gcc

[Fredriksson, Andreas]

Stefan (and the list),

I'm currently using Gentoo linux w/ gcc 3.2.2 for development.
We will examine the compatibility with various distributions
before the release and make a fully static version available if
it proves necessary.

As for CPU optimizations, we will probably go for a -march=i686
version, as it's easier to only handle one version at a time
and you need atleast a P3 to run the server at all.

Regards,
Andreas

-----Original Message-----
From: Stefan Siefert
To: andreas.fredriksson at dice.se
Sent: 4/2/2003 11:33 AM

Hi,

I have some questions regarding the gcc version. First of all, which gcc
version do you use exactly (cause of the change of abi between several
gcc 3
versions)? Is it possible to distribute some Processor optimized BF
(Server)
versions (we would be eager to test those for stability and performance
issues)?

Greetings,
Stefan Siefert
www.krawall.de
www.k-play.de

> -----Ursprüngliche Nachricht-----
> Von: Fredriksson, Andreas [mailto:andreas.fredriksson at dice.se]
> Gesendet: Montag, 31. März 2003 15:20
> An: 'bf1942 at icculus.org'
> Betreff: [bf1942] Linux server status report: 2003-03-31
>
>
>
> Hello people, here's the status update for last week.
>
> - Win32 and Linux code has been merged into a common codebase. This
>   has the side effect that we cannot instantly release a new
> 1.31 patch
>   since the head revision of our code contains changes not
> yet released
>   for win32.
>
>   We are considering a backport of all our fixes to 1.31 but I can't
>   comment on the status of that right now until management has had a
>   time to look at it. I'll keep you posted.
>
> - The Linux server is compiled with gcc3 now, and this has uncovered
>   some pretty nasty bugs that were masked out by gcc2/msvc. I'm still
>   sorting out library issues and how we are going to distribute the
>   package. Would a completely static version be acceptable for people
>   with older distributions?
>
> - The server now uses the same threaded UDP I/O subsystem as the win32
>   server but does so without increasing the thread priority
> of the listener
>   thread. We haven't had any problems with this yet but it
> might not work
>   out for some slower or very busy machines. We believe this
> is a good change
>   in the long run though.
>
> - We have not been able to reproduce the Bungee effect; it
> might be related
>   to the single threaded I/O used in the earlier betas.
>
> I'll keep you posted with information on upcoming releases
> and status reports.
>
> Regards,
> Andreas
>
>