Keyboard Shortcuts
Likes
- H390-Vm
- Messages
Search
This group is for all folks running the original IBM VM/370 Release 6 operating system (or later (e.g. VMTCE (Community Edition)) on Hercules. Like the other early IBM operating systems this version has always been in the public domain and so can be freely distributed. The base version as supplied by IBM is lacking in many facilities. IBM solved this by providing additional extension products which were licensed and so are not available. There are however many user enhancements available which can be installed. In addition, in order to get users up and running quickly updated "releases" of VM/370 included the most popular updates are available for download, so novices can start to learn VM without having to delve into the system internals. It is intended that this wiki will provide information on the base release and these updates.
The available versions are here :-
?
?
Re: Contributing to, Building, and Maintaining VM370CE
#VMCE
¿ªÔÆÌåÓý32.000 elements = 32.000 bits = 4k size (of course) ... sorry
about that
Am 27.08.2024 um 17:01 schrieb Bernd
Oppolzer via groups.io:
|
Re: Contributing to, Building, and Maintaining VM370CE
#VMCE
¿ªÔÆÌåÓýAm 27.08.2024 um 12:43 schrieb Dave
Wade:
In any case: the compiler can be downloaded as an AWS tape from the compiler website and installed by VM370CE users individually; the current version there is 2023.01 but I am working on an improvment, which will be version 2024.08, probably. (there is this problem with sets involving negative integers, see the Moshix video ... I am working on that; I need a completely new implementation of Pascal sets, which is AMODE 31 ready, BTW ... and it will support sets of up to 32.000 elements)Does that still work on a 370? Yes, I hope :-) in fact, I am doing the development on the Windows variant first, because this means creating and defining new P-Code instructions etc., then changeing the compiler so that it creates the new instructions and then implementing them in the P-Code interpreter, and then make the compiler run again, when it compiles itself using the new instructions. After that, the P-Code to 370 translator must be changed so that the new P-Code instructions create correct 370 instructions (or library routines, when needed). That said, there is a sort of problem with the maximum size of automatic storage for a procedure, which is 8k at the moment. A set with 32.000 elements needs 8k bytes (32.000 bits), so there is not much room for other variables in a procedure. The current sets were limited to 2000 bits, which is 256 bytes. That said: if you really use such large sets, you will run into trouble probably (when you're on the mainframe) ... but that's subject to testing :-) The real size of a set depends on how you define it. A set of CHAR, for example, has 36 bytes. A set of 0 .. 50 (where the elements can be all numbers from 0 to 50) needs 11 bytes (will be rounded to 12). Anyway: the old solution was plain wrong, and it was limited to AMODE 24 by design, which is not a good idea. Cheers, Bernd |
Re: Contributing to, Building, and Maintaining VM370CE
#VMCE
From: [email protected] <[email protected]> On Behalf Of Bernd Oppolzer via groups.io This may be a little topic drift, but ...OK That said: is there a person or group that I can contact when the new version is ready for deploying?I don't think so at present. Just this list In any case: the compiler can be downloaded as an AWS tape from the compiler website andDoes that still work on a 370? To this question: "I guess, first of all, is there really any serious interest in continuing the "Community Edition" idea?"OK noted Kind regards, have a nice dayDave Am 26.08.2024 um 20:59 schrieb William Denton via groups.io: I recently posted on asking these questions about how I should submit my VM updates for inclusion in a future CE "release": 1. Is there a master list of assigned update numbers? How do I figure out the "###" part of the update name"HRC###DK"???? 2. Is there a preferred way to package the files in an update? On my system, I keep all the files together in a .txt file in CMS READCARD format. How do the keepers of the CE releases prefer to receive proposed updates? 3. What sort of consensus/approval/voting process do I need to go through to share my stuff? Dave Wade responded that, since there really aren't any current processes, it might be a good idea to take the discussion to the wider group. So, how do we go about creating some sort of guidance/process for enhancing and maintaining future releases of VM370CE? I guess, first of all, is there really any serious interest in continuing the "Community Edition" idea? If so, then we probably could use at least a loose sort of "product management" strategy. Thoughts?? cheers, William Denton |
