1 INFO-VAX	Sat, 18 Nov 2006	Volume 2006 : Issue 634       Contents:! Re: Educational VMS IA64 licenses ! Re: Educational VMS IA64 licenses " Re: I can't install the LDAP login* Re: Purpose of uppercasing a RUN command ? Re: VMS Gaming History? = Re: What layered products are important for a VAX VMS system? = Re: What layered products are important for a VAX VMS system?  Re: Who called me?  F ----------------------------------------------------------------------  % Date: Fri, 17 Nov 2006 17:04:33 -0500 8 From: Stephen Hoffman <Hoff@HoffmanLabs-RemoveThis-.Org>* Subject: Re: Educational VMS IA64 licenses( Message-ID: <ejlblj$b64$1@pyrite.mv.net>   Malcolm Dunnett wrote:  M > Unfortunately I downloaded one of these (OPENVMS-I64-EDUOE) onto my RX2600  L > running VMS 8.3 but it doesn't appear to be recognized, I get the "no VMS ) > license active" errors when I boot VMS.  > O > I presume VMS doesn't know to check for the "EDUOE" type license, is there a   > patch to enable this?   9    You should have a OPENVMS-I64-FOE, OPENVMS-I64-EOE or  A OPENVMS-I64-MCOE license -- I don't remember which, but any will  H authorize base OpenVMS I64 -- in the stash of licenses received.  There D was a brief discssion here, and the EDUOE Educational and (though I D haven't looked for it) the HOE Hobbyist licenses would be for other I vendors or other products that want to tie to into that license and into  D that environment as part of their licensing.  If you don't have (at G least) the FOE, EOE or MCOE license PAK around, you probably won't get   anywhere useful.   ------------------------------  % Date: Fri, 17 Nov 2006 14:47:07 -0800 , From: "Malcolm Dunnett" <dunnett@mala.bc.ca>* Subject: Re: Educational VMS IA64 licenses Message-ID: <455e3b70@flight>   F "Stephen Hoffman" <Hoff@HoffmanLabs-RemoveThis-.Org> wrote in message " news:ejlblj$b64$1@pyrite.mv.net... > Malcolm Dunnett wrote: > G >> Unfortunately I downloaded one of these (OPENVMS-I64-EDUOE) onto my  L >> RX2600 running VMS 8.3 but it doesn't appear to be recognized, I get the 2 >> "no VMS license active" errors when I boot VMS. >>H >> I presume VMS doesn't know to check for the "EDUOE" type license, is   >> there a patch to enable this? > K >   You should have a OPENVMS-I64-FOE, OPENVMS-I64-EOE or OPENVMS-I64-MCOE  I > license -- I don't remember which, but any will authorize base OpenVMS  + > I64 -- in the stash of licenses received.   L Nope, all I got was the EDUOE license. I presumed that was meant to replace  the . FOE/EOE/MCOE license, but it isn't recognized.   ------------------------------  % Date: Fri, 17 Nov 2006 11:53:08 -0800 , From: "Malcolm Dunnett" <dunnett@mala.bc.ca>+ Subject: Re: I can't install the LDAP login  Message-ID: <455e12a9@flight>   D "Peter 'EPLAN' LANGSTOEGER" <eplan@langstoeger.at> wrote in message & news:455e0683$1@news.langstoeger.at...I > In article <455df394$1@flight>, "Malcolm Dunnett" <dunnett@mala.bc.ca>  	 > writes: D >>I'm trying to install the LDAP enabled login on VMS 8.3 Alpha ( as  ' >>So what is INSTALL complaining about?  > L > About a requirement you can't satisfy, so I assume, there is a bug in the  > kit. >    That would be my guess as well.   L FWIW I just tried the Itanium version and get exactly the same results (not  surprising I suppose)   E so the $64,000 question is how to get an updated kit ( or is there a   workaround ).    ------------------------------    Date: 17 Nov 2006 14:21:05 -0800- From: "Doug Phillips" <dphill46@netscape.net> 3 Subject: Re: Purpose of uppercasing a RUN command ? C Message-ID: <1163802065.460817.105530@k70g2000cwa.googlegroups.com>    Larry Kilgallen wrote:O > In article <ejkg9o$5aa$1@south.jnrs.ja.net>, david20@alpha2.mdx.ac.uk writes: t > > In article <OF7A40D9CD.6262BD35-ON85257229.004B367E-85257229.004B9753@metso.com>, norm.raphael@metso.com writes:H > >>I have no particular insights, but ISTM that since it is possible toJ > >>SET PROCESS/PARSE=EXTENDED with ODS5, then this may guarantee a match. > >> > > N > > SET PROCESS/PARSE=EXTENDED  on it's own will preserve case but will access' > > files in a case insensitive manner.  > > % > > SET PROCESS/CASE_LOOKUP=SENSITIVE  > > 7 > > would be needed to do case sensitive file accesses.  > > 9 > > But then they could have just written the commands in 
