4 The latest copy of this document is found at
6 http://kernel.org/git/?p=git/git.git;a=blob;hb=todo;f=TODO
12 - All non-binary commands will lose -script suffix in
13 $(bindir). The source to git-foo will be either git-foo.sh
14 or git-foo.perl in the source tree, and the documentation
15 will be in Documentation/git-foo.txt.
17 - The commands whose names have 'cache' to mean 'index file'
18 will get 'cache' in their names replaced with 'index'. For
19 git-fsck-cache and git-convert-cache, 'cache' will be
20 replaced with 'objects'.
22 - The commit walkers will have 'pull' in their names replaced
23 with 'fetch'. 'git-ssh-push' will become 'git-ssh-upload'.
25 - We continue to follow the convention to name the C source
26 file that contains the main program of 'git-foo' command
27 'foo.c'. That means we will have 'fsck-objects.c', for
30 - At this moment, I am not planning to rename the symbols used
31 in programs, nor any library sources. "cache.h" will stay
32 "cache.h", so does "read-cache.c". "struct cache_entry" and
33 "ce_match_stat()" will keep their names. We _might_ want to
34 rename them in later rounds but not right now.
36 - In 0.99.7, all renamed commands will have symbolic links in
37 $(bindir) so that old names continue to work. These backward
38 compatible names will not appear in documentation. The main
39 documentation, git(7) will talk about the new names but would
40 mention their old names as historical notes. Old environment
41 names defined in gitenv() will also be removed in this release.
43 - In 0.99.8, we do not install these backward compatible
44 symbolic links in $(bindir) anymore. The Makefile will have
45 a target to remove old symlinks from $(DESTDIR)$(bindir) you
46 can run manually to help you clean things up.
48 The timeframe for this is around Oct 1st, but I could be
49 talked into delaying the symlink removal if Porcelain people
50 find this schedule too tight.
53 What to expect after 0.99.7
54 ===========================
56 This is written in a form of to-do list for me, so if I say
57 "accept patch", it means I do not currently plan to do that
58 myself. People interested in seeing it materialize please take
65 * Accept patches from people who actually have done CVS
66 migration and update the cvs-migration documentation.
67 Link the documentation from the main git.txt page.
69 * Accept patches from people who were hit by shiny blue bat to
70 update the SubmittingPatches.
72 * Talk about using rsync just once at the beginning when
73 initializing a remote repository so that local packs do not
74 need to be expanded. I personally do not think we need tool
75 support for this (but see below about optimized cloning).
77 * Maybe update tutorial with a toy project that involves two or
80 * Update tutorial to cover setting up repository hooks to do
83 * Accept patches to finish missing docs.
89 * Tony Luck reported an unfortunate glitch in the 3-way merge.
90 Encourage discussions to come up with a not-so-expensive way
91 to catch the kind of ambiguities that led to his misery.
92 [Deathmatch between Daniel's and Fredrik's ongoing.]
94 * HPA has two projects, klibc and klibc-kbuild, that have large
95 set of overlapping files in different paths (i.e. one has many
96 renames from the other). There currently is no way for git to
97 help keep these two trees in sync, merging criss-cross between
98 them. The merge logic should be able to take advantage of
99 rename/copy detection smarts git-diff-* family has. Linus,
100 me, and Daniel outlined a smarter merge strategy for this.
103 * We might want to optimize cloning with GIT native transport
104 not to explode the pack, and store it in objects/pack instead.
105 We would need a tool to generate an idx file out of a pack
106 file for this. Also this itself may turn out to be a bad
107 idea, making the set of packs in repositories everybody has
108 different from each other.
110 * Libification. There are many places "run once" mentality is
111 ingrained in the management of basic data structures, which
114 * Maybe a pack optimizer.
116 * Maybe an Emacs VC backend.
122 * The recent commit walker safety patch may be too cautious and
123 appears to take forever when cloning. This may even be
124 infinitely looping in the code lifted from the old rev-list --
125 needs to be taken a look at [DONE INITIAL CUT].
127 * Encourage concrete proposals to commit log message templates
128 we discussed some time ago.
130 * Accept patches for more portability.
132 * strcasestr() in mailinfo. We may need compat/strcasestr.c;
133 this is bugging OpenBSD folks.
135 * Accept patches to cause "read-tree -u" delete a directory when
138 * Perhaps accept patches to introduce the concept of "patch flow
139 expressed as ref mappings" Josef has been advocating about.
141 * Perhaps accept patches to do undo/redo.
143 * Perhaps accept patch to optionally allow '--fuzz' in
146 * Allow 'git apply' to accept GNU diff 2.7 output that forgets
147 to say '\No newline' if both input ends with incomplete
150 * Maybe grok PGP signed text/plain in applymbox as well.
152 * Perhaps a tool to revert a single file to pre-modification
153 state? git-cat-file blob `git-ls-files | grep foo` >foo or
154 git-cat-file blob `git-ls-tree HEAD foo` >foo? What should
155 the command be called? git-revert is taken so is
158 * Enhance "git repack" to not always use --all; this would be
159 handy if the repository contains wagging heads like "pu" in
162 * Internally split the project into non-doc and doc parts; add
163 an extra root for the doc part and merge from it; move the
164 internal doc source to a separate repository, like the +Meta
165 repository; experiment if this results in a reasonable
166 workflow, and document it in howto form if it does.
168 * Make rebase restartable; instead of skipping what cannot be
169 automatically forward ported, leave the conflicts in the work
170 tree, have the user resolve it, and then restart from where it
173 * Output full path in the "git-rev-list --objects" output, not
174 just the basename, and see the improved clustering results in
175 better packing [Tried, but did not work out well].
177 * Remove obsolete commands [READY].
179 * Option to limit rename detection for more than N paths [READY].
181 * Option to show only status and name from diff [READY].
187 * 'git add --recursive'?
189 * 'git merge-projects'?
191 * 'git lost-and-found'? Link dangling commits found by
192 fsck-objects under $GIT_DIR/refs/lost-found/. Then
193 show-branch or gitk can be used to find any lost commit. [A
194 feeler patch sent out. Very underwhelming response X-<.]
196 Do not name it /lost+found/; that would probably confuse
197 things that mistake it a mount point (not our code but
200 * Add simple globbing rules to git-show-branch so that I can
201 say 'git show-branch --heads "ko-*"' (ko-master, ko-pu, and
202 ko-rc are in refs/tags/).
204 * We would want test scripts for the relative directory path
205 stuff Linus has been working on. So far, the following
206 commands should be usable with relative directory paths:
216 * In a freashly created empty repository, `git fetch foo:bar`
217 works OK, but `git checkout bar` afterwards does not (missing