On 37-01--10 02:59 PM, wrote: > Le jeudi 22 septembre 2011 à 17:59 +0100, Iain Lane a écrit : >> I've been working recently on upgrading libproxy to 0.47, the latest >> upstream. > > Seriously, don’t do that. > > The 0.3 series was already broken, and the 0.4 one is even worse. They > reimplemented again a non-standards-compliant HTTP parser, instead of > using an existing one. This library as a whole is a bucket of fail. > > We need to rip the interesting parts from it and put them in > glib-networking, using libsoup for the HTTP parts and libjscore for JS > parsing. Any approach that keeps libproxy in the archive is doomed to > produce gazillions of bugs.
And so, while we wait for perfection, instead of having functional we accept completely broken? And we prefer completely broken to functional simply because functional is not "perfect"? Seems to me that completely broken is even further from perfect though. Instead of being an idealist and standing in the way of users having a functional system, why not be a pragmatist and at least let users (who don't give a rat's ass for somebody's particular "ideals" of perfection, BTW) have something that works instead of something that breaks? b.
signature.asc
Description: OpenPGP digital signature
