00:00:04 but i'm pretty sure that it's at least okay, compared to other toolkits 00:00:05 thus, okayrad 00:00:12 it's a very lukewarm opinion 00:00:22 but implies... something greater... 00:00:25 or something. 00:01:37 my thoughts on Obj-C are pretty meh 00:01:37 "And so at 18 years of age, I moved from my tiny town of 800 people, changed my gender, moved to a new city, and tried to start a new life. I think I am failing (IAmA)." 00:01:41 I've got to wonder how the gender part fits in here 00:01:52 clearly to erase eir identity 00:01:58 ...although consciously changing your *gender* has to be some new superpower I've never heard of 00:02:07 i've only heard of correcting your sex to match your gender :-P 00:03:11 the terms aren't that well-defined 00:03:22 gender is mental, sex is physical 00:03:45 transgender is sex/gender mismatch, transsexual is prior sex/gender mismatch corrected by sex change 00:04:07 the text makes it clear it's just a mistake, but it's amusing 00:05:01 ehird: that's one definition and probably the most sensible one 00:05:30 it's the widely accepted one and the one all trans* (admittedly not very many) i've known have used /shrug 00:13:18 -!- MizardX has quit (Read error: 104 (Connection reset by peer)). 00:13:20 -!- MizardX- has joined. 00:13:46 -!- MizardX- has changed nick to MizardX. 00:56:01 * ehird considers ditching his VPS and getting a NearlyFreeSpeech.NET account... 00:56:02 Advantages: Less administration overhead, way cheaper (if I had a dynamic 300MB site that uses MySQL/InnoDB and served 10 GB a month, then it's around $7.50-$8 a month after the first month ($19.07 for the first month including domain registration)... so it's pretty much dirt cheap), tech savvy, minimal bullshit, they're pretty cool, good reputation, the neat thing where they mirror your site on a bunch of servers automatically and use the less populated ones 00:56:02 Disadvantages: Can't run overly-intensive processes, can't mess about with root, can't use things like FastCGI (just PHP or CGI), can't run IRC bots 00:56:13 it's a tough one :P 00:56:38 the administration thing is prolly a big deal though, my server's just sitting there because i haven't got it set up yet again 00:56:39 VPS? 00:57:50 virtual private server; basically you get root on a Xen instance 00:57:54 for like $20/mo 00:58:01 + bandwidth, storage etc 00:58:08 it's just like a dedi, except a bit slower and way cheaper 00:58:18 Not being able to run bots would suck. 00:58:20 * GregorR pets HackEgo. 00:58:27 GregorR: true. you know what they should do? 00:58:38 ... allow you to run bots? 00:58:39 ehird: why can't you run IRC bots? 00:58:44 GregorR: shush you, 00:58:48 ais523: because it's a web hosting service 00:59:00 i'd use it to host software repos, some useless bloggy thing, whatever 00:59:06 -!- augur_ has quit (Read error: 104 (Connection reset by peer)). 00:59:07 but anyway 00:59:13 there should be a service thing where: 00:59:25 Amazon's EC2 lets you spawn VPSes for an hourly rate or whatever, and they come with OSs and stuff 00:59:25 but 00:59:37 something like that, except instead of spawning a VPS, you just spawn a procses 00:59:39 *process 00:59:40 and it's cheaper 00:59:44 and still migrated across servers and shit 01:00:10 like, "run this process somewhere on your servers, I don't care, just give it an internet connection, a barebones Linux and some persistent storage" 01:00:43 then it'd just be like $ run-remote -d botinstance/ ./start irc.freenode.net 6667 01:01:05 it'd tar up botinstance, send it across, run ./start with those arguments, and persist any changes in the directory ./start is in (botinstance/) 01:01:10 and it could move across servers and shit automatically 01:01:25 basically like nohup, except it continues running after you shut the machine down :-P 01:01:42 (when it terminates you can download the final modified botinstance/) 01:02:29 