On Tue, May 10, 2005 at 11:10:10AM +0200, Rene Mayrhofer wrote: > Steve Langasek schrieb: > >>If that 2.3.x bug really only affects the newer (> 2.6.8) kernel, why > >>not just get 2.3.x pushed into sarge? Are there any other big issues > >>with it, that weren't in 2.2.x? Some people might certainly like the > >>agressive mode support, or 2.3.1's NAT-T fixes. Personally, 2.2.x is > >>fine for me though --- anything but 2.1.x for me :-) > Mainly because 2.3.x causes other openswan boxes to crash in some > (reproducable) cases - that's a pretty bad regression from 2.2.0 and I > keep bugging upstream with it for at least 3 months. No fix until now, > so we can't wait until it will be fixed. I would vote for 2.2.0-4. (or > even 2.2.0-5). > > Because 2.2.3 is no longer in the archive, and resurrecting new binaries via > > t-p-u gives us even less than the usual protection against breakage caused > > by a lack of testing. :/ > Does that mean that the only way to get the known stable 2.2.0-x back > into testing is an upload to unstable with an epoch? I really wouldn't > like to go that route if I can avoid it.... Well, AFAIK openswan has never actually been in testing /before/, so it's not a matter of getting it /back/; but yes, the upshot is that I'm not comfortable adding packages to testing via t-p-u. -- Steve Langasek postmodern programmer
Attachment:
signature.asc
Description: Digital signature