1 INFO-VAX	Tue, 26 Apr 2005	Volume 2005 : Issue 232       Contents: 6052 monografia 6052 Re: DECwindows: pointer policy4 Re: How to install DCPS 2.4 without stopping queues?4 Re: How to install DCPS 2.4 without stopping queues?4 Re: How to install DCPS 2.4 without stopping queues?4 Re: How to install DCPS 2.4 without stopping queues?4 Re: How to install DCPS 2.4 without stopping queues?7 Re: HP and IntelR Developer Workshops Update / Reminder  IP Failover: strange behaviour" Re: IP Failover: strange behaviour Re: Legato NetWorker4 Maybe HP should get out of the hardware business....8 Re: Maybe HP should get out of the hardware business....8 Re: Maybe HP should get out of the hardware business....8 Re: Maybe HP should get out of the hardware business....) Re: MySQL and PHP Authentication Protocol 2 Optimising Alpha Cobol application on OpenVMS IA646 Re: Optimising Alpha Cobol application on OpenVMS IA643 Re: RAW IP socket problem, recvfrom() never returns 3 Re: RAW IP socket problem, recvfrom() never returns  TK50 Re: TK50 Re: TK50 Re: TK50 Re: TK50: Re: [ECP V5.5A] Unrecognized ECP/Collect data file version: Re: [ECP V5.5A] Unrecognized ECP/Collect data file version: Re: [ECP V5.5A] Unrecognized ECP/Collect data file version: Re: [ECP V5.5A] Unrecognized ECP/Collect data file version: Re: [ECP V5.5A] Unrecognized ECP/Collect data file version  F ----------------------------------------------------------------------  % Date: Tue, 26 Apr 2005 06:23:10 +0100 * From: "monografia " <monografia@xfdgr.com> Subject: 6052 monografia 6052 > Message-ID: <DDAC168FB5C09749AD96D27804CA35D105AD2BCD@SETUBAL>  5            Monografias completas a partir de 250,00 !     D   monografia administrao direito turismo economia letras pedagogia   monografias teses  tese.  ! Sua monografia a partir de 250,00  http://monografiadireta.rg3.net    ------------------------------   Date: 26 Apr 2005 07:15:05 GMT2 From: "Dave Weatherall" <djw-nothere@nospam.nohow>' Subject: Re: DECwindows: pointer policy ? Message-ID: <DTiotGxQ0bj6-pn2-MPrEVhfjgsS0@dave2_os2.home.ours>   B On Tue, 19 Apr 2005 07:41:11 UTC, helbig@astro.multiCLOTHESvax.de 1 (Phillip Helbig---remove CLOTHES to reply) wrote:   ; > In article <d41ak3$dpd$1@news01.intel.com>, Ken Fairfield H > Actually, I am not very familiar with Windows.  What I want is "focus E > follows pointer" (and, now, I have it), but if I want to raise the  J > window, it should be sufficient to click ANYWHERE on it and not just on 
 > the border.    Small war-story from last week.   E Telephone rings and colleague says 'Dave, ATT isn't responding to my  C commands'. I ask the usual obvious questions, can't see an obvious  F reason so go down the two flights of stairs (escalators actually but IE need some dramatic licence), into the rig (c.f. lab) and look at the  = workstation screen and see immediately what is going on. Two  D colleagues working together. The one who has _logged in_ has 'focus F follows pointer', the other is used to 'focus selected by click'. The E latter was the one operating our test system (ATT) and had done what  E he always does - click the window and move the mouse out of the way.  D No focus, no commands going to ATT and so no response. Result - big 5 grin from me and two slightly embarrassed colleagues.   ? I suppose its no worse than not noticing that 'Hold Screen' is  	 selected.    --   Cheers - Dave W.   ------------------------------    Date: 25 Apr 2005 23:57:19 -0700* From: karl.rohwedder@gmx.de (Cluster-Karl)= Subject: Re: How to install DCPS 2.4 without stopping queues? = Message-ID: <6436beb5.0504252257.67a1179a@posting.google.com>   w JF Mezei <jfmezei.spamnot@teksavvy.com> wrote in message news:<1114464495.e0561ba037a451e8a1875af26a4a8b1f@teranews>...  > Larry Fahnoe wrote: E > > Now, with the transition to the PCSI kit for DCPS 2.4 it looks as D > > though I do not have the option to manage the queues myself as IK > > outlined above, and instead I'd have to shut printing down cluster wide % > > as I upgraded each system disk.    >  > ? > Extract the contents of the .PCSI package, and build your own B > installation procedure from the .pcsi script and various command+ > procedures that would be included inside.   B DCPS executes a preconfigure script named DCPS$PCSI_CONFIGURE.COM,F which defines a logical name DCPS$PCSI_PRECONFIGURE_ABORT, if the userA doesn't allow the stopping of the queues. The product description C (HP-AXPVMS-DCPS-V0204--1.PCSI$DESCRIPTION) then checks this logical  name and aborts.> May be you can copy the kit in reference format (prod cop dcps? /dest=.../form=ref), edit the description to skip the abort and : install the kit from the reference copy. I assume, this is unsupported, of course.   
