Menu
▾
▴
erc-help — Archive of problems, bugs, features, news
You can subscribe to this list here.
| 2001 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
(10) |
Nov
(46) |
Dec
(2) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2002 |
Jan
(5) |
Feb
|
Mar
(6) |
Apr
(2) |
May
(25) |
Jun
(2) |
Jul
(1) |
Aug
(5) |
Sep
(4) |
Oct
(7) |
Nov
|
Dec
(18) |
| 2003 |
Jan
(8) |
Feb
(1) |
Mar
(2) |
Apr
(4) |
May
(14) |
Jun
(32) |
Jul
(15) |
Aug
(23) |
Sep
(23) |
Oct
(22) |
Nov
(27) |
Dec
(24) |
| 2004 |
Jan
(18) |
Feb
(38) |
Mar
(32) |
Apr
(18) |
May
(70) |
Jun
(1) |
Jul
(21) |
Aug
(19) |
Sep
(32) |
Oct
(11) |
Nov
(11) |
Dec
(19) |
| 2005 |
Jan
(48) |
Feb
(13) |
Mar
(19) |
Apr
(25) |
May
(4) |
Jun
(23) |
Jul
(8) |
Aug
(13) |
Sep
(12) |
Oct
(17) |
Nov
(4) |
Dec
(5) |
| 2006 |
Jan
(31) |
Feb
(30) |
Mar
(28) |
Apr
(11) |
May
(21) |
Jun
(7) |
Jul
(12) |
Aug
(5) |
Sep
(7) |
Oct
(24) |
Nov
(2) |
Dec
|
|
From: Andreas F. <as...@vo...> - 2001-11-24 16:37:07
|
Today, Gergely Nagy <alg...@mi...> wrote: > What I did, was to modify erc-save-buffer-in-logs, to call a custom > defun, which will compute the actual filename. The only problem is > that this does not work if the logs are not saved through this > function, so, the buffer filenames should be set using this > customisable function, I think. I have modified the patch a little to conform more to naming conventions (-function, not -p which would make it a predicate), and also added the session's port to the function parameters. Your modified patch is below (with you added to the contributor's list): For my patch (see my previous message to this list): |
|
From: Gergely N. <alg...@bo...> - 2001-11-24 16:21:41
|
Greetings!
The hard-coded filenames of the ERC logs bothered me for some time, so
I sat down and wrote a smallish patch, which partly fixes the
problem.
What I did, was to modify erc-save-buffer-in-logs, to call a custom
defun, which will compute the actual filename. The only problem is
that this does not work if the logs are not saved through this
function, so, the buffer filenames should be set using this
customisable function, I think.
But that's way beyond my lisp knowledge. Anyway, the patch against
last nights CVS is attached below.
--- erc.el.orig Sat Nov 24 02:24:25 2001
+++ erc.el Sat Nov 24 14:49:23 2001
@@ -437,6 +437,17 @@
:group 'erc
:type 'boolean)
+(defcustom erc-generate-log-file-name-p 'erc-generate-log-file-name-default
+ "A function name to generate a log filename. The function must take
+four arguments: BUFFER, TARGET, NICK and SERVER. BUFFER is the buffer
+to be saved, TARGET is the name of the channel, or the target of the
+query, NICK is the current nick, and SERVER is the server the buffer
+is on."
+ :group 'erc
+ :type '(choice (const 'erc-generate-log-file-name-default)
+ (const 'erc-generate-log-file-name-old)
+ (function-item)))
+
(defvar erc-grab-buffer-name "*erc-grab*"
"The name of the buffer created by `erc-grab-region'.")
@@ -5181,6 +5192,13 @@
(buffer-enable-undo))
(set-buffer ob)))
+(defun erc-generate-log-file-name-default (buffer target nick server)
+ "Default log filename generator."
