Re: jack still broken

2005-09-23 Thread Henrique de Moraes Holschuh
On Fri, 23 Sep 2005, Thomas Bushnell BSG wrote: > > Why has it taken three months for the reverse dependencies to be rebuilt? Because of the gcc 4 transition and C++ transition, which screwed us all over since we needed to wait for the full dep-chain to be transitioned, and a lot of libs took thei

Re: jack still broken

2005-09-23 Thread Adeodato Simó
* Thomas Bushnell BSG [Fri, 23 Sep 2005 10:14:47 -0700]: > AFAICT there weren't bugs filed against those packages telling them to > be rebuilt, because the libjack people believe their libraries don't > conflict. http://people.debian.org/~adeodato/transition/libjack0.80.0-0 (And those are n

Re: jack still broken

2005-09-23 Thread Adam Heath
On Fri, 23 Sep 2005, Thomas Bushnell BSG wrote: > > No, you design the protocol from the beginning to have the first token be a > > protocol version, followed by a magic number(which could be different per > > version). > > Presumably it's too late for that. :) Stupid programmers. -- To UNSUBS

Re: jack still broken

2005-09-23 Thread Steve Langasek
On Fri, Sep 23, 2005 at 10:14:47AM -0700, Thomas Bushnell BSG wrote: > >> This has been going on for three months. > > Why has it taken three months for the reverse dependencies to be rebuilt? > > That shouldn't really be acceptable, whether or not there's some reason for > > (effective) conflicts

Re: jack still broken

2005-09-23 Thread Thomas Bushnell BSG
Steve Langasek <[EMAIL PROTECTED]> writes: > On Fri, Sep 23, 2005 at 12:38:24PM -0500, Adam Heath wrote: >> On Fri, 23 Sep 2005, Thomas Bushnell BSG wrote: > >> > Steve Langasek <[EMAIL PROTECTED]> writes: > >> > > How else do you handle the case of a library which implements a line >> > > protoco

Re: jack still broken

2005-09-23 Thread Thomas Bushnell BSG
Adam Heath <[EMAIL PROTECTED]> writes: > On Fri, 23 Sep 2005, Thomas Bushnell BSG wrote: > >> Steve Langasek <[EMAIL PROTECTED]> writes: >> >> > How else do you handle the case of a library which implements a line >> > protocol (presumably Unix sockets, in this case), and that protocol has >> > ch

Re: jack still broken

2005-09-23 Thread Steve Langasek
On Fri, Sep 23, 2005 at 12:38:24PM -0500, Adam Heath wrote: > On Fri, 23 Sep 2005, Thomas Bushnell BSG wrote: > > Steve Langasek <[EMAIL PROTECTED]> writes: > > > How else do you handle the case of a library which implements a line > > > protocol (presumably Unix sockets, in this case), and that

Re: jack still broken

2005-09-23 Thread Adam Heath
On Fri, 23 Sep 2005, Thomas Bushnell BSG wrote: > Steve Langasek <[EMAIL PROTECTED]> writes: > > > How else do you handle the case of a library which implements a line > > protocol (presumably Unix sockets, in this case), and that protocol has > > changed incompatibly? > > You implement the new pr

Re: jack still broken

2005-09-23 Thread Thomas Bushnell BSG
Steve Langasek <[EMAIL PROTECTED]> writes: > How else do you handle the case of a library which implements a line > protocol (presumably Unix sockets, in this case), and that protocol has > changed incompatibly? You implement the new protocol on a new port, so that both the old and new jackd can

Re: jack still broken

2005-09-22 Thread Steve Langasek
On Thu, Sep 22, 2005 at 12:00:22AM -0700, Thomas Bushnell BSG wrote: > Jack has been broken for three months now. > The problem is that the libraries libjack0.80.0-0 and libjack0.100.0-0 > declare exact version requirements against jackd: both of them > conflict with all but one version of jackd.

jack still broken

2005-09-22 Thread Thomas Bushnell BSG
Jack has been broken for three months now. The problem is that the libraries libjack0.80.0-0 and libjack0.100.0-0 declare exact version requirements against jackd: both of them conflict with all but one version of jackd. If they really are proper library packages, then they shouldn't conflict wi