Path: utzoo!mnetor!uunet!husc6!bbn!rochester!cornell!batcomputer!itsgw!imagine!pawl14.pawl.rpi.edu!jesup
From: jesup@pawl14.pawl.rpi.edu (Randell E. Jesup)
Newsgroups: comp.os.minix
Subject: Re: The Amiga Port
Message-ID: <597@imagine.PAWL.RPI.EDU>
Date: 31 Mar 88 03:07:31 GMT
References: <1483@louie.udel.EDU> <343@wsccs.UUCP> <1781@sugar.UUCP>
Sender: news@imagine.PAWL.RPI.EDU
Reply-To: beowulf!lunge!jesup@steinmetz.UUCP
Organization: RPI Public Access Workstation Lab - Troy, NY
Lines: 45

In article <1781@sugar.UUCP> karl@sugar.UUCP (Karl Lehenbauer) writes:
>Now hold on there, flamer.  You apparently haven't been reading
>comp.os.minix for long.  The idea of doing a Minix port for the 
>Amiga has been beaten to death.  The consensus has been that Minix needs 
>a way to relocate logical address zero and the Amiga can't (the ST can,
>hence an ST port is, or is nearly, available), making it a *big* job.  

	Sorry, not correct.  I am a former member of the amiga minix
mailing list.  Niether the ST nor the amiga can relocate la 0, but that
doesn't mean  you can't do minix, it just requires a bit of handwaving when
doing a fork().  The job would be almost the exact same work as went into
the ST version (actually less, since they've already done the 68000 conversion
stuff, and the fork() stuff).

>Further, it was widely agreed that massive amounts of development would 
>be required to approach the functionality of the Amiga under Commodore's 
>OS and toolset:  code for managing the blitter and copper, device 
>drivers, libraries, etc. etc. etc.  Also, the incentive to port a 
>multitasking exec to the Amiga isn't there nearly as much as for the 
>ST and PC, since the Amiga is already multitasking and the ST and
>PC aren't.

	BINGO!  That's the main reason why the project died out: a fairly large
amount of work to get something you already have.  Being able to run V7
compatible unix programs is nice, but most programs aren't v7-compatible.
A better solution is to use AmigaDos/exec/intuition/graphics libraries to
implement minix, and *poof* windowing V7 unix!  But still not really worth
the work to do it.

>  Having seen how Minix passes messages
>(single message rendezvous) versus how the Amiga does it (multi-message
>queuing, queueing without wait, also, lightweight processes), it is probable 
>that Minix will be slower than AmigaDOS on an Amiga, although the current
>version of AmigaDOS is bad enough at disk I/O that Minix could win on some
>I/O functions.

	Also very true.  I couldn't believe it when I realized that minix
messages didn't queue.

     //	Randell Jesup			      Lunge Software Development
    //	Dedicated Amiga Programmer            13 Frear Ave, Troy, NY 12180
 \\//	beowulf!lunge!jesup@steinmetz.UUCP    (518) 272-2942
  \/    (uunet!steinmetz!beowulf!lunge!jesup) BIX: rjesup

(-: The Few, The Proud, The Architects of the RPM40 40MIPS CMOS Micro :-)