+ (concat erc-log-channels-directory "/" (buffer-name buffer) ".txt"))
+(defun erc-generate-log-file-name-old (buffer target nick server)
+ "Old-style (#channel!nick@server.txt) log filename generator"
+ (concat erc-log-channels-directory "/" target "!" nick "@" server ".txt"))
+
(defun erc-save-buffer-in-logs (&optional buffer)
"When the logs are enabled, that is `erc-log-channels' is non-nil
and `erc-log-channels-directory' is a valid directory, appends the
@@ -5198,8 +5216,9 @@
erc-log-channels-directory
(erc-directory-writable-p erc-log-channels-directory))
(append-to-file (point-min) (point-max)
- (concat erc-log-channels-directory
- "/" (buffer-name buffer) ".txt"))
+ (eval (list erc-generate-log-file-name-p buffer
+ (erc-default-target) (erc-current-nick)
+ erc-announced-server-name)))
(erase-buffer)
(erc-display-prompt)
(goto-char (point-max)))
|
|
From: Andreas F. <as...@vo...> - 2001-11-24 16:05:05
|
<multipart sign=pgpmime> Hi all, I just created a patch containing a great lot of fixes for bad behavior, and nice new features as well. This is it: |
|
From: Mario L. <ml...@ho...> - 2001-11-23 16:53:30
|
sm...@wa... (Tijs van Bakel) writes: > Ok here is my enormous list of bugs and ideas. and here is a review, with done items marked. > > erc-switchb: > - An alternative to iswitchb-buffer for erc buffers only. The > following snippet shows the basic action, but the buffer names are too > long, and I see no easy way to make iswitchb use shorter alternatives > (iswitchb-read-buffer would return the abbreviated buffername, which > has to be mapped back again; this is A possibility, but perhaps there > are better alternatives?). See also the next point about buffer > names. This is in cvs, and works, sort of. Only rpoblem is that the user currently has to use iswitchb-default-keybindings, otherwise it doesnt work, and I still have no idea why. > Buffer names: > - Buffer names could perhaps be abbreviated to "#emacs, #emacs<2>, > #lisp, MisterX" to denote #emacs on 2 different servers, and a query > to MisterX. This would be useful both for switching and display in > the mode-line. I have not found a convincing method on how to exactly > name the buffers, though. Done. finished, working, rocking! > > Buffer list: > - There is now a list of buffers _per server_. It seems better to > have one global list for all buffers. It would be nice to make this a > property or alist, so that more data could be put in there. Think > about flags like "modified", "someone sent you a personal message > there", .. Don.e the function (erc-buffer-list) can be used to generate either a list of all buffers on all servers, or all buffers related to a specific process (server). > Inconsistency in feedback of input text: > - A message makes the entry prompt disappear, but for a /ME command it > doesn't. Should be easy to fix, but perhaps there's a reason for this > behaviour? antifuchs also reminded me of that today. Not done anything yet. > Timestamps: > - The code for adding of timestamps seems a bit messy. I > found this snippet repeated many times throughout the code: > > (if erc-timestamp-format > (format-time-string erc-timestamp-format > (current-time)) "") > > This should obviously be taken into a function. Perhaps it would be > better to have a hook to reformat a line of text though? Also, a > timestamp is not prepended to messages sent by yourself, nor for > server messages (topics &c.) I think other irc clients have a > timestamp-level value to select which messages should get a timestamp, > and which shouldn't. > [Note: I have done a bit of work on this already, but it doesn't > support timestamp levels yet. A generalization would be nice.] right, its better than it was before though already... > > Scrolling: > - "M-x term" manages to keep the entire window in use, having the > input line near the bottom of the window. This seems to be > implemented by having a _lot_ of (recenter (1- term-height)) calls all > over term.el. Not the most elegant method, but it is effective. > Could a somewhat cleaner method be used for erc, or should I just add > recenter calls everywhere where needed? I think you have some local code for that? > > Pop-up buffers: > - On incoming messages, with auto-query enabled, the query pops up, > and interrupts my typing. It'd be nice to have these buffers buried > by default (the mode-line change is enough). Nice customization > option -- seems I could hack this in somewhere. > [Note: Fixed, by adding a "bury" option to erc-join-buffer. It might > be nice to rename erc-join-buffer to erc-new-buffer.] done (by you) > > Persistent buffers: > - It would be nice to be able to save (C-x C-s) a buffer log, and > continue with it later. A message like ">> disconnected" could mark > where one picked