it'd be like a vps, but cheaper + way simpler 01:02:50 also, it'd bring more-or-less pay for what you use pricing to VPSes, which is rather difficult with a regular sort of VPS 01:04:01 in conclusion, I ramble a lot 01:04:33 ais523: btw, the technical reason you can't run an irc bot or fastcgi or whatever is presumably that your site is served through various servers at various times depending on load 01:04:45 ah, yes 01:04:55 and migrating processes would be a Bitch(TM), especially when for a web hosting service, IRC bots are irrelevant (they just use up resources) and FastCGI is relatively unused 01:05:08 although with heavy web frameworks getting more common, it'll be an issue for them 01:05:19 as Rails, Django etc don't play too nicely with a set-up-every-request CGI environment... 01:05:34 still, static files are nice for anything I'd want to put on there, so that doesn't bother me 01:05:36 it's mainly the irc bot thing 01:05:42 ais523: what do you think of my remote-process idea up there? ↑ 01:06:00 ehird: busy now, I'll look later 01:22:29 -!- comex has changed nick to CallForJudgement. 01:23:33 -!- CallForJudgement has quit (Nick collision from services.). 01:23:38 -!- comex has joined. 01:46:12 http://www.ticalc.org/archives/news/articles/14/145/145273.html 01:46:14 all ti signing keys factored 01:46:22 "Yes, due to a gigantic effort by the entire community, via the boinc project, we could factor keys within a week." 01:46:28 i can see they're putting the resources to good use! 01:47:09 "I thought this was going to be a mathematical analysis of the musical keys in which Atlanta-born rapper TI can sing. I have to admit I was somewhat disappointed." 01:47:30 The keys are on Wikileak 01:47:40 As well as elsewhere 01:48:11 *Wikileaks 01:49:04 So? 01:49:11 TI won't give a shit. 01:50:42 Okay, maybe they do: http://www.reddit.com/r/reddit.com/comments/9fkzd/censorship_on_reddit_yes_here_is_the_evidence/c0cl3sq 01:51:14 Retards. 01:51:33 Someone ought to get my Arch install working. 02:07:48 :| 02:13:37 X is still shit, btw. 02:21:52 * Sgeo remembers seeing people agree with The UNIX Hater's Manual (iirc) about X sucking. 02:23:37 The UNIX Hater's Handbook 02:23:40 But, uhh, no shit? 02:27:16 -!- jix_ has joined. 02:38:11 -!- jix has quit (Read error: 110 (Connection timed out)). 02:51:45 -!- Gracenotes has joined. 03:07:00 -!- CESSMASTER has quit ("☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃☃"). 03:21:06 -!- CESSMASTER has joined. 03:26:56 * coppro hands CESSMASTER some sewage 03:33:26 i think we've established it refers to ILLICIT SUBSTANCES. 03:34:31 -!- oerjan has quit ("And once you've consumed it, it becomes EXCESS"). 03:48:14 Sum jenkem 03:53:07 Leroy Jenkems. 03:58:36 -!- MizardX- has joined. 03:59:53 -!- Sgeo has quit ("Leaving"). 04:02:10 -!- comex has quit (hubbard.freenode.net irc.freenode.net). 04:02:10 -!- bsmntbombdood has quit (hubbard.freenode.net irc.freenode.net). 04:02:11 -!- cmeme has quit (hubbard.freenode.net irc.freenode.net). 04:02:11 -!- zbrown has quit (hubbard.freenode.net irc.freenode.net). 04:02:11 -!- EgoBot has quit (hubbard.freenode.net irc.freenode.net). 04:02:11 -!- MizardX has quit (hubbard.freenode.net irc.freenode.net). 04:02:42 -!- comex has joined. 04:02:42 -!- bsmntbombdood has joined. 04:02:42 -!- cmeme has joined. 04:02:42 -!- zbrown has joined. 04:02:42 -!- EgoBot has joined. 04:02:44 -!- MizardX- has changed nick to MizardX. 04:11:21 -!- CESSMASTER has quit (Read error: 104 (Connection reset by peer)). 04:17:51 -!- CESSMASTER has joined. 04:52:43 Aug 29 12:31:30 "Monster.Dildo.Cocks.XXX.DVDrip.XviD-GraceNotes" —reddit <-- I thought this was a confusing name for a porn vid! 04:52:58 ... oldest ... logreading ... ever 04:52:59 ---: I figured it out. 04:53:00 ---: It's a guy in a rooster costume handing out complimentary dildos to promote the opening of a new adult video store. 