INFO-VAX Mon, 28 Jan 2008 Volume 2008 : Issue 55 Contents: Re: Anyone interested in building a vms-like OS? Re: DECTerm and xterm Re: DECTerm and xterm Re: DECTerm and xterm Re: DECTerm and xterm (was: Open sourcing of VMS: bad precedent set) Re: DSPP Integrity remanufactured hardware offerings, no VMS option HELP: VMS as an NFS client to Windows or Linux - primer please Re: HELP: VMS as an NFS client to Windows or Linux - primer please Re: M$IE; was: DSPP Integrity remanufactured h/w... Re: Marketing of operating sytems Re: Open sourcing of VMS: bad precedent set Re: Open sourcing of VMS: bad precedent set Re: Open sourcing of VMS: bad precedent set RE: Open sourcing of VMS: bad precedent set PowerTerm 525 & eXcursion Re: Problem: DECterm and X.Org version: 1.4.0 Re: Problem: DECterm and X.Org version: 1.4.0 Re: Using disk space on NAS? Re: Using disk space on NAS? Re: Using disk space on NAS? ---------------------------------------------------------------------- Date: 27 Jan 2008 22:56:01 GMT From: billg999@cs.uofs.edu (Bill Gunshannon) Subject: Re: Anyone interested in building a vms-like OS? Message-ID: <604gg1F1p2h40U1@mid.individual.net> In article <479cd1a3$0$90264$14726298@news.sunsite.dk>, Arne Vajhøj writes: > Roger Ivie wrote: >> On 2008-01-27, Arne Vajhøj wrote: >>> Bill Gunshannon wrote: >>>> 2. Because they actually use the 30-some years of computing research >>>> rather than re-inventing things (badly) because of NIH syndrome. >>> Hmm. What OS's were it that revived pcc to avoid using the existing >>> gcc ? >> >> BSD's gripe with gcc is slightly more complex than NIH. > > If it was not because of NIH, then I think the problems > could have been resolved in a different way (they may have > to anyway, because I am not optimistic about the success > of the pcc revival). After their having used GCC for so many years NIH syndrome is rather unlikely. bill -- Bill Gunshannon | de-moc-ra-cy (di mok' ra see) n. Three wolves bill@cs.scranton.edu | and a sheep voting on what's for dinner. University of Scranton | Scranton, Pennsylvania | #include ------------------------------ Date: Sun, 27 Jan 2008 14:20:34 -0500 From: "FredK" Subject: Re: DECTerm and xterm Message-ID: Ericom and Attachmate have perfectly adequate terminal emulators for PCs - if you exclude VMS keyboard support - which isn't really a function of the terminal emulator as much as a keyboard driver issue. DECterm is a descendent of the VWS emulator, which in turn was a descendent of the Pro300 VT200 emulator. It is a VT400+ emulator - meaning it was never certified as a VT500 IIRC. The UNIX varient dxterm is roughly the same original emulation base as DECterm IIRC. DECterm doesn't draw glyphs. It draws text using X11 functions to draw text using standard font sets. There are no "higher level functions to draw text wth attributes". X11 provides a graphics context for setting the foreground/background colors and different routines for text that sets the background as part of text output. Nor does X11 provide for other attributes such as blinking. xterm is mostly junk written by people with only a hazy idea of ANSI standards at best. But like PCterm - they didn't really care much about compliance with a standard. "JF Mezei" wrote in message news:47981d4c$0$16240$c3e8da3@news.astraweb.com... > John E. Malmberg wrote: > >> It might be more useful to add the stuff that is missing to xterm and >> port xterm to VMS. > > NO ! > For one thing, there is no longer any development (or very little) for > DECwindows on VMS. The message is quite clear from the onwer of vMS that > VMS isn't for interactive use. > > Secondly and more importantly, DECterminal would provide a lot of value > to the world community by providing a complete VT emulation, scolling > bars. > > Since DEC developped the VT standard, making DECterm available would > further increase Digital's long lasting legacy to the world and make > even the young ones know that it was Digital that developped it. > > >> Fred Kleinsorge has posted before that the DECterm code has some special >> stuff to make the display update better. Maybe he could give some >> pointers on how to do this with xterm. > > If I recall, it was about what X routines were used to draw the stuff. > They were using a lower level call to X to draw glyphs instead of the > higher level calls to draw chartacters with attributes. > > Also, wasn't DECterm available on Tru64 too ? If so, it would mean that > the code inside wouldn't be *too* tied to VMS. ------------------------------ Date: Mon, 28 Jan 2008 02:30:26 GMT From: "John E. Malmberg" Subject: Re: DECTerm and xterm Message-ID: <67bnj.4622$9j6.4294@attbi_s22> FredK wrote: > xterm already has been ported to OpenVMS as far as I remember. That does not surprise me. I may go looking for it. The PERL on VMS self test for UTF-8 handling sends out a sequence that locks up a DECTerm. I do recall you posting before something about special code in DECterm to support screen updates. I think it was for smooth scolling. > I do not recall any request from any customer for UTF-8 support on DECterm. > Sheesh. Most of the world uses crappy VT100+ emulators from Microsoft. Is > the ANSI standards stuff for terminals still active? Pretty much. It seems that the PUTTY emulator terminal program has become the most popular one for LINUX and Windows, having both support for SSH and being free. The big drawback for Putty so far for me is a lack of support to emulate the missing LK keys from a PC keyboard. But it is open source, so it is possible for anyone to customize it. The other popular program is called screen, which allows the controlling terminal to detach and reattach to a session. One big difference between it and disconnectable terminals is that with screen, the session keeps on running while disconnected while screen buffers the output. > Has UTF-8 replaced ISO Latin-1? In the UNIX/LINUX world, it appears so. Open Source programs ported from LINUX are now assuming that the terminals are ANSI and UTF8, and also that the filenames can be in UTF-8. -John wb8tyw@qsl.network Personal Opinion Only ------------------------------ Date: Sun, 27 Jan 2008 21:57:04 -0500 From: =?ISO-8859-1?Q?Arne_Vajh=F8j?