2009-05-23 Swhack IRC Log

00:21:14 <nsh> phenny, ask sbp if there was .yi and when
00:21:14 <phenny> nsh: I'll pass that on when sbp is around.
00:21:51 <[bjoern]> answers are "no" and "very recently".
00:21:59 <nsh> cool
00:41:22 <deltab> clsn: Did you mean: antique thong
01:28:32 <deltab> * BuenGenio_ (n=Eugene@62.64.117.54) has joined #eurovision
01:28:32 <deltab> ‹BuenGenio_› wt... this channel is still going? is this a year-around support channel?? :P
01:28:32 <deltab> ‹slashtom› yes
01:28:32 <deltab> ‹slashtom› andrew lloyd webber has traumatised far too many in .uk
01:32:51 <nsh> plusplus
02:20:38 *** cre8radix (n=jamaerah@c-base/crew/cre8radix) has joined #swhack
02:20:40 <Monty2> welcome, cre8radix
02:22:17 <cre8radix> i'm not there yet, monty2
02:22:20 <Monty2> shampoo's Windows95 fondles porno viola >:)
02:22:47 <cre8radix> that doesn't get me off either, monty2
02:22:47 <Monty2> ]]]
02:23:07 <cre8radix> better, monty2, yet not the bomb
02:23:08 <Monty2> What does that suggest to you ?
02:23:54 <cre8radix> you're just not my type, i guess, or what would you make of it, monty2?
02:23:57 <Monty2> Prove to me that you don't attack bastards!
02:24:25 <cre8radix> well you still work... don't you, monty2?
02:24:27 <Monty2> people don't think "damn Arnia
02:24:50 <cre8radix> oh yes they do, monty2
02:24:50 <Monty2> You are sure.
02:25:01 <cre8radix> mosdef, monty2
02:25:01 <Monty2> Do you feel strongly about discussing such things ?
02:25:35 <cre8radix> you started that whole shit, monty2
02:25:36 <Monty2> screw you
02:25:46 <cre8radix> pussy!
02:50:44 *** jcgregorio (n=chatzill@adsl-072-148-043-048.sip.rmo.bellsouth.net) has joined #swhack
03:09:47 <nsh> .wik Rip van Winkle
03:09:48 <phenny> "'Rip Van Winkle' is a short story by the American author Washington Irving published in 1819, as well as the name of the story's fictional protagonist." - http://en.wikipedia.org/wiki/Rip_van_Winkle
03:17:44 <nsh> .ety laurel
03:17:46 <phenny> "c.1300, lorrer, from O.Fr. laurier, from L. laurus 'laurel tree,' probably related to Gk. daphne 'laurel' (for change of d- to l- see lachrymose), probably from a pre-IE Mediterranean language." - http://etymonline.com/?term=laurel
03:36:11 *** nsh- (n=nsh@host86-141-178-190.range86-141.btcentralplus.com) has joined #swhack
03:52:03 *** nsh has quit (Connection timed out)
04:50:52 *** cre8radix has quit ("your mama! >:)")
04:52:23 *** jcgregorio has quit ("ChatZilla 0.9.84 [Firefox 3.0.10/2009042523]")
05:45:35 *** lmorchard is now known as lmorchard|away
05:50:36 <[bjoern]> countdown
05:50:36 <Monty2> One large one and five small numbers: 75, 9, 4, 7, 3, 6. Your target is 934. You have thirty seconds, tick, tock, tick, tock...
05:50:57 <[bjoern]> .c (9)*(75+7*3)
05:50:58 <phenny> 9 * (75 + (7 * 3)) = 864
05:51:02 <[bjoern]> .c (9)*(75+7*4)
05:51:03 <phenny> 9 * (75 + (7 * 4)) = 927
05:51:06 <Monty2> Duh dum, duh dum, duh-da-da-dum! Your 30 seconds are up!
05:51:08 <Monty2> I found a solution in approximately 0 milliseconds. I found all 16 solutions in about 63 milliseconds.
05:51:10 <Monty2> ibot (75*3 + 7)*4 + 6
05:51:10 <phenny> (((75 * 3) + 7) * 4) + 6 = 934
05:51:25 <[bjoern]> countdown
05:51:25 <Monty2> One large one and five small numbers: 100, 7, 4, 1, 6, 9. Your target is 919. You have thirty seconds, tick, tock, tick, tock...
05:51:35 <[bjoern]> .c 9*(100+1)+7+4
05:51:36 <phenny> (9 * (100 + 1)) + 7 + 4 = 920
05:51:46 <[bjoern]> .c 9*(100+1)+6+4
05:51:47 <phenny> (9 * (100 + 1)) + 6 + 4 = 919
05:51:55 <Monty2> Duh dum, duh dum, duh-da-da-dum! Your 30 seconds are up!
05:51:56 <Monty2> I found a solution in approximately 0 milliseconds. I found all 10 solutions in about 125 milliseconds.
05:51:58 <Monty2> ibot (100 - 4 + 6)*9 + 1
05:51:59 <phenny> ((100 - 4 + 6) * 9) + 1 = 919
05:52:06 <[bjoern]> countdown
05:52:06 <Monty2> One large one and five small numbers: 50, 3, 3, 8, 10, 9. Your target is 756. You have thirty seconds, tick, tock, tick, tock...
05:52:25 <[bjoern]> .c (10+3)*(50+9)
05:52:25 <phenny> (10 + 3) * (50 + 9) = 767
05:52:36 <Monty2> Duh dum, duh dum, duh-da-da-dum! Your 30 seconds are up!
05:52:37 <[bjoern]> .c (10+3)*(50+9) - 8
05:52:37 <Monty2> I found a solution in approximately 0 milliseconds. I found all 32 solutions in about 63 milliseconds.
05:52:37 <phenny> ((10 + 3) * (50 + 9)) - 8 = 759
05:52:39 <Monty2> ibot (3*8 + 50 + 10)*9
05:52:39 <phenny> ((3 * 8) + 50 + 10) * 9 = 756
05:52:46 <[bjoern]> .c (10+3)*(50+9) - 8 - 3
05:52:47 <phenny> ((10 + 3) * (50 + 9)) - 8 - 3 = 756
05:52:50 <[bjoern]> countdown
05:52:50 <Monty2> One large one and five small numbers: 75, 2, 1, 1, 4, 8. Your target is 943. You have thirty seconds, tick, tock, tick, tock...
05:53:04 <[bjoern]> hmm
05:53:12 <[bjoern]> that does sound difficult Monty2
05:53:12 <Monty2> >:)
05:53:20 <Monty2> Duh dum, duh dum, duh-da-da-dum! Your 30 seconds are up!
05:53:22 <Monty2> Oh bugger, I don't think I can solve that one!
05:54:15 <realist> .wik Shakuntala Devi
05:54:15 <phenny> "Shakuntala Devi (Kannada: ಶಕುಂತಲಾ ದೇವಿ) is a calculating prodigy who was born on November 4, 1939[1]| in Bangalore, India." - http://en.wikipedia.org/wiki/Shakuntala_Devi
05:57:53 <[bjoern]> Did he bite you?
05:58:08 <realist> "In 1977 she extracted the 23rd root of a 201-digit number mentally."
05:59:36 <[bjoern]> somehow not doing that strikes me as the greater achievement.
05:59:57 <realist> "Alexis Lemaire merely twitched slightly as he worked out the 13th root of a random 200-digit number in his head in just 70.21 seconds"
06:00:18 <realist> -- http://www.medindia.net/News/Fastest-Human-Calculator-Too-Busy-To-Notice-Women-30592-1.htm
06:00:50 <[bjoern]> So long as they notice you...
06:01:05 <realist> So you dedicate your life to achieving what a computer can already do, only to never procreate
06:19:32 <[bjoern]> You don't want your children to break your records.
06:21:31 <[bjoern]> nice http://www.colbertnation.com/the-colbert-report-videos/181092/november-02-2005/c-span-coverage
06:40:43 <[bjoern]> .gs ergo is latin for *
06:40:44 <phenny> ergo is latin for *: therefore (8), " (6), ' (3)
06:40:54 <[bjoern]> .gs penis is latin for *
06:40:54 <phenny> penis is latin for *: tail (5)
06:41:09 <[bjoern]> .gs * is latin for idiot
06:41:10 <phenny> * is latin for idiot: " (3), which (2)
06:41:17 <[bjoern]> .gs * which is latin for idiot
06:41:18 <phenny> [bjoern]: (2)
06:49:21 *** lmorchard|away has quit (Remote closed the connection)
07:01:10 <[bjoern]> countdown
07:01:10 <Monty2> One large one and five small numbers: 25, 1, 1, 3, 6, 10. Your target is 157. You have thirty seconds, tick, tock, tick, tock...
07:01:22 <[bjoern]> .c 6*(25+1) +1
07:01:23 <phenny> (6 * (25 + 1)) + 1 = 157
07:01:33 <[bjoern]> too easy Monty2
07:01:37 <Monty2> In all honesty, weird corrosive milk-maid sleeps with bubbly talking games.
07:01:40 <Monty2> Duh dum, duh dum, duh-da-da-dum! Your 30 seconds are up!
07:01:41 <Monty2> I found a solution in approximately 0 milliseconds. I found all 804 solutions in about 63 milliseconds.
07:01:43 <Monty2> ibot 25*6 - 3 + 10
07:01:44 <phenny> (25 * 6) - 3 + 10 = 157
07:01:47 <[bjoern]> Oh I wish Monty2
07:01:47 <Monty2> You say oh you wish monty2 ?
07:01:57 <[bjoern]> Monty2: no.
07:01:59 <Monty2> Face it: fork-lift trucks is calibrated...?
07:02:05 <[bjoern]> countdown
07:02:06 <Monty2> One large one and five small numbers: 50, 2, 2, 6, 10, 3. Your target is 445. You have thirty seconds, tick, tock, tick, tock...
07:02:24 <[bjoern]> .c (6+3)*(50)-10/2
07:02:25 <phenny> ((6 + 3) * 50) - (10 / 2) = 445
07:02:36 <Monty2> Duh dum, duh dum, duh-da-da-dum! Your 30 seconds are up!
07:02:37 <Monty2> I found a solution in approximately 0 milliseconds. I found all 24 solutions in about 63 milliseconds.
07:02:39 <Monty2> ibot (50 - 6)*10 + 2 + 3
07:02:39 <phenny> ((50 - 6) * 10) + 2 + 3 = 445
07:02:42 <[bjoern]> countdown
07:02:42 <Monty2> One large one and five small numbers: 100, 10, 5, 7, 1, 7. Your target is 490. You have thirty seconds, tick, tock, tick, tock...
07:02:53 <[bjoern]> .c 5*(100) -10
07:02:53 <phenny> (5 * 100) - 10 = 490
07:03:00 <[bjoern]> that's way too easy Monty2
07:03:03 <Monty2> I reckon FooBot + legislation = canonical urine!
07:03:12 <Monty2> Duh dum, duh dum, duh-da-da-dum! Your 30 seconds are up!
07:03:14 <Monty2> I found a solution in approximately 0 milliseconds. I found all 1158 solutions in about 63 milliseconds.
07:03:15 <Monty2> ibot 10*7*7
07:03:16 <phenny> 10 * 7 * 7 = 490
07:03:21 <[bjoern]> nice solution
07:03:27 <[bjoern]> countdown
07:03:27 <Monty2> One large one and five small numbers: 100, 10, 7, 9, 9, 7. Your target is 969. You have thirty seconds, tick, tock, tick, tock...
07:03:52 <[bjoern]> .c 9*(100+7) + 7 - 9 + 10
07:03:53 <phenny> (9 * (100 + 7)) + 7 - 9 + 10 = 971
07:03:57 <Monty2> Duh dum, duh dum, duh-da-da-dum! Your 30 seconds are up!
07:03:58 <[bjoern]> .c 9*(100+7) + 7 - 10 + 9
07:03:59 <Monty2> I found a solution in approximately 16 milliseconds. I found all 48 solutions in about 63 milliseconds.
07:03:59 <phenny> (9 * (100 + 7)) + 7 - 10 + 9 = 969
07:04:00 <Monty2> ibot (100 + 7)*9 - 10 + 9 + 7
07:04:01 <phenny> ((100 + 7) * 9) - 10 + 9 + 7 = 969
07:04:31 <[bjoern]> I very much doubt the 1158 figure above Monty2
07:04:31 <Monty2> :(
07:13:00 *** [bjoern] has quit ("Quit")
07:37:38 *** procto has quit (kornbluth.freenode.net irc.freenode.net)
07:37:38 *** Morbus has quit (kornbluth.freenode.net irc.freenode.net)
07:37:38 *** tro has quit (kornbluth.freenode.net irc.freenode.net)
07:38:30 *** procto (n=boo@logarchy.org) has joined #swhack
07:38:30 *** Morbus (n=morbus@c-75-67-50-34.hsd1.nh.comcast.net) has joined #swhack
07:38:30 *** tro (i=trotek@k.thnx.org) has joined #swhack
07:38:30 <Monty2> howdy, Morbus
07:38:31 <Monty2> hi tro, how ya doing?
