[net.followup] Unix, Unixpeople, Usenix - from a non-compunerd's point of view...

earlw@pesnta.UUCP (20) (10/11/85)

Well, I think you have made a good point about Unix and the documentation.
Maybe what we need is net.docs where good documentation can be placed and
expired monthly.  The problem is getting the vendor to write and maintain 
the manuals and documents so the 'average' user can get useful information
without resorting to writing a small test program to understand what was
just read.

pdg@ihdev.UUCP (P. D. Guthrie) (10/11/85)

In article <2850@pesnta.UUCP> earlw@pesnta.UUCP (Earl Wallace) writes:
>Well, I think you have made a good point about Unix and the documentation.
>Maybe what we need is net.docs where good documentation can be placed and
>expired monthly.  The problem is getting the vendor to write and maintain 
>the manuals and documents so the 'average' user can get useful information
>without resorting to writing a small test program to understand what was
>just read.


What a great idea!! net.docs !! There is a real need for "how-to"
manuals on all aspects of unix-life.  I think that these documents could
be as useful as stuff in net.source for sites without a resident guru. 
What we all need to do is to share our knowledge and let more people in
on the "secrets" of Unix.  How about a "How to Port Software Between
Berkeley 4.x and System V.x" for starters.  Actually, this might go
better as a moderated group, just to keep duplications, and junk out. 
Then of course we would need a net.docs.d for discussion about
questionable points in submitted articles.

					Paul Guthrie
					ihnp4!ihdev!pdg

mark@cbosgd.UUCP (Mark Horton) (10/12/85)

In article <2850@pesnta.UUCP> earlw@pesnta.UUCP (Earl Wallace) writes:
>Well, I think you have made a good point about Unix and the documentation.
>Maybe what we need is net.docs where good documentation can be placed and
>expired monthly.  The problem is getting the vendor to write and maintain 
>the manuals and documents so the 'average' user can get useful information
>without resorting to writing a small test program to understand what was
>just read.

Well, I'm not sure I agree.  My shiny new 3B2 came with a stack of manuals
that fills up a 30 inch bookshelf (and took me all day to sort and assemble
into the binders and integrate the updates.)  Sun comes with a similar stack
of manuals.  So does Masscomp.  So does Xenix.  So does PC/IX.  Etc.

You know what?  I can never find anything in any of them.  When I want to
look something up, I'll dig out any reasonably old copy of the UNIX
programmers manual - either 4.1BSD or System III or System Vr1 will do
nicely.  4.2 would probably do fine too except that we can't get the
official Usenix ones, so I have a thick notebook which is unwieldy.
Or I'll just look in /usr/man.

Does this make me a guru?  Well, not really.  I did manage to memorize
what sections 1 through 8 stand for (and if this weren't a moving target
I would get my job done faster) and I do know what command I want info
about.  But I don't think I've ever had time to sit down and read through
30 inches of manuals to see what's available in the system.

This huge set of manuals seems to be viewed as a requirement by the
marketing establishment to get market acceptance.  I suppose it must
be useful to somebody.  But for a reference manual I want something
compact, like the UPM.  One binder (well, two if it won't fit.)

	Mark

lear@topaz.RUTGERS.EDU (eliot lear) (10/14/85)

> Well, I'm not sure I agree.  My shiny new 3B2 came with a stack of manuals
> that fills up a 30 inch bookshelf (and took me all day to sort and assemble
> into the binders and integrate the updates.)  Sun comes with a similar stack
> of manuals.  So does Masscomp.  So does Xenix.  So does PC/IX.  Etc.
> 
> You know what?  I can never find anything in any of them.

Neither can many people, especially those people that have not memorized
sections 1 through 8.  I like the idea of net.doc but how about something
more on the lines of net.questions (or net.answers or net.questions AND
net.questions.answers :-)).

					eliot
-- 

The opinions expressed above reflect no one's opinion save my own and are
so unique, they should be copyrighted!

