1 Fix the things that are causing the most pain to new users
2 ----------------------------------------------------------
3 1. A new import is tagging all messages as "inbox" -- total pain
5 2. Allow an easy way to get tags from directory names (if the user has them)
7 Emacs interface (notmuch.el)
8 ----------------------------
9 Enhance '+' and '-' in the search view to operate on an entire region
12 Fix '*' to work by simply calling '+' or '-' on a region consisting of
15 Add a global keybinding table for notmuch, and then view-specific
16 tables that add to it.
18 Add a '|' binding from the search view.
20 Add support for choosing from one of the user's configured email
21 addresses for the From line.
23 Make 'notmuch-show-pipe-message have a private history.
25 Add support for a delete keybinding that adds a "deleted" tag to the
26 current message/thread and make searches not return deleted messages
27 by default, (unless the user asks explicitly for deleted messags in
30 Add support to "mute" a thread (add a "muted" tag and then don't
31 display threads in searches by default where any message of the thread
34 Fix i-search to open up invisible citations as necessary.
36 Make '=' count from the end rather than from the beginning if more
37 than half-way through the buffer.
39 Fix to automatically wrap long headers (for RFC compliance) before
40 sending. This should probably just be fixed in message-mode itself,
41 (but perhaps we can have a notmuch-message-mode that layers this on
44 Implement Fcc and use it for all messages, (whether a new composition,
45 a reply, or a forward). This again may require a notmuch-message-mode
46 that extends message-mode.
48 Stop hiding the headers so much in the thread-view mode.
50 Allow opening a message in thread-view mode by clicking on either
53 Automatically open a message when navigating to it with N or P.
55 Change 'a' command in thread-view mode to only archive open messages.
57 Add a binding to open all closed messages.
59 Make all authors and subjects available to isearch, (hidden by default
60 but with magic expansion while isearching).
62 Fix notmuch-hello as follows:
64 1. Change the "notmuch" and message count in the welcome sentence to
67 2. Put the saved searches (if any) before the search bar.
69 3. When `notmuch-hello' (or even `notmuch' when it gets its new name)
70 is invoked directly, move to the first button, (go to point-min and
71 then call widget-forward). That is, if the user has any saved
72 searches, then point will be on the first one. If the user has no
73 saved searches, then point will be on the search bar.
75 4. Fix refresh of notmuch-hello to leave point in the same logical
76 place, (the same saved-search widget at least).
78 Change the 'a'rchive command in the thread view to only archive open
81 Emacs saved-search interface
82 ----------------------------
83 Here's a proposal Carl wrote (id:87einafy4u.fsf@yoom.home.cworth.org):
85 So what I'm imagining for the default notmuch view is something like
90 Notmuch search: _________________________________________
99 1 from:"someone special" and tag:unread
100 34 tag:notmuch and tag:todo
102 Click (or press Enter) on any search to see the results.
103 Right-click (or press Space) on any recent search to save it.
105 So the "saved searches" portion of the view is basically just what
106 notmuch-folder displays now. Above that there's an obvious place to
107 start a new search, (in a slightly more "web-browser-like" way than the
108 typical mini-buffer approach).
110 All recent searches appear in the list at the bottom automatically, and
111 there's the documented mechanism for saving a search, (giving it a name
112 and having it appear above).
116 Fix configure script to test each compiler warning we want to use.
120 Fix bash completion to complete multiple search options (both --first
121 and *then* --max-threads), and also complete value for --sort=
122 (oldest-first or newest-first).
124 notmuch command-line tool
125 -------------------------
126 Fix the --format=json option to not imply --entire-thread.
128 Implement "notmuch search --exclude-threads=<search-terms>" to allow
129 for excluding muted threads, (and any other negative, thread-based
130 filtering that the user wants to do).
132 Fix "notmuch show" so that the UI doesn't fail to show a thread that
133 is visible in a search buffer, but happens to no longer match the
134 current search. (Perhaps add a --matching=<secondary-search-terms>
135 option (or similar) to "notmuch show".) For now, this is being worked
136 around in the emacs interface by noticing that "notmuch show" returns
137 nothing and re-rerunning the command without the extra arguments.
139 Teach "notmuch search" to return many different kinds of results. Some
142 notmuch search --output=threads # Default if no --output is given
143 notmuch search --output=messages
144 notmuch search --output=tags
145 notmuch search --output=addresses
146 notmuch search --output=terms
148 Add a "--format" option to "notmuch search", (something printf-like
149 for selecting what gets printed).
151 Add a "--count-only" (or so?) option to "notmuch search" for returning
152 the count of search results.
154 Give "notmuch restore" some progress indicator.
156 Fix "notmuch restore" to operate in a single pass much like "notmuch
157 dump" does, rather than doing N searches into the database, each
158 matching 1/N messages.
160 Add a "-f <filename>" option to select an alternate configuration
163 Allow configuration for filename patterns that should be ignored when
166 Replace the "notmuch part --part=id" command with "notmuch show
167 --part=id", (David Edmonson wants to rewrite some of "notmuch show" to
168 provide more MIME-structure information in its output first).