07:50:38 *** leobard (n=Miranda@91-66-122-97-dynip.superkabel.de) has joined #swhack
07:50:44 *** leobard has quit (Remote closed the connection)
08:50:23 <sbp> yo
08:50:24 <phenny> sbp: 00:21Z <nsh> ask sbp if there was .yi and when
09:15:50 *** Arnia_ (n=jgeldart@87-194-158-123.bethere.co.uk) has joined #swhack
09:22:18 <sbp> phenny: tell nsh http://hansard.millbanksystems.com/commons/1939/sep/02/germany-and-poland-italian-proposals - Germany invades Poland, Chamberlain fails to declare war. after the his speech, before Greenwood's reply and unrecorded in Hansard, Leo Amery (on Chamberlain's benches!) shouts across to Arthur Greenwood: Speak for England!
09:22:18 <phenny> sbp: I'll pass that on when nsh is around.
09:24:08 *** sbp is now known as sbp_
09:24:13 *** sbp_ is now known as sbp
09:24:34 *** phenny has quit (Remote closed the connection)
09:24:56 *** phenny (i=sbp@59.176.232.72.static.reverse.ltdomains.com) has joined #swhack
09:30:59 *** cre8radix (n=jamaerah@c-base/crew/cre8radix) has joined #swhack
10:00:29 <sbp> oh, they revealed who performed the leak!
10:00:55 <sbp> oh, no, just a link in the chain:
10:00:57 <sbp> [[[
10:00:57 <sbp> Ex-SAS officer John Wick, who passed the data on from an anonymous source, said he had "no regrets" about exposing the expenses system's "rotten core".
10:01:06 <sbp> ]]] — http://news.bbc.co.uk/1/hi/uk_politics/8064868.stm
10:02:22 <sbp> heh, they have video of that MacKay meeting
10:17:39 *** dpawson (n=dpawson@dpawson.gotadsl.co.uk) has joined #swhack
10:18:18 <sbp> dpawson! been putting the bus pass to any more use?
10:18:57 <dpawson> :-) Starting to look further afield now! Getting adventurous!
10:19:09 <dpawson> What's Brighton like for public transport?
10:19:18 <sbp> you'll get addicted. it'll be National Express coaches next
10:19:25 <sbp> then it'll be coach tours across the continent
10:19:30 <dpawson> I don't think they're included?
10:19:38 <sbp> before you know it, you'll be smoking dope with hippies in a charabanc in Kashmire
10:19:40 <sbp> *Kashmir
10:20:03 <sbp> probably not, but you know what addictions are like
10:20:08 <dpawson> Kashmire (sp?) is definately not on the list
10:20:17 <sbp> (yeah, *Kashmir)
10:20:27 <dpawson> I found it quite fun, but then it was novel
10:20:28 <sbp> I was goinge all olde Englishe
10:20:40 <dpawson> who, you!! Never
10:20:41 <sbp> yeah, I was thinking about that on the train the other day
10:20:50 <sbp> I was looking out of the window a little bored
10:20:59 <dpawson> Commuter zombies
10:21:02 <sbp> and I thought wait, if this were my first train ride, I'd be all "WOAH, LOOK AT THAT!"
10:21:22 <sbp> so I tried to be more excited. but it was pretty hard
10:21:26 <dpawson> You should have seen my two children on their first tube ride! It was great.
10:21:32 <sbp> hehe, I'll bet
10:21:56 <dpawson> Are you still RDF free?
10:21:59 <sbp> see, that's a good example: some people hate the tubes, but I travel on them infrequently enough to still be excited by them
10:22:08 <sbp> yes, but the assault is mounting from all sides
10:22:18 <sbp> I suppose you've seen this new stuff they're trying to add to HTML 5?
10:22:19 <dpawson> Like many addictions?
10:22:24 <sbp> heh, indeed
10:22:29 <dpawson> Yes. Gets boring.
10:22:51 <sbp> it's quite amazing that it keeps continuing
10:22:58 <sbp> like every year, someone comes up with a data-in-html solution
10:23:02 <dpawson> rdfa could be interesting... but I wouldn't bet on it taking off
10:23:04 <sbp> that completely ignores all prior art
10:23:25 <sbp> don't forget grddl, which is a REC already...
10:23:43 <sbp> and then grddl itself eschewed a good five or so proposals of approaches beforehand
10:24:00 <dpawson> I think the uptake path has been set for html now. Only good when n browsers implement it?
10:24:18 <sbp> going way back, Murray Altheim and I proposed a thing using <meta> and <link> back in lik 2001 or 2002, and that even had a decent DTD (he did the m12n DTDs)
10:24:36 <sbp> yeah, but that's the general problem with all this data lark
10:24:48 <sbp> the HTML WG (well, the WHAT WG) say they're paving the cowpaths
10:24:56 <dpawson> Like the phrase
10:24:57 <sbp> but I don't really see this as a cowpath being paved
10:25:10 <dpawson> Oops, sorry, I thought you said cow pats
10:25:16 <sbp> it's more like paving the cowpats... exactly
10:25:48 <sbp> note how Hixie led his data-in-html use cases list with "say you have an iguana collection..."
10:26:01 <dpawson> I can only see the IE8 devs scratching their heads and wondering what to implement.
10:26:19 <dpawson> Lots of 'what if' scenarios. Not many from millions of users though.
10:26:35 <sbp> yeah. that's the problem. it's a kind of special pleading for usage
10:26:48 <sbp> deriving from possibilities, rather than a democracy of actualities
10:27:44 <dpawson> Trying to think back. Wasn't it Mozilla and Microsoft that by implementing X, basically got it taken up in HTML v.3+
10:27:52 <dpawson> I.e. what people were really using?
10:27:59 <sbp> in all my technological deals, with all the sites and projects I've got going on, there are only a small handful of things where I'd really like to be able to add data to HTML. one of these is that I have a kind of site which describes a load of locations, and I'd like to geotag the locations so that I can automatically create a map out of it
10:28:20 <sbp> yeah, back in those days it was implement first and specify later
10:28:26 <dpawson> Isn't that what XML is for?
10:28:31 <sbp> <marquee> and <blink> being the infamous examples of implementation-first
10:28:33 <dpawson> Oops. sorry. take that back
10:28:44 <dpawson> Which were not (widely) taken up?
10:28:48 <sbp> we can't use XML now, the WHAT WG banned it...
10:28:56 <dpawson> Oh!
10:29:26 <dpawson> Well then. We'd all better stop that silly markup stuff <grins/>
10:29:26 * sbp tests <marquee> and <blink> in Safari
10:29:33 <sbp> <marquee> still works, <blink> doesn't
10:29:38 <dpawson> Good
10:30:49 <sbp> even if XML were used, there is the common question of course about conformance to a specification, and it doesn't help if the specification is not then quite light on extensibility options
10:31:24 <sbp> for example if you wanted to extend m12n yourself, you had to be a DTD expert, and I believe they said somewhere that the general use case was expected to be big corporations, not your average webpage designer/hacker
10:32:10 <sbp> now on the other hand you can do something like <metadata> in SVG and say "here's an area where you can cram in whatever, go nuts, do what you like", which of course has the inherent danger that people will shoehorn in a bunch of rubbish this way
10:32:39 <dpawson> Extensible? My assumption is that if the data source is good/constant, then the receiver can hack on it to get what they want?
10:33:22 <sbp> what do you mean?
10:33:27 <dpawson> Isn't that the same issue though? Contents of <meta/> could be 'extensible'?
10:34:20 <dpawson> If you find a data source(xml) that gives you something interesting, and it remains stable over n weeks
10:34:33 <dpawson> then it's worth filtering it to a form you like and can work with.
10:34:52 <sbp> sure, but if you're using HTML 5 you don't have that extensibility option
10:35:10 <dpawson> So stay with XML. That's what its for?
10:35:25 <sbp> you do if you're using XHTML 5, but I have a feeling HTML 5 will be more popular because the error recovery is more forgiving as I understand it
10:35:48 <dpawson> Popular what for? Plain web pages?
10:35:59 <sbp> well the scenario is that you already have the majority of your data in HTML, and you just want to shoehorn in a little bit of data, such as in one of the very few actual scenarios that I have for data-in-html
10:35:59 <dpawson> Or Seans 'applications'?
10:36:04 <sbp> both
10:36:07 *** chris2 (n=chris@p5B1680D7.dip0.t-ipconnect.de) has joined #swhack
10:36:16 <sbp> there will be more HTML 5 on the web than XHTML 5
10:36:22 <sbp> perhaps by an order of magnitude. I suspect
10:36:38 <sbp> (true of my site as much as of anyone else's)
10:36:39 <dpawson> That's what bugs me. Just shoehorn a bit more in... Is that the right way?
10:37:00 <dpawson> Point to it from the html5
10:37:09 <dpawson> Here's my sexy xml.
10:37:17 <dpawson> metadata or whatever
10:38:20 <sbp> wouldn't that be somewhat inconvenient? this is geotagging information, so are you saying I ought to have <location xmlns="http://example.org/whatever" about="http://mypage.com/" latlong="85.901385,-91.981734"/> in a file for each input? or that I should have a huge centralised catalogue of all the data?
10:38:35 <sbp> the problem is that when the number of resources to maintain goes up, you end up with a sychronisation problem
10:38:43 <dpawson> true.
10:38:59 <sbp> if when I add a node I have to update two files rather than one, that doubles the burden
10:38:59 <cre8radix> :) heya
10:38:59 <dpawson> pulling from an href isn't that hard though.
10:39:04 <sbp> yo cre8radix
10:39:36 <dpawson> What if you generate both from a single source? Auto sync
10:39:38 <sbp> right, but the general way of engineering on the web to make it work is to be exceptionally lazy: far lazier than one would ordinarily deem necessary
10:39:50 <dpawson> Yep.
10:39:53 *** cre8radix has quit ()
10:40:12 <sbp> if you consider HTML in the early days, it worked because people could be extremely lax about syntax. DanC tried to make it very strict and SGML conformant, but it worked because people could make hideous errors and it would still more or less work
10:40:32 <sbp> if you look at robots.txt, if robots.txt had used SGML or XML, can you imagine it would have been supported almost universally as it is now?
10:40:41 <dpawson> Yep, but I think you want to be 'accurate', rather than lazy. True?
10:41:03 <dpawson> sbp != a new html hacker.
10:41:18 <sbp> well I don't think it's accurate to want to maintain all your data in one place with the minimal amount of keystrokes. I'd definitely call that laziness... :-)
10:41:34 <sbp> sure, but laziness definitely helps the newcomer
10:41:43 <dpawson> I'm less bothered about 'one place' when I'm auto building from a script?
10:41:47 <sbp> it helps the experienced and the inexperienced alike, I think
10:42:09 <sbp> sure, but mostly when people add data to a site they're not auto building
10:42:12 <dpawson> How many newbies would want to add geotags (without a gui to help them)
10:42:22 <sbp> what I'm doing is publishing knowledge which hasn't yet been published
10:42:29 <sbp> so I'm taking some pains to prepare that stuff
10:42:45 <dpawson> That's why I want to separate you from an html hacker building his first site?
10:43:00 <dpawson> So be lazy and script it!
10:43:01 <sbp> yeah, I'm not sure how universal this case is. it might be a case of pleading from possibilities again like the HTML WG seem to be; and it almost certainly is
10:43:22 <sbp> but it's one of the few cases that I can at least personally identify with and try to get to grips with in terms of how it should be done
10:43:45 <dpawson> I guess my position is that geotags (and other esoterics) have no place in plain ole html, so keep it out?
10:44:01 <sbp> scripting something is not the lazy way to do something, because you have to maintain a script, which (if you're being lazy) will be underdocumented; so you're constantly looking up how you did it, what kind of schema you're using, etc.
10:44:24 <sbp> well, that's what I'm wondering
10:44:27 <dpawson> One doesn't necessarily follow from the other ;-)
10:44:29 <sbp> I'm not entirely convinced either way
10:44:57 <sbp> because on the one hand I see the virtue of keeping some strict core like good old fashioned HTML 4.01, saying that basically people can't add random stuff
10:45:06 <sbp> and on the other hand I see the virtue of something like <metadata> in SVG
10:45:06 <dpawson> Until one of the working groups offer something that looks workable, I'd say keep it out is the only sensible option?
10:45:13 <sbp> I wonder if at the moment HTML might be too strict
10:45:21 <sbp> but that on the other hand we don't want to make it too lax either
10:45:30 <sbp> and what, therefore, would be the adequate balance between the two poles?
10:46:05 <sbp> well, they're going to do whatever they want. they've already put in a bunch of really bizarre looking stuff into HTML 5, and i wouldn't be surprised if that turns out to be in the final REC and implemented in the browsers
10:46:34 <dpawson> I can't see them having an 'extensible' block of anything midway through html?
10:46:42 <sbp> it would be frustrating if we had the power now to affect the process and did not take that opportunity. I do tend to think it's somewhat futile, but already I'm thinking about what should have been done circa 2003
10:46:51 <dpawson> The bizarre stuff is what makes me think it won't be taken up?
10:47:04 <sbp> well I don't mean exactly what <metadata> gives you
10:47:13 <sbp> won't be taken up: hopefully!
10:47:19 <sbp> but they also have a lot of social impact
10:47:20 <dpawson> :-)
10:47:26 <sbp> they started out as this strange faction
10:47:31 <sbp> and now they have a W3C WG
10:47:39 <sbp> they're trendy and loud
10:47:45 <sbp> which is a powerful combination
10:48:11 <dpawson> Don't forget we have n years of browser wars, Moz and IE experience to battle against sillies
10:48:44 <sbp> that's part of the problem: the browser manufacturers seem well represented in the WGs
10:48:59 <dpawson> Doesn't mean they'll implement it all though.
