On Thu, 19 Aug 2004 14:04:06 +0200, "Eduard Bloch" <[EMAIL PROTECTED]> said: > #include <hallo.h> > * Gertjan van Zwieten [Thu, Aug 19 2004, 12:01:12PM]: > > > I admit I was not aware of this perl one-liner you're > > referring to. I > > Such as... for x in *.deb; do perl -e 'read(STDIN,$a,2000); > $a=~s,^(.*?)\r?\n\r?\n,,ms; print $a, <STDIN>' < $x > pure/$x; done > > > didn't find any of this on the website, which by the way seems to be > > currently offline. Still, even with such script some problems > > remain. For instance apt-cacher's cache can't be easily turned into > > an offline cache, which can be useful when doing a fresh install. > > With replicator it's as simple as generating a local Packages.gz > > with dpkg-scanpackages. > > Okay. But I dont's see it as a disadvantage, and even then it would be > relatively easy to modify apt-cacher to use separate files for package > contents and HTTP headers.
Sure, apt-cacher can be modified to do this. But what do you conclude from that? I mean, I can see it could be a reason to not start another project, but http-replicator is already finished... Gertjan
