search

PDP-10 / its

Incompatible Timesharing System
Other
868 stars 84 forks source link

Test portable software on other OSs #1131

Open atsampson opened 6 years ago

atsampson commented 6 years ago

Some of the programs in the ITS source tree are designed to work across multiple operating systems; for example:

We could test this as part of the CI, using prebuilt distributions (or running machines) of the other OSs.

larsbrinkhoff commented 6 years ago

Sorry, the "todo" label doesn't have a good name. I have used it to tag things that are to be added here that existed in ITS at one point. As opposed to new bug fixes, new features, or entire new programs written or imported from elsewhere.

So far no one has come up with a better name, but I'm open to ideas!

larsbrinkhoff commented 6 years ago

Added MacHack VI.

rmaldersoniii commented 6 years ago

SUDS works on TOPS-20, at least the Stanford-hacked variety of TOPS-20, which is what you get both in MRC's Panda distribution and on the XKL Toad systems, although Panda and XKL made incompatible extensions later.

larsbrinkhoff commented 6 years ago

@kcr is testing some programs on TOPS-20.

larsbrinkhoff commented 6 years ago

Snyder's C compiler works on TOPS-20, and maybe WAITS too.

larsbrinkhoff commented 1 year ago

@n7275 has now tested loading a MIDAS.REL file that I assembled on ITS with DECSW=1. After me fumbling the file transfer a couple of times, I think we finally managed to get the file onto TOPS-10 6.x. Unfortunately typing LOAD MIDAS.REL resulted in some kind of file format error. Loading it with the DEC LOADER.SHR under ITS' DECUUO works ok.

rmaldersoniii commented 1 year ago

Sorry for the delayed response. I read PDP-10 e-mail on my phone, which does not allow me to change my e-mail address to respond, so I have to remember that someone is expecting me to chime in and do it from RMail.

Date: Mon, 20 Nov 2023 22:49:53 -0800 From: Lars Brinkhoff @.***>

@n7275 has now tested loading a MIDAS.REL file that I assembled on ITS with DECSW=1. After me fumbling the file transfer a couple of times, I think we finally managed to get the file onto TOPS-10 6.x. Unfortunately typing LOAD MIDAS.REL resulted in some kind of file format error. Loading it with the DEC LOADER.SHR under ITS' DECUUO works ok.

LOADER.SHR is ancient, dating back to the PDP-6. LINK.SHR (later LINK.EXE) dates from 1973 (Tops-10 5.x?). Issuing the LOAD command on a Tops-10 later than 4S72 will invoke LINK, not LOADER.

The internal format of .REL files differs greatly between LOADER and LINK.

Your result is unsurprising.

I think, though I am prepared to be entirely wrong, that WAITS uses a verson of LOADER, but it's going to expect a FAIL .REL file rather than MIDAS or MACRO. More research needed.

                                                            Rich
rmaldersoniii commented 1 year ago

I see that the 1973 version of LOADER.MAC is available at Trailing Edge:

http://pdp-10.trailing-edge.com/ap-c796e-sb/01/loader.mac.html

n7275 commented 1 year ago

I tried building this version of loader with the version of macro I have on 6.03.

Building loader works fine, and I'm able to run it. I was even able to load in a rel that I compiled with fortran-10, and have loader spit out a symbol table

but when I tried to load any of the midas.rel files, I get:

` .COMPIL LOADER.MAC MACRO: LOADER

EXIT

.EX LINK: Loading [LNKXCT LOADER Execution] *MIDAS.REL

?ILL. FORMAT BLOCK TYPE 040000 PROG. MIDAR FILE MIDAS.REL * `

Which looks virtual identical to what I get with the 6.03 versions of load or link

larsbrinkhoff commented 1 year ago

I'm starting to think maybe it's a file transfer error after all. There are a few steps needed to get the MIDAS.REL out of ITS and then into TOPS-10, so there's a chance some step was done wrong.

Here's what the start of the file looks like on ITS.

