Timeline


and

Feb 10, 2011:

11:59 PM Changeset [ab9bf01]release-1.10release-1.8release-1.9 by Anders Kaseorg <andersk@mit.edu>
Disable special handling of START and STOP characters (^Q and ^S) Signed-off-by: Anders Kaseorg <andersk@mit.edu>
11:59 PM Changeset [7488f27]release-1.10release-1.8release-1.9 by Anders Kaseorg <andersk@mit.edu>
Use STDIN_FILENO instead of 0 Signed-off-by: Anders Kaseorg <andersk@mit.edu>
9:10 PM Changeset [edd0be7]release-1.10release-1.8release-1.9 by David Benjamin <davidben@mit.edu>
Only handle CCs in messages sent directly to you. A zephyr with a CC line received on @SOME-REALM would still handle the CC logic in replycmd. This oddity could possibly be retrofited to allow @SOME-REALM to participate in CC'd messages, but that would require us to handle CC lines on messages received on *. In the meantime, let's not leave a half-baked not-feature. Also refactor the block slightly. The original formulation is confusing; it means to condition on the recipient of the message we reply to, but actually checks $to, the recipient we use in the reply. It just happens that in one case, the two are equal. The original code would assume that case and then change its mind if it was wrong.
9:08 PM Changeset [3f82515]release-1.10release-1.8release-1.9 by David Benjamin <davidben@mit.edu>
Rewrite aimwrite This now makes more gratuitous use of glib and should, as a bonus, be more sensible. In particular, we now consistently handle multi-word recipients with and without -m. Also avoid duplicating sending code.

Feb 9, 2011:

10:59 PM Changeset [6965867]release-1.10release-1.8release-1.9 by David Benjamin <davidben@mit.edu>
g_ptr_array_new_with_free_func is glib 2.22+ Use g_ptr_array_foreach instead. Reported-by: Adam Glasgall <adam@crossproduct.net>

Feb 8, 2011:

2:27 AM Changeset [8061a6a]release-1.10release-1.8release-1.9 by Anders Kaseorg <andersk@mit.edu>
Move glib libraries from LDFLAGS to LIBS This fixes linker errors on wheezy and natty: http://wiki.debian.org/ToolChain/DSOLinking Signed-off-by: Anders Kaseorg <andersk@mit.edu> Reviewed-by: David Benjamin <davidben@mit.edu>
12:39 AM 256Colors edited by andersk@mit.edu
Debian screen-256color underline bug (diff)

Feb 7, 2011:

9:21 PM Ticket #158 (./configure --without-zephyr && make fails with undefined reference to ...) closed by jgross@mit.edu
fixed: Fixed by commit f85278b8943b3c38ffc7ed08dfb9869f6cb7f1b9

Feb 6, 2011:

1:11 AM WikiStart edited by davidben@mit.edu
(diff)
1:11 AM release-notes/1.7.1 edited by davidben@mit.edu
(diff)
1:10 AM release-notes/1.7.1 created by davidben@mit.edu

Feb 5, 2011:

11:57 PM Changeset [63d9de9]release-1.7 by David Benjamin <davidben@mit.edu>
BarnOwl 1.7.1
11:54 PM Changeset [c268c9e]release-1.7 by David Benjamin <davidben@mit.edu>
Update do-release's output to match the wiki The actual wording we use has some differences in capitalization, etc.
11:52 PM Changeset [27705b65]release-1.7 by David Benjamin <davidben@mit.edu>
Append sender's realm to Cc'd unqualified names when replying Currently, if you reply to a personal from a user in another realm, that message was Cc'd to other people on their realm, and the sender didn't specify their realm, Barnowl will try to Cc the message to users of the same name in your realm. So, if you are 'bar@EXAMPLE.ORG' and you get a message from 'foo@EXAMPLE.COM' cc'd to 'bletch', if you try to reply, Barnowl will address it to foo@EXAMPLE.COM, cc'd to 'bletch', which would be bletch@EXAMPLE.ORG, which is clearly wrong. This patch appends the sender's realm to unqualified names in the message's Cc line when generating the reply zwrite command line. [davidben@mit.edu: Fix up some edge case handling]
11:28 PM Changeset [8d553bf]release-1.10release-1.8release-1.9 by David Benjamin <davidben@mit.edu>
Append sender's realm to Cc'd unqualified names when replying Currently, if you reply to a personal from a user in another realm, that message was Cc'd to other people on their realm, and the sender didn't specify their realm, Barnowl will try to Cc the message to users of the same name in your realm. So, if you are 'bar@EXAMPLE.ORG' and you get a message from 'foo@EXAMPLE.COM' cc'd to 'bletch', if you try to reply, Barnowl will address it to foo@EXAMPLE.COM, cc'd to 'bletch', which would be bletch@EXAMPLE.ORG, which is clearly wrong. This patch appends the sender's realm to unqualified names in the message's Cc line when generating the reply zwrite command line. [davidben@mit.edu: Fix up some edge case handling]