04:53:13 wat? 04:53:18 Actually, I had happened to type `quote and get 04:53:20 `quote 78 04:53:21 78| ??? Are the cocks actually just implanted dildos? Or are there monster dildos and cocks? Or are both the dildos and cocks monster? 04:53:23 I'm infinitely confused, dude 04:53:34 So I pasted that to somebody, then the log it came from. 04:53:44 And he figured out what it actually is :P 04:53:57 GregorR: What what actually is? 04:54:17 Monster.Dildo.Cocks.XXX.DVDrip.XviD-GraceNotes 04:54:27 Ah. 04:54:40 I'm trying to figure out how the rooster comes in, dude. 04:54:43 ZOMG TORRENT PLZ 04:54:47 Monster 04:54:56 Slereah: It's on scene FTP, bitch. 04:54:57 ... COCK 04:54:57 You fucking cock. 04:55:06 GregorR: Oooooooooooooooooooooooooooooooooh. 04:55:07 >_< 04:55:26 That still doesn't explain Monster. :P 04:55:49 I assume the dildos are monster. 04:56:11 Like, shaped like the penis of a monster? 04:56:14 Define "monster"! 04:56:21 `define monster 04:56:22 * an imaginary creature usually having various human and animal parts \ * giant: someone or something that is abnormally large and powerful \ * freak: a person or animal that is markedly unusual or deformed 04:56:28 IMAGINARY 04:56:32 So this whole story is... 04:56:35 a... a FABRICATION 04:56:40 NOOOOOOOOOOOOOOOOOOOOES 04:56:50 ;_; 04:57:03 Anyway, to solve the mystery: http://www.reddit.com/r/IAmA/comments/9fc4n/i_was_in_a_major_mp3_releasing_group_in_the/c0cjyyj?context=1 05:02:34 GregorR: how much do I need to bribe you to rant about how much my remote process thingy idea sucks? 05:02:53 Depends, how much does it suck? 05:02:54 OH 05:02:57 That, I recall. 05:03:09 That's actually an interesting idea, I was considering implementing it modulo the fact that I don't have the resources. 05:03:26 I guess you need to pay more to get criticism :D 05:03:47 -!- ehird has left (?). 05:03:49 -!- ehird has joined. 05:03:49 -!- ehird has left (?). 05:03:58 -!- ehird has joined. 05:04:01 I am clumsy 05:04:14 You keep falling ... out of the channel? 05:04:23 There's a trap door in the corner. 05:05:02 ^ 05:05:04 ` 05:05:04 No output. 05:05:07 ehird: you know your duty 05:05:10 push the ` into the ^ 05:05:13 um 05:05:13 and there will no longer be a trapdoor 05:05:23 Compose-^ ` 05:05:24 you're welcome 05:05:33 s/-/ / 05:07:56 Actually I wish more things had Compose. 05:08:06 I should set it up, but OTOH I could just get Arch woring. 05:09:12 Well, Arch works; X11 doesn't. 05:09:28 * ais523 wonders what ehird is doing online at 5am 05:09:42 I could ask the same of you 05:09:51 well, I'm on holiday 05:09:57 but IIRC, school's started already 05:10:00 You and your logic :P 05:13:04 * ehird wonders why there isn't a flipped-` character 05:13:15 Then we could have better ``dumb quotes[INVALID][INVALID] 05:13:58 * ais523 is annoyed at not being able to use control-click for autoscroll 05:14:12 eh? 05:14:17 my mousewheel has started acting up, so I can't really use it for middle-clicking very easily 05:14:24 so I've been control-clicking links to open them in new tabs 05:14:39 but unfortunately, the same substitution doesn't work on autoscroll 05:14:46 xmodmap or sth 05:14:53 (I have Firefox set to paste on middle-click for textboxes, scroll on middle-click for everything else) 05:14:57 also, you really need to replace that computer... 05:15:10 broken touchpad, associated mouse has broken button... :D 05:15:14 hmm... is it possible to map mouse buttons-modifier combinations to other mouse buttons? 05:15:20 yes, I'm pretty sure 05:15:21 and it's not completely broken, just intermittent 05:15:32 worst-case, just catch Ctrl-click in a daemon 05:15:37 and send a middle click event 05:16:15 oh, this is the third mouse I've been on 05:16:24 the second one broke, the first one was wireless and used up too much battery power 05:16:36 and ofc the touchpad didn't work for more than about a minute 05:17:04 ais523: remember how you were incredulous at people paying more than a few hundred pounds for a computer? 