1 INFO-VAX	Wed, 23 Mar 2005	Volume 2005 : Issue 163       Contents:M "Access Denied" Error from client connecting to Pathworks Advanced Server 6.1 P RE: "Access Denied" Error from client connecting to Pathworks Advanced Server 6.: (",) Do You Want To Know For Sure You Are Going To Heaven?- Re: DECnet, LAT, Cterm over IP/GRE - Timeout? ; Itanium Developer Workshop Schedule (rx1620 server for $2K) ( Need help on messed up VMS PCSI database7 Re: Need help to port VAX code to Alpha and to Itaninum 
 Re: opcdef
 Re: opcdef Remote queue on a linux server Re: resizing terminal window3 Re: Status bits for OFFLINE and MntVerifyTimeout ?? 3 Re: Status bits for OFFLINE and MntVerifyTimeout ??  Re: Windows anti-spyware5 Re: Word-11 font suddenly stopped working.  Now what?  Re: [Announce] FreeVMS 0.1.3 Re: [Announce] FreeVMS 0.1.3 Re: [Announce] FreeVMS 0.1.3B Re: [Java JNI] How to load a shared img without defining a symbol?  F ----------------------------------------------------------------------    Date: 22 Mar 2005 12:59:31 -0800/ From: "ropegun1@yahoo.com" <ropegun1@yahoo.com> V Subject: "Access Denied" Error from client connecting to Pathworks Advanced Server 6.1C Message-ID: <1111525171.536708.292850@z14g2000cwz.googlegroups.com>   D I am trying to get several PC clients running Pathworks 4 for Dos toD connect to an Alpha server running Pathworks 6.1 via DecNet.  I keepA getting an Error: Access Denied on the client PC's when trying to G connect to a share on the Alpha.  The server is a DS20E running OpenVms D 7.3-2.  We have an Alpha DS10L server running the same configurationF except the OS is 7.3-1 and it works fine. I have everything configuredG the same a far as I can tell.  I can't seem to find anything in the log G files indicating a problem.  Here are some of the things I have checked  so far. B Pwrk$guest and pwrk$default are the same as on the working server." Ran the admin utility and checked,F hostmap,services,computers,shares,trusts etc.  I think I have them all set up the same.< I am using Server Based Licensing according to the Pathworks documentation.4 I have a valid PWLMXXXCA07.03 license on the server.? The DS20E is set up as a primary domain controller.  We will be * replacing the DS10L with this  new server., All file and user permissions look the same.D I change searched the batch files on the clients and changed them to@ point to the new node name and Decnet address.  If I do a  "show= known nodes" from ncp on the client I can see the new server. E I have run out of ideas on things to check.  Any ideas on what may be F causing this "Access Denied Error"?   This is my first experience with= Pathworks so any help you guys could provide would be greatly  appreciated.   Thanks,  Heath Harber Special Metals Corp.   ------------------------------  % Date: Tue, 22 Mar 2005 18:02:25 -0500 ' From: "Main, Kerry" <kerry.main@hp.com> Y Subject: RE: "Access Denied" Error from client connecting to Pathworks Advanced Server 6. R Message-ID: <FD827B33AB0D9C4E92EACEEFEE2BA2FB59526C@tayexc19.americas.cpqcorp.net>   > -----Original Message-----9 > From: ropegun1@yahoo.com [mailto:ropegun1@yahoo.com]=20  > Sent: March 22, 2005 4:00 PM > To: Info-VAX@Mvb.Saic.Com = > Subject: "Access Denied" Error from client connecting to=20  > Pathworks Advanced Server 6.1  >=20F > I am trying to get several PC clients running Pathworks 4 for Dos toF > connect to an Alpha server running Pathworks 6.1 via DecNet.  I keepC > getting an Error: Access Denied on the client PC's when trying to < > connect to a share on the Alpha.  The server is a DS20E=20 > running OpenVms F > 7.3-2.  We have an Alpha DS10L server running the same configurationH > except the OS is 7.3-1 and it works fine. I have everything configuredA > the same a far as I can tell.  I can't seem to find anything=20  > in the log? > files indicating a problem.  Here are some of the things I=20  > have checked	 > so far. D > Pwrk$guest and pwrk$default are the same as on the working server.