up again. Something analogous like (desktop-read) antifuchs has done some log/save related things today, I didnt see the patch yet, so I have no idea if this is already covered in some way. > Nick change tracking in queries: > - Subject says it all :-) It seems that MIRC and some other irc > clients are capable of doing this, but I don't know the relevant > commands to send to a server in order to be notified of nick changes. > "Someone should research this" yow, thats still a problem. And I have no idea why you would sent to the server. If you are on the same channels as the other guy, it is obviously no problem... > Tracking channels [,#bug]: > - The extra comma is an obvious bug. It seems better (at least to me) > to have a _list_ of buffers that are changed, instead of using a > _string_ to store the changed buffers. When updating the modeline, > the list can be processed to produce the string, but the other way > around seems to be cumbersome, the current method is prone to errors. Still not fixed. > Misc.: > > - When quoting using "\C-RET", leading spaces should not be stripped > (try quoting a lisp function for instance) hmm, did you fix that. I didnt see that problem lately. > > - Transform "/join elisp" into "/join #elisp". Done (by you). > - Mimick the /quote feature of TinyFugue. I dont know what this means. > - Give some feedback when signing off. right, that would be useful... > > - DCC (zenirc seems to support dcc get, and dcc chat for incoming > chats) How would one interface with a subprocess in a compatible > way? Yow, dcc would be good, I read zenirc-dcc.el and it doesnt look to hard. One thing I would like to acomplish differently is that I'd like to have DCC as minor-mode for erc. zenirc currently implements a completely new major mode, and this looks like duplication to me. > > - Unset a topic > > - A way to toggle "/" into some other character. Sometimes one wants > to type text starting with a "/" character. really, it shouldnt be to hard, but do you really think people would use/need that? -- CYa, Mario <ml...@de...> Homepage(s): http://delysid.org | http://piss.at/ I think it's a new feature. Don't tell anyone it was an accident. :-) -- Larry Wall on s/foo/bar/eieio in <10...@jp...> |
|
From: Adam D. <du...@in...> - 2001-11-19 21:34:25
|
Hehe... I'm sorry but removing "(open-line 1)" from that position is displaying messages from other persons in the same line as the erc-prompt. This is no good :-(. One should check twice... sorry again. Bye. Adam. -- Adam Duck (du...@cs...) Bockenheimer Landstr. 135 / Zi. 211 60325 Frankfurt/Main ______________________________________ If you can spend a perfectly useless afternoon in a perfectly useless manner, you have learned how to live. - Lin Yutang |
|
From: Adam D. <du...@in...> - 2001-11-19 21:03:27
|
Well, there is a bug in ERC under XEmacs that puts a newline before erc-input-prompt every time I send a [return]. The code in question is: --- (let ((insert-end-position (point))) (goto-char (point-max)) (open-line 1) (goto-char (point-max)) --- I don't know why - because all the functions in this code are described to do the same in Emacs21 and XEmacs - but it only happens in XEmacs. Removing "(open-line 1)" seems to solve the problem - but it has, of course, unwanted effects on the Emacs-side. Well, I tried to do an "(insert "here")" before and after the `open-line' and they all end up at the end of the line in Emacs and in a new line in XEmacs. Is this my own setup or is somebody else seeing this? And: can someone add some code to fix this? In the sample configuration files they go like this: (defvar running-xemacs (string-match "XEmacs\\|Lucid" emacs-version)) I know this isn't a nice solution, but I cannot explain this behaviour. Bye. Adam. -- Adam Duck (du...@cs...) Bockenheimer Landstr. 135 / Zi. 211 60325 Frankfurt/Main ______________________________________ You can never tell which way the train went by looking at the track. |
|
From: Adam D. <du...@in...> - 2001-11-19 15:33:07
|
>>>>> "Mario" == Mario Lang <ml...@ho...> writes:
Mario> depending on if in XEmacs, device-sound-enabled-p is a
Mario> function or a variable, I am 1quite sure you want to do
Mario> either: (boundp 'device-sound-enabled-p) or (fboundp
Mario> 'device-sound-enabled-p)
Mario> Your code returns t in all cases, because it checks if
Mario> either t or nil, the return values or
Mario> device-sound-enabled-p are bound, And t and nil are always
Mario> bound in emacs lisp.
Mario> so, question remaining is: If device-sound-enabled-p a
Mario> function, or a variable?
Yes, you're right.
---
`device-sound-enabled-p' is a built-in function
(device-sound-enabled-p &optional DEVICE)
Documentation:
Return t if DEVICE is able to play sound. Defaults to selected device.