05:17:10 it's so they don't have to deal with that shit. 05:17:13 I still am 05:17:24 the point is, despite all your claims, this computer still works sufficiently well for me to not really consider changing 05:17:36 the fact is that the components break 05:17:39 -!- augur has joined. 05:17:52 you won't say that when the keyboard breaks beacuse of the same shoddy engineering to cut costs 05:17:57 *because 05:18:31 ehird: on my first computer (a BBC Micro B), neither the Y or B key worked 05:18:34 but I worked around it 05:18:41 by redefining a couple of the function keys 05:18:53 ais523: and surely you are not surprised that other, more sane people would consider it a better solution to buy a new one? 05:19:07 ehird: it cost about £20, it was junk someone else was throwing out 05:19:12 I didn't really expect it to work well 05:19:19 then you are using an invalid example. 05:19:55 when I buy, e.g. a cheap Dell laptop, I expect it to be a bit of a pain to get working, not have very good components, and break sooner than a more expensive alternative 05:20:12 when I buy, e.g. a ThinkPad, I expect it to work fine, have good components, last longer and cost several times more 05:20:17 it's a tradeoff 05:20:21 sufficiently sooner that the more expensive one lasts longer on average? 05:20:27 absolutely 05:20:32 thinkpads last like 10 yaers 05:20:35 *years 05:20:37 I'd expect people like you to upgrade often enough that you'd be upgrading before it broke, anyway 05:20:46 I mean, could you cope with a computer that was top of the range 10 years ago? 05:20:48 -!- Pthing has quit (Remote closed the connection). 05:21:00 probably not, but the point is that something that totally breaks later on, 05:21:04 will likely get defects a lot sooner 05:21:23 besides, things like input devices are very important, for obvious reasons 05:21:47 and lower-end hardware has a tendency to... disagree... 05:21:52 (with each other) 05:22:57 I'm just saying that it's understandable paying £800 for a sturdy, lightweight ThinkPad with long battery life, a good screen, an exceptional keyboard, the nub mouse, high-quality hardware and good compatibility, vs paying £300 for a low-end Dell 05:22:59 I'm still amused that it came with a Windows XP manual 05:23:01 despite running Linux 05:23:16 heh, the battery life is about 10 minutes on here 05:23:18 haha 05:23:21 if you don't mind fiddling a whole bunch, don't mind being inconvenienced at all and are on a tight budget, of course the Dell is superior 05:23:23 but otherwise... 05:23:45 I'm glad that it's fast at hibernating 05:24:09 you should be able to close a notebook's lid and pick it up immediately afterwards imo 05:24:27 I think that the suspend modes you get in linux are wrong 05:24:37 I don't, I don't like the chance of the hard drive being knocked 05:24:47 ais523: see, the high-end notebooks have shock protection on the harddrive 05:24:47 ehird: as opposed to? 05:24:51 that automatically parks the head 05:25:01 ehird: I know, and it even works on Linux as of a few months ago 05:25:11 coppro: It should first go to an ultra-low-power mode, solely to suspend to disk, and then it suspends to RAM 05:25:25 the difference is that it turns off immediately, instead of taking a second or two to suspend to anything 05:25:26 but, I don't trust it to work in, say, the middle of a distro upgrade 05:25:36 ais523: you're suspending in the middle of a distro upgrade? 