john@chance.UUCP (John R. MacMillan) (03/14/91)
|Is anyone working on a port of MINIX to the AT&T Unix PC? |The machine is also called the 3B1 or 7300. A friend and I have just started, using the Atari ST source as a base. We don't think it will be too hard, but we're both pretty busy, so it may be a while. When we have something working, I'll post a notice, and interested people can ask for the diffs.
ac999321@umbc5.umbc.edu (ac999321) (03/15/91)
In article <1991Mar14.054318.25699@chance.UUCP> john@chance.UUCP (John R. MacMillan) writes: >|Is anyone working on a port of MINIX to the AT&T Unix PC? >|The machine is also called the 3B1 or 7300. > >A friend and I have just started, using the Atari ST source as a >base. We don't think it will be too hard, but we're both pretty busy, WHAT???? Do you have this system confused with IBM's toy PC (and clones)? The AT&T _Unix_ PC is named the Unix PC because it was designed to run a derivative of AT&T's SysVR2, which comes with the machine. Just curious; why would you want to port Minix to this system when you already have a better version of Unix running on it?
dm@think.com (Dave Mankins) (03/15/91)
In article <5409@umbc3.UMBC.EDU> ac999321@umbc5.umbc.edu.UUCP (ac999321) writes: >Just curious; why would you want to port Minix to this system when you >already have a better version of Unix running on it? Two reasons: ``better'' is open to dispute --- quite a few of us curmudgeons still look back fondly on Version 7, with which MINIX is compatible. complete sources --- even though I have quite enough kernel hacking at work to keep me entertained, I still itch to change a few things on the 3b1 (getting rid of the 14 character filename limit is the biggest, adding a file-system driver that talks across a serial line to the *really big* disks at work, changing the screen driver so windows can stretch from the top to the bottom of the screen, and so they only have one-pixel borders...) -- david mankins (dm@think.com)
mvadh@cbnews.att.com (andrew.d.hay) (03/15/91)
In article <1991Mar14.054318.25699@chance.UUCP> john@chance.UUCP (John R. MacMillan) writes:
"|Is anyone working on a port of MINIX to the AT&T Unix PC?
"|The machine is also called the 3B1 or 7300.
"
"A friend and I have just started, using the Atari ST source as a
"base. We don't think it will be too hard, but we're both pretty busy,
"so it may be a while. When we have something working, I'll post a
"notice, and interested people can ask for the diffs.
while this is an interesting exercise, i think minix is deficient in a
few areas. for instance, no future version of minix will ever have
paging or swapping virtual memory; tanenbaum is actively opposed to
including it.
i'd be much more interested in a port of mach 3.0, or maybe amoeba...
i also have a Modest Proposal (TM): set up a non-profit corporation
under the umbrella of ASCRC (a small computer software foundation?).
get a source license from AT&T for the 3b1 (or convince them to turn
over the source, with the obligation to support the 3b1) *and* from
convergent/unisys for miniframe ctix. this would be funded by selling
shares to us on the net. there's another benefit from funding it this
way: as owners of the corporation, we'd have access to source (or at
least executables) in the corporation's possession without the need
for any kind of redistribution license. nothing would ever be sold!
why would we want to include ctix? well, it's an evolution of the
same code the 3b1 is based on; this makes it an easy (trivial?) port.
but it has some valuable features, like sVr3 compatibility and
loadable disk drivers...
--
Andrew Hay +------------------------------------------------------+
Ragged Individualist | JAAAAAAANE! |
AT&T-BL Ward Hill MA | HOW DO YOU STOP THIS CRAZY THING? |
a.d.hay@att.com +------------------------------------------------------+jkp@cs.HUT.FI (Jyrki Kuoppala) (03/15/91)
In article <5409@umbc3.UMBC.EDU>, ac999321@umbc5 (ac999321) writes: >Just curious; why would you want to port Minix to this system when you >already have a better version of Unix running on it? How about things like having the source, being able to fix bugs, being able to make the kernel better ? Minix is a good choice for this now, because it has good support from the net community and because there are no truly freely distributable alternatives (for the whole system, I mean, but that will probably change sometime before year 2000) currently available (hmm, don't know if Trix counts). Let's take an example: I wasn't satisfied with the 14 character file name length limitation on Minix. I just changed two constants in two include files, wrote a program to convert directories to the new format and rebuilt my system - almost by magic it has 62-character file names now. How do you that on the Unix-PC's OS ? //Jyrki
jaws@sequoia.cray.com (James A. Wheeler) (03/16/91)
In article <1991Mar15.130239.20389@cbnews.att.com> mvadh@cbnews.att.com (andrew.d.hay) writes: > >i also have a Modest Proposal (TM): set up a non-profit corporation >under the umbrella of ASCRC (a small computer software foundation?). >get a source license from AT&T for the 3b1 (or convince them to turn >over the source, with the obligation to support the 3b1) *and* from >convergent/unisys for miniframe ctix. this would be funded by selling [ ...Lots of good points deleted] I like this Idea, how many are willing to investigate? ------------------------------- J.A. Wheeler (jaws@brazos.cray.com) Cray Research, Inc. 5 Post Oak Park suite 1960 Houston, Tx 77027 (713)-297-7834
peter@ficc.ferranti.com (Peter da Silva) (03/16/91)
In article <5409@umbc3.UMBC.EDU> ac999321@umbc5.umbc.edu.UUCP (ac999321) writes: > Just curious; why would you want to port Minix to this system when you > already have a better version of Unix running on it? That was my reaction to the idea of a native port of MINIX to the Amiga. I can think of a few reasons: source availability, support woes, hack value. -- Peter da Silva. `-_-' peter@ferranti.com +1 713 274 5180. 'U` "Have you hugged your wolf today?"
