logs archiveIRC Archive / Freenode / #emacs / 2010 / January / 15 / 1
junis
why... w3m doesn't seem to recognize xml-sylesheet directive.... .o/
OberWrk
previously before restarting emacs I was using gud in compile mode and "Comint" mode appeared set and I was able to interact with the inline debugger in the compilation window. now it's back to RO and I can't seem to figure out what I did to make that work previously
consolers
junis -- set-fill-column is for M-q when you are editing
with w3m, w3m is a program that does formatting
you should have been more specific
w3m doesnt do stylesheets
or javascript
thank god
leo2007
consolers: do you use w3m to read rss?
consolers
no
i dont read rss, i think the technology is an insult to the intelligence when addressing the apparent problem it purportedly solves
junis
but it doesn't even do xsl sylesheets...
quotemstr
consolers: You're starting to sound like RMS.
ryoma
w3m does rss?
rss is one way, right? and it's a bunch of messages, kind of like nntp except not?
technomancy
in the words of a friend, "Listen, I know broken protocols. At least Mail has an RFC. Blogs just have 'various objections to Dave Winer.'"
RSS is astonishingly ill-conceived.
         

ryoma
the web really sucks in a lot of ways: silly protocols, lack of accessibility standards, stupid design and formatting decisions, lack of sexp.
quotemstr
ryoma: Life sucks too.
Consider the prostate.
consolers
not anywhere close tp the web
quotemstr
That's incrementalism for you.
e1f
'the starship mechanic' on tor is pretty good
consolers
teh web
ryoma
and you breathe through the same place you eat
quotemstr
Yes, RSS has problems.
Yes, the RFC822-derived protocols are far superior in every conceivable way.
But RSS is here to stay, so we might as well find ways to make it suck less.
There are a couple services that transform RSS from a poll-based mechanism or something based on asynchronous notifications, though their names escape me.
consolers
teh plague
ryoma
too bad they didn't decide to just use rfc822
leo2007
at the moment I use google reader
quotemstr
(Action) uses rss2email.
e1f
,t8 it en rigate
fsbot
striped
quotemstr
It works very well.
e1f
,t8 it en penne
fsbot
pens
ryoma
,t8 it en pasta
fsbot
pasta
ryoma
oh that tor
         

technomancy
there are a hell of a lot of reasonable objections to Dave Winer
rfc822 is its own special brew of crazy, but at least it's consistent and unambiguous.
OberWrk
,,comint
fsbot
I heard comint is [0] general COMmand INTerpreter in a window stuff
[1] see also ComintMode
OberWrk
,comintmode
fsbot
hmm, ComintMode is at http://www.emacswiki.org/cgi-bin/wiki.pl?ComintMode
technomancy
seriously though, any time you see the words "7-bit clean" you run like hell the other direction if you have a choice.
ryoma
i find >from to be annoying
leo2007
,rss2email
fsbot
From memory, rss2email is at http://www.aaronsw.com/2002/rss2email
quotemstr
Hahaha
consolers
t
ryoma, i use babyl . no > from in babyl format
mbox is braindamaged
and rmail-mbox indicates that emacs-devel is in sh*t
quotemstr
Heh. I replaced a coworker's shell with this today: http://pastebin.ca/1751600
You should have seen his reaction. :-)
(He just got to it now.)
e1f
http://www.tor.com/index.php?option=com_content&view=story&id=58363
consolers
unlike rss, mbox had a better alternative --- babyl
but the idiot developers of emacs devel are too strong in this age of kalki
ryoma
,kalki
fsbot
ryoma: Please, continue.
consolers
decline preceeding end of days
ryoma
fsbot: that's your job
fsbot
Try: crankytabsinsteadofspaces dillo dlgs-phrases ghey mad orgstruct quickstart tnas TWTAS uemacs http://www.emacswiki.org/emacs-pt/GlauberPrado
consolers
destroying actual value when creating imagined value in the minds of billions of first time users who have no clue of what is and isnt value
ryoma
emacs developers do that?
wow
quotemstr
Huh?
consolers
and introduce inefficiencies
jlf
the horror!
k-man
how much are blue ray dvdr disks these days?
quotemstr
k-man: More than tape?
leo2007
consolers: why do you think babyl is better than mbox?
k-man
quotemstr: tape is so boring though
quotemstr
Bah!
consolers
i dont "think"
quotemstr
All that whirring and grinding! The noise! The blinkenlichten!