170 Replace the "notmuch search-tags" command with "notmuch search
173 Fix to avoid this ugly message:
175 (process:17197): gmime-CRITICAL **: g_mime_message_get_mime_part: assertion `GMIME_IS_MESSAGE (message)' failed
176 Warning: Not indexing empty mime part.
178 This probably means adding a test case to generate that message,
179 filing an upstream bug against GMime, and then silencing the
180 notmuch-generated portion of the warning (so that once GMime is
181 fixed, this is all silent).
183 Simplify notmuch-reply to simply print the headers (we have the
184 original values) rather than calling GMime (which encodes) and adding
185 the confusing gmime-filter-headers.c code (which decodes).
189 Add an interface to accept a "key" and a byte stream, rather than a
192 Provide a sane syntax for date ranges. First, we don't want to require
193 both endpoints to be specified. For example it would be nice to be
194 able to say things like "since:2009-01-1" or "until:2009-01-1" and
195 have the other endpoint be implicit. Second we'd like to support
196 relative specifications of time such as "since:'2 months ago'". To do
197 any of this we're probably going to need to break down an write our
198 own parser for the query string rather than using Xapian's QueryParser
201 Make failure to read a file (such as a permissions problem) a warning
202 rather than an error (should be similar to the existing warning for a
205 Actually compile and install a libnotmuch shared library.
207 Fix to use the *last* Message-ID header if multiple such headers are
208 encountered, (I noticed this is one thing that kept me from seeing the
209 same message-ID values as sup).
211 Add support for configuring "virtual tags" which are a tuple of
212 (tag-name, search-specification). The database is responsible for
213 ensuring that the virtual tag is always consistent.
215 Indicate to the user if two files with the same message ID have
216 content that is actually different in some interesting way. Perhaps
217 notmuch initially sees all changes as interesting, and quickly learns
218 from the user which changes are not interesting (such as the very
219 common mailing-list footer).
221 Fix notmuch_query_count_messages to share code with
222 notmuch_query_search_messages rather than duplicating code. (And
223 consider renaming it as well.)
225 Provide a mechanism for doing automatic address completion based on
226 notmuch searches. Here was one proposal made in IRC:
228 <cworth> I guess all it would really have to be would be a way
229 to configure a series of searches to try in turn,
230 (presenting ambiguities at a given single level, and
231 advancing to the next level only if one level
232 returned no matches).
233 <cworth> So then I might have a series that looks like this:
234 <cworth> notmuch search --output=address_from tag:address_book_alias
235 <cworth> notmuch search --output=address_to tag:sent
236 <cworth> notmuch search --output=address_from
237 <cworth> I think I might like that quite a bit.
238 <cworth> And then we have a story for an address book for
241 Provide a ~me Xapian synonym for all of the user's configured email
244 Add symbol hiding so that we don't risk leaking any private symbols
245 into the shared-library interface.
247 Audit all libnotmuch entry points to ensure that all Xapian calls are
248 wrapped in a try/catch block.
250 Fix the "count" functionality to be exact as Olly explained in IRC:
252 ojwb> cworth: if you set the check_at_least parameter to the
253 database size, get_matches_estimated() will be exact
257 Implement support for "tag:*" to expand to all tags.
259 Fix "notmuch search to:" to be less confusing. Many users expect this
260 to search for all messages with a To: header, but it instead searches
261 for all messages with the word "to". If we don't provide the first
262 behavior, perhaps we should exit on an error when a configured prefix
263 is provided with no value?
265 Support "*" in all cases and not just as a special case. That is, "* "
266 should also work, as well as "* and tag:inbox".
268 Implement a syntax for requesting set-theoertic operations on results
269 of multiple searches. For example, I would like to do:
271 "tag:inbox" SET-SUBTRACT "tag:muted"
275 "tag:notmuch and <date-range>" SET-INTERSECT
276 "tag:notmuch and not (tag:merged or tag:postponed)"
278 See id:3wdpr282yz2.fsf@testarossa.amd.com for more details on the
279 use cases of the above.
283 Store a reference term for every message-id that appears in
284 References. We just started doing this for newly-added documents, but
285 at the next convenient database-schema upgrade, we should go back and
286 fix old messages to be consistent.
288 Start indexing the List-Id header, (and re-index this header for
289 existing messages at the next database upgrade).
291 Start indexing the message file's directory ana make it available for
292 search as "folder:" (and re-index this value for existing messages at
293 the next database upgrade).
295 Add support for the user to specify custom headers to be indexed (and
296 re-index these for existing messages at the next database upgrade).
300 Start testing --format=json.
302 Achieve 100% test coverage with the test suite.
304 Modularize test suite (to be able to run individual tests).
306 Summarize test results at the end.
308 Fix the insane quoting nightmare of the test suite, (and once we do
309 that we can actually test the implicit-phrase search feature such as
310 "notmuch search 'body search (phrase)'"
312 Test "notmuch reply" choosing the correct email address from the
313 Received header when no configured email address appears in To or Cc.
317 Audit everything for dealing with out-of-memory (and drop xutil.c).
319 Investigate why the notmuch database is slightly larger than the sup
320 database for the same corpus of email.