sup

A curses threads-with-tags style email client

sup-website.git

git clone https://supmua.dev/git/sup-website/

community/pipermail-archives/sup-devel/2010-06.txt (41610B) - raw

      1 From sup-bugs@masanjin.net  Tue Jun  1 15:28:53 2010
      2 From: sup-bugs@masanjin.net (anonymous)
      3 Date: Tue, 01 Jun 2010 19:28:53 +0000
      4 Subject: [sup-devel] [issue107] all new emails go to inbox
      5 In-Reply-To: <1275420532.84.0.199838497104.issue107@masanjin.net>
      6 Message-ID: <1275420532.84.0.199838497104.issue107@masanjin.net>
      7 
      8 
      9 New submission from anonymous:
     10 
     11 after checking out the latest git version, all new emails go to inbox now,
     12 although several sources are configured to be archived immediately.
     13 
     14 ----------
     15 messages: 242
     16 nosy: anonymous
     17 priority: bug
     18 ruby_version: 1.8.7 (2010-01-10 patchlevel 249)
     19 status: unread
     20 sup_version: git
     21 title: all new emails go to inbox
     22 
     23 _________________________________________
     24 Sup issue tracker <sup-bugs at masanjin.net>
     25 <http://masanjin.net/sup-bugs/issue107>
     26 _________________________________________
     27 
     28 From sup-bugs@masanjin.net  Wed Jun  2 18:06:32 2010
     29 From: sup-bugs@masanjin.net (Gaute Hope)
     30 Date: Wed, 02 Jun 2010 22:06:32 +0000
     31 Subject: [sup-devel] [issue108] Feature request: Jump to next message and
     32 	open in	thread-view
     33 In-Reply-To: <1275516359-sup-2170@dolk>
     34 Message-ID: <1275516359-sup-2170@dolk>
     35 
     36 
     37 New submission from Gaute Hope <eg at gaute.vetsj.com>:
     38 
     39 The attached patch adds the possibility to jump to the next message and
     40 open in one go in thread-view-mode.
     41 
     42 I often want to just skim through a thread; I might have been missing
     43 something, but it previously required me to locate the next message hit
     44 enter, then repeat..
     45 
     46 Now I can with C-n and C-p jump to the next, have it opened, hit again -
     47 it is closed if it was toggled, and continue through the thread.
     48 
     49 Please beware of bad ruby knowledge.
     50 
     51 ----------
     52 files: 0001-Jump-and-open-next-previous-message-with-C-n-and-C-p.patch
     53 messages: 243
     54 nosy: gauteh
     55 status: unread
     56 title: Feature request: Jump to next message and open in thread-view
     57 
     58 _________________________________________
     59 Sup issue tracker <sup-bugs at masanjin.net>
     60 <http://masanjin.net/sup-bugs/issue108>
     61 _________________________________________
     62 -------------- next part --------------
     63 A non-text attachment was scrubbed...
     64 Name: 0001-Jump-and-open-next-previous-message-with-C-n-and-C-p.patch
     65 Type: application/octet-stream
     66 Size: 4170 bytes
     67 Desc: not available
     68 URL: <http://rubyforge.org/pipermail/sup-devel/attachments/20100602/34b397b2/attachment.obj>
     69 
     70 From sup-bugs@masanjin.net  Wed Jun  2 20:18:23 2010
     71 From: sup-bugs@masanjin.net (Truxton Fulton)
     72 Date: Thu, 03 Jun 2010 00:18:23 +0000
     73 Subject: [sup-devel] [issue109] Crash when loading all threads (!!)
     74 In-Reply-To: <1275524303.69.0.689658146626.issue109@masanjin.net>
     75 Message-ID: <1275524303.69.0.689658146626.issue109@masanjin.net>
     76 
     77 
     78 New submission from Truxton Fulton <truxton at truxton.com>:
     79 
     80 I started sup
     81 There are 127945 messages in index
     82 I pressed L then chose a label that had 10122 messages
     83 sup displayed the first 60 threads
     84 I pressed !! to load all threads
     85 (because I want to apply 'a'rchive to all of them)
     86 after loading 546 threads, sup crashes (log attached).
     87 
     88 ----------
     89 files: exception-log.txt
     90 keyword: maildir
     91 messages: 244
     92 nosy: trux
     93 priority: bug
     94 ruby_version: 1.8.6
     95 status: unread
     96 sup_version: 0.11
     97 title: Crash when loading all threads (!!)
     98 
     99 _________________________________________
    100 Sup issue tracker <sup-bugs at masanjin.net>
    101 <http://masanjin.net/sup-bugs/issue109>
    102 _________________________________________
    103 -------------- next part --------------
    104 --- RuntimeError from thread: load threads for thread-index-mode
    105 wrong id called on nil
    106 /usr/lib/ruby/gems/1.8/gems/sup-0.11/lib/sup.rb:17:in `id'
    107 /usr/lib/ruby/gems/1.8/gems/sup-0.11/lib/sup/modes/thread-index-mode.rb:234:in `update'
    108 /usr/lib/ruby/gems/1.8/gems/sup-0.11/lib/sup/hook.rb:123:in `sort_by'
    109 /usr/lib/ruby/gems/1.8/gems/sup-0.11/lib/sup/modes/thread-index-mode.rb:234:in `each'
    110 /usr/lib/ruby/gems/1.8/gems/sup-0.11/lib/sup/modes/thread-index-mode.rb:234:in `sort_by'
    111 /usr/lib/ruby/gems/1.8/gems/sup-0.11/lib/sup/modes/thread-index-mode.rb:234:in `update'
    112 /usr/lib/ruby/gems/1.8/gems/sup-0.11/lib/sup/modes/thread-index-mode.rb:232:in `synchronize'
    113 /usr/lib/ruby/gems/1.8/gems/sup-0.11/lib/sup/modes/thread-index-mode.rb:232:in `update'
    114 /usr/lib/ruby/gems/1.8/gems/sup-0.11/lib/sup/modes/thread-index-mode.rb:643:in `__unprotected_load_n_threads'
    115 /usr/lib/ruby/gems/1.8/gems/sup-0.11/lib/sup/thread.rb:334:in `load_n_threads'
    116 /usr/lib/ruby/gems/1.8/gems/sup-0.11/lib/sup/index.rb:148:in `each_id_by_date'
    117 /usr/lib/ruby/gems/1.8/gems/sup-0.11/lib/sup/index.rb:239:in `each_id'
    118 /usr/lib/ruby/gems/1.8/gems/sup-0.11/lib/sup/index.rb:239:in `each'
    119 /usr/lib/ruby/gems/1.8/gems/sup-0.11/lib/sup/index.rb:239:in `each_id'
    120 /usr/lib/ruby/gems/1.8/gems/sup-0.11/lib/sup/index.rb:148:in `each_id_by_date'
    121 /usr/lib/ruby/gems/1.8/gems/sup-0.11/lib/sup/thread.rb:328:in `load_n_threads'
    122 /usr/lib/ruby/gems/1.8/gems/sup-0.11/lib/sup/modes/thread-index-mode.rb:640:in `__unprotected_load_n_threads'