05:25:38 tip: don't 05:25:44 ehird: I had to once 05:25:49 well, not suspending exactly 05:25:58 but the screen permablanked due to a bug 05:26:04 hardware bug, I think 05:26:08 see that? 05:26:14 so in the end, I had to REISUB it, in the middle of the upgrade 05:26:17 that's another issue with the hardware :-P 05:26:22 credit to Debian and possibly Ubuntu that it still worked afterwards 05:26:36 although I had to finish off the upgrade by giving the required apt commands at the command prompt 05:26:41 Alt+SysRq+stuff doesn't always work :( 05:26:51 ehird: the computer was still responding fine 05:27:00 i meant for me 05:27:04 the only reason I did it directly to the kernel is that the GUI would never have let me abort a distro upgrade 05:27:40 specifically, if X isn't recognizing your keyboard and mouse, and it's frozen, then you can't do anything with them 05:28:07 sounds reasonable 05:28:17 is SysRq handled at the X level, or the kernel level? 05:28:17 annoying, though 05:28:21 ais523: kernel, of course 05:28:22 kernel 05:28:27 but X steals the keyboard, I guess 05:28:31 yeah 05:28:32 and then fails to do anything with it 05:28:45 SysRq was originally invented so that brand-new multitasking operating systems had some way to tell them to switch processes 05:28:56 so that none of the other keys would conflict with existing programs 05:29:07 but then, people didn't use it, and invented alt-tab and so on instead 05:29:16 (and control-alt-delete, of course) 05:29:46 and to think control-alt-delete was originally security through obscurity (you could explain it to someone over the phone, but there's no chance they'd think of it by themselves, was the idea) 05:32:40 Doo doo. 05:32:44 ais523: haha, really? 05:32:54 what was it for at the time? 05:33:04 ehird: rebooting 05:33:09 but people weren't supposed to know that 05:33:12 why? 05:33:22 to keep techies employed? 05:33:25 possibly 05:33:30 and because they'd lose their data if they tried it 05:33:43 and companies nearly always misjudge the technical competence of their users one way or another 05:36:00 /me wonders why the inscrutable alt+unicode from windows and opt+{semi-random key/prefix} [key] caught on and compose prefix key didn't 05:36:18 I mean, Compose ' e makes more sense than Option-e e. 05:36:27 emacs uses C-x 8 for compose 05:36:32 which is easy and predictable, yet a pain to type 05:36:37 I mean, once you know about the C-x 8 05:36:49 and, interestingly, Word has its own compose mechanisms 05:37:00 e.g. Control-' e is é 05:37:06 hmm... I wonder what you use for ˝ with standard compose 05:37:10 despite control-alt-e being e in most of the rest of Windows 05:37:20 it's Option+j o to make ő here 05:37:34 but I don't know how X11 Compose analogises double characters 05:37:39 * ehird peeks at the file 05:38:12 ais523: what's ˝ called? 05:38:26 is that the hungarian double-acute? 05:38:29 or a tilde? 05:38:32 former 05:38:34 they look the same at this font size 05:38:39 and I don't know 05:38:40 the erdos accent 05:39:49 double-acute, apparently 05:39:49 :P 05:40:05 : "Ő" U0150 # LATIN CAPITAL LETTER O WITH DOUBLE ACUTE 05:40:12 so =, which makes no sense 05:40:18 still, compose is never any _worse_ than option 05:40:57 and Compose / o makes far more sense than devoting Option+o to ø 05:41:08 plus, Option+q is œ 05:41:10 I mean, what? 05:41:18 Compose o e only makes about, oh, ten times more sense 05:42:41 * ehird uses proper characters 05:43:04 ⌥q is œ, ⎄ o e is œ 05:43:11 I'm not sure how to type a composed oe with this keyboard mapping at all 05:43:17 although æ is altgr-a 05:43:27 basically the OS X system, then 05:43:35 which is better than windows but worse than compose 05:43:50 altgr-punctuation mostly composes the punctuation 05:44:04 although, with a random and weird key assignment, rather than the logical one 05:44:12 e.g. altgr-; composes a ' 05:44:47 heh 05:45:02 é = ⌥e e 05:45:16 because ´'s mostly used in é in english, i guess 05:45:55 and acute's the most common accent for an e altogether, I think 05:46:05 obviously 05:46:08 unless 05:46:08 unless 05:46:13 preëmptive 05:46:16 WHAT NOW BITCH :| 05:46:45 because hardly anyone writes it like that 05:46:51 :( 05:47:18 even though it's correct 05:47:42 I remember that a while back, I was talking how it's almost impossible to connect to most IRC networks without deviating from the relevant standards 05:47:45 for one thing, you'd use the wrong port 05:48:05 de jure standards are useless 05:48:20 irc rfcs spec a protocol that happens to have the same name :) 05:49:42 http://www.reddit.com/r/IAmA/comments/9kjaf/i_am_an_exextreme_programmer_i_did_it_for_6_years/ ;; weirdest definition of extreme programming ever 05:49:50 gahahahahahaha 05:54:22 -!