Date: Sun, 31 Aug 1997 17:49:25 -0400 (EDT)
Message-Id: <199708312149.RAA26964@delorie.com>
From: DJ Delorie <dj AT delorie DOT com>
To: djgpp-workers AT delorie DOT com
Subject: djgpp 2.02 alpha 970831
Precedence: bulk


I'm uploading a snapshot of what I got done this weekend (so far) to
Simtel.Net (for speed).  It will be in v2/alpha/970831/

Note that timezone stuff have their own zips now.

The knowledge base has a (mostly complete) list of changes.

----------

Issues that I did NOT resolve, that need more thought:

I still have a fork() implementation from <malcolm AT manawatu DOT gen DOT nz>
that nobody has commented on.

There were multiple stackavail candidates, none of them "complete".

Robert's symlink patches from 19 June: Why patch <stdlib.h>?  Why not
<process.h>?

redir: no changes.  I hope to find time to redo this completely, by
reading the original command line directly and manipulating it by
hand.

mktime() "fix" for dates within a daylight savings time change (i.e.
two possible time_t results): I think that this *is* an error, and
mktime should return -1.  However, the original problem still exists:
converting dos m/d/y to a unix time_t.  Would it be better to just
call mktime twice at the right spot: if it fails the first time, add
an hour, call mktime, then subtract an hour from the result.

There were some conflicting _use_lfn.c patches.  The ones that didn't
make it were the getenv() optimizations, since they seemed less agreed
upon.

access vs prn: we never decided if the patch it was a good idea.