 regards Kalle    ------------------------------  # Date: Tue, 26 Apr 2005 12:03:56 GMT / From: "Jeff Goodwin" <jgoodwin@maine.rrr-r.com> = Subject: Re: How to install DCPS 2.4 without stopping queues? 6 Message-ID: <M6qbe.8515$mG3.5641@twister.nyroc.rr.com>  7 "Larry Fahnoe" <fahnoe@fahnoetech.com> wrote in message = news:1114457372.316850.294530@o13g2000cwo.googlegroups.com...  > Hello, > C > We have a DCPS cluster with high availability requirements around G > printing.  A long time ago I set this cluster up so that all the DCPS H > queues use /AUTOSTART_ON with multiple nodes per queue.  Each queue isI > serviced by nodes in different datacenters with different system disks. I >  In the past this has made DCPS upgrades transparent as I could migrate G > printing to one node, upgrade DCPS on the idle node, and then reverse B > the process to upgrade the remaining node.  I never let the DCPS= > upgrade stop all the queues as I was managing this process.  > C > Now, with the transition to the PCSI kit for DCPS 2.4 it looks as B > though I do not have the option to manage the queues myself as II > outlined above, and instead I'd have to shut printing down cluster wide H > as I upgraded each system disk.  This is a problem for us and I'd like= > to know if there is a way I can avoid the outage during the # > installation process.  Any hints?   L Other have mentioned modifing the PCSI install procedure to bypass the queueH stopping and that'll work fine.  I did converse with Paul Anderson afterJ DCPS V2.4 came out about this same issue.  I suspect we'll have some queue+ stopping/not-stopping options in DCPS V2.5.   L I've field tested quite a few DCPS releases, but unfortunately, skipped thisF one.  If your site is heavily reliant on DCPS, I suggest you look into becoming a field test site.    -Jeff    ------------------------------    Date: 26 Apr 2005 08:15:58 -0700, From: "Larry Fahnoe" <fahnoe@fahnoetech.com>= Subject: Re: How to install DCPS 2.4 without stopping queues? C Message-ID: <1114528558.381388.214810@f14g2000cwb.googlegroups.com>   D I had not played with modifying a PCSI kit before, so thanks for theG pointer to get me started.  Some quick edits to DCPS$PCSI_CONFIGURE.COM   solves the problem quite nicely.  ; If Paul is lurking here, I'll add my vote to have 2.5 allow ? installations to continue without stopping queues.  If there is D pressure to make sure queues on nodes being upgraded do get stopped,C then perhaps asking for a the list of nodes running off the current G system disk and using that list to filter queues to be stopped would be ? in order.  Frankly, I think it is fine to allow installation to  continue with a warning.   --Larry    ------------------------------  # Date: Tue, 26 Apr 2005 15:31:35 GMT * From: Paul Anderson <paul.anderson@hp.com>= Subject: Re: How to install DCPS 2.4 without stopping queues? 5 Message-ID: <260420051131386589%paul.anderson@hp.com>   C In article <1114528558.381388.214810@f14g2000cwb.googlegroups.com>, + Larry Fahnoe <fahnoe@fahnoetech.com> wrote:   F > I had not played with modifying a PCSI kit before, so thanks for the1 > pointer to get me started.  Some quick edits to : > DCPS$PCSI_CONFIGURE.COM solves the problem quite nicely.  : Yes, you have discovered the way to do this for DCPS V2.4.  = > If Paul is lurking here, I'll add my vote to have 2.5 allow 4 > installations to continue without stopping queues.  D As Jeff Goodwin mentioned, he contacted me about this a while back. > DCPS V2.5 will allow three options:  Stop all queues (the V2.4F behavior), stop queues using the current system disk (recommended) and! stop no queues (not recommended).   E > If there is pressure to make sure queues on nodes being upgraded do F > get stopped, then perhaps asking for a the list of nodes running offD > the current system disk and using that list to filter queues to be > stopped would be in order.  D The new procedure will determine which queues are running on systemsB that use the current system disk.  You can choose to see a list ofE queues that will be stopped.  There's no option to stop only a subset  of those queues.  F > Frankly, I think it is fine to allow installation to continue with a
 > warning.  @ We will provide enough rope in the kit so you can hang yourself.   Paul   --    Paul Anderson   OpenVMS Engineering    Hewlett-Packard Company    ------------------------------    Date: 26 Apr 2005 08:54:53 -0700, From: "Larry Fahnoe" <fahnoe@fahnoetech.com>= Subject: Re: How to install DCPS 2.4 without stopping queues? C Message-ID: <1114530893.856846.203200@o13g2000cwo.googlegroups.com>    Paul,   E Have rope, will travel...  Your three option solution to the issue of D which queues to stop during an upgrade meets our needs well, no need! for any extra rope!  Many thanks.    --Larry    ------------------------------  # Date: Tue, 26 Apr 2005 10:11:39 GMT 5 From: rdeininger@mindspringdot.com (Robert Deininger) @ Subject: Re: HP and IntelR Developer Workshops Update / ReminderL Message-ID: <rdeininger-2604050611420001@user-uinj4bd.dialup.mindspring.com>  C In article <99eq611sqilq4sq512ep14bbi2kac8dq6q@4ax.com>, Joe Bloggs  <JBloggs@acme.com> wrote:   G >On 25 Apr 2005 07:44:48 -0500, koehler@eisner.nospam.encompasserve.org  >(Bob Koehler) wrote:  > E >>In article <6aoj61heciqi3gfkl0kclib1hiog4qha03@4ax.com>, Joe Bloggs  <JBloggs@acme.com> writes: >>> 6 >>> Namely, I wanted to judge the suitability of using> >>> the rx1620 for VMS development, out to 2-3 years from now. >>>  >>> thoughts/comments? >>H >>   2-3 years?  