1 INFO-VAX	Sat, 17 Apr 2004	Volume 2004 : Issue 213       Contents: Re: Back in the saddle> Re: How to turn linux into VMS - memory refresher for Dave ...> Re: How to turn linux into VMS - memory refresher for Dave ... Re: Linking XV for OpenVMS Re: Linking XV for OpenVMS Re: Linking XV for OpenVMS4 Re: Linux on its way out - unless you are a geek ...4 Re: Linux on its way out - unless you are a geek ...4 Re: Linux on its way out - unless you are a geek ... Re: PCanywhere and VAX VMS Re: PCanywhere and VAX VMS Re: PCanywhere and VAX VMS Re: VAX/VMS-WNT integration...  F ----------------------------------------------------------------------   Date: 16 Apr 2004 19:40:26 GMT+ From: "Doc." <doc.cypher@openvms-rocks.com>  Subject: Re: Back in the saddle 7 Message-ID: <Xns94CDDCBFFB668dcovmsrox@212.100.160.123>   ( Dan O'Reilly <dano@process.com> wrote in; news:6.0.0.22.2.20040416095128.021c5dd8@raptor.psccos.com:    H > I started programming on an old IBM batch-based system and punch cardsD > (don't remember the model).  3-day turnaround was considered greatE > time.  The first I programmed directly was a Perkin-Elmer Interdata G > system.  5 ASR33-type terminals and hex keypad to boot.  Booting took C > about 45 minutes (1+ pages of hex input on the keypad and 3 paper E > tapes to read).  No permanent storage; when you were done with your H > program and wanted to save it, you punched a new paper tape of it that5 > you had to read in next time you wanted to run it.   > ' > But heck, it was interactive!  <grin>   J C'mon, this sounds like something for Monty Python, cue the "when I was a  lad" shtuff. :-)  I I'm sure there's plenty of people here who can warp real-life experience  # into something highly entertaining.      Doc. --  G OpenVMS:     Eight out of ten hackers prefer *other* operating systems. G http://www.openvms-rocks.com    Deathrow Public-Access OpenVMS Cluster.    ------------------------------    Date: 17 Apr 2004 10:25:58 -0100* From: "Michael Kraemer" <M.Kraemer@gsi.de>G Subject: Re: How to turn linux into VMS - memory refresher for Dave ... 0 Message-ID: <40810636.MD-1.4.4.M.Kraemer@gsi.de>   > L >   I think we all have realized that the zero terminate strings of Unix mayE >   cause many problems with buffer overflows.  However, the projects M >   implementing internet services Unix or Linux know this problem.  They are M >   avoiding calling system calls which may return zero terminated strings of N >   unknown length, and they implement libraries with call interfaces with the@ >   same benifits as VMSes descriptors.  This makes most of this >   argument void.< >   So much more because VMS has a descriptor class for zero >   terminated strings, K >   and if you use that, then you have the same problems as Unix and Linux. L >   Further, it is up to the coders to make sure that the information in the> >   descriptors are right.  Thus, you are in a situation where >   security on both? >   platforms depend on coders coding to profesional standards.  > N > Of course, if you had used PL/I instead of C then you wouldn't have had this > problem in the first place.  >   E my PL/I memory is a bit rusty (10+ years ago, on IBM mainframes) over A this one, but I seem to remember that string returns of the kind  
 CHAR (*) VAR  ) weren't allowed, one had to use sth like   CHAR (255) VAR  ' instead, an invitation to buffer oflos. F Moreover, range checking (STRINGRANGE or -SIZE condition IIRC) weren't> the compiler default, for performance reasons. Such "features"B certainly do not enforce code security. Now of course things mightD have improved since then, but it shows that a language doesn't implyF security by itself. And, to add a few things, in those times it wasn't< uncommon to mess with string and array descriptors directly,7 to allow things which could not be done with legal PL/I 0 constructs, e.g. mapping of contiguous memory on? multi-dimensional arrays (IIRC). Such cheats were possible via  D (legal) PL/I pointer logic or, for performance reasons, via assembly subroutines.   ------------------------------    Date: 16 Apr 2004 15:39:03 -0700( From: bob@instantwhip.com (Bob Ceculski)G Subject: Re: How to turn linux into VMS - memory refresher for Dave ... = Message-ID: <d7791aa1.0404161439.466c3e8b@posting.google.com>   _ "Karsten Nyblad" <nospam@nospam.com> wrote in message news:<c5p1rs$1tbl$1@news.cybercity.dk>... 7 > "Bob Ceculski" <bob@instantwhip.com> wrote in message 9 > news:d7791aa1.0404160520.13ddd8f7@posting.google.com... G > > Yes, but first redesign and rewrite your unix to cleanly catagorize  > > and separateJ > > Kernel Mode from Supervisor Mode and from User Mode. Three modes are a > > minimum H > > for a correct ring protection system. The use of three or more rings > > happens toG > > be a fully patented methodology by OpenVMS Engineering. OpenVMS has 	 > > four. I > > OpenVMS also has 40 groups of higher mode functionality classified as 
 > > requiring  > > special named privileges.  > M > There is an important difference between Unix and VMS.  In Unix the command K > language interpreter is executed in a process separate from the processes N > executing user programs.  In VMS DCL is executed in the same process as userN > programs.  VMS needs supervisor mode to separate DCL and DCLs data from userF > programs.  Unix does not have that need because the command languageI > interpreter's code and data a protected by being in a separate process. K > Please tell me why the VMS is way of doing things is more secure than the  > Unix way.   ; and while you are checking out the cert counts for the last < ten years, check out this link below to TCPware, an IP stack8 for VMS ... notice the few certs listed had no or little= effect (access violation error) on VMS ... how did unix fare?   , http://www.process.com/techsupport/cert.html   ------------------------------  % Date: Fri, 16 Apr 2004 14:57:55 -0500 6 From: "Craig A. Berry" <craigberry@mac.com.spamfooler># Subject: Re: Linking XV for OpenVMS E Message-ID: <craigberry-584129.14575516042004@chi.news.speakeasy.net>   + In article <40800c76$1@obsidian.gov.bc.ca>, ,  "Alder" <PGDEHMKOKIMD@spammotel.com> wrote:     > H > It was my feeble attempt to resolve another set of "undefined" symbolsB > that I was getting before I added the "/NAMES=(AS_IS,SHORTENED)"D > qualifier.  They all related to the third-party imaging libraries:$ > LibJPEG, LibPNG, and LibTIFF, i.e. > 	 >     ... J >     LINK /EXEC=XV.EXE /NoDebug xv.obj,Sys$Disk:[]DECC_OPTIONS.OPT/Option- >     %LINK-W-NUDFSYMS, 70 undefined symbols: 1 >     %LINK-I-UDFSYM, JPEG_CALC_OUTPUT_DIMENSIONS 	 >     ... - >     %LINK-I-UDFSYM, PNG_CONVERT_FROM_TIME_T 	 >     ...  >     %LINK-I-UDFSYM, TIFFCLOSE 	 >     ... D >     %LINK-W-USEUNDEF, undefined symbol JPEG_CALC_OUTPUT_DIMENSIONS > referenced' >     in psect $LINK$ offset %X00000300 > >     in module XVJPEG file DISK$NET:[XV-3_10A-VMS]LIBXV.OLB;1	 >     ... J >     %MMK-F-ERRUPD, error status %X10648268 occurred when updating target > XV.EXE > B > The symbols from my other third-party library, LIBZ, are gettingD > resolved OK when I leave out the linker's "/NAMES" qualifier.  TheJ > difference seems to be that all the symbols in LIBZ are defined in lowerG > case, according to the output of the LIBRARY/NAMES/LIST command.  The F > symbols in the three imaging libraries appear to be defined in mixed > case.  > . > Any suggestions for what steps to take next?  F You need to compile everything, including all the libraries, with the G same /NAMES option (or without the option).  If one library or package  A requires /NAMES=(AS_IS,SHORTENED), then you have to use that for   everything you link against it.    ------------------------------    Date: 16 Apr 2004 15:52:57 -0600; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler) # Subject: Re: Linking XV for OpenVMS 3 Message-ID: <5ebstO7WSbsn@eisner.encompasserve.org>   X In article <40803845$1@obsidian.gov.bc.ca>, "Alder" <PGDEHMKOKIMD@spammotel.com> writes: > ) > Was that a dumb question, or what?  :^)  >   C    There are no dumb questions on c.o.v, there is only Andrew.  