    123 (eval):12:in `load_n_threads'
    124 /usr/lib/ruby/gems/1.8/gems/sup-0.11/lib/sup/modes/thread-index-mode.rb:624:in `load_n_threads_background'
    125 /usr/lib/ruby/gems/1.8/gems/sup-0.11/lib/sup.rb:76:in `reporting_thread'
    126 /usr/lib/ruby/gems/1.8/gems/sup-0.11/lib/sup.rb:74:in `initialize'
    127 /usr/lib/ruby/gems/1.8/gems/sup-0.11/lib/sup.rb:74:in `new'
    128 /usr/lib/ruby/gems/1.8/gems/sup-0.11/lib/sup.rb:74:in `reporting_thread'
    129 /usr/lib/ruby/gems/1.8/gems/sup-0.11/lib/sup/modes/thread-index-mode.rb:623:in `load_n_threads_background'
    130 /usr/lib/ruby/gems/1.8/gems/sup-0.11/lib/sup/modes/thread-index-mode.rb:694:in `__unprotected_load_threads'
    131 (eval):12:in `load_threads'
    132 /usr/lib/ruby/gems/1.8/gems/sup-0.11/lib/sup/modes/thread-index-mode.rb:673:in `load_all_threads'
    133 /usr/lib/ruby/gems/1.8/gems/sup-0.11/lib/sup/mode.rb:59:in `send'
    134 /usr/lib/ruby/gems/1.8/gems/sup-0.11/lib/sup/mode.rb:59:in `handle_input'
    135 /usr/lib/ruby/gems/1.8/gems/sup-0.11/lib/sup/buffer.rb:279:in `handle_input'
    136 /usr/lib/ruby/gems/1.8/gems/sup-0.11/bin/sup:279
    137 /usr/bin/sup:19:in `load'
    138 /usr/bin/sup:19
    139 
    140 From rlane@club.cc.cmu.edu  Wed Jun  2 23:57:41 2010
    141 From: rlane@club.cc.cmu.edu (Rich Lane)
    142 Date: Wed, 02 Jun 2010 23:57:41 -0400
    143 Subject: [sup-devel] [issue108] Feature request: Jump to next message
    144 	and open in thread-view
    145 In-Reply-To: <1275516359-sup-2170@dolk>
    146 References: <1275516359-sup-2170@dolk>
    147 Message-ID: <1275537191-sup-9518@zyrg.net>
    148 
    149 Excerpts from Gaute Hope's message of 2010-06-02 18:06:32 -0400:
    150 > 
    151 > New submission from Gaute Hope <eg at gaute.vetsj.com>:
    152 > 
    153 > The attached patch adds the possibility to jump to the next message and
    154 > open in one go in thread-view-mode.
    155 > 
    156 > I often want to just skim through a thread; I might have been missing
    157 > something, but it previously required me to locate the next message hit
    158 > enter, then repeat..
    159 > 
    160 > Now I can with C-n and C-p jump to the next, have it opened, hit again -
    161 > it is closed if it was toggled, and continue through the thread.
    162 > 
    163 > Please beware of bad ruby knowledge.
    164 
    165 Applied to master.
    166 
    167 From rlane@club.cc.cmu.edu  Wed Jun  2 23:58:20 2010
    168 From: rlane@club.cc.cmu.edu (Rich Lane)
    169 Date: Wed, 02 Jun 2010 23:58:20 -0400
    170 Subject: [sup-devel] [PATCH] Respect source.archived? in poll.
    171 In-Reply-To: <1274910492-13206-1-git-send-email-pi+sup@pihost.us>
    172 References: <1274910492-13206-1-git-send-email-pi+sup@pihost.us>
    173 Message-ID: <1275537478-sup-8837@zyrg.net>
    174 
    175 Applied to master.
    176 
    177 From rlane@club.cc.cmu.edu  Wed Jun  2 23:58:49 2010
    178 From: rlane@club.cc.cmu.edu (Rich Lane)
    179 Date: Wed, 02 Jun 2010 23:58:49 -0400
    180 Subject: [sup-devel] [PATCH] Allow toggle on Source.usual and
    181 	Source.archived
    182 In-Reply-To: <1273949121-16731-1-git-send-email-bwalton@artsci.utoronto.ca>
    183 References: <1273949121-16731-1-git-send-email-bwalton@artsci.utoronto.ca>
    184 Message-ID: <1275537512-sup-176@zyrg.net>
    185 
    186 Applied to master.
    187 
    188 From rlane@club.cc.cmu.edu  Thu Jun  3 00:27:04 2010
    189 From: rlane@club.cc.cmu.edu (Rich Lane)
    190 Date: Thu, 03 Jun 2010 00:27:04 -0400
    191 Subject: [sup-devel] email threading - tree vs. graph
    192 In-Reply-To: <20100525185026.GA11947@thialfi.home.net>
    193 References: <20100218114943.GB911@mjolnir> <1266730498-sup-78@tilus.net>
    194 	<20100525185026.GA11947@thialfi.home.net>
    195 Message-ID: <1275538158-sup-1305@zyrg.net>
    196 
    197 Excerpts from W. Trevor King's message of 2010-05-25 14:50:27 -0400:
    198 > I got some good feedback from Nicolas Pouillard on the Python tidbit I
    199 > posted, but after waiting optimisticly for some enterprising Rubist to
    200 > port it to Ruby and merge it into Sup, I've finally taught myself
    201 > enough Ruby to do it myself ;).  Here's DAG-supporting Sup (+ a few
    202 > glaring documentation updates)
    203 
    204 Thanks for the doc updates, I've applied them to master.
    205 
    206 I'm not yet convinced that supporting multiple parents is worth
    207 the complexity. How often do you get mails like this? What mail clients
    208 have UI for replying to multiple messages? I'd have to fake up some
    209 messages to actually see the full DAG support in action, so could you
    210 post a screenshot of your code viewing a thread where the new display
    211 helps?
    212 
    213 > We could also resurect the old indentation-style display in the thread
    214 > viewer, if people dislike my tig-style ascii graph.
    215 
    216 Actually, I think the tig-style display is very interesting and could be
    217 useful even without the generic graph support.
    218 
    219 From truxton@truxton.com  Thu Jun  3 04:27:34 2010
    220 From: truxton@truxton.com (Truxton Fulton)
    221 Date: Thu, 03 Jun 2010 01:27:34 -0700
    222 Subject: [sup-devel] [issue109] Crash when loading all threads (!!)
    223 In-Reply-To: <1275524303.69.0.689658146626.issue109@masanjin.net>
    224 References: <1275524303.69.0.689658146626.issue109@masanjin.net>
    225 Message-ID: <1275553467-sup-5430@terrapin.truxton.com>
    226 
    227 Hi sup developers!
    228 
    229 I've narrowed down the crash to the presence of 2 particular mails.
    230 I can reproduce the crash by creating a brand new sup config
    231 for a maildir containing only these 2 messages.  When sup starts
    232 up it gets the "wrong id called on nil" crash.  It seems that
    233 either message by itself is fine, but the combination of the
    234 two messages causes the problem.