 > > uppercase  > K > But writing the commands in uppercase would not have provided such a good I > warning to others who write command procedures (such as those assembled  > here).  9 JF's example looks like it's from TCPIP$SMTP_RECV_RUN.COM    Follow this logic:  / $     serv_id   = f$edit("tcpip$smtp","upcase") > $     serv_home = f$edit("sys$specific:[''serv_id']","upcase") $ !  $ ! purge smtp receiver logs $ !  $ purge: $     set default 'serv_home'   . There is a log file name created the same way:  9 $     lognam = f$edit("tcpip$smtp_recv_run.log","upcase")   5 And of course, JF's original "run" statement example.   D There is no good reason to perform f$edit on a constant argument. It7 makes as much sense as  $ ab = f$extract(0,2,"ABCDEFG")   E If the actual directory name is "SYS$SPECIFIC:[TCPIP$SMTP]" and can't ? be any other case, then use the frigging actual directory name.   B The fact that this has hard-coded names throughout doesn't make meE believe it was done to allow for the future use of logicals/variables 0 or to make future changes to the routine easier.  B This routine makes lexical calls to create symbols and does symbolA substitutions that accomplish nothing other than to waste cycles.   B This routine also defines two labels, "purge:" and "run:" that are= never referenced. Each is placed after a comment. Again, they  accomplish nothing positive.  G TCPIP$SMTP_RECV_RUN.COM is not an example of good DCL coding technique,  IMHO.   5 Not that I haven't written my share of crappy code;-)    ------------------------------  % Date: Fri, 17 Nov 2006 21:20:48 -0500 ' From: Dave Froble <davef@tsoft-inc.com>   Subject: Re: VMS Gaming History?9 Message-ID: <ncCdnSQTtfia88PYnZ2dnUVZ_t2dnZ2d@libcom.com>    glen herrmannsfeldt wrote: > Dave Froble wrote: >  > (snip) > I >> But if there is a case for a defense for me, my feeble memory is that  J >> the first microprocessors such as the 4004 and such came out in a 1981 J >> timeframe.  Could be wrong.  General time when the 10 was discontinued. > > > The 8080 came out in 1974, the 4004 was a few years earlier." > The 8086 came out in about 1979. > 	 > -- glen  >     I am getting too old and senile!  & Take me and put me out of this misery.  ' At least don't trust my memory anymore.    --  4 David Froble                       Tel: 724-529-0450> Dave Froble Enterprises, Inc.      E-Mail: davef@tsoft-inc.com DFE Ultralights, Inc.  170 Grimplin Road  Vanderbilt, PA  15486    ------------------------------  # Date: Fri, 17 Nov 2006 20:15:09 GMT % From: Rob Brown <mylastname@gmcl.com> F Subject: Re: What layered products are important for a VAX VMS system?E Message-ID: <Pine.LNX.4.61.0611171242290.31676@localhost.localdomain>   ) On Fri, 17 Nov 2006, Rich Alderson wrote:   D > DECnet and LAT are out.  The corporate network is strictly TCP/IP,  F So what if your VMS system talks to your TOPS-10 system using DECnet? C Have you got corporate network guys with nothing better to do than  3 check for "weird" packets and give you a hard time?   5 > and terminal server is something like a Cisco 2501.  > A > The way the Tops-10 2065 is hooked to the net for telnet is by  D > connecting its 16 terminal lines to two octopus cables on a 2501. D > I'm wondering if running an early version of VMS will require the  > same kind of hookup.  E My 5 seconds on google makes me think that the 2501 is a router.  If  E whatever you have does serial RS232 on one side and LAT on the other  F side (and you accept LAT), you can avoid that.  Since you reject LAT, C if your terminal server speaks TELNET and you get a VMS and TCP/IP  1 combination that works, you can still avoid that.   D My experience with VMS plus TCP/IP goes back to 6.2.  You should be D able to get a TELNET terminal server to work with that combination. 5 Others can speak to the question of earlier versions.   ? All of this presupposes that you have Ethernet in your 785 and  D Ethernet support in your chosen version of VMS.  