RDROYA01@ULKYVX.BITNET (Robert Royar) (09/22/86)
Re: an earlier posting about why Gnu *probably* wouldn't work on an ST. What I was really trying to say was GNU_EMACS needs the Lisp functions to remain such a wonderful package. I was responding to a question about when someone was planning to port GNU, without lisp, to the ST. I don't think that port would be worth it. It would be a step backwards for GNU. As for porting it to the ST. My 1040 only has about 896K left when it loads without DAs. Sounds like a lot doesn't it? Try loading a 600k application, up to 100K of source files, and then expect the ST to let you compile and link those modules without exiting. The linker can run in 128K, but with large compilations it needs 256K. Let's see 896 - 600 - 256 = 40. Given the problems with malloc and free on the ST, that 40K is gone in no time. BTW I have been using uEmacs 37 which includes a Cmode, a non-braindamaged word wrap and fill (i.e. it understands that a word includes punctuation marks when it wraps and that sentences are terminated with two spaces when it fills). This version also supports the store macro function which allows you to write routines that are a little more useful than the record keyboard macro can be but still no conditionals or loops. Now for the other side. Unipress has a Gosling EMACS that works on an IBM-PC (full memory I believe) and needs a hard drive. But like someone said, it doesn't support the full lisp that GNU has. And I use that lisp code and Info for my students who write essays with GNU_EMACS. Perhaps with some anti-C maneuvering (overlays on the hard drive), GNU_EMACS could be ported complete with "peaches en regalia" (F.Z.). *8^}
jmc@ptsfa.UUCP (Jerry Carlin) (09/23/86)
In article <8609221236.AA01198@ucbvax.Berkeley.EDU> RDROYA01@ULKYVX.BITNET (Robert Royar) writes: >...Try loading a 600k application, up to 100K of source files, and >then expect the ST to let you compile and link those modules without >exiting. Since Atari now has a FOUR meg machine (in Europe) what's the problem :-) Also, someone at the recent San Jose Atari Show had a prototype 2 meg upgrade for the 1040, so in a few months we'll all have the extra meg of memory that we could not think of a use for before and now can't live without! -- voice= 415 823-2441 uucp={ihnp4,dual,qantel}!ptsfa!jmc
braner@batcomputer.TN.CORNELL.EDU (braner) (09/24/86)
[] When I use Megamax C on a 1040ST, I set up a 576K RAM disk which nicely holds the compiler, linker, libraries, and a bunch of source and object files. I hate waiting for the disk, so I also have a communication program in the RAM disk. The RAM disk cannot be much smaller for serious use. In what's left (about 200K after GEMDOS and Micro-C-Shell take their tax) I can run the compiler OR the linker OR the editor, one at a time. Using microEMACS I can edit 2 or 3 source files simultaneously (about 20K each) before I run out of memory. Everything just fits and I DON'T WANT A BIGGER EDITOR! I want a smaller one! My version of microEMACS, which includes decent word wrap/fill and bracket-balance check, and RESIDENT help screens, is about 42K. Small for "real" editors, but a lot bigger than, say, the 4K RAM-resident screen editor I wrote for the Apple II - in machine language. The communication program I usually use on the ST is the PD "XMOTER(M)", because it has just the features I usually need, and is less than 5K long! - Moshe Braner
preece@ccvaxa.UUCP (09/25/86)
I'm curious how long it would take to load an 800K application into memory on the ST. -- scott preece gould/csd - urbana uucp: ihnp4!uiucdcs!ccvaxa!preece arpa: preece@gswd-vms