zaft@slced1.nswses.navy.mil (Gordon C Zaft) (03/16/91)
>In article <1991Mar15.130239.20389@cbnews.att.com> mvadh@cbnews.att.com (andrew.d.hay) writes: >> >>i also have a Modest Proposal (TM): set up a non-profit corporation >>under the umbrella of ASCRC (a small computer software foundation?). >>get a source license from AT&T for the 3b1 (or convince them to turn >>over the source, with the obligation to support the 3b1) *and* from >>convergent/unisys for miniframe ctix. this would be funded by selling I think this idea might be worth pursuing; at least, it would solve a number of long term problems in regards to the future of this machine. On the other hand, I'm not sure it's really worth it when you consider the costs involved.... What do others think? -- + Gordon Zaft | zaft@suned1.nswses.navy.mil + + NSWSES, Code 4Y33 | suned1!zaft@elroy.jpl.nasa.gov + + Port Hueneme, CA 93043-5007 | Phone: (805) 982-0684 FAX: 982-8768 + ** ..et resurrexit tertia die secundum scripturas, et ascendit in coelum.. ***
wwm@wa8tzg.mi.org (Bill Meahan) (03/17/91)
In article <1991Mar15.130239.20389@cbnews.att.com> mvadh@cbnews.att.com (andrew.d.hay) writes: >In article <1991Mar14.054318.25699@chance.UUCP> john@chance.UUCP (John R. MacMillan) writes: >"|Is anyone working on a port of MINIX to the AT&T Unix PC? >"|The machine is also called the 3B1 or 7300. >" >"A friend and I have just started, using the Atari ST source as a >"base. We don't think it will be too hard, but we're both pretty busy, >"so it may be a while. When we have something working, I'll post a >"notice, and interested people can ask for the diffs. > >while this is an interesting exercise, i think minix is deficient in a >few areas. for instance, no future version of minix will ever have >paging or swapping virtual memory; tanenbaum is actively opposed to >including it. > >i'd be much more interested in a port of mach 3.0, or maybe amoeba... > >i also have a Modest Proposal (TM): set up a non-profit corporation >under the umbrella of ASCRC (a small computer software foundation?). >get a source license from AT&T for the 3b1 (or convince them to turn >over the source, with the obligation to support the 3b1) *and* from >convergent/unisys for miniframe ctix. this would be funded by selling >shares to us on the net. there's another benefit from funding it this >way: as owners of the corporation, we'd have access to source (or at >least executables) in the corporation's possession without the need >for any kind of redistribution license. nothing would ever be sold! > >why would we want to include ctix? well, it's an evolution of the >same code the 3b1 is based on; this makes it an easy (trivial?) port. >but it has some valuable features, like sVr3 compatibility and >loadable disk drivers... > >-- >Andrew Hay +------------------------------------------------------+ >Ragged Individualist | JAAAAAAANE! | >AT&T-BL Ward Hill MA | HOW DO YOU STOP THIS CRAZY THING? | >a.d.hay@att.com +------------------------------------------------------+ Hear, hear!! I SECOND THE MOTION. All in favor .... Funny, I was just about to post a similar suggestion!! As long as the per-share price stays reasonable, say <$50 US [no, I'm not cheap, I just have a daughter in college and a wife in graduate school so I'm poverty-stricken until at least late 1992] we should ALL benefit. We could make this a stock-issuing company (still semi-private). People could purchase more than one share if they wished and non-shareholders could either purchase stock or pay a REASONABLE fee for anything developed by the company. What say, 3b1 folks, this makes REAL sense to me. Who knows, maybe out combined expertise could accomplish some things the big companies can't. -- Bill Meahan (WA8TZG) | Programming is simple: wwm@wa8tzg.mi.org OR | uunet!mailrus!sharkey!wa8tzg!wwm | All you have to do is put the right "Home for Cybernetic Orphans" | numbers in the right memory locations!