Re: Workshop Tapes - was Re: Examine RDR Files
toggle quoted message
Show quoted text
-----Original Message-----I believe it works for any tape i/o. I haven't looked at the code either, but it must put extra information on the tape because it correctly handles short blocks... So yeah, I think the chances are good that I'll be able to add support for thatOf course And as always, I can't say how long it'll take me to make these changes either.Very cool Fish. Whenever you have a moment. --Dave |
Re: Workshop Tapes - was Re: Examine RDR Files
Dave Wade wrote:
Fish wrote:[...] Great! I'll see what I can do!Does that sound like something that would make Hercules'sYes [...] Yes, I realize not being able to maintain position betweenIndeed! [...] One last wish. Would it be possible to support tapes writtenI can't say. I would need to know about the BLOKTAP format first. We have quite a few tapes written using this....Hmmm... You mean "BLOCKTAP"?! (Doh!) Silly me! You wrote "BLOKTAP" in one of your earlier emails, which I had never heard of before. I also did a "find" on all of my emails from this group for "bloktap" and "blocktap" and apparently there have been many emails containing both words which I'm presuming were all referring to the same thing: Steven Howes's BLOCKTAP command: * Yes? Correct? If so, then my answer changes to "probably". I haven't read through the entire thing yet (nor bothered to study the BLOCKTAP command's actual source code yet), but if all it is, is the same thing as the existing TAPE format but using 32K block sizes instead, then I have to say it certainly sounds simple enough. So yeah, I think the chances are good that I'll be able to add support for that tape format. But let me make those other changes first. They take priority. Once they're working properly, then I'll try to add support for BLOCKTAP format. And as always, I can't say how long it'll take me to make these changes either. My main focus right now is with helping to get 4.8 out the door. It takes priority right now. But when I find myself with a bit of free time here and there I'll certainly work on this change for you. Cool? -- "Fish" (David B. Trout) Software Development Laboratories mail: fish@... |
Re: Workshop Tapes - was Re: Examine RDR Files
Tony Harminc wrote:
Fish wrote:Because the Hercules vmfplc2.exe utility is not the same thing as the Hercules emulator. It's just a simple command-line program. Hercules must emulate a real tape device, and real tape devices can be positioned and will remain positioned until unloaded. Host command line commands can also position themselves to any point within a file, but once the utility exits, the next time you issue that same command line again, it has no idea what the previous command did. It's just a one-time-shot file processing command. For example: on Linux, I believe there is a tool/utility (i.e. command line command) called "head", yes? Entering the command "head -n 10 myfile.txt" will display the first 10 lines of the file. But entering the same command again, does not display the next 10 lines. It will display the exact same first 10 lines exactly like before. It has no idea what the previous command did. It has no idea where the previous command left off at. The same is true with Hercules's vmfplc2 utility. It doesn't run under the Hercules emulator itself. It runs under the host (Windows or Linux, etc). Each time you enter the command, it starts "fresh". Who is rewinding/"resetting" to loadpoint (but presumably notYour host operating system (Windows or Linux) in combination with the vmfplc2 executable itself. Each time it is run, it begins fresh. Each invocation knows nothing about any previous invocation. I guess in my experience it's very common on VM to use TAPE/VMFPLC2Well... VM's