no problems. > E Very strange.  I'm also using Netscape V4.76; the display today isn't F the same as what I was getting last night, but it's still not correct.E I get the pictures displayed, but no text except at the bottom of the E page.  "View Source" shows that it's all there, it's just not getting  displayed properly.   @ Must be something funky about my Netscape installation.  Thanks.   Hunter ------9 Hunter Goatley, Process Software, http://www.process.com/ ; <goathunter@GOATLEY.COM>     http://www.goatley.com/hunter/    ------------------------------  + Date: Sun, 04 Mar 2001 18:24:04 -0600 (CST) - From: Hunter Goatley <goathunter@goatley.com> : Subject: RE: Low Cost Alpha Price Drop - Now $850 Complete, Message-ID: <01K0T4BT1HJI8WVYSB@goatley.com>  , "Main, Kerry" <Kerry.Main@Compaq.com> wrote: >  >Hunter, > 
 Hi, Kerry.  E >Just remembered that a few weeks ago, I removed my Netscape Java and K >Javascript capability. Just turned it back on and indeed the site is a bit  >messed up.  > B Yep, disabling Javascript causes the page to display, although the+ display is different from that shown in IE.   B This probably explains the problem I've seen with other pages thatA load OK in IE but not Netscape.  It's especially frustrating when ( you're trying your best to never use IE!   Thanks, Kerry.   Hunter ------9 Hunter Goatley, Process Software, http://www.process.com/ ; <goathunter@GOATLEY.COM>     http://www.goatley.com/hunter/    ------------------------------  $ Date: Mon, 5 Mar 2001 13:33:01 +1100/ From: "Phil Howell" <phowell@snowyhydro.com.au> , Subject: Re: OpenVMS and cluster reliability1 Message-ID: <nTCo6.2937$o4.128592@ozemail.com.au>   0 Alan Greig <a.greig@virgin.net> wrote in message2 news:9unu9tco5af3nhkr9ij4u532rh0u6oll49@4ax.com...2 > On Fri, 2 Mar 2001 19:27:21 +1100, "Phil Howell"$ > <phowell@snowyhydro.com.au> wrote: > A > >The situation in 2001 is that tcp/ip is everywhere, almost any I > >system/application can stay up 99.9% of the time (if properly managed) < > >and you can't buy any erp type systems developed for vms.F > >(ie. systems that are cluster-aware, use the lock manager properly, > >and can do hot backups) > G > The aging MANMAN is fully cluster aware, uses DBMS and RDB and can do            ^^^^^ 8 I think that says enough - we have aging systems alreadyD > hot backups ($ DBO/BACKUP/MULTI/ONLINE) and is still developed andA > sold by CA on VMS. In the last 12 months we've had no unplanned                  ^^^  I think that says a bit more@ > downtime across the cluster related to VMS. We've had only oneH > software crash of one node which took it down for 4 mins 48 secs as itF > rebooted (LOCKMGRERR - known problem now worked around with the helpH > od DSN) Cluster uptime was almost at one full year (went live on AlphaC > switching from VAX in March 2000) when  a serious extreme weather D > related power outage in the middle of this Monday night eventuallyC > defeated our UPS and generator (and defeated my ability to get to D > work). System recovered itself unattended on power up but with twoD > disk failures (out of 24 19G disks) probably due to temperature inH > machine room dropping near freezing. Spares mapped into the mirrorsetsG > as replacements by the controller automagically. Compaq field service B > could not even get into or out of their offices (Livingston nearE > Edinburgh) let alone travel the 15 miles to reach us until the next  > day  > H > The NT team were still struggling with a backlog of problems caused byG > this outage four days later.. No major problems for the Unix team. As # > usual NT caused the most trouble.  >    ------------------------------  % Date: Mon, 05 Mar 2001 09:48:20 +0800 5 From: Netsurfer <netsurfer@sentosa.singaporemail.com> & Subject: Oracle/OVMS7.1 Memory tuning?8 Message-ID: <5nr5atsmfg62764t53mrkilk89o7d7rtsd@4ax.com>  2 Just upgraded from 256MB to 640MB on my Alpha4000.  F Without the original OpenVMS SysAdmin manual for the OpenVMS system, IB am totally lost in tuning the system after the increase of memory.  : Most of the parameters in MODPARAM.DAT are hardcoded by myA predecessor, and I don't really know if their values required anyPB modification. Running AUTOGEN don't really help as most of the key parameters had been fixed.  E I also having problems with tuning the Oracle 7.3.4, our only DBA hadtD left our company, and I am a non-qualified stand-in DBA. Where can IF get help for the SGA parameter tuning, any website with information ofA such? We don't have any maintenance support contract with Oracle.r  
