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 - In 0.99.8, we will still install the backward compatible
13 symbolic links in $(bindir). These will however be removed
16 git-ssh-push and git-ssh-pull pair is not going away within
17 this timeframe, if ever. Each of these old-name commands
18 continues to invoke its old-name counterpart on the other
22 What to expect after 0.99.8
23 ===========================
25 This is written in a form of to-do list for me, so if I say
26 "accept patch", it means I do not currently plan to do that
27 myself. People interested in seeing it materialize please take
28 a hint. Also whatever I marked "Perhaps" do not have to happen
29 if ever -- only if somebody cares enough and submits a clean
36 * Document the ref naming restrictions.
38 * Help Jon Loeliger to find place in the documentation to place
41 * Accept patches from people who actually have done CVS
42 migration and update the cvs-migration documentation.
43 Link the documentation from the main git.txt page.
45 * Talk about using rsync just once at the beginning when
46 initializing a remote repository so that local packs do not
47 need to be expanded. I personally do not think we need tool
48 support for this (but see below about optimized cloning).
50 * Maybe update tutorial with a toy project that involves two or
53 * Update tutorial to cover setting up repository hooks to do
56 * Accept patches to finish missing docs.
58 * Accept patches to talk about "Whoops, it broke. What's
61 * Accept patches to make formatted tables in asciidoc to work
62 well in both html and man pages (see git-diff(1)).
68 * We might want to optimize cloning with GIT native transport
69 not to explode the pack, and store it in objects/pack instead.
70 We would need a tool to generate an idx file out of a pack
71 file for this. Also this itself may turn out to be a bad
72 idea, making the set of packs in repositories everybody has
73 different from each other. [DONE; git-index-pack by Sergey,
76 * Git daemon, when deployed at kernel.org, might turn out to be
77 quite a burden, since it needs to generate customized packs
78 every time a new request comes in. It may be worthwhile to
79 precompute some packs for popular sets of heads downloaders
80 have and serve that, even if that could give more than the
81 client asks for in some cases. We will know about this soon
84 * Libification. There are many places "run once" mentality is
85 ingrained in the management of basic data structures, which
86 need to be fixed. [Matthias Urlichs is already working on
87 this: <pan.2005.10.03.20.48.52.132570@smurf.noris.de>; Post
90 * Maybe a pack optimizer.
92 Given a set of objects and a set of refs (probably a handful
93 branch heads and point release tags), find a set of packs to
94 allow reasonably minimum download for all of these classes of
95 people: (1) somebody cloning the repository from scratch, (2)
96 somebody who tends to follow the master branch head reasonably
97 closely, (3) somebody who tends to follow only the point
100 * Maybe an Emacs VC backend.
102 * 'git split-projects'? This requires updated 'git-rev-list' to
103 skip irrelevant commits.
104 Message-ID: <Pine.LNX.4.63.0509221617300.23242@iabervon.org>
106 * Look at libified GNU diff CVS seems to use, or libxdiff.
107 [Daniel has his own diff tool almost ready to start
108 integrating and testing; Post 1.0]
110 * Accept patches to fetch multiple objects by HTTP in parallel.
113 * Plug-in file-level merges [Post 1.0].
115 * Per-repository configuration mechanism [DONE by Linus].
121 * Decide the notation of "peeling the onion" operator, and
122 implement it in sha1_name.c. Perhaps postfix "^{}" to mean
123 "peel and expect anything", "^{blob}" to mean "peel and barf
124 unless blob". The current "^0" becomes shorthand for
127 * Quote the URL so that libcurl's metecharacter mechanism would
130 * Review the Makefile variables and exporting rules for them,
131 while looking at prefix passing by Kai Ruemmler [DONE].
133 * Review the 'sparse object database' change by Linus and move
134 the first phase of it to the "master" branch [DONE].
136 * Decide on mmap(). I am inclined to just stick to mmap
137 replacement by Johannes Schindelin and do nothing else right
138 now, except perhaps drop the writing-back support [DONE].
140 * Revisit Santi's patch to move commit temorary files out of the
141 working tree toplevel [DONE].
143 * More generally, review the use of temporary files again.
144 Assuming writable $GIT_DIR is more acceptable, but the
145 working tree toplevel may not be in a rare usage pattern.
147 * Perhaps accept more "want"s in upload-pack and do something
148 intelligent about it.
