00:07:53 * Sgeo tries Avidemux
00:08:05 * Sgeo needs a video editor, and a friend also needs a video editor
00:14:35 -!- Oranjer1 has changed nick to Oranjer.
00:29:49 <lament> anyone tried composing trance?
00:30:27 <ehiird> lament: first, learn how to compose a bar of trance
00:30:34 <ehiird> second, congratulations!
00:31:12 <ehiird> The general electronic music and music-in-general tools, I would imagine.
00:32:54 <Oranjer> well how does trance make trance?
00:33:05 <lament> ehiird: see, the reason you're not helpful is because you actually have no clue about this
00:33:21 <lament> which is why i'm not sure why you're even talking, and which is why i asked if anyone tried composing trance
00:33:42 <ehiird> I've never seen a genre-specific electronic music composition tool, however.
00:33:47 <ehiird> So I would be surprised if trance had one.
00:33:57 <Oranjer> well, it makes sense for those to exist
00:34:49 <Sgeo> http://normish.org/sgeo/quotes.txt
00:34:51 <HackEgo> 63|<fizzie> The thing is just to exist
00:35:42 <Sgeo> At least it reacted in the same minute
00:38:04 -!- sierinjs has quit (Read error: 110 (Connection timed out)).
00:41:08 <fizzie> I've been fiddling a bit with a data-driven trance-composing approach (some NMF sound source separation and a bit of clustering and a bit of grammar induction), but there are no real results yet, and that's probably not exactly what you meant.
00:47:29 -!- oerjan has joined.
00:48:22 <oerjan> <AnMaster> argh where is oerjan when you need him
00:58:49 <oerjan> deep in the atmosphere
00:59:21 <Oranjer> deep in the antipodean atmo?
01:02:02 <oerjan> well, everywhere on earth is the antipode of somewhere
01:03:00 <Oranjer> but we're speaking of deep according to the base at which we begin measuring the depth or shallowness of a statement or thought
01:05:00 <Oranjer> okay, then I shall declare I am
01:06:39 <oerjan> Praeterea censeo ergo sum
01:10:01 -!- Gracenotes has quit (Connection timed out).
01:11:53 -!- Gracenotes has joined.
01:23:20 <ehiird> I need to decide 32-bit vs 64-bit pretty fast.
01:24:45 <ehiird> Oranjer: But that leads to higher memory usage and more legacy crap. Plus it uses PAE, which is a dirty hack, to access >4 GiB of RAM (my only reason not to use 32-bit); if I'm in for such a hack, I can just use PAE in 32-bit mode.
01:24:59 <oerjan> 48-bit gets so little love
01:25:06 <ehiird> OTOH, it's more "native" on modern processors, and it does remove a little bit of the cruft. It also has more registers.
01:31:06 -!- Pthing has joined.
01:31:27 <SimonRC> Oh wow all these new debian packages I keep getting.
01:31:36 * SimonRC does an impression of a package from the "science" section:
01:31:41 <ehiird> Debian: the OS to use if you loooooooooooooooooooooooove new packages!
01:31:41 <SimonRC> libsmbxfrg - Redaltion of cross-FRGs to identify hectodyre sub-bindings
01:31:47 <SimonRC> SMBxFRG is a library for the hyperboline redaltion of semi-morphic-biased cross-FRGs. The hyperboline method is a AFL-based alternative to the more common centrate methodology of retrograde semi-gerentic FRG analysis, providing superior handling of candid-sequence ITF markov meshes, and capable of detecting beta-cyclomatic metrisation quotendates within IETA-chordic sub-hectodyres of the first, second, or fourth orders. The fixed points of the FRG-rectifica
01:31:54 <SimonRC> This package contains the Python bindings to the library.
01:32:02 <ehiird> SimonRC: But what about the flux capacitors?
01:32:27 <ehiird> Seriously, you have to wonder why the Debian folks haven't realised that having one huge monolithic repository isn't the solution and specialist third-party repositories are a good thing...
01:32:41 <SimonRC> I just ignore the subsections I don't care about
01:33:13 <SimonRC> I wish I had somewhere to put this sort of thing
01:33:35 <SimonRC> but I come up with it so rarely that it's not worth it
01:33:35 <ehiird> You think the Debian developers would listen?
01:33:59 <ehiird> Oh, wait — that package isn't real?
01:34:01 <SimonRC> I don't mind the current situation
01:34:27 <ehiird> It got cut off at "FRG-retifica", btw.
01:34:27 <SimonRC> although there are some packages that sound like that
01:34:46 <SimonRC> "FRG-rectification limits may be expressed as any catroliner polynomial of a JK-reductive D-space. Chordates may be loaded from standard HerACleS format files, version 2.2 or 3.0-3.5."
01:34:50 <SimonRC> "This package contains the Python bindings to the library."
01:35:17 * ehiird googles for PAE criticism
01:35:28 <SimonRC> maybe the line "* SimonRC does an impression of a package from the "science" section:" was a bit ambiguous
01:35:40 <ehiird> I thought you were just using an odd use of the word imppresrsion.
01:37:15 <ehiird> "PAE is a hack to get the OS to address 36 bits of physical memory address space, in order to overcome the 4GB limitation of 32-bit OS."
01:37:15 <ehiird> What a weak argument, considering that 64-bit OSs do it via PAE too.
01:37:22 <SimonRC> hm, "minbif" looks interesting
01:37:49 <ehiird> "IRC-friendly instant messaging client"
01:37:55 <ehiird> Which way? Like bitlbee or the other way?
01:38:27 <ehiird> IM is an inferior interface to IRC; and vice-versa too.
01:38:34 <ehiird> If you really want it, use Bitlbee; at least it's well-known.
01:38:37 <SimonRC> I think it lets you see IM as IRC
01:39:52 <ehiird> http://www.bitlbee.org/img/screenshot-mscchat.png
01:39:52 <ehiird> now THIS is a good use case
01:40:30 <ehiird> the cat in "lucumo: eek" made me lol
01:41:30 <Oranjer> is that doing what it looks like it is doing
01:42:01 <ehiird> Wait — PAE requires application support?
01:42:23 <ehiird> ... on one hand, fuck that shit, on the other hand, bleh!
01:46:57 <ehiird> a.out almost certainly doesn't support 64-bit.
01:59:02 <ehiird> "I don't think that Debian can really compete with Gentoo. Sure it might be okay, but when it comes to dependencies, you probably are still going to have to get them all on your own. Or is there something like portage in the Debian world as well?" -- funroll-loops.info
02:04:43 <Sgeo> Even I see the stupidity in that
02:08:06 <SimonRC> or Gren out of that book I forgot the name of
02:08:56 <Sgeo> I.. cannot believe I know that
02:10:44 <Sgeo> BTW, when I said Grendel before, I was referring to something else
02:11:24 -!- Kalagar has joined.
02:32:35 -!- adam_d has quit (Read error: 113 (No route to host)).
02:53:13 -!- Oranjer has left (?).
02:59:32 -!- FireFly has quit ("Later").
03:08:20 -!- BeholdMyGlory has quit (Remote closed the connection).
03:29:20 * Warrigal goes off to temporarily disable the connections between his brain and his plexuses.
03:35:51 -!- Kalagar has quit.
04:03:01 <ehiird> wtf http://en.wikipedia.org/wiki/Safe_sex_makespan
04:24:33 -!- Oranjer has joined.
04:27:37 -!- Jaykul has joined.
04:28:09 <Jaykul> Heh, have you guys seen this guy looking for an example of a "real world" program written in an esoteric language? http://stackoverflow.com/questions/1508581/is-there-any-practical-use-for-an-esoteric-language
04:29:55 <Sgeo> Mental benefits, I'd imagine
04:30:03 <Sgeo> Is a practical use
04:31:00 <Oranjer> it also provides *more* examples as to what a programming language *could* be, so you can then decide what it *should* be
04:31:13 <Jaykul> Well, despite the headline, he's not really looking for a practical esoteric language (that's an oxymoron), but rather for an example of someone writing a practical, real world app in an esoteric language (regardless of whether it was a good idea to use that language ;) )
04:31:35 <Oranjer> has anyone made a game in an esoteric language?
04:31:40 <Oranjer> or a game making program in one?
04:31:44 <pikhq> Oranjer: Yes; it's called Lost Kingdom.
04:32:03 <pikhq> Also, there's an IRC bot in Befunge.
04:33:10 <pikhq> http://zem.fi/~fis/fungot.b98.txt There's the source.
