INFO-VAX Thu, 08 Nov 2007 Volume 2007 : Issue 612 Contents: Re: Announcing: Jobs for the boys Re: Consolidated List of OpenVMS 30th Anniversary Articles on the Internet Inter Re: How do I write a SFTP SSH client - How to hook into TCPIP$SSH_SSH2.EXE Re: How do I write a SFTP SSH client - How to hook into TCPIP$SSH_SSH2.EXE Re: Informationweek mentions VMS 30 year anniversary Re: Informationweek mentions VMS 30 year anniversary Re: Informationweek mentions VMS 30 year anniversary Re: Informationweek mentions VMS 30 year anniversary Re: IO TIME vs COMputable Re: Minor Planet Center (was: Informationweek mentions VMS 30 year ) Re: MOUNT /ASSIST ---------------------------------------------------------------------- Date: Thu, 8 Nov 2007 20:40:46 +0800 From: "Richard Maher" Subject: Re: Announcing: Jobs for the boys Message-ID: Hi Jan-Erik, > Now, gSOAP doesn't do much att all, realy. : : : : : : > I would not call gSOAP a sledge-hammer, it's just > a couple of rather simple conversion tools! > Maybe there is a "client-only" version of gSOAP available, or maybe I'm looking at a different "gSOAP" in Wikipedia, but I personally find the following functionality to be excessive in a client API : - - The gSOAP Web services development toolkit offers an XML to C/C++ language binding to ease the development of SOAP/XML Web services in C and C/C++. - Includes stand-alone HTTP/1.1 and HTTPS secure Web Server. - gSOAP's memory management uses garbage collection so (deserialized) data can be cleaned up without a hassle. - Client and server (HTTP Web server and SOAP/XML engine included) - Offers Apache_mod, IIS, WinInet, CGI, and FastCGI interfaces. - Architecture features: integrated memory management with automatic leak detection in debug mode compiler-based XML serialization of native C and C++ data structures custom serializers and DOM support plug-ins for extensions (message logging, statistics, etc.) - The toolkit automatically serializes pointer-based data structure graphs, including cyclic graphs and pointers to derived class instances to support polymorphism. Is that really what you had in mind Jan-Erik? Polymorphic deserialization issues getting you down :-) Once again, to me this screams "WSIT is a pile of shit and we don't need your Java, JVM, or GC, 'cos we've got our own C/C++ baby, and it rocks!". Anyway, regardless of the merits of that argument, all I ask is that its proponents are cut loose from the sheltered-workshop that is HP/VMS these days, and are asked to fend for themselves. (Such artificial and counter-productive cross-subsidization is certainly illegal in the EU.) Bridgeworks was an abject failure by anyones's standards and WSIT is clearly going (gone!) the same way :-( Don't let them get away with it! Hold those bastards accountable! gSOAP *IS* being positioned as the WSIT replacement by the same talentless wankers that sold you Rally, COM, DECadmire, ACMSxp, ONC/RPC, DCE/RPC, BridgeWorks, Forte. Just say NO to ineptitude! Cheers Richard Maher "Jan-Erik Söderholm" wrote in message news:iBrYi.22$R_4.156@newsb.telia.net... > Richard Maher wrote: > > Hi Jan-Erik, > > Hi there... :-) > > > But personally given these requirements, yes, rather than take a gSOAP > > sledge-hammer to crack a nut (and in the absence of a supported solution) I > > would prefer to code a supportable solution from scratch. I doubt that we're > > talking rocket science here and all the SSL, XML-parsing/building and > > a.n.other infrastructure routines are all readily available. (Sadly "C" only > > so I'd have to get someone in :-) > > Now, gSOAP doesn't do much att all, realy. > > It takes the interface description (WSDL) as published > and produces header files and funtions stubs and whatever > to make writing those C, COBOL, Fortran apps easier. You get > your record descriptions and structures ready to use. > > I would not call gSOAP a sledge-hammer, it's just > a couple of rather simple conversion tools! > > Now, after that, it's just some libs to link against. > No Java, no ODS-5, none of the OS version limits of WSIT. > > Now I'm just lacking a tool that creates Rdb CREATE TABLE > statements from the WSDL... :-) > > Jan-Erik. ------------------------------ Date: Thu, 08 Nov 2007 09:37:40 GMT From: =?ISO-8859-1?Q?Jan-Erik_S=F6derholm?