    235 
    236 You can download these mails from my web server :
    237 
    238   http://www.truxton.com/bad2.tar.gz
    239 
    240 Thanks,
    241 
    242 -Truxton
    243 -- 
    244 
    245 From wking@drexel.edu  Thu Jun  3 06:21:10 2010
    246 From: wking@drexel.edu (W. Trevor King)
    247 Date: Thu, 3 Jun 2010 06:21:10 -0400
    248 Subject: [sup-devel] email threading - tree vs. graph
    249 In-Reply-To: <1275538158-sup-1305@zyrg.net>
    250 References: <20100218114943.GB911@mjolnir> <1266730498-sup-78@tilus.net>
    251 	<20100525185026.GA11947@thialfi.home.net>
    252 	<1275538158-sup-1305@zyrg.net>
    253 Message-ID: <20100603102109.GA1499@thialfi>
    254 
    255 On Thu, Jun 03, 2010 at 12:27:04AM -0400, Rich Lane wrote:
    256 > I'm not yet convinced that supporting multiple parents is worth
    257 > the complexity. How often do you get mails like this?
    258 
    259 I don't get mails with multiple in-reply-tos, but I send them ;).  I
    260 *do* get mails with in-text references to multiple messages.  For some
    261 mailing lists that I care about (be-devel), I go through and adjust
    262 in-reply-to so that it matches the in-text references.
    263 
    264 I think such graphs would be a nice interface to mailing lists, which
    265 gain members not familiar with the lists history.  Browsing a curated
    266 list, finding the critical background for a particular message would
    267 be trivial.
    268 
    269 > What mail clients have UI for replying to multiple messages?
    270 
    271 In Mutt, you can tag a bunch of messages, and then ;g to reply-to-all
    272 the tagged messages at once.  I started doing this to get the previous
    273 text from all the messages included in my reply, so I expect this sort
    274 of thing does occur occasionally in the wild.
    275 
    276 > I'd have to fake up some messages to actually see the full DAG
    277 > support in action, so could you post a screenshot of your code
    278 > viewing a thread where the new display helps?
    279 
    280 Attached is a small graph showing two related threads about
    281 bzr+windows.  The multiparent replies allow responses like Matt's,
    282   "we discussed this last year" message.
    283 
    284 -- 
    285 This email may be signed or encrypted with GPG (http://www.gnupg.org).
    286 The GPG signature (if present) will be attached as 'signature.asc'.
    287 For more information, see http://en.wikipedia.org/wiki/Pretty_Good_Privacy
    288 
    289 My public key is at http://www.physics.drexel.edu/~wking/pubkey.txt
    290 -------------- next part --------------
    291 A non-text attachment was scrubbed...
    292 Name: graph-threading.png
    293 Type: image/png
    294 Size: 13150 bytes
    295 Desc: not available
    296 URL: <http://rubyforge.org/pipermail/sup-devel/attachments/20100603/595e2c32/attachment-0001.png>
    297 -------------- next part --------------
    298 A non-text attachment was scrubbed...
    299 Name: not available
    300 Type: application/pgp-signature
    301 Size: 198 bytes
    302 Desc: not available
    303 URL: <http://rubyforge.org/pipermail/sup-devel/attachments/20100603/595e2c32/attachment-0001.bin>
    304 
    305 From michael+sup@stapelberg.de  Fri Jun  4 05:59:53 2010
    306 From: michael+sup@stapelberg.de (Michael Stapelberg)
    307 Date: Fri, 04 Jun 2010 11:59:53 +0200
    308 Subject: [sup-devel] [PATCH] Bugfix: Charset conversion needs to take place
    309 	*after* verifying inline GPG signatures
    310 Message-ID: <1275645515-sup-4706@midna.zekjur.net>
    311 
    312 Hi,
    313 
    314 still fixing some inline GPG problems ;-). This one is a little subtle: If the
    315 message arrives in a different encoding than UTF-8 (shame on the sender), the
    316 verification will fail because sup modifies the message by converting it to
    317 UTF-8. The attached patch fixes this.
    318 
    319 Best regards,
    320 Michael
    321 -------------- next part --------------
    322 A non-text attachment was scrubbed...
    323 Name: 0001-Bugfix-Charset-conversion-needs-to-take-place-after-.patch
    324 Type: application/octet-stream
    325 Size: 2799 bytes
    326 Desc: not available
    327 URL: <http://rubyforge.org/pipermail/sup-devel/attachments/20100604/3c41f615/attachment.obj>
    328 
    329 From sean@silentflame.com  Fri Jun  4 09:44:04 2010
    330 From: sean@silentflame.com (Sean Whitton)
    331 Date: Fri, 4 Jun 2010 14:44:04 +0100
    332 Subject: [sup-devel] Storing message tags and other Sup info as headers in
    333 	Maildir
    334 Message-ID: <20100604134404.GA28767@artemis.silentflame.com>
    335 
    336 Dear all,
    337 
    338 I've been checking back to the Sup website every few months for the past 
    339 year or so, waiting until Sup starts to look more stable and suitable 
    340 for regular use.  Like many people on this list, a great stumbling block 
    341 to adoption is the fact that Sup doesn't really let you access your 
    342 e-mail with anything other than Sup.  At the moment I use offlineimap 
    343 and read my mail with Mutt, but my phone and SquirrelMail can access the 
    344 Maildir just as well (by IMAP in the former case), and so everything 
    345 stays in sync and I can get to my e-mail from many places.
    346 
    347 Now, I am no real coder, and I've never written a line of ruby, but a 
    348 thought has occurred to me that I feel I should at least share, even if 
    349 it turns out to be completely impractical.  Why not store the 
    350 information associated with e-mails that is not rebuildable (that is, 
    351 tags, unread/read status, starred status, archived/killed/spam status) 
    352 as header lines (X-Sup-Tags: X-Sup-Status, and I guess read/unread could 
    353 be standard Maildir flags) in the e-mails themselves in the Maildir?
    354 
    355 This way, the Sup index could be rebuilt on multiple client machines 
    356 without any data actually going out of sync.  You'll still have a 
    357 punishing index rebuild every time you view your mail on a new machine, 
    358 but they'll never be the problem of things actually being wrong - tags 
    359 and stars and the like can all be propagated by offlineimap/IMAP.  The 
    360 indexer can rip out these flags into its index to maintain Sup's 
    361 professed speed, and then if they change, write them back into the 
    362 Maildir along with read/unread status and other flags.
    363 
    364 Does this exposition make sense?  Is this a practical way to 
    365 improve/create Sup's multi-client support?
