3
See also various low-level TODOs in the source code. Try looking in
4
the list archive or on gmane.org for previous discussion of these
7
These are classified by approximate size: an hour or less, a day or
8
less, and several days or more.
14
* Add of a file that was present in the base revision should put back
17
* Handle diff of files which do not have a trailing newline; probably
18
requires patching difflib to get it exactly right, or otherwise
19
calling out to GNU diff.
21
* Import ElementTree update patch.
23
* Syntax should be "bzr export -r REV".
27
* Plugins that provide commands. By just installing a file into some
28
directory (e.g. /usr/share/bzr/plugins) it should be possible to
29
create new top-level commands ("bzr frob"). Extensions can be
30
written in either Python (in which case they use the bzrlib API) or
31
in a separate process (in sh, C, whatever). It should be possible
32
to get help for plugin commands.
34
* Smart rewrap text in help messages to fit in $COLUMNS (or equivalent
37
* -r option should take a revision-id as well as a revno.
39
* "bzr info" could show space used by working tree, versioned files,
40
unknown and ignored files.
42
* "bzr info" should count only people with distinct email addresses as
43
different committers. (Or perhaps only distinct userids?)
45
* Tidier error for EPIPE: should be just "bzr: broken pipe" with no
46
other details because debugging information is rarely interesting.
48
* On Windows, command-line arguments should be glob-expanded__,
49
because the shell doesn't do this. However, there are probably some
50
commands where this shouldn't be done, such as 'bzr ignore', because
51
we want to accept globs.
53
__ http://mail.python.org/pipermail/python-list/2001-April/037847.html
55
* 'bzr ignore' command that just adds a line to the .bzrignore file
56
and makes it versioned.
58
* 'bzr help commands' should give a one-line summary of each command.
60
* Are any sanity checks useful in 'bzr ignore'?
65
* Display command grammar in help messages rather than hardcoding it.
67
* Change command functions into Command() objects, like in hct, and
68
then the grammar can be described directly in there. Since all
69
option definitions are global we can define them just once and
70
reference them from each command.
72
* Selective commit of only some files.
76
Status should be handled differently because it needs to report on
77
deleted and unknown files. diff only needs to deal with versioned
80
* Merge Aaron's merge code.
84
* Turn on stat cache code, and add optimization about avoiding
85
dangerous cache entries.
89
* More efficient diff of only selected files.
91
* Fix up Inventory objects to represent root object as an entry.
93
* Don't convert entire entry from
95
* Extract changes from one revision to the next to a text form
96
suitable for transmission over email.
100
* Write a reproducible benchmark, perhaps importing various kernel versions.
102
* Change test.sh from Bourne shell into something in pure Python so
103
that it can be more portable.
105
* Directly import diffs! It seems a bit redundant to need to rescan
106
the directory to work out what files diff added/deleted/changed when
107
all the information is there in the diff in the first place.
108
Getting the exact behaviour for added/deleted subdirectories etc
111
At the very least we could run diffstat over the diff, or perhaps
112
read the status output from patch. Just knowing which files might
113
be modified would be enough to guide the add and commit.
115
Given this we might be able to import patches at 1/second or better.
117
* Get branch over http.
119
* Pull pure updates over http.
121
* revfile compression.
123
* Split inventory into per-directory files.
125
* Fix ignore file parsing:
127
- fnmatch is not the same as unix patterns
129
- perhaps add extended globs from rsh/rsync
131
- perhaps a pattern that matches only directories or non-directories
133
* Expansion of $Id$ keywords within working files. Perhaps do this in
134
exports first as a simpler case because then we don't need to deal
135
with removing the tags on the way back in.
141
* Hooks for pre-commit, post-commit, etc.
143
Consider the security implications; probably should not enable hooks
144
for remotely-fetched branches by default.
146
* Pre-commit check. If this hook is defined, it needs to be handled
147
specially: create a temporary directory containing the tree as it
148
will be after the commit. This means excluding any ignored/unknown
149
files, and respecting selective commits. Run the pre-commit check
150
(e.g. compile and run test suite) in there.
154
* GUI (maybe in Python GTK+?)
156
* C library interface
added
removed
TODO
