[comp.os.minix] Proposal to split up comp.os.minix

mju@mudos.ann-arbor.mi.us (Marc Unangst) (02/26/89)

I've only been on this newsgroup for 5 months, but I've noticed that 
there are a lot of people that read this group that don't have both an 
ST and an IBM.  Thus, either the ST or the IBM binary/source postings 
are useless (depending on which machine you have).  Therefore, I think 
that it would be a good idea to split up comp.os.minix into 5 groups:

        comp.binaries.minix.ibm - IBM Minix binary postings.
        comp.binaries.minix.st  - Atari ST Minix binary postings.
        comp.sources.minix.ibm  - IBM Minix source postings.
        comp.sources.minix.st   - Atari ST Minix source postings.
        comp.os.minix           - Current comp.os.minix group, but 
                                  without binary or source postings.

The binary and sources groups wouldn't have to be moderated, or they 
could be moderated by a "mirror" that just packs up all submissions with 
appropriate headers, or they could be moderated by a real, live, human 
moderator.

Note that this is NOT a formal call for votes.  It is an attempt to 
gauge the net's feelings about such a split.

Also note that I have neither the time nor the resources to moderate a 
group, so the moderators (if any) would have to be someone other than 
me.

--  
Marc Unangst
UUCP          : mju@mudos.ann-arbor.mi.us
UUCP bang     : ...!uunet!sharkey!mudos!mju
UUCP bang alt.: ...!{ames, rutgers}!mailrus!clip!mudos!mju
Internet      : mju@mudos.ann-arbor.mi.us

willy@idca.tds.PHILIPS.nl (Willy Konijnenberg) (02/27/89)

In article <109.24078AE4@mudos.ann-arbor.mi.us> mju@mudos.ann-arbor.mi.us (Marc Unangst) writes:
>        comp.binaries.minix.ibm - IBM Minix binary postings.
>        comp.binaries.minix.st  - Atari ST Minix binary postings.
>        comp.sources.minix.ibm  - IBM Minix source postings.
>        comp.sources.minix.st   - Atari ST Minix source postings.
>        comp.os.minix           - Current comp.os.minix group, but 
>                                  without binary or source postings.
I think this is overdoing it a bit.
1. I would prefer to *discourage* binary postings.
   If you have a program that compiles on minix, post the source!
   There are, of course, these rare occasions where a program is
   usefull, can run on minix, but won't compile on minix.
2. Minix is unix, unix is supposed to encourage portable programming.
   Most sources should work on both PC, ST, Amiga, Transputer, 32032,
   or any other port of minix. If different architectures require
   minor modifications, then produce a single source with appropriate
   ifdefs, instead of different sources for every architecture.
   If source is not portable, then it must:
	be part of the kernel
	be a system administration tool
	contain bugs
So, *IF* we want to split, 2 groups should be enough. One for sources
and the like, the other for discussions, ideas, questions, etc.
I don't know if we really need to split.
Any ideas?

-- 
	Willy Konijnenberg		<willy@idca.tds.philips.nl>

henry@utzoo.uucp (Henry Spencer) (02/28/89)

In article <109.24078AE4@mudos.ann-arbor.mi.us> mju@mudos.ann-arbor.mi.us (Marc Unangst) writes:
>... either the ST or the IBM binary/source postings 
>are useless (depending on which machine you have)...

Well, I'd call the binary postings near-useless no matter what machine
you have, but I'm one of the old "gotta have the sources" types...
If you are concerned about fun things like viruses, there is good reason
for this attitude, mind you.

The bulk of the source postings are more or less machine-independent.

I'd say put the binaries in comp.binaries.ibm.pc or comp.binaries.atari.st;
that's where they belong, since the groups are *not* named c.b.msdos or
c.b.tos.  The rest can stay here for now.
-- 
The Earth is our mother;       |     Henry Spencer at U of Toronto Zoology
our nine months are up.        | uunet!attcan!utzoo!henry henry@zoo.toronto.edu

greyham@hades.OZ (Greyham Stoney) (02/28/89)

in article <109.24078AE4@mudos.ann-arbor.mi.us>, mju@mudos.ann-arbor.mi.us (Marc Unangst) says:
: 
:         comp.binaries.minix.ibm - IBM Minix binary postings.
:         comp.binaries.minix.st  - Atari ST Minix binary postings.
	       ^^^^^^^^ - How often do binarys get posted anyway????
:         comp.sources.minix.ibm  - IBM Minix source postings.
:         comp.sources.minix.st   - Atari ST Minix source postings.
:         comp.os.minix           - Current comp.os.minix group, but 
:                                   without binary or source postings.