04:59:18 <augur> let me tell you something interesting about negation in ASL :x
05:00:17 <Oranjer> no seriously, now I want to know
05:00:24 <Oranjer> now that google told me what ASL is
05:00:46 * pikhq probably knows it already.
05:01:15 <Oranjer> I would have never guessed
05:01:17 <augur> ASL has two ways of negating sentences
05:01:33 <augur> one is with an overt negation word
05:01:39 <augur> lets just call it not
05:01:49 <Oranjer> let's not and say we...oh, okay
05:02:41 <augur> and the other is with another word, call it NEG which has no overt realization
05:03:15 <augur> the word is there, but the way to "say" it is by not saying anything
05:03:42 <Oranjer> so how does the errr...reader realize it is being negated?
05:03:45 <augur> in both cases of negation, you have to use a THIRD "word", call it a scope indicator
05:03:57 <augur> that that does nothing more that indicate the scope of the negation
05:04:09 <augur> lets denote those by []'s
05:05:24 <augur> so an example of this negation stuff might be like so:
05:06:14 <augur> John [not] buy house == John is not buying a house
05:06:27 <augur> John [not buy house] == John is not buying a house
05:06:34 <augur> both of these are the same in meaning
05:07:21 <augur> but when you use the NEG word
05:07:28 <augur> you cannot get this:
05:07:32 <augur> John [NEG] buy house
05:07:39 <augur> John [NEG buy house]
05:08:10 <augur> where the negation scope marking extends to a small amount of time BEFORE the words "buy house"
05:08:22 <augur> as if NEG really is there taking up space but lacking in expression
05:10:10 <augur> so, interestingly, when there is an overt negation marker, you can omit the full scope marking, and by default it means it ranges over whatever its directly sister to
05:10:42 <augur> if you look at WH questions you get similar things
05:11:16 <augur> this one its a little trickier because you have two question morphemes, but neither of them are overt
05:11:51 <augur> however one of them causes overt effects, so lets call them Q for the one with no over effects, and q for the one with overt effects
05:12:30 <augur> where q forces the question word "who" to move to the end of the sentence
05:12:37 <augur> sentences that arent yes/no
05:13:00 <augur> so you also have a WH-question scope marker
05:13:35 <augur> {hate john q who}, as well as, hate john {q who}
05:13:43 <augur> but with the non-overt Q, you only get
05:13:59 <augur> you never get, who hate john {Q}
05:14:09 <augur> ir {who} hate john
05:17:15 <augur> you get this sort of thing all over the place with other phenomena too
05:17:51 <augur> also, get this, instead of using simple pronouns, because ASL is spatial, you just point
05:18:22 <augur> when you're talking about other people who arent there in the conversation, nor easilly pointed at
05:18:29 <augur> you just point in any random direction
05:18:40 <augur> and when you want to unambiguously refer to that person again, you point in the same direction
05:19:17 <augur> so you can unambiguously keep track of multiple unnamed participants by associating them with particular locations in the space around the conversers.
05:22:13 <Oranjer> sucks, though, if a conversation's referenced people gets so crowded that you screw up a reference
05:22:40 <augur> its a really need phenomena tho man
05:27:39 <Oranjer> what do you think of taking advantage of knowing Conway's Law and making a business that makes it's decisions using Consensus Decision Making?
05:27:57 <augur> whats conway's law
05:28:32 <Oranjer> an organization will almost certainly produce systems/products that match the internal structure of the organization
05:30:11 <Oranjer> http://en.wikipedia.org/wiki/Conway%27s_Law
05:30:44 <augur> <augur> i dont know.
05:31:22 <Oranjer> what do you think of me taking those exams for credits, instead of taking the courses?
05:31:46 -!- fax has quit ("Leaving").
05:32:17 <Oranjer> like, instead of taking the basic courses in college, I just take an exam that gives me the course's credit if I pass
05:33:46 <augur> stop being irritating.
05:34:13 <Oranjer> I want your opinion on taking those exams
05:35:19 <augur> im not a prophet or a fortune teller or anything!
05:37:21 <MizardX> Oranjer: Isn't the point that you learn as much as possible? If you already know the content of one cource, you might as well learn something else instead.
05:38:00 <MizardX> The credits are just to make sure you have learnt enough at the end
05:38:09 <Oranjer> so you're saying I should just take the tests?
05:38:41 <MizardX> No. I'm saying that you should study for some other cource, one that you don't already know.
05:39:35 <Oranjer> well, I'm talking about taking the tests for courses whose credits are required, *and* that I already know
05:40:06 <Oranjer> as in, either I take the course, or I take the test, I can't opt out
05:40:14 <MizardX> If you think you can pass the exam, you might as well take it.
05:40:45 <Oranjer> but I was concerned--is it seen as less of an accomplishment?
05:41:08 <MizardX> Does that matter? You get the credits.
05:41:29 <Oranjer> very well, I shall do every such exam I can find! huzzah!
05:42:40 <augur> MizardX is probably younger than you.
05:43:02 <augur> wait what someone my age?
05:43:09 -!- zzo38 has joined.
05:43:43 <zzo38> I asked on #bochs already, but I'm asking here too because more people are on here, in case anyone knows anything about Bochs or about operating systems or about MBR:
05:43:48 <zzo38> I created a disk image, and then I put at address $0: [B8 00 B8 8E C0 26 C6 06 FF 01 9A EB FE] and at address $1FE: [55 AA]. Why won't it load?
05:44:52 <zzo38> It just displays the VGA BIOS screen and then won't do anything.
05:44:57 <augur> algebraic datastructures
05:45:46 <Oranjer> what are these addresses zzo38?
05:46:51 <zzo38> These are the addresses in the file called c.img
05:47:15 <Oranjer> how uh how can you have addresses inside a file
05:47:30 <zzo38> I mean, the address of where the data is stored in the file. $0 is the beginning of the file
05:48:32 <MizardX> B8 00 B8 8E C0 26 C6 06 FF 01 9A EB FE is the code section of the MBR. 55 AA is the signature bytes.
05:49:04 <zzo38> Yes, at least that's what it is supposed to be, am doing something wrong?
05:49:12 <zzo38> It won't even load it. (I checked)
05:49:39 <zzo38> After I can get it to load, then I can see if the code is correct.
05:49:50 <zzo38> But, first, how do I load it?
05:55:04 <zzo38> No, that isn't how.
05:55:21 <augur> http://www.sciencedaily.com/releases/2009/10/091026125401.htm
05:55:28 <MizardX> add DWord ptr DS:[edx+0000FEEB],ebx
05:56:20 <Oranjer> sleep deprivation only does one thing to me, and that's make me sweat
05:56:41 <augur> it makes me sweat a bit too
05:56:48 <augur> a cold grimey sweat
05:57:06 <Oranjer> I can't feel it, /but I know it's there/
05:58:42 <augur> i wish i knew some organic chemistry
05:58:49 <augur> so i could experiment with this stuff
05:59:18 <augur> or start an illegal underground H+ organization x3
06:00:07 <zzo38> MizardX: It is supposed to be a 16-bit code however, because it is real-mode, isn't it supposed to?
06:00:30 <Oranjer> we can always start an invisible school
06:00:55 <augur> theres a freeschool in baltimore
06:01:26 <Oranjer> there's a reason you aren't taking organic chemistry there, isn't there?
06:02:19 <augur> its a school run by anarchists
06:15:32 -!- Jaykul has changed nick to Jaykul[AFK].
06:48:15 -!- Oranjer has left (?).
06:50:55 <zzo38> I have tried other disk images too, nothing works, I just get the BIOS screen and then it just does nothing after that.
06:57:47 <zzo38> I think I fixed it. It needs at least 2M memory allocated to work, I was allocating 1M
06:57:57 <zzo38> Now I will try again.
06:59:27 <zzo38> Why doesn't the documentation say you need at least 2M allocated?
06:59:40 -!- zzo38 has quit (Remote closed the connection).
07:10:43 -!- puzzlet_ has quit (Read error: 104 (Connection reset by peer)).
07:10:51 -!- puzzlet has joined.
07:14:04 -!- Sgeo has quit (Read error: 110 (Connection timed out)).
07:17:28 -!- bsmntbombdood_ has quit ("leaving").
07:18:33 -!- ehiird has quit (Remote closed the connection).
07:18:52 -!- ehird has joined.
07:19:01 <ehird> You know what JFS is?
07:19:11 <augur> Javascript File System! :o
07:19:36 <ehird> bsmntbombdood: okay
07:20:02 <oerjan> bsmntbombdood: are you as think as you drunk you are?
