2c about the development of ubuntu

[Ming Hua]

On Tue, Jan 03, 2006 at 07:49:08AM +0100, Ante Karamatic wrote:
> On Mon, 2006-01-02 at 23:20 +0100, Loïc Martin wrote:
> 
> > Sorry, but as an example scim bugs were reported about 2 weeks (at 
> 
> I was looking at that scim bug. It's an upstream error. This bug will be
> in Etch too (and every other newer distribution), if upstream doesn't do
> new version. IIRC, it doesn't compile well with gcc4. And, instaling
> scim breaks lot's of other programs. This is a known bug. And don't
> think we forgot it.

Hi Ante Karamatic,

I don't mean to criticize your comments, and I appreciate your looking
at the scim bug, but I think you got some facts wrong and I need to tell
my view on this issue.

There are at least two known bug related to the "scim in breezy causes
crashes" issue, and strictly speaking, neither of them is scim's fault.
One is about g++/libstdc++ having unversioned weak symbols, filed as
Debian #323216/#342198 and malone #2246.  The other is a bug in uim
which causes scim-setup to crash at startup time, filed as malone #5896.

Both bugs are clearly analyzed, reported to upstream (gcc and uim) and
accepted, and have workarounds.  The latter is already fixed in dapper
according to my test.  There definitely may be other bugs that causes
crashes, but nobody has taken the time to analyze them, or at least
provide a clear description to reproduce them yet.

Ming, Debian scim maintainer
2006.01.03