This seems a little silly..... we have an O/S with the FULL source, and we
want to post binaries?.
-- 
# Greyham Stoney:    (disclaimer not necessary: I'm obviously irresponsible)
# greyham@hades.nucleus.oz - Ausonics.        +61 2 428-6476 (my_phone@work)
# replys WILL bounce; try:    greyham@utscsd.oz - Uni of Technology, Sydney.
# WARNING: Reply mail is VERY broken at present. Any replys to utscsd.oz pls

bds@lzaz.ATT.COM (B.SZABLAK) (02/28/89)

In article <198@ssp11.idca.tds.philips.nl>, willy@idca.tds.PHILIPS.nl (Willy Konijnenberg) writes:
>    If source is not portable, then it must:
> 	be part of the kernel
> 	be a system administration tool
> 	contain bugs

Well, there are some tools that are not system administration tools, but still
have machine dependencies. For example, the Minix ST debugger I posted 2 months
ago (mdb) includes a 68000 disassembler, and knowledge of some kernel
data structures. Probably 50% would have to be discarded in porting it to
the PC.

By the way, is there any interest in a repost? There have been significant
bug fixes and minor enhancements to the original post. The ptrace() system
call patches (part of the recent post by Howard Johnson) is required.

Finally, I haven't seen that many binaries go by on this newsgroup, and
then usually with good reason (If its too big to compile on a basic .5 Meg ST
for example, a binary is nice - gcc comes to mind [even 1 Meg won't do]).
I hope people aren't confusing compressed and uuencoded source postings as
binaries...

APEARSON%WAYNEST1.BITNET@cunyvm.cuny.edu (Patrick B. Haggood) (03/01/89)

On the discussion of discouraging binaries:

On on hand, I am kinda lazy in implimenting tools (I like making applications
, not boot-strapping every utility that floats my way, killing hours just to
get a seconds-saving convenience utility working on my system).

However....

Minix is a good OS, but it is primarily a LEARNING OS.  That is, people should
be drawn to it because they wish to understand it (I was).  If I wanted another
OS-in-the-box, I woulda bought OS/9 or Idris (yauc, yet another U**x clone).  S
o the posting of source is consistent with the Minix ideal of available source
for understanding by the user.  Therefore, I endorse the posting of source, and
only posting binaries on very very implicit circumstances.

ncoverby@ndsuvax.UUCP (Glen Overby) (03/04/89)

In article <109.24078AE4@mudos.ann-arbor.mi.us> mju@mudos.ann-arbor.mi.us (Marc Unangst) writes:
>I've only been on this newsgroup for 5 months, but I've noticed that
>there are a lot of people that read this group that don't have both an
>ST and an IBM.  Thus, either the ST or the IBM binary/source postings
>are useless (depending on which machine you have).  Therefore, I think
>that it would be a good idea to split up comp.os.minix into 5 groups:

There have been many calls for a split, and I expect that there will be many
more, and until the volume and disorganization gets much more severe than it
is now, they will all get shot down.

I see the following types of postings in this group:
        (1) questions, problems
        (2) bug reports and fixes
        (3) new "unofficial" programs (i.e. not from Andy or Johan)
        (4) official updates (from Andy or Johan)

Splitting the questions and problems into separate groups for the Atari ST
and IBM PC would be counterproductive, I think.  We are both running one
base system (the Minix sources) and many of the problems will exist in both
systems.  I, for one, would read both groups even though I am stuck running
only MINIX-PC.  If the groups were split, I think that there would be a lot
of crossposting.

Bug reports fall into the same category as (1).  Both groups should remain
unmoderated.

It would nice to make separate groups out of the last two categories, IF
they were moderated.  I don't think there is the volume to merit a group for
both machines, and I don't think there is much use to!  Anybody halfway
serious about Minix SHOULD be tracking both groups to see what the other
half of the brain is doing.  Why subscribe to a mild form of facism by have
a moderator?  Well, the postings can be serialized so that receivers can be
assured that they've gotten everything, non-source postings can be
eliminated ("straight signal"), and archives can be maintained with a
consistent organization.

As an archive site maintainer, I find myself doing a lot of the work which I
believe a moderator would do, but I work in bursts, when I have time,
normally on a long weekend from school.  Thats not too good for a Usenet
moderator.

The latter category would fit in somewhat similar to
comp.bugs.4bsd.ucb-fixes; that is, ONLY fixes from Andy and Johan.  So far,
they haven't been doing these updates that often and that group could be
merged with a moderated Minix sources group, probably making Andy and Johan
sub-moderators with posting privileges (sort of like the Australian
sub-moderator for comp.sources.unix).

Binaries should be banished from the group except in special cases, one
being programs which CANNOT be compiled under Minix (like elle and kermit).
I consider the recent binary postings all "noise".  Give me the sources and
I'll compile the things myself!  You don't have to be RMS to want sources!

In summary, I think the content of comp.os.minix fits into two categories:
discussions and bug reports and sources and updates.  Even though two
different machines are being discussed, they have a LOT in common and do
benefit from being discussed in one place; the volume isn't really that
high, and rn "kill" files do a nice job of eliminating fluff.  Sources, on
the other hand, would benefit from another group WITH a moderator.  Without
one, it would be mere chaos and we'd be no better off than we are now.
--
Glen Overby     <ncoverby@plains.nodak.edu>
                uunet!ndsuvax!ncoverby
                ncoverby@ndsuvax (Bitnet)