10:49:03 <sbp> Moz and IE didn't battle against insanity
10:49:07 <sbp> they battled against one another
10:49:10 <dpawson> :-)
10:49:12 <sbp> IE added <marquee>
10:49:17 <sbp> Netscape added <blink>
10:49:32 <sbp> they didn't care at all about adding features that they thought would endear people to their browsers
10:49:47 <sbp> even though they were clearly bonkers
10:50:01 <dpawson> They aren't doing that now though... are they? Or does market share count as the same
10:50:39 <sbp> on not implementing it all: I think that things like the Acid Test show that browsers are becoming increasingly inclined to try to implement things according to entire specification, because compliance with the specification is getting to be trendy now
10:50:51 <sbp> yeah, I think browsers still basically have market share in mind
10:50:53 <dpawson> Which is a good thing.
10:51:04 <sbp> it's good only if the specifications are good
10:51:41 <sbp> this seems to be founded on the idea that specifications are inherently good because they're specifications, and of course one only has to look at XSD to know that's definitely not necessarily the case
10:51:43 <dpawson> Which brings it full circle to the WG
10:51:49 <sbp> right
10:52:20 <dpawson> Wonder if TimBL would veto it if seen as bad.
10:52:34 <dpawson> Who defines bad though.
10:52:47 <sbp> if it wasn't for Björn, he would have progressed the new XHTML m12n specifications
10:52:55 <sbp> despite a huge catalogue of errors that weren't addressed therein
10:53:17 <sbp> (they had to recind the PERs after he wrote a huge sarcastic email about the ignoral)
10:53:30 <sbp> (having already pursued all the regular professional channels)
10:53:56 <dpawson> Two WGs is one too many I guess.
10:54:23 *** libby_ (n=libby@94-169-110-24.cable.ubr22.aztw.blueyonder.co.uk) has joined #swhack
10:54:31 <dpawson> Hi libby_
10:55:41 <sbp> one HTML WG has always been one too many, traditionally
10:55:45 <sbp> two is presumably two too many
10:55:56 <dpawson> :-) That's a view!
10:56:19 <dpawson> Freeze at 4.x?
10:56:37 <sbp> well I can imagine there being a good HTML WG
10:56:46 <sbp> but I can't imagine how to ensure it
10:56:52 <dpawson> Imagine being the operative word?
10:56:55 *** libby_ is now known as libby
10:56:58 <sbp> like I can't imagine how to ensure a good government
10:56:58 <sbp> yeah
10:57:11 <libby> hey dpawson! (and sbp!)
10:57:18 <sbp> and it's not like the HTML WG is democratically elected
10:57:57 <sbp> yo libby!
10:58:18 <dpawson> Have you seen burningbirds comments on her attempts to sway the WG?
10:58:23 <sbp> nope?
10:58:59 <dpawson> http://tr.im/mbA6  etc
11:00:10 <sbp> these apply in the more important question of governance of the people, but I don't see why they don't also apply to the HTML WG for example as a microcosm group which is affecting millions' of people's interests: http://www.mhs.ox.ac.uk/blackboard/benn-l.htm
11:00:12 <dpawson> http://tr.im/mbAg on rdfa + microdata. She's trying to get stuck in.
11:01:11 <sbp> yeah, forking doesn't really sound reasonable
11:01:12 <dpawson> Very true Sean.
11:01:37 <sbp> "go form your own government, and see if you can attract more popular support! (but you can't vote)"
11:01:40 <dpawson> Some very loud voices on the WGs though.
11:02:20 <sbp> I suppose the idea is that one may always revolt
11:02:22 <dpawson> I thought WGs were supposed to justify their content through requirements though? Or doesn't that count with html, due history?
11:02:48 <dpawson> You might think they're revolting, I couldn't possibly comment!!!
11:02:51 <sbp> but revolution doesn't fix small problems, only large problems — and then only problems which are seen as large vox populi
11:02:54 <sbp> hehe
11:03:12 <sbp> well, they form the requirements themselves
11:03:23 <dpawson> exactly. Without review
11:03:31 <sbp> they don't have to prove the requirements are requirements
11:03:40 <dpawson> Then iterate as the mood takes them
11:03:44 <sbp> and they're not held accountable if they were wrong
11:03:46 <dpawson> process issue?
11:03:48 <sbp> aye
11:03:52 <sbp> of course a process issue
11:03:57 <dpawson> tim?
11:04:25 <sbp> Tim makes mistakes, as we saw with the PERs
11:04:31 <dpawson> Public review of requirements, moderated by Tim or .... another group?
11:04:40 <dpawson> No one's perfect
11:04:54 <dpawson> He has a good batting average though
11:05:06 <sbp> well one of the things that the W3C suggested about the PERs controversy is that they have a public group to discuss progressions of specifications
11:05:17 <sbp> yeah, but I think what we see is more problems than TimBL can cope with
11:05:28 <sbp> his ideas on various Semantic Web ideas for example were simply unheeded
11:05:37 <dpawson> I guess that could come down to who shouts the loudest?
11:05:46 <sbp> and he was powerless to direct things in the way he saw fit, because of his institutional commitments
11:05:56 <dpawson> He's not quite the benevolent dictator is he.
11:06:12 <sbp> nope, not the same case as LWall or GvR
11:06:13 <dpawson> His load is simply too big these days.
11:06:27 <sbp> in a sense he did do much worse than those guys
11:06:29 <dpawson> Use TAG?
11:06:32 <sbp> he also had a bigger invention, though
11:06:45 <dpawson> == more responsibility
11:07:43 <sbp> well he seemed to lose capital early on, I mean in the sense of being perceived to be the "head of the project". I'm not sure why that happened... perhaps he was simply too nice about it, though LWall seems to be nice about it with regards to perl and that didn't run out of his overall control
11:07:57 <dpawson> SWMBO calls. Lunch is ready... when I set the table. TTFN Sean
11:08:02 <sbp> I suspect the problem is roughly that he lost control of the implementations
11:08:04 <sbp> c'ya!
11:08:21 <sbp> with perl and python, LWall and GvR wrote the standard implementations
11:08:37 <sbp> but with the WWW, the standard implementation was written by other people
11:08:59 <sbp> so control was decentralised in that sense early on
11:09:42 *** nsh (n=nsh@host86-141-178-190.range86-141.btcentralplus.com) has joined #swhack
11:09:42 <Monty2> howdy, nsh
11:14:35 *** nsh- has quit (Success)
11:15:52 <sbp> I think the problem is that Sam Ruby, with his forking comment, is using an unrealistic assessment of what HTML is: a very large technology, a kind of supertechnology. forking may be appropriate for GNU sized projects or even smaller, but who could fork HTML?
11:16:41 <sbp> similarly, the analogy of government is not an entirely useful one. a government is something that bears on human rights and so forth, and though HTML may be a supertechnology and have a great impact on people, it is still a technology and the people who make it are not in the realm of governing
11:26:43 *** nsh- (n=nsh@host86-141-178-190.range86-141.btcentralplus.com) has joined #swhack
11:49:26 <sbp> .title http://www.jenitennison.com/blog/node/102
11:49:28 <phenny> sbp: Evolving Standards | Jeni's Musings
11:51:54 <sbp> http://wiki.whatwg.org/wiki/FAQ#HTML5_should_support_a_way_for_anyone_to_invent_new_elements.21
11:56:11 *** nsh_ (n=nsh@host86-141-178-190.range86-141.btcentralplus.com) has joined #swhack
11:56:41 *** nsh has quit (Nick collision from services.)
11:56:43 *** nsh- has quit (Nick collision from services.)
11:56:46 *** nsh_ is now known as nsh
12:06:12 *** nsh- (n=nsh@host86-141-178-190.range86-141.btcentralplus.com) has joined #swhack
12:06:33 *** nsh_ (n=nsh@host86-141-178-190.range86-141.btcentralplus.com) has joined #swhack
12:07:12 *** nsh has quit (Nick collision from services.)
12:07:20 *** nsh_ is now known as nsh
12:34:38 <nsh> whurp, swhackpals?
12:34:38 <phenny> nsh: 09:22Z <sbp> tell nsh http://hansard.millbanksystems.com/commons/1939/sep/02/germany-and-poland-italian-proposals - Germany invades Poland, Chamberlain fails to declare war. after his speech, before Greenwood's reply and unrecorded in Hansard, Leo Amery (on Chamberlain's benches!) shouts across to Arthur Greenwood: Speak for England!
12:35:12 <nsh> cool
12:42:40 <nsh> sbp, is you tehbouts?
12:43:29 <sbp> nsh: yo!
12:43:33 <nsh> Greenwood was pretty prophetic mang
12:43:49 <sbp> see also the next day's Hansard
12:43:51 <nsh> why haven't i heard of this frood
12:43:55 <nsh> silly noneducation
12:44:23 <sbp> the declaration of war is especially worth reading
12:44:28 <sbp> it's very English and reserved
12:44:38 <sbp> "we will protect Poland's interests as promised", sort of thing
12:44:42 <sbp> not "WAR ON GERMANY! GRR!"
12:46:10 <nsh> nice
12:46:58 <nsh> we should do a national hansard tour, sbp
12:47:02 <nsh> apply for some funding
12:47:05 <nsh> and recreate great sittings
12:47:17 <sbp> oh man, that'd be great
12:47:23 <sbp> do it all on Roombas
12:47:25 <nsh> as part of the bbc's "give a fuck about democracy" season
12:47:28 <nsh> word
12:47:29 <sbp> hehe
12:48:03 <nsh> looking at http://www.georgehart.com/virtual-polyhedra/kepler.html
12:48:52 <sbp> “In, Harmonice Mundi, Kepler also defined the class of convex uniform polyhedra which we call the Archimedean solids, unfamiliar with the fact that Archimedes had defined the class already.”
12:49:06 <sbp> (apart from the coolness of the quote, the unnecessary use of commas is funny)
12:50:13 <nsh> i still tend to use commas in this way
12:50:25 <nsh> because i tend to write as transcript of speech
12:50:38 <nsh> rather than dactyloprose
12:50:54 <nsh> Fuller was the same
12:51:09 <nsh> i would write most things in one sentence if the world would let me get away with it
12:51:18 <nsh> with the occasional semicolon for affect [sic]
12:54:36 <sbp> ah; yes; well; me too; to an extent; but; you know
12:54:49 <sbp> I also like the old convention of colon chains
12:55:19 <sbp> so you demonstrate something thus: and then you demonstrate another thing: and so lead to your conclusion: which may have a kind of twist as follows: YEAH MANG!
12:55:25 <nsh> yeah, i'm partial to them too
12:55:27 <nsh> but
12:55:38 <nsh> i tend to use them with preceding spaces in this usage
12:55:51 <sbp> preceding spaces?
12:55:55 <sbp> imitating old type?
12:55:58 <nsh> so i would partition my sentence : thusly, to provide an example : for the logical delineation of ideas
12:56:09 <nsh> which is strange, because usually i hate spaces before punctuation
12:56:12 <nsh> but with the colon it's ok
12:56:32 <nsh> generally only in poetry though
12:57:26 <nsh> i think it was laurence of arabia who, in the dedication to the seven pillars thanked someone (who] for contributing all the semi-colons
12:57:29 * nsh checs
12:57:37 <sbp> hmm. looks like one of those a : b :: p : q things
12:57:47 <sbp> which I tend to associate with American tests for some reason
12:58:13 <nsh> ah: The book so written passed in 1921 into proof; where it was fortunate in
12:58:13 <nsh> the friends who criticized it. Particularly it owes its thanks to Mr. and
12:58:13 <nsh> Mrs. Bernard Shaw for countless suggestions of great value and diversity:
12:58:13 <nsh> and for all the present semicolons.
12:58:21 <sbp> I should hire someone to check Americanisms
12:58:23 <sbp> ha
12:58:34 <nsh> yeah, makes me smile
12:58:47 <sbp> you know about Shaw's shed, I presume?
12:58:55 <nsh> nuppers
12:59:08 <sbp> well Shaw had a shed, in which he used to do much of his work
12:59:15 <sbp> out in the back of his garden
12:59:24 <nsh> oh, nice
12:59:27 <sbp> and he named the shed London
12:59:29 <nsh> haha
12:59:33 <sbp> so that when someone came to call on him...
12:59:41 <sbp> his wife could answer the door
12:59:43 <nsh> looking at http://scriptuality.blogspot.com/2009/05/shaws-shed.html
12:59:45 <nsh> lolol
12:59:47 <sbp> and say: sorry, Mr. Shaw is away in London
12:59:48 <nsh> "He's in london"
12:59:49 <nsh> haha, nice
12:59:51 <sbp> yeah
12:59:51 <sbp> hehe
12:59:56 <nsh> cleverness
13:00:29 <sbp> a modern writer would of course call their shed Milton Keynes
13:00:37 <nsh> hehe
13:01:04 <nsh> what's that double-barreled place name..
