Path: utzoo!utgpu!jarvis.csri.toronto.edu!cs.utexas.edu!execu!sequoia!rpp386!aubrey From: aubrey@rpp386.cactus.org (Aubrey McIntosh) Newsgroups: comp.lang.modula2 Subject: Re: Re: DECUS Modula-2 ??? Message-ID: <17874@rpp386.cactus.org> Date: 7 Feb 90 01:03:51 GMT References: Reply-To: aubrey@rpp386.UUCP (Aubrey McIntosh) Organization: vima, Austin TX Lines: 38 In article Modula2 List writes: >Aubrey McIntosh asks: > >Sorry, didn't mean to start a war. The copy I have of the M-code source >contains neither an interpreter nor a linker. There was a version available >only in binary form (for PCs) with both, but it was terribly slow. > >Grant Guenther I thought this went as e-mail. Sigh. I have the "c18" source, and the "Provo" interpreter, which I believe to be the two products. I agree that they are slow under the interpreter. I just didn't want to spend much time only to end up in some obscure corner that you had already found out about. I have disassembled (de-compiled) the entire version of Medos-2 supplied with the Provo interpreter, including the linking-loader. I would like to correspond with folks who have licenses to either the C18 compiler or who have the provo interpreter. I presume that I was the last person alive to actually use the interpreter. I have run it through a spiffy disassembler (Sourcer) and compared it with the Modula-2 interpreter given in the c18 license. It is certain that they are close kin. I have a pencil/paper architecture for register use on the 8088 that should substantially increase the speed of the interpreter. There are some neat undocumented additions, such as execute native code procedure, do RS-232 in/out, and others I've forgotten. But then again, I have other M2 compilers now. -- Aubrey McIntosh comp.os.minix, comp.lang.modula2, soc.culture.celtic Austin, TX 78723 1-(512)-452-1540 (v)