00:01:28 <sbp> * sbp wonders where the logs go now
00:01:32 <sbp> .time
00:01:32 <datum> Wed, 11 Feb 2004 00:01:32 GMT
00:01:53 <sbp> ah, 11. good
00:03:07 <thelsdj> its funny that in moving to arizona, i'm not as much bothered by being 2 hours off from EST (where i used to live) but that i'm 7 hours of GMT rather than 5
00:04:15 <sbp> heh!
00:04:27 <sbp> what's most affecting you about being further off of GMT?
00:05:02 <thelsdj> for one thing i'm used to the essential mix being on at 9pm but now i have to remember its 7pm lol
00:05:33 <thelsdj> ie 2am in england
00:05:44 <AaronSw> great, now the locking isn't working at all
00:05:45 <sbp> that's so great
00:05:52 <sbp> I should make you an honourary European
00:05:55 <sbp> well, Brit
00:06:20 <thelsdj> i've been plotting how to get out of this damn country
00:06:40 <sbp> * sbp found that a plane ticket worked for him
00:06:43 <AaronSw> oh, never mind. ir was lockf
00:07:58 <sbp> Aaron: I'm doing the experiment about folder uniqueness that you mentioned on my own websites
00:08:37 <sbp> hmm... what's the easiest way to print out duplicate lines in a sorted file...
00:08:47 <sbp> ah, uniq -d
00:09:19 <thelsdj> whats this folder uniqueness experiment?
00:09:49 <sbp> $ wc -l web-folders.txt
00:09:49 <sbp> 613 web-folders.txt
00:09:52 <sbp> $ uniq -d web-folders.txt | wc -l
00:09:52 <sbp> 79
00:09:57 <sbp> hang on and I'll explain...
00:10:15 <AaronSw> ok, but a lot of that's stuff like index.html and style.css
00:10:24 <sbp> $ python -c "print (79.0/613)*100"
00:10:25 <sbp> 12.8874388254
00:10:31 <sbp> no. it only counts the directories
00:10:50 <sbp> I did find web -name '*' -type d
00:11:24 <AaronSw> that doesn't sound right. what's uniq -cd say?
00:11:44 <AaronSw> err uniq -cd | sort -rn
00:12:05 <sbp> mainly two
00:12:08 <sbp> the most is nine
00:12:20 <sbp> nine "test" folders, nine "notes" folders
00:12:39 <AaronSw> hm. it'd be cool if you published the output
00:12:46 <AaronSw> i'd be curious to see it
00:12:51 <sbp> sure. hang on
00:13:25 <sbp> http://infomesh.net/200X/web-dupes-count.txt
00:13:50 <AaronSw> if most of them are 2 then you can slap a 2 on the end of the name and be fine
00:13:52 <sbp> this is across 11 sites
00:13:55 <sbp> heh, heh
00:14:22 <AaronSw> oh, this is 13% duplicated, so most of the time you're fine
00:14:36 <AaronSw> across sites isn't really fair either
00:14:54 <AaronSw> but 87% success is pretty good
00:14:57 <sbp> it's pretty fair. I'm doing this because I'm thinking about just having one account
00:15:01 <sbp> spin!
00:15:06 <sbp> 87% fat free!
00:15:11 <sbp> :-)
00:15:32 <AaronSw> seriously, if you only needed to think about a better name 13% of the time, i think that's success
00:15:36 <sbp> www doesn't really count--you can strike that one
00:15:47 <sbp> yeah, it's not *bad*
00:15:49 <AaronSw> neither do all the date dirs
00:15:55 <AaronSw> right?
00:15:59 <sbp> plus I wouldn't actually be having it completely flat
00:16:02 <sbp> nope, they don't either
00:16:16 <sbp> though there's only two, oddly
00:16:28 <AaronSw> i see more than 2
00:16:34 <sbp> months don't count
00:17:14 <AaronSw> why do you have two pwyky dirs
00:17:23 <AaronSw> and a lot of these are duplications across months which would be one dir normally
00:17:27 <AaronSw> so those don't count
00:17:31 <AaronSw> nor do duplications across sites
00:17:33 <sbp> one on infomesh for the code, one on f2o for the test
00:17:46 <sbp> that's a possibility
00:17:49 <AaronSw> that doesn't count
00:17:58 <sbp> I'd guesstimate that the actual figure would be a half, maybe a third of this
00:18:05 <sbp> yeah it does
00:18:05 <AaronSw> i don't think many of these are real conflicts
00:18:11 <sbp> I *would* want two dirs for pwyky
00:18:18 <AaronSw> why?
00:18:23 <AaronSw> oh, i see
00:18:27 <sbp> compare http://infomesh.net/2... right
00:18:43 <sbp> pwyky-doc would be good, or pwyky-test
00:18:59 <sbp> my-pwyky for pwyky-test
00:19:03 <sbp> * sbp shrugs
00:19:11 <AaronSw> i declare victory
00:19:18 <sbp> the problem is, if you choose a crap name, you're stuck with it for years
00:19:21 <AaronSw> but i still hate flock, grr
00:19:30 <AaronSw> sbp, redirects
00:19:36 <sbp> I really just want to rm -rf everything and start again
00:19:37 <sbp> bleh
00:19:42 <sbp> then I have to maintain old and new
00:19:47 <sbp> through fiddly redirects
00:19:48 <AaronSw> ?
00:20:00 <sbp> I mean, delete all the websites, not the stuff
00:20:04 <sbp> and make a new structure
00:20:34 <AaronSw> i don't understand the old and new complaint. you put a couple things in msci, realize it'd be better called misc, move the things to misc and redirect msci to misc
00:20:54 <sbp> so you have to maintain the redirect file
00:20:54 <AaronSw> (if mod_speling doesn't do it for you)
00:20:57 <sbp> heh
00:21:12 <AaronSw> maintain the redirect file? you add a line each time you rename a directory
00:21:14 <AaronSw> not hard
00:21:21 <sbp> you must admit it'd be cleaner if you got it right the first time
00:21:35 <AaronSw> or, if you have an ideal web server, you do:
00:21:35 <AaronSw> mv msci misc
00:21:35 <AaronSw> ln -s misc msci
00:21:46 <sbp> then people don't have to worry about which location is definitive, or update their links if the redirect is via HTTP perm moved response
00:21:50 <AaronSw> of course it's cleaner not to leave msci around, but i don't see it as a problem
00:22:11 <sbp> might be if you want to use msci in the future
00:22:23 <sbp> it's not so much a problem for me, but look at a site the size of W3C
00:22:27 <AaronSw> so then you just redirect those files
00:22:28 <sbp> and it could be a problem for me
00:22:34 <sbp> I need to go through this list more carefully
00:22:42 <sbp> hmm
00:22:56 <sbp> then you have two projects sharing one directory
00:22:58 <thelsdj> i have my rss file scp'd to my hosting provider hourly from my home machine and then ln -s'd like 3 times because i changed my weblog a few times before and changed the directory hah
00:23:12 <AaronSw> you don't have two projects, you have a project and a ghost
00:23:17 <AaronSw> it's a haunted directory
00:23:21 <sbp> hehheh
00:23:33 <sbp> * sbp likes the concept, and likes thelsdj's illustration
00:23:52 <sbp> I don't know what it is about haunted directories...
00:23:55 <sbp> they just spook me
00:24:01 <AaronSw> heh
00:24:16 <AaronSw> thelsdj, with ln -s you should only need to do it once
00:24:34 <thelsdj> between 2 websites i have 7 duplicates but basically they are just mirrors of what i have on my home machine's /var/www
00:25:15 <sbp> I like how vorpal has a web folder. I have that too but in a different place
00:25:23 <sbp> neither of us use apache's default location
00:25:31 <sbp> nor do we use each other's location
00:25:37 <thelsdj> hah
00:25:40 <thelsdj> i use the default on debian
00:25:59 <sbp> and plenty of people can't decide between /usr/local/bin and /usr/bin and /opt for e.g. firefox
00:26:42 <thelsdj> i've always said /opt is evil
00:27:06 <thelsdj> if i install from a binary package for my distro then it goes in /usr if its from source or a 3rd party binary its /usr/local
00:27:09 <sbp> in fact, now I just ln it to my home tools directory, so that it doesn't have root priviliges
00:27:25 <AaronSw> when i do-over vorpal, i'm going to use /public
00:27:27 <AaronSw> ala djb
00:27:28 <sbp> but that's a problem since to install XPI extensions you have to run a su copy
00:27:35 <sbp> do-over?
00:27:42 <AaronSw> sorry, redo
00:27:43 <sbp> thelsdj: I agree that /opt is evil
00:28:01 <thelsdj> who's silly idea was opt in the first place?
00:28:13 <sbp> thelsdj: so where would you put python? where would you put firefox?
00:28:34 <sbp> $idiot
00:28:52 <thelsdj> depends on how i install it, i haven't installed firefox because i'm waiting for a debian package
00:29:01 <thelsdj> i also have python installed from debian package
00:29:04 <thelsdj> so its all /usr
00:29:08 <sbp> I think I prefer web, Aaron. why /public? just because djb's using it?
00:29:18 <AaronSw> yes
00:29:19 <sbp> hmm
00:29:23 <thelsdj> small apps that compile to just like 1 binary file i usually put in ~/bin
00:29:31 <AaronSw> and because it's not just for web
00:29:33 <thelsdj> or scripts i've written
00:30:02 <sbp> I'm quite disappointed I got sbp instead of miscoranda.com
00:30:07 <thelsdj> i have some weird scripts in my ~/bin that i don't even remember the purpose of
00:30:13 <sbp> it's like "you're never getting another folder here ever. muahahahaha!"
00:30:27 <sbp> plus I can never remember
00:30:32 <sbp> plus I use domain names
00:30:36 <sbp> but really I should use /http
00:30:42 <sbp> instead of /web... I think
00:30:58 <sbp> thelsdj: ooh!