07:20:46 <bsmntbombdood> i know that's backwards but i read it perfectly the first time :D
07:22:37 <oerjan> the fact that your spelling is still perfect is suspicious. but maybe you have a spell checker.
07:24:48 <bsmntbombdood> i bet you've never licked something with a air dish number of 4
07:25:12 <oerjan> what the heck is an air dish number
07:27:37 <ehird> bsmntbombdood: let's get married!
07:27:53 <ehird> okay we're married now
07:27:55 <ehird> bsmntbombdood: i want a divorce!
07:29:11 <ehird> ejaculatory alimony.
07:45:37 <oerjan> yes. one might say that.
07:47:08 -!- puzzlet has quit (Remote closed the connection).
07:47:13 -!- puzzlet has joined.
07:56:51 <oerjan> and with internet access to boot
07:59:59 -!- clog has quit (ended).
08:00:00 -!- clog has joined.
08:00:20 <oerjan> nothing. civilization is an illusion.
08:00:38 <oerjan> also, i should probably go to bed soon.
08:03:04 <oerjan> that's because it's a secret. i should probably not have told you.
08:03:47 <ehird> bsmntbombdood: mix the tea, the bagel, irish cream, and some alcohol-based beverage together
08:04:09 <oerjan> argh. that is _definitely_ not civilized.
08:17:33 <ehird> bsmntbombdood: so what
08:17:40 <ehird> you can never have too much alcohol!
08:37:12 -!- ehird has quit.
09:18:15 -!- adam_d has joined.
09:47:00 -!- oerjan has quit ("leaving").
10:25:11 <AnMaster> 08:00:38 <oerjan> also, i should probably go to bed soon.
10:36:37 -!- MizardX has quit ("restart").
10:42:20 -!- MizardX has joined.
12:13:22 <fizzie> How old-fashioned; it's been out almost a day or so already.
12:19:38 -!- FireFly has joined.
12:40:01 -!- rodgort has quit (Client Quit).
12:40:11 -!- rodgort has joined.
13:09:57 -!- zzo38 has joined.
13:12:46 <zzo38> I should try to assign this power to my character in D&D game, please read: http://zzo38computer.cjb.net/dnd/other_stuff/my_rule_1.txt
13:29:12 -!- zzo38 has quit (Remote closed the connection).
13:31:05 -!- BeholdMyGlory has joined.
13:42:56 <oklopol> sleep deprivation makes me sweet
11:47:48 -!- clog has joined.
11:47:48 -!- clog_ has joined.
14:02:36 -!- clog has quit (Read error: 110 (Connection timed out)).
14:08:43 -!- sierinjs has joined.
14:09:41 <sierinjs> if the pointer of block in brainfuck is 0 and a < occours inside [], then it exits from the [], right?
14:14:09 <oklopol> if the IP encounters a < in the prorgram, the MP moves one step left on the tape
14:15:30 <AnMaster> "GRUB 2 is the default boot loader for new installations with Ubuntu 9.10 RC, replacing the previous GRUB "Legacy" boot loader."
14:15:39 <AnMaster> though "Existing systems will not be upgraded to GRUB 2 at this time, as automatically reinstalling the boot loader is an inherently risky operation."
14:25:41 <sierinjs> oklopol: mkaaaay, but what's IP/MP?
14:26:04 <oklopol> instruction pointer, memory pointer
14:28:08 <sierinjs> yeah, but if < happens within [] and MP is 0?
14:29:47 <oklopol> there's no what if, there are no exceptions
14:30:30 <sierinjs> erm, m'okay, but that doesn't explain much
14:32:09 <oklopol> by MP = 0 do you mean you're at the leftmost cell? that's implementation dependent
14:32:32 <oklopol> but that has nothing to do with whether you're in a loop so i assumed you meant the value of the cell and not the pointer
14:33:14 <sierinjs> i'm making my own brainfuck interpreter ^_^
15:00:58 -!- fax has joined.
15:11:03 <sierinjs> are there any other conditions when interpreter exits from [] than when MP = 0?
15:40:55 <fizzie> AnMaster: Debian did the GRUB 2 update so that they added a single entry to the GRUB 1 menu to chainload GRUB 2 with; then you could test it out, and if it worked, say "sudo update-from-grub-legacy" to stick GRUB 2 to the MBR. (I guess it's still a bit risky, but better.)
15:55:56 -!- ais523 has joined.
15:56:24 <AnMaster> ais523, http://omploader.org/vMm51YQ bootchart for laptop on jaunty
15:56:46 <AnMaster> oh and I tried karmic in a VM. What the hell is up with gdm? You can no longer change the theme of it easily it seems
15:57:07 * AnMaster ended up googling and doing some stuff from the vt as root to be able to fix it
15:59:12 <AnMaster> this seems so very unlike ubuntu
15:59:31 <AnMaster> and horrible looking splash screen when you log in
16:02:13 <ais523> yes, the theme seems impossible to change altogether
16:02:45 <ais523> I'm slightly annoyed that the login screen lists all the users/usernames, finding the correct username is a speedbump for people trying to use a computer incorrectly
16:02:51 <AnMaster> ais523, see http://www.ubuntumini.com/2009/09/hack-karmics-gdm-login-screen.html
16:02:55 <ais523> although, I'm not sure if you can make that login screen come up remotely
16:03:32 <AnMaster> http://forums.fedoraforum.org/showthread.php?t=205633 (fedora yes, but the approach works for me under ubuntu too):
16:03:36 <AnMaster> sudo gconftool-2 --direct --config-source xml:readwrite:/etc/gconf/gconf.xml.defaults --type bool --set /apps/gdm/simple-greeter/disable_user_list true
16:04:06 <AnMaster> ais523, still that instead requires you to click a button before entering user name
16:04:30 * AnMaster decides to check if you can make it use another login manager, from xfce or maybe kdm
16:05:26 <ais523> AnMaster: I have both gdm and kdm installed
16:05:35 <ais523> Ubuntu handled this by telling kdm to exit immediately upon loading
16:05:44 <ais523> which is kind-of amusing, because upstart has it set to respawn
16:05:54 <ais523> so when the computer loads, gdm loads once, kdm loads about 10 times
16:06:01 <ais523> before upstart notices the loop and stops respawning it
16:06:08 <AnMaster> ais523, this seems rather silly
16:06:30 <ais523> (this reminds me of Ubuntu trying to disable the beep on shutdown by blacklisting the PC speaker kernel module....)
16:06:42 <AnMaster> ais523, btw does karmic fix it properly?
16:06:52 <ais523> AnMaster: the beep doesn't occur, I'm not sure why
16:06:57 <ais523> I think the shutdown sequence is entirely different
16:07:01 <ais523> meaning that the bug doesn't happen
16:07:47 <AnMaster> [ ! -f /etc/X11/default-display-manager -o "$(cat /etc/X11/default-display-manager 2>/dev/null)" = "/usr/sbin/gdm" ]
16:07:54 <AnMaster> weird line to have in the upstart gdm script
16:08:05 <AnMaster> I'm not entirely sure what that is supposed to do
16:08:10 <AnMaster> it seems the result of the test is never used
16:09:22 <fizzie> Is it a shell script or just some "execute commands and stop if a command fails" script?
16:10:22 -!- FireFly has quit (Read error: 60 (Operation timed out)).
16:10:55 <fizzie> It might be the latter, in which case it makes sense.
16:11:31 <fizzie> (I don't really know anything about upstart.)
16:11:32 <AnMaster> synaptic thinks tzdata is deprecated
16:11:40 <AnMaster> as in, not available in karmic any longer
16:11:47 <AnMaster> oh and lots of available packages depends on it
16:11:55 <AnMaster> and apt-get refuses to reinstall it even when told to
16:13:21 <fizzie> That's strange; there's tzdata 2009o-1ubuntu2 in karmic according to packages.ubuntu.com.
16:13:58 <fizzie> I guess I should try out the Koala in the iBook some of these days. Don't really remember what it had installed.
16:14:06 <AnMaster> fizzie, that is the one I have installed
16:14:27 <AnMaster> there are *two* ones listed for me
16:14:58 <AnMaster> fizzie, http://pastebin.ca/1650514
16:15:04 <AnMaster> can you make any sense of that?
16:16:23 -!- FireFly has joined.
