holloway@drivax.UUCP (Bruce Holloway) (01/30/88)
In article <1988Jan24.012514.26587@gpu.utcs.toronto.edu> pete@gpu.utcs.UUCP (Peter Santangeli) writes: > I believe what Jim has come up against here is the "soon to be as >famous as the 40 folder limit" bug in the file creation system. > It seems that the programmer who wrote the GEMDOS routine to find >the first free allocation block on a device was as clueless as the guy who >designed the directory system. > This brings up an interesting idea. Gemdos is certainly a usable >system, but... > We have a limited number of folders (static). > My hard drive creates new files slower than my floppy. > The OS doesn't do any sector level buffering. >These features remind me painfully of my experiences with TRSDOS on a >1977 trs-80! A long, long time ago, a company named Atari wanted a windowing system for its hot new computer. They came to Digital Research, since unlike some other companies, we had experience designing operating systems for 680x0 based computers. And we had Crystal (which became GEM). And we already had an 'okay' 68000 operating system, CP/M-68K, and we were (then) working on a multitasking, multiuser version of the same called "CDOS/68K". But GEM ran only on MS-DOS machines, and sorta depended on that file structure. Hmmm. So we considered changing GEM... But then a systems programmer said that _he_ could write a new operating system in a month that would look just like MS-DOS from a functional standpoint, and would be great and perfect and peachy-keen. And he did - pretty much alone. And since Atari was in such a hurry, they took what we had before all the horrid problems in it cropped up. And later we fixed all those problems and then forgot about it. I don't think there's anyone here still working on GEMDOS. *** NOTE: THESE COMMENTS, ET.AL., ARE IN NO WAY SANCTIONED, CONDONED, OR REPRESENTATIVE OF DIGITAL RESEARCH, INC. THEY MAY EVEN BE WRONG. I JUST WANTED TO EXPLAIN WHY AND HOW DRI COULD RELEASE SUCH A SUBSTANDARD O/S. *** - Bruce -- ******************************************************************************* * Bruce Holloway - Terminal Netnews Addict uunet!amdahl!drivax!holloway * * ALBATROSS, ATARI*TROS @ Plink ALBATROSS @ Delphi * *******************************************************************************
apratt@atari.UUCP (Allan Pratt) (02/03/88)
in article <3005@drivax.UUCP>, holloway@drivax.UUCP (Bruce Holloway) says: > But then a systems programmer said that _he_ could write a new operating > system in a month that would look just like MS-DOS from a functional > standpoint, and would be great and perfect and peachy-keen. And he did - > pretty much alone. And since Atari was in such a hurry, they took what we > had before all the horrid problems in it cropped up. > > And later we fixed all those problems and then forgot about it. I don't think > there's anyone here still working on GEMDOS. I take issue with the second paragraph: it took DRI (in the person of that wonder-programmer mentioned above) a *long* time to come up with anything new, and when they did, it still had bugs. I noticed one in the first day of testing it. (Dirty buffers would get flushed to disk at odd times, later than you expect.) We (Atari) decided not to trade a set of known bugs for a set of unknown ones, so we didn't adopt this "DOS 1.1" from DRI. While there may be nobody at DRI working on GEMDOS, there sure is somebody at Atari working on it. Stay tuned for more details. ============================================ Opinions expressed above do not necessarily -- Allan Pratt, Atari Corp. reflect those of Atari Corp. or anyone else. ...ames!atari!apratt
mj@myrias.UUCP (Michal Jaegermann) (02/05/88)
Bruce Holloway gave us his version of origins of GEMDOS. One would
really would like to believe in all these stories about one mythical
evil system programmer. There is only one fly in this ointment. DRI,
as I understand, is responsible for a lot of other "goodies" on ST.
Let us see. First GEM itself. Is also full of bugs (should I include
a well-known list). It does practically nothing else but puts slow and
unwieldy graphics shell on the top of your computer. This shell has not
a slightest clue what is a search path, environement variable, etc. Can
handle one file only at a time (with an exception of copying - but pray
that you do not have grab more files then slots in a window). What is
more GEM was crippled from the very start because of missing integral
part - an infamous GDOS. Therefore no proportional fonts, no device
independent output, no (actually) multiple virtual workstations. In
other words - nothing from all these things for which one really suffers
and learn to live with cutesy, SLOW, graphics interfaces. You may
continue this story on your own.
Ok, wo what else we got from DRI. Well known Alcyon C compiler in
which nobody bothered to fix even such simple things like error values
returned from every pass. Accompanying "documentation" and tools were
of similar quality. A recent message from Neil Harris indicates that
currently Atari values Alcyon C for $20, which seems to be generous.
This was followed by the next DRI product included with ST, namely
Dr. Logo. Probably one of worst Logo implementations which I ever
seen. Both buggy (it rally cannot handle any more involved recursion)
and, once again, awfully slow. Logo on 1MHz 9900 (TI99/4a) is much
faster and better done then DR.LOGO on 8MHz 68000. Though the first one
is a little bit more cramped, due to somehow smaller memory available.
DR.LOGO is also no good for anything else, but some turtle play (if you
are really patient). Reason - even rudimentary file access was
conveniently forgotten.
This LOGO was also graced with an editor which can be overtyped by a one
finger typist. I guess that one has to write it in interpreter written
in slow BASIC to achieve that effect.
The only redeeming feature in all of this is ST BASIC. It was so
awfull, that nobody wanted to use it at all and everybody switched
immediately to other languages. It this was a part of a master plan to
force good habits onto microcomputer users, then it succeded
admirably. Only I am not sure if DRI was responsible also for that.
If all of above was one-month doing of a busy system programmer, then
this guy, for sure, had a quite hot month.
