[HECnet] installing LPs on RSX

[Johnny Billquist]

On 2015-04-29 17:24, Mark Matlock wrote:
>
>> On Apr 28, 2015, at 7:11 PM, Johnny Billquist <bqt at softjar.se> wrote:
>>
>>> On 2015-04-29 02:37, Mark Matlock wrote:
>>>
>>>>> On Apr 28, 2015, at 11:55 AM, Johnny Billquist <bqt at softjar.se> wrote:
>>>>>
>>>>> On 2015-04-28 20:10, Mark Matlock wrote:
>>>>>
>>>>>
>>>>>>> On Apr 27, 2015, at 5:31 PM, Johnny Billquist <bqt at softjar.se> wrote:
>>>>>>>
>>>>>>> On 2015-04-28 02:15, Mark Matlock wrote:
>>>>>>>     BP2 with I/D space would be worth some effort to bring back from extinction. With the high speed of Simh and E11 with modern CPUs, if one could figure out a way to reassemble the tape blocks in the various combinations the see if BRU could read it successfully or not in some automated way, then just let the CPU have at it. It can't be worse than breaking the Enigma code in World War II.
>>>>>>
>>>>>> A few thousand blocks. The number of permutations can be interesting... Also, it is not totally trivial to see if you got it all right...
>>>>>
>>>>> The process to see if you got the right combination of blocks is the key to whether an automated process could be used. I guess you would have BRU try to read a virtual tape and produce a listing, but the listing could be correct, but still have some blocks out of order in some small file stored on the tape.
>>>>
>>>> The BP2 tape is not a BRU tape. Had it been, I already have a fix.
>>>
>>> It is really too bad that wasn't a BRU tape.
>>
>> Yeah... I sometimes wonder why BP2 is rather odd. It uses its own installation procedures instead of the standard tools that most other layered products use, and it also behaves rather different than any other tool (it don't exit with any status for example, so you cannot tell if a compilation went ok or not programatically).
>
> It really would be great to have the source code for the layered products to be able to fix issues like that.

*Yes*
Oh, how I wish I had some clue on where to find those...
So far, that search have been really futile. :-(

>>>> It's a plain DOS tape. DOS tapes are extremely stupid. Basically, anything goes. There is nothing telling you if the data comes out right or not. It's essentially implied just by being able to read the block.
>>>> (. Why would you ever get blocks in a different order than written? .)
>>>
>>> Indeed!   I wonder if the original tape that Tim Shoppa had still exists?
>>
>> Nope. Or at least Tim don't have it. I asked several years ago, and he had returned the tapes after dumping them, and this was close to 15 years ago...
>> I don't know if the guy he originally got the tapes from still have them. I'm not sure Tim even have any contact with the person, or even remember who it was.
>
> Bruce Mitchell had a few tapes that were a bit rare. I think he told me once that he had the SPD-11 (RSX system performance monitoring system) tape which I've never seen anywhere. I know I have asked him about FMS-11 V1.0 which has RSX KED. (Later versions of FMS didn't include it)

Bruce Mitchell seems to refuse to speak to me. Not sure why.
You're thinking of SPM-11 by the way. It would be an interesting tools 
to play with.

Yes, as for KED, I heard it would be on some FMS distribution, but all 
the ones I have lack it. Not that I've really been missing KED in the 
first place. If I had some more time, I'd focus on getting AMIS working 
instead. I have the sources, and it's a pretty decent EMACS clone. There 
is a version working under RSTS/E, but that version sources seems to be 
lost. Newer version needs some hacking, since it by default is just a 
bit too large.
(Written mostly in Pascal.)

>>>>>>>      Also, I meant to say that the PDP-11 C V1.2 is corrupt but the V1.1 is okay as I recall. Also, DECUS C works quite well, but I don't recall which DECUS tape it is on.
>>>>>>
>>>>>> I did some changes/improvements to DECUS C many years ago. I believe my changed version should be available with ftp from Update.
>>>>>
>>>>> I will need to look for your updated DECUS C on the Update ftp. Part of the challenge from the DECUS tapes is determining if you have the latest version of a particular software. Thanks for bringing your version of C to my attention.
>>>>
>>>> No worries. My fixes included making it all work correctly in I/D space, and then I did a half-hearted attempt at adding a void datatype.
>>> The I/D space fix for C would be great. Sometime ago I was trying to fix APL-11 to use I/D space. A guy had done this and put it on one of the DECUS tapes, but it was linked to an old FCSRES and when I tried to rebuild it, I got errors from TKB that I could not figure out. That was one of the things I was working on in my guest account on MIM that you graciously provided. Sometime I need to take another look at that. I think it is probably a simply fix, but I don't understand .odl files so it hasn't gotten fixed. In APL having I/D space is a big deal because it enlarges the available workspace by two fold.
>>
>> I can probably help with that, if you just can catch me when I have some time...
>
> Thanks, I'll capture exactly what the errors are and get it to you in a brief email so hopefully it won't take much of your time to determine what is needed.

Feel free. Always happy to help.

>>>>>> I should probably make a kit disk or something available from MIM/Madame as well. If I only had time...
>>>>>>
>>>>>> Too bad noone is willing to pay for me doing this kind of work. I could spend full time just fixing things for years... :-)
>>>>> That is too bad. There are a number of things in RSX that I'm sure you could improve. I really do appreciate what you've done on TCP/IP.   Thanks!
>>>>
>>>> Plenty of work to do... So little time. :-)
>>>> Anyway, it's great that people find TCP/IP useful...
>>>> I think the FTP code is good enough to be left for a while now. Next for proper telnet both in and out...
>>> I have a nice terminal emulator on my iMac that works well with Simh telnet to DZ ports but it would be great to telnet directly to the real PDP-11/83 system I have. The FTP file access has been fantastic to great .dsk files over to the PDP-11/83. I use the SCSI2SD device and a UC07 to have two pseudo RA92s that provide plenty of space and high reliability.
>>
>> By the way, you can also transfer TPC files. There is a slight trick to it, but it works.
>> You need the TPC tool from the DECUS library, and also a tool called ATT.
>> Once you have them, run ftp on the RSX system, and fetch TPC files in block mode (really, block mode, not binary mode). Once you have them on the RSX system, use ATT to change the file attributes. The first two words should be 1002, leave all the other header words alone.
>> After that you can use the DECUS TPC tool to read the file, and write it to a virtual tape.
>
> Thanks for that information! I have been planning to transfer some .tpc files to the PDP-11/83 and did not know about the changes needed on the file attributes. I could have wasted some time trying to figure this out.

You are welcome. I became curious about doing just this a couple of days 
ago, and located ATT. It's available now at MIM::DU:[ATT].

There are several versions of TPC around. Unfortunately, they all have 
small irritating issues. I'm almost tempted to write a new version 
myself... The code is pretty ugly.

	Johnny