| |
| |
| |
|
Page: 1 2 3 4 5 6 7
<cliini> I'm thinking about many CL things! :)
<beach> care to share?
<cliini> I can open a buffer with occlusion code in it if you want.
<beach> I am not sure what to say to that
<cliini> It's my html generator.
<beach> ah, what does it generate the html from?
<cliini> From rather freeform text, pretty much looking like email.
<dan_b> is that the successor to smartupdown or whatever it was called?
<cliini> Yeah, it's a successor to that.
<dan_b> I can't decide whether writing cl weblog software is in any way a sensible use of my time
<cliini> I'm currently thinking about user extensibility in the source language.
<cliini> Maybe something like {foobar 1 2 "quux"} which would end up calling a function foobar, or perhaps something more complex, with a read phase and execute phase separated.
<pvaneynd> hagabaka: did you rm -rf ~/.slime ?
<cliini> Actually I think the separation is necessary. This could be used for stuff like implementing smart tables.
<cliini> (Multiple possible backing stores for table data, possibly generate multiple representations of the table at one go.)
<cliini> I've been thinking about this, and I think the primary purpose of free-time hacking is to prevent me going insane from the idiotic stuff at work.
<hagabaka> cool, that made it work
<hagabaka> thanks pvaneynd
<cliini> So whether some project X is a good or a bad use of my time is only a secondary concern.
<cliini> If I didn't have a family and a mortgage I'd probably make aggressive career moves. :)
<baetis-fly> mortgages ****...
<dan_b> well, if it stops you going insane it's a good use of your time by definition
<cliini> Without mortgages we'd all rent from evil capitalists. ;)
<baetis-fly> true true :)
<Xach> esr's blog software is written in cml2
<dan_b> hopefully my evil capitalist has fixed the hole in my^Whis ceiling today
<cliini> Did it rain in?
<dan_b> yes
<cliini> Ouch.
<dan_b> flat roof, and I rather suspect blocked gutter
<cliini> beach, I have some neat things in the design to make it more interesting. :-)
<chibikeandy> ****s
<chibikeandy> my landlord let a hole in our bathroom ceiling go for months
<cliini> Like I have a real intermediate representation, as a tree looking a bit like (html () (head () (title () "foobar" ...
<cliini> Then the user can arrange for his code to munge that tree before it's written out.
<cliini> I'm not very good at user interface, so the user interface is a bunch of stuff you can call. Actually it works pretty nicely in CL.
<cliini> No, that's wrong: actually it works pretty nicely because of SLIME.
<alpheus> I've written an araneida handle-request-response method, specialized only on the handler argument. It handles POST fine, but I get a 404 on GET. Is there something obviously wrong with what I've done?
<dan_b> is it returning non-NIL ?
<alpheus> afaict it never gets called ... I've inserted a (break) as the first form.
<cliini> Oh, and the unstated ***umption is that this will be used to generate the html to files, from which any web server can serve it. This is probably not hard and fixed, but that's how I've been using it.
<dan_b> nothing wrong with files
<dan_b> well, ok, there is, but they're pretty serviceable beasts even so
<cliini> Apache serves a file with about 1/10000th the resources that my code has written the file.
<cliini> So using a file system as a cache happens to work reasonably well, in this case.
<cliini> In a former life I implemented a purely database backed web system, and it made me allergic to such things.
<beach> cliini: so the user interface you are talking about is used to manipulate the "internal" html before it is written out?
<cliini> Yeah.
<beach> sounds like it ought to be an interactive CLIM app :)
<cliini> Well, after you've made your code work properly, I imagine you will only touch the source files and press a button which will run your entry code and everything happens nicely .:)
<hagabaka> emacs/slime doesn't automatically indent when I edit lisp files...how do i turn it on?
<cliini> But perhaps the initial development would be nice to do in a CLIM system.
<nyef> hagabaka: At a guess, apropos for 'electric'?
<hagabaka> ...what does apropos mean?
<mgr> good evening
<levi> hagabaka: You need to remap the enter key to to 'newline-and-indent in your lisp-mode-map
<hagabaka> oh...
<baetis-fly> hagabaka: or just get used to doing ctrl-j instead of enter.
<beach> mgr: good evening
<hagabaka> oh
<hagabaka> that looks like an easier temporary solution
<beach> hagabaka: I agree with baetis-fly. C-j is even easier to find than CR
<hagabaka> heh
<baetis-fly> anything to give my pinkies a break is good.
<hagabaka> I use a pinky for Ctrl though
<hagabaka> But my pinkies are strong!!!1
<levi> It's not very hard to remap the key. Just look at someone's .emacs and copy the relevant part out. ;)
<baetis-fly> my left pinky is the strongest muscle in my body.
<hagabaka> ok, i'll do that sometime
<hagabaka> does anyone have time to help me debug or spot the error in a homework?
<nyef> Nope, so sorry, it's against policy to help with homework. ^_-
<hagabaka> :(
<levi> hagabaka: Notice nyef's wink. You might try posting your code on lisppaste and asking for help with it again.
<hagabaka> ok
<baetis-fly> minion: lisppaste
<lisppaste> To use the lisppaste bot, visit http://paste.lisp.org/new/lisp and enter your paste.
<Azure_Ag> Is there any resource around that might give some guidance as to which backend to use with UnCommon Web?