16:17:29 <AnMaster> it also says that some packages are unused and suggests using autoremove for them
16:17:47 <AnMaster> openoffice.org-l10n-sv openoffice.org-l10n-en-gb openoffice.org-help-en-gb gimp-help-common openoffice.org-l10n-common
16:17:47 <AnMaster> gimp-help-en gimp-help-sv openoffice.org-help-sv
16:18:57 <fizzie> They have probably changed the dependency types so that those no longer autoinstall. (Maybe. If so, you can always tag those as manually installed.)
16:19:09 <fizzie> I don't really know how to use Synaptic, though, since I always just use aptitude.
16:19:29 <AnMaster> fizzie, apt-get for me unless I don't know how to do it that way... then I use synaptic
16:20:15 <fizzie> Aptitude's better than apt-get at dependency-handling; you can browse resolution suggestions and so on.
16:20:40 <fizzie> How's it going with the Ubuntu Software Center or whatnot? Wasn't that in karmic already?
16:21:08 <AnMaster> fizzie, yeah it is even worse than the old gnome thingy for that
16:21:18 <AnMaster> a lot more clicks to select something for install
16:21:31 <AnMaster> you seem to only be able to install one at a time
16:21:43 <AnMaster> no "select check boxes, then click install
16:23:27 -!- Jaykul[AFK] has changed nick to Jaykul.
17:00:21 <Pthing> !bf [-]+++++++++++>[-]>[-]>>[-]+>>[-]<[-]++++++++++[>+++<-]>++<<<<<<[>[>>>+<+<<-]>>[<<+>>-]<[>>+<+<-]>[<+>-]>>[-]>>[-]>[-]>[-]>[-]>[-]>[-]<<<<<<<[-]>>[-]++++++++++[<<++++>>-]<<++++++++<[>>>+>>+<<<<<-]>>>>>[<<<<<+>>>>>-]<<<<<[>>>>+>+<<<<<-]>>>>>[<<<<<+>>>>>-]<<[>>+>[-]<[>+>+<<-]>>[<<+>>-]<---------->+<[>-<[-]]>[<[-]>[-]<<[-]<---------->>>>+<[-]]<<<<-][-]>>>>[-]<[-]>>[<+<+>>-]<<[>>+<<-]>>[<<<<<<<+>>>>>+>>-]<<[>>+<<-]>>[<<
17:00:21 <Pthing> <<<+>>>+>>-]<<[>>+<<-]<<<[<<.>>[-]]>>>>[<<<<<<->>>>>->-]<[>-<-]<<[<<<+>>>>>+<<-]>>[<<+>>-]<<<<<.>.<<<<<[-]>[<+>-]>>[<<+>>-]<<<<-]
17:10:45 <AnMaster> Pthing, you can easily compress that a bit though
17:10:52 <AnMaster> for instance, the initial [-] isn't required
17:11:06 <Pthing> i just saw it somewhere and wondered what it did
17:11:29 <Pthing> would do, but i'm doing something else
17:15:30 -!- Slereah_ has joined.
17:17:27 <ais523> Pthing: paste it and set EgoBot to the pastebin
17:18:13 <AnMaster> great, after trying to switch to kdm, X refuses to start
17:18:17 <AnMaster> with completely unrelated errors
17:18:24 <AnMaster> yet switching back to gdm fixes it
17:19:41 -!- Slereah has quit (Read error: 60 (Operation timed out)).
17:20:51 <ais523> !bf +++++++++++>>>>+>++++++++++[>+++<-]>++<<<<<<[>[>>>+<+<<-]>>[<<+>>-]<[>>+<+<-]>[<+>-]>>[-]>>[-]>[-]>[-]>[-]>[-]>[-]<<<<<<<[-]>>[-]++++++++++[<<++++>>-]<<++++++++<[>>>+>>+<<<<<-]>>>>>[<<<<<+>>>>>-]<<<<<[>>>>+>+<<<<<-]>>>>>[<<<<<+>>>>>-]<<[>>+>[-]<[>+>+<<-]>>[<<+>>-]<---------->+<[>-<[-]]>[[-]<[-]<[-]<---------->>>>+<[-]]<<<<-]>>>[-]>[-]>[<+<+>>-]<<[>>+<<-]>>[<<<<<<<+>>>>>+>>-]<<[>>+<<-]>>[<<<<-]<<[<<<+>>>>>+<<-]>>[<<+>>-
17:20:53 <ais523> ]<<<<<.>.<<<<<[-]>[<+>-]>>[<<+>>-]<<<<-]
17:21:05 <ais523> I was busy peephole-optimising by han
17:21:24 <AnMaster> ais523, paste it and set EgoBot to the pastebin!
17:22:16 <ais523> !bf http://pastebin.ca/raw/1650570
17:22:30 <ais523> well, that was anticlimatic
17:22:44 <fax> !fungot style
17:23:19 <EgoBot> help: General commands: !help, !info, !bf_txtgen. See also !help languages, !help userinterps. You can get help on some commands by typing !help <command>.
17:23:31 <AnMaster> fax, and you are confusing the bots
17:23:42 <ais523> (fungot isn't here...)
17:23:50 <fax> fungot ^style
17:24:04 <fizzie> And you can run longer programs on fungot with the text-variable-thing.
17:24:29 <fizzie> "IRC read failed."; must've been a netsplit, or break in my connection.
17:24:41 <ais523> !bf ,[.,]!Hello, world!
17:24:56 <ais523> oh, EgoBot doesn't accept that sort of input
17:25:27 <fizzie> And now my hostname is wrong, aw.
17:25:30 <ais523> !bf ++++++++++++++++++++++++++++++++[.+]
17:25:36 <EgoBot> !"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~
17:25:45 <ais523> (confining it to printables will probably produce a more readable output)
17:25:46 -!- fizzie has quit ("jumpin' jumpin'").
17:25:48 -!- fizzie has joined.
17:26:14 -!- fungot has joined.
17:26:51 <fizzie> ^str 0 set +++++++++++>>>>+>++++++++++[>+++<-]>++<<<<<<[>[>>>+<+<<-]>>[<<+>>-]<[>>+<+<-]>[<+>-]>>[-]>>[-]>[-]>[-]>[-]>[-]>[-]<<<<<<<[-]>>[-]++++++++++[<<++++>>-]<<++++++++<[>>>+>>+<<<<<-]>>>>>[<<<<<+>>>>>-]<<<<<[>>>>+>+<<<<<-]>>>>>[<<<<<+>>>>>-]<<[>>+>[-]<[>+>+<<-]>>[<<+>>-]<---------->+<[>-<[-]]>[[-]<[-]<[-]<---------->>>>+<[-]]<<<<-]>>>[-]>[-]>[<+<+>>-]<<[>>+<<-]>>[<<<<<<<+>>>>>+>>-]<<[>>+<<-]>>[<<<<-]
17:26:51 <fungot> Set: +++++++++++>>>>+>++++++++++[>+++<-]>++<<<<<<[>[>>>+<+<<-]>>[<<+>>-]<[>>+<+<-]>[<+>-]>>[-]>>[-]>[-]>[-]>[-]>[-]>[-]<<<<<<<[-]>>[-]++++++++++[<<++++>>-]<<++++++++<[>>>+>>+<<<<<-]>>>>>[<<<<<+>>>>>-]<<<<<[>>>>+>+<<<<<-]>>>>>[<<<<<+>>>>>-]<<[>>+>[-]<[>+>+<<-]>>[<<+>>-]<---------->+<[>-<[-]]>[[-]<[-]<[-]<---------->>>>+<[-]]<<<<-]>>>[-]>[-]>[<+<+>>-]<<[>>+<<-]>>[<<<<<<<+>>>>>+>>-]<<[>>+<<-]>>[<<<<-]
17:27:04 <fizzie> ^str 0 add <<[<<<+>>>>>+<<-]>>[<<+>>]<<<<<.>.<<<<<[-]>[<+>-]>>[<<+>>-]<<<<-]
17:27:20 <ais523> fizzie: I tried running it locally, it used up 100% of my CPU and didn't do anything obvious
17:27:35 <fizzie> Oh, well. That's not so interesting, then.
17:27:52 <fizzie> Anyway, something like that can be used to run longer programs.
17:28:07 <fizzie> (Unless I've broken it.)
17:28:46 -!- Asztal has joined.
17:29:01 <ais523> we need a fingerprint that creates Befunge VMs
17:29:14 <ais523> so that you could safely run arbitrary Befunge-98 without it breaking out and affecting the rest of the program
17:44:15 -!- FireFly has quit ("Later").
17:44:46 -!- fax has quit ("Leaving").
