Kevin Smith <[EMAIL PROTECTED]> wrote: > Walter Landry wrote: > > I have another question. Does mercurial really support plain-jane > > HTTP servers? All of the docs mention setting up a cgi script or > > running "hg serve", and my limited tests didn't work. > > > > I have been told repeatedly that it does. However, my understanding is > that you have to specify the old http protocol, so a URL starting with > "http://" won't work. > > http://www.selenic.com/pipermail/mercurial/2005-June/001620.html
It does not seem to be working. [EMAIL PROTECTED]:~/arx/other_vc$ hg clone old-http://www.selenic.com/hg/ hg_http requesting all changes adding changesets adding manifests adding file revisions modified 0 files, added 0 changesets and 0 new revisions For some reason it does not download everything. Using the plain "http" url works fine. > > The reason this comes up is that I was thinking of how to optimize > > network traffic with the new hashes for revisions format, and I wanted > > to see what mercurial does. > > Ah. I know the git folks had some fancy stuff going before they threw it > all out and did packing. I don't think they've put fast http pulling > back in yet. My memory was that they were talking about parallel pulls, > and some http 1.1 (?) feature where you could request multiple pages > with one get command? I wan't paying close attention so I could be all > wrong. > > Do you think plain http pulls would be any slower with the rev-hashes > than what we have today? If not, I don't think it's worth worrying > about. If so, I don't think it's worth worrying about yet :-) No slower than what we have now, but right now it is too slow. Fixing it requires futzing with the archive format, so I want to bundle all of the format changes together. Cheers, Walter _______________________________________________ Arx-users mailing list [email protected] http://lists.nongnu.org/mailman/listinfo/arx-users
