logs archiveIRC Archive / Freenode / #emacs / 2010 / June / 1 / 1
JordiGH
build system wants to install .so in /usr/bin
qDot_
Sometimes binaries and shared libraries like to be close to each other
And sometimes people don't understand rpaths.
JordiGH
What folly be this?
And they're not called libfoo.so
(Action) wonders if it's worth fixing this.
sigjuice
I am trying to pick a different TAGS file based on what file is currently visited.
is there a hook that fires upon visiting file where I might be able to change tags-table-file?
faceface
when I go to save this file, emacs asks me to choose an encoding
I'd rather just save 'normal'
Select coding system (default iso-2022-jp-2): ?
Yes, I am a moron.
ð#, Taiwan ?
technomancy
sigjuice: find-file-hook, probably
faceface
These default coding systems were tried: utf-8 However, none of them safely encodes the target text.
hekevintran
i can use C-j to indent to the appropriate depth
how do you un-indent?
jabot
hekevintran: kill-rectangle ?
         

sigjuice
technomancy: I think I misunderstood what "visit" means. I guess I am looking for a hook that fires when a new buffer is selected.
technomancy
sigjuice: maybe make that var buffer-local?
faceface
Downloaded: 562 files, 2.1M in 2.6s (839 KB/s)
sorry
wrong chan
leo2007
is setcar a special form? the doc-string says no.
bpalmer
why do you think it would be?
damd
special form?
,special form
fsbot
CL_note--special_forms_compared -- at http://www.gnu.org/software/emacs/elisp-manual/html_node/elisp_120.html#IDX370
damd
what is a "special form"?
leo2007
bpalmer: I'm not sure if the first arg is evaled.
bpalmer
damd: to explicate by example: if , or, and
damd
"lazy" forms?
sohail
I thought special forms were builtins
bpalmer
damd: that's the most common way they're special, yep.
sohail
i guess I was wrong
bpalmer
any case, it's a form that isn't a macro but is still special.
damd
heh, got it
bpalmer
sohail: not all subrps count as special forms.
sohail: like, forward-char is done in C but is basically just a normal function call.
sohail
I see
bpalmer
leo2007: and yes.
leo2007
bpalmer: yes, it is evaled ?
For example, (setcar kill-ring "xyz"), if kill-ring is evaled to a list how can kill-ring get changed?
         

bpalmer
it doesn't mess with the variable
it messes with the cons cell
xristos
does anyone see anything wrong with this: http://paste.lisp.org/display/111017
leo2007
bpalmer: I always want to write (setcar 'kill-ring ....) just like (set '....)
bpalmer
set works on a symbol; setcar works on a cons cell
leo2007
bpalmer: does this imply when lisp sees a list, it knows the addresses of a cons cell i.e. addresses of its car and cdr.
ggole
leo2007: a cons cell is essentially two mutable locations.
bpalmer
leo2007: well, when C code does, certainly. It's arguable whether lisp knows it.
ggole
Lisp can't observe addresses exactly, because they may have to be relocated.
leo2007
I want to understand when setcar changes the car, how is that change picked up by the symbol kill-ring
ggole
It isn't
kill-ring continues to have the same value (in C-land, it continues to point to the same address)
leo2007
ggole: (setcar kill-ring "xyz") modifies kill-ring
ggole
It does not
leo2007
I have tested it a few times.
ggole
It modifies the list structure that kill-ring refers to.
leo2007
what info is used by kill-ring to refer to the list structure other than the head of the list?
ggole
At a low level, kill-ring is a symbol that contains in it's symbol-value slot the address of a cons cell.
leo2007
yes, I understand this bit.
ggole
When you eval (setcar kill-ring ...), the lisp evaluator will evaluate kill-ring to get at that address. This is the value which setcar observes.
Then it uses that address to mutate the first half of the cell.
sepult
why does starting an inferior lisp make erc reconnect too ?
ggole
kill-ring itself, the symbol, is entirely unaffected.
leo2007
ggole: ok, thanks. I think I have the picture now.
ggole
leo2007: the key point is that setcar never sees any part of the symbol.
This is why, for example, you need to quote the argument for add-to-list.
leo2007
ggole: yeah, that's why I thought it might be a special form.
ggole
Special forms can get at symbols directly because they see the list structure returned by the reader, which contains symbol addresses.
But not functions like setcar.
leo2007
ggole: is this somewhere in the info?
ggole
Hmm, there's a page about the evaluator somewhere. One moment.
leo2007
ggole: thanks in advance.
Anyone running Emacs under gnu/linux?
Can you paste image from other apps into a emacs buffer?
ggole
M-: (info "(elisp)Evaluation") is what I'm thinking of, but it doesn't describe things at a low level.
leo2007: by the way, I had some trouble figuring out what causes the edebug problem.
leo2007
ggole: never mind. I think I more or less understand.
ggole: do you mind sending a bug report along with your findings?
ggole
Yeah, I'll do that.
leo2007
ggole: i think labels also has problems but I don't have a reproducable case.
e1f
"Calibre, for ebooks, is one of those amazing open source projects that is tremendously useful, very well done, and has the user interface of a lawn mower designed by a blind cake decorator."
leo2007
hehe
indio
Anyone using twit.el?
leo2007
ggole: can your emacs paste image?
indio
How can I see @ tweets ?
e1f
,twit
fsbot
[0] I was just trying to help! Won't be bothering *you* again... Pfft!