13:01:09 <nsh> norton-subcourse, i want to say
13:01:10 <nsh> but it's not
13:01:18 <nsh> something like that though
13:01:24 *** Arnia (n=jgeldart@87-194-158-123.bethere.co.uk) has joined #swhack
13:01:28 <nsh> yo arnia
13:01:50 <nsh> wait, maybe it was then
13:01:53 <nsh> but there
13:01:59 <nsh> s probably an even cooler name
13:02:10 <nsh> that norton-subcourse is eclipsing in my brain
13:02:27 <sbp> there's a place called Windy-gates, in Fife
13:02:52 <nsh> we live near a "Six-mile-bottom"
13:03:00 <nsh> .o maps six-mile-bottom
13:03:01 <phenny> nsh: Sorry, no such service. See http://code.google.com/p/phenny-ws/wiki/ServiceDefinitions
13:03:13 <nsh> phenny, tell nslater .o maps please
13:03:13 <phenny> nsh: I'll pass that on when nslater is around.
13:03:30 <nsh> oh wait, i was asking for a scripting project because i was planning to learn perl6
13:03:32 <nsh> maybe i should do it
13:03:40 <nsh> hmm, seems a bit productive though
13:03:56 <nsh> would be cool on the other hand
13:04:00 * nsh takes a shit on it, br
13:04:01 <nsh> b
13:04:09 <sbp> there's a Banchory-Devenick in Aberdeenshire
13:05:53 <nsh> nice
13:06:14 <sbp> Burgh-clere in Hampshire
13:06:44 <sbp> Congres-bury, North Somerset
13:06:55 <sbp> Countes-thorpe, Leicestershire
13:07:12 <sbp> Englesea-brook, Cheshire
13:07:36 <nsh> mmm
13:07:39 <sbp> Havering-atte-Bower? awesome
13:07:43 <sbp> .wik Havering-atte-Bower
13:07:44 <phenny> "Havering-atte-Bower is a village and outlying settlement of the London Borough of Havering,[1]| located 15 miles (24 km) north east of Charing Cross and close to the Greater London boundary." - http://en.wikipedia.org/wiki/Havering-atte-Bower
13:07:46 <nsh> plusplus
13:08:10 <sbp> Hole-in-the-Wall in Hereford and Worcester
13:08:16 <nsh> haha
13:08:21 <nsh> must be a good place to earn money
13:08:26 <sbp> hehe
13:08:29 <sbp> Hook-a-Gate, Shropshire
13:08:37 <nsh> i hope someone gets directed there by an errant satnav someday
13:08:43 <nsh> (hole-in-the-wall, that is]
13:08:57 <nsh> when looking for an ATM
13:09:11 <sbp> Praze-an-Beeble, Cornwall
13:09:29 <sbp> a Summer-house in Darlington
13:09:43 <sbp> Tavern-spite, Pembrokeshire
13:10:06 <nsh> i like tarvern-spite
13:10:10 <nsh> sp
13:10:28 <nsh> btw, have you read The Meaning of Liff?
13:10:34 * nsh still hasn't gotten around to it
13:11:45 <sbp> yeah
13:12:05 <nsh> isgút?
13:12:14 <sbp> it's okay, obviously a bit of a novelty book
13:12:18 <nsh> aye
13:12:20 <sbp> the publishing equivalent of a funny webpage
13:12:25 * nsh tehnod
13:12:48 <nsh> btw sbp, what is the degree of your propensity to earn the bank-of-england envifying rate of 0.02 percent interest again?
13:13:09 <sbp> en Français, svp?
13:13:21 <nsh> if this doesn't inflate significantly (ie, at all] i'm disposed to get it for finngirl's birthday http://cgi.ebay.co.uk/ws/eBayISAPI.dll?ViewItem&ssPageName=STRK:MEWAX:IT&item=130307739100
13:13:53 <nsh> but ze figures of entallyment do not commensurate in bankland :-/
13:14:07 <sbp> oh, sure, I can do that again for a superpenny hidden amongst the remuneration
13:14:12 <nsh> cool
13:14:20 <sbp> did you check that extra account that you forgot you had?
13:14:38 <nsh> yeah, it's all razored to middlemars now
13:14:43 <sbp> heh, heh
13:14:55 <sbp> as Rumsfeld might say, try to make all your unknown unknowns into known unknowns
13:15:03 <nsh> hopefuly getting paid for the phone sales work i did a few weeks ago
13:15:15 <nsh> as i hassled mr slick about it
13:15:22 <sbp> what were you selling? gilt bonds?
13:15:26 <nsh> haha
13:15:27 <sbp> I can imagine you selling gilt bonds
13:15:30 <nsh> totalz
13:15:37 *** Arnia_ has quit (Read error: 110 (Connection timed out))
13:15:43 <nsh> nah, it was convincing ex customers of a scottish hotel to go back
13:15:57 <nsh> thought it'd be easy, but it turns out the place was the basis for faulty towers
13:16:01 <sbp> haha
13:16:34 <nsh> and i actually cringed a few times (also, one shudder] at how much of a marketer voice the guy i was working for had
13:16:40 <sbp> I'd quite like to have stayed at Fawlty Towers really
13:16:43 <nsh> made me feel oily and wrong inside
13:16:44 <nsh> me too
13:16:46 <sbp> wouldn't have eaten the food though
13:16:58 <nsh> i'd have taken a cow to precud it
13:17:01 <sbp> nor stood underneath the moose
13:17:10 <nsh> man, when i'm rich, i will only eat via cow stomaches
13:17:17 <sbp> hehe
13:17:28 <nsh> SOLID PLATINUM COWS
13:17:37 <sbp> with diamond encrusted udders
13:17:57 <nsh> .gs solid platinum *
13:17:58 <phenny> solid platinum *: iphone (9), in (6), rope (5), bathed (4), band (4), tanzanite (3)
13:17:59 <nsh> totlalz
13:18:06 <nsh> .wik tanzanite
13:18:07 <phenny> "Tanzanite is the blue/purple variety of the mineral zoisite which was discovered in the Meralani (Merelani) Hills of Northern Tanzania in 1967, near the city of Arusha." - http://en.wikipedia.org/wiki/Tanzanite
13:18:12 <nsh> ok
13:18:20 <nsh> .wik zoisite
13:18:21 <phenny> "Zoisite is a calcium aluminium hydroxy sorosilicate belonging to the epidote group of minerals." - http://en.wikipedia.org/wiki/Zoisite
13:18:55 *** nsh changed the topic to: "Swhack: a calcium aluminium hydroxy sorosilicate belonging to the epidote group of minerals"
13:19:03 <sbp> Calaluhydroxysorosilicate
13:19:11 <nsh> that's us
13:19:21 <sbp> .wik Epidote
13:19:21 <phenny> "Epidote is a calcium aluminium iron sorosilicate mineral, Ca2Al2(Fe|3+|;Al)(SiO4)(Si2O7)O(OH), crystallizing in the monoclinic system." - http://en.wikipedia.org/wiki/Epidote
13:19:32 <sbp> we crystalise in the monoclinic system?
13:19:51 * nsh considers apprpopriating "epidote" as a pseud-detecting device
13:20:02 <nsh> slip it in some sentences, and if they pretend to know what it means
13:20:10 <nsh> they get teh derisionlulz
13:20:18 <nsh> sbp, Shhhh loggy....
13:20:20 <sbp> mispronounce it for extra luls
13:20:24 <nsh> word
13:20:36 <nsh> epidoté
13:20:42 <sbp> ARRIVA!
13:20:48 <nsh> (emphasis on the second syllable, of course]
13:20:53 <nsh> brbs
13:21:47 <sbp> so, <meta> can currently go anywheres in flow content in HTML 5
13:21:55 <sbp> as long as it has an itemprop attribute
13:23:26 <sbp> so you can be all:
13:23:26 <sbp> <div item>
13:23:26 <sbp> <meta itemprop="shizzle" content="dizzle">
13:23:26 <sbp> <p>Shizzle? Dizzle!</p>
13:23:26 <sbp> </div>
13:25:13 *** Morbus has quit ("http://www.disobey.com/")
13:26:14 <sbp> so they require registration for meta/@name
13:26:24 <sbp> but meta/@itemprop is free to use any value
13:28:12 <sbp> [[[
13:28:13 <sbp> Has anybody polled Dave Raggett, Arnaud Le Hors, or Ian Jacobs for their
13:28:13 <sbp> opinion on RDFa for HTML4? Isn't the spec for RDFa in XHTML a separate
13:28:13 <sbp> document from the XHTML specs?
13:28:13 <sbp> If it turns out that the the spec text is written, and that the
13:28:13 <sbp> consensus is that right place for such is in the HTML5 spec itself, and
13:28:17 <sbp> it turns out that Ian doesn't respect such consensus, then I am
13:28:19 <sbp> committed to resolve that problem.
13:28:22 <sbp> ]]] - Sam Ruby, http://lists.w3.org/Archives/Public/public-html/2009May/0243
13:38:23 <nsh> this arcana is nonlucent to me
13:39:05 <nsh> (just been teaching my ten month old nephew to pickpocket]
13:39:18 <sbp> yeah, you've got to pick a pocket or two
13:39:25 <sbp> oh, you've got to pick a pocket or two
13:40:42 <nsh> verily the truth ye sooth
13:41:01 <sbp> it's just funny to see them apparently accidentally design TMTOWTDI
13:41:09 <sbp> when they were probably intended TOOOWTI
13:41:16 <sbp> *intending
13:41:30 <sbp> because the above is basically no different to:
13:41:39 <sbp> <div item data-schizzle="dizzle">
13:41:45 <sbp> <p>Shizzle? Dizzle!</p>
13:41:46 <sbp> </div>
13:41:54 <nsh> .wik TMTOWTDI
13:41:54 <phenny> "There's more than one way to do it (TMTOWTDI or TIMTOWTDI, pronounced 'Tim Toady') is a Perl motto." - http://en.wikipedia.org/wiki/TMTOWTDI
13:42:04 <nsh> memoryfail
13:42:09 <nsh> yeah
13:42:14 <nsh> it's probably inevitable though
13:42:15 <Monty2> 9
13:42:21 <nsh> quite, monty
13:42:26 <Monty2> see why that it's up from all</LimitExcept> for almost half hour working groups offer some data lark
13:42:31 <sbp> well they seemed to shoehorn this new feature in pretty quickly
13:42:36 <sbp> without really thinking it all through
13:42:47 <sbp> probably just to provoke discussion though
13:42:55 <sbp> I mean, it's not like it's at Last Call yet
13:43:08 <nsh> Internet standards group in hasty revision shocker...
13:43:13 <sbp> heh, heh
13:43:57 <nsh> so, what does the above mean, sbp?
13:44:04 <nsh> about registrations and meta tags
13:44:14 <nsh> it might be diverting to understand
13:46:27 <sbp> that's one of the funny controversies
13:46:46 <sbp> they require people wanting to register new <meta name="blahblah" ...> values on a wiki on their website
13:47:03 <sbp> er, to do so on a wiki on...
13:47:09 <nsh> aren't meta tags just for search engines and such?
13:47:23 <sbp> yeah, they're pretty universally ignored and unimportant
13:47:29 <nsh> right
13:47:44 <sbp> though I tend to use "created" for example to show when I made a document
13:47:50 <nsh> aye
13:47:53 <sbp> now in their scheme, I'd have to push that through on their wiki
13:48:09 <sbp> which they wouldn't allow, because I use non-ISO standad date formats
13:48:16 <sbp> like today I might add:
13:48:20 <nsh> so it wouldn't validate if you used a meta name that doesn't accord to the wiki specs?
13:48:31 <sbp> <meta name="created" content="23rd May 2009">
13:48:36 <sbp> right
13:48:44 <nsh> but if you use an infoprop it would
13:48:49 <sbp> and anybody wanting to validate has to look on the wiki
13:48:49 <nsh> so it's pointless?
13:48:49 <sbp> right
13:48:51 <nsh> ok
13:48:52 <sbp> right
13:48:54 <nsh> yay i grok
13:49:08 <sbp> but with itemprop you have to set an item somewhere
13:49:22 <nsh> .wik itemprop
13:49:23 <phenny> "Authenticity refers to the truthfulness of origins, attributions, commitments, sincerity, devotion, and intentions." - http://en.wikipedia.org/wiki/Authenticity
13:49:27 <nsh> ok
13:49:43 <sbp> so it's not a particularly clean design
13:49:58 <sbp> like, you can't do this:
13:50:00 <sbp> <head>
13:50:05 <sbp> <meta name="created" content="23rd May 2009">
13:50:05 <nsh> HTML in "not elegant" shocker... ;-]
13:50:07 <sbp> </head>
13:50:10 <sbp> heh, heh
13:50:13 <nsh> [sry]
13:50:24 <sbp> well I mean not clean even above and beyond how usually unclean it is
13:50:30 * nsh nods
13:50:30 <sbp> like, you can't do this obvious simple thing here
13:50:44 <sbp> but again, they might resolve it, dunno. don't care too much
13:50:49 <sbp> you can already type the bytes
13:50:58 <sbp> as long as browsers don't start blowing up about it, it's cool
13:51:02 <nsh> yeah
13:51:12 <nsh> also, why not just have "nonauthenticated section" in webpages
13:51:16 <nsh> sorry
13:51:18 <nsh> nonvalidated
13:51:27 <nsh> so you can have extra-standards portions
13:51:36 <nsh> which browsers can be all lesseifaire about
13:51:43 <nsh> probably an invitation to crap
13:52:45 <nsh> [god i hate french]
13:53:21 <nsh> laissez-faire
13:55:50 <sbp> they used to kinda have that
13:55:53 <sbp> it was called <XMP>
13:56:07 <sbp> don't think it made it as far as HTML 2
13:56:12 *** nsh- has quit ()
13:56:21 * nsh nods
13:56:58 <sbp> oh, XMP did make it in, but not in the old SGML-breaking way of course
13:57:18 <sbp> HTML was mostly created at random, of course
13:57:28 <sbp> it was just some old documentation language they had at CERN
13:57:31 <sbp> and TimBL added <a>
13:57:41 *** Morbus (n=morbus@c-75-67-50-34.hsd1.nh.comcast.net) has joined #swhack
13:58:33 <sbp> woah, there's a spec!