 Thank you.        - Regards,  	 Netsurfer-        ====R For any personal email replies, please remove " sentosa. " from my E-mail address.   ------------------------------  % Date: Sun, 04 Mar 2001 21:07:53 -0500s- From: JF Mezei <jfmezei.spamnot@videotron.ca>t Subject: Piping / file filters, Message-ID: <3AA2F4E3.AECE1B32@videotron.ca>  A The postcript language has a feature which I find quite valuable.   K When reading data from a file ( usually its equivalent to SYS$INPUT ), thatl< data is usually encoded using different encoding mechanisms.  J What the language allows you to do is to "stack" the decoders (or encoders5 when you are writing to a file) on top of each other.   M For instance, I can have a basic ascii filter with a self-defined enf-of-file N market such as "JF-LOVES-CHOCOLATE". Inside that file, the data may be encodedL as ASCII-85 or pure hexadecimal, and inside of that, the data may be encoded and comressed as JPEG data.   K When the "image" command requests to read some data, the data automaticallyoK passes through all those filters and "image" never knows nor sees the asciis; data in the file, it only sees an uncompressed bitmap data.c  L The data does not go through intermediate storage, hence the use of the termM "filter". Another neat feature is that you can define the file to be an arrayrM that contains your image or the standard input stream (or an actual file) ands, this is transparent to the "image" operator.  N If the image statement asks for 25 bytes, then it will be given 25 bytes, evenK if at the other end, 50 bytes are read from the file. (or vice versa, if atp< the end, only 20 bytes are read and then uncompressed to 50)  G In the case of VMS, RMS provides a very basic filtering when you access J indexed files with sequential commands, or when you attempt to read "text" from a DDIF file.e  N I am curious if there would be any standard/established ways to implement such3 filtering inside a programming language such as C ?8   ------------------------------  % Date: Mon, 05 Mar 2001 04:15:45 +0000C) From: Christof Brass <brass@infopuls.com>-" Subject: Re: Piping / file filters, Message-ID: <3AA312F1.AE87FAC9@infopuls.com>   JF Mezei wrote:  > C > The postcript language has a feature which I find quite valuable.  > M > When reading data from a file ( usually its equivalent to SYS$INPUT ), thatn> > data is usually encoded using different encoding mechanisms. > L > What the language allows you to do is to "stack" the decoders (or encoders7 > when you are writing to a file) on top of each other.t > O > For instance, I can have a basic ascii filter with a self-defined enf-of-filegP > market such as "JF-LOVES-CHOCOLATE". Inside that file, the data may be encodedN > as ASCII-85 or pure hexadecimal, and inside of that, the data may be encoded > and comressed as JPEG data.1 > M > When the "image" command requests to read some data, the data automaticallyDM > passes through all those filters and "image" never knows nor sees the ascii = > data in the file, it only sees an uncompressed bitmap data.  > N > The data does not go through intermediate storage, hence the use of the termO > "filter". Another neat feature is that you can define the file to be an arrayrO > that contains your image or the standard input stream (or an actual file) and . > this is transparent to the "image" operator. > P > If the image statement asks for 25 bytes, then it will be given 25 bytes, evenM > if at the other end, 50 bytes are read from the file. (or vice versa, if at > > the end, only 20 bytes are read and then uncompressed to 50) > I > In the case of VMS, RMS provides a very basic filtering when you access L > indexed files with sequential commands, or when you attempt to read "text" > from a DDIF file.  > P > I am curious if there would be any standard/established ways to implement such5 > filtering inside a programming language such as C ?   & Better use an objectoriented language.) Better read "Design Patterns" by the GoF.s6 You could implement this with the "Decorator" pattern. Have fun doing this in C! ;-)n   ------------------------------   End of INFO-VAX 2001.127 ************************