- notostraca has joined. 05:54:28 -!- GregorR has changed nick to Gregor. 05:55:06 -!- notostraca has quit (Client Quit). 05:55:54 * Gregor now possesses +1 NICK OF AWESOME 05:56:03 you got it dropped? 05:56:14 -!- ehird has changed nick to gr. 05:56:18 Indeedidoo. 05:56:23 I AM GREGOR R 05:56:34 No, I'm Spartacus. 05:56:35 Last seen 11 weeks ago! 05:56:42 My surname is not Hurd. 05:56:44 Wonder if that's droppable. 05:56:52 So I can extort Gregor for it. 05:56:54 -!- gr has changed nick to e. 05:57:08 I don't want "gr" 05:57:09 Las seen 4 weeks ago! 05:57:11 -!- e has changed nick to ehird. 05:57:11 I'm quite happy with "Gregor" 05:57:12 *last 05:57:15 One day I shall have the nick e. 05:57:29 *Last, actually 06:03:54 http://imgur.com/6NO0J.png 06:03:58 * ehird disappears in a puff of logic 06:18:35 Mehhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhh 06:53:52 hmm... there isn't much good CSS support in text mode browsers 06:54:04 even elinks fails at any sort of layout or indeed anything beyond colours 07:02:17 I haven't seen anything like Erlang's lists-are-magically-strings in another language <-- char* in C? Also yeah it is just about how io:format() prints it given the ~w or ~p format. You can force print it as a list or as a string in either case 07:02:35 char * isn't a string and it doesn't display as a string. 07:02:46 You can write it out as bytes if you call functions specifically for that purpose. 07:02:56 But it is not "a string", and there is no stringification in C. 07:06:02 ehird, indeed. This is same. io:format's ~p format writes it out as a string if it seems to make sense 07:06:12 No. This is not the same hting at all. 07:06:12 and ~p is what the REPL uses 07:06:14 *thing 07:06:22 char* doesn't intrinsically display as anything 07:06:25 ehird, it is still a list internally, like char* idoesn't change 07:06:27 One is the inspected, human-readable output of a REPL, one is dumping bytes. 07:06:38 you can use %s to print it string-style, but ofc that's not the only way to print it 07:06:44 Comparing them is either purposeful folly or astounding denseness. 07:06:45 ais523, well, there is no "intrinsically" here either, except using the format ~p (or ~w) 07:06:55 which tries to autoformat them in a nice way 07:06:55 Yes, which is THE STANDARD WAY TO INSPECT AN OBJECT. 07:07:05 And is USED by the REPL. 07:07:16 of course you can use ~s to force one to be printed as a string for example 07:07:36 a REPL printing ASCII-like things as strings might be useful; I know gdb always tries to print char* as a string, for instance 07:07:41 escaping nonprintables 07:07:52 not when those things are regular integer lists 07:07:55 ehird, what about char* not being stringy hm, check how gdb shows them 07:07:59 which is the standard way there 07:08:01 or at least indistinguishable from them 07:08:03 it shows it as a string 07:08:08 with lots of escapes in 07:08:17 (if it isn't "string-y" 07:08:19 ) 07:08:25 AnMaster: I'm really not interested in talking to you about this, any more than I am to a brick wall 07:08:45 gprolog's weird, in that 'this is a string' and "this is an integer list taking the ASCII values of the relevant characters" 07:08:49 and they're two different concepts 07:08:56 ais523, ok THAT is weird 07:09:06 That's not weird at all. 07:09:11 'foo' is not a string. 