13:58:39 <sbp> someone put a spec of it from 1986 on the web!
13:59:25 <sbp> http://cdsweb.cern.ch/record/997909/files/cer-002659963.pdf
14:02:53 <nsh> nice
14:03:15 <sbp> this is a great read
14:03:37 * nsh doesn't dare open acrobat
14:03:43 <nsh> but will read when in nonfail os
14:03:51 <sbp> ah! XMP was from this!
14:05:46 <nsh> .wik XMP
14:05:46 <phenny> "Extensible Metadata Platform, Adobe's standard for processing and embedding metadata in various file formats" - http://en.wikipedia.org/wiki/XMP
14:05:52 <nsh> oh i see
14:06:04 <sbp> ooh, SGMLguid was just one of the types
14:06:35 <nsh> THE SECRETS OF MARKUP ARE REVEAL'T TO OUR ENTREPID EPIDOTÉ
14:07:03 <nsh> WHAT NEXT SHALL BEFALL THE SWHACKER?
14:07:53 <sbp> funny, there was an <H0>
14:08:03 <sbp> as well as <H1>, <H2>, etc.
14:11:12 * nsh lulz at http://batteriesfeelincluded.blogspot.com/2009/05/309.html
14:11:19 <nsh> .title
14:11:20 <phenny> nsh: Batteries Feel Included: 309
14:11:29 <nsh> mehtitle
14:14:25 <sbp> phenny: tell nslater * nsh lulz at http://batteriesfeelincluded.blogspot.com/2009/05/309.html
14:14:25 <phenny> sbp: I'll pass that on when nslater is around.
14:24:42 *** Arnia has quit (Read error: 110 (Connection timed out))
14:29:43 <sbp> oh funny, you can do this:
14:29:44 <sbp> <p item>
14:29:44 <sbp> <a itemprop="about" href=""></a>
14:29:44 <sbp> <meta itemprop="key" content="value">
14:29:44 <sbp> </p>
14:30:04 <sbp> (or <div>)
14:30:26 <sbp> (in fact, absolutely div. can't do it in p)
14:30:38 <sbp> <div item>
14:30:38 <sbp> <a itemprop="about" href=""></a>
14:30:38 <sbp> <meta itemprop="key" content="value">
14:30:38 <sbp> </div>
14:30:43 *** Arnia (n=jgeldart@0-16-cb-bd-57-e.it.wlan.dur.ac.uk) has joined #swhack
14:32:36 <Arnia> Wow, this is laziness: http://www.stocklayouts.com/
14:32:56 <Arnia> So now you can combine stock images into a stock layout, with stock textures and a stock colour theme
14:33:06 <nsh> that's efficient reuse
14:33:09 <nsh> stupid to pay for, of course
14:33:17 <nsh> but the idea is sound
14:34:34 <sbp> hmm:
14:34:35 <sbp> [[[
14:34:36 <Arnia> Hm
14:34:36 <sbp> For each item type type of item, run the following substeps:
14:34:36 <sbp> If type is not an absolute URL, then let type be the result of concatenating the string "http://www.w3.org/1999/xhtml/custom#" with type.
14:34:36 <sbp> Generate the following triple:
14:34:36 <sbp> subject subject predicate http://www.w3.org/1999/02/22-rdf-syntax-ns#type object type
14:34:47 <sbp> ...
14:34:47 <sbp> For each name name in element's property names, run the following substeps:
14:34:48 <sbp> If name is equal to the string "about", skip this name.
14:34:48 <sbp> Otherwise, if name is not an absolute URL, then let name be the result of concatenating the string "http://www.w3.org/1999/xhtml/custom#" with name.
14:34:50 <sbp> Generate the following triple:
14:34:51 <sbp> subject subject predicate name object value
14:34:53 <sbp> ]]] - HTML 5
14:34:59 <sbp> so they're putting the item types into the same namespace as the itemprop values?
14:41:07 <sbp> ah! you can use link inline
14:41:16 <sbp> chuckle, even better then
14:43:45 <sbp> so you can do this:
14:43:46 <sbp> <head item>
14:43:46 <sbp> <link itemprop="about" href="">
14:43:46 <sbp> <meta itemprop="key" content="value">
14:43:46 <sbp> </head>
14:44:00 <nsh> ah well, missed out on that eeepc for finngirl
14:44:08 <sbp> nsh: d'oh
14:44:10 <nsh> forgot my numbers not working makes fast bidding difficult
14:44:14 <nsh> had to use on-screen keyboard
14:44:17 <nsh> did not win.
14:44:20 <sbp> heh, ouch
14:44:23 <nsh> yeah, no matter
14:44:36 <nsh> will likely find one cheaper when i can actually afford it anyway
14:44:40 <nsh> so positive economy
14:50:35 *** libby has quit (Read error: 113 (No route to host))
14:54:40 *** dpawson_ (n=dpawson@dpawson.gotadsl.co.uk) has joined #swhack
15:10:45 *** dpawson has quit (Read error: 113 (No route to host))
15:33:54 *** dpawson_ has quit ("Leaving")
15:35:54 *** kpreid has quit ()
15:36:16 *** kpreid (n=kpreid@cpe-67-249-58-190.twcny.res.rr.com) has joined #swhack
16:13:24 <sbp> chuckle. I love old textbooks:
16:13:25 <sbp> --
16:13:25 <sbp> Pray, Sir, why do you chufe to ufe fuch a long hard Word as
16:13:25 <sbp> Phantafmatography, I can hardly fpeak it ?
16:13:25 <sbp> --
16:13:49 <sbp> er, Google Books interpreted the long s'es as f, of course. silly thing
16:18:41 <sbp> [[[
16:18:42 <sbp> [53]
16:18:42 <sbp> Of a Scholler.
16:18:42 <sbp> A Scholler as he was blowing the fire in a winter night, his bellowes nose fell out, Gentlemen quoth he, it must needs bee cold weather when the bellowes no[s]e doth droppe.
16:18:52 <sbp> ]]] — http://www.mun.ca/alciato/jests/jest2.html
16:19:48 *** Arnia has quit (Read error: 110 (Connection timed out))
16:28:43 <sbp> ah, MacKay will stand down at the next election
16:36:49 <sbp> hmm, Snow Leopard looks most likely to ship in August
16:37:16 <sbp> but it looks like they'll tell us about it at WWDC
16:39:16 <sbp> phenny: "Die Feuerschützen-Gesellschaft zu Basel"?
16:39:17 <phenny> sbp: "The fire-protecting society Basel" (de to en, translate.google.com)
16:43:51 <sbp> another reason to dislike git:
16:43:52 <sbp> $ ls -1 /usr/local/bin/git-* | wc -l
16:43:52 <sbp>   141
16:44:23 *** jetscreamer (n=jetscrea@unaffiliated/jetscreamer) has joined #swhack
16:46:10 <sbp> it's amazing how much complete crap is in my /usr/local/bin
16:46:32 <kpreid> /usr/local/rubbish_bin
16:46:54 <sbp> hehe
16:47:01 <kpreid> these days I like to do the thing of putting mostly symlinks in /usr/local/bin, so I at least know where the stuff came from
16:47:22 <sbp> interesting idea. but where do you install to originally?
16:47:30 <sbp> I assume most of these come from configure/make processes
16:47:39 <sbp> so presumably you're passing --prefix to configure, right?
16:48:49 <kpreid> yup
16:49:14 <kpreid> Usually something like --prefix /Stuff/pkgname/
16:49:39 <kpreid> There is also software (e.g. GNU stow) to automate this, but I've never bothered with it
16:49:47 *** nsh has quit (Connection timed out)
16:50:04 <sbp> oh, that's an interesting idea
16:50:24 <sbp> that's a very interesting idea
16:50:24 <kpreid> the other thing you get out of it is trivial uninstalling:
16:50:35 <sbp> rm -rf /Stuff/packagename
16:50:40 <kpreid> blow away the package directory then all the symlinks which name it
16:50:45 <sbp> aye
16:50:58 <sbp> man, that's clever shit
16:51:04 <kpreid> also means that you can do stuff like run the installers as arbitrary accounts since they own the tree
16:51:08 <sbp> I was thinking about just maintaining a ~/local
16:51:17 <sbp> then I'd be reasonable sure it wasn't arsing about with stuff in /
16:51:25 <kpreid> so you can arrange so that you KNOW the installer hasn't spewed random cruft to wherever it thinks it is right
16:51:25 <Monty2> another day but "global" design docs?
16:51:25 <sbp> and I could do it all without the superuser account
16:51:46 <sbp> oh, that's a point. so don't even use your own account?
16:51:51 <sbp> use a guest account, sort of thing?
16:52:03 <kpreid> perhaps, but I just like not doing my make/make install as root
16:52:14 <sbp> sure, that's what I meant about ~/local
16:52:15 <kpreid> stuff it all in one bin that isn't uid 0
16:52:23 <kpreid> also not my personal account either
16:52:24 <sbp> but your plan has the extra benefit of the package division
16:53:23 <kpreid> oh yeah, and my recipe for maintaining symlinks of that sort is ln -fish
16:53:42 <sbp> ln -fish?
16:53:48 * sbp checks the man page
16:55:22 <sbp> -f[orce], -i[nteractive] (overridden by -f), -s[ymbolic] link, -h (don't follow symlinks regarding the file you're linking to)
16:55:23 <sbp> cool
16:55:46 * sbp looks at http://www.pathname.com/fhs/pub/fhs-2.3.html#USRLOCALLOCALHIERARCHY
16:56:04 <kpreid> yes, it's not entirely sensible but it is easy to remember :)
16:56:09 <sbp> right
16:56:13 <sbp> like du -cks
16:56:21 <sbp> ducks and fish
16:56:34 <kpreid> -i is to permit overwriting for updating, and -h because it's just the right thing
16:56:35 <sbp> okay, so you can't sensibly install the packages into ~/local/packagename
16:56:50 <sbp> so now I'm wondering how to deploy your system cleanly
16:56:59 <kpreid> why's that?
16:57:19 <sbp> well it says:
16:57:22 <sbp> “No other directories, except those listed below, may be in /usr/local after first installing a FHS-compliant system.”
16:57:38 <sbp> and the intention of ~/local is to have it act like /usr/local
16:57:43 <sbp> perhaps I should eschew that altogether
16:57:52 <sbp> because what you're actually doing is creating a series of *locals*
16:57:59 <sbp> ~/locals would be sensible
16:58:07 <sbp> ~/locals/packagename/stuff-goes-here
16:58:21 <sbp> or even just ~/packages
16:58:43 <sbp> which would you prefer? ~/locals or ~/packages?
16:58:55 <kpreid> oh I see
16:58:59 <kpreid> I would ~/packages
16:59:00 <sbp> of course that means you either symlink in /usr/local/bin, or ~/bin or whatever
16:59:17 <kpreid> I do systemwide installs and I use /Stuff for no particular reason
16:59:49 <sbp> yeah, if I were doing systemwide stuff and using the Mac Capitals Thing, I'd probably do /Packages
17:00:02 <sbp> not sure whether I want to use /usr/local/bin or ~/bin
17:00:21 <sbp> I also have scripts of my own that I want to bung in there
17:00:35 <sbp> not sure whether to put them in ~/packages/inamidst/bin or something and then symlink them
17:02:02 <kpreid> I wouldn't
17:02:15 <sbp> whyn't so?
17:02:47 <sbp> I'm also wondering about one-shot installs
17:02:55 <sbp> for example, pngout is just a single binary
17:02:59 <sbp> but I didn't write it
17:06:06 *** nsh_ (n=nsh@host86-141-178-190.range86-141.btcentralplus.com) has joined #swhack
17:06:24 <sbp> also I think I'm inclined more towards ~/scripts than ~/bin anyway
17:06:41 <nsh_> unless they're binary scripts
17:06:48 <nsh_> WRITTEN IN HOPSCODE
17:06:54 <sbp> yeah, in which case ~/binscripts
17:06:55 *** nsh_ is now known as nsh
17:07:06 <nsh> k
17:07:22 * nsh goes back to familytiems
17:07:45 <nsh> also, all my keys appear to now work1234567890-=
17:07:49 <nsh> OooOooO
17:08:16 <sbp> you can make quick bids!
17:08:28 <sbp> even backspace?
17:08:38 <sbp> if backspace worked, there wouldn't be any = on your list
17:14:38 <nsh> yeah, i readded the ==
17:14:42 <nsh> wasn't sure you'd grok
17:14:47 <nsh> sorry
17:15:13 <nsh> has been too much deals with teh berkpeeps
17:15:24 <deltab> http://smcv.pseudorandom.co.uk/2009/05/dbusgproxy-300.png
17:15:57 <nsh> nice
17:15:57 <sbp> bet Monty has loads of those calls
17:16:01 <Monty2> err, value and something :) webmachine+couchappstart is insufficient."?