= Subject: Re: Consolidated List of OpenVMS 30th Anniversary Articles on the Internet Inter Message-ID: JF Mezei wrote: >> InternetNews.Com: OpenVMS at 30: Still Going Strong >> by Sean Michael Kerner >> http://www.internetnews.com/ent-news/article.php/3707631 > > This one is very good. Should be at the top of the list. > > One general comment though: it says that the next release of VMS is the > end of 2009. That is a bit far in my opinion unless we're talking about > V 9.0 with major changes. -OpenVMS V8.4, Alpha/Integrity, late 2008 or early 2009 -OpenVMS V8.n FRS: +18-24 months Platforms: Alpha and Integrity So V8.n(ext) is aprox late 2010 or early 2011. Where did you found that about V9.0 ??? (From presentations at the VMS Tech Upd.) Jan-Erik. ------------------------------ Date: Thu, 08 Nov 2007 02:17:14 -0800 From: Trefor Subject: Re: How do I write a SFTP SSH client - How to hook into TCPIP$SSH_SSH2.EXE Message-ID: <1194517034.666168.124530@i38g2000prf.googlegroups.com> On 7 Nov, 18:29, Jose Baars wrote: > I have been looking into more or less the same problem. > > Writing your own FTP client was doable, although harder than you would > think at first glance, as some FTP servers will hang for unknown > reasons, > and you do want to detect this. > > I needed a callable SSH filetransfer library, to be called from C or > COBOL, > and a sftp client that made it possible to login to a remote system in > batch > using sftp using password in stead of public key authentication, > something > the TCP/IP services sftp client will not let you do. > > I did not have any luck in finding a usable public domain SSH library > to > use as the basis for this callable SSH library, and I really didn't > want to > try to implement the SSH protocol myself, to be honest I find the > RFC's > rather hard to understand, and what I understood at times > contradictory. > Hopefully you have more luck, and probably you are smarter than I > am :-) > > I did find a couple of Java SSH libraries that allowed me to build a > SFTP > client that accepts both public key and password to be used as > authentication to a remote system. > > This is were I found these Java libraries: > > -http://sourceforge.net/projects/jsch/ > -http://www.ganymed.ethz.ch/ssh2/ > > Hope this helps, > Jose I was looking at the OpenVMS SFTP program. It spawns sys$system:tcpip $ssh_ssh2.exe which does the SSH bit which makes sense. It appears to call it with the following parameters. $mc sys$system:tcpip$ssh_ssh2.exe -v -x -a -o "passwordprompt %U@%H's password:" - -o "authenticationnotify yes" my_user@my_host.somewhere.co.uk -s sftp which makes sence. But I've not managed to work out how the SFTP program hooks into tcpip$ssh_ssh2.exe to send the SFTP protocol stuff. Anyone any ideas? Trefor ------------------------------ Date: Thu, 8 Nov 2007 09:12:08 -0500 From: "Richard Whalen" Subject: Re: How do I write a SFTP SSH client - How to hook into TCPIP$SSH_SSH2.EXE Message-ID: > $mc sys$system:tcpip$ssh_ssh2.exe -v -x -a -o "passwordprompt %U@%H's > password:" - > -o "authenticationnotify yes" my_user@my_host.somewhere.co.uk -s sftp Here's an explaination of what happens with the above: SFTP2 creates a subprocess and uses a couple of VMS mail boxes (C pipes) to communicate with it. -v is verbose mode - which is really debug level 2 -x allows X11 forwarding, which probably doesn't matter for SFTP2 -a allows agent forwarding, which can be used in some types of authentication -o is for options, in this case the password prompt which is specified as user@host's password: SSH2 has a default value for this of Password -o "authenticationnotify yes" says that SFTP2 wants to be notified about the result of authentication with the string AUTHENTICATED YES -s sftp requests the sftp subsystem to be run on the remote system after successful authentication. The SFTP2 client will write to one of the mailboxes, SSH2 will encrypt the data and transport it to the remote system where it will be decrypted and the sftp subsytem will read it from it's stdin as its command stream. The sftp subsystem writes its responses to its stdout, which is encrypted, passed back, decrypted and sent through the other mailbox to the SFTP2 client. In the SFTP protocol any data that pertains to the command is shipped with it. ------------------------------ Date: Thu, 8 Nov 2007 11:13:58 +0000 (UTC) From: david20@alpha2.mdx.ac.uk Subject: Re: Informationweek mentions VMS 30 year anniversary Message-ID: In article , =?ISO-8859-1?Q?Jan-Erik_S=F6derholm?