It's exciting!
consolers
leo2007 i've given the reasons here many times
ryoma
sounds contradictory to give reasons for not thinking
leo2007
(Action) missed that
consolers
what difference does it make
ryoma
consolers: do you use emacs 21?
consolers
no i patch emacs
leo2007
I use gnus so I don't care what they do with rmail
consolers
and i use older ngnus-0.10 which has babyl support
because i save to babyl from gnus.
ryoma
interesting
you can edit headers in babyl and it will keep the whole set
but then, maybe it keeps 2 sets all the time anyway
?
consolers
yeah you can retrieve an md5-verifiable copy of the actual message as it was received in babyl
with mbox, it'll recode and lose your data
shabble
can anyone think of a package or snippet that uses ido to select from a list of static options, that I could steal for inspiration?
I'm getting nowhere reading docstrings
ryoma
ido does that
consolers
bah my emacs-w3m isnt working when i change w3m-use-cookies to t
ryoma
read the source
technomancy
shabble: ido-completing-read
leo2007
,ido-completing-read
fsbot
[google] http://www.emacswiki.org/emacs/external-abook.el
http://www.emacswiki.org/emacs/RecentFiles
jkr_
Is there a proper way to write a hook that bails out of a function if a condition isn't met? Having (error "foo") as the result of the condition seems to work, but I'm not sure if it's kosher.
harovali1
how can I edit a file through ssh in another machine ?
leo2007
,news
fsbot
NeWS is, like, [0] the Network-extensible Widget System. Think X11, but better.
[1] Think X11, but deader, more like.
r4yner
harovali1: tramp
shabble
harovali1: ,,tramp
fsbot
TRAMP is, like, [0] an Emacs library, included in Emacs 22, that implements transparent remote file access over a number of protocols (ssh, sudo, rsync, ftp, ...) <http://emacswiki.org/wiki/TrampMode>, <http://savannah.nongnu.org/projects/tramp> ..[Type ,more]
jlf
harovali1: see also ,,sshfs
fsbot
sshfs is [0] SSH Filesystem, which some might prefer to use instead of TRAMP.
[1] http://fuse.sourceforge.net/sshfs.html,
[2] for OS X: http://code.google.com/p/macfuse/
shabble
technomancy: ooh, that looks like the one
ralph__
Does anyone know how to write the function "yesterday" in elisp?
Or to add an integer number to a date?
harovali1
thanks !
ralph__
I know about format-time-string, but this won't tell me how to add, say, 12 days to today's date.
Thanks.
jkr_
ralph__: I have it in some function somewhere. Let me hunt it down for you...
ggole
jkr_: conditionally returning from a function can be done with (unless cond (rest-of-function))
jkr_
ggole: thanks -- but there's a pre-existing function in another package (message.el), and it has hooks at various points along the process (sending mail). I want to write a function that will be run as a hook, and will stop the mailing process if a condition isn't met.
I could wrap the whole thing with defadvice, but I want to check the condition at the point of the hook, when a few routines have already been run.
ggole
In that case, error is the way to go
Although read the docstring to see if there is already a way to do that
consolers
ah type mismatch in custom-set-var w3m-c-accept-domains
jkr M-x apropos ^message.*hook
jkr_
ralph__: (decode-time (time-add (current-time) (seconds-to-time (* 60 60 24))))
ralph__: that will add one day. You can put day-seconds in a let statement, and multiply it accordingly.
consolers
jkr which hook were you using?
i guess it doesnt matter you can always call message-kill-buffer
jkr_
consolers: it was message-send-mail-hook. I just wanted a last-second confirmation (like in feedmail). But if I enter n to the confirm y-or-n-p (say because I notice a last-second misspelling), I still want to be in the buffer.
