Make rpm spec less explicit Fixes: * current buildrpm failing on Fedora because of unexpected zeroconf extensions * unowned directories hgext/*/

Merge with crew

Catch both IOError and OSError, fix regression introduced by 8046f0a070a6

inotify: remove unused imports (thanks pyflakes)

color: don't run status twice for -n Just cut off the status characters.

color: fix status -n

keyword: do not shadow builtin format (detected by pychecker)

hooks: restore io correctly on exception

protocol: use changegroupsubset() if possible (issue1389) Due to the fix to the pull race, to avoid sending unnecessary changesets, use changegroupsubset if possible. This will increase the load on the server.

Merge with crew

email: add References field in the header This hopefully helps clients get the threading right.

coal/paper: link to shortlog instead of log from error pages

coal/paper: move actual templates from coal to paper

coal/paper: show branch name in changeset view

coal/paper: display tags and branch in filelog page

coal/paper/monoblue: show branch names in hgweb graph

coal/paper: display branch names in log

hgweb: support broken IIS 5 behavior with .cgi in PATH_INFO

manifest: fix _search() corner-case It failed when searching the empty string in the null revision.

Improved error message for log --follow Mads Kiilerich notes that the existing message was unclear when a nonexistent file was supplied.

Improve the monoblue theme for hgwebdir

import: add similarity option (issue295)

mq: recover more gracefully from interrupted qrefresh (issue1216)

mq: use atomictempfiles during patch refresh

mq: create patch header class to abstract header manipulation

mq: gracefully abort qpush/qgoto to guarded patch (issue1186)

More verbose logging when filemerge searches for merge-tool Previously it was very hard to find out what was going on when the expected merge tool wasn't used. This patch tries to improve that. hg merge -v now shows which tools were searched for but not found.

templater: return data in increasing chunk sizes Currently hgweb is not streaming its output -- it accumulates the entire response before sending it. This patch restores streaming behaviour. To avoid having to synchronously write many tiny fragments, this patch also adds buffering to the template generator. Local testing of a fetch of a 100,000 line file with wget produces a slight slowdown overall (up from 6.5 seconds to 7.2 seconds), but instead of waiting 6 seconds for headers to arrive, output begins immediately.