Re: Is the bugdatabase alive?

[Andrew Haley <aph at redhat dot com>]

Mark Wielaard writes:
 > Hi,
 > 
 > On Sun, 2003-02-16 at 20:10, ŽØyvind Harboe wrote:
 > > I submitted 9051 a while ago and as near as I can tell,
 > > the problem is gone in 3.3 (http://gcc.gnu.org/cgi-bin/gnatsweb.pl).
 > > 
 > > In ignorance of the proper procedure, I thought I'd post the
 > > information here.
 > 
 > Thanks for reminding. I know I looked at it but couldn't figure out what
 > was going on. It seems that giving -Os does some optimization that is
 > not working for gcj produced programs.

This isn't terribly surprising, as there are some known bugs in -Os
that generate incorrect code, and have been for some time.

 > Strangely I just get a
 > NullPointerException not an abnormal program termination. But I couldn't
 > debug it since adding -g makes it disappear.

That is supposed never to happen.  Simply turning on debugging is not
supposed to change the code.

 > (Will add that info to the bug now.)
 > 
 > > Whats the deal with the bugdatabase? I get the impression 
 > > that it is dusting in the corner. 
 > 
 > Some bugs are, but it is actively used to track the state of the
 > project.
 > 
 > > I got a grand total of 80 bugs when I listed all libgcj bugs,
 > > and few of the bugs where assigned.
 > 
 > Bugs only get assigned when someone is actually working on them.
 > 
 > > My impression is that gcj is maintained by various people making
 > > the changes they need and submitting patches, rather than a group
 > > of programmers fixing bugs that mortal users find.

Well, yes.  There isn't a benevolent patrician paying programmers to
work on fixing bugs that people send in.  However, from time to time I
have fixed bugs from the bug database and so have other people.

Andrew.