    366 
    367 S
    368 
    369 -- 
    370 Sean Whitton / <sean at silentflame.com>
    371 OpenPGP KeyID: 0x3B6D411B
    372 http://seanwhitton.com/
    373 
    374 -------------- next part --------------
    375 A non-text attachment was scrubbed...
    376 Name: not available
    377 Type: application/pgp-signature
    378 Size: 836 bytes
    379 Desc: Digital signature
    380 URL: <http://rubyforge.org/pipermail/sup-devel/attachments/20100604/d85f5e9b/attachment.bin>
    381 
    382 From michael+sup@stapelberg.de  Fri Jun  4 18:17:54 2010
    383 From: michael+sup@stapelberg.de (Michael Stapelberg)
    384 Date: Sat, 05 Jun 2010 00:17:54 +0200
    385 Subject: [sup-devel] [PATCH] Decode messages according to their
    386 	Content-Transfer-Encoding
    387 Message-ID: <1275689764-sup-3384@midna.zekjur.net>
    388 
    389 Hi,
    390 
    391 quote of the commit message:
    392     Decode messages according to their Content-Transfer-Encoding
    393         
    394     This is necessary for MIME-messages (for example as part of multipart/signed)
    395     which are encoded in base64.
    396 
    397 Best regards,
    398 Michael
    399 -------------- next part --------------
    400 A non-text attachment was scrubbed...
    401 Name: 0001-Decode-messages-according-to-their-Content-Transfer-.patch
    402 Type: application/octet-stream
    403 Size: 1433 bytes
    404 Desc: not available
    405 URL: <http://rubyforge.org/pipermail/sup-devel/attachments/20100605/b307d5aa/attachment.obj>
    406 
    407 From rlane@club.cc.cmu.edu  Fri Jun  4 22:54:24 2010
    408 From: rlane@club.cc.cmu.edu (Rich Lane)
    409 Date: Fri, 04 Jun 2010 22:54:24 -0400
    410 Subject: [sup-devel] [PATCH] Decode messages according to their
    411 	Content-Transfer-Encoding
    412 In-Reply-To: <1275689764-sup-3384@midna.zekjur.net>
    413 References: <1275689764-sup-3384@midna.zekjur.net>
    414 Message-ID: <1275706048-sup-6981@zyrg.net>
    415 
    416 Excerpts from Michael Stapelberg's message of 2010-06-04 18:17:54 -0400:
    417 >     Decode messages according to their Content-Transfer-Encoding
    418 >         
    419 >     This is necessary for MIME-messages (for example as part of multipart/signed)
    420 >     which are encoded in base64.
    421 
    422 Do we need to do this in other places too?
    423 
    424 From rlane@club.cc.cmu.edu  Fri Jun  4 22:55:08 2010
    425 From: rlane@club.cc.cmu.edu (Rich Lane)
    426 Date: Fri, 04 Jun 2010 22:55:08 -0400
    427 Subject: [sup-devel] [PATCH] Bugfix: Charset conversion needs to take
    428 	place *after* verifying inline GPG signatures
    429 In-Reply-To: <1275645515-sup-4706@midna.zekjur.net>
    430 References: <1275645515-sup-4706@midna.zekjur.net>
    431 Message-ID: <1275706494-sup-497@zyrg.net>
    432 
    433 Applied to master.
    434 
    435 From michael+sup@stapelberg.de  Sat Jun  5 06:55:39 2010
    436 From: michael+sup@stapelberg.de (Michael Stapelberg)
    437 Date: Sat, 05 Jun 2010 12:55:39 +0200
    438 Subject: [sup-devel] [PATCH] Decode messages according to their
    439 	Content-Transfer-Encoding
    440 In-Reply-To: <1275706048-sup-6981@zyrg.net>
    441 References: <1275689764-sup-3384@midna.zekjur.net>
    442 	<1275706048-sup-6981@zyrg.net>
    443 Message-ID: <1275735302-sup-799@midna.zekjur.net>
    444 
    445 Hi Rich,
    446 
    447 Excerpts from Rich Lane's message of 2010-06-05 04:54:24 +0200:
    448 > Do we need to do this in other places too?
    449 Not sure, unfortunately. I?d guess only for MIME messages, for other
    450 messages the Transfer-Encoding seems to work already.
    451 
    452 Best regards,
    453 Michael
    454 
    455 From rlane@club.cc.cmu.edu  Mon Jun  7 11:55:59 2010
    456 From: rlane@club.cc.cmu.edu (Rich Lane)
    457 Date: Mon, 07 Jun 2010 11:55:59 -0400
    458 Subject: [sup-devel] [PATCH] Decode messages according to their
    459 	Content-Transfer-Encoding
    460 In-Reply-To: <1275735302-sup-799@midna.zekjur.net>
    461 References: <1275689764-sup-3384@midna.zekjur.net>
    462 	<1275706048-sup-6981@zyrg.net>
    463 	<1275735302-sup-799@midna.zekjur.net>
    464 Message-ID: <1275926137-sup-8653@zyrg.net>
    465 
    466 Excerpts from Michael Stapelberg's message of 2010-06-05 06:55:39 -0400:
    467 > Excerpts from Rich Lane's message of 2010-06-05 04:54:24 +0200:
    468 > > Do we need to do this in other places too?
    469 > Not sure, unfortunately. I?d guess only for MIME messages, for other
    470 > messages the Transfer-Encoding seems to work already.
    471 
    472 Ok, applied to master.
    473 
    474 From snaipperi@gmail.com  Mon Jun  7 23:14:08 2010
    475 From: snaipperi@gmail.com (Matti Eiden)
    476 Date: Tue, 8 Jun 2010 06:14:08 +0300
    477 Subject: [sup-devel] Storing message tags and other Sup info as headers
    478 	in Maildir
    479 In-Reply-To: <20100604134404.GA28767@artemis.silentflame.com>
    480 References: <20100604134404.GA28767@artemis.silentflame.com>
    481 Message-ID: <AANLkTikI63bB_jx3EJsJ15iObHWLZJf_x0d8iw-gDrXP@mail.gmail.com>
    482 
    483 Hi Sean,
    484 
    485 I was expecting somebody else to reply in this but since I'm not
    486 seeing anyone doing so, I'll give it a shot. I only recently
    487 discovered sup, and I'm not a Ruby programmer, however I got quite a
    488 bunch of experience with Python (now the rest of the mailing list can
    489 proceed flaming me). The following post is purely speculative.
    490 
    491 Shortly put, I don't expect to see any change on this matter, at least
    492 in the near future. There is and has been a tool, sup-sync-back which
    493 should reflect the changes back to mbox/maildir but the way it works
    494 is far from ideal, I guess. Anyway..
    495 
    496 Let's get one thing straight first, sup doesn't really use specific
    497 status flags/tags or such for mails. Instead, every piece of
    498 information is in the labels. A couple of labels are predefined:
    499 Attachment, Deleted, Draft, Inbox, Killed, Sent, Spam, Starred and
    500 Unread. Technically, a message that is archived is simply lacking the
    501 label "Inbox". Rest of the labels are user-defined.