---
So it has to be "(fboundp 'device-sound-enabled-p)". I apologise, but
I thought I had tested it with Emacs21. Strangely, it returned `nil'
when I evaluated "(boundp (device-sound-enabled-p))".
But even more strange: I can't reproduce this now ;-).
>> (device-sound-enabled-p)) (play-sound-file filepath)
Mario> another remaining question is: do you really want this? I
Mario> mean, is play-sound-file in XEmacs non-blocking, or not? If
Mario> not, be warned that emacs is unresponsive while the
Mario> soundfile is played.
Yes, I'd really like to have this feature, as this is my (nearly) only
chance to get sound in Windows. This function does not hang XEmacs -
at least not (emacs-version) "XEmacs 21.5 (beta2) \"artichoke\"
[Lucid] (i586-pc-win32) of Sun Aug 05 2001 on NEVERYON". I tried it
with "baywatch_full.wav" and I could type while it was playing.
So I suppose it doesn't do that in other versions, too (I've been using
XEmacs for 3 years now - starting with 21.1 - and I've never seen
another behaviour).
If you have other disbelieves: Let me know. But I really think it's a
big advantage for XEmacs-users (especially on MS-Windows).
Bye. Adam.
--
Adam Duck (du...@cs...)
Bockenheimer Landstr. 135 / Zi. 211
60325 Frankfurt/Main
______________________________________
I considered atheism but there weren't enough holidays.
|
|
From: Mario L. <ml...@ho...> - 2001-11-19 15:09:21
|
Adam Duck <du...@in...> writes: > Well, I don't have write-access to CVS and I also don't know what to > do with it if I had it. I've inserted some code into `erc-play-sound' > that makes my life with XEmacs on WinXP easier... > > (cond ((and filepath (file-exists-p filepath)) > (if (and (boundp (device-sound-enabled-p)) depending on if in XEmacs, device-sound-enabled-p is a function or a variable, I am 1quite sure you want to do either: (boundp 'device-sound-enabled-p) or (fboundp 'device-sound-enabled-p) Your code returns t in all cases, because it checks if either t or nil, the return values or device-sound-enabled-p are bound, And t and nil are always bound in emacs lisp. so, question remaining is: If device-sound-enabled-p a function, or a variable? > (device-sound-enabled-p)) > (play-sound-file filepath) another remaining question is: do you really want this? I mean, is play-sound-file in XEmacs non-blocking, or not? If not, be warned that emacs is unresponsive while the soundfile is played. -- CYa, Mario <ml...@de...> Homepage(s): http://delysid.org | http://piss.at/ "I don't think so," said Ren'e Descartes. Just then, he vanished. |
|
From: Adam D. <du...@in...> - 2001-11-19 14:32:15
|
Why does `erc-keywords' search for keywords in the whole line (including nick-names)? I don't want to highlight all messages a ".*duke.*" says but only that ones in which he states "duke". Can someone enlighten me why including nicknames is the better solution? Or even better: can someone fix this? ;-). Bye. Adam. -- Adam Duck (du...@cs...) Bockenheimer Landstr. 135 / Zi. 211 60325 Frankfurt/Main ______________________________________ I considered atheism but there weren't enough holidays. |
|
From: Adam D. <du...@in...> - 2001-11-19 14:15:12
|
Well, I don't have write-access to CVS and I also don't know what to
do with it if I had it. I've inserted some code into `erc-play-sound'
that makes my life with XEmacs on WinXP easier...
(cond ((and filepath (file-exists-p filepath))
(if (and (boundp (device-sound-enabled-p))
(device-sound-enabled-p))
(play-sound-file filepath)
(start-process "erc-sound" nil "/bin/tcsh" "-c"
(concat erc-play-command " " filepath))))
(t (beep)))
All I'm trying to say is: _please_ include this into erc.el :-).
Bye. Adam.
--
Adam Duck (du...@cs...)
Bockenheimer Landstr. 135 / Zi. 211
60325 Frankfurt/Main
______________________________________
If it's zero degrees outside today and it's supposed to be twice
as cold tomorrow, how cold is it going to be?
|
|
From: Adam D. <du...@in...> - 2001-11-19 13:52:18
|
>>>>> "Mario" == Mario Lang <ml...@ho...> writes:
Mario> 2. question is, do we want to set
Mario> erc-session-server to the name the server announces, or do
Mario> we want to add an additional variable,
Mario> erc-announced-server-name and use that in the modeline?
