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
9 What to expect from now on
10 ==========================
12 This is written in a form of to-do list for me, so if I say
13 "accept patch", it means I do not currently plan to do that
14 myself. People interested in seeing it materialize please take
15 a hint. Also whatever I marked "Perhaps" do not have to happen
16 if ever -- only if somebody cares enough and submits a clean
23 * tree entries in index? -- sorry, stalled
25 * "intent to add" index entries? -- together with the above
28 * Plug-in file-level merges. On the other hand, we may not even
29 need this; just tell people to run "xxdiff -U" on the working
30 tree files (or kompare).
36 * Libification. There are many places "run once" mentality is
37 ingrained in the management of basic data structures, which
38 need to be fixed. [Matthias Urlichs was already working on
39 this: <pan.2005.10.03.20.48.52.132570@smurf.noris.de>, but I
40 do not know what happened to his efforts]
42 * Lazy clones that can be controlled by the user, ranging from
43 totally on-demand a la CVS/SVN to "cache down to this old
44 commit so that I can make full use of git on at least recent
45 history". This need a lot of work in making tools to exit
46 gracefully when they hit unavailable objects while offline.
52 * git grep should be able to omit prefix when invoked from a
53 subdirectory. Either make it an option or default with
54 --full-name like ls-files does.
56 * duplicated refspec given to "fetch-pack a a a" makes it emit
57 strange error message because it triggers the "match only
58 once" logic. Maybe strip the dups on the input side
60 <20060608073857.GA5072@informatik.uni-freiburg.de>).
62 * map only parts of huge packfiles and LRU fragments of them.
63 People are starting to try git on projects with deep history
64 and/or many objects, e.g. Mozilla.
66 * upload-pack support for start fetching from any valid point on
67 the history, not just published refs. (Erik W. Biederman
68 <m164jc9ekx.fsf@ebiederm.dsl.xmission.com>)
70 * git-daemon side support for virtual hosting. Client side
71 is ready in 1.4.0 (Jon Loeliger <1149610100.23938.75.camel@cashmere.sps.mot.com>).
73 * teach git-upload-pack not to ack-continue early when the
74 client has roots it does not know about but it already has
75 learned the fork points for all the requested heads
76 (Ralf Baechle <20060524131022.GA11449@linux-mips.org>).
78 * Per user .gitconfig across repositories -- ongoing.
80 * Encourage competition between annotate vs blame. Maybe come
81 up with some nontrivial test cases.
83 * Subprojects. Try "gitlink" -- sorry, stalled.
85 * Rebase and checkout -m should be able to use recursive
86 strategy as well. These commands currently do not work across
89 * Decide what to do about rebase applied to merged head. One
90 extreme is to allow rebase if "rev-list ours..theirs" gives
91 anything. This loosens the current merge-base based approach.
92 The other extreme is to refuse rebase if "rev-list
93 theirs..ours" contains any merge commit, which was discussed
96 <43CC695E.2020506@codeweavers.com>
98 * Decide what the right thing to do upon an empty merge commit,
99 when both branches happen to have obtained the same set of
100 changes through different history. Not recording such keeps
101 the history simpler, and the next merge would soon create a
102 true merge commit anyway, but this does not feel quite right.
104 <20060114021800.4688.qmail@web31803.mail.mud.yahoo.com>
106 * diff stopping at the first output; qgit wants to know if this
107 tree has any A or D from the other tree and nothing else.
108 Would help internal tree-diff in rev-list as well.
110 * daemon --strict-symlink.
112 * daemon --base-path does not apply automatically to whitelist
113 somehow feels wrong. If somebody cares enough, accept
116 * Perhaps detect cloning request in upload-pack and cache the
117 result for next cloning request until any of our refs change.
119 * Maybe grok PGP signed text/plain in applymbox as well.
125 * git-proxy should be spawned with sh -c 'command' $1 $2.
127 * Maybe a true git-proxy command that reads the first request
128 pkt-line, and redirects the request to its real destination.
130 * test scripts for the relative directory path stuff.
132 * In a freshly created empty repository, `git fetch foo:bar`
133 works OK, but `git checkout bar` afterwards does not (missing