150 * Perhaps detect cloning request in upload-pack and cache the
151 result for next cloning request until any of our refs change.
153 * Perhaps send less "want"s from fetch-pack.
155 * Encourage concrete proposals to commit log message templates
156 we discussed some time ago.
158 * Accept patches to cause "read-tree -u" delete a directory when
159 it makes it empty [DONE].
161 * Perhaps accept patches to do undo/redo.
163 * Perhaps accept patch to optionally allow '--fuzz' in
166 * Allow 'git apply' to accept GNU diff 2.7 output that forgets
167 to say '\No newline' if both input ends with incomplete
170 * What to do with TABs and LFs in pathnames without breaking GNU
171 patch? [DONE -- go with GNU patch extension proposed by Paul].
173 * Adjust apply.c to proposed GNU patch extension that quotes \n
174 and \t in C style, inside "". [DONE]
176 * Adjust diff.c to the same. [DONE]
178 * Maybe grok PGP signed text/plain in applymbox as well.
180 * Perhaps a tool to revert a single file to pre-modification
181 state? People with BK background know this operation as
182 'clean'. 'git checkout [-f] ent [path...]' was suggested by
183 Matthias Urlichs which sounds a natural extention to what the
184 command currently does.
186 * Enhance "git repack" to not always use --all; this would be
187 handy if the repository contains wagging heads like "pu" in
190 * Internally split the project into non-doc and doc parts; add
191 an extra root for the doc part and merge from it; move the
192 internal doc source to a separate repository, like the +Meta
193 repository; experiment if this results in a reasonable
194 workflow, and document it in howto form if it does.
196 The point is to make it possible to fork that part off to
197 somebody else; then I do not have to maintain Documentation
198 directory myself anymore, just like I simply slurp the latest
199 gitk from Paul and not worry about it ;-).
201 * Make rebase restartable; instead of skipping what cannot be
202 automatically forward ported, leave the conflicts in the work
203 tree, have the user resolve it, and then restart from where it
204 left off [mechanism mostly done].
206 * Output full path in the "git-rev-list --objects" output, not
207 just the basename, and see the improved clustering results in
208 better packing [Tried, but did not work out well].
210 * Updated git-changes-script Jeff Garzik needs [Inquiry for
211 external spec sent out with a quick hack. Will know if that
212 is what he needs hopefully soon].
214 * An mechanism to ignore filesystem mode bits altogether [DONE].
220 * Disallow [\001-\040\177] byte values from ref names. Also we
221 need to disallow ':' (used in refspec), '^' and '~' (postfix
222 "peel the onion" operators), and '..' ("ref1..ref2" notation
223 becomes ambiguous otherwise) [DONE].
225 * Update fetch-pack and clone-pack to ignore funny refs from the
226 other end, while making sure peek-remote does not discard them.
229 * Update upload-pack to send tag^{}. This would hopefully help
230 Pasky's automated tag tracking, and also Martin's findtags.
233 * short SHA1 naming is not enforcing uniqueness. Should fix [DONE].
235 * 'git repack' can be DOSed. Should fix [DONE].
237 * Stop installing the old-name symlinks [POSTPONED, but before 1.0].
239 * 'git merge-projects'?
241 Subject: Re: Merges without bases
242 References: <1125004228.4110.20.camel@localhost.localdomain>
243 Date: Thu, 25 Aug 2005 15:26:36 -0700
244 Message-ID: <7vvf1tps9v.fsf@assigned-by-dhcp.cox.net>
246 * 'git lost-and-found'? Link dangling commits found by
247 fsck-objects under $GIT_DIR/refs/lost-found/. Then
248 show-branch or gitk can be used to find any lost commit. [A
249 feeler patch sent out. Very underwhelming response X-<.]
251 Do not name it /lost+found/; that would probably confuse
252 things that mistake it a mount point (not our code but
255 * Add simple globbing rules to git-show-branch so that I can
256 say 'git show-branch --heads "ko-*"' (ko-master, ko-pu, and
257 ko-rc are in refs/tags/).
259 * We would want test scripts for the relative directory path
260 stuff Linus has been working on. So far, the following
261 commands should be usable with relative directory paths:
271 * In a freashly created empty repository, `git fetch foo:bar`
272 works OK, but `git checkout bar` afterwards does not (missing