So just exiting the message-send-mail-with-sendmail function (in my case) with an error seems to be the only way to do this without it either (a) going on and sending the mail, or (b) getting rid of it.
consolers
ya prolly
the error handler should be set up in the scope of the C-c C-c call to send the mail
you could replace your send function to something that (catch 'foo (message-send-and-exit)) and in your hook you could (throw 'foo)
jkr_
Yeah... the problem is that I'm trying to avoid keeping my own hacked copy of half the functions in message.el. The more I can do with hooks, the better.
consolers
should be set up in the *enclosing scope of the C-c etc
jkr_
But that does sound more proper...
consolers
but thats a given with hacking anycode which larsi has touched
its a mess
to get any behaviour that he has not anticipated and coded
dont know if throw-catch will work just as well in an around advice on message-send-and-exit
Surgo
Is there a way to tell if the current function you are in was called interactively?
jkr_
Hmm, well I'll experiment with putting the throw-catch in the around advice. Eventually I guess I'll just have to start keeping my private version, but I'm trying to see if I can avoid it as long as possible.
consolers
C-h f interactive-p
jkr_
consolers: thanks for the suggestions
consolers
dont thank me!
Surgo
I'll think you anyway. Thanks consolers!
rgr
does emacs load windmove.el by default?
Surgo
Hmm, any readable way to pretty-print a let block with multiple variables so that the variables are clearly distinguishable from the evaluation block?
watermind
whenever I try to save it asks me if I want to add a newline at end of file
this drives me crazy :S
can I turn this off?
snogglethorpe
argh
i'm wrapping a present
but the wrapping paper isn't big enough
thus the awkward "use little bits of other wrapping paper to fill in the gaps" style
junis
but w3m now when i read my mklamadhyamakak with parallel translation w3m does not handle the devangar+... .o/
sshelagh
just ask one of your neighbours to help you wrap
snogglethorpe: japanese are the best wrappers I know
junis
bulgarians are the best wrappers...
sshelagh
but I don't know any bulgarians
junis
... but the ones that unwrap wond be much happy...
consolers
i decided to use unicode roman instead of unicode devanagari
i still need a typesetting solution to typeset translations
technomancy
"I'm super K, and I'm hear to say: I like to make rap in a different way"
consolers
tex cant do it --- when you want text in one row and translation in 2nd row
and balance those across the whole chapter
junis
what is tex... i now make overlays all over
sshelagh
I remember going into a bookstore and asking about the "the Latex companion"
consolers
(w3m does handle unicode devanagari though)
sshelagh
The young man was releaved to see it was a *computer* book
consolers
heh
tex algorithms are useless for parallel texts
snogglethorpe
sshelagh: cmon he had been hoping for a date!
consolers
the problem is there isnt an alternative to tex
junis
if it does why mu mantras are now rows of boxes...
sshelagh
snogglethorpe: no it was definitely relief
consolers
because emacs thinks it doesnt have the font
snogglethorpe
sshelagh: well, it can be stressful asking a complete stranger on a date,...
sshelagh
consolers: I'm sure I half remember a latex package that deals with translations
but maybe different encodings are too much for it to handle
consolers
no the problem isnt encoding,
tex typesets one page at a time, the problem is to balance parallel texts across multiple pages
sshelagh
So you have one language on one page and the translation on the opposing page?
snogglethorpe
i suppose machines now days could deal with laying out the entire book at once instead...
consolers
sshelagh or better in balanced rows
so one could read each continously
sshelagh
(Action) has wrestled with latex in the past
consolers
and in critical texts, you have more rows
sshelagh
not wearing it mind!
I wonder it one of the poetry critic packages might be what you are looking for
They have lines and following critiques which might mimic what you need
« prev 1 2 3 4 5 6 7 8 next »