OK, I know HP isn't DEC, but it took well over a decadeL >>   before VMS enineering stopped trying to fit VMS on the original 11/780. > 2 >right. was just wondering what folks thought the B >chances for the rx1620, to wind up 'unsupported' like the rx2600.    D rx2600 is NOT unsupported.  There are 4 supported CPU types, and VMSH expects to support those 4 types indefinitely.  The only unsupported CPUI is the 900 MHz Mckinley.  A McKinley CPUs can be upgraded (in-box) to any 6 of the other 4 types to get a supported configuration.  D Where did you get the impression that rx2600 isn't supported by VMS?   ------------------------------    Date: 26 Apr 2005 00:32:18 -0700% From: ksidibaba@gmail.com (ksidibaba) ' Subject: IP Failover: strange behaviour = Message-ID: <3dd11af5.0504252332.72d19069@posting.google.com>    Hi all,   E just trying to get to grisp with VMS, we are encountering quite a few A unexpected problems, one of which being IP-Failover. We are using C OpenVMS 7.3-2, with all recommended actual patches on 2 DS15, *not* ? participating in an OpenVMS cluster, both connected to the same D switch. On both machines, 2 NICs are configured to take part into IPB Failover. As long as both machines are up and running, IP failoverD behaves the way one would expect it to work (i.e. pulling one of theB network cables from one active NIC results in the other one takingE over). But when one of the machine is powered down, the other machine B starts losing connectivity. It looks as if the system is trying toD reconfigure the IP address of the failover-enabled NICs. Is this theC expected behaviour? that would imply that if a machine breaks down, B the other one would lose connectivity (!). Any help on that matter would be very much appreciated!  Thanks in advance,   Karim    ------------------------------    Date: 26 Apr 2005 15:09:11 +01006 From: peter@langstoeger.at (Peter 'EPLAN' LANGSTOEGER)+ Subject: Re: IP Failover: strange behaviour , Message-ID: <426e5997$1@news.langstoeger.at>  e In article <3dd11af5.0504252332.72d19069@posting.google.com>, ksidibaba@gmail.com (ksidibaba) writes: F >just trying to get to grisp with VMS, we are encountering quite a few6 >unexpected problems, one of which being IP-Failover.   5 Please consider LAN failover (instead of IP Failover) ' $ MCR LANCP SET DEVICE/FAILOVER_SET=...    --   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: Tue, 26 Apr 2005 07:57:11 -0500 ( From: Wayne Sewell <wayne@tachysoft.com> Subject: Re: Legato NetWorker / Message-ID: <00A42DBE.03D65B0D.3@tachysoft.com>   & >Date: Mon, 25 Apr 2005 19:56:16 -0700 >Subject: Re: Legato NetWorker) >From: Jeff Cameron <roktsci@comcast.net>  >X-Newsgroups: comp.os.vms0 >Message-ID: <BE92FBE0.CB2C%roktsci@comcast.net> >MIME-Version: 1.0    L >The biggest flaw with tapesys is that they provide their own hacked versionI >of Backup, so if a VMS patch comes out with a new backup.exe you have to * >wait until tapesys hacks the new version.  K Huh??  What millenium are you in?  Yes, the original version of tapesys did I this in the 1980s, but not for at least a decade, probably much longer.        > L >TapeControl achieves the same functionality by intercepting backup (as wellL >as initialize, mount and dismounts [dcl and system services]) doing what itM >has to do to manage the tape volumes, then passes the operation onto the VMS  >native facility.  >   K Non-ancient versions of tapesys use pure native vms backup as well.  I have M never seen the patching thing in operation, and my first contact with tapesys K was in 1992.  The only reason I even knew about the patching was because of I historical residue in the source tree, which I cleaned up many years ago.     J >Legato networker does the same too, but using open tape format instead of >VMS backup savesets.  >   J You make that sound like a benefit.  Many people here feel that it is not,M given that you cannot *restore* your disks without legato being in the loop.  G If vms savesets are the medium, only standalone backup is required.  No  network, no foreign system.   F Actually the "intercepting backup...doing what it has to...passes the > operation onto the VMS native facility" thing is outmoded too.  O tapesys uses the system service and rtl intercepts when running on vms versions H prior to 7.1.  For vms 7.1 and later, it does not use backup.exe at all,N instead using the callable backup api (backupshr).  This way *no* hacks of anyN kind are needed, including system service intercepts.  Everything is done withL calls to straight hp-supplied code.  All of the tapesys-specific stuff is in
 callbacks.  M Does tapecontrol use the backup API?  Does networker?  SLS does not appear to 0 (no reference to backupshr in the backup image).  I The system service intercept thing will not work on itanium, at least not D without a massive recoding effort.  The mechanism for system serviceM dispatching and for calling procedures is radically different.  AFAIK, no one M has yet worked out the technique for the system service intercepts, including L vms engineering.  There is a project in progress for this, at the request ofM myself (I have another product that still uses intercepts) and others, but it @ is still in a preliminary stage and has produced no actual code.  O I suspect that this is one of the major reasons that SLS is not being ported to 