= Subject: Re: DECTerm and xterm Message-ID: <479d4479$0$90276$14726298@news.sunsite.dk> John E. Malmberg wrote: > FredK wrote: >> Has UTF-8 replaced ISO Latin-1? > > In the UNIX/LINUX world, it appears so. > > Open Source programs ported from LINUX are now assuming that the > terminals are ANSI and UTF8, and also that the filenames can be in UTF-8. With .NET then Windows is also moving towards UTF-8. There are good drivers for it in the increasingly importance of the asian markets. Arne ------------------------------ Date: Sun, 27 Jan 2008 14:10:41 -0500 From: "FredK" Subject: Re: DECTerm and xterm (was: Open sourcing of VMS: bad precedent set) Message-ID: xterm already has been ported to OpenVMS as far as I remember. I do not recall any request from any customer for UTF-8 support on DECterm. Sheesh. Most of the world uses crappy VT100+ emulators from Microsoft. Is the ANSI standards stuff for terminals still active? Has UTF-8 replaced ISO Latin-1? "John E. Malmberg" wrote in message news:OAUlj.729$v.51@attbi_s22... > Thomas Dickey wrote: >> Arne Vajhøj wrote: >>> Thomas Dickey wrote: >>>> JF Mezei wrote: >>>>> DECterm, being a proper implementation of the VT standards with most >>>>> of >>>>> the VT bells and whistles (except tracing of control characters) would >>>>> be a great donation to the open source world, replacing Xterm which is >>>>> really really basic. >>>> too little, too late - DECterm doesn't support UTF-8 >> >>> If it were open sourced (which I do not find likely to happen, but >>> let us assume) then would it be hard to add ? >> >> It'd be a fair-sized chunk of work to do (a few months for someone >> working >> at it). > > It might be more useful to add the stuff that is missing to xterm and port > xterm to VMS. > > Fred Kleinsorge has posted before that the DECterm code has some special > stuff to make the display update better. Maybe he could give some > pointers on how to do this with xterm. > > -John > wb8tyw@qsl.network > Personal Opinion Only > ------------------------------ Date: Sun, 27 Jan 2008 16:19:56 -0800 (PST) From: Sue Subject: Re: DSPP Integrity remanufactured hardware offerings, no VMS option Message-ID: On Jan 24, 10:31=A0am, Rich Jordan wrote: > On Jan 23, 6:27 pm,Sue wrote: > > > On Jan 23, 2:31 pm, Rich Jordan wrote: > > > > DSPP is providing refurbished low end Integrity servers at a pretty > > > good price to DSPP members, along with a quantity break. =A0There are > > > three options: one HPUX, one Linux, and one Windows. =A0HPUX and Windo= ws > > > come with what appear to be licensed copies of their respective OS's. > > > > Its great to see these being provided at pretty nice prices, but it > > > would have been especially nice to have a VMS option. > > > I saw this post and am tracking it down. > > >Sue > > Sue, > =A0 =A0 =A0thanks. =A0I don't know if we would actually be in a position t= o > buy one (if it happened a few months from now I'd really press for it, > but right now we're too darn busy to even worry about it). =A0It would > sure be nice to have a second itanium as a test box since our 2620 is > in production and hard to change now. Thats ok would like to see it fixed anyway Sue ------------------------------ Date: Mon, 28 Jan 2008 15:53:30 +1100 From: Gremlin Subject: HELP: VMS as an NFS client to Windows or Linux - primer please Message-ID: <13pqnulhsbico24@corp.supernews.com> OK, so perhaps we can start and do a primer for VMS8.2 as a client to NFS on Windows2003 and/or Linux. Starting point: - I have an NFS server on Windows2003R2 and SuSE Linux and Solaris10 and Reh Hat EL5 and can connect any client OS to them using anonymous access rights - except VMS (which I can't even connect when I specify username/GID/UID). I feel it is safe to say the NFS servers are working OK on all the OSes. - I can ping the server names from VMS - the files in the NFS share area are full access to anyone, including anonymous - the directory is the same - tcpip sho proxy shows my VMS username with UID 0 GID 0 outbound to all the NFS servers - Windows mapping shows the "unix" username the same as my VMS username and it is mapped to a Windows domain administrator account - The Windows domain administrator account has full access to all files (I checked) and also the share and all the directories. - I can issue the following commands to any OS noted above and the command(s)complete OK: -tcpip mount DNSF0: data /host="servername" /path="NFSshareName" /system /struct=5 /noadf -tcpip mount DNSF0: data /host="servername" /path="NFSshareName" /struct=5 /noadf /gid=0 /uid=0 -tcpip mount DNSF0: data /host="servername" /path="NFSshareName" /struct=5 /noadf /gid=-2 /uid=-2 where DNSF0: can also be DNSF1: or any other integer to give it the same mount point each time where NFSshareName has been the NFS share as exported by the OS, or the actual path (/f/data/nfs/) - I cannot successfully do dir dnsf0:[000000] - I typically get device timeout....however, thanks to a reply on a similar post a couple of days ago, I added to /NOADF flag and now I get -RMS-E-PRV, insufficient privilege or file protection violation, AND now Windows is logging mapping errors - still works fine with Solaris, SuSE etc.... So, the /NOADF flag got rid of the timeout but exposed a mapping flaw(?). On the Windows side, anonymous access is allowed on the share and my VMS username is mapped to a Windows Domain Administrator account with full access to everything - so why the mapping error...... ------------------------------ Date: Mon, 28 Jan 2008 00:57:21 -0500 From: JF Mezei Subject: Re: HELP: VMS as an NFS client to Windows or Linux - primer please Message-ID: <479d6fd7$0$16197$c3e8da3@news.astraweb.com> I don't recall how I succeeded this yet. But this is what SHOW MOUNT/FULL shows: > $ tcpip show mount/full > _DNFS1:[000000] mounted > BRAKES.vixenation.ca:/ > Transport TCPIP-UDP Writing Enabled > Read/write size 8192/8192 Write conversion Enabled > RPC timeout 0 00:00:01.00 ADF usage USE,UPDATE,CREATE > RPC retry limit 4 Fileids Unique > Attribute time 0 00:00:15.00 Server type UNIX > Directory time 0 00:00:30.00 Advisory Locking Disabled > Cache Validation MODIFY TIME Default user [DECNET] > Superuser No Default UID,GID 501,501 501,501 is the uid/gid of my account on the mac. So by default, the mac lets the vms machine act as me in terms of file accesses. But I also have a proxy for both incoming and outgoing mapping my VMS username to 501,501, and mapping SYSTEM to 0,0 To get ODS5 capabilities, you need to specify the server type of UNIX, even if the doc says it is there by default. ------------------------------ Date: Mon, 28 Jan 2008 02:17:23 GMT From: VAXman- @SendSpamHere.ORG Subject: Re: M$IE; was: DSPP Integrity remanufactured h/w... Message-ID: In article , Keith Parris writes: > > >Bob Koehler wrote: >> Meanwhile my kids are running into trouble with their friends >> emailing them .docx, .xlsx, ... files from Office Vista. > >I haven't heard of an "Office Vista" version -- you are probably >receiving documents from Office 2007. There's a free compatability pack >from Microsoft to allow earlier versions of Office to read the new >document formats: >http://office.microsoft.com/en-us/products/HA101686761033.aspx > >Or you could install OpenOffice from http://OpenOffice.org/ Sending them as PDF would make them more *universally* "compatable". (sic) -- VAXman- A Bored Certified VMS Kernel Mode Hacker VAXman(at)TMESIS(dot)COM "Well my son, life is like a beanstalk, isn't it?" http://tmesis.com/drat.html ------------------------------ Date: Sun, 27 Jan 2008 16:29:20 -0800 (PST) From: Sue Subject: Re: Marketing of operating sytems Message-ID: <41d077d2-1780-4651-8061-754bd46e339f@v4g2000hsf.googlegroups.com> On Jan 21, 12:09=A0pm, Sue wrote: > On Jan 15, 3:26=A0am, JF Mezei wrote: > > > Apple is holding a little shingding in San Francisco this week. Steve > > Job's keynote address should be available for viewing later this > > afternoon athttp://www.apple.com. He is bound to make a great sales > > pitch for our favourite operating system (OS-X :-) > > > Wasn't there a time when DECUS events were larger than Macworld ? > > > Imagine if VMS had been allowed to be marketed, andSuewould have made > > similar keynote adresses in front of thousands =A0with worldwide media > > reporting on it ? > > I am flattered JF, to tell the truth I am in awe of SteveJobs. > > sue By the way the keynote is excellent and is avaible on the net for viewing. He has 4 main annoucments and it takes one hour 23 minutes. Sue ------------------------------ Date: Sun, 27 Jan 2008 14:26:06 -0500 From: "FredK" Subject: Re: Open sourcing of VMS: bad precedent set Message-ID: I see no point in "Open Sourcing" DECterm. The VT500 was essentially the pinnacle of terminals - with the advent of non-terminal GUIs (Windows/Mac/Browsers/etc) - what is the point? Sure DECterm/dxterm could be open sourced for Linux I suppose... but I don't see some huge demand or requests streaming in that would make it worth going through the hoops to create a Linux version and all the red tape it would involve to give up the intellectual property. "madcrow" wrote in message news:8d17a76e-af83-473e-b5a0-2e867914ebe6@f10g2000hsf.googlegroups.com... > On Jan 23, 7:12 am, Thomas Dickey wrote: >> JF Mezei wrote: >> > DECterm, being a proper implementation of the VT standards with most of >> > the VT bells and whistles (except tracing of control characters) would >> > be a great donation to the open source world, replacing Xterm which is >> > really really basic. >> >> too little, too late - DECterm doesn't support UTF-8 >> >> -- >> Thomas E. Dickeyhttp://invisible-island.netftp://invisible-island.net > > Given it's likely role of emulating terminals where full compliance > with real VT specs is needed, it would never NEED to do UTF-8. There's > not many "legacy systems" that use Unicode after all... ------------------------------ Date: Sun, 27 Jan 2008 15:57:05 -0800 (PST) From: Sue Subject: Re: Open sourcing of VMS: bad precedent set Message-ID: <6461952c-207e-4a9b-b94b-6d8ce337484c@f47g2000hsd.googlegroups.com> On Jan 27, 1:39=A0am, JF Mezei wrote: > Phillip Helbig---remove CLOTHES to reply wrote: > > > mentality---but they certainly use it. =A0DECterm and TPU are part of wh= at > > they are paying for. =A0People with support contracts can get bugs fixed= . > > MPE was also profitable, and people were paying support contracts. But > on Sept 7 2001, when Carly unveiled her wedding plans, MPE was put on > the chopping block. Same for Tru64. > > People at higher levels like Staller/Livermore/Hurd only look at > Powerpoint level of detail. If they see a dwonward curve in number of > customer, or a downward curve in profits, or a downward curve on "return > on investment", then they will decide that the product no longer has a > viable future and will act to get rid of it. > > They may not kill VMS like they did Alpha MPE or Tru64. But reducing > development staff, preventing marketing to the world and telling ISVs to > move off VMS is not a sign of HP really wanting those remaining profits > generated by VMS to continue forevere (and certaintly not grow). > > The valiant staff such asSueare doing their dardnest within the VMS > group to counter the negative direction set by upper management and > hoping to prove upper management wrong by showing good results despite > upper management's attempts otherwise. > > Remove a few key champions (such asSue) from the VMS group, and you > might see VMS go downhill very fast. > > In the end, those champions (I'll assumeSueisn't alone in this) are > fighting an uphill battle, swimming against the current. And in the end, > HP top magement are the ones who set the direction of that current. They > have decided