Feb 4, 2011:

5:33 PM Changeset [22662d7]release-1.7 by David Benjamin <davidben@mit.edu>
Don't reset colorpairs in the middle of drawing Resetting colorpairs while drawing the mainwin causes the existing contents in a popwin to refer to invalid color pairs. We used to draw the mainwin first and redraw the contents of each window from scratch. Moving to libpanel in 1.6 changed this, so background colors occasionally bled into your popwin. This changes the colorpair logic to only trigger when we need to, and to forcibly dirty every window if needed. NOTE: if we don't have enough color-pairs to draw the current screen, this will draw everything twice. But it will probably almost never happen. Reported-by: Alex Dehnert <adehnert@mit.edu> Reviewed-by: Alejandro R. Sedeño <asedeno@mit.edu>
5:33 PM Changeset [b1ed3e2]release-1.7 by David Benjamin <davidben@mit.edu>
Refactor some common code in colorpair_mgr Reviewed-by: Alejandro R. Sedeño <asedeno@mit.edu>
5:32 PM Changeset [099597c]release-1.10release-1.8release-1.9 by David Benjamin <davidben@mit.edu>
Don't reset colorpairs in the middle of drawing Resetting colorpairs while drawing the mainwin causes the existing contents in a popwin to refer to invalid color pairs. We used to draw the mainwin first and redraw the contents of each window from scratch. Moving to libpanel in 1.6 changed this, so background colors occasionally bled into your popwin. This changes the colorpair logic to only trigger when we need to, and to forcibly dirty every window if needed. NOTE: if we don't have enough color-pairs to draw the current screen, this will draw everything twice. But it will probably almost never happen. Reported-by: Alex Dehnert <adehnert@mit.edu> Reviewed-by: Alejandro R. Sedeño <asedeno@mit.edu>
5:32 PM Changeset [e062f97]release-1.10release-1.8release-1.9 by David Benjamin <davidben@mit.edu>
Refactor some common code in colorpair_mgr Reviewed-by: Alejandro R. Sedeño <asedeno@mit.edu>
3:49 PM Changeset [4d24650]release-1.10release-1.8release-1.9 by David Benjamin <davidben@mit.edu>
Make owl_text_indent return a new string You can't really use it correctly as it is.
3:49 PM Changeset [e3c8332]release-1.10release-1.8release-1.9 by David Benjamin <davidben@mit.edu>
Use GString in owl_cmddict_execute_argv While we're at it, provide a correctly quoted command-line.
3:49 PM Changeset [c6b1782]release-1.10release-1.8release-1.9 by David Benjamin <davidben@mit.edu>
Use g_strjoinv in parsing zwrite lines
3:49 PM Changeset [0290b2a]release-1.10release-1.8release-1.9 by David Benjamin <davidben@mit.edu>
Use GString instead of allocating a giant buffer in owl_parseline
3:49 PM Changeset [df3a1f4]release-1.10release-1.8release-1.9 by David Benjamin <davidben@mit.edu>
Replace realloc logic in owl_zephyr_getsubs with GString
3:49 PM Changeset [9de316d1]release-1.10release-1.8release-1.9 by David Benjamin <davidben@mit.edu>
No need to NULL-terminate argument vectors owl_parseline does it for us now.
3:49 PM Changeset [7e111f4]release-1.10release-1.8release-1.9 by David Benjamin <davidben@mit.edu>
Replace owl_fmtext's buffer management with GString reallocs are hard. Let's go shopping^Wuse GString!
3:49 PM Changeset [d3c318b]release-1.10release-1.8release-1.9 by David Benjamin <davidben@mit.edu>
Don't stick random NULs when waddnstr exists
3:49 PM Changeset [dfe5829]release-1.10release-1.8release-1.9 by David Benjamin <davidben@mit.edu>
Be more accurate in allocating argument vector in owl_command_view 50 is a pretty random number.
3:49 PM Changeset [fc7481a]release-1.10release-1.8release-1.9 by David Benjamin <davidben@mit.edu>
strndup is pretty awesome and doesn't involve adding 5
3:49 PM Changeset [65c753e]release-1.10release-1.8release-1.9 by David Benjamin <davidben@mit.edu>
Use a GPtrArray to manage the growing list in owl_parseline It's much nicer than dealing with piles of owl_realloc. Also ensure that our argvs are NULL-terminated so we may used the various *v glib functions.
3:49 PM Changeset [d3941a0]release-1.10release-1.8release-1.9 by David Benjamin <davidben@mit.edu>
Use g_strjoinv in owl_global_set_startupargs It's much nicer than dealing with string lengths, although it does mean we lose const-correctness because glib just gives up on that with lists of strings. Also make a copy of the argument list because getopt_long messes around with argument lists anyway.