Oops... C-c C-c was wrong in that case :)
Well, the problem is that a user does not have the slightest
possibility to see the announced server-name. So, they cannot set up
the variable correctly and will think my code is a bunch of crap.
I personally would like to see the server _I am_ connected to and not
the server I tried to connect to. I also found out that
irc.de.quakenet.org connects me to irc3.de.quakenet.org. It was only
my regexp that saved me in that case.
AFAIC displaying `erc-announced-server-name' in the modeline would be
a good solution. You can always look to which server you wanted to
connect by C-h v `erc-session-server'. No Information gets lost.
Bye. Adam.
--
Adam Duck (du...@cs...)
Bockenheimer Landstr. 135 / Zi. 211
60325 Frankfurt/Main
______________________________________
Join hands across the seas. Stop continental drift.
|
|
From: Mario L. <ml...@ho...> - 2001-11-19 11:12:12
|
Adam Duck <du...@in...> writes: > Well, I was implementing an auto-join feature like this: > > (defun erc-channels-autojoin (proc parsed) > (let ((host (elt parsed 1)) > (nick (elt parsed 2)) > (ajl erc-autojoin:channels)) > ... > My list looks like this: > erc-autojoin:channels (list (list ".*\\.quakenet\\.org" "#spontanctf" "#toefte") > (list "tu-ilmenau\\.de" "#fastmp3.info")) > > Btw, I hooked this function to `erc-server-376-hook' "End of /MOTD". very well done. > > It works fine with "irc.de.quakenet.org" - where the mode-line states > the same as `erc-session-server' and the variable `host' I get by parsing > "parsed". > But there seems to be a problem with irc.tu-ilmenau.de: > here the mode-line says "irc.tu-ilmenau.de" and so does > erc-session-server; but I get "Tu-Ilmenau.DE" into the function > `erc-channels-autojoin'. It wouldn't be that bad if a user - i.e. me > - could see the server they are in to set `erc-autojoin:channels' but > all they can see is the mode-line string which does not match the > string I am given in the function. 1. the modeline contains always the same as erc-session-server, because it actually displays the value of that variable :) 2. question is, do we want to set erc-session-server to the name the server announces, or do we want to add an additional variable, erc-announced-server-name and use that in the modeline? (defun erc-server-004 (proc parsed) (let ((server-name (aref parsed 4))) ;; or 1, not sure whats better (setq erc-session-server server-name))) play with that if you like. > Can anyone help? > > Bye. Adam. > -- > Adam Duck (du...@cs...) > Bockenheimer Landstr. 135 / Zi. 211 > 60325 Frankfurt/Main > ______________________________________ > > Take my advice, I don't use it anyway. > > > _______________________________________________ > Erc-help mailing list > Erc...@li... > https://lists.sourceforge.net/lists/listinfo/erc-help > -- CYa, Mario <ml...@de...> Homepage(s): http://delysid.org | http://piss.at/ "When the going gets tough, the tough get empirical." -- Jon Carroll CYa, Mario <ml...@de...> Homepage(s): http://delysid.org | http://piss.at/ ma...@bd... (no longer valid - where are you now, Martin?) -- from /usr/src/linux/drivers/cdrom/mcd.c |
|
From: Adam D. <du...@in...> - 2001-11-17 13:40:29
|
Well, I was implementing an auto-join feature like this: (defun erc-channels-autojoin (proc parsed) (let ((host (elt parsed 1)) (nick (elt parsed 2)) (ajl erc-autojoin:channels)) ... My list looks like this: erc-autojoin:channels (list (list ".*\\.quakenet\\.org" "#spontanctf" "#toefte") (list "tu-ilmenau\\.de" "#fastmp3.info")) Btw, I hooked this function to `erc-server-376-hook' "End of /MOTD". It works fine with "irc.de.quakenet.org" - where the mode-line states the same as `erc-session-server' and the variable `host' I get by parsing "parsed". But there seems to be a problem with irc.tu-ilmenau.de: here the mode-line says "irc.tu-ilmenau.de" and so does erc-session-server; but I get "Tu-Ilmenau.DE" into the function `erc-channels-autojoin'. It wouldn't be that bad if a user - i.e. me - could see the server they are in to set `erc-autojoin:channels' but all they can see is the mode-line string which does not match the string I am given in the function. Can anyone help? Bye. Adam. -- Adam Duck (du...@cs...) Bockenheimer Landstr. 135 / Zi. 211 60325 Frankfurt/Main ______________________________________ Take my advice, I don't use it anyway. |