that VMS has no strategic long term importance. They have > mentioned to the media many times already -- and it is now a consistant > message -- that they only care about the installed base, hoping to > capture a good portion of them whe =A0they migrate off VMS. > > The move by Cerner is an indication that HP wants to quicken the pace of > the slow winding down of VMS. > > > Initially, DECnet Phase IV and FMS were not going to be ported to > > Itanium. =A0Customer pressure changed that. > > When HP's instincts are to not port anything unless there is customer > pressure, it says a lot about HP's true desires. A company like Apple or > Microsoft will actively try to expand their market and actively develop > NEW applications. HP is more than happy to kill any VMS product unless > there is too much resistance from the remaining customers. > > On May 7th 2002, when they finally announced that VMS wouldn't be killed > (but also included the famous Stallard memo), they committed to > continuing the Compaq "Plan of Record". In other words: continued the > Curly policy of trying to kill VMS. Because in the end, that is what it > was all about. Many here know how close Culry came to killing VMS a year > before he killed Alpha. > > When a company has a well established policy of ignoring a product, > refusing the fact that a product could be far more profitable if > marketed, it means that the company is convinced that the product has no > long term future and it isn't worth doing anything to shore it up. > > Consider that many VMS staff here agreed with the opinion that > advertising VMS wouldn't yield good enough results. (despite the short > lived VMS renaissance that happened as a result of Marcello convincing > Curly to not kill VMS and give it a chance, that renaissance turned > negative growth into nearly double digit growth in about 6 months, proof > that advertising was at the time still quite effective for VMS). > > The fact that VMS isn't yet officially dead is only testament to the > amount of momentum it had back in the 1980s and despite all the brakes > put on it since then, it isn't dead yet. =A0Those brakes have absorbed > almost all the inertia now, and there really isn't much left. > > Just because Gartner's schedule for VMS' demise was wrong doesn't give > VMS eternal life. =A0It is a false sense of security to base the future of= > VMS on the fact that it has survived all those attacks in the last > decades. The first couple of bullets may not have instantly killed VMS, > but eventually, it will have lost enough blood to no longer be viable. > And I feel (unfortunatly) that VMS has now reached that point. > > With the disapearance of the medical business, it doesn't leave much. > > Consider stock exchanges: HP owns both Tandem and VMS. With OMX' future > in question, HP will be tempted to move VMS based exchanges to Tandem. > It does't make much sense for HP to continue to spend money on 2 > competing systems. It would be more profitable if the stock exchanges > were all on the same system. Same customers and half the development costs= . > > > Define "values". =A0I would say they value that which makes them profit,= > > and that includes VMS and parts of VMS even if they are relatively > > static. > > To HP, the remaining value of VMS is as a back end system. User > interface on VMS is not a value for HP. They keep on bragging about web > based system management. (aka: Windows explorer as the user interface on > Windows). > > So the various user interface applications on VMS could be donated as > open source without changing any value HP still sees in VMS. > > Lets face it, the only real asset VMS has left is the clustering. And HP > was quite clear that it wasn't worth porting it to anything since > Veritas was good enough. Dear Folks, You know how much I care for you and for VMS. We just had and Ambassadors meeting where Martin Fink called in. And while I value the fact that at most every posting on COV the future of VMS is discussed. We do have a roadmap which includes hardware and software futures. in the last 6 months we have had more press than ever before. Based on your request we had Martin Fink on Customer call, more technical web casts than ever before. We are moving forward and at the point we are just another one of the OS's in HP. We want to be judged on our merit. Reality. We are a business, we are not free, our engineers are (in my opinion) the best in the industry hence deserve a pay check. What do you want? Do not flame me. Do not give me un realistic statments like advertising, its not going to happen. Give me something, me Sue can do. And I will do my best to do it. Tell me where the problem is and I will try and find someone to fix it. I know your job depends on VMS so does mine. ------------------------------ Date: Sun, 27 Jan 2008 21:19:20 -0500 From: JF Mezei Subject: Re: Open sourcing of VMS: bad precedent set Message-ID: <479d3c29$0$15759$c3e8da3@news.astraweb.com> Sue wrote: > You know how much I care for you and for VMS. [font size=5000] Yes [/font] > We do have a roadmap which includes hardware and software > futures. So did Alpha on 24-JUN-2001. Road maps change. Changes can be subtle, progressive or radical (as in the case of Alpha). The attitude of the vendor towards a product is far more indicative of the vendor's true commitment to that product. > Based on your request we had Martin Fink on Customer call, I take it that from HP management's point of view, C.O.V. is the only place where there are complaints about how HP is mishandling VMS ? I take it that people like Stallard and Livermore tell Hurd to dismiss complaints sent to him as complaints from weirdos ? I take it upper managemenent are told that real VMS customers are perfectly happy with the way HP is handling VMS ? > Do not flame me. Our hopes would be that you would act as a router to send the flames to the right people within HP. None of the flames are aimed at you. I think people genuinely love