07:09:13 It is an atom. 07:09:14 there's an intlist-string casting function 07:09:16 Like Lisp's 'foo. 07:09:23 well, atoms and strings are indistinguishable in Prolog 07:09:24 "foo" is a string. 07:09:28 Yes, I know the terms differ, 07:09:38 whereas an intlist is something quite different 07:09:39 but I'm using standard terms to clarify what they *are*. 07:09:43 fair enough 07:09:49 most of the string-manipulation functions work on atoms 07:10:00 although, gprolog has a limit to the number of atoms a program can use 07:10:06 so I suppose you need lists for TCness 07:10:41 Well, old Lisps used |ATOMS TO PRINT OUT TEXT TOO|. 07:11:13 * ehird 's mind is bludgeoned with the most awesome idea ever 07:11:14 ais523, anyway the point is that erlang's ~p format (which is used by REPL indeed) will try to print anything in a nice way. This means it will even try to indent nicely if it goes multi-line for a tuple or such. And it tries to print lists that seems string-like (only printable chars and newline, plus possibly some other ones) as strings, while if it is not string like, it prints it like a [1,2,3] st 07:11:14 yle list. 07:11:20 ais523: At least SWI-Prolog mixes in a third thing, which is a separate "string" data type; you can tell it to read doublequotes as strings instead of lists of integers, or to read `foo` as string, and there's atom_to_string and such what you'd expect. 07:11:20 It's a decentralised purely functional package manager. 07:11:26 actually, come to think of it, intlists don't need string-manipulation primitives 07:11:27 This must be what a hangover feels like. 07:11:30 Except intellectual. 07:11:33 ehird: ok, that is a great idea 07:11:53 fizzie: most versions of Prolog let you customize what the various quoting styles mean 07:11:54 ehird, nice idea 07:11:58 so my guess is, it isn't particularly standardised 07:12:11 and they're all doing it like that for compatibility 07:12:24 Latest packages direct from the source! A web of trust! Perfectly trustable rollbacks, upgrades! Great dependency handling! Automatic upgrades (with bad upgrades flagged by the web of trust, I guess)! 07:12:51 you should be able to live-run updates without upgrading 07:12:59 as in, run updates without installing them, live-CD style 07:13:00 ais523: Sure, that's easy. 07:13:04 yep 07:13:13 ehird, web of trust? Hm... 07:13:19 interesting idea 07:13:38 I'm tempted to add the capability-security-hacked-up-by-using-a-separate-account-for-each-program, but that'd be (a) slow for short-running command line programs and (b) a bit too hacky and revolutionary to work without a bunch of futzing. 07:13:55 ehird, a bit hard for someone knowing no one using the system to get into though 07:14:07 Still, the unmodified idea on its own is almost enough to turn into zzo and make my own distribution for. 07:14:29 AnMaster: Yeah, that's true; and I just realised that relying on it to flag good/bad updates is trying to make trustability technological, which as SSL demonstrates doesn't work. 07:14:33 And speaking of strings, still; Matlab does strings as vectors of small integers. But it does use the "char" element-data-type for those vectors, and then determines whether to print as string or as vector based on that. 07:14:45 You can sneak a hole into Debian anyway, so it's not like it's any worse. 07:15:06 and Windows/OS X users rarely upgrade to find a hole, at least I've never really heard of it happening to decent sotware 07:15:14 (well, hole = malicious program) 07:15:34 ehird, also it would be quite a pain to maintain for the distro maintainer, it seems likely that for large such webs of trust the trusted upgrades may diverge, and take a long time to reach everyone 07:15:48 ehird: and there was that attempt to sneak a trojan into the kernel which was noticed 07:15:50 The distro maintainer wouldn't do packages. 07:15:52 generally 07:15:54 -!- FireFly has joined. 07:15:59 which would be kind of, err..., bad for critical security upgrades 07:16:05