17:16:08 <nsh> totals
17:16:47 <sbp> deltab: /usr/local/bin, ~/bin, or ~/scripts? for 1) single file programs I made, 2) single file programs other people made, and 3) programs in ~/packages (installed via configure/make)
17:17:16 <deltab> undecided
17:17:22 <sbp> ooh, played
17:17:41 <sbp> item three being kpreid's genius suggestion, I should note
17:19:03 <deltab> there's also single file programs that I've tinkered with
17:19:32 <sbp> hmm. I'd count those in 2), I think
17:19:40 <deltab> at home those all end up in ~/bin; at work, because others may want to use them, some things are in /usr/local/bin
17:20:07 <sbp> yeah, I'm not so worried about multi-user
17:20:56 <sbp> trying to make a list of packages I actually tend to use
17:21:07 <sbp> and which aren't on OS X by default (it's got python, rsync, etc.)
17:21:20 <sbp> not sure that all of these aren't in the default install, but anyway:
17:21:21 <sbp> advcomp
17:21:21 <sbp> exiftool
17:21:21 <sbp> hg/mercurial
17:21:21 <sbp> lynx
17:21:21 <sbp> optipng
17:21:23 <sbp> python3
17:21:25 <sbp> rakudo
17:21:27 <sbp> svn/subversion
17:21:29 <sbp> wget
17:21:33 <sbp> funny how three of these are PNG compression packages. heh
17:21:38 <sbp> er, only two actually
17:21:46 <sbp> didn't include pngout, because that's a single-file install
17:25:26 *** nsh_ (n=nsh@host86-141-178-190.range86-141.btcentralplus.com) has joined #swhack
17:25:37 <sbp> I sorta wonder too if gmake is needed
17:25:47 <sbp> because OS X make tends to be not very helpful
17:30:21 <sbp> hmm, another nice thing about kpreid's system is that you can easily move the old packages out of the way to install the latest version (and move 'em back again in the new version is borkens)
17:30:53 <chris2> leopard make is gnu make...
17:32:39 <sbp> chris2: oh? cool
17:32:57 <chris2> % which make
17:32:57 <chris2> /usr/bin/make
17:33:13 <chris2> % make --help | tail -1
17:33:13 <chris2> Report bugs to <bug-make@gnu.org>
17:33:22 <sbp> hmm, it seems I have git installed because (at least) I need it for rakudo
17:33:27 <sbp> which has a bizarre installation process
17:33:42 <sbp> cf.
17:33:43 <sbp> 20:04:43 <sbp> $ git clone git://github.com/rakudo/rakudo.git
17:33:43 <sbp> 20:05:27 <sbp> $ cd rakudo
17:33:43 <sbp> 20:06:28 <sbp> $ perl Configure.pl --gen-parrot
17:33:43 <sbp> — http://swhack.com/logs/2009-03-28#T20-04-43
17:33:49 <kpreid> "rakudo" is one of those wait-how-do-you-spell-it-names
17:33:52 <kpreid> "raduko?"
17:33:57 <chris2> how is that bizarre?
17:33:57 <sbp> hehe, yeah
17:34:00 <sbp> I had problems
17:34:27 <sbp> chris2: well there isn't a bundle to download, and git is messy
17:34:33 <chris2> git rocks :P
17:34:35 <sbp> chris2: and --gen-parrot is strange (though helpful)
17:34:46 <sbp> again:
17:34:47 <sbp> $ ls -1 /usr/local/bin/git-* | wc -l
17:34:47 <sbp>   141
17:34:53 <chris2> update your git,
17:34:55 <chris2> you git :P
17:35:10 <chris2> % ls ~/bin/git-* | wc -l
17:35:10 <chris2>    3
17:35:12 <sbp> how many binaries does it install now? only 133?
17:35:22 * sbp checks his version
17:35:28 <chris2> % ls ~/libexec/git-core/git-* | wc -l
17:35:28 <chris2>   137
17:35:36 <chris2> % git --version
17:35:36 <chris2> git version 1.6.0.14.g55dc
17:35:37 <sbp> $ git --version
17:35:37 <sbp> git version 1.5.2.1
17:35:43 <chris2> you need 1.6 for that
17:35:49 <sbp> ~/libexec?
17:35:49 <chris2> i could update as well
17:35:56 <chris2> yeah
17:36:07 <sbp> whazzat?
17:36:07 <chris2> i installed to ~
17:36:14 *** nsh has quit (Connection timed out)
17:36:15 <sbp> ah!
17:36:22 <chris2> binaries not meant for direct use
17:36:29 <chris2> look into /usr/libexec
17:37:49 <sbp> I had the FHS open in a tab somewhere...
17:38:26 <sbp> no "libexec" in http://www.pathname.com/fhs/pub/fhs-2.3.html
17:38:52 <chris2> too bad ;)
17:39:06 <sbp> considering that there's already /bin, /usr/bin, and /usr/local/bin, I must say I don't *really* see the point of /usr/libexec...
17:39:17 <sbp> not to mention /opt/bin, heh
17:39:23 <chris2> not cluttering up your path
17:39:29 <chris2> age-long history under unix
17:39:43 <sbp> wonder why it didn't make it into the FHS?
17:39:46 <xover> libexec is for binaries that you don't use directly; think plugins, scripts, etc.
17:39:46 <chris2> its not like the git guys invented it
17:39:46 <phenny> xover: 19 May 09:56Z <Arnia> tell xover to email it to me at my gmail address and I'll try and get to it in the next couple of days
17:40:00 <sbp> no, I'm not blaming git
17:40:10 <sbp> just saying I hadn't heard of it and it's not in the FHS...
17:40:34 <chris2> debian apparently puts stuff to /usr/lib now
17:40:38 <xover> It may be a Red Hat thing.
17:40:47 <sbp> Wikipedia sez: “Dynamically loaded libraries go in /usr/libexec and similar directories.”
17:42:12 <xover> You may be better off looking at the definitions given by RPM and the auto* tools.
17:44:07 *** Arnia (n=jgeldart@0-16-cb-bd-57-e.it.wlan.dur.ac.uk) has joined #swhack
17:44:07 <Monty2> howdy, Arnia
17:45:46 *** leobard (n=Miranda@91-66-122-97-dynip.superkabel.de) has joined #swhack
17:47:51 <sbp> hmm, Panther has xmlwf installed by default
17:48:06 <sbp> xover: what's "RPM"?
17:48:14 * sbp ducks
17:48:55 <kpreid> .gcs xmlwf xmlwtf
17:48:56 <phenny> xmlwf (3,460), xmlwtf (692)
17:49:10 <kpreid> .gc htmlwf htmlwtf
17:49:11 <phenny> htmlwf htmlwtf: 0
17:49:15 <kpreid> .gcs htmlwf htmlwtf
17:49:16 <phenny> htmlwtf (1,200), htmlwf (8)
17:49:23 <kpreid> .gcs jsonwf jsonwtf
17:49:24 <phenny> jsonwtf (10), jsonwf (0)
17:58:39 *** leobard has quit ("Miranda IM! Smaller, Faster, Easier. http://miranda-im.org")
18:03:59 <sbp> hmm, you can't seem to configure where exiftool is installed to
18:04:04 <sbp> because it uses perl Makefile.PL
18:04:13 <sbp> which takes neither --help nor --prefix options
18:05:43 <sbp> hmm, and mercurial doesn't have a ./configure either
18:05:46 <sbp> its first step is make
18:06:43 <sbp> and $ python setup.py --help is... unhelpful
18:07:10 *** leobard (n=Miranda@91-66-122-97-dynip.superkabel.de) has joined #swhack
18:07:46 <sbp> $ tar -jxvf lynx2.8.6.tar.bz2
18:07:46 <sbp> lynx2-8-6
18:07:47 <sbp> (fail)
18:12:48 <sbp> optipng's configure script:
18:12:48 <sbp> $ ./configure -help
18:12:48 <sbp> Usage:
18:12:48 <sbp>   ./configure [options]
18:12:48 <sbp> Options:
18:12:49 <sbp>   -h, -help        Show this help
18:12:51 <sbp>   -with-system-zlib    Use the system-supplied zlib
18:12:53 <sbp>   -with-system-libpng   Use the system-supplied libpng
18:12:55 <sbp> Environment variables:
18:12:57 <sbp>   CC           C compiler command
18:12:59 <sbp> baffling lack of --prefix support
18:23:49 <sbp> okay, I just tried installing seven packages
18:24:01 <sbp> using ~/packages
18:24:04 <sbp> * advancecomp - installed fine
18:24:18 <sbp> * exiftool - failed because uses perl Makefile.PL, and no --prefix support
18:24:34 <sbp> * mercurial - failed because uses make with no configure step
18:24:41 <sbp> * lynx - installed fine
18:24:57 <sbp> * optipng - failed because its configure script doesn't support --prefix
18:25:04 <sbp> * python3 - installed fine
18:25:08 <sbp> * wget - installed fine
18:25:28 <sbp> disappointingly low success rate there
18:25:49 <xover> .wik .rpm
18:25:50 <phenny> "RPM Package Manager is a package management system[1]|." - http://en.wikipedia.org/wiki/.rpm
18:25:54 <sbp> technically you don't have to install exiftool
18:26:04 <sbp> you can just move the directory to ~/packages/exiftool
18:26:15 <sbp> but it doesn't follow the /usr/local conventions in that case
18:27:31 <sbp> (for the record I know what RPM is, and I've even used it believe it or not...)
18:28:12 <sbp> also for the record, I have 194/5673 ports installed in MacPorts
18:28:21 <sbp> and most of those nearly 200 packages I don't really use
18:31:36 <sbp> wow:
18:31:36 <sbp> $ du -cks -h /opt
18:31:36 <sbp> 3.2G  /opt
18:31:36 <sbp> 3.2G  total
18:31:56 <sbp> $ du -cks -h ~/packages
18:31:56 <sbp> 78M  /Users/sbp/packages
18:31:56 <sbp> 78M  total
18:32:07 <sbp> (not that I managed to install many packages in that experiment)
18:32:15 <sbp> going to try a more tricky looking one now: ffmpeg
18:33:35 <sbp> $ svn checkout svn://svn.ffmpeg.org/ffmpeg/trunk ffmpeg
18:33:36 *** nsh_ has quit (Connection timed out)
18:36:24 *** nsh_ (n=nsh@host86-141-178-190.range86-141.btcentralplus.com) has joined #swhack
18:40:16 *** [bjoern] (n=bjoern@dslb-094-223-180-085.pools.arcor-ip.net) has joined #swhack
18:40:31 <sbp> LO, A STRANGER ARRIVES
18:40:38 *** lisppaste2 has quit (Remote closed the connection)
18:40:46 <sbp> LO, A BOT FALLS
18:40:54 <[bjoern]> I am a strangest.
18:41:00 <[bjoern]> No comparatives, please.
18:41:04 <sbp> superlativest
18:41:22 <sbp> I'm installing ffmpeg (again)
18:41:27 <sbp> into ~/packages/ffmpeg
18:41:35 <sbp> to test how well packages can cope with --prefix
18:41:35 <[bjoern]> "Pizzza Man Dressed As Superhero Comes To Rescue (with video)"
18:41:42 <sbp> so far, only 4/7 are of wins
18:41:52 <[bjoern]> that's much
18:41:58 <sbp> any Pizzza Woman?
18:42:08 <[bjoern]> mail had non-text content
18:42:20 <sbp> worked: advancecomp, lynx, python3, wget
18:42:28 <sbp> didn't work: exiftool, mercurial, optipng
18:42:52 <sbp> exiftool because it's perl, and perl makefiles can't take a --prefix
18:42:59 <sbp> mercurial because it's python and see above
18:43:01 <[bjoern]> actually yes they do
18:43:04 <sbp> and optipng just because it sucks
18:43:08 <sbp> oh? how do you specify that?
18:43:15 <sbp> I tried perl Makefile.PL --prefix
18:43:18 <sbp> and it was like what
18:43:24 <sbp> so I tried perl Makefile.PL --help
18:43:27 <sbp> and it was like what
18:43:29 <sbp> so I gave up
18:44:02 <[bjoern]> Makefile.PL PREFIX=...
18:44:14 <[bjoern]> perldoc perlmodinstall
18:44:49 <sbp> * ffmpeg - installed fine
18:45:12 <sbp> [bjoern]: ah, interesting
18:45:13 <sbp> trying this
18:46:12 <sbp> $ perl Makefile.PL PREFIX=/Users/sbp/packages/exiftool
18:46:12 <sbp> Checking if your kit is complete...
18:46:12 <sbp> Looks good
18:46:12 <sbp> Writing Makefile for Image::ExifTool
18:46:20 * sbp checks the Makefile
18:46:43 <sbp> looks good
18:46:59 <sbp> great, it worked! thanks
18:47:08 <sbp> now, I wonder if mercurial has a similar hidden option
18:47:15 <sbp> hey, so we're at 6/8
18:47:39 <[bjoern]> It is I, reader of the PODs, knower of the documentation.
18:47:50 <sbp> http://www.selenic.com/mercurial/wiki/UnixInstall#Per-user_installation
18:47:51 <sbp> hehe
18:48:41 <sbp> these instructions look a bit dubious...