$ > Ran the admin utility and checked,H > hostmap,services,computers,shares,trusts etc.  I think I have them all > set up the same.> > I am using Server Based Licensing according to the Pathworks > documentation.6 > I have a valid PWLMXXXCA07.03 license on the server.A > The DS20E is set up as a primary domain controller.  We will be , > replacing the DS10L with this  new server.. > All file and user permissions look the same.F > I change searched the batch files on the clients and changed them toB > point to the new node name and Decnet address.  If I do a  "show? > known nodes" from ncp on the client I can see the new server. G > I have run out of ideas on things to check.  Any ideas on what may be H > causing this "Access Denied Error"?   This is my first experience with? > Pathworks so any help you guys could provide would be greatly  > appreciated. >=20	 > Thanks,  > Heath Harber > Special Metals Corp. >=20  
 Heath -=20   A few misc thoughts - @ - can you "set host x" via decnet from the client to the server?= - if passwords in play, are you using upper or lower case?=20 F - are their routers in between client and server? To eliminate network= devices, put clent on small local hub with server and retest. G - is TCPIP installed/enabled on the client? If yes, can you telnet from  client to server?    Regards   
 Kerry Main Senior Consultant  HP Services Canada Voice: 613-592-4660  Fax: 613-591-4477  kerryDOTmainAThpDOTcom (remove the DOT's and AT)=20  $ "OpenVMS has always had integrity .. Now, Integrity has OpenVMS .."   ------------------------------    Date: 22 Mar 2005 13:43:24 -0800 From: Ron038551@yahoo.com C Subject: (",) Do You Want To Know For Sure You Are Going To Heaven? B Message-ID: <1111527804.585946.98930@g14g2000cwa.googlegroups.com>  8 http://www.want-to-be-sure.blogspot.com << Click On Link   ------------------------------  + Date: Tue, 22 Mar 2005 21:36:10 +0000 (UTC) 7 From: moroney@world.std.spaamtrap.com (Michael Moroney) 6 Subject: Re: DECnet, LAT, Cterm over IP/GRE - Timeout?( Message-ID: <d1q34a$84o$1@pcls4.std.com>  / Scott Fisher <scott.fisher@remmele.com> writes:    >Situation: H >We have a WAN (T1) with Alphas, DECservers (DS100, DS200), and windows I >boxes.  Everything is fine.  We run IP, DECnet, MOP, etc across the wan  H >without problems.  The DS100 downloads across the wan without problem. G >We want to move to MPLS which requires that all packets are IP  thus  F >the problem with the DS100 and DECnet.  We have an older application G >that requires VTs and attached printers and it will not be going away  J >for a few years.  Also, we rely on VMS Copy, Cterm, and submit/remote on I >the remote alphas.  I know we can do things with ftp, telnet, dcl, etc,  4 >but it would be nice not to have to mess with that.  I >3.    We can connect from same VT to Alpha-B (across the MPLS with GRE)  I >and it will connect just fine.  PROBLEM:  After about 20-30 minutes, it  I >will disconnect from the service  the VT will talk to the DS100, but  E >we are logged out of alpha-B.  This problem happens at varying time    >intervals but 100% of the time.  G LAT assumes a LAN situation.  It makes many assumptions (latency etc.)  I that are true in a LAN but not necessarily so on a WAN.  I guess your T-1 F is good enough so that that doesn't matter but encapsulating LAT on IPF makes things bad enough to matter.  You can replace LAT with telnet IP4 sessions, newer DECservers also know IP (see below).  C >5.    From VT on DS100, connect to Alpha-A, the $set host alpha-B  H >(across the network), connects fine, but times out after 5-30 minutes. 0 >Again, times vary but happens 100% of the time.  J DECnet can deal with WANs just fine but in this situation it thinks all isG on LANs so it probably makes similar assumptions as LAT does.  You may  ; want to look into DECnet-over-IP (requires DECnet Phase V).   H >So, it seems like we have some kind of a timeout going on, but I cant E >think of where to go next.  I apologize for the long message, but I   >think the data is important.   