    502 
    503 Standard maildir format already provides following flags by default:
    504 (S)een, (T)rashed, (D)raft. In addition flags that sup doesn't need;
    505 Passed, Replied and Flagged.
    506 
    507 Dovecot (an IMAP server) provides user defined flags for maildirs. The
    508 flags are lowercase letters ranging a-z (up to 26 different), and
    509 seems like it works OK with the maildir. (
    510 http://wiki.dovecot.org/MailboxFormat/Maildir ). This could be one
    511 nice option if a limit of some 20 user tags aren't too few.
    512 
    513 Example maildir mail flags:
    514 :2,Sacd
    515 
    516 S - standard maildir flag: Seen, MUA would label as "Read"
    517 a - custom flag - MUA would label as "Archived"
    518 b - custom flag - MUA would label as "Starred"
    519 d - custom flag - User would label as "Work"
    520 
    521 This is of course kinda opposite to sup's current situation where
    522 messages are defined as Unread or Inbox, but anyway..
    523 
    524 
    525 Personally I don't like the idea of MUA messing up with the email
    526 headers as you suggested. Plus, this would result in a lot of
    527 fragmentation on a hard disk with tens of thousands of emails, if on
    528 initial sync sup would need to write label information in the middle
    529 of every email. And as always when writing, there's a risk of data
    530 loss, which would require additional measures.
    531 
    532 This area interests me, however, and I guess I'm gonna make some
    533 experiments how the flagging thing works in action, and if it makes
    534 other mail clients break. (I doubt it does, otherwise Dovecot wouldn't
    535 be using it, huh?).
    536 
    537 Regards,
    538 Matti
    539 
    540 From damien.leone@fensalir.fr  Wed Jun  9 08:53:16 2010
    541 From: damien.leone@fensalir.fr (Damien Leone)
    542 Date: Wed, 09 Jun 2010 14:53:16 +0200
    543 Subject: [sup-devel] [PATCHES] reply-mode, signatures and account selector
    544 Message-ID: <1276087951-sup-7816@mailer>
    545 
    546 Sup guys,
    547 
    548 Three patches are attached to this email, the commit messages should
    549 be self explanatory but here are some more details:
    550 
    551 As the first patch changes the way headers are handled in reply-mode
    552 there is a regression in the before-edit hook possibilities.
    553 Before, we could write something like:
    554 
    555 if header["Cc"] == "foo"
    556    header["Bcc"] = "bar"
    557 end
    558 
    559 Hence, switching the reply-to type using the selector would have
    560 changed the Bcc field if a type using Cc was selected.
    561 
    562 This would be no longer possible since there is now only one "set" of
    563 headers running the hook, however if you properly wrote your reply-to
    564 hook you can select the reply type you want and the previous code
    565 will be working as expected.
    566 
    567 On the other hand (in my opinion), the reply-mode now handles better
    568 its job, that is to say changing To and Cc without interfering with
    569 other fields that might have been edited manually by the user.
    570 
    571 I asked rlane about this on IRC:
    572 > <dleone> what is the reason for the before-edit hook being runned on @bodies and @header for each Reply-To types?
    573 > <rlane> that's so that when you switch reply-to type the right signature/etc is displayed
    574 
    575 I see no regression regarding this in the patch, so it should be okay.
    576 
    577 Other patches add an account selector in edit-mode (it is useful to me
    578 and I saw requests for such a feature) and a better signature handling
    579 regarding the edit_signature option.
    580 
    581 Regards,
    582 -------------- next part --------------
    583 A non-text attachment was scrubbed...
    584 Name: 0001-reply-mode-improve-the-way-headers-are-handled.patch
    585 Type: application/octet-stream
    586 Size: 5335 bytes
    587 Desc: not available
    588 URL: <http://rubyforge.org/pipermail/sup-devel/attachments/20100609/e3f92192/attachment-0003.obj>
    589 -------------- next part --------------
    590 A non-text attachment was scrubbed...
    591 Name: 0002-Add-account_selector-config-option.patch
    592 Type: application/octet-stream
    593 Size: 3643 bytes
    594 Desc: not available
    595 URL: <http://rubyforge.org/pipermail/sup-devel/attachments/20100609/e3f92192/attachment-0004.obj>
    596 -------------- next part --------------
    597 A non-text attachment was scrubbed...
    598 Name: 0003-edit-mode-change-the-way-signatures-are-handled.patch
    599 Type: application/octet-stream
    600 Size: 2855 bytes
    601 Desc: not available
    602 URL: <http://rubyforge.org/pipermail/sup-devel/attachments/20100609/e3f92192/attachment-0005.obj>
    603 
    604 From damien.leone@fensalir.fr  Wed Jun  9 09:15:11 2010
    605 From: damien.leone@fensalir.fr (Damien Leone)
    606 Date: Wed, 09 Jun 2010 15:15:11 +0200
    607 Subject: [sup-devel] [PATCHES] reply-mode,
    608 	signatures and account selector
    609 In-Reply-To: <1276087951-sup-7816@mailer>
    610 References: <1276087951-sup-7816@mailer>
    611 Message-ID: <1276089233-sup-510@mailer>
    612 
    613 I noticed an error in the third patch.
    614 Please consider the one attached to *this* email.
    615 
    616 Regards,
    617 
    618 Excerpts from Damien Leone's message of mer. juin 09 14:53:16 +0200 2010:
    619 > Sup guys,
    620 > 
    621 > Three patches are attached to this email, the commit messages should
    622 > be self explanatory but here are some more details:
    623 > 
    624 > As the first patch changes the way headers are handled in reply-mode
    625 > there is a regression in the before-edit hook possibilities.
    626 > Before, we could write something like:
    627 > 
    628 > if header["Cc"] == "foo"
    629 >    header["Bcc"] = "bar"
    630 > end
    631 > 
    632 > Hence, switching the reply-to type using the selector would have
    633 > changed the Bcc field if a type using Cc was selected.
    634 > 
    635 > This would be no longer possible since there is now only one "set" of
    636 > headers running the hook, however if you properly wrote your reply-to
    637 > hook you can select the reply type you want and the previous code
    638 > will be working as expected.
    639 > 
    640 > On the other hand (in my opinion), the reply-mode now handles better
    641 > its job, that is to say changing To and Cc without interfering with
    642 > other fields that might have been edited manually by the user.
    643 > 
    644 > I asked rlane about this on IRC:
    645 > > <dleone> what is the reason for the before-edit hook being runned on @bodies and @header for each Reply-To types?
    646 > > <rlane> that's so that when you switch reply-to type the right signature/etc is displayed
    647 > 
    648 > I see no regression regarding this in the patch, so it should be okay.
    649 > 
    650 > Other patches add an account selector in edit-mode (it is useful to me
    651 > and I saw requests for such a feature) and a better signature handling
    652 > regarding the edit_signature option.