<nyef> Azure_Ag: Yes. Use whichever works.
<piso> #ucw
<mgr> beach: hello! Sorry, telephone..
<Azure_Ag> Thank you.
<nyef> Reading the trivial-template code, I see what has to be a waste of a perfectly good tagbody.
<nyef> "butter to catch problems soon"?
<mgr> beach: actually I could commit my little info-pane patch. But I'm not sure where the right place for the presentation translator is.. (see my recent email to climacs-devel)
<rtoy_> Hmm. I can't seem to find the clhs section that says a declaration name can't be the name of a deftype, defstruct, or condition.
<rtoy_> But I found it just yesterday.
<nyef> Okay, that really is trivial templating. I'm not sure I like the implementation, but I certainly like the idea.
<kpreid> where is this implementation?
<kpreid> oops. never mind.
<lisppaste> hagabaka pasted "a calculator with LET and LET*" at http://paste.lisp.org/display/12729
<hagabaka> that's the homework i'm having problem with...
<nyef> I'm impressed. You managed to completely mess up the syntax highligher with that. ^_^
<hagabaka> :(
<hagabaka> i could paste it again with ;; for the comments at the beginning
<nyef> No, no... Done is done.
<nyef> Don't want to waste paste numbers, they're a non-renewable resource. ^_-
<levi> Heh, it thinks all the code is inside a string.
<nyef> Except for the strings, yes.
<rudi> mgr: I see what you mean wrt my change to the mcclim cursors - my test case involved text fields only :(
<nyef> As I said, I'm quite well impressed.
<lisppaste> hagabaka annotated #12729 with "sorry about bad comments" at http://paste.lisp.org/display/12729#1
<kpreid> hagabaka: how does it fail?
<z0d> nyef: you can annotate, so paste numbers are quasi-reusable
<hagabaka> oh yeah
<hagabaka> when I use multiple variable bindings with let* in Tiny, which uses extend-env-sequential, it says the first variable is not defined when evaluating the second one...
<mgr> rudi: Will you have another look on the problem? :) That would be nice.
<nyef> Nothing immediately springs to mind unless you're building a bogus environment.
<rudi> mgr: I'm looking at changing the cursor appearance inside (setf port-keyboard-input-focus)
<rudi> that should work for all cases.
<nyef> Try running extend-env-sequential with only one binding and see if the resultant environment looks sane.
<kpreid> hagabaka: as a matter of readability, I suggest wrapping (destructuring-bind (name expr) (first bindings) ...) around the call to extend-environment
<hagabaka> it works when I use only one binding
<hagabaka> destructuring-bind?
<nyef> clhs destructuring-bind
<specbot> http://www.lispworks.com/reference/HyperSpec/Body/m_destru.htm
<nyef> kpreid: I'd consider just using a let for a destructuring that simple.
<hagabaka> hmm, how would i use it for this program?
<mgr> rudi: You are responding to my message on 15.10.2005, 20:17: "The way it is done is wrong. flip-cursor-output [called from (setf cursor-state) :around; that seems to be the reason for these (letf (((cursor-state cursor) nil))] always sees only :HOLLOW.", aren't you?
<hagabaka> i mean, i only need to call the extend-environment with a list of (var val) forms...
<rudi> mgr: I noticed that interactor panes always have a hollow cursor.
<mgr> rudi: yes
<rudi> text-field and text-area switch to hollow/filled via armed and disarmed events
<rudi> I think stream panes, not being gadgets, don't have these
<nyef> (destructuring-bind (name expr) (first bindings) (extend-environment env (list name) (list (teval expr))))
<mgr> rudi: hm, but I've had the same problem with the text-fields of the ACCEPT calls inside the ACCEPTING-VALUES dialogue of my file selector.
<mgr> (and I still have)
<rudi> ah. well, switching cursor appearance during keyboard input focus changes should work in all cases.
<hagabaka> hmm
<rudi> mgr: once I have something, can I send you a patch so you can try it on your code?
<mgr> rudi: That's nice. I didn't really like the fact that every text-field always had a black, filled cursor.
<hagabaka> nyef, does that replace the recursion of extend-env-sequential?
<mgr> rudi: Yes, of course.
<chandler`> beach: for what it's worth, I hit enter with my right hand index finger. Then again, I'm not typing on a standard keyboard here at work.
<kpreid> hagabaka: no, it's a better way of writing the call to extend-environment within it
<nyef> hagabaka: No, just the call to extend-environment.
<hagabaka> oh
<rudi> cool, bbl
<kpreid> hagabaka: you might try loading your code, doing (trace extend-env-sequential), then trying what breaks
<hagabaka> ok
<nyef> The comment about eval-binding is wrong.
<hagabaka> about the form of expr?
<nyef> Yah..
<nyef> And the comment above extend-env-sequential is wrong.
<hagabaka> oh, didn't change it
<hagabaka> i'm confused what the format should be now...
<hagabaka> it should be something like (let ((a 1) (b 2)) (- b a))
<nyef> I don't see anything else obviously wrong here.
<rtoy_> Yay! NFS servers and clearcase down. Time to go home!
Return to lisp
or
Go to some related
logs:
checking for X... configure: error: Can't find X i
alcohol error 1402
politics
roma
|
|