H From what I've seen, IP (the TCP protocol in particular) isn't as robustE in detecting timeouts/lost packets as other protocols such as LAT and G DECnet.  LAT will detect a lost packet & request retransmission and the G user usually sees a 1 second hang or a very quick disconnect (from both F ends).  DECnet often takes longer but both ends seem to agree that theJ link is disconnected.  On the other hand, all the time I see the followingJ on telnet:  Session hangs for a long time, finally the local side gives upE and disconnects, I telnet back in and the old session is still there, J thinking it's still connected.  Something isn't dealing with lost packets/ NAKs/whatever correctly.  I Seems that you're getting the disadvantages of IP without any advantages. I Perhaps you should stay with the old configuration until your application  is retired?  --   -Mike    ------------------------------  % Date: Tue, 22 Mar 2005 20:56:24 -0500 ' From: "Main, Kerry" <kerry.main@hp.com> D Subject: Itanium Developer Workshop Schedule (rx1620 server for $2K)R Message-ID: <FD827B33AB0D9C4E92EACEEFEE2BA2FB59527C@tayexc19.americas.cpqcorp.net>  = Readers of this list might be interested in the following:=20    ++++++++++++++++++++++++++  H http://h21007.www2.hp.com/dspp/bus/bus_BusDetailPage_IDX/1,1252,7255,00. htmlE "Last year's developer series finished as a resounding success, so HP G and Intel(r) have continued their combined efforts, creating a complete C suite of products and services to support developer adoption of the  Intel Itanium(r) architecture.  F As part of the Developer Workshop, HP is making an HP Integrity rx1620? Server available to attendees for a special price of $2,000 (US E dollars). This price includes admission to the HP and Intel Developer @ Workshop. After three days of interactive lectures and proctored? hands-on labs in HP-UX 11i, SUSE Linux (64-bit), HP OpenVMS, or ; Windows(r) Server 2003 environments, you'll leave with your F Itanium-based applications porting efforts well underway.  Plus, an HPH Integrity rx1620 Server featuring the Intel Itanium 2 processor, will beE shipped with the appropriate operating system and the latest software  development tools.  A You'll also have membership in both the HP Developer and Solution E Partner Program (DSPP) and the Intel Early Access Program. Everything : you need to complete the migration of your products to theD performance-leading, industry-standard architecture. Servers will beC shipped directly to you after the workshop so you can continue your + porting efforts to Itanium architecture.=20    April 26 - 28, 2005  Tampa, Florida   July 12 - 14, 2005 Washington, DC   August 23 - 25, 2005 San Diego, California    October 18-20, 2005  Houston, Texas  E For more information, including workshop agendas, course descriptions " and registration information go toE http://www.hp.com/go/itaniumdeveloperforums or call 1-888-647-7755 or  603-644-3791 ex.587. ++++++++++++++++++++++++++   Regards   
 Kerry Main Senior Consultant  HP Services Canada Voice: 613-592-4660  Fax: 613-591-4477  kerryDOTmainAThpDOTcom (remove the DOT's and AT)=20  $ "OpenVMS has always had integrity .. Now, Integrity has OpenVMS .."   ------------------------------  # Date: Tue, 22 Mar 2005 23:16:53 GMT , From: david-m-jones@nospam.net (David Jones)1 Subject: Need help on messed up VMS PCSI database > Message-ID: <FH10e.41797$wl4.1538117@twister.southeast.rr.com>  N I instaled a product a designated hard drive. The drive then crashed before I I made a full backup of it. So I could not re-install it as PRODUCT utility F said it was already installed. I could not PRODUCT REMOVE it since theK the installed directory did not now exist. I searched the Internet for some K help and tried PRODUCT REGISTER on that product. It removed the product but G now PRODUCT SHOW PRODUCT lists zero products. The product pcsi database H files still exist in sys$system and at lease one main pcsi database file' has the complete (I think) information.   