000000:  000006000002
000001:  000000000000
000002:  000345253525
000003:  000014000000
000004:  000003000001
000005:  200000000000
000006:  422042400000
000007:  000001000002
000010:  040000000000
000011:  000000000041
000012:  264000000271
000013:  000001000002
000014:  040000000000
000015:  000000000125
000016:  000000415534
000017:  000001000022
000020:  252505252525
000021:  000000400000
000022:  100000402715
000023:  401000401200
000024:  401000401667
000025:  200000402411
000026:  400000401421
000027:  000000000000
000030:  200000402405
000031:  401000402003
000032:  200000402540
000033:  400000401367
000034:  200000402303
000035:  200000402252
000036:  100000402653
000037:  200000402255
000040:  401000401431
000041:  200000402367
000042:  401000404311
larsbrinkhoff commented 1 year ago

Another data point: I have updated MACRO-10 on ITS from some old version to "MACRO %53A(1152) 19-JUL-1979". This version outputs .REL files that works with the ancient LOADER on ITS.

I also have some version of LINK.SHR on ITS, and it's happy about loading MIDAS.REL.

n7275 commented 1 year ago

When I'm home later today, I will compare with the first few words I have on my end.

On Thu, Nov 23, 2023, 02:33 Lars Brinkhoff @.***> wrote:

I'm starting to think maybe it's a file transfer error after all. There are a few steps needed to get the MIDAS.REL out of ITS and then into TOPS-10, so there's a chance some step was done wrong.

Here's what the start of the file looks like on ITS.

000000: 000006000002 000001: 000000000000 000002: 000345253525 000003: 000014000000 000004: 000003000001 000005: 200000000000 000006: 422042400000 000007: 000001000002 000010: 040000000000 000011: 000000000041 000012: 264000000271 000013: 000001000002 000014: 040000000000 000015: 000000000125 000016: 000000415534 000017: 000001000022 000020: 252505252525 000021: 000000400000 000022: 100000402715 000023: 401000401200 000024: 401000401667 000025: 200000402411 000026: 400000401421 000027: 000000000000 000030: 200000402405 000031: 401000402003 000032: 200000402540 000033: 400000401367 000034: 200000402303 000035: 200000402252 000036: 100000402653 000037: 200000402255 000040: 401000401431 000041: 200000402367 000042: 401000404311

— Reply to this email directly, view it on GitHub https://github.com/PDP-10/its/issues/1131#issuecomment-1823919441, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAAYP3Z4X6RFZNM6XMPENY3YF333HAVCNFSM4FNQXBO2U5DIOJSWCZC7NNSXTN2JONZXKZKDN5WW2ZLOOQ5TCOBSGM4TCOJUGQYQ . You are receiving this because you were mentioned.Message ID: @.***>

n7275 commented 1 year ago

This is what I get:

0/ 000006,000002 000000,000001 000345,253524 3/ 000014,000001 000003,000000 200000,000000 6/ 422042,400000 000001,000002 040000,000001 11/ 000000,000041 264000,000270 000001,000002 14/ 040000,000001 000000,000124 000000,415534 17/ 000001,000023 252505,252524 000000,400001 22/ 100000,402714 401000,401201 401000,401667 25/ 200000,402411 400000,401420 000000,000001 30/ 200000,402405 401000,402002 200000,402541 33/ 400000,401367 200000,402302 200000,402253 36/ 100000,402653 200000,402255 401000,401431 41/ 200000,402367 401000,404310 000001,000022 44/ 252425,250025 000000,400021 400000,000202 47/ 200000,402460 401000,403645 400000,401366 52/ 000000,000001 401000,401662 200000,404621 55/ 200000,402416 400000,401365 401000,401546 60/ 400400,401564 000000,000000 000000,000000 63/ 000000,000000 400000,401427 400000,401422 66/ 263740,000056 000001,000022 202000,000000 71/ 000000,400042 263740,000076 263740,000076 74/ 254000,000166 263740,000076 263740,000076 77/ 263740,000076 263740,000076 263740,000076 102/ 263740,000041 263740,000076 263740,000076 105/ 263740,000100 263740,000074 263740,000076 110/ 263740,000076 263740,000076 263740,000076

When compared with yours: image

rcornwell commented 1 year ago

Should this not be:

COMPILE LOADER.MAC

START *MIDAS.REL