00:30:58 <thelsdj> hmmm ripper.py now whats that
00:31:06 <sbp> publish them, publish them, publish them
00:31:15 <sbp> I'm obsessed with people's ~/bin scripts lately
00:31:43 <sbp> I have about 100, and I'm currently going through them and smartening them up for release
00:31:49 <thelsdj> ah ok, now i remember, it takes an mp3 and two time signatures and outputs a wav file of the audio between the time sigs (currently you have to put in the exact # of milliseconds
00:32:02 <sbp> oh great!
00:32:09 <sbp> see that's the kind of thing I wanna see :-)
00:33:01 <sbp> cd && tar -cf bin.tar bin && gzip -9 bin.tar && send-to-www-archive bin.tar.gz
00:33:32 <sbp> echo K PLZ THX
00:33:59 <thelsdj> ok now i have no fucking clue what this one is
00:34:17 <thelsdj> http://jane.no-ip.com/BINiebabies/findstrange.pl
00:34:26 <sbp> * sbp thinks this is a lesson to document your code
00:35:06 <thelsdj> ooh right
00:35:17 <thelsdj> it looks at a list of mp3 filenames and outputs the ones that don't fit to known naming standards
00:35:41 <sbp> odd. I think someone on #joiito was asking us to debug some very similar code recently
00:38:09 <thelsdj> http://jane.no-ip.com/BINiebabies/ is the scripts in my ~/bin that i wrote, theres lots of other weird scripts all over my ~/ and /var/www
00:40:11 <sbp> thanks!
00:40:14 <thelsdj> i took too long to release intelliaudio, just the other day i noticed an app in debian that does the exact same thing
00:40:22 <verbosus> *** verbosus has quit ("Client exiting")
00:40:35 <thelsdj> and probably much better heh
00:41:01 <thelsdj> actually i think intelliaudio has a freshmeat page, along with one i wrote called projdiary
00:41:30 <thelsdj> projdiary took a look at an advogato project and output an html pages of the last N diary entry for each memeber of the project
00:42:19 <AaronSw> sbp, should I publish my currently reading list?
00:42:27 <Jaramir> hello there :o)
00:42:52 <sbp> * sbp finds his proxy MIME$ hack working wonderfully
00:42:56 <sbp> AaronSw: of course!
00:42:58 <sbp> hi there Jaramir
00:43:25 <Jaramir> :o)
00:44:33 <AaronSw> http://www.allconsuming.net/xml/users/currently_reading.aaronsw.js
00:44:53 <sbp> er
00:45:01 <Jaramir> document.write?
00:45:13 <sbp> * sbp uses text/html$ on it... so great
00:45:46 <AaronSw> MIME$?
00:45:46 <sbp> try http://allconsuming.net/weblog.cgi?url=http://www.aaronsw.com/
00:45:50 <sbp> yeah
00:46:30 <sbp> I set up proxy so that when I come across a file like http://example.org/perl.pl and it's being served with some very stupid MIME type, I can visit http://text_plain$example.org/perl.pl to rectify the problem
00:46:35 <Jaramir> cool, "watches weblogs for books" :o)
00:46:36 <sbp> dirty, dirty hack but it works so well...
00:46:55 <Jaramir> just curios how many are reading LoTR..
00:46:57 <AaronSw> hm, the js is missing a book
00:47:02 <sbp> read it
00:47:03 <sbp> yeah
00:48:12 <AaronSw> Hm, I don't get the cover of this map book
00:48:33 <thelsdj> ah interface
00:48:36 <thelsdj> that book rocks
00:48:49 <thelsdj> make sure to check out the other "stephen bury" book cobweb
00:52:04 <Jaramir> seeya!
00:52:07 <AaronSw> Douglas Allchin: '
00:52:07 <AaronSw> By the end, the reader is well prepared to appreciate the cover image: a shadow of Pinnochio's profile cast across the arcadian "Meadowdale Estates," surrounded by poisionous sites.'
00:52:15 <sbp> hmm. perhaps the best classification scheme is the one that results in the least cross-directory links for other classifications
00:52:20 <Jaramir> *** Jaramir is now known as Jara[zZ]
00:52:24 <sbp> c'ya Jara[zZ]
00:52:52 <sbp> * sbp wonders how to easily measure that
00:53:20 <AaronSw> hierarchical classification schemes are stupid
00:54:14 <AaronSw> rave reviews: "This book was my sole reading material on a camping trip, and if it hadn't been a gift from a friend, I would have used it, page by page, to add heat to the campfire."
00:55:10 <AaronSw> Another review is entirely about the overuse of the elipse in the book
00:55:25 <thelsdj> bahaha
00:58:28 <thelsdj> http://allconsuming.net/weblog.cgi?url=http://jane.no-ip.com/blog/
01:00:05 <sbp> ah, it seems that you need subsets in order to optimize
01:00:28 <sbp> because it doesn't matter how *many* are cross-categories, it's how many cross-categories there are
01:02:08 <sbp> this is my test set: [[[
01:02:08 <sbp> n3s - sw code
01:02:09 <sbp> swintro - sw doc
01:02:09 <sbp> htmlhistory - other doc
01:02:09 <sbp> bproxy - other code
01:02:09 <sbp> wypy - other code
01:02:11 <sbp> cwminfo - sw doc
01:02:13 <sbp> swhackfaq - other doc
01:02:15 <sbp> pwyky - other code
01:02:17 <sbp> enquire - other doc
01:02:19 <sbp> pyrple - sw code
01:02:21 <sbp> ]]]
01:02:25 <sbp> and my two example classifications:
01:02:27 <sbp> i.
01:02:29 <sbp> /sw - n3s swintro cwminfo pyrple
01:02:31 <sbp> /other - htmlhistory bproxy wypy pwyky enquire swhackfaq
01:02:33 <sbp> ii.
01:02:35 <sbp> /doc - swintro htmlhistory cwminfo swhackfaq enquire
01:02:37 <sbp> /code - n3s bproxy wypy pwyky pyrple
01:03:42 <deltab> there's a discussion on xdg-list at the moment about whether mathematics is a science
01:03:53 <deltab> sbp: makes sense
01:04:08 <thelsdj> what is xdg-list's normal topic?
01:04:08 <sbp> of course, the /doc page could be split up across sw and other... hmm...
01:04:22 <sbp> .google xdg-list
01:04:23 <datum> xdg-list: https://listman.redhat.com/mailman/listinfo/xdg-list
01:04:41 <sbp> that sounds pretty hillarious
01:04:52 <deltab> thelsdj: freedesktop standards
01:05:41 <thelsdj> haha
01:06:14 <thelsdj> hmm the archives don't show february
01:06:51 <sbp> * sbp tries https://listman.redhat.com/cgi-bin/search.cgi?q=mathematics+science&ps=20&o=0&m=all&wm=wrd&wf=222210&ul=%2Farchives%2Fxdg-list to no avail
01:07:33 <sbp> hmm
01:07:48 <sbp> perhaps I could have a central list of project names and their keywords
01:08:12 <sbp> so for example, I could then have n3s at /n3s, but because its profile is "sw code", it'd also appear in /sw and /code
01:08:32 <sbp> (and display its other keywords and links to those lists)
01:08:43 <deltab> you could replace 'other' by 'needs better categorization'
01:08:51 <sbp> heh. yes...
01:09:08 <sbp> though then it'd turn up in /needs, /better, and /categorization :-)
01:09:35 <sbp> in fact, other could just be deleted
01:10:01 <sbp> so "swintro - sw doc; htmlhistory - doc; pwyky - code" etc.
01:10:28 <sbp> as long as they turn up on *one* of the pages, that's fine
01:10:50 <sbp> so really it's classes and instances
01:11:03 <sbp> /n3s is an instance folder, belonging to both /sw and /code
01:11:21 <sbp> which implies that I should have a /folder class... :-)
01:11:36 <sbp> (actually, that's not a bad idea: but /categories or /keywords would be better)
01:11:57 <sbp> ugh. class plural problem
01:12:46 <thelsdj> http://euclid.trentu.ca/math/sb/misc/mathsci.html
01:15:44 <sbp> hmm. /index
01:18:26 <thelsdj> the root directory of my website is such a mess, the one on my home box has 139 files/dirs at the first level
01:18:41 <sbp> yeah. that's a problem with this scheme
01:18:45 <sbp> mess at the first level...
01:18:57 <sbp> the problem is: 7+/-2
01:19:05 <sbp> or is that 5+/-2...
01:19:09 <thelsdj> the one on my hosting service has 112 at the first level heh
01:19:26 <sbp> in any case, anything in double digits is bad for choosing and browsing
01:19:41 <thelsdj> basically my problem is i just drop stuff in there so people can see it and never really bother to sort it much
01:19:44 <sbp> and anything in 1,2,3 is a waste of space
01:19:56 <sbp> yeah
01:20:02 <sbp> I guess that's what 200X is for
01:20:12 <sbp> and misc for d8uv, and 2002 for Aaron since he doesn't increment
01:20:27 <sbp> more of a problem for local filesystems, though
01:20:41 <sbp> since on the web you rarely see straight directory lists
01:20:54 <sbp> well, you do, but they're stupid and not required
01:21:15 <thelsdj> that'd be interesting, always have the front page of your website be basically a directory listing to force yourself to sort it
01:21:25 <sbp> hmm. yeah
01:21:46 <sbp> I've always admired Hickson and Syrreni's sorting procedures
01:21:48 <deltab> a good thing about folders on the Mac and Amiga is that you can visually group icons
01:21:53 <sbp> but their sites still look horrid
01:22:10 <deltab> Windows too? I can't remember
01:22:12 <sbp> hmm. screenshot example, deltab?
01:22:19 <sbp> you can group icons by type
01:22:24 <thelsdj> OS X isn't as good as mac used to be for that because it wastes too much space between icons
01:22:38 <sbp> (type meaning file extension)
01:22:47 <deltab> I mean you can move them into a clump
01:23:21 <sbp> there's no fixed grid if that's what you mean...