E Does anyone know how to rebuild the pcsi database so it will list the  installed products?   $ Thanks much in advance for any help. David Jones    ------------------------------  % Date: Tue, 22 Mar 2005 16:13:09 -0500 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> @ Subject: Re: Need help to port VAX code to Alpha and to Itaninum+ Message-ID: <42408A5D.5C4A5A3@teksavvy.com>    Bob Koehler wrote:H >    The ANSI C standard now states that for any array named x, the nameJ >    alone and the name preceded by & both return the address of the firstG >    element of x.  This didn't use to be, so if you are compiling with < >    different standards you can expect different behaviour.  C I tend to no trust these things and do &x[0] to make sure I get the ; address of first element, especially when x is a structure.   E I think that ANSI C broke a big rule with char arrays being pointers, $ and not really alowing &mychararray.  A So when not so sure, I use &mychararray[0].  (although for simple H character arrays, this is not an issue, but again, within structures, it can be inteesting).    ------------------------------  % Date: Tue, 22 Mar 2005 15:24:34 -0500 3 From: Steve Lionel <Steve.Lionel@REMOVEintelME.com>  Subject: Re: opcdef 8 Message-ID: <vav041lg3gi6o5oiuqkuftd0ga3knqsnmk@4ax.com>  I On 22 Mar 2005 06:48:05 GMT, "Dave Weatherall" <djw-nothere@nospam.nohow>  wrote:  C >> That's CDEC$ or !DEC$.  The WARN=NOALIGN is what is wanted here.  > E >Thanks Steve - I felt something wrong when I typed it. My brain was  G >not up to programming speed after spending  2 weeks giving a training   >course.  D I thought you folks might get a kick out of this.  You know that theM DEC/Compaq Fortran compiler sources were brought over to Intel in 2001 (along L with most of the Fortran compiler engineers, such as myself.)  Well, there'sI this DEC$ prefix in the directive syntax.  You all know what it stood for N originally.  When Compaq bought DEC, nobody suggested changing it to CPQ$, butM we did get some questions at Intel about it, suggesting that we should change M it (ignoring that we'd still have to support the old syntax for compatibility > - that and what three letters would be appropriate for Intel?)  M The response of Fortran project leader Stan Whitlock was delightful.  He told K them that they misunderstood, and that the prefix meant "Directive Enhanced J Compilation" and it had to stay as-is.  And as thus it is described in the Intel Fortran documentation.     Steve Lionel Software Products Division Intel Corporation 
 Nashua, NH  8 User communities for Intel Software Development Products"   http://softwareforums.intel.com/ Intel Fortran Support 7   http://developer.intel.com/software/products/support/    ------------------------------  % Date: Tue, 22 Mar 2005 15:56:40 -0500 ' From: Dave Froble <davef@tsoft-inc.com>  Subject: Re: opcdef 0 Message-ID: <114111b4dme89e2@corp.supernews.com>   Steve Lionel wrote:   F > I thought you folks might get a kick out of this.  You know that theO > DEC/Compaq Fortran compiler sources were brought over to Intel in 2001 (along N > with most of the Fortran compiler engineers, such as myself.)  Well, there'sK > this DEC$ prefix in the directive syntax.  You all know what it stood for P > originally.  When Compaq bought DEC, nobody suggested changing it to CPQ$, butO > we did get some questions at Intel about it, suggesting that we should change O > it (ignoring that we'd still have to support the old syntax for compatibility @ > - that and what three letters would be appropriate for Intel?)  H Pointy haired bosses everywhere, and all with equal intellegence .. zero  @ How many members of the team were ready for a nervous breakdown?  O > The response of Fortran project leader Stan Whitlock was delightful.  He told M > them that they misunderstood, and that the prefix meant "Directive Enhanced L > Compilation" and it had to stay as-is.  And as thus it is described in the > Intel Fortran documentation.   