    653 > 
    654 > Regards,
    655 
    656 --
    657 Damien Leone <damien.leone at fensalir.fr>
    658 
    659 Web: http://dleone.fensalir.fr/
    660 GPG: 0x82EB4DDF
    661 -------------- next part --------------
    662 A non-text attachment was scrubbed...
    663 Name: 0003-edit-mode-change-the-way-signatures-are-handled.patch
    664 Type: application/octet-stream
    665 Size: 3257 bytes
    666 Desc: not available
    667 URL: <http://rubyforge.org/pipermail/sup-devel/attachments/20100609/aca1d795/attachment.obj>
    668 
    669 From sean@silentflame.com  Wed Jun  9 13:42:08 2010
    670 From: sean@silentflame.com (Sean Whitton)
    671 Date: Wed, 9 Jun 2010 18:42:08 +0100
    672 Subject: [sup-devel] Storing message tags and other Sup info as headers
    673  in Maildir
    674 In-Reply-To: <AANLkTikI63bB_jx3EJsJ15iObHWLZJf_x0d8iw-gDrXP@mail.gmail.com>
    675 References: <20100604134404.GA28767@artemis.silentflame.com>
    676 	<AANLkTikI63bB_jx3EJsJ15iObHWLZJf_x0d8iw-gDrXP@mail.gmail.com>
    677 Message-ID: <20100609174207.GA22445@artemis.silentflame.com>
    678 
    679 Hi Matti,
    680 
    681 Thanks for the informative reply.  IMAP user-defined flags sound like an 
    682 excellent way to go about this, but of course, not everyone uses dovecot 
    683 (dunno why!).  Let me know how you get on.
    684 
    685 On Tue, Jun 08, 2010 at 06:14:08AM +0300, Matti Eiden wrote:
    686 > I was expecting somebody else to reply in this but since I'm not
    687 > seeing anyone doing so, I'll give it a shot. I only recently
    688 > discovered sup, and I'm not a Ruby programmer, however I got quite a
    689 > bunch of experience with Python (now the rest of the mailing list can
    690 > proceed flaming me). The following post is purely speculative.
    691 > 
    692 > Shortly put, I don't expect to see any change on this matter, at least
    693 > in the near future. There is and has been a tool, sup-sync-back which
    694 > should reflect the changes back to mbox/maildir but the way it works
    695 > is far from ideal, I guess. Anyway..
    696 
    697 I keep hearing about sup-sync-back but I'm not sure really what it does.  
    698 Where can I find information about it?
    699 
    700 S
    701 
    702 -- 
    703 Sean Whitton / <sean at silentflame.com>
    704 OpenPGP KeyID: 0x3B6D411B
    705 http://seanwhitton.com/
    706 
    707 -------------- next part --------------
    708 A non-text attachment was scrubbed...
    709 Name: not available
    710 Type: application/pgp-signature
    711 Size: 836 bytes
    712 Desc: Digital signature
    713 URL: <http://rubyforge.org/pipermail/sup-devel/attachments/20100609/8fdaecac/attachment.bin>
    714 
    715 From gregor@hoffleit.de  Thu Jun 10 09:05:07 2010
    716 From: gregor@hoffleit.de (Gregor Hoffleit)
    717 Date: Thu, 10 Jun 2010 15:05:07 +0200
    718 Subject: [sup-devel] Bug with searching for multiple labels at once
    719 In-Reply-To: <1276171687-sup-6993@sam.mediasupervision.de>
    720 References: <1276098490-sup-8738@sam.mediasupervision.de>
    721 	<1276170636-sup-6551@r61>
    722 	<1276171687-sup-6993@sam.mediasupervision.de>
    723 Message-ID: <1276175002-sup-9199@sam.mediasupervision.de>
    724 
    725 Indeed, I just noticed that the NewUserGuide.txt is much more up-to-date
    726 than the Wiki (http://sup.rubyforge.org/wiki/wiki.pl?SearchingMail).
    727 
    728 As a start, I copied the information from the NewUserGuide to that Wiki
    729 page. Still, somebody who knows better should update that page.
    730 Furthermore, I marked the information in "Testing Xapian" as obsolete in
    731 the Wiki.
    732 
    733 
    734 Still, it is not clear to me what happens without explicit AND or OR
    735 operator.
    736 
    737 "label:somelabel label:anotherlabel" seems to be equivalent to
    738 "label:somelabel OR label:anotherlabel".
    739 
    740 "label:ruby-talk subject:\[ANN\]" seem to be equivalent to
    741 "label:ruby-talk AND subject:\[ANN\]".
    742 
    743 The Xapian QueryParser documentation (http://xapian.org/docs/queryparser.html)
    744 is not clear about this either.
    745 
    746 
    747 Regards,
    748     Gregor Hoffleit
    749 
    750 From sup-bugs@masanjin.net  Sun Jun 13 03:11:19 2010
    751 From: sup-bugs@masanjin.net (anonymous)
    752 Date: Sun, 13 Jun 2010 07:11:19 +0000
    753 Subject: [sup-devel] [issue110] RuntimeError from thread: load threads
    754 	for	thread-index-mode
    755 In-Reply-To: <1276413079.42.0.0334970771348.issue110@masanjin.net>
    756 Message-ID: <1276413079.42.0.0334970771348.issue110@masanjin.net>
    757 
    758 
    759 New submission from anonymous:
    760 
    761 The first time I tried sup, with no mail in maildir, it opened without any problem.
    762 When I tried to open sup after fetchmail and procmail, with a few hundred
    763 messages, it crashed while it was opening, as shown in the attachment.