Michal Jaegermann
...alberta!myrias!mj
---------------------------------------------------------------------
Disclaimer: All of the above are purely my personal opinions. My
employer even does not know what is ST ( :-) ).
--------------------------------------------------------------------landon@apple.UUCP (Landon Dyer) (02/06/88)
> And we already had an 'okay' 68000 operating system, CP/M-68K, and we were > (then) working on a multitasking, multiuser version of the same called > "CDOS/68K". But GEM ran only on MS-DOS machines, and sorta depended on that > file structure. Hmmm. So we considered changing GEM... Actually, CDOS was DRI's "black hole" project. It kept eating manpower, and a lotta people wanted off. This is relevant to: > But then a systems programmer said that _he_ could write a new operating > system in a month that would look just like MS-DOS from a functional > standpoint, and would be great and perfect and peachy-keen. And he did - > pretty much alone. And since Atari was in such a hurry, they took what we > had before all the horrid problems in it cropped up. The mists of history part... GEMDOS was originally written by Jason "Born to Code" Loveman, a programmer on DRI's CDOS effort. So sick of working on file systems during the day, he went home at night to spend time writing his own file system on his Compac. HE did it HIS way. CP/M-68K was "okay", but that didn't mean stellar. That means we were looking for just about any alternative (real folders sure sounded better than 16 user areas, though we didn't know about the 40-folder limit at the time). So we did a port of GEMDOS for January 1984 CES; it was in ROM at the show, though it wasn't pretty, and it was part of the reason we didn't let anyone touch the very first six working STs. Over the next few months GEMDOS gradually got more solid. We made a disk release of the system in June of 1984, though we weren't too happy about it. The next five months before the first ROM release saw little improvement in the file system. DRI had fired Jason, and some new people at DRI were struggling valiantly with his idiosyncratic brain-child. We saw few updates from DRI, though we squashed quite a few bugs on our own. My favorite line to describe the state of GEMDOS (and TOS in general) is still Han Solo's from The Empire Strikes Back: "Whew! And I thought they smelled bad ... on the outside!" > And later we fixed all those problems and then forgot about it. I don't think > there's anyone here still working on GEMDOS. DRI did not fix "all those problems." No one has (unless it's Allan Pratt). It is very clear (though not very interesting, now) that DRI has abandonded GEMDOS. Probably just as well... -Landon -- I speak for me.
alexande@drivax.UUCP (Mark Alexander) (02/09/88)
Landon Dyer's summary is accurate in the broad sense, although some
of the unimportant details weren't quite right.
I think it's a mistake to blame just one individual for the problems
with GEMDOS. For sure, having programmers that aren't motivated to
produce good work doesn't help. But that project also suffered from
extremely tight, over-optimistic schedules that couldn't be met
without compromising quality.
Maybe DRI marketing types thought the Atari deal would bring them the
glory and bucks they lost when Microsoft got in good with IBM. Frankly,
it didn't work. After reality set in, it was apparent that DRI would
never make much money off the Atari. That's probably why they never
put much effort into really fixing GEMDOS.
Disclaimer: I don't speak for DRI. These are my own opinions
and observations.
--
Mark Alexander ...{hplabs,seismo,sun,ihnp4}!amdahl!drivax!alexande
"Bob-ism: the Faith that changes to meet YOUR needs." -- Bobbraun@drivax.UUCP (Kral) (02/09/88)
In article <3046@drivax.UUCP> alexande@drivax.UUCP (Mark Alexander) writes: > >Landon Dyer's summary is accurate in the broad sense, although some >of the unimportant details weren't quite right. > As one of the last people to work on this project (with one exception), I can only add my support to most of the comments from Atari and DRI sources, particularly Landon's quote from Star Wars. If Atari has only fixed a few bugs in that system, then they are to be commended from it. As Mr. Pratt probably knows all too well, it is far to easy to swap new bugs for old ones without re-writting major chunks. Mark sums up the "motivation" behind the results: mostly marketing types working out deals with unrealistic schedules behind them. Bruce's comments on the origin, while simplistic, are not too far off base. If generalized, they apply to much of the system that someone else (sorry, forgot the name) was complaining about (eg, Alcyon came, er, for next to nothing, and was "already complete"). As most programmers know, if you do a bad (read: hurried) job the first time, it will most likely take longer and longer to get seemingly small bugs fixed. My sympathies go out to Mr. Pratt. DISCLAIMER: complaints and congratulations come from me, not DRI. -- kral [THERE ARE NO ORDINARY MOMENTS] 408/647-6112 ...{ism780|amdahl}!drivax!braun [You learn by what you do, Not by what you *don't* do] DISCLAIMER: If DRI knew I was saying this stuff, they would shut me d~-~oxx
wes@obie.UUCP (Barnacle Wes) (02/13/88)
In article <7340@apple.UUCP>, landon@apple.UUCP (Landon Dyer) writes: > It is very clear (though not very interesting, now) that DRI has abandonded > GEMDOS. Probably just as well... It would seem that Atari is abandoning GEMDOS as well. A.P. has fixed much of what is wrong with it, and there are rumors of a disk OS being distributed to developers for testing, but Atari still hasn't gotten the blitter ROMS out to the unwashed masses yet. Andy Tanenbaum has said he should have the ST version of Minix out some time this spring. I may never go back to GEMDOS. Anything I don't like about Minix, I can fix, or add! P.S. - Landon, what does Apple have you working on? An ST emulator for the Mac II? :-) -- /\ - " Against Stupidity, - {backbones}! /\/\ . /\ - The Gods Themselves - utah-cs!utah-gr! / \/ \/\/ \ - Contend in Vain." - uplherc!sp7040! / U i n T e c h \ - Isaac Asimov - obie!wes