And now secure forever.  :-)   ------------------------------    Date: 22 Mar 2005 21:35:14 -0800" From: mrs2001@libero.it (Maurizio)' Subject: Remote queue on a linux server = Message-ID: <1d5d30a9.0503222135.52bbdaaa@posting.google.com>    HI,    I need your help.   ' I've a Digital PC with OpenVms V.7.1-2.   I Some weeks ago, I'd to change my HD but my image backup wasn't very good.   ? So, after restore, I'd some problems with a remote print queue. 3 I've a linux print server with an OCE 9400 Plotter.   " I defined print queue in this way:   $$$ init/queue oce8 /on=oce8::  < But when I try to start my queue, this message is displayed:   $$$ start/queue oce8: %JBC-E-NOSUCHNODE, remote node is unknown or not reachable  % I ping this host (oce8) successfully:   F UCX> ping oce8                                                         %UCX-I-LOOPACT, OCE8 is alive      Please help me   Thanks,    md'a   ------------------------------    Date: 22 Mar 2005 11:40:33 -0800 From: drew.shelton@sematech.org % Subject: Re: resizing terminal window C Message-ID: <1111520433.759840.185410@l41g2000cwc.googlegroups.com>    Tom Linden wrote: D > Have a somewhat vexing nuisance. I us PuTTY on a W2K box to accessG > my VMS cluster.  Each box has a different background color associated E > with it, so visually I can see which system I am on.  Of course, it  isA > also on the xterm title bar and the dcl prompt.   Now the PuTTY  settingsF > are the same, but on a 7.3 systems both VAX and Alpha  the number of  	 > columns B > displayed  increases as the window is resized, but not on 7.3-1, 7.3-2 or > 8.2 * > systems.  Any suggestions on how to fix?  G I've never used Putty to connect to VMS, but since it's X-Windows, have @ you checked the DECW$TERMINAL_DEFAULT.DAT files on the different& systems to make sure they're the same?   Drew  L ============================================================================A Drew Shelton                            drew.shelton@sematech.org < VMS Systems Administrator               office: 512-356-7575< Sematech                                fax:    512-356-7600 2706 Montopolis Drive A Austin, TX 78741-6499                I speak for myself only, not 	 Sematech. D     "There are two major products that come out of Berkeley: LSD and UNIX. D     We don't believe this to be a coincidence." - Jeremy S. AndersonL ============================================================================   ------------------------------  + Date: Tue, 22 Mar 2005 22:35:09 +0000 (UTC) P From: helbig@astro.multiCLOTHESvax.de (Phillip Helbig---remove CLOTHES to reply)< Subject: Re: Status bits for OFFLINE and MntVerifyTimeout ??$ Message-ID: <d1q6it$r8q$1@online.de>  A In article <1111494934.024911.9680@o13g2000cwo.googlegroups.com>, & "Shankar" <passhan@gmail.com> writes:   G >    One of my applications needs to know if the disk I query for is in F > OFFLINE or MntVerifyTimeout states. I am using the SYS$GETDVI system> > call to query the characteristics of the disk. I can see theC > UCB$V_MNTVERIP and UCB$V_MNTVERPND status bits. Is there any way, F > through SYS$GETDVI, to determine if the disk is either in OFFLINE or > MntVerifyTimeout states ?    $ mntvfy = %x4000  $ valid  = %x0800  $ sts = f$getdvi(vol,"sts") C $ if (sts .and. mntvfy) .eq. mntvfy then ... ! Disk is in MntVerify J $ if (sts .and. valid) .ne. valid then ...   ! Disk is in MntVerifyTimeout   ------------------------------    Date: 22 Mar 2005 21:45:06 -0800# From: "Shankar" <passhan@gmail.com> < Subject: Re: Status bits for OFFLINE and MntVerifyTimeout ??C Message-ID: <1111556706.332727.156720@g14g2000cwa.googlegroups.com>   G Thanks Dave and Phillip. Do we assume that if a device is not available + for use, then it means that it is OFFLINE ?    Thanks,  Shankar    ------------------------------  % Date: Tue, 22 Mar 2005 16:58:43 -0500 # From: "John Smith" <a@nonymous.com> ! Subject: Re: Windows anti-spyware , Message-ID: <dOednVdKa7-BCN3fRVn-jA@igs.net>   Dave Weatherall wrote:D > On Tue, 22 Mar 2005 00:59:35 UTC, Paul Sture <paul.sture@decus.ch> > wrote: >  >> Dave Weatherall wrote:  >>