VMFPLC2 command is not interactive. It, like Hercules's vmfplc2 command, is not written to be interactive. It, just like Hercules, does one thing each time and that's it. But one of the things it is able to do is issue commands to tape devices, which are then processed by the Hercules emulator, which then positions the emulated tape device as requested and REMEMBERS (for as long as Hercules remains up and running!) where that tape device is positioned at. So the next time the VMFPLC2 is issued, its "input file" (tape device) is still position where the previous command left off at. If the file that Hercules's vmfplc2 command was reading was actually a real physical tape device, then yeah, it would behave pretty much identically to VM's, because that's the way real tape devices behave. But Hercules's vmfplc2 command is not reading a real tape device. It's reading an ordinary binary Windows/Linux disk file. And Windows/Linux doses not remember where each command "left off at"! It's not an emulator! Is that clear now? Does all that make any sense? -- "Fish" (David B. Trout) Software Development Laboratories mail: fish@... |
Re: Contributing to, Building, and Maintaining VM370CE
#VMCE
¿ªÔÆÌåÓýThis may be a little topic drift, but ... I am the contributor of the New Stanford Pascal compiler, which
is part of the VM370CE. Last time (when the compiler was first included) I had some
conversations with Bob Bolch (IIRC) about That said: is there a person or group that I can contact when the
new version is ready for deploying? In any case: the compiler can be downloaded as an AWS tape from
the compiler website and To this question: "I guess, first of all, is there really any
serious interest in continuing the "Community Edition" idea?" Kind regards, have a nice day Bernd
Am 26.08.2024 um 20:59 schrieb William
Denton via groups.io:
|
Re: OT: SDFVM access
¿ªÔÆÌåÓýDrew, It appears to be down tonight. (well its midnight here). Do you have any input? Dave ? From: [email protected] <[email protected]> On Behalf Of Drew Derbyshire via groups.io
Sent: Sunday, August 4, 2024 4:05 AM To: [email protected] Subject: [h390-vm] OT: SDFVM access ? The VM/SP image now running at SDF has moved to planet.sdf.org port 24. ? Also, for getting permanent accounts on SDFVM: ?
I had no input to this policy.? (I found out with the rest of SDF world). ? -ahd- |
Re: Contributing to, Building, and Maintaining VM370CE
#VMCE
¿ªÔÆÌåÓýTony, ? Pretty sure it is not rocket science to produce an EXEC to take a user mode and convert it to an OFFICIAL mod¡ ? Dave ? ? ? ? From: [email protected] <[email protected]> On Behalf Of Tony Harminc via groups.io
Sent: 26 August 2024 20:24 To: [email protected] Subject: Re: [h390-vm] Contributing to, Building, and Maintaining VM370CE #VMCE ? On Mon, 26 Aug 2024 at 15:18, Dave Wade via <dave.g4ugm=[email protected]> wrote:
? The only problem I see with that is that it's traditional to flag new or changed source lines with the fix ID, and if that's not known until someone else assigns it... ? Tony H. |
Re: Contributing to, Building, and Maintaining VM370CE
#VMCE
On Mon, Aug 26, 2024 at 14:59 William Denton via <williamedenton=[email protected]> wrote:
I've been opening GitHub Issues at either ?or? for everything I discover, and then attaching fixes when I figure out how to fix them. My VM updates are all RAPxxxcc, 'cause my initials are RAP ?. ?"cc" is the component code - DK, DS, DM, or DT.? And I use AUXLCL, 'cause there's already DMSLCL CNTRL that specifies that.? So, for example, see.? And I've been assuming that whomever collects things into a release will renumber them as HRCyyycc, where xxx and yyy are likely to be different.? I have a few small EXECs that make renumbering updates easier.
I've been keeping mine on the LCL disk, with the expectation that I'll delete them when the corresponding HRC updates come out in the next release.? And, of course, I've been opening Issues, which at least one person has pulled a fix from.
I ought to leave that as a question for Dave et al.? But as a very old friend of Bob Bolch, I'll suggest that his technique would have been to accept every bugfix, if written will enough, and to ?discuss feature contributions on this list.? Some of you may recall the discussions about BREXX and CMS.