 itanium.    N In contrast, tapesys is already running on itanium and performing backups.  ItO is still in test, but it appears to be fully operational.  I haven't received a ' bug report from the test guys in weeks.     L It would be a pretty big rewrite of a backup module to switch to the API andL off the system service intercepts.  We went through this exercise years ago,' when the API first came out in vms 7.1.    Wayne O =============================================================================== N Wayne Sewell, Tachyon Software Consulting  (281)812-0738   wayne@tachysoft.com; http://www.tachysoft.com/www/tachyon.html and wayne.html    O =============================================================================== P Jake Blues:"You traded the Caddy for a microphone? ...... Okay, I can buy that."   ------------------------------  % Date: Tue, 26 Apr 2005 09:24:47 -0400 # From: "John Smith" <a@nonymous.com> = Subject: Maybe HP should get out of the hardware business.... , Message-ID: <VJqdnTI4Wum93PPfRVn-2Q@igs.net>  J ....and just sell operating systems. If HP can't/won't build a workstationE for VMS at a competitive price-point, please don't hobble us with the K inability to run VMS on hardware that are price/performance leaders (please < don't bring up the $2000 DSPP 'deal' as a counter argument).  : The rationalization of going Itanic in the first place wasI 'commoditization', but at this juncture Opteron is arguably the commodity K processor of choice for most OEM's, and it will get cheaper and faster more  quickly than Itanic.  ' ---------------------------------------   7 IBM Among First To Offer Dual-Core Opteron Workstations    Mon Apr 25, 1:23 PM ET Kirk L. Kroeker  newsfactor.com  L IBM on Friday introduced the IntelliStation A Pro 6217 workstation, a systemD that will ship with the new dual-core AMD (NYSE: AMD - news) Opteron processors.   K IBM also announced that the IBM eServer 326 1U server line will include the  new dual-core chips.  J "AMD64 dual-core technology builds on the already strong value propositionK our standards-based IntelliStation workstation provides to customers," said J Bob Lenard, director of the IntelliStation product line at the IBM Systems and Technology Group.   F "Besides the addition of dual core, IBM engineered the new A Pro to beL faster, more agile, cooler and quieter with more graphical ability, allowingL the performance and reliability levels that IBM continuously meets to remain# at the highest level," said Lenard.   F In addition to the new dual-core processors, the A Pro 6217 offers PCIK Express graphics from Nvidia and 3Dlabs with dual-display support right out  of the box.   H According to IBM, the new 3DLabs Wildcat Realizm graphics chipset, whichK ships with the IBM workstation, has achieved some of the highest benchmarks  in the industry.  C The system supports up to 16 GB of PC3200 ECC memory and comes with @ Microsoft Windows XP Professional or Red Hat Enterprise Linux 3.  H The workstation will be available beginning April 26, 2005 for US$3,259.             --L OpenVMS - The never advertised operating system with the dwindling ISV base.   ------------------------------    Date: 26 Apr 2005 07:27:37 -0700* From: "Alan Greig" <greigaln@netscape.net>A Subject: Re: Maybe HP should get out of the hardware business.... C Message-ID: <1114525657.903535.291030@f14g2000cwb.googlegroups.com>    John Smith wrote: @ > ....and just sell operating systems. If HP can't/won't build a workstation G > for VMS at a competitive price-point, please don't hobble us with the E > inability to run VMS on hardware that are price/performance leaders  (please > > don't bring up the $2000 DSPP 'deal' as a counter argument). > < > The rationalization of going Itanic in the first place wasA > 'commoditization', but at this juncture Opteron is arguably the 	 commodity A > processor of choice for most OEM's, and it will get cheaper and  faster more  > quickly than Itanic.  A Seems that AMD have managed to make a very scalable system - much G better than the 64 bit Xeon. Intel can't allow AMD to get too far ahead D so there's every reason to expect future Intel X86-64 bit processorsC that scale well and outperform Itanium in large SMP configurations.   F The release of the HP benchmarks makes sense of something I read in anC HP employee blog which seemed to hint at such an astonishingly well D performing SMP Opteron system. Makes me wonder when/if we see X86-64 variants of Integrity. --  
 Alan Greig   ------------------------------  % Date: Tue, 26 Apr 2005 11:45:03 -0400 # From: "John Smith" <a@nonymous.com> A Subject: Re: Maybe HP should get out of the hardware business.... , Message-ID: <E-ydnUtUa-Cc__PfRVn-oQ@igs.net>   Alan Greig wrote:  > John Smith wrote: A >> ....and just sell operating systems. If HP can't/won't build a A >> workstation for VMS at a competitive price-point, please don't ? >> hobble us with the inability to run VMS on hardware that are B >> price/performance leaders (please don't bring up the $2000 DSPP! >> 'deal' as a counter argument).  >>= >> The rationalization of going Itanic in the first place was B >> 'commoditization', but at this juncture Opteron is arguably the@ >> commodity processor of choice for most OEM's, and it will get/ >> cheaper and faster more quickly than Itanic.  > C > Seems that AMD have managed to make a very scalable system - much C > better than the 64 bit Xeon. Intel can't allow AMD to get too far A > ahead so there's every reason to expect future Intel X86-64 bit @ > processors that scale well and outperform Itanium in large SMP > configurations.  > H > The release of the HP benchmarks makes sense of something I read in anE > HP employee blog which seemed to hint at such an astonishingly well F > performing SMP Opteron system. Makes me wonder when/if we see X86-64 > variants of Integrity.    	 <cynical> L With zero advertising for VMS, only slightly more for NSK and HP-UX, HP doesJ not seem to be concerned with growing the market for Itanic-based systems.5 Pretty soon there'll be no need for Integrity at all.   