 >> <big snip>  >>; >>  > The same reason Acrobat reader stays at 5. V6 says it " >>> needs IE6 to install - bah !!! >>> B >> Dunno about the Windows version, but on my Mac I've changed theB >> default for .pdf files to Acrobat 5, because Acrobat 6 takes an >> _age_ to fire up. >> > E > That indeed was the other reason I decided against it for myself. I F > did go thro' the nausea on one of our machines at a customer site. AG > stand-alone PC that I want the customer to be able to use to read the D > PDF's of my user manuals, while he's working on VMS (we're back onD > group-topic <G>) on an adjacent VT520. I forgot to take a cover-CDE > with Acrobat 5, found 6 on that of the magazine I'd just bought and E > had to go thro' the install IE before Acrobat routine to get it all 9 > on. The policy for the machine had been to keep IE off.     L Many apps will only need IE as a prerequisite to run the installer - usuallyK just an installer thing so you can view the readme html and they don't want L to worry about Mozilla or Opera (why I don't know). But after the install isG complete it is usually possible to remove IE and the app will run fine. 9 Don't know if IE is required to run the uninstall though.   E If you have a spare machine around get your PC support people to try: ( a)  installing the app with IE installed b) uninstall IE  c) see if the app works / d) try uninstalling the app without IE present.   / Steps c) and d) should work fine in most cases.    --F OpenVMS - The never advertised operating system with the dwindling ISV base.    ------------------------------  # Date: Wed, 23 Mar 2005 00:07:15 GMT L From: winston@SSRL.SLAC.STANFORD.EDU (Alan Winston - SSRL Central Computing)> Subject: Re: Word-11 font suddenly stopped working.  Now what?6 Message-ID: <00A41281.AD88AA41@SSRL.SLAC.STANFORD.EDU>  U In article <4240346C.3070408@MMaz.com>, "Barry Treahy, Jr." <Treahy@MMaz.com> writes: - >Alan Winston - SSRL Central Computing wrote:  >  >>VMSers --  >>E >>VMS 7.3-2 on DS20E.  Word-11 5.0, essentially unchanged since 1995.  >>N >>(Apologies for Word-11 specific terminology, but I figure you probably can'tL >>help me if you don't know Word-11, so I'm not going to a lot of trouble to >>define my terms.)  >>K >>All of a sudden, documents using Font 352 no longer print.  When I try to Q >>DT them, the process aborts with an "Error 10: cannot download font to printer" O >>(which seems peculiar when there's no printer involved yet).  A document that % >>printed last month won't print now.  >> >>   >>G >Never used WP on VMS, just Mass-11, but is there a remote possibility  ( >that the font was corrupted some how?    I I guess.  The font file is there and it's still the same size, but I can    get a new copy and put it there.  ! Can you pull the font files from  F >an earlier backup where you know the product worked fine?  Also, try J >making a copy of the problem document, changing the font to one that you H >know works, and then try printing, because your document might also be $ >ill and therefore need restoring...  I Did that.  It printed.  That's how I diagnosed that the problem was with  	 the font.    Thanks,    -- Alan    ------------------------------    Date: 22 Mar 2005 21:32:22 +0100( From: Andreas Davour <ante@update.uu.se>% Subject: Re: [Announce] FreeVMS 0.1.3 4 Message-ID: <cs9u0n3o209.fsf@Psilocybe.Update.UU.SE>  = koehler@eisner.nospam.encompasserve.org (Bob Koehler) writes:   a > In article <cs9u0n62dv7.fsf@Psilocybe.Update.UU.SE>, Andreas Davour <ante@update.uu.se> writes:  > > E > > Sourceforge is a place to host your software project, not a tool.  > >  > 1 >    Yes, but doens't it assume CVS or some such?   D It's one of the available features, but not something you'll have to use.     H >    Do you think they could actually take VMS INSTAL'able save sets and >    not break them?  / Packaged as a ZIP or a tarball, yes I think so.    /andreas     --  A A: Because it fouls the order in which people normally read text. ' Q: Why is top-posting such a bad thing?  A: Top-posting. ; Q: What is the most annoying thing on usenet and in e-mail?    ------------------------------  + Date: Tue, 22 Mar 2005 16:46:21 +0000 (UTC) $ From: JKB <knatschke@koenigsberg.fr>% Subject: Re: [Announce] FreeVMS 0.1.3 ; Message-ID: <slrnd40iuu.23n.knatschke@rayleigh.systella.fr>    Le 22-03-2005,  propos de Re: [Announce] FreeVMS 0.1.3, (  Bob Koehler crivait dans comp.os.vms :a > In article <cs9u0n62dv7.fsf@Psilocybe.Update.UU.SE>, Andreas Davour <ante@update.uu.se> writes:  >>  D >> Sourceforge is a place to host your software project, not a tool. >>   > 1 >    Yes, but doens't it assume CVS or some such?   < 	Sure, but we have both ftp _and_ CVS... and a mailing list.  	 	Regards,    	JKB   --  J En plus c'est simple, je fais ce genre de trucs en g77 depuis des annes :J il suffit d'crire un wrapper en C. Et comme a, j'ai le meilleur des deuxN mondes : la rigueur quasi-monacale du Fortran, et l'exubrance ptulante du C.   ------------------------------   Date: 23 Mar 2005 02:40:03 GMT( From: bill@cs.uofs.edu (Bill Gunshannon)% Subject: Re: [Announce] FreeVMS 0.1.3 , Message-ID: <3ac383F680fcoU1@individual.net>  3 In article <d1$t7WRhtXBC@eisner.encompasserve.org>, 0 	Kilgallen@SpamCop.net (Larry Kilgallen) writes: > F > The fact that a language is versatile does not excuse the programmer > from taking care.   ( I've been saying that about C for years.   bill   --  J Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolvesD bill@cs.scranton.edu     |  and a sheep voting on what's for dinner. University of Scranton   |A Scranton, Pennsylvania   |         #include <std.disclaimer.h>       ------------------------------  + Date: Tue, 22 Mar 2005 20:24:07 +0000 (UTC) . From: klewis@LUMINA.MITRE.ORG (Keith A. Lewis)K Subject: Re: [Java JNI] How to load a shared img without defining a symbol? . Message-ID: <d1put7$qts$1@newslocal.mitre.org>   Bernhard <lithium@nospam_gmx.at> writes in article <Xns9620E840CDE42lithiumnospamgmxat@212.33.32.148> dated Mon, 21 Mar 2005 21:53:02 +0000 (UTC):
 >Hi again!L >After some investigation and experimenting I can say that loading a native - >lib/img directly from the JAR does not work. H >A workaround is to extract the native lib from the JAR on install time.G >One problem remains though. Under OpenVMS it seems I have to define a  L >symbolic name for the native lib -pointing to the absolute location on the K >disk -, so that OpenVMS finds it, when using the "System.load" Java call.  M >For a highly dynamic, plugin-oriented application like we are implementing,  L >it is unwanted to define a symbol for every installed JAR that uses native J >code, since this would require us to do a OS-dependent system call every  >time. > H >Is there another way to make OpenVMS find the correct native lib? Do I  >always need a symbol?K >I tried to set the java.library.path, but no avail. Under other OSs, this   >would be sufficient: H >System.load("foo") tries to find "foo.dll" on the java.library.path in 5 >Windoze, for Unix this call searches for "libfoo.so"   K Hmmm, sounds like this java.library.path thing is a feature that Sun should # document and HP should implement...   J For a workaround, I would guess that the default location is SYS$LIBRARY. F Try putting your shareable image (.EXE) containing the JNI routines inM SYS$COMMON:[SYSLIB].  Second guess would be SYS$SYSTEM (SYS$COMMON:[SYSEXE]).   A One thing -- there are security issues in allowing users to place G arbitrarily-named images in that directory (replacing a system-standard L image, for example, could cause the user's code to be executed in privileged! mode on the next system startup).   0 --Keith Lewis              klewis {at} mitre.org> The above may not (yet) represent the opinions of my employer.   ------------------------------   End of INFO-VAX 2005.163 ************************