1 TODO items for public-inbox
3 (Not in any particular order, and
4 performance, ease of setup, installation, maintainability, etc.
5 all need to be considered for everything we introduce.)
7 * general performance improvements, but without relying on
8 XS or pre-built modules any more than we currently do.
9 (Optional Inline::C and user-compiled re2c acceptable)
11 * mailmap support (same as git) for remapping expired email addresses
13 * support remapping of expired URLs similar to mailmap
14 (coordinate with git.git with this?)
16 * HTTP, IMAP, NNTP, POP3 proxy support. Allow us to be a frontend for
17 firewalled off (or Tor-exclusive) instances. The use case is
18 for offering a publicly accessible IP with a cheap VPS,
19 yet storing large amounts of data on computers without a
20 public IP behind a home Internet connection.
22 * support HTTP(S) CONNECT proxying to IMAP/NNTP/POP3 for users with
25 * DHT (distributed hash table) for mapping Message-IDs to various
26 archive locations to avoid SPOF.
28 * optional Cache::FastMmap support so production deployments won't
29 need Varnish (Varnish doesn't protect NNTP or IMAP, either)
31 * dogfood and take advantage of new kernel APIs (while maintaining
32 portability to older Linux, free BSDs and maybe Hurd).
34 * dogfood latest Xapian, Perl5, SQLite, git and various modules to
35 ensure things continue working as they should (or better)
36 while retaining compatibility with old versions.
38 * Support more of RFC 3977 (NNTP)
39 Is there anything left for read-only support?
41 * Configurable linkification for per-inbox shorthands:
42 "$gmane/123456" could be configured to expand to the
43 appropriate link pointing to the gmane.io list archives,
44 likewise "[Bug #123456]" could be configured to expand to
45 point to some project's bug tracker at http://example.com/bug/123456
47 * configurable synonym and spelling support in Xapian
49 * Support optional "HTTPS Everywhere" for mapping old HTTP to HTTPS
50 links if (and only if) the user wants to use HTTPS. We may also
51 be able to configure redirects for expired URLs.
53 Note: message bodies rendered as HTML themselves must NOT change,
54 the links should point to an anchor tag within the same page,
55 instead; giving the user options.
57 * configurable constants (index limits, search results)
59 * handle messages with multiple Message-IDs (done for v2, doable for v1)
61 * handle broken double-bracketed References properly (maybe)
62 and totally broken Message-IDs
64 cf. https://public-inbox.org/git/20160814012706.GA18784@starla/
66 * improve documentation
68 * linkify thread skeletons better
69 https://public-inbox.org/git/6E3699DEA672430CAEA6DEFEDE6918F4@PhilipOakley/
71 * Further lower mail parser memory usage. We still slurp entire
72 message bodies into memory and incur 2-3x overhead on
73 multipart messages. Inline::C (and maybe gmime) could work.
75 * use REQUEST_URI properly for CGI / mod_perl2 compatibility
76 with Message-IDs which include '%' (done?)
78 * handle unencoded slashes from user-generated URLs properly
79 https://public-inbox.org/meta/20230217085255.xcsaoozloz2yuxil@pengutronix.de/
81 * better test cases, make faster by reusing more setup
84 * large mbox/Maildir/MH/NNTP spool import (in lei, but not
85 for public-facing inboxes)
87 * MH import support (read-only, at least)
89 * Read-only WebDAV interface to the git repo so it can be mounted
90 via davfs2 or fusedav to avoid full clones.
91 davfs2 needs Range: request support for this to be feasible:
92 https://savannah.nongnu.org/bugs/?33259
93 https://savannah.nongnu.org/support/?107649
95 * Contribute something like IMAP IDLE for "git fetch".
96 Inboxes (and any git repos) can be kept up-to-date without
99 * Improve bundle support in git to make it cheaper to host/clone
100 with dumb HTTP(S) servers.
102 * Expose targeted reindexing of individual messages.
103 Sometimes an indexing bug only affects a handful of messages,
104 so it's not worth the trouble of doing a full reindex.
106 * code repository integration (cgit: done, TODO: gitweb, etc...)
108 * migration path to v2 (making it transparent for "git fetch"
109 may not be possible, but "public-inbox-fetch" will handle it)
111 * imperfect scraper importers for obfuscated list archives
112 (e.g. obfuscated Mailman stuff, Google Groups, etc...)
114 * improve performance and avoid head-of-line blocking on slow storage
115 (done for most git blob retrievals, Xapian needs work)
117 * allow optional use of separate Xapian worker process to implement
118 timeouts and avoid head-of-line blocking problems. Consider
119 just-ahead-of-time builds to take advantage of custom date parsers
120 (approxidate) and other features not available to Perl bindings.
122 * integrate git approxidate parsing into Xapian w/o spawning git
124 * HTTP(S) search API (likely JMAP, but GraphQL could be an option)
125 It should support git-specific prefixes (dfpre:, dfpost:, dfn:, etc)
126 as extensions. If JMAP, it should have HTTP(S) analogues to
127 various IMAP extensions.
129 * scalability to tens/hundreds of thousands of inboxes
131 - inotify-based manifest.js.gz updates
135 * lei - see %CMD in lib/PublicInbox/LEI.pm
136 (there's a truckload here..)
138 * make "git cat-file --batch" detect unlinked packfiles so we don't
139 have to restart processes (very long-term)
141 * linter to check validity of config file
143 * linter option and WWW endpoint to graph relationships and flows
144 between inboxes, addresses, Maildirs, coderepos, newsgroups,
145 IMAP mailboxes, etc...
147 * pygments support - via Python script similar to `git cat-file --batch'
148 to avoid startup penalty. pygments.rb (Ruby) can be inspiration, too.
150 * highlighting + linkification for "git format-patch --interdiff" output
152 * highlighting for "git format-patch --range-diff" output
153 (linkification is too expensive, as it requires mirroring)
155 * support UUCP addresses for legacy archives
157 * support pipelining as an IMAP/NNTP client for -watch + lei
159 * expose lei contents via read/write IMAP/JMAP server for personal use
161 * git SHA-256 migration/coexistence path
163 * decode RFC 3676 format=flowed + DelSp properly (see mflow (mblaze), mutt, ...)