 </cynical>     --F OpenVMS - The never advertised operating system with the dwindling ISV base.    ------------------------------    Date: 26 Apr 2005 09:29:59 -0700* From: "Alan Greig" <greigaln@netscape.net>A Subject: Re: Maybe HP should get out of the hardware business.... C Message-ID: <1114532999.905072.196360@z14g2000cwz.googlegroups.com>    John Smith wrote:    > <cynical> F > With zero advertising for VMS, only slightly more for NSK and HP-UX, HP does C > not seem to be concerned with growing the market for Itanic-based  systems.7 > Pretty soon there'll be no need for Integrity at all.   C Have a look at Sun's claimed benchmark figures for their own 8 core F Opteron based system. 	http://www.sun.com/servers/entry/v40z/index.jsp  F       04/21/2005 Top result on 8-thread SPEC OMPM2001 HPC benchmarkNew  F       04/21/2005 World Record floating point performance for all 8-way x86-compatible systemsNew     C A quick glance shows that it isn't just the SAP benchmarks the dual A core Opterons excel at. Even the floating point performance beats = Itanium. Tried to find some benchmark figures for the 16 core G configuration but nobody seems to have posted them yet even though some D vendors claim to be shipping already. Sun say they will support that configuration later this year. --  
 Alan Greig   ------------------------------  % Date: Tue, 26 Apr 2005 09:08:58 +0100 0 From: Chris Sharman <chris.sharman@sorry.nospam>2 Subject: Re: MySQL and PHP Authentication Protocol4 Message-ID: <d4ksuq$66b$1$8302bc10@news.demon.co.uk>   Kevin.R.Morris@gmail.com wrote:   ) > The answer to my question was found at:  > 3 > http://dev.mysql.com/doc/mysql/en/old-client.html  >   # Yes - I got bitten by this one too. G There is a later version available (php_mysql_mysql_*.zip iirc) but it   doesn't work with openssl097e.   Chris    ------------------------------    Date: 26 Apr 2005 07:50:07 -0700% From: andrew.rycroft@intrinsitech.com ; Subject: Optimising Alpha Cobol application on OpenVMS IA64 C Message-ID: <1114527007.188421.142590@g14g2000cwa.googlegroups.com>    Hi,   G We have just migrated a Cobol application from OpenVMS Alpha to OpenVMS " IA64. We used the Cobol command :-   $ COBOL/STANDARD=3  E All the code compiled, and we have linked it and it runs. We now want A to optimise it. Does anyone have any suggestions on what compiler @ switches to use, or what we need to do to optimise it for IA64 ?   Thanks Andrew   ------------------------------  # Date: Tue, 26 Apr 2005 16:56:18 GMT & From: John Reagan <john.reagan@hp.com>? Subject: Re: Optimising Alpha Cobol application on OpenVMS IA64 0 Message-ID: <Soube.4523$dp.953@news.cpqcorp.net>  & andrew.rycroft@intrinsitech.com wrote: > Hi,  > I > We have just migrated a Cobol application from OpenVMS Alpha to OpenVMS $ > IA64. We used the Cobol command :- >  > $ COBOL/STANDARD=3 > G > All the code compiled, and we have linked it and it runs. We now want C > to optimise it. Does anyone have any suggestions on what compiler B > switches to use, or what we need to do to optimise it for IA64 ? >  > Thanks > Andrew >    $ HELP COBOL /OPTIMIZE  C Unless you had some DCL symbol for COBOL, you already received the  " default of full IA64 optimization.  G Unless you find problems, I think you're done.  It really is that easy   for many people.   --   John Reagan / HP Pascal/{A|I}MACRO for OpenVMS Project Leader  Hewlett-Packard Company    ------------------------------    Date: 26 Apr 2005 11:56:18 +0200. From: huber@NOBODY-mppmu.mpg.de (Joseph Huber)< Subject: Re: RAW IP socket problem, recvfrom() never returns+ Message-ID: <q1rJhvLlijPx@vms.mppmu.mpg.de>   g In article <1114444174.643752.30010@o13g2000cwo.googlegroups.com>, "tolik" <khusid@ulticom.com> writes: G > What are the errors are you getting? Did you use this line to compile 	 > server? @ >  gcc -o serv -DOVMS -D_DECC_V4_SOURCE -D_SOCKADDR_LEN server.c' o.k. compiles now (VMS 7.3-1, DECC 6.5)     > , > listen and accept on raw IP sockets? :-) ?G > Network layer is connection-less, unlike your typical transport layer  > (TCP for example).  ; Sorry, You're completly right, I overlooked the RAW socket, U and Yes, see also the same behaviour, (client and server on VMS, TCPIP services 5.3): Q No errors,Sockets visible with  TCPIP SHOW DEV, but no data received, no clue ...    --  6    Joseph Huber, Muenchen  http://www.huber-joseph.de/   ------------------------------    Date: 26 Apr 2005 04:28:29 -0700" From: "tolik" <khusid@ulticom.com>< Subject: Re: RAW IP socket problem, recvfrom() never returnsC Message-ID: <1114514909.873378.105830@o13g2000cwo.googlegroups.com>   D One thing that will make it work if I have my server program bind to1 one of the IP interfaces on the box, for example: / sin.sin_addr.s_addr = inet_addr("172.25.5.80"); A However, this does not work (but works fine on Solaris and Linux) ( sin.sin_addr.s_addr = htonl(INADDR_ANY);E BTW, INADDR_ANY is defined to be zero, thats why I memset sockaddr_in ? structure to zero in my sample programs.  I've tried explicitly D initializing it with INADDR_ANY, and got the same results.  To me itG looks like a bug in VMS.  The user should be able to bind to all the IP F interfaces on the box, unless HP wants to be "different" from the rest of the world.....    ------------------------------   Date: 26 Apr 2005 13:20:16 GMT1 From: bill@triangle.cs.uofs.edu (Bill Gunshannon) 
 Subject: TK50 , Message-ID: <3d6tggF6rmdkpU1@individual.net>   I am back in begging mode.  A Is there anyone who has TK50 drives (and maybe even a TK50Z) that = they are interested in getting rid of?  Knowing that they are @ getting rarer every day, I think it would be a real good idea if@ I could stockpile a few as I expect to continue to need them for some time yet.  A Oh yeah.  Anybody got old TK50 tapes they need to get rid of too?    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: 26 Apr 2005 09:55:17 -0400% From: kludge@panix.com (Scott Dorsey)  Subject: Re: TK50 + Message-ID: <d4lh85$jmi$1@panix2.panix.com>   ) Bill Gunshannon <bill@cs.uofs.edu> wrote:  >I am back in begging mode.  > B >Is there anyone who has TK50 drives (and maybe even a TK50Z) that> >they are interested in getting rid of?  Knowing that they areA >getting rarer every day, I think it would be a real good idea if A >I could stockpile a few as I expect to continue to need them for  >some time yet.   C I may have one or two, but to be honest they are very easy machines % to repair.  There isn't much to them.   D The only problem are the heads, which were made by Applied MagneticsD and are no longer available.  A couple head companies have said theyA would make replacements for me if I could order a couple hundred.   B >Oh yeah.  Anybody got old TK50 tapes they need to get rid of too?  C Probably.  How many do you need?  I think we threw out a couple big  stacks recently. --scott  --  6 "C'est un Nagra.  C'est suisse, et tres, tres precis."   ------------------------------  % Date: Tue, 26 Apr 2005 10:27:49 -0400 2 From: "Stanley F. Quayle" <squayle@insight.rr.com> Subject: Re: TK50 , Message-ID: <426E17A5.1652.779D05@localhost>  / On 26 Apr 2005 at 13:20, Bill Gunshannon wrote: H > Is there anyone who has TK50 drives (and maybe even a TK50Z) that theyH > are interested in getting rid of?  Knowing that they are getting rarerH > every day, I think it would be a real good idea if I could stockpile a= > few as I expect to continue to need them for some time yet.   D I have two stand-alone SCSI TK50 drives (are those the TK50Z?) that D you can have cheap.  They don't work, but might be good for parts.  # At least the power supplies work...   D If you need drives that work, I've had good success with Joel Marty D of Global-IT (http://www.global-itcorp.com/products.php?id=&cid=23).  9 > Anybody got old TK50 tapes they need to get rid of too?   B I have a big batch of used TK50's, since my CHARON-VAX [Shameless F Plug (tm)] customers use them for conversions sometimes. Joel has new " 3M tapes still in the shrink wrap.  E Of course, why are you keeping TK50's around?  Wouldn't it be better  > to burn the data to CD's?  Even an old RRD40 will read CD-R's 
 (usually).  ? [And, of course, I could convert the whole mess to CHARON-VAX.]   
 --Stan Quayle  Quayle Consulting Inc.  