I expect that such F support exists in version 4.x.  I can't guess about earlier versions. @ I think Ethernet was pretty new in 1985.  I know that the first 5 networks I saw in the late '70s did not use Ethernet.   G If you don't have Ethernet support in your chosen version of VMS, then  ? I think you cannot avoid having the octopus cables you mention.   G But on the other hand, you will be free to run DECnet, and nobody will  E know, since you won't be on the corporate network.  Before Ethernet,  @ DEC built networks using various parallel and serial interfaces.  @ I won't ask about the wisdom of having your publicly accessible * display systems on your corporate network.   - Rob      --    B Rob Brown                        b r o w n a t g m c l d o t c o m6 G. Michaels Consulting Ltd.      (780)438-9343 (voice)4 Edmonton                         (780)437-3367 (FAX)2                                   http://gmcl.com/   ------------------------------  % Date: Fri, 17 Nov 2006 16:15:28 -0500 8 From: Stephen Hoffman <Hoff@HoffmanLabs-RemoveThis-.Org>F Subject: Re: What layered products are important for a VAX VMS system?( Message-ID: <ejl8pg$aa4$1@pyrite.mv.net>   Rich Alderson wrote:  Q > DECnet and LAT are out.  The corporate network is strictly TCP/IP, and terminal ( > server is something like a Cisco 2501.  5    Cisco routers can bridge DECnet, LAT, MOP and SCS.   ;    Though if you want to go ancient, you use DECnet DDCMP.  C Point-to-point or multipoint if you go way back, or later Ethernet.       Less ancient: LAT.       Recent: IP.  O > The way the Tops-10 2065 is hooked to the net for telnet is by connecting its Q > 16 terminal lines to two octopus cables on a 2501.  I'm wondering if running an < > early version of VMS will require the same kind of hookup.  G    If DECnet-Plus is feasible, you can operate DECnet over IP, meaning  F that the local Cisco network guys won't have to figure out how to set  the Cisco box to bridge it.   H    FWIW, Cisco router boxes can transfer LAT and SCS and MOP and DECnet E just fine in this bridging mode, and -- if you want it -- Cisco even  I offers a DECnet router.  The Cisco boxes can even selectively bridge and  G route a mixture of protocols on the same LAN, because the box can look  C at the packet headers and decide to pass the packet through to the  F router or through to the bridge.  (The Cisco guys tend to call this a  Virtual LAN, FWIW.)   C    Or you load a more recent node with both DECnet and TCP/IP, and  H bridge it.  That was a common solution, particularly when you needed to 0 run in a configuration where you didn't have IP.  P >>> I've got install tapes for 6.0 and 6.2.  I've heard (in c.o.v., as I recall)M >>> that 7.3 will run on a 785, but we might not choose to push the envelope. 0 >>> We're really looking in the other direction.L >> There are not too many things that have been dropped from VMS.  I cannot J >> think of any.  Running a new version won't cause you to lose anything.  >> It will avoid problems. >>J >> V6.2 was a rather good version.  If you don't want to go to V7.2, then N >> I'd suggest V6.2.  I'd avoid anything between those two.  Personal opinion.  H    Running on a newer version will get you better reliability, and more E current products.  Bluntly, the earliest releases of TCP/IP Services  D (UCX) were somewhat shaky.  Running older software means you get to H chase both older hardware AND older software.  But if that is something & you want to try, well, by all means...  P > For historical purposes, I may want to run a 5.x or 4.x system, or even 3.2 ifQ > I correctly remember that that was the last version with the old interface that O > looked like Tops-10 and the PDP-11 and PDP-8 UIs.  But all of that depends on $ > what I can find.  First, licenses!  G    If you get current licenses and/or permission, don't bother digging  E around further as LMF did not exist prior to V5.0, and -- other than  C DECnet -- there were no particular license keys.  The ancient V4.*  < DECnet license keys are on the OpenVMS Freeware V4.0 distro.  -    Here are the VAX/VMS V4.