01:23:27 <deltab> right
01:23:37 <sbp> (unless you turn on auto-arrange)
01:23:41 <deltab> and things stay where you put them
01:23:44 <sbp> yeah
01:23:47 <thelsdj> windows only does that well on the desktop i think
01:23:55 <sbp> yeah...
01:24:49 <thelsdj> i wonder if anyone is still using the rss scraper i wrote for suprnova.org months ago, i'm not even sure it still works
01:24:55 <sbp> we need huge flexible 150 degree 3ft high touch screen monitors, really
01:25:00 <sbp> it'd allow some great interfaces
01:25:02 <thelsdj> definatly
01:25:23 <deltab> I once read about making pipelines using Mac folders
01:25:43 <sbp> .google pipelines using Mac folders
01:25:44 <datum> pipelines using Mac folders: http://www.osxfaq.com/Tutorials/LearningCenter/
01:25:52 <sbp> doesn't look promising
01:26:07 <deltab> you put the filters in a line, then drop something onto the program at the end
01:26:32 <deltab> the programs examine the window they're in and their names to work out what to do
01:26:42 <deltab> it was a DDJ article
01:28:22 <sbp> aha: http://www.ddj.com/documents/s=961/ddj9605b/
01:28:56 <arcon> what's wrong with heirarchal classification?
01:29:06 <arcon> doh, damn scroll bar ;P
01:29:11 <sbp> hierarchies are evil
01:29:22 <arcon> how so?
01:29:37 <thelsdj> i think hierarchies are only evil if the interface to access them is evil
01:29:40 <sbp> see above!
01:29:46 <arcon> damn ;P
01:29:57 <sbp> more accurately, being forced to use hierarchies is evil
01:30:07 <sbp> some applications are naturally hierarchial
01:30:20 <sbp> but for things like website organization, keywords seem to be much more natural
01:30:32 <sbp> arcon: look at my sw & other vs. code & doc example
01:30:40 <sbp> and my proposed keyword-ish solution
01:30:43 <deltab> I've also set up pipelines on the icon bar in RISC OS; but the programs weren't connected: I had to drag the output of each into the next
01:30:59 <sbp> hehe. manual pipelining
01:31:46 <deltab> yeah, but easier because the icons were in order :-)
01:31:57 <deltab> less far to drag too
01:32:09 <sbp> that should've been an article
01:32:32 <sbp> I think that now you've released code, the next thing you need to do is an article
01:32:52 <sbp> and I want to see Python code too
01:33:06 <deltab> about what?
01:33:29 <arcon> so the problem with heirarchies are the lock you into a solid definition?
01:33:40 <sbp> hey, you can write the next miscoranda post about hierarchives and file organization if you want :-)
01:33:48 <sbp> * sbp tries to think of a serious suggestion...
01:33:56 <sbp> I'm not sure. what're you into at the moment?
01:34:16 <deltab> sbp: that's something I've thought about quite a bit
01:34:19 <sbp> arcon: yeah. you only expose one possible view of the system, and often you want a different view
01:34:36 <sbp> then you should have a good answer!
01:34:43 <deltab> sbp: most recently in response to http://bigfractaltangle.com/archive/2004/02/09.jsp
01:34:46 <arcon> so what if you use a heirarchial tree for storing with a semantic map on top?
01:34:56 <arcon> like a database with views
01:35:16 <deltab> arcon: why have the tree?
01:35:28 <arcon> because you need a way to store the data without duplication?
01:35:50 <arcon> for example, I use a heirarchy to store files locally
01:35:52 <sbp> sorry, I misunderstood what you were responding to
01:35:59 <arcon> Images, Movies, Notes, Programs, etc.
01:36:01 <sbp> but still... that's even more interesting
01:36:29 <arcon> I have been playing with giving the fles semantic references so I can browse them by meaning instead of heirarcy
01:36:45 <arcon> I could store them by meaning but initially (before they are given meaning) I won't have a place to store them
01:36:56 <arcon> heh , s/fles/files/
01:37:15 <sbp> also, people are loathe to add metadata
01:37:22 <arcon> plus, if a file has multiple meanings it would have to be stored in two locations...
01:37:28 <sbp> Seth Nickell found this with GNOME storage: you have to glean the data from the files themselves
01:37:51 <arcon> isn't a heirarcy like a minimal for of metadata?
01:37:55 <sbp> so my advice is that, for example, Notes is redundant because each note you make should just be in a file that has two attributes: its creation time and its content
01:38:00 <sbp> yes
01:38:08 <sbp> but it's a bad one
01:38:18 <arcon> what would be a better way to store my files?
01:38:31 <sbp> probably nothing on current OSes
01:38:34 <sbp> I have to sadly admit
01:38:37 <arcon> heh
01:38:45 <arcon> continue witht eh NOtes bit, I'm interested
01:38:50 <deltab> sbp: depends on the interface - if it's lots of fill-in-this-form, yes
01:38:56 <arcon> currently my notes are just a date and content, the date is the file name
01:38:56 <sbp> well, for example, I have this program on my computer:
01:39:14 <sbp> [[[
01:39:14 <sbp> $ cat note
01:39:15 <sbp> #!/bin/sh
01:39:15 <sbp> # Don't forget that cygwin sh is actually ash--irritating
01:39:15 <sbp> DATETIME=$(date +%Y%m%d-%H%M%S)
01:39:15 <sbp> echo "$@" > ${HOME}/notes/${DATETIME}.txt
01:39:17 <sbp> echo Made note at ~/notes/${DATETIME}.txt
01:39:19 <sbp> ]]]
01:39:31 <sbp> snap
01:39:41 <arcon> ok, that's very similar to my process
01:39:56 <sbp> I didn't always do it that way, though
01:40:06 <sbp> in fact, I have an entire article floating about on the "other" way
01:40:29 <sbp> in fact, I worked solidly for half a year on the notes problem
01:40:40 <arcon> hehe, I am just starting to fuss over it
01:40:41 <sbp> produced thousands of lines of code to try to get somewhere on it
01:41:00 <deltab> sbp: wow, you're worse than Morbus :-)
01:41:04 <arcon> hehehe
01:41:10 <arcon> that's bad
01:41:30 <sbp> it basically came down to discussing it with shellac et al. in Bristol and realising that I needed to rewrite the OS
01:41:34 <arcon> I am currently working on something to 'glue' together my 'cruft' - notes, misc files, etc.
01:42:04 <arcon> since I host some sites locally my plan was to have my notes be fed into an app that would make them simple xml fragments
01:42:07 <sbp> arcon: did you see what I was talking about the other day, attempting to annotate files based on their iNodes?
01:42:11 <sbp> er, inodes. heh
01:42:14 <sbp> Mac moment there
01:42:20 <arcon> then the files would be in a dir and a script would throw it all onto a webpage
01:42:29 <arcon> sbp, nope, I missed that
01:42:29 <sbp> well, Apple
01:42:33 <arcon> I was in and out a lot yesterday
01:42:41 <sbp> XML fragments? hmm
01:42:59 <arcon> like, the xml would be <note date="xxxx"></note>
01:43:03 <sbp> anyway, the basic idea is that you'd be able to assign notes and keywords and so forth to files
01:43:06 <arcon> with something in the note of course
01:43:12 <sbp> please... don't both with XML for *that*
01:43:14 <arcon> and an xsl to transform that into a page
01:43:14 <sbp> er, bother
01:43:17 <sbp> ew
01:43:30 <thelsdj> sbp, i think 200_ would be good instead of 200X heh
01:43:39 <arcon> I was tring to keep it 'code free'
01:43:41 <sbp> <date> <space> <entry> would be so much easier
01:43:46 <sbp> XSLT is code
01:43:52 <arcon> I could just have it spit out html fragments
01:43:55 <sbp> thelsdj: suggestion noted... :-)
01:44:22 <sbp> xalan sometransform.xsl is no different to perl somescript.pl
01:44:25 <thelsdj> its a lot better than 200- or 200 i think
01:44:25 <arcon> I've been doing a craplod of playing around with xml/xslt and it kind of leaked in to be honest
01:44:45 <sbp> * sbp advises strongly against it
01:44:52 <arcon> I waws planning on having the browser parse it
01:45:04 <sbp> using XML where XML is not needed is a very bad thing to get started on
01:45:08 <sbp> even worse for RDF
01:45:11 <arcon> hehe
01:45:33 <sbp> huge Keep It Simple Stupid violation
01:45:44 <arcon> so then if I stick with a flat file and the date as the name, and use something to parse that into a web page of all my notes that's fine
01:45:49 <sbp> XML is the anti-KISS
01:45:57 <arcon> the concept is a bigger concern than the implementaiton
01:45:58 <sbp> yeah
01:46:02 <deltab> sbp: except where is n't, yes
01:46:04 <sbp> yeah
01:46:10 <arcon> true dat, XML is never simple
01:46:23 <deltab> it can be simpler than the alternative
01:46:24 <sbp> yeah. XML is useful sometimes...
01:47:05 <deltab> if you want to combine data from many files, it's probably easy if they're in XML rather than each in its own format
01:47:19 <deltab> er, easier
01:47:27 <sbp> hmm
01:47:29 <arcon> in this case they would all be in text format, I rarely find the need for markup of notes
01:47:39 <arcon> they are 90% brain dump before I forget it
01:48:00 <sbp> not so sure about that. if you look at complex XML applications that need to be merged, you find you still basically need an application for parsing each of them into a better condition
01:48:18 <sbp> format probably makes it a little worse
01:48:30 <sbp> I'm just saying it's not as much of a solution as it seems
01:48:43 <sbp> I think back to the infoset-in-RDF things...
01:48:44 <arcon> xml enforces parsing standardization, not content standardization
01:48:56 <arcon> tha'ts the problem I run into
01:49:01 <sbp> yeah
01:49:15 <deltab> why do you want to describe an infoset in RDF?
