[comp.os.minix] MINIX, questions and future

bl@infovax.UUCP (Bj|rn Larsson) (11/16/89)

I have MINIX 1.2, aquired from Prentice-Hall a long time ago, and
I have just started (finally) to play with it. It works reasonably
on my two-floppy 386 machine (if I turn down the clock to 8 Mhz).
I would like to start playing with the sources and modify them,
and so far building boot diskettes and compiling the libraries
(using TurboC 2.0 on DOS) seems to work. However, since my version
is so old, I have the following questions (to AST and you others).
Please bear with me if I just re-iterate the most common ones...

1)	I have seen mention about 1.3, 1.4* and 2.0. Of these, my
	understanding is that 1.3 comes from Prentice-Hall now.
	Is there an 'official' 1.4x version? I think I've seen
	that AST is working on 'getting 2.0 out'. When will that
	be (approximately)?

2)	Is it possible to say something about which parts of MINIX
	have been most changed since 1.2? I've read about new drivers
	for RS232 and for winchesters. Amoeba (which I seem to under-
	stand is a networking interface) is not part of MINIX, I
	take it, even if it can run with it? With changes to MINIX,
	I mean the OS itself like above, not the various utilities/
	applications etc that run under it. But, are sources to the
	compiler passes available nowadays?

3)	Do later 'official' versions of MINIX source provide ready-
	made support for generating MINIX using Turbo C under DOS?
	This would be my development environment of choice. C86 is
	not a very common compiler, and PCIX systems are probably
	even more scarce. And in my case, doing the work under MINIX
	using 2 floppy drives is a nuisance. For me, a hard disk
	is essential to work efficiently - and 1.2 doesn't work with
	my HD. (This is not an important point - I can modify for
	TC without problems - but it would save a few hours if I
	don't have to. And the first thing would be to get the HD
	functional).

4)	Does AST 'approve' new versions? My impression is that
	many people make their own modifications to the project
	and then post the results. Sometimes their different con-
	tributions must conflict. So who decides what is the
	'current official' version, which is guaranteed (hrmm)
	to work on at least a majority of machines. Or is there
	no such thing? My feeling is there should be a MINIX co-
	ordinator (not necessarily AST - I'm sure he has his hands
	full) (BTW, not me either - I have hands AND feet full,
	and probably not the ability!)

5)	Would it be too much to request a source posting (maybe
	in compressed form) of the 'current' kernel / mm / fs, and
	the C and start-up library? Applying patches (and which
	ones are 'official'?) to old versions has a big potential
	for errors when multiple generation patches are involved,
	and when contradicting patches float around. As I see it,
	posting these items would still use much less UseNet bandwidth
	than for example the mamooth postings that sometimes appear
	in comp.sources.unix (no harm intended, mr $alz). A posting
	would provide a firm ground to stand on, and clean out dead-
	ends that may have been introduced by intermiediate work.

	Probably such postings would be made in ARC or ZOO format,
	since I believe almost everybody have these - but that's my
	personal point of view.

6)	How big is the time-lag for the latest version at AST as
	compared to the one that can be gotten from PH? Also, how
	reliable are they (PH) nowadays? When I and another guy at
	work 	ordered 1.2, one of us got 1.2 and one got 1.1 - in
	the same shipping container!

7)	My impression from reading comp.os.minix is that the initial
	big interest in MINIX is diminishing. Fewer people seem to
	post their views and modifications, and those who do tend to
	be a small 'insider' group. Do you out there think there's
	some justification in this impression of mine and, if so,
	why do you think it is?

	My personal answer to this is that if there's this small
	group working with MINIX development, other people may feel
	on the outside, and therefore they lose interest. This is
	particularly true since almost nobody has access to what is
	currently going on, and therefore it is impossible to influ-
	ence the direction things are going. There are no easy solu-
	tions to this of course. (I'm really exposing myself to
	flames, aren't I?)

8)	On technical matters: My floppy driver doesn't work when the
	CPU runs 20MHz. It fails reading disk blocks, both with 80-
	and 40-track diskettes (both work at 8MHz). Under DOS every-
	thing is OK. Anybody has a fix or can guess why this is so?

Answers / opinions about the above are appreciated. Please no flames -
if someone takes offence, please let me know in a civilized way. I
realize that my views may not fit in with the MINIX philosophy. I
want to be able to experiment like everybody else, but I also don't
want to work with sources that are 3 generations old, with bugs that
have since been cleaned out. In other words, I want to know what the
corrent and 'official' MINIX version is, and I would like to see some
stricter order in the MINIX activities... to avoid re-inventing wheels.

							Bjorn

-- 
 ====================== InfoVox = Speech Technology =======================
 Bjorn Larsson, INFOVOX AB      :      ...seismo!mcvax!kth!sunic!infovax!bl
 Box 2503                       :         bl@infovox.se
 S-171 02 Solna, Sweden         :         Phone (+46) 8 735 80 90