|
From: Adam D. <du...@in...> - 2001-11-17 13:23:25
|
Well, the function I mean is: I recently - yesterday - down-loaded erc from cvs and found that (defun erc-server-MODE (proc parsed) (let* ((sspec (aref parsed 1)) (tgt (aref parsed 2)) (mode (mapconcat 'identity (delete nil (subseq parsed 3)) " ")) ... is causing errors. The problem is, that `delete' only works on lists - at least in XEmacs - so I get an error every time someone changes the mode. Has anyone seen this? Anyway, `delete*' does the job right. Bye. Adam. -- Adam Duck (du...@cs...) Bockenheimer Landstr. 135 / Zi. 211 60325 Frankfurt/Main ______________________________________ Whoa...I did a 'zcat /vmlinuz > /dev/audio' and I think I heard God... |
|
From: Mario L. <ml...@ho...> - 2001-11-12 10:00:15
|
Benjamin Drieu <bd...@ap...> writes: > Not really a patch but a code snippet that you might consider useful : > it allows to switch to the first active channel, according to > erc-modified-channel-string. Ahh, nice. I have an idea here, so I will just go and comment: Do you use Emacs in X/with mouse? Then the following would be even more useful: 1. Have a function like yours which takes an argument (the number of the entry). 2. Have a local-map in the global-mode-string where the ctracked channels are shown which makes every channel clickable. On click, the window where the mode-line belongs to switches buffer to that buffer. What do you think? > It works fine but it is not very satisfying since it only parse this > string instead of using a real channel list paradigm. Humpf, another > item in my todo list ;-) Someone should really rewrite channel tracking so that a list is used instead of string. Would make that function simpler, and most probably remove the leading-comma bug. > (defun erc-jump-to-first-active-channel () > "Jump to the first active channel." > (interactive) > (let* ((regexp "^ \\[,?\\(#?[^,]*\\)\\(,\\|]$\\)") > (result (string-match regexp erc-modified-channel-string)) > (begining (and result (match-beginning 1))) > (end (and result (match-end 1))) > (channel (and result (substring erc-modified-channel-string begining end)))) > (if channel > (switch-to-buffer (erc-get-buffer erc-session-server (downcase channel)))))) hmm, the last line needs to look like this for erc >1.107: (switch-to-buffer (erc-get-buffer channel))))) ahh, and it is flawed anyway, because channel tracking doesn't handle multiple servers yet. -- CYa, Mario <ml...@de...> Homepage(s): http://delysid.org | http://piss.at/ |
|
From: Mario L. <ml...@ho...> - 2001-11-12 01:22:09
|
OK! Issues closed in the last few days: 1. erc-mode|header-line-format: You can now customize erc-mode-line-format as well as erc-header-line-format 2. Short buffer names: erc uses now short buffer names, thanks to smoke! M-x iswitchb-default-bindings RET C-x b and switching between many channels should be very convinient again. 3. erc-buffer-list handling is now channeled through the function erc-bufferlist. For those who wants to write erc-switch function, thats the thing you wanted! -- CYa, Mario <ml...@de...> Homepage(s): http://delysid.org | http://piss.at/ Hey, if pi == 3, and three == 0, does that make pi == 0? :-) -- Larry Wall in <199...@wa...> |
|
From: Benjamin D. <bd...@ap...> - 2001-11-11 00:54:18
|
Not really a patch but a code snippet that you might consider useful :
it allows to switch to the first active channel, according to
erc-modified-channel-string.
It works fine but it is not very satisfying since it only parse this
string instead of using a real channel list paradigm. Humpf, another
item in my todo list ;-)
(defun erc-jump-to-first-active-channel ()
"Jump to the first active channel."
(interactive)
(let* ((regexp "^ \\[,?\\(#?[^,]*\\)\\(,\\|]$\\)")
(result (string-match regexp erc-modified-channel-string))
(begining (and result (match-beginning 1)))
(end (and result (match-end 1)))
(channel (and result (substring erc-modified-channel-string begining end))))
(if channel
(switch-to-buffer (erc-get-buffer erc-session-server (downcase channel))))))
Hope this helps.