17:54:13 -!- BeholdMyGlory has quit (Read error: 104 (Connection reset by peer)).
18:04:07 -!- Sgeo has joined.
18:39:33 -!- Kalagar has joined.
18:46:58 -!- Oranjer has joined.
18:53:11 -!- oerjan has joined.
19:00:34 <oerjan> iwc = irregular webcomic
19:00:53 <Oranjer> actually, I had forgotten, thanks
19:00:55 <AnMaster> oerjan, seems he forgot to close one <b> tag
19:01:17 <AnMaster> "News: I am away from the Internet from 31 Oct to 3 Nov. I will not be reading or responding to e-mail during this time. The comics should update as normal, but if anything goes wrong, I won't be able to fix it."
19:01:27 <AnMaster> why is it things always break when he is away
19:01:37 <oerjan> um i haven't read it yet
19:02:06 <AnMaster> fizzie, ais523: something is broken with firefox in ubuntu karmic
19:02:17 <AnMaster> ais523, it doesn't use system wide setting for hinting and AA
19:02:26 <AnMaster> much more blurry and slightly subpixelishg
19:03:14 <AnMaster> http://www.ubuntu-inside.me/2009/07/howto-fix-firefox-35s-font-hinting.html
19:07:50 * oerjan didn't realize it was 11:11 GMT before
19:08:55 <oerjan> iwc update time. that "no reason" in the faq seems rather dubious now
19:09:20 <AnMaster> oerjan, it is 11:11 *somewhere* if you publish it at 11 past
19:09:59 <oerjan> yes but it could be a WWI reference
19:10:19 <oerjan> alas, it seems WWI ended at 11:00, not 11:11
19:10:25 <Deewiant> http://en.wikipedia.org/wiki/11:11_%28numerology%29
19:11:21 <ais523> it should have been 11:11:11 1/9
19:12:41 <oerjan> i cannot say i have noticed that particular coincidence. obviously it will now start cropping up all over the place :)
19:14:41 <oerjan> and just now i discovered wikipedia has deleted the page on the Baader-Meinhof phenomenon. Coincidence? I think not!
19:16:00 <AnMaster> oerjan, Baader-Meinhof phenomenon?
19:16:01 <oerjan> also, i am joking. despite believing in synchronicity.
19:16:23 <oerjan> AnMaster: YOU WILL NEVER KNOW NOW, WHAT, WITH IT BEING DELETED
19:16:26 <Deewiant> Well, you're not joking in that it was deleted, some months ago IIRC.
19:16:27 * AnMaster knows what Baader-Meinhof was, but not what the "Baader-Meinhof phenomenon" is/was
19:16:33 <oerjan> unless you use the power of the google
19:18:10 <oerjan> BUT THAT WOULD BE CHEATING
19:18:47 -!- Pthing has quit (Remote closed the connection).
19:19:46 -!- fax has joined.
19:21:31 <AnMaster> ais523, another karmic issue: you know the menu for logout/shutdown and such?
19:21:44 <AnMaster> why is there a status setting thingy at the top of it
19:21:57 <AnMaster> ais523, how do you get rid of it?
19:22:03 <ais523> the options are grayed out if there aren't any running programs that care
19:22:09 <AnMaster> I don't want it to clutter the menu
19:22:09 <ais523> and I don't know; it's just the one menu option
19:22:39 <ais523> there's probably some way to turn it off somewhere
19:22:43 <AnMaster> ais523, also switching to kdm seems like only sane option under karmic
19:22:44 <ais523> and it'll probably be on the Web by now
19:23:19 <AnMaster> ais523, what is the name of the sub menu in English. For some reason that single menu is i18ned to Swedish here...
19:23:43 <AnMaster> ais523, thanks. Oh what is the odd letter icon thingy for
19:24:03 <ais523> what odd letter icon thingy?
19:24:20 <AnMaster> ais523, argh: https://bugs.launchpad.net/ubuntu/+source/indicator-applet/+bug/447964
19:24:34 <AnMaster> ais523, it seems linked to evolution, But it is in the indicator applet thingy
19:24:42 <AnMaster> which iirc is used for other (useful) stuff
19:24:49 <AnMaster> so not sure how to get rid of the evolution icon
19:24:59 <oerjan> AnMaster: maybe it's not a missing <b>. maybe the web page just has an evil beard
19:25:24 <ais523> it's empathy it's linked to, not evolution
19:25:36 <ais523> that means new mail arrived, I think
19:25:49 <ais523> just turn off evolution-alarm-notifier in the services thing
19:27:12 * AnMaster tries to uninstall the relevant packages
19:27:49 <oerjan> evolution-alarm-notifier sounds badass. like, it goes off if somewhere evolves giant man-eating squirrels...
19:28:15 <AnMaster> oerjan, no. It is just a boring enterprisy gorupwareific thing
19:28:28 * oerjan swats AnMaster -----###
19:28:37 <AnMaster> god damn. Even uninstalling didn't help
19:29:37 <AnMaster> mail-notification - mail notification in system tray
19:29:38 <AnMaster> mail-notification-evolution - evolution support for mail notification
19:30:16 <AnMaster> oh, claims that isn't installed
19:31:44 <AnMaster> ah I guess it is the indicator-messages package
19:32:03 <AnMaster> indicator-applet* indicator-applet-session* indicator-messages* indicator-session* ubuntu-desktop*
19:32:06 -!- zzo38 has joined.
19:33:00 <ais523> AnMaster: why do you think I can help?
19:33:26 <zzo38> Do you think this is good, and what level adjustment (if any), and what other stuff should I write on this file? http://zzo38computer.cjb.net/dnd/other_stuff/my_rule_1.txt
19:33:46 <AnMaster> ais523, you used ubuntu longer. But anyway I need somewhere to release my irritation. Speaking with you works.
19:35:46 -!- ehird has joined.
19:36:05 <zzo38> And, how bad is this game, in your opinion: http://www.digitalmzx.net/wiki/index.php?title=Super_ASCII_MZX_Town
19:37:31 <ais523> zzo38: hard to tell from that description
19:37:40 <ais523> (to your second question)
19:38:01 <zzo38> You can try play it, if you want to.
19:38:21 <zzo38> http://zzo38computer.cjb.net/mzx1/ASCMZXTO/ASCMZXTO.ZIP
19:38:29 <zzo38> http://zzo38computer.cjb.net/mzx1/mzx_extended/
19:38:40 <zzo38> That's all you need to run it
19:38:55 <ais523> for your first question, I don't think you'd persuade many people to play that class even with an LA of -19
19:39:24 <zzo38> That is not a class or race, it is a add-on. And I would very much like to apply it to my character if the DM wants to
19:40:04 <zzo38> The DM wants to think about LA too however
19:40:12 <ais523> it just seems like a way to die pretty quickly
19:40:12 -!- Oranjer has left (?).
19:40:18 <zzo38> So if I can get help, we can figure it out. I think -19 is too low
19:40:35 <zzo38> And I think I can survive with this
19:40:48 <zzo38> And even use it to my benefit in strange ways
19:40:57 <ais523> you're probably unlikely to run at anything but LA 0
19:41:02 <ais523> negative LAs freak DMs out
19:41:06 <ais523> and positive would clearly be unfair
19:41:15 <zzo38> The DM actually said LA -20.
19:41:23 <zzo38> However I said that's too low (in my opinion)
19:42:04 <ais523> anyway, I don't see how you're going to get out of the problem that you take damage every hour and have no natural healing
19:42:16 <ais523> you're going to wear the clerics on your team dry
19:42:36 <zzo38> I don't take damage every hour. It happens only in stuff in case of the things listed, in the past hour.
19:42:45 <zzo38> So if I avoid those things, I won't take that damage.
19:42:59 <zzo38> And, we have no clerics in our party, however I have some healing powers
19:43:37 <zzo38> (The ones about items/artifacts are if you have carried/used that item in the past hour, not counting inaccessible items due to transformation)
19:44:20 <zzo38> And, many of these things can be suppressed temporarily due to the use of Permanent Max HP Loss Actions
19:45:39 <zzo38> My DM actually *suggested* negative LA. I disagreed but he insists
19:47:46 <zzo38> However, he won't be in this country today.
19:47:50 <coppro> negative LA seems like a truly terrible idea
19:48:36 <zzo38> However, you could still add up everything and if the total is negative increase it to +0. But I'm not sure how best it is though
19:49:04 <coppro> Who needs decent racial traits when you can cast wish at level 1?
