From mboxrd@z Thu Jan  1 00:00:00 1970
Message-ID: <f884497a8f5ccfb72c05e70c3e14e652@proxima.alt.za>
To: 9fans@cse.psu.edu
Subject: Re: [9fans] GNU Make
Date: Wed,  2 Jun 2004 13:13:39 +0200
From: lucio@proxima.alt.za
MIME-Version: 1.0
Content-Type: text/plain; charset="US-ASCII"
Content-Transfer-Encoding: 7bit
Topicbox-Message-UUID: 9013ab26-eacd-11e9-9e20-41e7f4b1d025

> Messages and Codes was notoriously unhelpful in many cases.

Sure, but was the problem with the concept or with the implementation?
In my particular case, my concern lies largely with system library
functions, a finite error message space.  In fact, given that a
significant portion of the errors are concurrent with 9P, it ought to
be trivial to rationalise at least those error messages into a finite
set.

I'm suggesting that discipline in the use of error messages would be
constructive and that one mechanism to encourage such discipline would
be the use of indexing.  I appreciate that it then becomes difficult
to escape the enforced rigidity, but one can add the escape mechanism
a priori.

I actually fail to see any other disadvantages of a disciplined
approach.  I do see very clearly where the Plan 9 approach becomes a
nightmare.

++L