--
o Benjamin Drieu: bd...@ap...
dr...@bo...
o APRIL: http://www.april.org/
|
|
From: Gergely N. <alg...@de...> - 2001-11-07 21:05:16
|
Hello there! I got this wishlist bug against the Debian erc package. Since I cannot do anything about it myself, I'm sending it to the list and Mario, in hope that someone will comment on it. Subject: Bug#118645: All text in a single buffer To: su...@bu... From: Neal H Walfield <ne...@cs...> Date: Wed, 07 Nov 2001 21:14:21 +0100 Package: erc Severity: Wishlist It would be nice if erc could use a single buffer for all channels (or select channels) similar to zenirc. This is useful when idling on several channels. Thanks. |
|
From: Fabien P. <pe...@li...> - 2001-11-07 18:24:17
|
Wed, 07 Nov 2001 19:15:41 +0100, tu as dit : > Here is a small patch to help nick completion. Oups :) change (case-fold-search nil) to (completion-ignore-case t) and you're ok. Sorry :) -- Fabien Penso <pe...@li...> | LinuxFr a toujours besoin de : http://perso.LinuxFR.org/penso/ | http://linuxfr.org/dons/ |
|
From: Mario L. <ml...@ho...> - 2001-11-07 16:48:26
|
Benjamin Drieu <bd...@ap...> writes: > Mario Lang <ml...@ho...> writes: > > > What about erc-action-format? > > (setq erc-action-format "* %n %a") > > with %n ) nick, %l = login, %h = host, %a = action.. did I forget something? > > It sounds great ! > > Here is another suggestion that you may have thought about before : > what about using ``erc-interpret-controls'' to define colors in that > kind of format strings (i.e. I am really in love with that cuuuuuute > pink star that I used to see in my former IRC client ;-)). wooaaa, cool. Thanks. I didnt think about that. and smoke was already searching for a way to get colours in there. Thats a really good solution! -- CYa, Mario <ml...@de...> Homepage(s): http://delysid.org | http://piss.at/ Remember though that THERE IS NO GENERAL RULE FOR CONVERTING A LIST INTO A SCALAR. -- Larry Wall in the perl man page |
|
From: Mario L. <ml...@ho...> - 2001-11-07 16:30:44
|
Hello. In erc cvs rev. 1.91, I tried to modularize erc a bit. I removed erc-fill-region calls from the core functions and added erc-fill-region to erc-insert-hook instead. Unfortunately, this results in a serious problem. erc-fill-region changes the region, e.g. the end position is probably something else after erc-fill-region was executed. So, we would need to have erc-fill-region at the end of erc-insert-hook. This already conflicts with one function smoke is writing, which also required to be the last one on this hook. So we need to find a nice solution for this. Here is what I can propose: 1. Write erc-run-hook-with-args in a way that if one of those hook functions returns a list which length is same as number of args, call the next hook with those values. 2. make new buffer-local variables named erc-insert-start and erc-insert-end. Those should be set wherever erc-insert-hook gets called, and those should be markers. erc-insert-hook would get called without any args then, and use the erc-insert-start and erc-insert-end markers to determine the region. If erc-fill-region inserted some text then, the markers would reflect that for the next function. 3. Wherever erc-insert-hook gets called, first, create a temp-buffer, insert text there, call the insert-hook, extract the resulting text, and insert it where appropriate. 4. Move erc-fill-region back into the core functions. I would like to avoid this solution obviously. So, all you lisp hackers, please please comment on this, and perhaps more solutions. I am somehow a lisp newbie and I am not sure which solution is most flexible for future use. Comments? -- CYa, Mario <ml...@de...> Homepage(s): http://delysid.org | http://piss.at/ |
|
From: Benjamin D. <Ben...@al...> - 2001-11-07 14:01:01
|