01:49:42 <sbp> deltab: because then you can manipulate XML files using RDF rules
01:49:48 <sbp> it's pure joy
01:50:00 <deltab> yeah, but is someone actually doing that?
01:50:05 <sbp> yeah me
01:50:07 <sbp> and Connolly
01:50:11 <deltab> ah, what for?
01:50:25 <sbp> well it never really worked all that well because CWM is so buggy
01:50:31 <sbp> but I was aiming for validation
01:50:42 <sbp> not sure what DanC was doing--a hundred and one things
01:50:45 <deltab> * deltab nods
01:50:53 <deltab> makes sense
01:51:15 <deltab> validation being about whether something satisfies rules
01:51:22 <sbp> I even wrote a schema language for it, and it wasn't bad. I think it was back when you had DTD or WXS, and so I aimed somewhere in between--I guess I got roughly into RELAX NG territory
01:51:24 <sbp> right
01:51:38 <sbp> it's like schematron, done properly
01:51:52 <sbp> .google XSLIR
01:51:53 <datum> XSLIR: http://infomesh.net/2002/03-10xslir/README.html
01:52:03 <sbp> there you go. implementation bad, concept good
01:52:18 <deltab> what benefits does it bring?
01:52:32 <sbp> the schema language is documented here: http://infomesh.net/2002/03-10xslir/xslir.n3
01:52:47 <sbp> actually, I think that the main benefit was just in that the schema language didn't suck
01:52:58 <sbp> the RDF side of it was a terrible pain in the ass, obviously, because it didn't work
01:53:06 <sbp> on the other hand: the model had to be made very clear
01:53:11 <sbp> so it wasn't as arbitrary
01:53:47 <deltab> and it didn't work because of problems with cwm?
01:53:56 <sbp> RDF forces you to think from the model first, and that's much better than retrofitting meaning to a syntax. for example, even the infoset itself was created *after* XML 1.0 syntax, and I think if the infoset had come first, it would've lessened a lot of the PI etc. crap
01:53:59 <sbp> yes
01:54:17 <sbp> there were some horrible list handling bugs which probably haven't even been fixed yet
01:54:35 <sbp> I filed some bugs when I started to run into them, but the deeper I got, the worse the bugs
01:54:43 <sbp> and eventually there was no way to work around it
01:54:51 <deltab> what sort of things?
01:55:08 <sbp> [[[
01:55:08 <sbp> There are a few little workarounds in the code (some of them due to
01:55:08 <sbp> quite severe bugs in CWM)
01:55:13 <sbp> ]]] - http://lists.w3.org/Archives/Public/www-rdf-logic/2002Mar/0008
01:55:16 <sbp> just searching... 'ang on
01:56:07 <sbp> this looks like one relevant one: http://lists.w3.org/Archives/Public/www-archive/2002Mar/0024
01:57:16 <deltab> * deltab watches Enterprise
01:57:26 <sbp> don't blame you
01:57:59 <arcon> heh, time to go find some dinner and hack on some stuff ;)
02:01:34 <thelsdj> I so hate people who don't provide a feed that has the full text of their posts
02:02:03 <sbp> heh, yeah
02:02:13 <sbp> * sbp is glad to say that he provides the full text of his
02:02:19 <sbp> but I worry that it's too much
02:02:20 <sbp> oh well
02:02:22 <thelsdj> yes you do
02:02:24 <thelsdj> * thelsdj pets
02:02:28 <sbp> heh, heh
02:02:54 <thelsdj> hmm up to 24 subscriptions in my latest attempt at following my aggregator
02:03:12 <sbp> cool
02:06:32 <thelsdj> * thelsdj continually reloads the usps.com tracking page
02:07:57 <Ash> splodge: stop
02:13:22 <sbp> hi Morbus
02:13:28 <sbp> * sbp blogs Jim's PUT wiki: http://miscoranda.com/88
02:13:29 <Ash> splodge: OFF
02:13:35 <Ash> splodge: OFF
02:13:48 <sbp> splodge: on
02:13:50 <Ash> splodge: on
02:13:51 <Ash> splodge: on
02:13:51 <Ash> splodge: on
02:13:51 <sbp> heh, heh
02:13:51 <Ash> splodge: on
02:13:51 <Ash> splodge: on
02:13:54 <Ash> splodge: on
02:13:59 <Ash> splodge: on
02:14:01 <Ash> splodge: on
02:14:07 <Ash> are you logging or what, splodge
02:14:09 <Ash> MAKE UP YOUR MIND
02:14:11 <sbp> Morbus: your call has been registered
02:14:21 <sbp> Morbus: please call back later for confirmation
02:14:29 <sbp> * sbp waits for Ash to be done
02:14:39 <sbp> ...
02:14:41 <sbp> ...
02:14:43 <sbp> ...
02:14:46 <sbp> lalala
02:14:48 <sbp> LALALALALA
02:15:02 <sbp> splodge, grep LALALALA
02:15:07 <splodge> I'm logging. I found 2 answers for 'LALALALA'
02:15:08 <splodge> 0) 2004-02-11 02:15:02 <sbp> splodge, grep LALALALA
02:15:09 <splodge> 1) 2004-02-11 02:14:48 <sbp> LALALALALA
02:15:12 <thelsdj> hmmm heres an interesting concept, i don't want all my mp3s to sit in my web dir, but i'm always sharing a list of them with people so i'm always doing a 'find /mnt/data/mp3 > /var/www/mp3s.txt' or 'find /mnt/data/mp3/Mixed > /var/www/mixes.txt" i wonder if anyones built a general mechanism for keeping up to date lists of files but not allowing access to the files themselves
02:15:25 <sbp> man. I'm going to be checking like very hour now to see if it's loggign
02:15:49 <sbp> thelsdj: just set up a cron job to do that or something
02:15:53 <sbp> Morbus: what's the problem?
02:16:03 <Morbus> i'm back to my old two column problem.
02:16:08 <sbp> ah
02:16:12 <Morbus> need both columns the same height.
02:16:16 <sbp> oh man
02:16:23 <sbp> that's easy enough
02:16:32 <Morbus> example?
02:17:44 <sbp> sure. hang
02:18:38 <thelsdj> sbp, sure but what if someone wants a list of my trance mp3s? i don't want to have a cron job for every possible list
02:18:55 <sbp> ah, I see
02:19:06 <sbp> can you extract the data from the files?
02:19:17 <thelsdj> i've already got them sorted in folders
02:19:41 <thelsdj> so what i'd want is a general purpose web tool that when given a dir would recursivly list all files
02:19:48 <thelsdj> i was just curious if it'd been already done
02:20:35 <d8uv> *** d8uv (d8uv@16-67-237-24.gci.net) has joined #swhack
02:20:44 <d8uv> Liquid Molten Death.
02:21:28 <thelsdj> reminds me of my drink idea, 'Death on the Rocks'
02:22:21 <sbp> feck
02:22:28 <sbp> I know I saw a damn good example somewhere
02:24:21 <thelsdj> 2 oz. arsenic, 3 oz. sweet & sour, 3 oz. coke, 2 ice cubes
02:24:24 <sbp> Morbus: whyn't use a standard like http://glish.com/css/9.asp ?
02:24:55 <Morbus> sbp: because the sidebar shouldn't expand into the background color.
02:25:06 <Morbus> http://63.173.138.175/~morbus/libdb/index.cgi
02:25:15 <Morbus> that was one suggestion that failed.
02:25:36 <sbp> what's wrong with /~morbus/libdb/index.cgi?
02:25:43 <Morbus> it looks like ass.
02:25:44 <splodge> * splodge is already logging
02:25:48 <splodge> * splodge is already logging
02:25:49 <Morbus> the yellow at the bottom?
02:25:49 <Morbus> dumb.
02:25:54 <splodge> * splodge is already logging
02:26:02 <splodge> * splodge is already logging
02:26:10 <splodge> * splodge is logging
02:26:30 <sbp> er
02:26:38 <sbp> you mean the yellow underneath the white section?
02:26:50 <Morbus> yeah
02:27:04 <sbp> okay. let me try to fix it
02:27:10 <sbp> firefox's view source really sucks ass btw
02:27:50 <sbp> oh... this is driving me nuts
02:28:35 <sbp> * sbp wonders if the proxy is causing it
02:29:54 <sbp> where on earth is the CSS file?
02:31:32 <sbp> okay, got it
02:32:44 <Morbus> Now Playing: 'The Fall' from Jerry Goldsmith's album 'The Omen' (Rating: 4/5).
02:33:04 <sbp> argh, you took the yellow out?
02:33:06 <deltab> yay, Jerry Goldsmith
02:33:08 <sbp> put it back, put it back!
02:35:33 <sbp> that '<div class="clear"> </div>' thing at the bottom was what was causing it
02:35:49 <Morbus> that was required for IE5, and some other reason too.
02:35:51 <sbp> I just commented it out and it worked
02:36:00 <sbp> sigh
02:36:04 <Morbus> indeed
02:36:11 <sbp> <div class="main clear">
02:36:34 <sbp> ouch
02:36:37 <sbp> no, don't do that
02:37:27 <sbp> I dunno
02:37:40 <sbp> I subscribe to the Aaron Swartzian school of page design at the moment
02:37:42 <Morbus> yeah, its like, css sucks.
02:37:45 <Morbus> which is?
02:38:02 <sbp> Gill Sans for the headings, let the browser figure out the rest
02:38:28 <sbp> actually, I've just redesigned my homepage to be a quite snazzy CSS design
02:38:41 <sbp> but that's the most decent design I've done for ages
02:38:47 <sbp> and it probably looks crap in IE5 etc.
02:38:52 <sbp> * sbp shrugs
02:39:24 <sbp> bah, it looks crap in IE *6*
02:41:00 <Morbus> * Morbus chuckles.
02:41:14 <sbp> * sbp fixes that quickly
02:42:03 <Morbus> css sucks.
02:42:08 <Morbus> i can't believe this is so difficult.