18:48:53 *** lisppaste2 (n=lisppast@common-lisp.net) has joined #swhack
18:49:26 <sbp> there's a $ make local
18:49:48 <sbp> wonder what that does
18:50:12 <sbp> I'll try 'em both
18:50:49 <sbp> oh wait, I didn't read the first one properly
18:50:51 <sbp> can't do that
18:50:58 <sbp> so it'll have to be the make local thing
18:52:22 <[bjoern]> Now you wish it was a Perl module.
18:52:46 <sbp> ...a little bit, yes!
18:52:49 <sbp> hmm, there's this:
18:52:50 <sbp>  install   - install program and man pages to PREFIX (/usr/local)
18:52:58 <sbp> but the documentation indicates that you have to munge PYTHONPATH too
18:53:08 <sbp> because basically it's going to install it whereever your python libs are
18:53:38 <sbp> and if you're not doing that, then python can't find it
18:54:32 <[bjoern]> You to use one of the many similar things for Perl too, or use "use lib ..." at runtime
18:54:38 <[bjoern]> have
18:55:02 <sbp> I suppose there needs to be a homedir equivalent of site-packages which can be symlinked to
18:55:45 *** jsled has quit ("Terminated with extreme prejudice - dircproxy 1.0.5")
18:56:07 <sbp> hmm, going to try this:
18:56:07 <[bjoern]> He's like totally still using dircproxy
18:56:22 <sbp> $ make install PREFIX=/Users/sbp/packages/mercurial
18:56:36 <sbp> yeah, pfft. jsled ain't even paying attention
18:57:55 <sbp> apparently this was an alias for:
18:58:08 <sbp> python setup.py build
18:58:08 <sbp> python setup.py install --prefix="/Users/sbp/packages/mercurial" --force
18:58:25 <sbp> but anyway, it seems to have worked
18:58:34 *** jetscreamer has quit ()
18:58:57 <sbp> so, 7/8. only optipng has failed
18:58:58 <sbp> $ du -cks -h ~/packages
18:58:59 <sbp> 123M  /Users/sbp/packages
18:58:59 <sbp> 123M  total
18:59:24 <sbp> I haven't actually tested the binaries. maybe I should do that!
19:01:18 <sbp> $ ln -fish ~/packages/advancecomp/bin/advpng
19:01:18 <sbp> $ ln -fish ~/packages/exiftool/bin/exiftool
19:01:18 <sbp> $ ln -fish ~/packages/ffmpeg/bin/ffmpeg
19:01:18 <sbp> $ ln -fish ~/packages/lynx/bin/lynx
19:01:18 <sbp> $ ln -fish ~/packages/mercurial/bin/hg
19:01:18 <sbp> $ ln -fish ~/packages/python3/bin/python3.0
19:01:20 <sbp> $ ln -fish ~/packages/wget/bin/wget
19:01:44 <sbp> exiftool doesn't work
19:01:53 <sbp> "Can't locate Image/ExifTool.pm in @INC"
19:02:19 <sbp> o reader of PODs, knower of documentation! what is to be done about said problem?
19:04:17 <sbp> (mercurial fails for a similar reason, but I know why and how to fix that)
19:05:03 <sbp> o [bjoern] of [bjoern]s, within the chan
19:05:09 <sbp> give bold instructions, if thou can!
19:06:37 <sbp> thinking about it, a ~/local would be helpful for this
19:06:48 <sbp> could link the packages into ~/local/lib/python2.5/site-packages
19:06:54 <sbp> and add that to PYTHONPATH
19:07:04 <sbp> and presumably a similar fix for perl
19:07:17 <xover> perl -Idirectory
19:07:29 <sbp> xover: no per-command solutions
19:07:34 <sbp> need something envar related
19:09:06 <xover> @INC (module search path) is set at perl compile time; put your modules somewhere in @INC, add -I to the perl command, set $PERL5LIB, or add "use lib 'dir';" to the program.
19:09:45 <xover> The path of least resistance is usually to not install Perl modules outside of the compiled-in default @INC paths.
19:10:14 <xover> In particular, I know p5p and the Apple Darwin folks put some effort into making this stuff Just Work when you don't try to be clever.
19:11:40 <sbp> okay. looks like I need to symlink in ~/local/lib/perl5/5.8.8?
19:12:14 <sbp> the version stuff is not very pretty
19:12:19 <sbp> same with python
19:12:34 <sbp> PYTHONPATH=~/local/lib/python2.5/site-packages
19:12:34 <sbp> PERL5LIB=~/local/lib/perl5/5.8.8
19:12:38 <xover> Which perl are you using?
19:12:49 <sbp> $ perl --version
19:12:50 <sbp> This is perl, v5.8.8 built for darwin-2level
19:13:07 <xover> Yes, but is the system default one or something homebrew?
19:13:14 <xover> +it
19:13:16 <sbp> dunno, doesn't really matter
19:13:41 <sbp> it's not in ~/packages, if that's what you mean
19:14:05 <sbp> if you've not been following, I'm trying to implement a scheme that kpreid suggested
19:14:28 <sbp> where he ./confifugre --prefix=~/packages/packagename software packages
19:14:30 <xover> Oh. You're actively /trying/ to be clever. Nevermind then.
19:14:31 *** lmorchard (n=lmorchar@66.103.254.157.static.a2webhosting.com) has joined #swhack
19:14:47 <sbp> no, not being clever
19:14:49 <xover> Best of luck to ye laddie.
19:14:52 <sbp> basically the idea is: a non-root install
19:15:13 <sbp> and it just doesn't seem like python or perl support that without some pain
19:15:27 <sbp> all of the C programs work fine
19:15:35 <sbp> (except for optipng, inexplicably)
19:15:57 <xover> Perl at least supports that just fine…
19:16:00 <sbp> xover: for example, if I were installing this on crschmidt's machine where I don't have root...
19:16:06 <sbp> xover: okay, how?
19:16:16 <xover> …if only it was perl and not its modules you were trying to install.
19:16:26 * kpreid hasn't tried doing this scheme with perl or python modules
19:16:27 <sbp> see, there you go! :-)
19:16:46 <sbp> kpreid: the only problem actually seems to be that version management gets a little tricky
19:17:00 <sbp> kpreid: as I say above, basically it necessitates doing this:
19:17:01 <sbp> PYTHONPATH=~/local/lib/python2.5/site-packages
19:17:01 <sbp> PERL5LIB=~/local/lib/perl5/5.8.8
19:17:03 * xover hugs RPM…
19:17:09 <sbp> and then linking the modules as you would link the binaries
19:17:20 <sbp> if you can symlink the binaries, you can symlink the modules...
19:17:30 <sbp> but you symlink the binaries into a single dir
19:17:41 <sbp> whereas where you symlink the modules depends on your python/perl version
19:17:44 <sbp> which is sorta annoying
19:18:02 <sbp> because I suspect there there'll be a lot of backwards compatibility
19:18:12 <sbp> like, surely perl 5.8.9 is going to support perl 5.8.8 modules...
19:18:36 <sbp> on that metric, symlinking into ~/local/lib/perl5 would perhaps be more reasonable
19:18:41 <sbp> but then you see the problem in python
19:19:06 <sbp> python2.5 can support python2.4 stuff, but python3.0.1 can't support python2.5
19:19:25 <sbp> hmm
19:19:45 <sbp> I mean, you wouldn't do ~/local/lib/python2/site-packages
19:19:46 *** kpreid has quit (Remote closed the connection)
19:20:31 <xover> perl -e 'print join "\n", @INC'
19:20:43 <sbp> yeah, did that
19:20:48 <sbp> it's a confusing mess of stuff
19:21:03 <xover> Take a look at the dirs in there and take a guess at how it handles versioning.
19:21:15 <sbp> I tried...
19:21:37 <xover> i.e. there's no particular reason to put your modules into a versioned directory.
19:22:03 <sbp> hmm. all of mine are versioned except two:
19:22:03 <sbp> $ perl -e 'print join "\n", @INC'
19:22:03 <sbp> /opt/local/lib/perl5/5.8.8/darwin-2level
19:22:03 <sbp> /opt/local/lib/perl5/5.8.8
19:22:03 <sbp> /opt/local/lib/perl5/site_perl/5.8.8/darwin-2level
19:22:03 <sbp> /opt/local/lib/perl5/site_perl/5.8.8
19:22:05 <sbp> /opt/local/lib/perl5/site_perl
19:22:07 <sbp> /opt/local/lib/perl5/vendor_perl/5.8.8/darwin-2level
19:22:09 <sbp> /opt/local/lib/perl5/vendor_perl/5.8.8
19:22:11 <sbp> /opt/local/lib/perl5/vendor_perl
19:22:13 <sbp> and they have site_perl and vendor_perl
19:22:25 <[bjoern]> Versions when upgrading matter mostly if you have XS modules and go from 5.8 to 5.10, otherwise there should be no problem
19:22:25 <sbp> not quite sure what the differences are there
19:22:35 <[bjoern]> (if you make that switch, you have to recompile all your xs modules)
19:22:40 *** eel has quit (Read error: 104 (Connection reset by peer))
19:22:58 <sbp> well that's not too bad
19:23:34 <sbp> hmm, on helios it's more clear:
19:23:35 <sbp> sbp@helios:~$ perl -e 'print join "\n", @INC'
19:23:35 <sbp> /etc/perl
19:23:35 <sbp> /usr/local/lib/perl/5.10.0
19:23:35 <sbp> /usr/local/share/perl/5.10.0
19:23:35 <sbp> /usr/lib/perl5
19:23:37 <sbp> /usr/share/perl5
19:23:39 <sbp> /usr/lib/perl/5.10
19:23:41 <sbp> /usr/share/perl/5.10
19:23:43 <sbp> /usr/local/lib/site_perl
19:23:55 <[bjoern]> Using cpan you can make a bundle module for all your installed modules, then you just install the bundle with all dependencies and you have all your modules reinstalled, on such an occasion.
19:24:31 <sbp> actually, I should see where it installed to in ~/packages/exiftool, thinking about it
19:25:26 <sbp> hmm:
19:25:27 <sbp> lib/perl5/site_perl/5.8.8/File
19:25:27 <sbp> lib/perl5/site_perl/5.8.8/File/RandomAccess.pm
19:25:27 <sbp> lib/perl5/site_perl/5.8.8/File/RandomAccess.pod
19:25:27 <sbp> lib/perl5/site_perl/5.8.8/Image
19:25:29 <sbp> lib/perl5/site_perl/5.8.8/Image/ExifTool
19:25:31 <sbp> lib/perl5/site_perl/5.8.8/Image/ExifTool/AFCP.pm
19:25:33 <sbp> lib/perl5/site_perl/5.8.8/Image/ExifTool/AIFF.pm
19:25:35 <sbp> lib/perl5/site_perl/5.8.8/Image/ExifTool/APE.pm
19:25:37 <sbp> etc.
19:25:49 <sbp> so technically, ~/usr/local/lib/perl5/site_perl/5.8.8
19:25:54 <sbp> and I need to link File and Image
19:26:06 <sbp> er, ~/local/lib/perl5/site_perl/5.8.8
19:26:41 <[bjoern]> I would imagine you can set the env var using something dynamic that would give you the right dirs.
19:26:49 <[bjoern]> ala X=`Y`
19:27:08 <sbp> that would be nice
19:28:04 <sbp> this has a fairly acceptable symmetry:
19:28:05 <sbp> PYTHONPATH=~/local/lib/python2.5/site-packages
19:28:05 <sbp> PERL5LIB=~/local/lib/perl5/site_perl
19:28:15 <sbp> what's not going to work is when something else installs to File/blahblah
19:28:38 <[bjoern]> Things would be easier if you had rebuilt Perl and Python yourself for local use, I would imagine.
19:28:49 <sbp> because if I'm soft linking to ~/packages/exiftool/lib/perl5/site_perl/5.8.8/File, obviously it's going to be hard to mix another File in with it...
19:28:58 <sbp> yeah, that's a good point
19:30:00 <sbp> even then I'm not quite sure how it'd work in this system. hmm
19:30:40 <sbp> I guess when you install a perl module, you'd have to use prefix=~/packages/perl5
19:30:48 <sbp> sorry, PREFIX=~/packages/perl5
19:31:04 <sbp> though normally you'd use ~/packages/exiftool
19:31:09 <sbp> gonna try this
19:31:28 <sbp> at least it's easy to remove exiftool!
19:31:31 <sbp> rm -rf ~/packages/exiftool
19:31:34 <[bjoern]> If you build your own perl you can specify where the modules go and never worry about it again..
19:32:30 <sbp> until you install perl again...
19:32:36 *** kpreid (n=kpreid@cpe-67-249-58-190.twcny.res.rr.com) has joined #swhack
19:32:45 <sbp> I had this problem with python
19:32:58 <sbp> you install python2.1, 2.2, 2.3, 2.4, 2.5
19:33:04 <sbp> and along the way you install various packages
19:33:18 <sbp> and you go mad remembering which package is installed with which python version
19:33:34 <sbp> but that's not peculiar to the ~/packages system; that's a problem with root installation too
19:34:34 <[bjoern]> I usually only upgrade Perl along with reinstalling the whole system...
19:34:43 <sbp> hmm, strange configure option:
19:34:46 <sbp> ./Configure -des -Dprefix=$HOME/localperl
19:35:29 <sbp> good job they didn't call it "configure"
19:35:32 <sbp> and I guess that's why...
19:36:14 <sbp> (5.10, by the way)
19:36:43 <sbp> Build Perl with MAD? [n]
19:37:05 <sbp> <socks.h> NOT found. (NOT found?)