    764 
    765 ----------
    766 files: exception-log.txt
    767 messages: 252
    768 nosy: anonymous
    769 priority: bug
    770 ruby_version: ruby 1.8.7 (2009-06-12 patchlevel 174) [x86_64-linux]
    771 status: unread
    772 sup_version: v0.11
    773 title: RuntimeError from thread: load threads for thread-index-mode
    774 
    775 _________________________________________
    776 Sup issue tracker <sup-bugs at masanjin.net>
    777 <http://masanjin.net/sup-bugs/issue110>
    778 _________________________________________
    779 -------------- next part --------------
    780 --- RuntimeError from thread: load threads for thread-index-mode
    781 invalid source 2
    782 /usr/lib64/ruby/gems/1.8/gems/sup-0.11/lib/sup/index.rb:197:in `build_message'
    783 /usr/lib64/ruby/gems/1.8/gems/sup-0.11/lib/sup/index.rb:148:in `each_id_by_date'
    784 /usr/lib64/ruby/gems/1.8/gems/sup-0.11/lib/sup/thread.rb:332:in `call'
    785 /usr/lib64/ruby/gems/1.8/gems/sup-0.11/lib/sup/thread.rb:332:in `load_n_threads'
    786 /usr/lib64/ruby/gems/1.8/gems/sup-0.11/lib/sup/index.rb:148:in `each_id_by_date'
    787 /usr/lib64/ruby/gems/1.8/gems/sup-0.11/lib/sup/index.rb:239:in `each_id'
    788 /usr/lib64/ruby/gems/1.8/gems/sup-0.11/lib/sup/index.rb:239:in `each'
    789 /usr/lib64/ruby/gems/1.8/gems/sup-0.11/lib/sup/index.rb:239:in `each_id'
    790 /usr/lib64/ruby/gems/1.8/gems/sup-0.11/lib/sup/index.rb:148:in `each_id_by_date'
    791 /usr/lib64/ruby/gems/1.8/gems/sup-0.11/lib/sup/thread.rb:328:in `load_n_threads'
    792 /usr/lib64/ruby/gems/1.8/gems/sup-0.11/lib/sup/modes/thread-index-mode.rb:640:in `__unprotected_load_n_threads'
    793 (eval):12:in `load_n_threads'
    794 /usr/lib64/ruby/gems/1.8/gems/sup-0.11/lib/sup/modes/thread-index-mode.rb:624:in `load_n_threads_background'
    795 /usr/lib64/ruby/gems/1.8/gems/sup-0.11/lib/sup.rb:76:in `reporting_thread'
    796 /usr/lib64/ruby/gems/1.8/gems/sup-0.11/lib/sup.rb:74:in `initialize'
    797 /usr/lib64/ruby/gems/1.8/gems/sup-0.11/lib/sup.rb:74:in `new'
    798 /usr/lib64/ruby/gems/1.8/gems/sup-0.11/lib/sup.rb:74:in `reporting_thread'
    799 /usr/lib64/ruby/gems/1.8/gems/sup-0.11/lib/sup/modes/thread-index-mode.rb:623:in `load_n_threads_background'
    800 /usr/lib64/ruby/gems/1.8/gems/sup-0.11/lib/sup/modes/thread-index-mode.rb:694:in `__unprotected_load_threads'
    801 (eval):12:in `load_threads'
    802 /usr/lib64/ruby/gems/1.8/gems/sup-0.11/bin/sup:231
    803 /usr/bin/sup:19:in `load'
    804 /usr/bin/sup:19
    805 
    806 From michael+sup@stapelberg.de  Wed Jun 16 14:11:47 2010
    807 From: michael+sup@stapelberg.de (Michael Stapelberg)
    808 Date: Wed, 16 Jun 2010 20:11:47 +0200
    809 Subject: [sup-devel] =?utf-8?b?W1BBVENIXSBEb27igJl0IGRpc3BsYXkgIi4uLiIg?=
    810 	=?utf-8?q?after_snippets_which_are_displayed_completely?=
    811 Message-ID: <1276711781-sup-4102@midna.zekjur.net>
    812 
    813 Hi,
    814 
    815 quote from the commit message:
    816     Don?t display "..." after snippets which are displayed completely
    817     
    818     Short mails (for example: "Yes, the date works for me.") often can
    819     be displayed completely in the snippet. However, before this patch,
    820     sup abbreviated the snippet even though it was not abbreviated.
    821 
    822 
    823 Best regards,
    824 Michael
    825 -------------- next part --------------
    826 A non-text attachment was scrubbed...
    827 Name: 0001-Don-t-display-.-after-snippets-which-are-displayed-c.patch
    828 Type: application/octet-stream
    829 Size: 2085 bytes
    830 Desc: not available
    831 URL: <http://rubyforge.org/pipermail/sup-devel/attachments/20100616/1123bd1c/attachment.obj>
    832 
    833 From michael+sup@stapelberg.de  Tue Jun 22 11:40:45 2010
    834 From: michael+sup@stapelberg.de (Michael Stapelberg)
    835 Date: Tue, 22 Jun 2010 17:40:45 +0200
    836 Subject: [sup-devel] [PATCH] inline-gpg: call text_to_chunks on the text
    837 	before/after the GPG part
    838 Message-ID: <1277221190-sup-946@midna.zekjur.net>
    839 
    840 Hi,
    841 
    842 quote from the commit message:
    843 
    844     inline-gpg: call text_to_chunks on the text before/after the GPG part
    845         
    846     This is necessary for stupid mailers which produce TOFU mails
    847     containing unquoted inline gpg mails *argh*.
    848 
    849 
    850 Best regards,
    851 Michael
    852 -------------- next part --------------
    853 A non-text attachment was scrubbed...
    854 Name: 0001-inline-gpg-call-text_to_chunks-on-the-text-before-af.patch
    855 Type: application/octet-stream
    856 Size: 2719 bytes
    857 Desc: not available
    858 URL: <http://rubyforge.org/pipermail/sup-devel/attachments/20100622/2456115a/attachment-0001.obj>
    859 
    860 From sascha-pgp@silbe.org  Tue Jun 29 03:50:25 2010
    861 From: sascha-pgp@silbe.org (Sascha Silbe)
    862 Date: Tue, 29 Jun 2010 09:50:25 +0200
    863 Subject: [sup-devel] [PATCH] fix reference to EncodingUnsupportedError
    864 Message-ID: <1277797825-8181-1-git-send-email-sascha-pgp@silbe.org>
    865 
    866 This fixes the following error:
    867 
    868 ./lib/sup/message.rb:473:in `message_to_chunks': uninitialized constant Redwood::Message::EncodingUnsupportedError (NameError)
    869 
    870 Signed-off-by: Sascha Silbe <sascha-pgp at silbe.org>
    871 ---
    872  lib/sup/message.rb |    2 +-
    873  1 files changed, 1 insertions(+), 1 deletions(-)
    874 
    875 diff --git a/lib/sup/message.rb b/lib/sup/message.rb
    876 index 4e6bbeb..1385585 100644
    877 --- a/lib/sup/message.rb
    878 +++ b/lib/sup/message.rb
    879 @@ -470,7 +470,7 @@ private
    880          when "7bit", "8bit", nil
    881            m.body
    882          else
    883 -          raise EncodingUnsupportedError, encoding.inspect
    884 +          raise RMail::EncodingUnsupportedError, encoding.inspect
    885          end
    886          body = body.normalize_whitespace
    887          payload = RMail::Parser.read(body)
    888 -- 
    889 1.6.5
    890 
    891 
    892 From sascha-pgp@silbe.org  Tue Jun 29 04:04:54 2010
    893 From: sascha-pgp@silbe.org (Sascha Silbe)
    894 Date: Tue, 29 Jun 2010 10:04:54 +0200
    895 Subject: [sup-devel] [PATCH] Don't choke when scanning message with unknown
    896 	encoding
    897 Message-ID: <1277798694-8642-1-git-send-email-sascha-pgp@silbe.org>
    898 
    899 This fixes the following error:
    900 
    901 ./lib/sup/message.rb:473:in `message_to_chunks': "7BIT" (RMail::EncodingUnsupportedError)
    902 
    903 when running sup-sync on a folder that contains a mail with these headers:
    904 
    905 Content-Transfer-Encoding: 7bit
    906 X-Mime-Autoconverted: from 8bit to 7bit by courier 0.60
    907 
    908 Signed-off-by: Sascha Silbe <sascha-pgp at silbe.org>
    909 ---
    910  lib/sup/message.rb |    2 +-
    911  1 files changed, 1 insertions(+), 1 deletions(-)
    912 
    913 diff --git a/lib/sup/message.rb b/lib/sup/message.rb
    914 index 1385585..a13fc0c 100644
    915 --- a/lib/sup/message.rb
    916 +++ b/lib/sup/message.rb
    917 @@ -259,7 +259,7 @@ class Message
    918            rmsg = source.load_message(source_info)
    919            parse_header rmsg.header
    920            message_to_chunks rmsg
    921 -        rescue SourceError, SocketError => e
    922 +        rescue SourceError, SocketError, RMail::EncodingUnsupportedError => e
    923            warn "problem getting messages from #{source}: #{e.message}"
    924            ## we need force_to_top here otherwise this window will cover
    925            ## up the error message one
    926 -- 
    927 1.6.5
    928 
    929 
    930 From sascha-pgp@silbe.org  Tue Jun 29 04:12:05 2010
    931 From: sascha-pgp@silbe.org (Sascha Silbe)
    932 Date: Tue, 29 Jun 2010 10:12:05 +0200
    933 Subject: [sup-devel] [PATCH] fix crash in sup-dump if the default sent
    934 	source is used
    935 Message-ID: <1277799125-8696-1-git-send-email-sascha-pgp@silbe.org>
    936 
    937 This fixes a crash in sup-dump if the index contains a "sent" message and
    938 no "sent" folder has been explicitly configured in the config file
    939 (so it hasn't been added to sources.yaml).