 ----------- Stanley F. Quayle, P.E. N8SQ  +1 614-868-1363 3 8572 North Spring Ct., Pickerington, OH  43147  USA 0 stan-at-stanq-dot-com       http://www.stanq.com   ------------------------------    Date: 26 Apr 2005 08:45:13 -0700 From: tomarsin2015@comcast.net Subject: Re: TK50 C Message-ID: <1114530313.901910.164340@o13g2000cwo.googlegroups.com>   C How many TK50 tapes do you want?? Just send me your address. I will $ even pick up the tab for shipping!!!   ------------------------------    Date: 26 Apr 2005 08:53:28 -0700 From: tomarsin2015@comcast.net Subject: Re: TK50 C Message-ID: <1114530808.447895.197460@o13g2000cwo.googlegroups.com>   C How many TK50 tapes do you want?? Just send me your address. I will $ even pick up the tab for shipping!!!   ------------------------------    Date: 26 Apr 2005 00:03:06 -0700* From: karl.rohwedder@gmx.de (Cluster-Karl)C Subject: Re: [ECP V5.5A] Unrecognized ECP/Collect data file version < Message-ID: <6436beb5.0504252303.853bc8e@posting.google.com>  h peter@langstoeger.at (Peter 'EPLAN' LANGSTOEGER) wrote in message news:<426d387f@news.langstoeger.at>...C > Since I upgraded ECP from V5.5 to V5.5A, I noted a little problem : > (which most likely existed before, but I haven't noted). >  > 1) ECP$PERF_DATA on ODS5 Disk ? > 2) CPC File written during startup (@ECP$LIBRARY:ECP$MANAGER)  >  > $ DIRECTORY/SINCE  > ' > Directory DISK$VMSSYS:[ECP$PERF_DATA]  > 3 > LUNA_2005APR25_1.cpc;1				<--- note the lowercase  >  > Total of 1 file, 4471 blocks. I > $! ECP doesn't like lowercase in filenames (with strange error message) 6 > $ PLAN ANA/CPC_VMS_FILE=LUNA_2005APR25_1.CPC/ANA=X.XR >  Enterprise Capacity Performance Analyzer V5.5A-25 (c) 1997, 2005 Hewlett-Packar3 > d Development Company, L.P., All Rights Reserved. 9 > ECP-WARNING  Unrecognized ECP/Collect data file version ; > $! And though it tells WARNING it aborts and does nothing * > $! Even in GUI (/MOTIF) the same problem. > $! Rename file to all uppercase and it works8 > $ RENAME/LOG LUNA_2005APR25_1.CPC LUNA_2005APR25_1.CPCQ > %RENAME-I-RENAMED, DISK$VMSSYS:[ECP$PERF_DATA]LUNA_2005APR25_1.cpc;1 renamed to:3 > DISK$VMSSYS:[ECP$PERF_DATA]LUNA_2005APR25_1.CPC;1y6 > $! But how to rename the file while you are in GUI ?6 > $ PLAN ANA/CPC_VMS_FILE=LUNA_2005APR25_1.CPC/ANA=X.XR >  Enterprise Capacity Performance Analyzer V5.5A-25 (c) 1997, 2005 Hewlett-Packar3 > d Development Company, L.P., All Rights Reserved.h% > ECP-INFO  ECP/Analyze processing... 1 > ECP-INFO  Writing ECP/Analyze report to x.x ...: >  > This leads to the questions: > = > 1) Why is the filename case important ? (bad programming ?)fN > 2) Is the startup process in extended parse style ? (haven't noticed before)7 > 3) Did anybody of you already notice (and told HPQ) ?s > E > Fix is to replace ".cpc" with ".CPC" in ECP$LIBRARY:ECP$MANAGER.COMn' > of course, but will HPQ fix it, too ?g >  > -EPLAN    F The only DCL procedure in SYS$STARTUP... on our V8.2 system, that setsE the parse  style, is GNV$STARTUP.COM. It doesn't reset it, may be youP  called GNV$STARTUP  during boot?  = On V8.2 the datacollector of ECP is not used (it uses the TDCV< collection files instead), so I haven't noticed this before.  
 regards KalleE   ------------------------------    Date: 26 Apr 2005 02:26:00 -0700% From: Bart.Zorn@xs4all.nl (Bart Zorn)eC Subject: Re: [ECP V5.5A] Unrecognized ECP/Collect data file versionE= Message-ID: <a98cd882.0504260126.75d4930e@posting.google.com>-  h peter@langstoeger.at (Peter 'EPLAN' LANGSTOEGER) wrote in message news:<426d387f@news.langstoeger.at>...C > Since I upgraded ECP from V5.5 to V5.5A, I noted a little problem8: > (which most likely existed before, but I haven't noted). >  > 1) ECP$PERF_DATA on ODS5 Disk ? > 2) CPC File written during startup (@ECP$LIBRARY:ECP$MANAGER)s >  > $ DIRECTORY/SINCE  > ' > Directory DISK$VMSSYS:[ECP$PERF_DATA]0 > 3 > LUNA_2005APR25_1.cpc;1				<--- note the lowercase- >  > Total of 1 file, 4471 blocks.7I > $! ECP doesn't like lowercase in filenames (with strange error message):6 > $ PLAN ANA/CPC_VMS_FILE=LUNA_2005APR25_1.CPC/ANA=X.XR >  Enterprise Capacity Performance Analyzer V5.5A-25 (c) 1997, 2005 Hewlett-Packar3 > d Development Company, L.P., All Rights Reserved.O9 > ECP-WARNING  Unrecognized ECP/Collect data file version.; > $! And though it tells WARNING it aborts and does nothingr* > $! Even in GUI (/MOTIF) the same problem. > $! Rename file to all uppercase and it works8 > $ RENAME/LOG LUNA_2005APR25_1.CPC LUNA_2005APR25_1.CPCQ > %RENAME-I-RENAMED, DISK$VMSSYS:[ECP$PERF_DATA]LUNA_2005APR25_1.cpc;1 renamed toe3 > DISK$VMSSYS:[ECP$PERF_DATA]LUNA_2005APR25_1.CPC;1 6 > $! But how to rename the file while you are in GUI ?6 > $ PLAN ANA/CPC_VMS_FILE=LUNA_2005APR25_1.CPC/ANA=X.XR >  Enterprise Capacity Performance Analyzer V5.5A-25 (c) 1997, 2005 Hewlett-Packar3 > d Development Company, L.P., All Rights Reserved.o% > ECP-INFO  ECP/Analyze processing...i1 > ECP-INFO  Writing ECP/Analyze report to x.x ...w >  > This leads to the questions: > = > 1) Why is the filename case important ? (bad programming ?) N > 2) Is the startup process in extended parse style ? (haven't noticed before)7 > 3) Did anybody of you already notice (and told HPQ) ?a > E > Fix is to replace ".cpc" with ".CPC" in ECP$LIBRARY:ECP$MANAGER.COM ' > of course, but will HPQ fix it, too ?  >  > -EPLAN  4 I think it is a matter of priorities. Have a look atE ECP$LIBRARY:ECP$MANAGER.COM, it looks like an excersise in the use of-F F$GETQUI and F$CVTIME. When you put so much effort in that there is no$ spare time to do other things right.  C I still believe it is possible to create source code which compiles4A cleanly, without warnings which propagate to the linker. But thateF seems too much to ask these days. Let alone implementing things right:C filenames MUST be case INsensitive! Only really braindead operatingi6 systems have case sensitivity. Even Weendoze does not!  	 Bart Zorna   ------------------------------    Date: 26 Apr 2005 15:29:39 +01006 From: peter@langstoeger.at (Peter 'EPLAN' LANGSTOEGER)C Subject: Re: [ECP V5.5A] Unrecognized ECP/Collect data file versions, Message-ID: <426e5e63$1@news.langstoeger.at>  e In article <a98cd882.0504260126.75d4930e@posting.google.com>, Bart.Zorn@xs4all.nl (Bart Zorn) writes:i5 >I think it is a matter of priorities. Have a look atPF >ECP$LIBRARY:ECP$MANAGER.COM, it looks like an excersise in the use ofG >F$GETQUI and F$CVTIME. When you put so much effort in that there is no % >spare time to do other things right.P  * Wow, that's cynical. Where is the smiley ?  D >I still believe it is possible to create source code which compilesB >cleanly, without warnings which propagate to the linker. But thatG >seems too much to ask these days. Let alone implementing things right:rD >filenames MUST be case INsensitive! Only really braindead operating7 >systems have case sensitivity. Even Weendoze does not!i  ' Do you perhaps know whom to blame ? ;-):L btw. U**X is case sensitive, this means you call it more braindead than M$ ?   -- t Peter "EPLAN" LANGSTOEGERr% Network and OpenVMS system specialiste E-mail  peter@langstoeger.atF A-1030 VIENNA  AUSTRIA              I'm not a pessimist, I'm a realist   ------------------------------    Date: 26 Apr 2005 15:06:10 +01006 From: peter@langstoeger.at (Peter 'EPLAN' LANGSTOEGER)C Subject: Re: [ECP V5.5A] Unrecognized ECP/Collect data file versionc, Message-ID: <426e58e2$1@news.langstoeger.at>  i In article <6436beb5.0504252303.853bc8e@posting.google.com>, karl.rohwedder@gmx.de (Cluster-Karl) writes:iG >The only DCL procedure in SYS$STARTUP... on our V8.2 system, that setseF >the parse  style, is GNV$STARTUP.COM. It doesn't reset it, may be you! >called GNV$STARTUP  during boot?s  M Kalle, you won a BINGO. Exactly one line before invoking ECP$STARTUP.COM :-))rM Moved GNV$STARTUP.COM now to the end (in addition to the fix in ECP$MANAGER).m  > >On V8.2 the datacollector of ECP is not used (it uses the TDC= >collection files instead), so I haven't noticed this before.s  ? Thanks for reminding me, but this machines are still V7.3-2 ;-)v   -EPLAN  G PS: Please, (VMS) engineering, can we have uppercase filenames in _all_nC your DCL scripts (and non-multiplatform sourcecode programs, too) ?aM I don't like to see unneccessary mixed case filenames on my (system) disks...o -- o Peter "EPLAN" LANGSTOEGERe% Network and OpenVMS system specialists E-mail  peter@langstoeger.atF A-1030 VIENNA  AUSTRIA              I'm not a pessimist, I'm a realist   ------------------------------    Date: 26 Apr 2005 06:32:10 -0700$ From: "AEF" <spamsink2001@yahoo.com>C Subject: Re: [ECP V5.5A] Unrecognized ECP/Collect data file version8C Message-ID: <1114522330.432441.158190@g14g2000cwa.googlegroups.com>R    Peter 'EPLAN' LANGSTOEGER wrote:C > Since I upgraded ECP from V5.5 to V5.5A, I noted a little problemo: > (which most likely existed before, but I haven't noted). [...]r. > $! Rename file to all uppercase and it works8 > $ RENAME/LOG LUNA_2005APR25_1.CPC LUNA_2005APR25_1.CPCF > %RENAME-I-RENAMED, DISK$VMSSYS:[ECP$PERF_DATA]LUNA_2005APR25_1.cpc;1
 renamed to3 > DISK$VMSSYS:[ECP$PERF_DATA]LUNA_2005APR25_1.CPC;1.6 > $! But how to rename the file while you are in GUI ?     Open a DECterm!  :-)  B Doesn't DECwindows still have Fileview? You can't do it with that?     [...]d   ------------------------------   End of INFO-VAX 2005.232 ************************