Path: utzoo!utgpu!water!watmath!clyde!att!ucbvax!STONY-BROOK.SCRC.SYMBOLICS.COM!jrd From: jrd@STONY-BROOK.SCRC.SYMBOLICS.COM (John R. Dunning) Newsgroups: comp.sys.atari.8bit Subject: Relocatable Object Code Format Message-ID: <19880828160154.2.JRD@MOA.SCRC.Symbolics.COM> Date: 28 Aug 88 16:01:00 GMT Sender: daemon@ucbvax.BERKELEY.EDU Organization: The Internet Lines: 39 Date: Thu, 25 Aug 88 22:30:49 EDT From: jhs@mitre-bedford.ARPA Yes, I think it IS John Dunning who is hoping to "whomp up" a relocatable format and linking loader. Yup, that's me. I've heard of other relocatable object formats, but have never succeeded in getting hold of specs on any of them, and don't know of anyone else who claims to be working on them. In my opinion, it is PRECISELY BECAUSE the 6502 has very poor support for relocatability that a relocatable object code format and a linking, relocating loader ARE NEEDED. Precisely. If it was easy to get that functionality, nobody would be worried about it. With such facilities, code can be relocated to the absolute load addresses during the linkage process, all under the automatic control of the linker. Linking will require a little computation, and the relocatable format will be a little more "bulky" than the final, absolute load modules that result. However, once such machinery is in hand, it will become practical to build REAL languages for the 8-bitters. Yup, that's my thinking exactly. Heck, even a version of ACTION! with the capability of building relocatable object libraries would be a terrific advance! For that matter, relocatable assembly language libraries would be pretty neat too. Yup. In other mail I mentioned my plan to build a linker for the object format I'm working on. The obvious extension is to arrange it so that it can extract modules from libraries, and I've left room in my format definitions for that sort of thing. I'd be interested in reading a "product review" of ICD's format. Me too.