Yes, at least on my part. Ross |
Re: Workshop Tapes - was Re: Examine RDR Files
toggle quoted message
Show quoted text
-----Original Message-----A scan of the waterloo tape I have produces nearly 4000 lines of output, so yes, loess might be more.... You'd rather have the ability to limit the SCAN function toWell of course I am greedy so a range would be good, so something like -files n<-m> Well yes... How about this: make the <ctlfile> an optional argument for the SCANSounds good. Then to easily LOAD only the file(s) you want, you could simply edit that fileYes, I mean for the Waterloo tape it may not be necessary to delete but yes it would be good... Does that sound like something that would make Hercules's vmfplc2 utility aYes Yes, I realize not being able to maintain position between commands creates a design challenge... Hopefully my proposed change will make creating/using <ctlfile>'s a little bitI think that¡¯s a good compromise. Have an auto generated control is a big plus. One last wish. Would it be possible to support tapes written using BLOCKTAP. We have quite a few tapes written using this.... .. the source is on here... --Dave |
Re: Workshop Tapes - was Re: Examine RDR Files
On Mon, 26 Aug 2024 at 15:05, Fish Fish via <david.b.trout=[email protected]> wrote: [...] Just as an explanation, Herc's vmfplc2 will never be as easy to use as VM's, due to the need for a <ctlfile> due to the difference in the processing hosts. VM's VMFPLC2 tool's filesystem is quite different from your Hercules host's, as well as its data format too (EBCDIC vs. ASCII) and the fact that on VM, it's able to physically position the physical input tape and it will stay there until the next command is issued, whereas on your Hercules host, each time a new command is issued, the input tape always starts out positioned at loadpoint each time (not to mention the need to be able to "translate" a CMS filename to/from a Linux/Windows host filename for the DUMP and LOAD functions). Um, why does a tape in these circumstances not stay where it was left? Who is rewinding/"resetting" to loadpoint
(but presumably not unloading) each time, and why? I guess in my experience it's very common on VM to use TAPE/VMFPLC2 to browse quite interactively, and the Hercules version seems to be more oriented to batch operation where you know in advance exactly what you want to do. Tony H. |
Re: Contributing to, Building, and Maintaining VM370CE
#VMCE
On Mon, 26 Aug 2024 at 15:18, Dave Wade via <dave.g4ugm=[email protected]> wrote:
The only problem I see with that is that it's traditional to flag new or changed source lines with the fix ID, and if that's not known until someone else assigns it... Tony H. |
Re: Contributing to, Building, and Maintaining VM370CE
#VMCE
¿ªÔÆÌåÓýHi Folks William, thanks for posting on here.. ?
? No thoughts on the rest. The original idea was for quick releases. That hasn¡¯t happened¡ ? Dave ? From: [email protected] <[email protected]> On Behalf Of William Denton via groups.io
Sent: 26 August 2024 19:59 To: [email protected] Subject: [h390-vm] Contributing to, Building, and Maintaining VM370CE #VMCE ? I recently posted on asking these questions about how I should submit my VM updates for inclusion in a future CE "release":
Dave Wade responded that, since there really aren't any current processes, it might be a good idea to take the discussion to the wider group. ? So, how do we go about creating some sort of guidance/process for enhancing and maintaining future releases of VM370CE? I guess, first of all, is there really any serious interest in continuing the "Community Edition" idea? If so, then we probably could use at least a loose sort of "product management" strategy. ? Thoughts?? ? cheers, William Denton |
Re: Workshop Tapes - was Re: Examine RDR Files
Dave Wade wrote:
[...] I actually find your VMFPLC2 a tad un-helpful on these tapes.Touch¨¦. :) So Hercules's SCAN always lists the contents of the *entire* tape, which you'd rather it not do? You'd rather have the ability to limit the SCAN function to only ONE specific physical (tapemark delimited) file? Yes? And you'd also rather not have to manually build the ctlfile in order to LOAD only the specific file(s) you want? Yes? How about this: make the <ctlfile> an optional argument for the SCAN function, which, when specified, specifies the name of the <ctlfile> that you want SCAN to automatically *create* for you instead? (i.e. when specified, the tape's contents would not be shown, but rather would simply be written to the specified output file instead. Then to easily LOAD only the file(s) you want, you could simply edit that file and either delete (or comment out) all of the file(s) you *didn't* want, and then specify that <ctlfile> in your LOAD command. Does that sound like something that would make Hercules's vmfplc2 utility a little easier to use? -------------------- Just as an explanation, Herc's vmfplc2 will never be as easy to use as VM's, due to the need for a <ctlfile> due to the difference in the processing hosts. VM's VMFPLC2 tool's filesystem is quite different from your Hercules host's, as well as its data format too (EBCDIC vs. ASCII) and the fact that on VM, it's able to physically position the physical input tape and it will stay there until the next command is issued, whereas on your Hercules host, each time a new command is issued, the input tape always starts out positioned at loadpoint each time (not to mention the need to be able to "translate" a CMS filename to/from a Linux/Windows host filename for the DUMP and LOAD functions). Thus the need for a <ctlfile>. Hopefully my proposed change will make creating/using <ctlfile>'s a little bit easier. I'd appreciate your thoughts, Dave. Thanks! -- "Fish" (David B. Trout) Software Development Laboratories mail: fish@... |
Contributing to, Building, and Maintaining VM370CE
#VMCE
I recently posted on asking these questions about how I should submit my VM updates for inclusion in a future CE "release":
Dave Wade responded that, since there really aren't any current processes, it might be a good idea to take the discussion to the wider group.