02:42:24 <sbp> I'm thinking about using two giant images on mine
02:42:46 <sbp> GIFs, too
02:42:53 <sbp> just to rebel
02:43:06 <sbp> GIFs get layouts *perfect*
02:43:13 <sbp> just set the alt and you're fine
02:43:29 <sbp> they don't re-wrap to well, though
02:43:32 <sbp> oh!
02:43:33 <Morbus> [21:42] <Atua2> I hear you, Morbus
02:43:33 <Morbus> [21:43] <Morbus> reallly? i haven't opened my mouth for like five hours.
02:43:50 <sbp> but they do display unicode on any browser that supports them
02:43:51 <sbp> heh
02:44:17 <Morbus> [21:43] <Atua2> one more dumb question before I'm done
02:44:20 <Morbus> [21:43] <Atua2> --how fast do you type?
02:44:24 <Morbus> [21:43] <Morbus> as fast as i need to ;)
02:51:21 <sbp> freaky. the images actually work really well
02:51:42 <thelsdj> I've finally found my favorite use for bittorrent, downloading foreign movies that don't even end up in the foreign section of the video stores
02:55:49 <mediovia__> Hey, Morbus. We met a long time ago when I was a doid
02:56:16 <Morbus> a doid?
02:56:23 <mediovia__> thelsdj: thanks for the link to the experimental bittorrent client - very useful
02:56:24 <Morbus> have you graduated to hemmorhoid?
02:56:34 <mediovia__> Morbus: was my nick ages ago
02:56:37 <mediovia__> heh
02:59:09 <splodge> * splodge is already logging
02:59:20 <mediovia__> I tried reinstalliing Amphetadesk a few weeks ago, but had some problems, as I remember it, with dependencies fro zlib, or zlib1g, would that make sense?
02:59:36 <mediovia__> * mediovia__ finishes lame linkdump at http://okdisco.com/wiki.asp?page=LinkDump and considers making for bed
02:59:48 <Morbus> mediovia__: comment "use Compress::Zlib" in AmphetaDesk.pl
03:01:14 <mediovia__> Morbus: ah, I'll give that a go, thanks
03:02:23 <mediovia__> MS VPC for Mac Elevation of privilege vulnerability, but probably noone here uses it
03:02:45 <d8uv>
03:03:00 <d8uv> Gah. My computer is teh suck.
03:03:30 <sbp> man. nobody spends as much time on their homepages as I do, and yet everyone else's sucks much less than mine
03:03:38 <sbp> maybe too much effort is what borks it
03:03:47 <sbp> but then I look at older version and think that no, they were pretty bad
03:04:12 <sbp> presentation is everything is presentation is everything is presentation
03:04:41 <sbp> this new one's getting towards good
03:04:43 <sbp> but!
03:04:49 <sbp> I have the organization problem now
03:05:23 <sbp> do I point people to my code and documentation, my semantic web stuff and web architecture stuff, or what?
03:05:30 <sbp> weblog and outside stuff...
03:05:31 <sbp> eek
03:05:40 <sbp> classification problem overload
03:06:18 <sbp> what soup of sentences will be most optimal to people coming to the page
03:06:23 <sbp> that's what keeps bugging me
03:06:28 <thelsdj> mediovia__, have you read eastern standard tribe yet?
03:06:30 <sbp> unsolvable, but... must... solve...
03:07:08 <d8uv> Sem web, Web arch, Code + Doc.
03:07:15 <d8uv> That's what I think.
03:07:26 <thelsdj> i like the idea of having a header on your blog that doesn't change, introducing people to what you do and links to various stuff thats not the blog
03:07:27 <sbp> and then no link to miscoranda? or email and IRC links?
03:07:32 <thelsdj> yet have all the blog entries underneath
03:07:46 <sbp> the problem is that the current design has a single toolbar at the top with limited width
03:07:57 <sbp> thelsdj: that's another thing... why do so many people have their blogs as their homepages?
03:08:05 <sbp> I could never, never have a blog as a homepage
03:08:18 <sbp> I don't think anything else says "hi I suck" harder than that
03:08:28 <thelsdj> well depends on how much information you want on your homepage
03:08:49 <sbp> I suppose
03:08:55 <sbp> the scary corner case is deltab
03:09:08 <sbp> I take it that privacy concerns overwhelm him most
03:09:18 <sbp> maybe tav caught it from deltab
03:09:26 <thelsdj> i have the most silly homepage ever
03:09:31 <thelsdj> take a look at http://jane.no-ip.com/
03:09:33 <thelsdj> theres like nothing there
03:09:44 <thelsdj> just quick links i put cause i was tired of remembering some urls
03:09:52 <mediovia__> thelsdj: no, don't think I have unless it's Cory Doctorow's novel that was part published iin Salon
03:10:00 <thelsdj> and half the links don't work anymore
03:10:21 <sbp> wow, that's pretty brutal
03:10:37 <sbp> reminds me of John Cowan's... except, even smaller. I've seen quite a few of that size, though
03:10:49 <sbp> actually, I like those because they direct people to other stuff that matter
03:11:02 <sbp> though it might be a good idea to have just a little bit about you
03:11:14 <sbp> but what do I know? deltab's in complete contradiction to Aaron. it's unsettling
03:11:30 <thelsdj> well i was just pointing out that i don't really have a homepage haha
03:11:38 <sbp> even Morbus doesn't seem to particularly care, and he's Mr. Code Shui. but then there's always Google...
03:11:39 <sbp> yeah
03:12:14 <sbp> * sbp gets frightened in case people use ".google Sean B. Palmer" as his homepage...
03:13:51 <sbp> /sbp/ is the second highest ranked page on infomesh.net, though, after swintro
03:14:29 <Morbus> *** Morbus has quit (Client Quit)
03:16:11 <sbp> I guess what you do is more important than how you talk about what you do
03:16:17 <thelsdj> hehe ".google Adam Wendt" is almost just as bad because MT regenerated my page when there weren't any new posts to put on it so its basically empty
03:16:28 <sbp> whoops
03:16:39 <thelsdj> except for the sidebar
03:16:43 <thelsdj> so it looks a bit out of place
03:17:59 <mediovia__> * mediovia__ just uses the full text search on his site, and that hardly ever get him what he wants
03:33:24 <sbp> woah. 2004-02-06 was the fifth largest #swhack log ever
03:33:40 <sbp> 2004-02-10 is also well up there
03:34:04 <thelsdj> nice
03:38:18 <thelsdj> am i the only person that has more than 1 computer on my desk and hates not having an easy way to copy and paste from one computer to another? heh
03:38:55 <sbp> do you have them networked?
03:38:58 <sbp> should be easyish
03:39:16 <thelsdj> yea, my brain is working on ideas right now
03:40:13 <sbp> burp.sh:
03:40:16 <thelsdj> would want it to be cross platform hrm
03:40:26 <sbp> * save clipboard to ~/.clipboard
03:40:29 <sbp> slurp.sh:
03:40:43 <sbp> * read from /remotepath/.clipboard
03:40:47 <sbp> that's a problem
03:40:50 <sbp> what platforms?
03:41:19 <thelsdj> right now linux and os x, i guess i could mount an nfs path on os x or something
03:41:28 <sbp> yeah
03:42:08 <thelsdj> most of the time i'm just wanting to pull up urls that i have on one machine so i'm thinking of a bookmarklet that works like a stack, push on one browser, pop on another
03:42:18 <sbp> ah
03:42:20 <sbp> good idea
03:43:28 <mediovia__> which is why I use a wiki with bookmarklets, used to use KnowNow (javascript push)
03:43:52 <mediovia__> * mediovia__ goes back to EST
03:43:59 <thelsdj> heh
03:46:01 <thelsdj> i was having so much fun with xnest yesterday
03:46:27 <thelsdj> http://jane.no-ip.com/xnest.jpg
03:46:46 <thelsdj> also xnest2.jpg
03:53:20 <thelsdj> wow nedit's python mode doesn't auto ident after a line ending with :
04:01:08 <thelsdj> what simplest way to check if a file exists in python? i haven't had to think in python for quite a while heh
04:07:39 <thelsdj> nevermind, found os.path.isfile
04:10:15 <arcon_> *** arcon_ (~arcon@dsl027-161-088.atl1.dsl.speakeasy.net) has joined #swhack
04:10:23 <arcon_> *** arcon_ has quit (Read error: 104 (Connection reset by peer))
04:45:32 <tom> *** tom (~tom@ppp130-178.lns1.adl2.internode.on.net) has joined #swhack
04:48:48 <tom> *** tom has left #swhack
04:50:28 <thelsdj> proof of concept: http://jane.no-ip.com/pushpop/pushpop.py
04:54:54 <d8uv> thelsdj: I don't think it works.
04:56:01 <d8uv> Popping doesn't show the latest link on the stack.
04:57:34 <thelsdj> d8uv just fixed it
04:58:06 <thelsdj> (i think) hehe
04:58:20 <d8uv> Neat.
04:58:25 <d8uv> That's cool.
04:58:39 <thelsdj> yea, and it would have weird results with multiple people using it hahaha
04:58:46 <thelsdj> so its better for just inside your lan
05:23:03 <arcon> *** arcon has quit ("Leaving")
05:24:53 <mediovia__> neato. bbiab.
05:35:25 <mediovia__> links, links, links
05:36:11 <mediovia__> I only go to bed when I'm really tired, and as soon as I put my head down and relax I think "Wouldn't it be nice to have some links to go with this?"
05:38:06 <mediovia__> Actually went a bit further and got one of those lenses off ebay for making cheap tv projectors, thought I'd prop it over the laptop screen and project a scrolling page onto the ceiling
05:47:18 <AaronSw> their school newspaper is called the interrobang: http://tailsteak.netherweb.com/archive.php?num=161
05:47:21 <AaronSw> how cool is that?
