What's outstanding

[Junio C Hamano <gitster@pobox.com>]

Here is the index of topics that I am aware of; some of them may not be
anywhere in git.git.


Regressions?

* fetch with "refs/*:refs/*" errors out erroneously
  http://article.gmane.org/gmane.comp.version-control.git/77335

I think tonight's patch would fix it; let's see how well it goes...

* "git fetch" shows error when dangling symref exists at the remote
  but does not really error out
  http://article.gmane.org/gmane.comp.version-control.git/76658

* "git fetch" does not exit with non-zero status when it failed to update
  some refs due to non-ffness
  http://article.gmane.org/gmane.comp.version-control.git/77178



New features that won't be merged before 1.5.5 final.

* ff=only (Sverre Hvammen Johansen)
  http://article.gmane.org/gmane.comp.version-control.git/78250

  With vastly improved documentation and justification compared to the
  previous rounds, I am beginning to actulaly like this series.

* more help backends (Christian Couder)
  http://article.gmane.org/gmane.comp.version-control.git/78150

  With the documentation clarification the overall idea looked fine.

* cvsserver updates (Damien Diederen with help by Frank Lichtenheld)
  http://article.gmane.org/gmane.comp.version-control.git/78188

  Looked cleanly done.

* use single database to serve more than one cvsserver instances (Josh Elsasser)
  http://article.gmane.org/gmane.comp.version-control.git/78366

  Looked cleanly done.

* rewriting annotated tags in filter-branch (Brandon Casey)
  http://article.gmane.org/gmane.comp.version-control.git/78286

  I haven't been heavily involved in this program, neither as a user nor a
  designer, so I am not absolutely sure about the details, but the idea
  looked clean and sane.

* git lost-found vs git fsck --lost-found (me)
  http://article.gmane.org/gmane.comp.version-control.git/78267

* core.inithook (Dscho)
  http://article.gmane.org/gmane.comp.version-control.git/78123

  This would hopefully give a cleaner way for mingw folks to futz with
  core.autocrlf without breaking existing setups.

* receive.localBranches = { refuse | allow } (Dscho)
  http://article.gmane.org/gmane.comp.version-control.git/78065

  The infamous "do not push into a repository with work tree".

* autosetting core.ignorecase (Dmitry Potapov)
  http://article.gmane.org/gmane.comp.version-control.git/78176

  Follow-up to Linus's "case insensitive" git

* mailinfo extention to extract Message-ID and others (Anton Gladkov)
  http://article.gmane.org/gmane.comp.version-control.git/78006

* add -p with mode change (Jeff King)
  http://article.gmane.org/gmane.comp.version-control.git/78340

* rebase -p
  http://article.gmane.org/gmane.comp.version-control.git/78074

  I highly suspect that this is much deeper than just the implementation.

* synopsys: use {} instead of () for grouping alternatives (Jari Aalto)
  http://article.gmane.org/gmane.comp.version-control.git/72243

  I sent an invitation for this topic earlier, but we are at -rc2 and it
  is too late for 1.5.5 now.

* "[alias] st = status" and "cd .git && git st" (Jeff King)
  http://article.gmane.org/gmane.comp.version-control.git/72327

* use "assume unchanged" bit to implement narrow checkout
  http://article.gmane.org/gmane.comp.version-control.git/77046

* send-pack vs receive-pack history exchange (me)
  http://article.gmane.org/gmane.comp.version-control.git/77643

* zlib abstraction (Marco)
  http://article.gmane.org/gmane.comp.version-control.git/72262

* git --index-file=<foo> <cmd> (Linus)
  http://article.gmane.org/gmane.comp.version-control.git/77332