[1] at http://www.emacswiki.org/cgi-bin/wiki.pl?TwIt
indio
I want to see @msgs of the guys I'm following
ggole
leo2007: dunno, I never tried
leo2007
ggole: I observe emacs's memory usage jumps a lot, when pasting images.
ggole
I don't know anything about the image handling in emacs. If you see a worryingly large amount of allocation, bring it up on emacs-devel.
By the way, labels are handled specially in edebug. If you use them in any non-canonical way, edebug probably isn't going to like it.
leo2007
ggole: so sometimes I just use lambda instead.
ggole
leo2007: let me know if you can reproduce a problem.
leo2007
ggole: OK.
sword
ARGH
I hate statistics
shabble
along with 84% of the rest of the world!
madsy
"I belong to the other 13%", etc
rbarraud
90% of everything is garbage
madsy
http://bash.org/?77904
rbarraud
(yayy for GC ;-) )
technomancy
pretty sure it's closer to 95%
sword
Y'all shut up and stop mocking me
offby1
mock mock
mock lisp
ggole
It's like lisp, but with the parens backwards.
Man, that looks funny
Still quite readable.
sword
That would be a truly evil trick to play on a Lisp user
Mess with their syntax table and watch heads asplode
quotemstr
Okay.
Magit question: I'm in "Changes:" and I hit tab to read a diff. It's a long one, so I have to scroll down to read it.
xristos
how do i bind symbols to values at runtime ?
do i have to use progv from cl ?
quotemstr
I forgot to set the mark before scrolling. How can I get back up to the header line so I can turn off display of the diff?
xristos: It's explained in the elisp manual.
elisp is already dynamically-bound so progv is mostly useless
qDot_
I've been wondering what diff block movement is myself, heh
I run into that problem to, and usually do it some stupid way.
Anyways, less git, more bbq, bbiab.
technomancy
quotemstr: I wish I knew. if you figure it out let me know
magit has a handful of minor peculiarities like that
quotemstr
I'm sure if I hack it up, I can come up with a go-to-previous-header function.
technomancy
quotemstr: you can use 1, 2, 3, etc to jump to sections
quotemstr
I'd really prefer to see the diff in a separate buffer, TBH
yonkeltron
hello
does emacs support elisp?
i understand it's a language for writing xbox games
quotemstr
(Action) flays yonkeltron.
yonkeltron
quotemstr:
quotemstr: does it support directX15?
damd
cool, emacs 24 will use guile
finally!
quotemstr
No it won't.
damd
troll failed
technomancy
damd: sweet, I can finally use call/cc to implement my super-flexible bookmarks framework!
(Action) waits for quotemstr's head to asplode
quotemstr
You're just trolling.
Everyone knows Emacs 24 will be based on org-mode.
xristos
quotemstr: if there are no existing bindings and i use set instead of progv won't i in a sense be making global variables ?
quotemstr
xristos: Use LET.
ggole
(Action) ponders the intersection of continuations and dynamic scope
xristos
i can't create the symbol names at runtime with let
or can i?
ggole
You can create them with intern
pjb``
xristos: why would you want?
ggole
Or if you already have them, but not in a let, then just make the let form.
xristos
i don't have the symbol names i want to bind
i have to create them at runtime
quotemstr
ggole: Let's make every variable a global variable.
ggole
Like `(let ((,sym1 "foo")) (,sym2 "bar")) (concat ,sym1 ,sym2))
pjb``
xristos: there's progv to do that. (require 'cl).
xristos
e.g pattern matcher finds 4 matches, i need to create $1 $2 $3 $4 and bind them
pjb``
(progv (generate-$-variables (length matches)) matches (do-something-with-$-variables))
xristos
pjb``: i know about progv, i am just asking because quotemstr mentioned it's not needed
quotemstr
Fair enough, but in that case, I'd just use a hash table.
ggole
What kind of pattern matching is this? Aren't the match variables usually statically known?
Or is this some Prolog thing?
xristos
something like that
ggole
Right. Well, have fun.
quotemstr
Let's resolve function reference at compile time.
bpalmer
let's use pass-by-name semantics.
quotemstr
Let's not support function arguments or return values.
shabble
Let's pass-by-parcel
tommylommykins
emacs noob question: How to clear all the text in a buffer?
bpalmer
erase-buffer
quotemstr
C-x h C-w
tommylommykins
:D
Thanks
quotemstr
"`repeat' ignores commands bound to input events. Hence the term "most recently executed command" shall be read as "most recently executed command not bound to an input event".
Why?
fsbot
Why not?
bpalmer
,df repeat
fsbot
repeat is an interactive compiled Lisp function in `repeat.el'.
It is bound to C-x z.
(repeat REPEAT-ARG)
Repeat most recently executed command.
With prefix arg, apply new prefix arg to that command; otherwise,
use the prefix arg that was used before (if any). ..[Type ,more]
bpalmer
quotemstr: probably because it wants to let people repeat interesting things
not movement.
quotemstr
But it does repeat movement.
"Interesting" is what repeat-complex-command is for.
bpalmer
(I have no repeat; I do have repeat-complex-command)
quotemstr
Ah, XEmacs.
xristos
is there a minor mode that allows me to save the history of comint-based modes like ielm ?
quotemstr
xristos: Have you checked the wiki?
xristos
there is desktop-mode which seems a little too much
quotemstr
,apropos savehist
fsbot
2 matches. (savehist savehist-mode)
quotemstr
,df savehist
fsbot
Doh! Symbol's function definition is void: savehist
quotemstr
,savehist
fsbot
quotemstr: savehist is, like, [0] the feature of saving minibuffer history to an external file after exit
« prev 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 next »