Message-Id:	<199809071354.PAA36430@ieva06.lanet.lv>
From:	"Andris Pavenis" <pavenis AT lanet DOT lv>
To:	Eli Zaretskii <eliz AT is DOT elta DOT co DOT il>
Date:	Mon, 7 Sep 1998 15:59:05 +0300
MIME-Version:	1.0
Subject:	Re: (should be: egcs-1.1a uploaded)
CC:	djgpp-workers AT delorie DOT com
References:	<Pine DOT A32 DOT 3 DOT 91 DOT 980907114645 DOT 61918A-100000 AT ieva06 DOT lanet DOT lv>
In-reply-to:	<Pine.SUN.3.91.980907124930.23961D-100000@is>

Date sent:      	Mon, 7 Sep 1998 12:56:03 +0300 (IDT)
From:           	Eli Zaretskii <eliz AT is DOT elta DOT co DOT il>
Subject:        	Re: (should be: egcs-1.1a uploaded)

> 
> On Mon, 7 Sep 1998, Andris Pavenis wrote:
> 
> > I didn't have any crashes (or other problems) with egcs-1.1 compiled using
> > slightly patched DJDEV202
> 
> Crashes are not the only problems you could have from djdev202, and your 
> system is only one of all possible combinations out there.  IMHO, 
> djdev202 is not tested enough yet to be used in production-quality 
> software.  A combination of a (possibly) buggy compiler with a (possibly) 
> buggy library used to build that compiler, and a (possibly) buggy code 
> you compile with that compiler is IMHO too much for many programmers, 
> even the adventurous ones ;-).

Of course I'm not thinking about crashes only. Of course some 
other tests could be Ok. For examples I think You could try to 
build EMACS (I'm not using it, so it would be difficult for me to test 
the results). Also some other large packages would be nice
(as I already said I have tried RHIDE and also my own programs -
about 2 Mb source in C++).
 
> > So I suggest that we can release it as it is without rebuilding. I
> > only need to modify archive gcc-e11b.zip to have support of djdev201.
> 
> It's up to you.  You asked for opinions and I gave you mine.
> 

I understand that it would be more safe to rebuild using djdev201.
Therefore I'm asking here to test it with available sources to see if 
there will be any serious problems.

Andris

- Raw text -