Mail Archives: djgpp/1997/09/06/12:47:50
From: | pjfarley AT dorsai DOT org (Peter J. Farley III)
|
Newsgroups: | comp.os.msdos.djgpp
|
Subject: | Suggestion for future DJGPP development -- depend on bash
|
Date: | Wed, 03 Sep 1997 03:36:00 GMT
|
Organization: | None
|
Lines: | 32
|
Message-ID: | <340cd19f.6787527@snews.zippo.com>
|
NNTP-Posting-Host: | news.newsdawg.com
|
To: | djgpp AT delorie DOT com
|
DJ-Gateway: | from newsgroup comp.os.msdos.djgpp
|
For discussion:
In attempting a rebuild of a patched gcc, I have discovered what I
feel are avoidable problems in an LFN=y environment, and have a
suggestion for future gcc development and distribution that I feel
would resolve these problems compatibly with 8.3 requirements.
With the advent of a real bash in the DJGPP environment, it seems to
me *not* to make a great deal of sense not to depend on a working
binary of bash as a pre-condition for source-building gcc. To restate
that last bit more clearly, my suggestion is this: Make it a
requirement for gcc source building that a binary of DJGPP bash is
required to perform the build.
The advantages are obvious: The unix configuration scripts can all be
made compatible and operable with DJGPP bash, which is a much easier
job than trying to duplicate them in bat files, and (all?) LFN issues
can be assumed to be addressed by the pre-existing bash environment.
This allows all zipped source files to be configured with original gcc
long names, while retaining 8.3 compatibility with non-LFN
environments.
In addition, shifting now will make the transition to new versions
(whenever they appear) of gcc more painless and therefore quicker to
implement.
I can't think of any disadvantages, but my experience is admitedly
limited. I'm open to further rational discussion of any you can
envision.
----------------------------------------------------
Peter J. Farley III (pjfarley AT dorsai DOT org)
- Raw text -