05:49:21 <AaronSw> the comic is pretty funny too
05:50:04 <AaronSw> this guy claims to have trademarked "Interrobang": http://www.davetepper.net/
05:50:37 <AaronSw> while providng his site under a CC license
05:51:44 <d8uv> AaronSw: That's messed up.
05:52:38 <AaronSw> sbp, when are we going to see this fabled new homepage design?
05:53:08 <AaronSw> @@update: I noticed we even got Larry Lessig using @@.
05:53:17 <AaronSw> Actually, he only uses one @
05:53:20 <AaronSw> but it's close enough.
05:54:49 <sbp> sorry. you can see it now if you want
05:55:00 <sbp> I was just hacking on what'll be the world's shortest wiki. muahahaha
05:55:14 <sbp> d8uv, Aaron: if either of you are around, you can help me think of a name for it
05:55:18 <sbp> the language is PHP
05:55:23 <sbp> and don't ask
05:55:27 <thelsdj> sbp, you see my proof of concept?
05:55:29 <d8uv> PHP?!
05:55:29 <sbp> let me upload the new homepage...
05:55:31 <thelsdj> haha
05:55:32 <thelsdj> how short?
05:56:01 <sbp> wc gives: 3 11 307
05:56:07 <sbp> but I'm working on getting that shorter
05:56:24 <sbp> 4 lines actually; 3 just counts line breaks
05:56:37 <AaronSw> ok, this is getting scary now
05:56:51 <sbp> wait till you see it
05:57:03 <AaronSw> let me give you a name
05:57:10 <thelsdj> ooh my bittorrent tracker has surpassed 10k downloads
05:57:52 <thelsdj> i mean 10k completed downloads of files hosted on my tracker, not 10k downloads of a tracker i wrote
05:57:58 <deltab> wIcky
05:58:19 <thelsdj> wicka wicka weee
05:58:35 <AaronSw> wicky - heh!
05:58:39 <d8uv> Phwysical?
05:58:41 <AaronSw> err wIcky
05:58:49 <deltab> or w-icky
05:58:59 <d8uv> Oh. Didn't see the serifs. That's cool.
05:59:17 <AaronSw> ickywicky
05:59:46 <deltab> wickle
06:00:34 <thelsdj> wickly
06:01:06 <deltab> hmm, I don't know what that means
06:02:42 <thelsdj> ooh how about wikke
06:02:48 <thelsdj> dict it
06:03:00 <sbp> Aaron, what's the PHP extension on Vorpal?
06:03:09 <sbp> uh. not that I'm asking for any particular reason
06:03:11 <deltab> heh
06:03:30 <thelsdj> i think 'Wikke' would be great
06:03:32 <d8uv> You should call it "fdisk"
06:03:40 <sbp> rm-rf
06:04:26 <sbp> oh, I uploaded my homepage too. man, so busy I forgot about that
06:04:44 <sbp> [URI deleted]
06:04:56 <sbp> Aaaaron!
06:05:54 <thelsdj> sbp, for some reason that font doesn't look right to my eyes
06:05:59 <thelsdj> (the main text font)
06:06:16 <sbp> not used to serif, eh?
06:06:52 <d8uv> sbp: MiniWi. A lame reference BUT STILL
06:07:24 <thelsdj> no i guess i'm not, its like i can't read it without noticing the font
06:08:41 <d8uv> thelsdj: I can't read much text without instantly trying to ID the font either.
06:10:42 <sbp> argh, what the... I can't get PHP working
06:10:58 <sbp> Aaron, it'd be really really helpful if you were around right now
06:11:01 <sbp> * sbp does an Aaron dance
06:11:11 <sbp> Georgia
06:11:22 <d8uv> sbp: Gimme the code, I'll paste it up to the d8uv.com server.
06:11:47 <sbp> okay... er, but the thing is that it requires quite a bit of fucking around
06:12:01 <d8uv> Hmm...
06:12:09 <thelsdj> sbp, holdon a sec
06:12:11 <sbp> and also that shell_exec is enabled
06:12:39 <d8uv> Hmm... thelsdj could probably help you more.
06:13:07 <thelsdj> sbp, just create a public_html
06:13:11 <thelsdj> should work outa the box
06:13:38 <thelsdj> speed should be faster now if it was real lagged
06:15:01 <sbp> ah, okay
06:15:21 <thelsdj> i'll check the config file to make sure .htaccess works in there
06:15:44 <sbp> hey great, it seems to be working...
06:16:02 <thelsdj> should work
06:16:11 <thelsdj> and url should be jane.no-ip.com/~sbp/
06:16:17 <sbp> hehe. nightmare scenario: -bash: nano: command not found
06:16:21 <sbp> yeah, I'm already there. ooh
06:17:19 <thelsdj> try nano again
06:17:20 <d8uv> Heh. nano is the rock.
06:17:30 <thelsdj> apt-get install rocks
06:17:32 <sbp> it's okay; I'm scp-ing now
06:17:37 <sbp> aww... sorry
06:17:45 <thelsdj> hehe
06:17:51 <sbp> nano will probably be very useful in a mo' though...
06:18:06 <sbp> oh sweet, it's working
06:18:21 <sbp> and your PHP is loose enough to get this all running
06:18:23 <d8uv> Heh. I have far too many linux problems to even talk about Linux in general.
06:18:28 <d8uv> I'm not qualified.
06:18:33 <thelsdj> heh
06:18:39 <thelsdj> nice
06:18:48 <thelsdj> my system is VERY loose
06:18:54 <thelsdj> i'm supprised i don't get rooted every week
06:19:02 <sbp> heh, heh
06:25:18 <sbp> thelsdj: help? "You don't have permission to access /~sbp/put.py on this server."
06:25:21 <sbp> I've chmod 755'd it
06:26:26 <thelsdj> k holdon
06:26:33 <sbp> thanks
06:26:39 <sbp> * sbp appreciates this very much, bwt
06:26:41 <sbp> er, btw
06:27:45 <thelsdj> ok, fixed
06:27:52 <sbp> thanks
06:28:00 <thelsdj> no prob
06:37:09 <sbp> sigh, now the tricky part... battling Apache
06:37:33 <thelsdj> for write access or what?
06:37:52 <thelsdj> or you doing some weird URI thing?
06:38:00 <sbp> I need to basically map all of the requests for files that exist to put.py
06:39:28 <sbp> bzzt. "Internal Server Error"
06:39:43 <sbp> do I have my own logs, or could you possibly tail them for me, please?
06:39:50 <thelsdj> i think you just spelled something wrong
06:39:57 <thelsdj> Invalid command 'RewriteEnging',
06:39:59 <sbp> ah, you're right
06:40:00 <sbp> heh
06:40:09 <sbp> ta
06:40:23 <thelsdj> * thelsdj keeps a tail -f on the error_log
06:40:59 <sbp> good idea
06:42:09 <sbp> * sbp tries to find where he put QuickPut... argh
06:44:39 <sbp> ack. PUT is forbidden. any way I can get around that just locally?
06:44:48 <sbp> really, I only want to allow it for the put.py script
06:45:23 <thelsdj> hmm
06:45:28 <thelsdj> probably holdon
06:45:32 <sbp> ta
06:47:02 <thelsdj> k try again
06:47:26 <sbp> whoo!
06:47:29 <sbp> wow, you're good at this
06:48:16 <thelsdj> grep httpd.conf is my friend
06:49:42 <sbp> heh, heh
06:52:52 <sbp> * sbp takes a quick peek at RFC 2616...
06:53:56 <thelsdj> hah
06:57:05 <thelsdj> malformed header from script. Bad header=200 OK: /home/sbp/public_html/put.py
06:59:36 <thelsdj> still not right
07:05:20 <sbp> ack
07:05:33 <sbp> yeah, I'm trying to get it to return custom responses
07:05:41 <Ash> malformed butter from fridge. Bad butter=200 RANCID:
07:05:43 <sbp> the odd thing is that 201 seems to be working and yet 404 isn't
07:05:53 <Ash> What are you doing, sbp?
07:05:58 <splodge> * splodge is already logging
07:06:03 <sbp> wiki contest again
07:06:05 <Ash> did you know that people have written these things called 'http servers'
07:06:05 <thelsdj> the only one i ever saw errors for was 200
07:06:12 <Ash> oh great
07:06:16 <Ash> wikiwikiwanking
07:06:17 <sbp> $ GET --head http://jane.no-ip.com/~sbp/AnotherWikiName
07:06:17 <sbp> HTTP/1.0 200 OK
07:06:27 <sbp> it really should be 404ing...
07:07:03 <thelsdj> hmmm
07:07:56 <thelsdj> i got Not Found:
07:08:42 <sbp> really?
07:08:43 <sbp> hmm
07:08:45 <thelsdj> yea
07:08:57 <sbp> but then it's not being passed to the document that I called for ErrorDocument
07:08:57 <thelsdj> now i'm getting the php
07:09:13 <sbp> that's because I changed put.py...
07:09:16 <thelsdj> ah ok
07:09:18 <sbp> * sbp is fudging it
07:09:25 <thelsdj> i did get 404 once
07:10:02 <sbp> but not the PHP too: I suspect that one Apache hands off to the CGI, it doesn't care about what response code it gives any more, as long as there is one it'll pipe it out
07:10:33 <thelsdj> hmm yea that's too bad
07:10:38 <sbp> oh my goodness, it's working
07:10:41 <thelsdj> hahaha
07:10:55 <sbp> if you've got Mozilla, you can follow along!
07:11:38 <thelsdj> oh my
07:12:34 <sbp> I really ought to speak to Jim tomorrow and see if he can resolve that bug
07:12:52 <thelsdj> this hurts my brain
07:13:00 <sbp> heh!
07:13:57 <thelsdj> ah yes when you edit it adds and duplicates?
07:14:06 <sbp> * sbp sits back and relaxes a bit now, knowing that he's gonna probably womp the perlers tomorrow
07:14:07 <sbp> yeah...
07:14:22 <thelsdj> oh wait, not always
07:14:57 <sbp> really? when doesn't it duplicate?