8-)    ------------------------------  % Date: Fri, 16 Apr 2004 15:12:39 -0700 * From: "Alder" <PGDEHMKOKIMD@spammotel.com># Subject: Re: Linking XV for OpenVMS + Message-ID: <40805a56$1@obsidian.gov.bc.ca>   5 "Alder" <PGDEHMKOKIMD@spammotel.com> wrote in message % news:40803845$1@obsidian.gov.bc.ca... B > "Bob Koehler" <koehler@eisner.nospam.encompasserve.org> wrote in message / > news:nGK+S3qMP7c1@eisner.encompasserve.org... 7 > > In article <40800c76$1@obsidian.gov.bc.ca>, "Alder" & > <PGDEHMKOKIMD@spammotel.com> writes: > > >    > H > Any ideas why the map file is showing the LibTIFF location differently > than the others? >   < Scratch that question.  The docs for the LINK utility state:  E File. Full file specification of the input file, including device and C directory. If the specification is longer than 35 characters, it is E shortened by dropping the device portion of the file specification or A both the device and directory portions of the file specification.   D Which of course still begs the question of the undefined symbols for TIFF.    Cheers,    Terry    ------------------------------  % Date: Fri, 16 Apr 2004 15:22:36 -0400 * From: Chuck Chopp <ChuckChopp@rtfmcsi.com>= Subject: Re: Linux on its way out - unless you are a geek ... : Message-ID: <efWfc.35802$UC4.25042@bignews2.bellsouth.net>   Andrew Harrison wrote:   >>>>I >>>> I do agree that clustering on every other server platform has been a @ >>>> joke. I still don't see a distributed lock manager in otherH >>>> clustering solutions to allow for concurrent access to file systems< >>>> from multiple clustered nodes.  The most popular of theA >>>> Windows/NetWare/Linux clustering solutions provide more of a ? >>>> fail-over rather than a loosely-coupled type of clustering G >>>> environment.  I laugh everytime I see one of those other operating < >>>> systems being touted as having clustering capabilities. >>>> >>> % >>> Well you haven't looked very far.  >>>  >>> How about looking here.  >>> > >>> http://www.sun.com/storage/software/data_mgmt/performance/ >>> + >>> Incedentally this is not a new product.  >>>  >>> Who is laughing now ?  >>>  >> >>C >> He still is laughing.  Like he said, he found the others a joke.  >> >>' > I think you need to re read his post.  > @ > He said he didn't find anything that allowed concurrent access/ > to filesystems from multiple clustered nodes.  > ; > He has now been shown one and his laughing days are over.  > 9 > Unless you think he has actually tried QFS ! do you ???  > 	 > Regards  > Andrew Harrison  >     ) Sun StorEdge Performance Suite Software -   G  From what I see, it is a storage system add-on via an additional file  L system that can be installed, not a true fully-featured clustering solution J that is an integral part of the operating system.  It is aimed at SAN and J NAS storage, but does it also allow for concurrent read/write access from M multiple cluster nodes regardless of when the storage is locally attached to  E a subset of the cluster nodes where each node has multiple redundant   physical paths to the storage?  F I don't see it as being functionally equivalent to OpenVMS Clustering J technology.  Perhaps I simply set too high a standard when it comes to my M expectations of what a "clustering solution" provides in the way of features  J and functionality.  I spent so many years working with OpenVMS Clustering F that now I'm spoiled.  When I refer to a cluster, I'm talking about a E loosely-coupled configuration where multiple nodes all boot the same  M operating system and coordinate access to all shared cluster resources via a  I distributed lock manager.  Furthermore, all of the coordinated access to  L resources is integrated as part of the operating system such that there are M absolutely no differences in how applications are coded in order for them to  L run on either a standalone node or on one or more nodes in a cluster.  This M includes not needing to migrate data onto special volumes that run a special  I file system.  Other than having to enable MSCP serving of disks, there's  M nothing else that needs to be done for a cluster member to serve out locally  G attached disks to other cluster members.  