19:49:33 <zzo38> I know, that's why total LA should never be negative
19:51:00 <ais523> coppro: a level 1 wish would require an LA of -16
19:51:04 <ais523> which seems kind-of unlikely
19:51:15 <ais523> I know there was an effort to weaken goblins to the point where LA -1 was balanced, but they didn't manage it
19:51:18 <ehird> ais523: hi hypocrite
19:51:25 <AnMaster> ais523, zzo38 what is this LA? thing?
19:51:32 <ais523> ehird: what are you going to accuse me of hypocriticality about?
19:51:33 <coppro> AnMaster: LA is Level Adjustment
19:51:33 <zzo38> LA=Level Adjustment.
19:51:49 <coppro> it's basically a crutch to allow players to play stronger races by reducing the number of class levels they get
19:51:53 <zzo38> It means that your XP and starting money is calculated due to a different level than the actual one
19:51:57 <ehird> ais523: <ais523> why don't we talk about esolangs more! <ais523> I wish this channel was on topic! <ais523> can we talk about esolangs?
19:52:01 <ehird> ais523: <ais523> blah blah blah DnD or whatever
19:52:05 <ais523> seems a bit unlikely, really
19:52:13 <ais523> also, this channel is zzo38's while he's here
19:52:19 <coppro> (actually, you wouldn't be able to cast wish at level 1 anyway due to XP requirements, but still...)
19:52:22 <zzo38> For example, LA+1 means your XP is calculated due to your HD level + 1
19:53:03 <zzo38> This channel is not mine, if it was it would clearly have a + sign at the beginning of its name instead of #
19:53:26 * coppro loves 3.5e. It's a stellar exapmle of how NOT to design a game
19:53:56 <zzo38> I happen to like 3.5e however there are some things wrong, that I try to fix by writing Icosahedral RPG instead
19:54:34 <ehird> wow, at least i was outright aggressive about it as opposed to passive-aggressive
19:55:11 <ais523> AnMaster: languages are really annoying when you find you need an operator on all objects to see if they're a particular constant or not
19:55:21 <ehird> ais523: i don't think you do</peanut gallery>
19:55:34 <AnMaster> wow I believe X on karmic just crashed when the screen was locked
19:55:36 <ais523> ehird: I can't figure out how to parse that sentence in content
19:55:48 <AnMaster> and now the lock screen option is gone
19:55:51 <ehird> the is-it-a-constant
19:55:58 <ehird> don't think you need it
19:56:17 <AnMaster> ais523, does your karmic has an option in the logout/shutdown menu to lock the screen?
19:56:21 <ais523> ehird: well, the issue is, the only way to find out what an object is is to send messages to it
19:56:22 <ais523> the messages are also options
19:56:28 <ais523> AnMaster: yes, and it works
19:56:41 <ehird> AnMaster: you have to login via gdm
19:56:41 <ais523> AnMaster: well, you are messing with that menu...
19:56:44 <ehird> not start x yourself.
19:56:52 <AnMaster> ais523, no I wasn't. I gave up on it
19:56:52 <ais523> ehird: aha, because it's gdm that handles the locking
19:57:03 <ehird> AnMaster: then tough shit.
19:57:05 <coppro> AnMaster: I've been using karmic for months now and never had anything of the sort happen - and I'm on KDE
19:57:23 <AnMaster> ehird, issue is with gdm in karmic being horrible
19:57:26 <ehird> expecting gnome to integrate with kdm is quite laughable.
19:57:32 <ehird> AnMaster: is it? i've used it and it was absolutely fine.
19:58:09 <ais523> coppro: AnMaster has decided that the new gdm is intolerably bad
19:58:10 <AnMaster> ehird, you can't set a better colour scheme without a log of hacks to begin with (that default brown diarrhoea look is quite ugly!)
19:58:21 <ehird> coppro: anmaster uses ubuntu because it "just works" and then changes everything about it and complains when it braeks.
19:58:22 <AnMaster> sure there are hacks to get rid of it
19:58:22 <ais523> AnMaster: it isn't brown by default, it's greyscale
19:58:31 <zzo38> So, is LA -1 for this file I wrote good enough (my character's LA is already +1 so that simply cancels it out). I don't believe it should be lower.
19:58:32 <AnMaster> you need to click a button before entering user name
19:58:54 <ehird> AnMaster: if that's diarrhoea colour to you, by the way, you have bowel problems
19:58:58 <ehird> i've never had orange diarrhoea.
19:59:17 <ehird> oh, i forgot, your screen makes everything look utterly wrong.
19:59:24 <ehird> right, sorry, it's orange in fact, you see.
19:59:36 <zzo38> Whatever number it is I will add it to the file before printing it out a second time
20:00:33 <AnMaster> ehird, hm you have /usr/share/gdm/themes/HumanList/background.png as the bg image?
20:00:51 <ehird> Show me a screenshot, I'm not booting into Ubuntu.
20:01:11 <AnMaster> ehird, since this is in a vm it will take a sec
20:01:27 <zzo38> And, you can discuss scoring of computer games? (Just any computer games in general, I mean)
20:01:45 <zzo38> Some is I think the scoring is wrong or is OK but could be improved, usually I try to improve it
20:02:08 <ais523> zzo38: make sure there isn't some repetitive action you can take to increase your score arbitrarily high
20:02:13 <ais523> like Death farming in NetHack
20:02:25 <ehird> is that where you farm Death himself
20:03:03 <zzo38> ais523: I always keep rtack of this. However, many games do not.
20:03:09 <ehird> 08:21:08 <AnMaster> fizzie, yeah it is even worse than the old gnome thingy for that
20:03:10 <ehird> 08:21:10 <AnMaster> IMO
20:03:10 <ehird> 08:21:18 <AnMaster> a lot more clicks to select something for install
20:03:10 <ehird> 08:21:23 <AnMaster> or deselect it
20:03:10 <ehird> 08:21:24 <AnMaster> in fact
20:03:11 <ehird> 08:21:31 <AnMaster> you seem to only be able to install one at a time
20:03:12 <ehird> 08:21:36 <AnMaster> or deinstall one at a time
20:03:14 <ehird> 08:21:43 <AnMaster> no "select check boxes, then click install
20:03:17 <ehird> you can go to another package and install it while the other is installing
20:03:19 <ais523> ehird: yes, you kill-equivalent him repeatedly
20:03:19 <ehird> and it is not more clicks
20:03:35 <ais523> as in, actually killing Death permanently doesn't really make sense, but you can repeatedly knock him out
20:03:41 <ehird> type to search, double click the package, click Install.
20:03:46 <zzo38> Occasionally, it is possible (with very minor changes to the game) to fix it by requiring that you have to complete the entire game but with the lowest total score possible, instead of the highest.
20:03:57 <ehird> with the old one: type to search, tick the package, click install, click "Close" at the end
20:04:10 <ehird> it's the same amount
20:04:17 <ehird> the button changes to uninstall
20:04:19 <ehird> instead of install
20:04:37 <AnMaster> ehird, no. I went to category and then only thing was an arrow button that you showed an info screen about the app
20:04:44 <AnMaster> then there the uninstall button was
20:05:00 <AnMaster> clearly more clicks than clicking check boxes in the old category lists
20:05:01 <zzo38> For one thing, I think the ADOM scoring could definitely be improved a lot
20:05:03 <ehird> Which is why you search when you want to uninstall something, because obviously you already know what it is.
20:05:21 <AnMaster> ehird, actually no. I wanted to see "what useless crap is there here to get rid of"
20:06:16 <ehird> how about, if you're going to go against ubuntu's philosophy and be a malcontent about every-fucking-thing they've done, stop using ubuntu
20:06:37 <ehird> you say you want it to "just work" but clearly you're not happy with that considering how long you spend changing everything
20:06:44 <AnMaster> hm that background is yellow-brownish on my desktop
20:06:49 -!- oerjan has quit ("Later").
20:07:11 <zzo38> Yes, you don't have to use Ubuntu, there are various others, you can even write your own if you prefer
20:07:40 <ehird> AnMaster: The word for "yellow-brownish" is "orange".
20:07:50 <AnMaster> ehird, what about the splash screen
20:07:54 <AnMaster> that is brown on my desktop too
20:08:03 <ehird> Define splash screen.
20:08:06 <AnMaster> ehird, this http://omploader.org/vMm55MQ
20:08:23 <AnMaster> (different res, just picked one randomly from that directory)
20:08:23 <ehird> That's very-dim-browny-pinky-purpley.