Jan 26, 2011:

1:31 AM Jabber edited by davidben@mit.edu
Fix up formatting a bit (diff)

Jan 25, 2011:

1:08 AM Changeset [337383e]release-1.7 by David Benjamin <davidben@mit.edu>
Update copyright notices for 2011
1:07 AM Changeset [b03c714]release-1.10release-1.8release-1.9 by David Benjamin <davidben@mit.edu>
Update copyright notices for 2011

Jan 24, 2011:

10:57 PM Changeset [9a3cdca]release-1.7 by David Benjamin <davidben@mit.edu>
Stick the local realm onto the argument of viewuser if needed A regression caused by 811644f3912f5278edb8dcb964547f041976150e. Now that owl_function_zuserfilt expects a longuser, its callers need to take care of adding the realm where needed. Reported-by: Benjamin Kaduk <kaduk@mit.edu> Reviewed-by: Alejandro R. Sedeño <asedeno@mit.edu>
10:57 PM Changeset [e06f413]release-1.10release-1.8release-1.9 by David Benjamin <davidben@mit.edu>
Stick the local realm onto the argument of viewuser if needed A regression caused by 811644f3912f5278edb8dcb964547f041976150e. Now that owl_function_zuserfilt expects a longuser, its callers need to take care of adding the realm where needed. Reported-by: Benjamin Kaduk <kaduk@mit.edu> Reviewed-by: Alejandro R. Sedeño <asedeno@mit.edu>

Jan 23, 2011:

3:27 AM Changeset [74312ad]release-1.10release-1.8release-1.9 by David Benjamin <davidben@mit.edu>
Update do-release's output to match the wiki The actual wording we use has some differences in capitalization, etc.
3:27 AM WikiStart edited by davidben@mit.edu
Punt a bunch of needless caps while I'm at it (diff)
3:12 AM WikiStart edited by davidben@mit.edu
drop some caps (diff)

Jan 20, 2011:

8:07 PM WikiStart edited by davidben@mit.edu
(diff)
8:02 PM release-notes/1.7 created by davidben@mit.edu
7:59 PM Changeset [42115bf9]release-1.10release-1.8release-1.9 by David Benjamin <davidben@mit.edu>
Reimplement owl_function_typefilt with proper quoting Probably not actually a big deal, but may as well. [nelhage@nelhage.com: Use owl_string_build_quoted]
7:59 PM Changeset [9d4c398]release-1.10release-1.8release-1.9 by David Benjamin <davidben@mit.edu>
Reimplement owl_function_zpunt using proper quoting [nelhage@nelhage.com: Use owl_string_build_quoted]
7:59 PM Changeset [a5f477c]release-1.10release-1.8release-1.9 by David Benjamin <davidben@mit.edu>
Reimplement owl_function_zuserfilt with proper quoting [nelhage@nelhage.com: Use owl_string_appendf_quoted]
7:59 PM Changeset [c426bc2]release-1.10release-1.8release-1.9 by David Benjamin <davidben@mit.edu>
Rewrite owl_function_classinstfilt with proper quoting The current ad-hoc code fails to handle tabs, which allows an attacker to force a failed parse or run an arbitrary perl function. [nelhage@nelhage.com: Use owl_string_appendf_quoted]
7:59 PM Changeset [f47696f]release-1.10release-1.8release-1.9 by David Benjamin <davidben@mit.edu>
Add owl_string_appendf_quoted for easy construction of command lines.
7:59 PM Changeset [443dcfa]release-1.10release-1.8release-1.9 by David Benjamin <davidben@mit.edu>
Add a test for zuserfilt
7:59 PM Changeset [b31252d]release-1.10release-1.8release-1.9 by David Benjamin <davidben@mit.edu>
Add a few unit tests for classinstfilt They currently fail because our quoting is buggy.
7:59 PM Changeset [c7fe23e]release-1.10release-1.8release-1.9 by David Benjamin <davidben@mit.edu>
If a smartfilter fails to parse, handle the error Otherwise we segfault. Of course, we should always succeed in creating a filter, but we sometimes don't because of quoting bugs.
6:04 PM Changeset [b35144c]release-1.7 by David Benjamin <davidben@mit.edu>
BarnOwl 1.7
Note: See TracTimeline for information about the timeline view.