[lear@topaz.rutgers.edu]
[{allegra,seismo}!topaz!lear]

hans@erisun.UUCP (Hans Albertson) (10/15/85)

[]
To some extent I have to agree. 
To wit: Some, quite a few, of UNIX' proponents are not terribly well
	educated in computer science.
	The awe some seem to feel for the gods, Ritchie, Kernighan,
	Bourne et al is annoying.
But: Not efficient? What version are you using? 4.1 was reasonably
     efficient for the VAX, even in comparison with VMS V2. 4.2 is
     not too bad compared to VMS V4.1. Both trade efficiency for
     functionality. 
     Bad doc? I personally find the man pages of UNIX easier on the
     mind than the diarrhoea of DEC's manual shelves. Difference of
     temperament, no doubt. One thing: While a DEC manual will repeat
     any piece of info some 3-4 times, a UNIX man page is mostly non
     redundant.
     Usenix not on par with ACM. Well. Of course not. Did you expect
     it to? You should compare a Usenix conf with say DECUS. It compares
     quite well, actually.
     UNIX fanatics refusing to discuss other OSes? My favourite is
     TOPS 20. Any day. Not that i've used it as much as UNIX, but
     I recognize excellence when I see it. Soon I may have a chance to
     check on ITS. I have a feeling I'll like that. IBM's VM seems
     like a good design too. BUT, unlike UNIX they are NON-PORTABLE.
     Viciously so. I'll discuss your ears to cinders if you like...
     I was weened on VMS, but it's system calls are just too many,
     with millions of parameters, and they all have to be there every
     time. And it's file system imposes formats on files. IMPOSES!
     not permits. IMPOSES.

Then again, maybe you were just joking.

	-- 
Hans Albertsson, USENET/uucp: {decvax,philabs}!mcvax!enea!log-hb!hans
Real World:  TeleLOGIC AB, Box 1001, S-14901 Nynashamn,SWEDEN

ron@brl-sem.ARPA (Ron Natalie <ron>) (10/15/85)

> Neither can many people, especially those people that have not memorized
> sections 1 through 8.  I like the idea of net.doc but how about something
> more on the lines of net.questions (or net.answers or net.questions AND
> net.questions.answers :-)).
> 
> 					eliot
How about net.pinheads?

dc@datlog.UUCP ( David Crone ) (10/16/85)

In article <354@ihdev.UUCP> pdg@ihdev.UUCP (55224-P. D. Guthrie) writes:
>In article <2850@pesnta.UUCP> earlw@pesnta.UUCP (Earl Wallace) writes:
>>Well, I think you have made a good point about Unix and the documentation.
>>Maybe what we need is net.docs where good documentation can be placed and
>>expired monthly.
>
>
>What a great idea!! net.docs !! There is a real need for "how-to"
>manuals on all aspects of unix-life.  I think that these documents could
>be as useful as stuff in net.source for sites without a resident guru. 
>What we all need to do is to share our knowledge and let more people in
>on the "secrets" of Unix.  How about a "How to Port Software Between
>Berkeley 4.x and System V.x" for starters.  Actually, this might go
>better as a moderated group, just to keep duplications, and junk out. 
>Then of course we would need a net.docs.d for discussion about
>questionable points in submitted articles.
>
>					Paul Guthrie
>					ihnp4!ihdev!pdg

I think both Paul and Earl are missing the point!!!!

UNIX is now a COMMERCIAL ( :-) ) operating system used by manufacturers,
software houses and believe it or not END USERS of applications.

It's time that AT&T and Berkeley revamped/rewrote the system documentation 
so anybody could at least have some understanding of the operation of the
utilities without a) writing test programs b) relying on UNIX 'gurus' or 
c) looking at the source code (if your're lucky enough to have the licence) .

The idea of having a news directory to disseminate 'good' documenation without
making it available to all through the proper channels(AT&T and Berkeley) is to
me completely worthless. Remember the UNIX community is larger than the net.!!! 


UNIX's long term existence now rests within the commercial arena. Such companies
that are using or wish to use UNIX cannot/will not (ultimately) rely on the
guru principle of operation and the lack of good OS documentation could
jeopardise the future of the system.


		Dave (19th Nervous Breakdown) Crone

(These opinions are mine alone, and do not necessarily reflect those of my
 employers or fellow-workers)

nick@inset.UUCP (Nick Stoughton) (10/18/85)

In article <1542@cbosgd.UUCP> mark@cbpavo.UUCP (Mark Horton) writes:
>You know what?  I can never find anything in any of them.  When I want to
>look something up, I'll dig out any reasonably old copy of the UNIX
>programmers manual - either 4.1BSD or System III or System Vr1 will do
>nicely.  4.2 would probably do fine too except that we can't get the
>official Usenix ones, so I have a thick notebook which is unwieldy.
>Or I'll just look in /usr/man.

I know the feeling all too well...I've resorted to looking in a
V5 (yes, version 5, not Sys V) on more than one occasion, though
I must admit that the commonest manual I use is V7. Even the distributed
4.2 manuals are too thick to fit into a single binder, which is what I
really want. Trouble is, the more that goes into a system, the more
documentation there has to be with it.