07:15:05 <thelsdj> hmm check out foo
07:15:08 <thelsdj> it hasn't duplicated once
07:15:11 <sbp> yeah, I am at the moment
07:15:13 <thelsdj> just keeps adding to the beginning
07:15:25 <sbp> oh, right
07:15:32 <sbp> yeah, when the box clears itself it'll prepend
07:15:46 <sbp> when the box doesn't clear itself, so you're editing the text that's in there, it'll duplicate
07:15:57 <thelsdj> yea
07:16:01 <sbp> really you should be able to see the source in the box, edit it, and not have it duplicate...
07:16:03 <sbp> ideally
07:16:09 <irc.freenode.net> *** irc.freenode.net has changed the topic to: WikiWikiWank
07:16:10 <sbp> let's just say it's buggy at the mo'
07:16:12 <thelsdj> * thelsdj nods
07:16:12 <sbp> at least it works
07:16:13 <sbp> heh, heh
07:16:44 <thelsdj> hmmmm
07:16:55 <sbp> you can put in HTML, btw
07:17:02 <thelsdj> yea i noticed that
07:17:08 <sbp> and WikiNames get automatically converted to links
07:17:11 <sbp> cool
07:17:26 <sbp> hehe, this is so fun
07:17:50 <thelsdj> i don't see any link after the box like the HomePage talks about
07:18:16 <sbp> ah, it only occurs in pages that are linked to
07:18:37 <sbp> so if you create a link to NewPage on HomePage, and then go to create NewPage, it'll have a reference saying "HomePage"
07:19:01 <sbp> it's bending the backlinks rules a bit, but the perlers have bended the rules quite a bit too...
07:19:08 <thelsdj> hehe
07:19:32 <sbp> who'd've thought it? JavaScript, PHP, and Shell...
07:20:00 <thelsdj> haha
07:20:24 <sbp> "Code by Jim Ley and Sean B. Palmer, test site with thanks to Adam Wendt. We're very, very sorry..."
07:20:32 <thelsdj> hahahah
07:20:52 <sbp> I can't wait to write out the installation instructions...
07:21:12 <sbp> "1) Make sure you have a PUT handler set up if you don't already (but I'm sure you have one already... don't you?!)"
07:21:59 <sbp> "2) Make sure that your PHP installation is iffy and allows shell_exec(). Don't ask questions. No, you won't be r00ted."
07:22:32 <sbp> actually, as far as I can tell, this is very secure
07:22:57 <sbp> since the PUT handler is doing all the saving and it's *very* strict, it's better than the Perl ones, which just allow saving over the entire filesystem, AFAICT
07:23:41 <sbp> the handler only allows alphabetical filenames, so it can only write to the current directory, and won't overwrite even the scripts
07:24:15 <sbp> 296 characters...
07:26:33 <thelsdj> ah now i see the only php is for the backlinking
07:26:51 <d8uv> sbp: You have a test page I haven't bothered to look for?
07:27:41 <thelsdj> d8uv http://jane.no-ip.com/~sbp/AnyWikiName
07:28:29 <d8uv> Woah. That's cool.
07:28:45 <thelsdj> and the source is:
07:28:46 <thelsdj> <b><textarea onchange="x=new XMLHttpRequest();x.open('PUT',location);x.send(this
07:28:46 <thelsdj> .value.replace(/(([A-Z][a-z]+){2,})/g,'<a href=\u0022$1\u0022>$1</a>')+this.
07:28:46 <thelsdj> parentNode.innerHTML);setTimeout('location.reload()',999)"></textarea></b><?echo
07:28:47 <thelsdj> `n=$_SERVER[REQUEST_URI]&&echo \$(grep -l \${n##*/} *)`?>
07:28:51 <thelsdj> 4 lines
07:28:58 <sbp> and I've figured out the bug, too!
07:29:03 <thelsdj> nice
07:29:09 <thelsdj> or atleast its 4 lines on my terminal
07:29:56 <d8uv> Hmm... I see a bug on the anywiki page, but that's to be expected this early.
07:30:41 <d8uv> But this seems very awesome.
07:31:34 <deltab> why "n=$_SERVER[REQUEST_URI]"?
07:31:54 <thelsdj> to grep for links to that page
07:32:35 <deltab> why give a different name to a variable you only use once?
07:32:58 <thelsdj> thats in the shell
07:33:01 <sbp> because that gets interpolated
07:33:11 <thelsdj> its setting an evironment variable
07:33:13 <deltab> doesn't PH pass on the CGI vars?
07:33:14 <sbp> then the second variable is, as thelsdj says, in the shell
07:33:19 <sbp> no...
07:34:04 <thelsdj> hmm its no longer uploading for me, you twiddling with it?
07:34:10 <sbp> yep
07:34:12 <thelsdj> k hehe
07:34:50 <sbp> see: http://jane.no-ip.com/~sbp/env.php/TestingTheory
07:34:59 <sbp> env.php is <?echo `echo \${REQUEST_URI##*/}`?>
07:38:38 <sbp> okay, it's up and working again: and this time, no duplicates!
07:38:44 <sbp> the file size is a little bigger, though
07:38:49 <sbp> I did manage to save one char
07:38:52 <sbp> but I lost a few too...
07:39:08 <sbp> 309 chars now
07:39:43 <sbp> new source: [[[
07:39:43 <sbp> <b><textarea onchange="x=new XMLHttpRequest();x.open('PUT',location);x.send(this
07:39:43 <sbp> .value.replace(/(([A-Z][a-z]+){2,})/g,'<a href=\u0022$1\u0022>$1</a>')+'<b>'+
07:39:44 <sbp> this.parentNode.innerHTML+'</b>');setTimeout('location.reload()',999)"
07:39:44 <sbp> ></textarea></b><?echo`n=$_SERVER[REQUEST_URI]&&echo \$(grep -l \${n##*/} *)`?>
07:39:44 <sbp> ]]]
07:40:09 <sbp> perhaps I can do something about those \u0022 things, and the 999
07:41:07 <deltab> \x22?
07:41:21 <sbp> ah, yeah
07:41:25 <sbp> heh, I'd just tested that
07:41:27 <deltab> or \'
07:41:32 <sbp> hmm
07:41:35 <sbp> that's a good idea
07:42:31 <Ash> and to think that sbp complains about not getting anything productive done
07:42:35 <Ash> GEE I WONDER WHY THAT HAPPENS
07:42:42 <deltab> hehe
07:44:52 <sbp> heh, heh
07:44:57 <sbp> this is productive enough
07:46:05 <Ash> uh huh
07:46:08 <thelsdj> hmmm what about all the bold tags
07:46:10 <Ash> you just keep telling yourself that sbp
07:46:11 <Ash> ;-)
07:46:30 <sbp> the bold tags are to prevent the duplication
07:46:35 <thelsdj> ah
07:46:37 <sbp> Jim's hack: I misimplemented it the first time
07:46:42 <sbp> current source: [[[
07:46:42 <sbp> <b><textarea onchange="x=new XMLHttpRequest();x.open('PUT',location);x.send(
07:46:42 <sbp> this.value.replace(/(([A-Z][a-z]+){2,})/g,'<a href=\'$1\'>$1</a>')+'<b>'+this.
07:46:42 <sbp> parentNode.innerHTML+'</b>');setTimeout('location.reload()',99)"></textarea>
07:46:42 <sbp> </b><?echo`n=$_SERVER[REQUEST_URI]&&echo \$(grep -l \${n##*/} *)`?>
07:46:43 <sbp> ]]]
07:47:02 <sbp> having 99 instead of 999 means that it doesn't refresh right sometimes
07:47:06 <sbp> but... another bug
07:48:05 <sbp> shame XMLHttpRequest().open('PUT',location).send won't work
07:48:20 <sbp> perhaps one can leave the "new" out?
07:48:41 <sbp> though I'm sure Jim would've done that if he could...
07:49:43 <sbp> nope. doesn't work
07:50:06 <thelsdj> what if the php code doesn't get executed the first time and then if your php handler handled all files then that would be in there... hmm
07:51:32 <sbp> oh! we're so stupid
07:51:42 <sbp> the quotes around the value of the a/@href aren't required
07:51:42 <thelsdj> we are?
07:51:47 <thelsdj> haha
07:52:15 <sbp> what do you mean by the PHP code not being executed the "first time"?
07:53:00 <thelsdj> you get it to be PUT into the wiki page instead of the output of it or something
07:54:16 <thelsdj> oooh
07:54:22 <thelsdj> put it before the closing </b>
07:54:35 <thelsdj> put </b> after >?
07:54:39 <thelsdj> er ?>
07:54:54 <thelsdj> that way the initial linkback is saved
07:55:11 <sbp> ooh, good idea!
07:55:30 <sbp> done
07:56:04 <thelsdj> nice
07:56:27 <sbp> ha, great, it works. good suggestion
07:56:51 <sbp> 297 characters at the moment
07:57:22 <sbp> hmm. I think Jim mentioned a different event handler...
07:57:33 <sbp> splodge: grep Jim.*on
07:57:40 <splodge> I'm logging. I found 107 answers for 'Jim.*on' (showing 0...4)
07:57:41 <splodge> 0) 2004-02-11 07:57:33 <sbp> splodge: grep Jim.*on
07:57:42 <splodge> 1) 2004-02-11 07:57:22 <sbp> hmm. I think Jim mentioned a different event handler...
07:57:43 <splodge> 2) 2004-02-11 07:48:41 <sbp> though I'm sure Jim would've done that if he could...