you and really appreciate your devotion to VMS. > Do not give me un realistic statments like > advertising, its not going to happen. When you say "unrealistic", do you mean that it is not within your powers to get advertising done, or that it is unrealistic to expect HP to change their mind and start advertising VMS ? ------------------------------ Date: Sun, 27 Jan 2008 20:49:10 -0600 From: "Forster, Michael" Subject: RE: Open sourcing of VMS: bad precedent set Message-ID: <1abf01c86158$59de3d91$37146a8d@mcwcorp.net> Interesting from watching both of your viewpoints and also mine. It is a = carefull dance both sides play and are allowed to comment on. I too am = dependent on VMS and also a database and language from 1969. Why would the average CIO be savy or aware of VMS abilities and = strength? -----Original Message----- From: "JF Mezei" To: "Info-VAX@Mvb.Saic.Com" Sent: 01/27/08 8:25 PM Subject: Re: Open sourcing of VMS: bad precedent set Sue wrote: > You know how much I care for you and for VMS. [font size=3D5000] Yes [/font] > We do have a roadmap which includes hardware and software > futures.=20 So did Alpha on 24-JUN-2001. Road maps change. Changes can be subtle, progressive or radical (as in the case of Alpha). The attitude of the vendor towards a product is far more indicative of the vendor's true commitment to that product. > Based on your request we had Martin Fink on Customer call, I take it that from HP management's point of view, C.O.V. is the only place where there are complaints about how HP is mishandling VMS ? I take it that people like Stallard and Livermore tell Hurd to dismiss complaints sent to him as complaints from weirdos ? I take it upper managemenent are told that real VMS customers are perfectly happy with the way HP is handling VMS ? > Do not flame me.=20 Our hopes would be that you would act as a router to send the flames to the right people within HP. None of the flames are aimed at you. I think people genuinely love you and really appreciate your devotion to VMS. > Do not give me un realistic statments like > advertising, its not going to happen. When you say "unrealistic", do you mean that it is not within your powers to get advertising done, or that it is unrealistic to expect HP to change their mind and start advertising VMS ? ------------------------------ Date: Sun, 27 Jan 2008 20:47:43 +0100 From: Marc Van Dyck Subject: PowerTerm 525 & eXcursion Message-ID: Is there a Vista-compatible version of those two softwares available somewhere ? What is the vista compatibility status of the versions distributed on the OpenVMS consolidated distribution CDs ? Thanks in advance, -- Marc Van Dyck ------------------------------ Date: Sun, 27 Jan 2008 14:06:00 -0500 From: "FredK" Subject: Re: Problem: DECterm and X.Org version: 1.4.0 Message-ID: Mr FredK isn't doesn't know everything. Suprise. Random pauses over a network connection? Shock. I'll also note that a SunRay isn't even an X11 device. DECterm is nothing more or less than a Xlib/Motif application. I'd have to look at the code to see if their is anything special it does when it gets a focus event - but I can't think of anything offhand. The paths involved in X11 are long even when local. Networking just makes it worse. A mouse click on a remote system is handled by its device driver, sent up to its X11 server, packaged into an X11 event packet and sent to a network socket. After TCPIP delivers this correctly, it ends up in a network driver on OpenVMS. In the mean time, the local server would have assigned focus if needed and also sent a focus event to its socket. The network driver on VMS would hand the packet up the stack until it gets to the DECwindows transport layer which delivers the packets to the application - which then does something appropriate with them. So latency becomes an issue. Now if the window manager is also running on OpenVMS - then it also has it's own role to play. "JF Mezei" wrote in message news:479ca47d$0$4323$c3e8da3@news.astraweb.com... > Mark Daniel wrote: > >>> Have you tried to do a SHOW PROC/CONT on both the DECW$TE_xxxx >>> controller process as well as the _FTA4: (the DCL process) while this is >>> happening ? >> >> Yes. Seems quiescent. > > One would need Mr FredK in here to describe what could happen. > > When an application on VMS displays on a remote X terminal, does the > process on VMS take care of all of the X protocol from the application > code all the way to the final QIO that sends a packet out the ethernet ? > > In the case of DECterm however, there are 2 processes involved. The > DECterm controller process and the actual DCL process. Between the > controller and DCL process, I believe it is the pseudo terminal > interface being used (the FTA devices). This should not be affected by > the typoe of X terminal being used. > > But the controller process might be waiting for something to happen, > hence the freeze. Do you have multiple DECterms opened ? (if I remember > correctly, a decterm can support 4 DCL sessions). > > Would the controller (the DECW_TExxx) be in LEF state or HIB during the > stall ? > >> I was running the xev on VMS for the XP based display server and on >> FreeBSD for the X.Org based server. It shouldn't matter considering xev >> is interacting with the X display server not the client system. > > If XEV is running on VMS, it is using the older VMS X routines to send > and receive to the X terminal. So yes it matters in that it may or may > not know about certain events. > > On the other hand, XEV on my VMS (8.3 alpha) was able to show me the > editres event, even though applications on VMS don't seem to respond to > those packets. > > >> that xev would be showing the essentials of what was going on >> under-the-hood. A pause is odd. An ACCVIO or some other bugcheck >> perhaps. An indeterminate pause? Hmmm. > > You need a magic incantation to get Mr FredK to come out of the woodwork > and give you guidance on this. Perhaps there are magic logical names > that