= writes: >david20@alpha2.mdx.ac.uk wrote: > >> I wouldn't put too much faith in their accuracy. > >Which dosn't meen a thing. >It's the visability of VMS that counts for >those that we *want* to see the article. > I agree about VMS visibility but the comment above was in response to Bill Gunshannon's question about the accuracy of the articles figure of 300,000 systems. David Webb Security team leader CCSS Middlesex University >Jan-Erik. ------------------------------ Date: Thu, 8 Nov 2007 11:27:38 +0000 (UTC) From: david20@alpha2.mdx.ac.uk Subject: Re: Informationweek mentions VMS 30 year anniversary Message-ID: In article , JF Mezei writes: >Syltrem wrote: >> http://www.informationweek.com/news/showArticle.jhtml?articleID=202801794 >> > >Merci for the pointer. > >Better late than never. Interesting however how the article starts out >by going out of its way to paint VMS as an old operating system when >Unix is in fact older than VMS. > As far as I can see it doesn't explicitly mention Unix and in an article about the 30th anniversary it would be difficult to not mention it's age. In general I think it is a reasonable article - I wasn't aware that the Smithsonian Astrophysical Observatory Minor planet center was using a VAX/VMS cluster. David Webb Security team leader CCSS Middlesex University >As to the mention of 8.3 in another article, hasn't 8.3 been out for >quite some time already ? (I realise there there is to be some mini >IA64 release to support some new IA64 contraptions, but will this also >result in an Alpha release ? ) > >As for 8.3 on the vax, perhaps we could use the article to force HP to >produce 8.3 for VAX , threathening to sue them or false advertising if >they don't :-) :-)( :-) :-) :-) ------------------------------ Date: Thu, 08 Nov 2007 07:28:37 -0600 From: Ron Johnson Subject: Re: Informationweek mentions VMS 30 year anniversary Message-ID: On 11/08/07 05:27, david20@alpha2.mdx.ac.uk wrote: > In article , JF Mezei writes: >> Syltrem wrote: >>> http://www.informationweek.com/news/showArticle.jhtml?articleID=202801794 >>> >> Merci for the pointer. >> >> Better late than never. Interesting however how the article starts out >> by going out of its way to paint VMS as an old operating system when >> Unix is in fact older than VMS. >> > As far as I can see it doesn't explicitly mention Unix and in an article about > the 30th anniversary it would be difficult to not mention it's age. > > In general I think it is a reasonable article - I wasn't aware that the > Smithsonian Astrophysical Observatory Minor planet center was using a VAX/VMS > cluster. The article was contradictory. What the article's author wrote was "VAX", but at the bottom of the article, the scientist wrote (something on the order of) "we see no need to leave Alpha". -- Ron Johnson, Jr. Jefferson LA USA Give a man a fish, and he eats for a day. Hit him with a fish, and he goes away for good! ------------------------------ Date: Thu, 08 Nov 2007 09:03:44 -0500 From: Chip Coldwell Subject: Re: Informationweek mentions VMS 30 year anniversary Message-ID: david20@alpha2.mdx.ac.uk writes: > In general I think it is a reasonable article - I wasn't aware that the > Smithsonian Astrophysical Observatory Minor planet center was using a VAX/VMS > cluster. They were still using it when I worked there. The Center for Astrophysics (combination of the Harvard College Observatory, the Harvard Department of Astronomy, and the Smithsonian Astrophsical Observatory, all crammed into 60 Garden St in Cambridge, Mass http://www.cfa.harvard.edu/) was a solid VMS house when I started working there circa 1989. There was a "computation facility" (CF) that ran the facility-wide cluster. At some point, a decision was made to move everything to SunOS/SPARC, and that caused quite a bit of fragmentation. There were isolated VMS clusters running for years after the CF standard workstation became a Sun box. There was a group in theoretical astrophysics who insisted on Digital Unix because they liked doing numerics on Alpha. The minor planet center was one of the isolated VMS clusters that remained after the switch to Sun, a switch that was nearly completed just in time for Linux