From: Graham Toal To: edinburgh-computer-history@yahoogroups.com Date: Fri, 25 Oct 2002 14:17:07 -0500 Subject: [ed-comp-hist] Roderick's EMAS2900 tape Reply-To: edinburgh-computer-history@yahoogroups.com JGH mailed me the contents of the EMAS2900 tape today to put on the archive: There were two tapesets: one was version 1.15 of the subsystem source, which I've put here: http://www.gtoal.com/athome/edinburgh/emas/emas2/subsystem/total0115s.imp (the later version 3.02 of the same source is also in that directory) The other is the EMAS2900 boot tape - it would appear to be a Ram image ready to load into some address and then jump to an appropriate starting point. The raw tape is here: http://www.gtoal.com/athome/edinburgh/emas/emas2/boottape/boottape.bin and a 'dumpfile' ascii listing (12Mb) is here: http://www.gtoal.com/athome/edinburgh/emas/emas2/boottape/boottape.asc The data is the same repeated EMPTY block until offset 000cc000. I don't suppose any of our members who are familiar with the 29xx architecture, and EMAS, have the time or inclination to start a 29XX emulation project? (One name that springs to mind who has both the skill *and* the documentation is Bill Laing ;-) ) We have all the pieces we'd need - binary boot tape and all the sources. Writing emulators is one of the more enjoyable programming projects you can do (I know, I've done two already) but they need to be done by someone who already knows the architecture well and that ain't me this time. I scoured the net and cannot find any existing 29XX emulators, which is a real shame because if one existed we'd pretty much be ready to run. We'd be in a better position to emulate the Amdahl EMAS because there already exists an emulator for that, *but* we don't have as full a set of sources, nor do we have the 370 Imp80 compiler. So there would be quite a lot more work involved in getting to the point where we could actually use the existing emulator. (find imp80, cross-compile it, generate EMAS binaries, build 370 boot tape - maybe also need to build an emas filing system on disk and populate it with the cross-compiled binaries). Not twisting any arms, just throwing it out as an idea if anyone is interested. Meanwhile, here's a note from Gordon on the tape contents. Graham ----------------------------------------------------------------------- The analysis of the tape was: Tape density = 1600 bpi. 1 block of 264 bytes Tape mark 124 blocks of 4096 bytes Tape mark Tape mark 648 blocks of 4096 bytes Tape mark Tape mark The first block was a tape label block containing the label "EMAS". (This would normally be 80 bytes, and was padded to 264 by 16_81's) The two files on the tape are attached to this message. Although both were fixed 4096 byte blocks on the tape, they will just be byte streams in the attached .TAR file. I don't have Roderick or Bob's e-mail addresses at work so you may want to forward this to them. My question to Roderick is - is this what you expected from the tape. Again the tape read without any errors. Best Regards, Gordon Hughes