20:08:34 -!- Kalagar has quit.
20:08:34 <AnMaster> ehird, that is the thing used for the ubuntu splash
20:09:14 <AnMaster> ehird, further, even the orange look of the gdm background is ugly as fuck
20:09:45 <ehird> Nobody gives a fuck what you think about the artwork!
20:09:54 <ehird> If you hate it so much you're going to die oh god, don't fucking use Ubuntu!
20:10:10 <ehird> We get it! Ubuntu is uglier than babies thrown in a blender!
20:10:11 <AnMaster> ehird, of course you can work around it. What about the user list? ais523 didn't like that either
20:10:16 <Sgeo> SCP-413: THE BUILDINGS NEXT DOOR AND ACROSS THE STREET DONT BELIEVE IM A PRODUCTIVE MEMBER OF SOCIETY AND THAT I WILL GIVE AWAY THEIR PLANS
20:10:16 <Sgeo> Dr. ███████: What plans?
20:10:16 <Sgeo> SCP-413: THAT WE ATTACK TOMRROW
20:10:36 <ehird> AnMaster: why do you continually use the argument "AIS523 DID IT TOO EXCEPT 10X LESS ANNOYINGLY"?
20:10:45 <ehird> it's highly unconvincing, even if i accept its blatant appeal to authority
20:10:47 <ais523> ok, this argument about the gdm background is confusing me
20:10:55 <ais523> AnMaster says it's orange, ehird says it's purple, I say it's grey
20:11:05 <AnMaster> ais523, it is not gray. No way.
20:11:06 <ehird> he says it's brown
20:11:15 <ais523> same colour, technically speaking
20:11:21 <ehird> shut up, both of you, you both use crappy TN laptop displays
20:11:28 <ehird> I know what colour it is :-P
20:11:31 <ehird> the middle bit is browny
20:11:35 <ehird> then it doees to crimsony at the sides
20:11:42 <ehird> and there's sort of a hint of purple in the halo of the bottom bit
20:12:15 <AnMaster> and the gray scale thing is the shutdown thingy
20:13:02 <AnMaster> ais523, anyway. The login list can as I mentioned be forced off by gconf stuff. But then you get a button to click before you can enter user name
20:13:25 * AnMaster assumes ehird will find is silly that anyone wanted to hide user list
20:13:53 <ais523> I don't, the user-list sort-of assumes people will primarily use the mouse
20:14:04 <ehird> you can type your password
20:14:15 <ais523> umm, there's more than one user on the list?
20:14:15 <AnMaster> ehird, I want to have to type user name too.
20:14:24 <ehird> ais523: arrow keys?
20:14:29 <ais523> and IIRC, there's no obvious way to tell which user is selected
20:15:44 <zzo38> Ubuntu is all many different kind of problems, I have to write my own different distribution instead for better
20:15:54 <ais523> ah, got it, it's in reverse video
20:15:58 -!- coppro has quit (Remote closed the connection).
20:16:03 <ais523> next issue: I've used computers so much that I don't even /notice/ reverse video...
20:16:28 <AnMaster> ais523, what would cause the reverse video? screen tilt?
20:16:29 -!- coppro has joined.
20:17:08 <ais523> AnMaster: repeatedly switching between black-background and white-background programs and/or websites
20:17:10 <zzo38> Ubuntu has been called "Windows Linux Edition" in some cases
20:17:31 <AnMaster> ais523, what exactly do you mean by reverse video here?
20:17:54 <AnMaster> ais523, or you mean it does that for selection?
20:17:57 <ais523> AnMaster: black on white rather than white on black
20:18:03 <ais523> zzo38: have you ever seen Linspire?
20:18:10 <ais523> and yes, it does that for selection
20:18:28 <zzo38> I have seen Linspire too, but FreeGeek uses Ubuntu
20:19:27 <AnMaster> ehird, what do you have against a "enter username" text input?
20:19:34 <ehird> Nothing, and I never said I did
20:19:41 <AnMaster> ehird, you acted like you did above
20:20:05 <ehird> I have something about you being a whiny bitch for hours because you chose a distro that you said you wanted to just work and complain when you fiddle with shit and expect that from everything in direct contradiction
20:21:13 <AnMaster> err "complain when you fiddle with shit and expect that from everything in direct contradiction"
20:21:27 <AnMaster> I complain that the new gdm is much dumbed down
20:21:42 <ais523> I actually really like it, apart from the user list
20:21:55 <ehird> You said you wanted something that "just worked" instead of fiddling about, and that's why you picked Ubuntu. So you installed it and promptly fiddled with everything and continued doing so.
20:21:56 <ais523> things like having dropdowns rather than putting everything behind menus
20:22:14 <ehird> Now you're complaining that you can't keep doing this, thus exposing that no, you really don't just want something that just works.
20:22:20 <AnMaster> ehird, I still want to be able to at least select background and such. ...
20:22:23 <ehird> So stop complaining or switch to something fiddly.
20:22:32 <ehird> Go and whine on the bug tracker, then.
20:22:38 <zzo38> I don't like user list on logon, either. Although currently I don't have Linux, I still turned off the welcome menu and set it to not keep the last username, require CTRL+ALT+DELETE, and a few oter things
20:22:44 <AnMaster> ehird, I subscribed to an open bug about it
20:22:47 <ehird> You complain about me complaining about software here instead of to the software authors; at least I don't repeat it for hours.
20:22:54 <AnMaster> one of about, uh, 500 or so subscribing users
20:22:56 <zzo38> Anyone who has Windows but prefers this way must have done so
20:23:14 <ehird> That's like 90% of Ubuntu users.
20:23:27 <AnMaster> ehird, considering the majority doesn't report bugs or don't have accounts or such.
20:23:30 <ehird> It's not like malcontents are the type to subscribe and users who only have a mild preference wouldn't generally bother.
20:23:48 <AnMaster> "and users who only have a mild preference wouldn't generally bother."
20:23:58 <AnMaster> that is what I said on the line above partly
20:24:37 <AnMaster> ehird, anyway I fail to see why locking screen with kdm would be an issue. locking screen in KDE when using gdm works
20:24:44 <ais523> umm, require control-alt-delete on login? for Linux?
20:24:51 <ehird> zzo38 uses Windows.
20:25:00 <ehird> because he hasn't got around to writing his entirely own distro with his own everything yet.
20:25:43 <AnMaster> yay found a solution for screen locking. xlock
20:25:44 <zzo38> The reason I don't use Linux is because I already have Windows, once I get a new computer or this Windows stops or whatever, I will use Linux next instead
20:25:50 <AnMaster> now to integrate it with stuff
20:26:01 <AnMaster> ais523, that effectively kills X though
20:27:05 <ehird> http://tools.suckless.org/slock
20:27:07 <ehird> For screen locking.
20:27:26 <AnMaster> ehird, like, screen is auto locked when you close lid
20:28:00 <AnMaster> ehird, that is the feature I want of screen locking. Locking for close lid, suspend to disk/ram
20:28:13 <AnMaster> (yeah closing lid *does* suspend to ram the way I set it up
20:28:20 <AnMaster> ehird, and still use kdm not gdm
20:29:05 <ehird> I linked you to a screen locker, I'm not interested in hearing your numerous demands with the implicit and false implication that it cannot do those things.
20:29:31 <AnMaster> ehird, it needs to be hooked up to acpid or such then
20:30:15 <ehird> Incidentally, XLock considered harmful: http://www.jwz.org/xscreensaver/versus-xlock.html
20:30:26 <ehird> Ignore the parts about the screensaver prts, naturally.
20:32:48 <AnMaster> slock is a simple screen locker utility for X,
20:33:40 <ehird> "I can't press all the fun buttons! How on earth will I display the password asterisks in 72-point Impact now! THIS SOFTWARE SUCKS! It doesn't satisfy my control OCD. Add new bloat in it, quick, just so I can change how the bloat operates!"
20:34:09 <AnMaster> just found the disagreement on options a bit funny
20:34:14 <ehird> -v isn't an option
20:34:19 <ehird> it doesn't change how the software works
20:34:23 <ehird> (sorry; I misunderstood your comment.)
20:34:27 <ehird> AnMaster: yes, but it said options, not flags.
20:34:41 <AnMaster> ehird, usually they are the same in man pages
20:34:52 <AnMaster> ehird, and -v does change the way it works
20:35:08 <AnMaster> instead of locking the screen it prints out version and copyright
20:35:16 <ehird> It doesn't change the way it locks the screen; it isn't an option about locking the screen.