Here is a small patch against erc.el 1.92 that corrects a minor problem. There *does* exist women on IRC, but ERC doesn't seem to know that. Seeing a friend ``renaming himself'' every time she changes her nick really shocked me ;-). So I changed the NICK notice to ``is now known as'' (à la xchat) and added a timestamp that was missing. Hope this helps. -- o Benjamin Drieu: Ben...@fr... bd...@ap... o Alcove: http://www.alcove.com/ o APRIL: http://www.april.org/ ===File ~/erc-1.92.diff===================================== --- erc-1.92.el Wed Nov 7 14:48:44 2001 +++ erc.el Wed Nov 7 14:46:50 2001 @@ -3286,7 +3286,10 @@ (list nick login host) (list nn)) (cons (erc-make-notice - (format "%s (%s@%s) has renamed himself to %s" + (format "%s%s (%s@%s) is now known as %s" + (if erc-timestamp-format + (format-time-string erc-timestamp-format + (current-time)) "") nick login host nn)) bufs))))) @@ -3312,7 +3315,10 @@ (list nick nil nil) (list nn)) (cons (erc-make-notice - (format "%s (%s@%s) has renamed himself to %s" + (format "%s%s (%s@%s) is now known as %s" + (if erc-timestamp-format + (format-time-string erc-timestamp-format + (current-time)) "") nick "???" "???" nn)) bufs))))) ============================================================ |
|
From: Benjamin D. <bd...@ap...> - 2001-11-06 11:24:13
|
Mario Lang <ml...@ho...> writes: > What about erc-action-format? > (setq erc-action-format "* %n %a") > with %n ) nick, %l = login, %h = host, %a = action.. did I forget something? It sounds great ! Here is another suggestion that you may have thought about before : what about using ``erc-interpret-controls'' to define colors in that kind of format strings (i.e. I am really in love with that cuuuuuute pink star that I used to see in my former IRC client ;-)). Cheers, Benjamin -- o Benjamin Drieu: bd...@ap... dr...@bo... o APRIL: http://www.april.org/ |
|
From: Mario L. <ml...@ho...> - 2001-11-05 18:34:49
|
sm...@wa... (Tijs van Bakel) writes: > Mario Lang <ml...@ho...> writes: > > > So, We'd need to handle updating erc-buffers (the global variable) > > in all places where a buffer is created or removed. Even > > a kill-buffer should remove it from erc-buffers (thats currently not > > implemented with erc-buffer-list either, and generates weird behaviour > > if the User kills a erc buffer and uses erc afterwards for something else. > > Currently, killing a erc buffer leaves a buffer in erc-buffer-list > > which is <Deleted buffer> and that confuses erc quite a bit.) > > I hope DEFADVICE'ing kill-buffer is sufficient? What about kill-buffer-hook? > > My main concern is the data structure we should use for a global > > erc-buffers variable. It should be possible to > > determine/destinguish process buffers from channel > > buffer. destinguishing channel buffers from info-buffers or query > > buffers would also be useful later on. Any ideas? > > I think a plain list of Emacs buffers would be best. One can then > always use local buffer variables to store extra information. Right, Terrible loop, but maybe most efficient code You will need some helper functions then, like erc-buffer-list-same-session-server and the like. -- CYa, Mario <ml...@de...> Homepage(s): http://delysid.org | http://piss.at/ I'm still waiting for the advent of the computer science groupie. |
|
From: Mario L. <ml...@ho...> - 2001-11-05 15:11:38
|
Fabien Penso <pe...@li...> writes: > 30 Oct 2001 21:52:36 +0100, tu as dit : > > > I'll send my big TODO list later, this is a short mail to see who's > > already subscribed :-) > > Hi. > > Something I dislike with erc, it gets really difficult to make a long > discussion with someone, as your private messages appears in the status > bar for a short time. I had to C-x 4 b *Messages* to not miss any of > them. Have a look at erc-auto-query, try playing with /query nickname. Third suggestion is, have a look at the erc-process buffer (thats the buffer where the server messages arive). If you dont have a open query buffer, private messages arrive there, so you shouldnt need to use *Messages* > Also it would be nice to be cable to use TAB to have erc put /m > <nickname of last query> so you don't have to manually type it. Again, have a look at /query nickname. > Also I would prefer to see the lines I type in the chan, appear in the > buffer as it was from someone else, to be sure it has been sent. Hmm, CVS version does this for most things (except /me). Could you give a bit more detail here. What exactly is happening... -- CYa, Mario <ml...@de...> Homepage(s): http://delysid.org | http://piss.at/ Linux: The OS people choose without $200,000,000 of persuasion. -- Mike Coleman |
197 messages has been excluded from this view by a project administrator.