would make the debugging easier. > > BTW, do the pauses happen only when you change focus to the decterm ? Or > do they happen also once you are in focus and type away various commands ? > > Another thing you may try: > Type an extremely log file on the decterm. Leave to another window, and > return to the decterm. See if the scrolling pauses for those seconds, or > if the scrolling of text continues even if the display is unresponsive. ------------------------------ Date: Mon, 28 Jan 2008 07:55:16 +1030 From: Mark Daniel Subject: Re: Problem: DECterm and X.Org version: 1.4.0 Message-ID: <13pptodqptccpfb@corp.supernews.com> FredK wrote: > Mr FredK isn't doesn't know everything. Suprise. > > Random pauses over a network connection? Shock. I'm a little disappointed to be dismissed that casually. Particularly with the general care I take in making posts to this forum. This in addition to the amount of effort spent investigating and reporting the issue. Of course I could easily be mistaken on all points. > I'll also note that a > SunRay isn't even an X11 device. Fair enough. I've been caught thinking like this before. Of course you *can* use an X application on a system separate to the Sun system being used to host the SunRay (just to be sure, to be sure, i.e. [system+X-application]->X11->[Sun-host+SunRay-server]->ALP->[SunRay-client]). And users do. And across internal LANs. And users have used DECterms in this manner. Without this issue. I did describe it as a 'thin client'. We don't use X-terminals any longer. And certainly don't host sessions on VMS anymore due to the XDM issue I posted on some time ago http://groups.google.com/group/comp.os.vms/browse_thread/thread/2bb10af59be7be61/8c110014b9bcc58c http://groups.google.com/group/comp.os.vms/msg/8c110014b9bcc58c > DECterm is nothing more or less than a Xlib/Motif application. I'd have to > look at the code to see if their is anything special it does when it gets a > focus event - but I can't think of anything offhand. That's no problem. This is a community forum and few expect HP staff to be active unless it is in their own (and/or perhaps VMS') best interest. I was hoping for someone posting, "yup, I've seen that before - happens when you ...", and if I was *really* lucky, "... you can workaround it by ...". > The paths involved in X11 are long even when local. Networking just makes > it worse. A mouse click on a remote system is handled by its device driver, > sent up to its X11 server, packaged into an X11 event packet and sent to a > network socket. After TCPIP delivers this correctly, it ends up in a > network driver on OpenVMS. In the mean time, the local server would have > assigned focus if needed and also sent a focus event to its socket. The > network driver on VMS would hand the packet up the stack until it gets to > the DECwindows transport layer which delivers the packets to the > application - which then does something appropriate with them. So latency > becomes an issue. Now if the window manager is also running on OpenVMS - > then it also has it's own role to play. I understand the basics of the X Windows System and the X11 protocol. Although I was remiss in not explicitly introducing my description with "this does not appear to be a network issue because", I did make some effort to explain it was confined to the DECterm application. That no other common DECW applications exhibit the same behaviour. But just in case it was not obvious here is a quotation from the original post: >> Under X.Org 1.4.0 running on FreeBSD 6.3-RELEASE using KDE 3.5.8 as the >> desktop, a DECterm running on VMS V8.3 (and on V7.3, so it's not a >> recent VMS phenomenon) experiences short (one to a few seconds) pauses >> during focus events. These are inconsistent; not occuring every time >> the DECterm is focused or used but they do happen a *lot* and makes >> DECterm unusable (a pity). Other DECWindows applications >> (EDIT/TPU/INT=DECW, Mail, Clock, Mozilla) do not experience the same >> issue. This seems confined to the DECterm and its terminal component >> (pull-down menus seem to behave normally). It does not happen in >> EDIT/TPU/INT=DECW which *looks* like it is based on a similar GUI >> code-base. It is not a KWin issue as it behaves the same way under twm. You would imagine if it was an environment issue then it would be seen with all DECW applications, especially something text-oriented such as EDIT/TPU/INT=DECW or Mail composition. Likewise, I did make some (obviously not enough) effort to eliminate host and network differences by specifying the data came from a single VMS (Alpha) and a single dual-boot Athlon system: >> Two sets of data are provided. 1) is an X Window System application >> running under Win-XP. It does not exhibit the problem. 2) is the >> problematic Xorg system under FreeBSD. Both these environments are run >> on the same (dual-boot) Athlon-64 box. I did not explicitly specify this test environment is a network of three systems on a single 100Mbps Belkin switch. A PWS500 with VMS V8.3, the XP/FreeBSD Athlon-based PC, and a notebook. The DECterm pauses are definitely *not* network related except perhaps in an unimaginably bizarre fashion. Occam's Razor would suggest it is the X.Org 1.4.0 display server. However, as the DECterm is the only application to exhibit this behaviour of several DECW tried on the VMS system it would also suggest some peculiarity of DECterm in particular rather than of DECW in general. Thanks for your input. > "JF Mezei" wrote in message > news:479ca47d$0$4323$c3e8da3@news.astraweb.com... > >>Mark Daniel wrote: >> >> >>>>Have you tried to do a SHOW PROC/CONT on both the DECW$TE_xxxx >>>>controller process as well as the _FTA4: (the DCL process) while this is >>>>happening ? >>> >>>Yes. Seems quiescent. >> >>One would need Mr FredK in here to describe what could happen. >> >>When an application on VMS displays on a remote X terminal, does the >>process on VMS take care of all of the X protocol from the application >>code all the way to the final QIO that sends a packet out the ethernet ? >> >>In the case of DECterm however, there are 2 processes involved. The >>DECterm controller process and the actual DCL process. Between the >>controller and DCL process, I believe it is the pseudo terminal >>interface being used (the FTA devices). This should not be affected by >>the typoe of X terminal being used. >> >>But the controller process might be waiting for something to happen, >>hence the freeze. Do you have multiple DECterms opened ? (if I remember >>correctly, a decterm can support 4 DCL sessions). >> >>Would the controller (the DECW_TExxx) be in LEF state or HIB during the >>stall ? >> >> >>>I was running the xev on VMS for the XP based display server and on >>>FreeBSD for the X.Org based server. It shouldn't matter considering xev >>>is interacting with the X display server not the client system. >> >>If XEV is running on VMS, it is using the older VMS X routines to send >>and receive to the X terminal. So yes it matters in that it may or may >>not know about certain events. >> >>On the other hand, XEV on my VMS (8.3 alpha) was able to show me the >>editres event, even though applications on VMS don't seem to respond to >>those packets. >> >> >> >>>that xev would be showing the essentials of what was going on >>>under-the-hood. A pause is odd. An ACCVIO or some other bugcheck >>>perhaps. An indeterminate pause? Hmmm. >> >>You need a magic incantation to get Mr FredK to come out of the woodwork >>and give you guidance on this. Perhaps there are magic logical names >>that would make the debugging easier. >> >>BTW, do the pauses happen only when you change focus to the decterm ? Or >>do they happen also once you are in focus and type away various commands ? >> >>Another thing you may try: >>Type an extremely log file on the decterm. Leave to another window, and >>return to the decterm. See if the scrolling pauses for those seconds, or >>if the scrolling of text continues even if the display is unresponsive. -- Blindness to suffering is an inherent consequence of natural selection. Nature is neither kind nor cruel but indifferent. [Richard Dawkins; The Devil's Chaplain] ------------------------------ Date: 27 Jan 2008 20:35:17 GMT From: healyzh@aracnet.com Subject: Re: Using disk space on NAS? Message-ID: Jur van der Burg <"lddriver at digiater dot nl"> wrote: > If you can reproduce this with LD V9.0 then please let me know the details, > preferably with a reproducer. NFS connected container files have had limited > testing, and there could still be a problem. The Server in this test is running OpenBSD. $ ld version %LD-I-VERSION, LD version V9.0, module X-9 built on May 6 2007 14:55:47 -LD-I-DRIVERVERSION, Driver version: 6-MAY-2007 14:56:17.84 -LD-I-SYSINFO, Node: MONK, Hardware: COMPAQ Professional Workstation XP1000, VMS version: V8.3 $ ld create dnfs2:[test]user2.dsk/size=10000000 $ show dev ld Device Device Error Volume Free Trans Mnt Name Status Count Label Blocks Count Cnt $1$LDA0: (MONK) Online 0 $1$LDA1: (MONK) Mounted 0 CDROM 884257 1 1 $1$LDA2: (MONK) Online 0 $ ld connect dnfs2:[test]user2.dsk lda3: $ show dev ld Device Device Error Volume Free Trans Mnt Name Status Count Label Blocks Count Cnt $1$LDA0: (MONK) Online 0 $1$LDA1: (MONK) Mounted 0 CDROM 884257 1 1 $1$LDA2: (MONK) Online 0 $1$LDA3: (MONK) Online 0 $ init lda3: user2 $ mount/noassist lda3: user2/system %MOUNT-I-MOUNTED, USER2 mounted on _$1$LDA3: (MONK) $ show dev ld Device Device Error Volume Free Trans Mnt Name Status Count Label Blocks Count Cnt $1$LDA0: (MONK) Online 0 $1$LDA1: (MONK) Mounted 0 CDROM 884257 1 1 $1$LDA2: (MONK) Online 0 $1$LDA3: (MONK) Mounted 0 USER2 1611232 1 1 $ $ LD CREATE DNFS2:[TEST]TEST.DSK/SIZE=4000000 $ LD CONNECT DNFS2:[TEST]TEST.DSK LDA4: $ INIT LDA4: TEST $ MOUNT/NOASSIST LDA4: TEST /SYSTEM %MOUNT-I-MOUNTED, TEST mounted on _$1$LDA4: (MONK) $ SHOW DEV LDA4: Device Device Error Volume Free Trans Mnt Name Status Count Label Blocks Count Cnt $1$LDA4: (MONK) Mounted 0 TEST 3999792 1 1 $ TCPIP SHOW VERSION HP TCP/IP Services for OpenVMS Alpha Version V5.6 on a COMPAQ Professional Workstation XP1000 running OpenVMS V8.3 $ LD CREATE DNFS2:[TEST]TEST2.DSK/SIZE=8000000 $ LD CONNECT DNFS2:[TEST]TEST2.DSK LDA5: $ INIT LDA5: TEST2 $ MOUNT/NOASSIST LDA5: TEST2 /SYSTEM %MOUNT-I-MOUNTED, TEST2 mounted on _$1$LDA5: (MONK) $ SHOW DEV LDA5: Device Device Error Volume Free Trans Mnt Name Status Count Label Blocks Count Cnt $1$LDA5: (MONK) Mounted 0 TEST2 7999696 1 1 -- Next I deleted LDA4: and LDA5: to free up diskspace on the NFS server. $ LD CREATE DNFS2:[TEST]TEST3.DSK/SIZE=12000000 $ LD CONNECT DNFS2:[TEST]TEST3.DSK LDA4: $ INIT LDA4: TEST3 $ MOUNT/NOASSIST LDA4: TEST3 /SYSTEM %MOUNT-I-MOUNTED, TEST3 mounted on _$1$LDA4: (MONK) $ SHOW DEV LDA4: Device Device Error Volume Free Trans Mn t Name Status Count Label Blocks Count Cn t $1$LDA4: (MONK) Mounted 0 TEST3 3611184 1 1 $ It looks like I'm limited to 4Gb, additionally in spite of some info in "Ask the Wizard" it appears the NFS client only supports NFS V2 (as per the TCPIP V5.6 SPD). There is one advantage to a 4Gb limit, namely that the image will fit on a DVD-R blank for backups. The problem is that I've not been on 4GB drives for years, so I'd have to split up disks to get down to that size. So I need to deal with that, or investigate living on an NFS share (not sure how practical that is). Or I can simply wait for OpenVMS 8.4 and see if the client finally supporting NFS V3 or iSCSI make it into the release for the Alpha platform. Of course getting this working in some manner would give me needed justification for building a full fledged home NAS system. Here is my other problem. The following shows the state that LDA1: is in after I accidentally ripped the NFS share out from under the Logical Disk that had been attached to it. $ show dev ld Device Device Error Volume Free Trans Mnt Name Status Count Label Blocks Count Cnt $1$LDA0: (MONK) Online 0 $1$LDA1: (MONK) Mounted 