Likewise, cluster nodes with  L multiple physical paths to the disks will coordinate their access via those J physical paths while simultaneously coordinating with other cluster nodes > that access the disks via network-based cluster interconnects.    M As for lumping Sun in at the start of the thread... it was just to "get your  M goat"... and it did.  I hadn't seen you posting much here lately and figured  J that you'd bark most ferociously if a little flame bait popped up on your 0 radar screens.  You took it hook, line & sinker.       --   Chuck Chopp   8 ChuckChopp (at) rtfmcsi (dot) com http://www.rtfmcsi.com  @ RTFM Consulting Services Inc.     864 801 2795 voice & voicemail2 103 Autumn Hill Road              864 801 2774 fax Greer, SC  29651  , Do not send me unsolicited commercial email.   ------------------------------    Date: 16 Apr 2004 15:41:00 -0600; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler) = Subject: Re: Linux on its way out - unless you are a geek ... 3 Message-ID: <WDyJPeZ9L6KU@eisner.encompasserve.org>    In article <c5oh4n$9i3$1@new-usenet.uk.sun.com>, Andrew Harrison SUNUK Consultancy <Andrew_No.Harrison_No@nospamn.sun.com> writes: >  > How about looking here.  > < > http://www.sun.com/storage/software/data_mgmt/performance/ > ) > Incedentally this is not a new product.  >  > Who is laughing now ?   B    Folks who needn't limit themselves to fiber-network SANs.  The =    pages product you site claims the product to be limited to ?    fiber-network SANs.  That's not what everybody has or needs.   C    A couple decades ago you could have compared this to VAXclusters H    being limited to CI.  So Solaris is only a couple decades behind VMS,1    not bad seeing how UNIX had a one decade lead.    ------------------------------  # Date: Sat, 17 Apr 2004 17:16:27 GMT ! From: Nigel Barker <nigel@hp.com> = Subject: Re: Linux on its way out - unless you are a geek ... 8 Message-ID: <cco180duusrjja1ush7j33kapp1op35ma9@4ax.com>  E On Thu, 15 Apr 2004 15:14:31 +0100, Andrew Harrison SUNUK Consultancy . <Andrew_No.Harrison_No@nospamn.sun.com> wrote:   >Chuck Chopp wrote: = >Why would we also spend loads of money developing StarOffice   J Should't that be 'spend loads of money purchasing the application from theP original German developers then spend more spiffing it up into a product that we can make money out of'?   > >HP for example talks a wonderfull talk about Linux but really@ >only provides commodity platforms to run Linux and what exactly# >is the incremental value in that.    M Glad to see that you finally acknowledge Itanium as a commodity platform. Not P many other sources for big honking SMP Linux machines (if that's what you want).   -- Nigel Barker Live from the sunny Cote d'Azur    ------------------------------  % Date: Sat, 17 Apr 2004 10:25:22 +0200 , From: "Hans Vlems" <hvlems.dotweg@zonnet.nl># Subject: Re: PCanywhere and VAX VMS 9 Message-ID: <c5qppd$4o2cr$1@ID-143435.news.uni-berlin.de>   C "Dominic Olivastro" <DOlivastro@ChiResearch.com> schreef in bericht ? news:32b02$4080351d$44a5e110$18043@msgid.meganewsservers.com... G > Does anyone have any experience with PCAnywhere on a computer that is  > connected to VAX VMS?  >  > Here is our setup:C > 1.  Several PC's connected via Ethernet to each other and the VAX L > 2.  The VAX is running Pathworks server and the PC's are running PathworksG > client software.  The PC's often use PowerTerm to connect to the VAX. 8 > 3.  A remote computer for an employee working at home.D > 4.  One PC at work is running PCAnywhere (the host) and the remote computerI > is running PCAnywhere (the client).  The remote computer calls up via a  > modem. >  > Here is what we found:= > 1.  The remote computer can access the host via PCAnywhere. C > 2.  The remote can run any software on the host (eg, Excel, Word, 	 homegrown  > software, etc)H > 3.  The remote can NOT run PowerTerm, the terminal emulator that comes with4 > Pathworks and allows the user to log into the VAX. > I > Of course, the third point is the problem.  Does anyone have experience  with > this?  >  > Dom  >  > --   > Dominic Olivastro  > CHI Research, Inc  > # > web:   http://www.ChiResearch.com  > fax:     1-856-546-9633 ! > voice: 1-856-546-0600 (ext 224) $ > email:  DOlivastro@ChiResearch.com >  > L There's no reason it shouldn't work. I used that setup a couple of years agoH to solve a queue problem on a VAX. I used PCanywhere to connect to an NTK system and ran the VT320 emulator that comes (came?) with Pathworks. The NT - system used LAT to connect to the VAX though. L Could it be that the PC that runs the PCanywhere server is confused, IP wise that is?L The VAX obviously runs LAT. Even if it isn't running yet, LAT may be startedA with @sys$startup:lat$startup and live with the default settings. 7 Run Powertem over LAT (or DECnet) and see what happens.    Hans   ------------------------------  % Date: Fri, 16 Apr 2004 13:07:57 -0700 & From: Tom Crabtree <tccrab@sunset.net># Subject: Re: PCanywhere and VAX VMS , Message-ID: <c5pefc01hfe@enews3.newsguy.com>  G > Does anyone have any experience with PCAnywhere on a computer that is  > connected to VAX VMS? < I've got a similar setup, but with Alpha's instead of VAX's. > M > 3.  The remote can NOT run PowerTerm, the terminal emulator that comes with 4 > Pathworks and allows the user to log into the VAX.F We use Hummingbird Exceed, which works just fine for us, but you will ' need to have TCP/IP running on the VAX.  Why is PowerTerm not working?    TomC   ------------------------------  % Date: Fri, 16 Apr 2004 17:45:18 -0400 6 From: "Dominic Olivastro" <DOlivastro@ChiResearch.com># Subject: Re: PCanywhere and VAX VMS E Message-ID: <d8b06$40805125$44a5e110$19797@msgid.meganewsservers.com>   3 "Tom Crabtree" <tccrab@sunset.net> wrote in message & news:c5pefc01hfe@enews3.newsguy.com...I > > Does anyone have any experience with PCAnywhere on a computer that is  > > connected to VAX VMS? > > I've got a similar setup, but with Alpha's instead of VAX's.   > > J > > 3.  The remote can NOT run PowerTerm, the terminal emulator that comes with6 > > Pathworks and allows the user to log into the VAX.G > We use Hummingbird Exceed, which works just fine for us, but you will ) > need to have TCP/IP running on the VAX.  > Why is PowerTerm not working?   I When the client computer tries to run PowerTerm on the host, the client's J connection to the host just dies, and the user is forced to call-up again.C I've never heard of Hummingbird Exceed, and I don't want to get new / software.  Is anyone using PCAnywhere on a VAX.    Dom " mailto::DOlivastro@CHIResearch.com   ------------------------------  % Date: Sun, 18 Apr 2004 03:02:59 +0000 8 From: David McKenzie <david.mckenzie@paradigm-shift.biz>' Subject: Re: VAX/VMS-WNT integration... C Message-ID: <40814fa1$0$18988$c30e37c6@lon-reader.news.telstra.net>    Chris Doran wrote:  : > contracer11@uol.com.br (Shiva MahaDeva) wrote in message; > news:<ddf392ea.0404140631.6a8f778e@posting.google.com>... G >> I'd like issue the command "at" in a WNT system from VMS system...is G >> there any way ? I can issue Solaris commands from VMS using RSH, but  >> could I use RSH with WNT ?  > F > I had RSH from the NT4.0 Server resource kit working for many years.E > It enabled me to do editing on VMS (how do people live without file G > versions?) and run a command procedure to FTP files to NT and compile C > and link them. ISTR I had to do a bit of experimentation with the ? > login name to get it working, and the required syntax changed H > mysteriously when I got a new NT machine. I can provide a copy of thisF > command procedure (or another to do the same thing on HP-UX). But beC > warned it will need a lot of customisation. I also have some baby G > command procedures that FTP files to/from NT to parallel directories.  > F > The bad news is that none of this works on W2k Workstation (it mightB > on W2k Server). Worse, it stopped working on NT whilst trying toB > reinstall Oracle, and I've never managed to get RSH going again, > though FTP still does. >  > Chris   & cygwin rsh ssh and more (and its free)   --   David McKenzie  remove rugby  - Web:            http://www.paradigm-shift.biz 7 Mail            David.McKenzie@paradigm-shift.rugby.biz    ------------------------------   End of INFO-VAX 2004.213 ************************