?
So, how do we go about creating some sort of guidance/process for enhancing and maintaining future releases of VM370CE? I guess, first of all, is there really any serious interest in continuing the "Community Edition" idea? If so, then we probably could use at least a loose sort of "product management" strategy.
?
Thoughts??
?
cheers,
William Denton |
Re: Workshop Tapes - was Re: Examine RDR Files
toggle quoted message
Show quoted text
-----Original Message-----Generally yes, but the workshop tapes all have "abstract listing" as the first file... I actually find your VMFPLC2 a tad un-helpful on these tapes. On VM you can do a "VMFPLC2 SCAN" which lists the contents of the first physical file, so up to the first tape mark. This only has the "ABSTRACT LISTING" so then a "VMFPLC2 REW" and a "VMFPLC2 LOAD" to get the contents. You can then read that, and use the FSF , SCAN, REW and LOAD to position the tape, check the contents of that physical file and load it.... .. on windows you do a scan and you get all the tape, about 300 files, then you need to build a control file to get the ones you want. I think it would be good to have options to skip so many tape marks and only list so many physical files.. .. but that¡¯s just me... Dave |
Re: Workshop Tapes - was Re: Examine RDR Files
Dave Wade wrote:
I think the old ones are in VMFPLC2 format, (might be TAPE),Unless you already know exactly what files are on the tape, I would not recommend doing that. The "LOAD" function reads the files from the tape and actually loads them onto your minidisk. Instead, I would do a "SCAN", which reads the tape and just lists the files that are on the tape but does not actually load them. Also, if you were not aware, Hercules comes with a vmfplc2 utility too (*), which, while completely different from VM's VMFPLC2 (or TAPE) commands, has similar functionality but is designed as a simple Linux or Windows host command-line command instead, that allows you to examine (SCAN) an existing .aws/.het VMFPLC2/TAPE format tape file, or create such a tape from host files (DUMP function), or LOAD files from such a tape onto your host system. Basically it provides the same basic functionality as VM's command, but is designed to be run your host as a Hercules offline utility. Documentation for how to use it is in our README.VMFPLC2.md file: * (You might consider installing some type of "markdown viewer" browser plugin/extension so our README files are properly formatted/displayed in your browser.) So I would personally do a Hercules vmfplc2.exe SCAN function, and then, is there is a file on the tape called "ABSTRACT LISTING", then I would run the utility again and LOAD that file onto your Windows/Linux host and examine (view) it there. Then you'll know exactly what file(s) you'll want to LOAD onto your VM/CE system. But then maybe that's just me. :) --------------------- (*) FYI: (shameless plug) My HercGUI product () also makes Hercules's VMFPLC2 utility much easier to use too. Some of the later ones are in BLOKTAP format.Is that different from TAPE format? If it is, then I don't think Hercules's vmfplc2.exe utility will be able to handle it. It only supports the newer VMFPLC2 format or the older TAPE format. I've never heard of BLOKTAP format before. Is BLOKTAP format documented anywhere? Maybe Hercules's utility can be updated to support that VM tape format too? HTH -- "Fish" (David B. Trout) Software Development Laboratories mail: fish@... |
Workshop Tapes - was Re: Examine RDR Files
Doug,
I think the old ones are in VMFPLC2 format, (might be TAPE) , mount the tape on VM, do a VMFPLC2 LOAD and there is an ABSTRACT LISTING file showing the content.
Some of the later ones are in BLOKTAP format. I think we now have a process to read those, I'll have to do some searching.
Dave? |