19:38:41 * [bjoern] build-perl-yourself conclusion, on windows anyway, was "Works. Now, never try that again"
19:38:50 <sbp> chuckle
19:38:56 <[bjoern]> Lately I suck more at typing than usual.
19:39:16 <sbp> yeah, what's all this jibbo-jabbu
19:40:30 <sbp> I had made python to ~/packages/python3
19:40:40 <sbp> but I'm gonna remake it to ~/packages/python3.0.1 I think
19:40:43 <sbp> just to make it clear
19:41:53 <sbp> basically I think the idea is that if you're likely to want to install programming language modules, add a version number to the programming language package
19:43:03 <[bjoern]> Gotta wonder why package manager writers have not found a unified solution for this...
19:43:26 <sbp> partisan politics
19:44:37 <sbp> I think this is fairly clean so far
19:44:44 <sbp> 5/8 are working
19:45:15 <[bjoern]> Your 0x10 problem is getting worse too btw.
19:45:17 <sbp> hopefully gonna be 9/10 when perl5.10 and python2.5 are installed (with exiftool and mercurial)
19:45:25 <sbp> hehe. yeah, I can't see that I've typed it
19:45:33 <sbp> ...........
19:45:56 <sbp> I'm using a Windows keyboard plugged into OS X
19:46:11 <sbp> whenever I brush the Right Click button, it sends an 0x10 for some reason
19:46:19 <[bjoern]> You should plug OS X into a Windows keyboard.
19:46:33 <sbp> oh that would just be terrible
19:46:40 <sbp> I'd be typing 01x0s or something
19:47:09 <sbp> --
19:47:10 <sbp> $ make install
19:47:10 <sbp> make install.perl install.man STRIPFLAGS= DESTDIR=""
19:47:10 <sbp> make[1]: Entering directory `/Users/sbp/stuff/packages/perl-5.10.0'
19:47:10 <sbp>     AutoSplitting perl library
19:47:10 <sbp> ./miniperl -Ilib -e 'use AutoSplit; \
19:47:12 <sbp>         autosplit_lib_modules(@ARGV)' lib/*.pm
19:47:14 <sbp> --
19:47:17 <sbp> (miniperl?)
19:47:39 <[bjoern]> .gc "sold * soul to God"
19:47:40 <phenny> "sold * soul to God": 125
19:47:47 <[bjoern]> bootstrapping
19:47:57 <[bjoern]> Perl needs Perl to build Perl
19:48:00 <sbp> I can imagine it installing perl1 first
19:48:04 <sbp> then using that to install perl2...
19:48:34 <sbp>  /Users/sbp/packages/perl5.10/man/man1/perlguts.1
19:48:47 <sbp> perlguts is what you get when a perl program explodes?
19:50:04 <sbp> okay, installing python2.6.2 too
19:50:36 <[bjoern]> Perl internals you were afraid to ask about.
19:50:37 <sbp> perl5.10 seems to have worked
19:51:23 <sbp> now to try installing ExifTool again...
19:53:28 <sbp> $ ~/bin/perl Makefile.PL PREFIX=/Users/sbp/packages/perl5.10
19:55:50 <sbp> $ ln -fish ~/packages/perl5.10/bin/exiftool
19:56:13 <sbp> seems to work!
19:57:10 *** Morbus has quit ("http://www.disobey.com/")
19:57:16 <sbp> now for python and mercurial
19:57:49 <deltab> what's -h?
19:58:05 <sbp> deltab: if the file you're linking to is a symlink, don't follow it
19:58:10 <sbp> just symlink to the symlink
19:58:33 <sbp> (kpreid said something like "it's just the Right Thing")
19:58:43 <deltab> -f and -i?
19:58:46 <kpreid> er, NO
19:58:48 <sbp> yeah,
19:58:53 <kpreid> it affects not following the *target*
19:58:56 <sbp> er, yeah, they're mutually contradictory
19:58:58 <sbp> but -f wins out
19:59:09 <kpreid> e.g. if -h is not given and you ln -s foo bar, and bar is a symlink, the target of bar will be replaced
19:59:09 <sbp> kpreid: hmm?
19:59:27 <sbp> ah!
19:59:30 <sbp> woah, that is scary
19:59:58 <sbp> ...so yeah, I agree that -h really is the Right Thing :-)
20:00:45 <deltab> only an issue if you have ln overwrite existing destinations
20:01:14 <sbp> which you do if you have -f
20:01:49 <sbp> so yeah, ln -s would be okay
20:02:11 <sbp> but if you're adding -f you probably want -h too
20:02:17 * deltab nods
20:02:20 <sbp> and if you've got f, s, and h, why not i...
20:03:03 <deltab> ah, well, I choose options for their effects, not the words they can spell :-)
20:03:15 <sbp> weirdo
20:04:02 <sbp> funnily enough, I used to have alias ducks='du -cks'
20:04:10 <sbp> but I didn't use it. I still kept doing du -cks
20:04:22 <sbp> it's not as fun when you're not being clever with the flags
20:06:46 <sbp> okay, for mercurial I think I'm going to need to do this:
20:06:50 <sbp> $ make install PYTHON=/Users/sbp/bin/python PREFIX=/Users/sbp/packages/python2.6.2
20:06:55 <sbp> or... I could run the commands myself
20:08:05 <sbp> which would be:
20:08:06 <sbp> ~/bin/python setup.py build
20:08:06 <sbp> ~/bin/python setup.py install --prefix=/Users/sbp/packages/python2.6.2 --force
20:10:50 <sbp> $ ln -fish ~/packages/python2.6.2/bin/python
20:11:37 * sbp tries mercurial with the make install command
20:11:46 *** nsh_ has quit (Connection timed out)
20:12:19 <sbp> excellent, it seems to have used the right python and installation prefix
20:12:49 <sbp> $ ln -fish ~/packages/python2.6.2/bin/hg
20:13:28 <sbp> great, it works!
20:13:40 <sbp> there, 9/10 packages working
20:13:53 <sbp> $ ls ~/packages
20:13:54 <sbp> ./       advancecomp/  lynx/      python2.6.2/  wget/
20:13:54 <sbp> ../       ffmpeg/     perl5.10/    python3.0.1/
20:14:03 <sbp> $ ls ~/bin
20:14:03 <sbp> ./       exiftool@    lynx@      python3.0@
20:14:03 <sbp> ../       ffmpeg@     perl@      wget@
20:14:03 <sbp> advpng@     hg@       python@
20:15:24 <sbp> the only one that didn't work is optipng
20:15:24 <Monty2> k sleep :) heya
20:15:30 <sbp> and in the configure script, it says:
20:15:30 <sbp> # This a hand-crafted, experimental, and, well, uncommented configure script.
20:15:30 <sbp> # Your feedback will be appreciated.
20:20:36 *** xodium (n=link@193.157.66.173) has joined #swhack
20:21:29 * sbp configures anyway and looks at the makefile
20:22:18 <sbp> hardcoded into src/scripts/gcc.mak:
20:22:19 <sbp> prefix=/usr/local
20:23:15 <sbp> so I guess if I just edited that...
20:24:09 *** leobard has quit ("Miranda IM! Smaller, Faster, Easier. http://miranda-im.org")
20:24:57 <sbp> $ find optipng
20:24:57 <sbp> optipng
20:24:57 <sbp> optipng/bin
20:24:57 <sbp> optipng/bin/optipng
20:24:57 <sbp> optipng/man
20:24:58 <sbp> optipng/man/man1
20:25:00 <sbp> optipng/man/man1/optipng.1
20:25:02 <sbp> great!
20:25:18 <sbp> $ ln -fish ~/packages/optipng/bin/optipng
20:25:26 <sbp> 10/10
20:28:22 <sbp> another nice thing about this is that you can roughly see individual package sizes:
20:28:23 <sbp> $ du -cks -h *
20:28:23 <sbp> 3.9M  advancecomp
20:28:23 <sbp> 33M  ffmpeg
20:28:23 <sbp> 1.5M  lynx
20:28:23 <sbp> 208K  optipng
20:28:25 <sbp> 59M  perl5.10
20:28:27 <sbp> 89M  python2.6.2
20:28:29 <sbp> 72M  python3.0.1
20:28:31 <sbp> 512K  wget
20:28:33 <sbp> 259M  total
20:28:39 <sbp> obviously perl has exiftool (small) and python2.6 has mercurial (large)
20:28:57 <sbp> interesting that python is bigger than perl though
20:29:30 <[bjoern]> you must delete 3MB
20:29:51 <sbp> heh, yeah
20:30:07 <sbp> [bjoern]: here:
20:30:08 <sbp> $ du -cks -h [^a]*
20:30:08 <sbp> 33M  ffmpeg
20:30:08 <sbp> 1.5M  lynx
20:30:08 <sbp> 208K  optipng
20:30:08 <sbp> 59M  perl5.10
20:30:10 <sbp> 89M  python2.6.2
20:30:12 <sbp> 72M  python3.0.1
20:30:14 <sbp> 512K  wget
20:30:16 <sbp> 256M  total
20:30:26 <[bjoern]> cheat
20:30:29 <sbp> heh, heh
20:35:00 <sbp> hmm, let's really play this shit. git and then rakudo?
20:35:17 <[bjoern]> naughty
20:36:55 <sbp> http://kernel.org/pub/software/scm/git/git-1.6.3.1.tar.bz2
20:37:24 *** libby (n=libby@94-169-110-24.cable.ubr22.aztw.blueyonder.co.uk) has joined #swhack
20:38:26 <sbp>     $ make prefix=/usr all doc info ;# as yourself
20:38:26 <sbp>     # make prefix=/usr install install-doc install-html install-info ;# as r
20:38:26 <sbp> oot
20:38:37 <[bjoern]> .oot
20:38:53 <sbp> or, as if that wasn't confusing enough:
20:38:53 <sbp>     $ make configure ;# as yourself
20:38:53 <sbp>     $ ./configure --prefix=/usr ;# as yourself
20:38:53 <sbp>     $ make all doc ;# as yourself
20:38:55 <sbp>     # make install install-doc install-html;# as root
20:39:17 <[bjoern]> .gcs ["Ocarina Of Time" oot] [oot]
20:39:18 <phenny> oot (1,080,000), "Ocarina Of Time" oot (26,400)
20:39:44 <sbp> going to do it the second way
20:40:46 *** nsh_ (n=nsh@host86-141-178-190.range86-141.btcentralplus.com) has joined #swhack
20:41:00 *** xodium has quit ("Leaving.")
20:42:25 <sbp> will be fun to see how many git-* executables there are
20:42:27 <sbp> and where it puts them
20:42:54 <sbp> hmm:
20:42:55 <sbp> /usr/bin/perl Makefile.PL PREFIX='/Users/sbp/packages/git'
20:42:55 <sbp> Writing perl.mak for Git
20:46:51 <MoiraA> hello sbp
20:47:25 <sbp> hey MoiraA
20:47:49 <MoiraA> how's things?
20:49:07 <sbp>   XMLTO git-add.1
20:49:07 <sbp> Unknown option: n
20:49:07 <sbp> Usage: head [-options] <url>...
20:49:07 <sbp>   -m <method>  use method for the request (default is 'HEAD')
20:49:09 <sbp> chuckle
20:53:36 <deltab> (more) HIGNFY!
20:58:08 <nsh_> now this is weird
20:58:11 <nsh_> the keys stopped working again
20:58:19 <nsh_> must be a connection thing rather than dampness then
20:58:33 <nsh_> or some moisture shifted and something on the line is stuck again
20:58:44 <nsh_> chore
20:59:38 <[bjoern]> perhaps they just hate you.
21:01:27 <nsh_> also a possibility
21:02:01 *** nsh_ has quit ()
21:02:33 <[bjoern]> .gc "Downloadable Index of the Internet"
21:02:33 <phenny> "Downloadable Index of the Internet": 4
21:17:15 *** jeffarch has quit (Read error: 104 (Connection reset by peer))
21:18:12 *** jeffarch (n=jeff@70-41-156-138.cust.wildblue.net) has joined #swhack
21:19:19 *** Arnia has quit (Read error: 110 (Connection timed out))
21:22:53 *** Arnia (n=jgeldart@87-194-158-123.bethere.co.uk) has joined #swhack
21:22:53 <Monty2> lo Arnia
21:42:24 *** scarless (n=cock@h75n3-um-a31.ias.bredband.telia.com) has joined #swhack
21:47:13 <[bjoern]> .gs hold an erection for *
21:47:14 <phenny> hold an erection for *: long (4), hours (4), up to 45 (3), so long (3), sexual orgasm (3)
22:23:16 *** nsh (n=nsh@wikipedia/nsh) has joined #swhack
22:41:25 *** nsh has quit (Remote closed the connection)
23:02:12 *** libby has quit (Read error: 113 (No route to host))
23:09:11 *** nsh (n=nsh@host86-141-178-190.range86-141.btcentralplus.com) has joined #swhack
23:33:56 *** jsled (n=jsled@dsl195.burlvtma.sover.net) has joined #swhack
23:33:57 *** nsh has quit (Remote closed the connection)
23:35:09 *** nsh (n=nsh@host86-141-178-190.range86-141.btcentralplus.com) has joined #swhack
23:35:09 <Monty2> howdy, nsh
23:35:27 <nsh> surp monty
23:35:29 <Monty2> thanks Wohali is misleading