to come along and be the next sexy new thing. Nobody liked paying the price differential between SPARC workstations and PCs that could run Linux, and there was tremendous pressure on the CF to add support for Linux as I left to work elsewhere. Chip -- Charles M. "Chip" Coldwell "Turn on, log in, tune out" Somerville, Massachusetts, New England ------------------------------ Date: Thu, 08 Nov 2007 10:25:28 +0100 From: "e.e" Subject: Re: IO TIME vs COMputable Message-ID: Hein RMS van den Heuvel a écrit : > On Nov 7, 11:20 am, "e.e" wrote: >> Bob Koehler a écrit : >> >>> In article , "e.e" writes: >>>> i know that the IO time = ELapsed - CPU Time. > > Just FYI... You know you posted this question twice right? > The other topic received more replies > > http://groups.google.com/group/comp.os.vms/browse_frm/thread/d3bf97983c5842d4/ef85ca1421021c6b?hl=en#ef85ca1421021c6b > > Hein. > yes i saw.. a cancel of this post doesn't work.. everything cannot work fine like VMS.. ;-) Thanks. Eric ------------------------------ Date: 8 Nov 2007 06:53:43 -0600 From: Kilgallen@SpamCop.net (Larry Kilgallen) Subject: Re: Minor Planet Center (was: Informationweek mentions VMS 30 year ) Message-ID: In article , david20@alpha2.mdx.ac.uk writes: > In general I think it is a reasonable article - I wasn't aware that the > Smithsonian Astrophysical Observatory Minor planet center was using a > VAX/VMS cluster. The last I knew, that set of machines has only a few VAXen left and the bulk of machines in the cluster are all Alpha. Certainly in 1990 (the year cited for Gareth Williams starting) it was VAXen. By the way, local lore around Boston is that the Smithsonian (not necessarily the Minor Planet Center) was responsible for DEC adding H-Floating as an option for the VAX 11/780s. One of their people was featured in Time or Newsweek magazine a few years back in the case of an asteroid that did not hit Earth after all. Presumably if it _had_ hit earth, there would have been an even bigger story (or no story at all). I got a tour of this operation once, and for anyone who is concerned about the reputation of VMS in the face of false predictions, data at the Minor Planet Center comes from a worldwide set of corresponding astronomers who submit a sighting if they find something that has not been reported before. Certainly I can appreciate that just seeing something once in the sky makes it hard to predict its future path. They need to get other locations reporting the same thing at different times before figuring out whether it is time to call Bruce Willis. At the time of my tour, the vast body of code was written in Fortran. Those who know me have guessed that I did _not_ suggest switching to C or Java for greater buzzword compliance. ------------------------------ Date: Thu, 08 Nov 2007 05:14:19 -0800 From: AEF Subject: Re: MOUNT /ASSIST Message-ID: <1194527659.971749.207940@v23g2000prn.googlegroups.com> On Nov 8, 12:26 am, Jim Mehlhop wrote: > Robert Jarratt wrote: > > I had a MOUNT command with an implicit /ASSIST in my SYSTARTUP_VMS.COM. > > When I changed the disk around this MOUNT command failed and VMS sent > > messages to the console asking that this be fixed. However I was unable to > > log in to the machine to fix the issue, presumably because the login limit > > had not been opened up. What surprised me though was that I could not log in > > to the console either, and there was no way to do a clean shutdown. > > Fortunately I was able to put the old disk back and then change the MOUNT to > > specify /NOASSIST, so that the startup could complete. > > > Now this is a hobbyist system so the consequences are irrelevant, but I am > > curious as to how you would get out of this situation in a production > > environment without doing what I was obliged to do. Did I miss something > > (other than not specifying /NOASSIST?) > > > Thanks > > > Rob > > Not sure why everyone makes this so difficult. > > In a production environment you would do a conversational boot then at > sysboot> set startup_p1 "min" > sysboot> cont > > Then log in at the console and modify systartup_vms.com > > then run sysgen reset startup_p1 "" and reboot Why are *you* making it more difficult! "cont"? C'mon. A simple C will do! :-) No problem. AEF ------------------------------ End of INFO-VAX 2007.612 ************************