07:57:44 <splodge> 3) 2004-02-10 *** irc.freenode.net has changed the topic to ÒHome of JibberJim: convincing people everywhere to make some spangly Y-Fronts and Rhinestone boots for their husbandsÓ
07:57:45 <splodge> 4) 2004-02-10 <JibberJim> the books related to http://www.thursdaynext.com/jurisfiction/ are good on that they're shipping UltraWord 9 IIRC which has such features in their stories
07:58:43 <thelsdj> splodge: grep Jim.*on.*=
07:58:45 <splodge> I'm logging. I found 12 answers for 'Jim.*on.*=' (showing 0...4)
07:58:47 <splodge> 0) 2004-02-11 07:58:43 <thelsdj> splodge: grep Jim.*on.*=
07:58:48 <splodge> 1) 2004-02-10 <JibberJim> <b><textarea onchange="x=new XMLHttpRequest();x.open('PUT',location);x.send(
07:58:49 <splodge> 2) 2004-02-10 <JibberJim> <textarea onchange="x=new XMLHttpRequest();x.open('PUT',location);x.send(
07:58:50 <splodge> 3) 2004-02-10 <JibberJim> <b><textarea onchange="x=new XMLHTTPRequest();x.open('PUT',location);x.send(
07:58:51 <splodge> 4) 2004-02-10 <JibberJim> <div><textarea onchange="x=new XMLHTTPRequest();x.open('PUT',location);
08:00:01 <sbp> aha: 2004-02-10.txt:<JibberJim> onchange of the textarea! onclick might be better, saves a character!
08:00:03 <sbp> ew: Disconnecting: Bad packet length 1099485581.
08:00:08 <sbp> ssh threw up
08:01:25 <thelsdj> lol
08:02:16 <sbp> eek
08:02:20 <sbp> onclick works... kinda
08:02:27 <thelsdj> hmm
08:02:28 <sbp> but at the cost of making it really really hard to use
08:02:33 <thelsdj> yea
08:02:38 <thelsdj> go back to onchange for now
08:02:42 <sbp> because if you click in the textarea it uploads--so you have to use tab
08:02:44 <sbp> yeah...
08:06:10 <sbp> the cool thing is that this is basically original code. most of the contest up until now has gone thus: "steal the current winner's code, reduce by three characters, proclaim yourself the new winner"
08:06:25 <sbp> the grep -l idea was stolen from one of the current winners
08:06:29 <sbp> and that's it
08:08:23 <sbp> hmm. the only way I can think of shortening it is if there's a better way to get vars from the environment
08:08:55 <sbp> `` seems to be basically an interpolated string, so I don't think that doing functions in there will work
08:09:01 <sbp> though one could assign outside and then use it...
08:09:04 <thelsdj> maybe php has a shorter version of $_SERVER[REQUEST_URI]
08:09:26 <thelsdj> like $REQUEST_URI or something
08:09:47 <sbp> hmm
08:09:52 <mediovia__> onBlur? click outside the textarea to PUT, maybe
08:10:03 <thelsdj> also
08:10:09 <thelsdj> can't we replace && with just a space?
08:11:10 <sbp> ooh, I think you're right about $REQUEST_URI
08:11:14 <sbp> and onBlur might work too
08:11:22 <sbp> && can't be replaced with a space, I don't think
08:11:27 <thelsdj> try it
08:11:50 <sbp> $ pqr=pqr echo $pqr
08:11:53 <sbp> [EOF]
08:11:56 <sbp> $ pqr=pqr&&echo $pqr
08:11:56 <sbp> pqr
08:11:58 <sbp> [EOF]
08:12:01 <thelsdj> hmm
08:13:00 <thelsdj> how is it then that you can do like GCC=gcc ./configure or shit like that i know i've seen it before when i compile programs
08:13:10 <sbp> great! it all works
08:13:20 <thelsdj> oh must have to do with the shell interpreter
08:13:20 <sbp> hmm. that's odd. I'm not sure
08:13:25 <sbp> perhaps the fact that echo is a builtin
08:13:29 <thelsdj> bash does something interesting
08:13:32 <sbp> * sbp nods
08:13:54 <sbp> onblur and $REQUEST_URI are working
08:14:11 <sbp> * sbp notes "Charles Goodier" for future acknowledgemental reference
08:14:33 <sbp> 286 chars
08:15:19 <mediovia__> this thing deserves its own javascript microbrowser
08:16:00 <sbp> hehheh
08:16:27 <sbp> deltab: is there no way to get bash to expand a string where the n is in ${n##*/}?
08:17:42 <sbp> except for bash -c
08:17:44 <sbp> that doesn't count
08:17:48 <sbp> :-)
08:18:30 <thelsdj> wait a minute
08:19:07 <thelsdj> why the heck arn't we doing echo \$(grep -l $REQUEST_URI *)?
08:22:10 <sbp> because you have to strip off everything before and including the last / of the REQUEST_URI
08:22:11 <sbp> but!
08:22:16 <sbp> I think I've saved a few more characters
08:22:22 <sbp> it seems that JavaScript supports $&
08:22:34 <thelsdj> hah
08:23:38 <sbp> yep, it does. whoo!
08:23:58 <sbp> heh, heh. just wait till I tell Jim all of this...
08:24:35 <thelsdj> wait how did that save any characters? wasn't it already $1 $1?
08:24:48 <sbp> because now we don't need parens around the whole match
08:24:53 <thelsdj> aaah
08:25:01 <sbp> /(([A-Z][a-z]+){2,})/g -> /([A-Z][a-z]+){2,}/g
08:25:04 <thelsdj> right
08:25:18 <sbp> I know you got it; just noting for the logs and anyone else watching
08:25:26 <thelsdj> :)
08:25:49 <sbp> * sbp tries to omit the "echo" from <?echo...?>, but it doesn't work...
08:25:53 <sbp> too much to ask, I suppose
08:26:40 <thelsdj> wait
08:26:44 <thelsdj> there is a shortcut for that i htink
08:26:53 <thelsdj> oh right but not with ``
08:27:03 <thelsdj> hmmm
08:27:26 <thelsdj> ooh
08:27:28 <thelsdj> take out echo
08:27:31 <thelsdj> and put a .
08:27:36 <sbp> ooh
08:27:36 <thelsdj> before and after the `
08:27:39 <thelsdj> 's
08:27:44 <thelsdj> thats the trick i believe
08:27:49 <thelsdj> like <?.$foo.?>
08:27:58 <thelsdj> prints out $foo
08:28:27 <sbp> Parse error: parse error in .../echotest.php on line 1
08:28:38 <thelsdj> doh
08:29:44 <thelsdj> hmm my test with just $foo didn't work either
08:29:51 <thelsdj> * thelsdj goes to look for the real syntax
08:30:34 <sbp> * sbp flicks through http://www.php.net/manual/en/
08:31:21 <thelsdj> ooooh
08:31:26 <thelsdj> <?=$foo?>
08:31:36 <sbp> oh!
08:31:39 <sbp> heh, I just found that too
08:31:53 <thelsdj> nice
08:32:01 <thelsdj> jesus:/var/www/bm# cat > echotest.php
08:32:01 <thelsdj> <?=`echo blah`?>
08:32:01 <thelsdj> jesus:/var/www/bm# php echotest.php
08:32:01 <thelsdj> blah
08:32:01 <thelsdj> jesus:/var/www/bm#
08:32:46 <thelsdj> 282 hehe
08:33:07 <sbp> yay!
08:33:17 <sbp> 281 in fact, since we'll chop of the trailing \n too
08:37:58 <sbp> oh, duh, this works too: <?=`echo \$(n=$REQUEST_URI;grep -l \${n##*/} *)`?></b>
08:38:02 <sbp> saves a single character
08:38:19 <thelsdj> ah nice
08:39:08 <sbp> basename works too, of course, but is one character longer...
08:40:49 <libby> *** libby (~libby@82-32-4-244.cable.ubr01.azte.blueyonder.co.uk) has joined #swhack
08:41:02 <thelsdj> what about PHP_SELF?
08:41:22 <sbp> oh! and I think I can get rid of all the $() and echo trickery since I was only employing that for exec
08:41:25 <sbp> PHP_SELF?
08:41:43 <thelsdj> $PHP_SELF instead of $REQUEST_URI
08:42:10 <sbp> ooh, I think so, yes
08:44:08 <sbp> yep. down to 268
08:44:12 <thelsdj> bahah
08:44:38 <thelsdj> the funny thing about this one is its actually readable compaired to the perl entries
08:46:46 <sbp> yes!
08:51:37 <deelan> *** deelan (~chatzilla@host131-95.pool8020.interbusiness.it) has joined #swhack
08:51:42 <thelsdj> the biggest string of text thats repeated in there is textarea but i can't find a way to save any from that
08:53:05 <sbp> heh: ' In PHP 4.2.0 and later, the default value for the PHP directive register_globals is off. This is a major change in PHP. Having register_globals off affects the set of predefined variables available in the global scope. For example, to get DOCUMENT_ROOT you'll use $_SERVER['DOCUMENT_ROOT']'
08:53:14 <sbp> yeah... I'm just reading through the docs to try to find a way
08:54:18 <deelan> i was told that globals are evil in PHP. i try to avoid PHP altogether
08:54:28 <thelsdj> so do we
08:54:37 <thelsdj> but its helping us make the worlds smallest wiki
08:54:58 <thelsdj> down to 269 characters
08:56:20 <deelan> hehe. i see. i've just read the JS PUT trick wiked
08:57:05 <sbp> the shame about the PHP manual is that it's all nice and neat and security conscious and so forth
08:57:38 <thelsdj> omg this might be a silly idea
08:57:56 <thelsdj> but what if we left off the </textarea> tag? or would that make the javascript not work?
08:58:06 <thelsdj> or would that just plain not work anyways
08:58:13 <sbp> so we have quite a few installation notes: In PHP 4.2.0 and later, register_globals must be turned on. `` must be allowed, so shell_exec must be allowed...
08:58:22 <sbp> I think that that would bork the JavaScript
08:58:31 <thelsdj> try it for the heck of it haha
08:59:11 <thelsdj> might bork the this.parentNode.innerHTML or something
08:59:23 <thelsdj> but it just might not
08:59:36 <sbp> heh. the "</b>" appears inside the textarea for me using Firefox 0.8