john@chance.UUCP (John R. MacMillan) (03/18/91)
|>A friend and I have just started, using the Atari ST source as a |>base. We don't think it will be too hard, but we're both pretty busy, | |WHAT???? Do you have this system confused with IBM's toy PC (and clones)? |The AT&T _Unix_ PC is named the Unix PC because it was designed to run |a derivative of AT&T's SysVR2, which comes with the machine. Damn! Now I know why I haven't been able to get Lotus 1-2-3 to work for the past 3 1/2 years! :-) Seriously, I really do mean we're porting Minix to the AT&T UNIX PC, aka 3B1 aka PC7300. Wouldn't be much work to do a port where all you had to do was buy the floppies for the target machine, and stuff them in, now, would it? :-) |Just curious; why would you want to port Minix to this system when you |already have a better version of Unix running on it? A few reasons. I don't get to do any kernel work at work, and I don't have source for the UNIXpc OS. Also, at the time we decided to do the port, it didn't look like any future upgrades from AT&T would be forth- coming, so it would have been one way to support future hardware development with the machine. Also ``better'' is a matter of opinion. Personally, I feel modern commercial Unices have too much bloat, and even if upgrades come, they're not likely to trim down any, so the only way for me to get a ``better'' UNIX is to start small, and do it myself.
dlb5404@tamuts.tamu.edu (Daryl Biberdorf) (03/18/91)
I'd be interested in this proposed incorporation scheme, but I have
an idea. Wouldn't it be better to appoint someone to collect names
to see how many people are interested rather than cluttering up the
group? If the original suggester of this idea isn't willing, I am.
Let me know....
--Daryl Biberdorf, dlb5404@{rigel,tamuts}.tamu.edu
Texas A&M Universityjohn@chance.UUCP (John R. MacMillan) (03/18/91)
[ I've cut comp.os.minix out of this -- JRM ] |while this is an interesting exercise, That's mostly why we're doing it. |... i think minix is deficient in a |few areas. for instance, no future version of minix will ever have |paging or swapping virtual memory; tanenbaum is actively opposed to |including it. He's not going to put it in the official distribution, because he wants to keep the source similar and smaill on all platform, and wants to support the lowest common denominator. That doesn't mean we can't add it to our port. It makes for a more interesting exercise that way. :-) |i'd be much more interested in a port of mach 3.0, or maybe amoeba... Go ahead, I'd like to see them too. :-) Both are probably a little on the large size for our little machine, and neither of is terribly UNIX-ish, at their lowest level. The Mach micro-kernel is just micro by modern kernel standards, and what's currently available doesn't include the UNIX emulation (which also adds quite a bit to the size). Plus mach is just a kernel; no utilities. Many are available from GNU, but that makes for a lot more work. As for various plans to enable us to get our grimy little mitts on the UNIX PC source, I really hope they go well, but I'm not waiting.
tom@afthree.as.arizona.edu (Thomas J. Trebisky) (03/19/91)
jaws@sequoia.cray.com (James A. Wheeler) writes: >mvadh@cbnews.att.com (andrew.d.hay) writes: >> >>i also have a Modest Proposal (TM): set up a non-profit corporation >>under the umbrella of ASCRC (a small computer software foundation?). >>get a source license from AT&T for the 3b1 (or convince them to turn >>over the source, with the obligation to support the 3b1) *and* from >>convergent/unisys for miniframe ctix. >I like this Idea, how many are willing to investigate? I too like this idea (I admire the original motive of the Minix port workers, namely to have the sources for the system they are running.) Getting the system V stuff for the 3B1 series into the hands of someone able to turn it to the public good would be wonderful. Another twist - what about a port of 4.3BSD to the Unix PC (note: I don't want to enter into a potential firestorm of System V versus BSD debate here.) I have been taking a hard look at porting 4.3BSD to the CT miniframe for some time now (and was thinking of pressing my newly acquired unix-PC into service on this project as development systems). I am beginning to learn that the 3b1 and miniframe are almost brothers. Sooooo anything along these lines would be admirable. -- Tom Trebisky ttrebisky@as.arizona.edu (Internet) Steward Observatory University of Arizona Tucson, Arizona