0 CDROM 884257 1 1 $1$LDA2: (MONK) Online 0 $1$LDA3: (MONK) Mounted 0 USER2 1611232 1 1 $ LD CREATE DNFS2:[TEST]CDROM2.DSK/SIZE=921600 $ LD CONNECT dnfs2:[test]CDROM2.DSK LDA2: %LD-F-DUPUNIT, Duplicate unitnumber If you need additional data, please let me know. Thanks, Zane ------------------------------ Date: Sun, 27 Jan 2008 21:39:00 +0100 From: Jur van der Burg <"lddriver at digiater dot nl"> Subject: Re: Using disk space on NAS? Message-ID: <479cebe4$0$85794$e4fe514c@news.xs4all.nl> > $1$LDA2: (MONK) Online 0 > $1$LDA3: (MONK) Mounted 0 USER2 1611232 1 > 1 > $ LD CREATE DNFS2:[TEST]CDROM2.DSK/SIZE=921600 > $ LD CONNECT dnfs2:[test]CDROM2.DSK LDA2: > %LD-F-DUPUNIT, Duplicate unitnumber LDA2 is still connected to a file. which one? Do a LD SHOW/ALL to get the info. Jur. healyzh@aracnet.com wrote: > Jur van der Burg <"lddriver at digiater dot nl"> wrote: >> If you can reproduce this with LD V9.0 then please let me know the details, >> preferably with a reproducer. NFS connected container files have had limited >> testing, and there could still be a problem. > > The Server in this test is running OpenBSD. > > $ ld version > %LD-I-VERSION, LD version V9.0, module X-9 built on May 6 2007 14:55:47 > -LD-I-DRIVERVERSION, Driver version: 6-MAY-2007 14:56:17.84 > -LD-I-SYSINFO, Node: MONK, Hardware: COMPAQ Professional Workstation XP1000, > VMS > version: V8.3 > $ ld create dnfs2:[test]user2.dsk/size=10000000 > $ show dev ld > > Device Device Error Volume Free Trans > Mnt > Name Status Count Label Blocks Count > Cnt > $1$LDA0: (MONK) Online 0 > $1$LDA1: (MONK) Mounted 0 CDROM 884257 1 > 1 > $1$LDA2: (MONK) Online 0 > $ ld connect dnfs2:[test]user2.dsk lda3: > $ show dev ld > > Device Device Error Volume Free Trans > Mnt > Name Status Count Label Blocks Count > Cnt > $1$LDA0: (MONK) Online 0 > $1$LDA1: (MONK) Mounted 0 CDROM 884257 1 > 1 > $1$LDA2: (MONK) Online 0 > $1$LDA3: (MONK) Online 0 > $ init lda3: user2 > $ mount/noassist lda3: user2/system > %MOUNT-I-MOUNTED, USER2 mounted on _$1$LDA3: (MONK) > $ show dev ld > > Device Device Error Volume Free Trans > Mnt > Name Status Count Label Blocks Count > Cnt > $1$LDA0: (MONK) Online 0 > $1$LDA1: (MONK) Mounted 0 CDROM 884257 1 > 1 > $1$LDA2: (MONK) Online 0 > $1$LDA3: (MONK) Mounted 0 USER2 1611232 1 > 1 > $ > > $ LD CREATE DNFS2:[TEST]TEST.DSK/SIZE=4000000 > $ LD CONNECT DNFS2:[TEST]TEST.DSK LDA4: > $ INIT LDA4: TEST > $ MOUNT/NOASSIST LDA4: TEST /SYSTEM > %MOUNT-I-MOUNTED, TEST mounted on _$1$LDA4: (MONK) > $ SHOW DEV LDA4: > > Device Device Error Volume Free Trans > Mnt > Name Status Count Label Blocks Count > Cnt > $1$LDA4: (MONK) Mounted 0 TEST 3999792 1 > 1 > $ TCPIP SHOW VERSION > > HP TCP/IP Services for OpenVMS Alpha Version V5.6 > on a COMPAQ Professional Workstation XP1000 running OpenVMS V8.3 > > $ LD CREATE DNFS2:[TEST]TEST2.DSK/SIZE=8000000 > $ LD CONNECT DNFS2:[TEST]TEST2.DSK LDA5: > $ INIT LDA5: TEST2 > $ MOUNT/NOASSIST LDA5: TEST2 /SYSTEM > %MOUNT-I-MOUNTED, TEST2 mounted on _$1$LDA5: (MONK) > $ SHOW DEV LDA5: > > Device Device Error Volume Free Trans > Mnt > Name Status Count Label Blocks Count > Cnt > $1$LDA5: (MONK) Mounted 0 TEST2 7999696 1 > 1 > > > -- Next I deleted LDA4: and LDA5: to free up diskspace on the NFS server. > > $ LD CREATE DNFS2:[TEST]TEST3.DSK/SIZE=12000000 > $ LD CONNECT DNFS2:[TEST]TEST3.DSK LDA4: > $ INIT LDA4: TEST3 > $ MOUNT/NOASSIST LDA4: TEST3 /SYSTEM > %MOUNT-I-MOUNTED, TEST3 mounted on _$1$LDA4: (MONK) > $ SHOW DEV LDA4: > > Device Device Error Volume Free Trans > Mn > t > Name Status Count Label Blocks Count > Cn > t > $1$LDA4: (MONK) Mounted 0 TEST3 3611184 1 > 1 > $ > > It looks like I'm limited to 4Gb, additionally in spite of some info in "Ask > the Wizard" it appears the NFS client only supports NFS V2 (as per the TCPIP > V5.6 SPD). > > There is one advantage to a 4Gb limit, namely that the image will fit on a > DVD-R blank for backups. The problem is that I've not been on 4GB drives > for years, so I'd have to split up disks to get down to that size. So I > need to deal with that, or investigate living on an NFS share (not sure how > practical that is). Or I can simply wait for OpenVMS 8.4 and see if the > client finally supporting NFS V3 or iSCSI make it into the release for the > Alpha platform. > > Of course getting this working in some manner would give me needed > justification for building a full fledged home NAS system. > > > Here is my other problem. The following shows the state that LDA1: is in > after I accidentally ripped the NFS share out from under the Logical Disk > that had been attached to it. > > $ show dev ld > > Device Device Error Volume Free Trans > Mnt > Name Status Count Label Blocks Count > Cnt > $1$LDA0: (MONK) Online 0 > $1$LDA1: (MONK) Mounted 0 CDROM 884257 1 > 1 > $1$LDA2: (MONK) Online 0 > $1$LDA3: (MONK) Mounted 0 USER2 1611232 1 > 1 > $ LD CREATE DNFS2:[TEST]CDROM2.DSK/SIZE=921600 > $ LD CONNECT dnfs2:[test]CDROM2.DSK LDA2: > %LD-F-DUPUNIT, Duplicate unitnumber > > If you need additional data, please let me know. > > Thanks, > Zane > ------------------------------ Date: 27 Jan 2008 21:48:44 GMT From: healyzh@aracnet.com Subject: Re: Using disk space on NAS? Message-ID: Jur van der Burg <"lddriver at digiater dot nl"> wrote: > > $1$LDA2: (MONK) Online 0 > > $1$LDA3: (MONK) Mounted 0 USER2 1611232 1 > > 1 > > $ LD CREATE DNFS2:[TEST]CDROM2.DSK/SIZE=921600 > > $ LD CONNECT dnfs2:[test]CDROM2.DSK LDA2: > > %LD-F-DUPUNIT, Duplicate unitnumber > LDA2 is still connected to a file. which one? Do a LD SHOW/ALL to get the info. $ LD SHOW/ALL %LD-I-CONNECTED, Connected _$1$LDA1: to DNFS2:[TEST]CDROM.DSK;1 %LD-F-DETECTEDERR, Detected fatal error -SYSTEM-F-DEVNOTMOUNT, device is not mounted This is because I accidentally unmounted the NFS drive. Even remounting the NFS drive won't let the file be accessed. Is there a way to kill LD and restart it? I looked but wasn't able to find anything on that. Zane ------------------------------ End of INFO-VAX 2008.055 ************************