* keys for DECnet:   ;    http://h71000.www7.hp.com/freeware/freeware40/netrtg040/   F    These keys work ONLY on VAX/VMS V4.*, and do not license DECnet on  V5.0 or later.  G    If you go too early in your releases, you loose access to Ethernet.  E Off the top, I don't recall which release first supported the DEUNA.  H (Circa V3.4?, with the introduction of DECnet Phase IV?)  And the older D controllers are Ethernet controllers, not IEEE 802.3 controllers -- E there were features added in 802.3 that didn't exist in the original  H Ethernet specs -- the heartbeat settings were what messed up most folks.  C    The old interface is the Monitor Console Routines (MCR) command  H interface, and that was built into OpenVMS (VMS) up through V3.* and it F became the RSX layered product at V4.0.  There were pieces of OpenVMS G (VMS) that were compatibility-mode images, and most everything finally  H "went native" at V4.0.  The RSX product went over to Mentec, as part of 8 the sale of the PDP hardware and software some eons ago.   ------------------------------  % Date: Fri, 17 Nov 2006 15:47:57 -0500 8 From: Stephen Hoffman <Hoff@HoffmanLabs-RemoveThis-.Org> Subject: Re: Who called me? ( Message-ID: <ejl75u$9pf$1@pyrite.mv.net>  H    What VAXman and Tom Linden are discussing does not appear to be what   you want, nor not what you seek.  F    For best reliability, I would tend to code this module information H into the application directly (C has preprocessor defines for this, and D this basic information can be coded into most any language), as the B symbolic information is simply not available in all cases for all I images.  (And that approach can also be more portable.)  On OpenVMS, the  A symbolic information gets stripped out of production images, for  F instance, and can be present or absent depending on how the object or G the image is built.  (You can see this case when an image crashes, and   no symbols are shown.)  I    The debugger and the traceback tool can certainly get at the embedded  I call-stack and traceback information -- when it is present -- and can be  G called from within an application.   John Reagan has posted an example  G of dynamically activating the debugger, and there's another an example  / of signaling SS$_DEBUG over in the OpenVMS FAQ  % <http://www.hoffmanlabs.org/vmsfaq/>.   I    As for generating the traceback, there's a TBK$ routine that has been  , documented in the OpenVMS V8.3 manuals, see I <http://h71000.www7.hp.com/DOC/83final/4493/4493pro_070.html>.  Routines  G that attempt to symbolize are also included, as are some code examples. A There's an older TBK$SHOW_TRACEBACK around, but AFAIK that's not   documented.  For that, see  1 <http://h71000.www7.hp.com/wizard/wiz_5175.html>.   H    There may well be more or different coding examples over in the ITRC B Search Assistant area (formerly Natural Language Search Assistant A (NLSA), formerly AskQ), as that is where such usually get posted.   C    If you want to go after the call stack and symbolic information  F yourself (whether directly, or even using the RTL), expect to have to H code it for each architecture you plan to support -- see the invocation I context routines in the LIB$ RTL as a start.  The RTL calls in this area   are platform-specific.  G    What you are seeking here is typically non-modular and non-portable  G and platform-specific, and it's also something that the OpenVMS hasn't  C typically made getting at very easy.  (And I'd not expect it to be  H consistent from one platform to the next; this stuff is deep within the H layer of OpenVMS that tends to change when it gets ported.)  And as for D non-modular, I've seen folks that code conditionally using the call D stack, basically looking back through the call stack and coding the C called routine to act differently depending on how or where it got  D called.  (This is a variation of the classic computer science "come * from" transfer of control statement. :-) )  G    I'd ask you to elaborate here.  What are you up to, and why are you   after this information?    ------------------------------   End of INFO-VAX 2006.634 ************************