    940 
    941 Signed-off-by: Sascha Silbe <sascha-pgp at silbe.org>
    942 ---
    943  bin/sup-dump |    7 +++++++
    944  1 files changed, 7 insertions(+), 0 deletions(-)
    945 
    946 diff --git a/bin/sup-dump b/bin/sup-dump
    947 index d6a06e4..953b6e5 100755
    948 --- a/bin/sup-dump
    949 +++ b/bin/sup-dump
    950 @@ -19,9 +19,16 @@ Usage:
    951  EOS
    952  end
    953  
    954 +$config = Redwood::load_config Redwood::CONFIG_FN
    955  index = Redwood::Index.init
    956  Redwood::SourceManager.init
    957  index.load
    958 +Redwood::SentManager.init $config[:sent_source] || 'sup://sent'
    959 +if(s = Redwood::SourceManager.source_for Redwood::SentManager.source_uri)
    960 +  Redwood::SentManager.source = s
    961 +else
    962 +  Redwood::SourceManager.add_source Redwood::SentManager.default_source
    963 +end
    964  
    965  index.each_message :load_spam => true, :load_deleted => true, :load_killed => true do |m|
    966    puts "#{m.id} (#{m.labels.to_a.sort_by { |l| l.to_s } * ' '})"
    967 -- 
    968 1.6.5
    969 
    970 
    971 From sascha-ml-email-sup-devel@silbe.org  Tue Jun 29 04:54:22 2010
    972 From: sascha-ml-email-sup-devel@silbe.org (Sascha Silbe)
    973 Date: Tue, 29 Jun 2010 08:54:22 +0000
    974 Subject: [sup-devel] Will sup blow up after adding 10k sources?
    975 Message-ID: <1277801600-sup-3476@twin.sascha.silbe.org>
    976 
    977 Hi!
    978 
    979 While fixing the sup-dump vs. default sent source bug (see posted patch) I stumbled over the following pieces of code:
    980 1. SentManager uses source_id of 9998 for the default sent source
    981    (lib/sup/sent.rb:52)
    982 2. DraftManager uses a fixed source_id of 9999
    983    (lib/sup/draft.rb:13,49)
    984 3. SourceManager skips the source_id of SentManager and DraftManager
    985    while computing the id for a new source (lib/sup/source.rb:203)
    986 
    987 Does that mean sup will blow up once I have more than ~10k sources? I'm already at > 5k sources (containing > 680k messages, took > 2 days to import), so this isn't just an academic exercise...
    988 
    989 Wouldn't it be better to (*) use small fixed source ids for SentManager/DraftManager and let SourceManager start above a number of reserved ids (say: 50)? Is there something in sup that depends on the non-special source ids not to have "holes" (or will waste resources in that case)?
    990 
    991 Another approach would be to make the SentManager/DraftManager source ids dynamic and add them to sources.yaml.
    992 
    993 Sascha
    994 
    995 (*) Read: Would you accept a patch that implements that new logic?
    996 
    997 --
    998 http://sascha.silbe.org/
    999 http://www.infra-silbe.de/
   1000 -------------- next part --------------
   1001 A non-text attachment was scrubbed...
   1002 Name: signature.asc
   1003 Type: application/pgp-signature
   1004 Size: 490 bytes
   1005 Desc: not available
   1006 URL: <http://rubyforge.org/pipermail/sup-devel/attachments/20100629/a6666680/attachment.bin>
   1007 
   1008 From bwalton@artsci.utoronto.ca  Tue Jun 29 09:15:56 2010
   1009 From: bwalton@artsci.utoronto.ca (Ben Walton)
   1010 Date: Tue, 29 Jun 2010 09:15:56 -0400
   1011 Subject: [sup-devel] Will sup blow up after adding 10k sources?
   1012 In-Reply-To: <1277801600-sup-3476@twin.sascha.silbe.org>
   1013 References: <1277801600-sup-3476@twin.sascha.silbe.org>
   1014 Message-ID: <1277817081-sup-1172@pinkfloyd.chass.utoronto.ca>
   1015 
   1016 Excerpts from Sascha Silbe's message of Tue Jun 29 04:54:22 -0400 2010:
   1017 
   1018 Hi Sascha,
   1019 
   1020 > Wouldn't it be better to (*) use small fixed source ids for
   1021 > SentManager/DraftManager and let SourceManager start above a number
   1022 > of reserved ids (say: 50)? Is there something in sup that depends on
   1023 > the non-special source ids not to have "holes" (or will waste
   1024 > resources in that case)?
   1025 
   1026 I agree that this would have been a better choice early on.  Assign
   1027 the lowest possible id's to the Sent/Draft sources and then hand out
   1028 id's dynmaically from there.
   1029 
   1030 > Another approach would be to make the SentManager/DraftManager
   1031 > source ids dynamic and add them to sources.yaml.
   1032 
   1033 I'd have to look, but changing these id's after they've been used will
   1034 cause problems requiring a resync, I believe.
   1035 
   1036 I'm sure others that have been more active in this part of the source
   1037 lately can provide more info too.
   1038 
   1039 HTH.
   1040 -Ben
   1041 --
   1042 Ben Walton
   1043 Systems Programmer - CHASS
   1044 University of Toronto
   1045 C:416.407.5610 | W:416.978.4302
   1046 
   1047