20:35:27 <ehird> It's a meta-option, so to speak; it's at the level of the application binary, rather than its inner function.
20:35:29 <AnMaster> ehird, it changes it to *not* lock the screen
20:35:33 <ehird> Admittedly the man page might just be a typo.
20:35:43 <AnMaster> ehird, the man page was made by debian btw
20:36:00 <ehird> Oh. Well that'd explain it.
20:36:16 <ehird> I wouldn't think it actually needs a manpage but there youg o.
20:36:48 <AnMaster> ehird, debian is rather fanatic about that
20:37:04 <ehird> Yes; if only they were any good at writing manpages.
20:39:14 <ehird> Gah, why isn't there a mount you can do in userspace just on things you own.
20:39:44 <ehird> So I could do $ mkdir blah; usmount -o loop -t ext2 fs blah
20:39:53 <ehird> I guess I could write usmount as a setuid thing.
20:41:19 <ehird> No, a regular full-blooded mount, with the files owned by the user.
20:41:33 <ehird> But the destination directory, and source device/files, must be accessible by the user.
20:41:43 <ehird> So, pretty much just for loopbacks.
20:41:49 <AnMaster> ehird, it would need to make sure options like nodev,nosuid and such are passed too
20:42:02 <ehird> /linuxrc a symlink to /bin/busybox; queer.
20:42:16 <AnMaster> ehird, if that is an initrd: why
20:42:24 * AnMaster thought initramfs were used since ages
20:42:34 <AnMaster> and I'm pretty sure initrd never was ext2
20:42:38 <ehird> I'm looking at the stage1 stali rootfs.
20:42:53 <AnMaster> and /linuxrc is for initrd unless I misremember
20:43:07 <ehird> So it's for initrd, then.
20:44:15 <AnMaster> ehird, initrd is more complex from a kernel point of view. Since the kernel itself has to start /sbin/init after the /linuxrc process exits. With initramfs the initial process (/init unless I misremember) never exits but should end up executing the real init at the end
20:44:31 <ehird> That's a minor detail; initramfs is, I belieeve, more cmoplex than initrd in itself.
20:44:43 <ehird> Anyway, eventually I believe they'll have enither.
20:44:48 <ehird> My distro certainly won't.
20:44:52 <ehird> Writing my own init will be fun.
20:45:06 <ehird> Even the uber-minimal BSD-style ones (like BusyBox's which doesn't even support runlevels) use inittabs and stuff.
20:45:21 <ehird> Mine will be awesome and use shell scripts only! Mwahaha.
20:45:25 <ehird> AnMaster: so the question is, why symlink it to busybox?
20:45:32 <ehird> Maybe busybox has a linuxrc command
20:45:39 <ehird> I was thinking it'd start a shell or whatever
20:45:46 <ehird> Maybe it treats linuxrc as init
20:45:49 <ehird> and starts busybox init
20:46:08 <AnMaster> ehird, initramfs are gziped cpio archives. initrds are gziped file system images
20:46:34 <ehird> Are you sure that linuxrc is initrd?
20:46:44 <ehird> Pretty sure it's just an old name for the init process thingy.
20:46:55 <ehird> AnMaster: Are you sure that it isn't that initrd starts via linuxrc?
20:46:57 <AnMaster> ehird, I would have expected /linuxrc to be a shell script
20:47:06 <ehird> I wouldn't; it wasn't on my fucked up Debian netbook thing.
20:47:08 <AnMaster> "Once the initial root file system is up, the kernel executes /linuxrc as its first process. When it exits, the kernel assumes that the real root file system has been mounted and executes "/sbin/init" to begin the normal user-space boot process."
20:47:50 <AnMaster> debian netbook? Oh the one where you were dumped into root login?
20:52:20 <ehird> Incidentally, configuring a kernel is tiring.
20:52:35 <ehird> It's hard to set up a good environment to bootstrap a distro...
20:52:46 <AnMaster> ehird, first one on a given hardware yes. After that you can just do make oldconfig to check for changes usually
20:52:48 <ehird> I wonder if I could write an init and drop it in my Arch fs and have it wor
20:53:06 <ehird> AnMaster: I'm not configuring for specific hardware, it's for my distro.
20:53:08 <AnMaster> and the occasional make menuconfig when you need to change some specific setting
20:53:15 <AnMaster> ehird, ah, general. Most as m then?
20:56:27 <ehird> AnMaster: No modules.
20:56:45 <AnMaster> so then all drivers are built in
20:56:51 <ehird> Not all, just a small amount.
20:57:10 <ehird> No it isn't, there's going to be barely anything in there.
20:57:17 <ehird> It'll be smaller than any kernel you're running, almost certainly.
20:57:34 <ehird> For instance, none of the big ELF code... just uber-simple a.out. That's a pretty big drop right there.
20:57:48 <AnMaster> ehird, probably. But that means it will be limited to a few hardware options
20:57:52 <ehird> No support for 83429239487234 filesystems... just JFS (the main filesystem), FAT, ext and a few others.
20:57:56 <ehird> AnMaster: No, I'm going generic.
20:58:04 <ehird> The main non-generic hardware driver you need is for graphics.
20:58:08 <ehird> So I'll have packages for that.
20:58:15 <ehird> kernel-nvidia, kernel-radeonhd, etc.
20:58:37 <ehird> AnMaster: SATA I'll probably include the drivers that do best with the most common hardware.
20:58:44 * AnMaster needs the VIA SATA stuff for example on his desktop. And AHCI on the laptop.
20:59:04 <ehird> It won't be very hard to add your own, since the build environment stuff will be easily downloadable so it's just a quick menuconfig and then using my mkfile to build the package.
20:59:08 <ehird> But I'll include a few.
20:59:22 <ehird> I estimate the kernel will be something like 5 MiB.
21:00:17 <ehird> My Arch kernel here is 601 KiB + 1.8 MiB
21:00:22 <ehird> (kernel26.img and vmlinuz26)
21:00:36 <AnMaster> ehird, arch does use modules though
21:00:47 <AnMaster> ehird, also with modules built in you won't need any initrd
21:00:48 <ehird> I was giving a lot of credence to my major minimalist powers.
21:00:53 <ehird> AnMaster: Yep, I won't have any
21:00:58 <AnMaster> like on my desktop. No initrd there
21:01:09 <AnMaster> ehird, unless you want harddisk encryption or lvm or such
21:01:25 <ehird> I might include disk encryption.
21:01:41 <ehird> I might have a tool to compile and install a kernel with predefined configurations.
21:02:15 <ehird> # mkkernel encryption radeonhd
21:02:30 <ehird> And that'll add a package to your system called kernel-encryption-radeonhd.
21:02:33 <ehird> (Sorted alphabetically.)
21:02:43 <ehird> And become the kernel.
21:03:05 <ehird> Basically like Gentoo's USE flags, but on a much smaller scale and just for the kernel.
21:03:37 <ehird> I don't plan to support too much, though; LVM is quite unlikely.
21:03:44 <ehird> Especially since you can't shrink JFS anyway.
21:03:54 <ehird> (Which is a shame, but I haven't found something as good as JFS in other areas.)
21:03:59 <AnMaster> ehird, encryption of / does however require an initrd/initramfs
21:04:10 <ehird> AnMaster: Or a separate boot partition.
21:04:20 <AnMaster> ehird, _and_ a separate boot partition
21:04:32 <ehird> I can probably make a tiny tiny stub initramfs.
21:04:37 <AnMaster> ehird, after all, the boot loader has to be able to load the initrd from somewhere
21:05:30 <AnMaster> or you could have a user space /boot/decrypt-and-init and then have /boot as the original root fs and do some strange mount tricks... Oh and you would need to use init=/decrypt-and-init on the kernel command line
21:05:30 <ehird> Anyway, for a stock configuration, including X11 and window manager startup (login skipped; window manager will probably be dwm, so, minimalist), as soon as lilo hands over to the kernel, I expect to be able to finish boot in 2 seconds.
21:05:42 <AnMaster> (initrd is a less hackish solution then)
21:05:48 <ehird> On an SSD? Let's say 1.6 seconds.
21:06:05 <ehird> AnMaster: initrd is ext2 only, though, isn't it? Does it work if just the boot partition is ext2 only, I wonder?
21:06:29 <AnMaster> ehird, hm? initrd is an ext2 file system image that is gzip compressed
21:06:37 <ehird> Right, that's what I meant.
21:06:44 <AnMaster> as for /boot, it could be some other fs supported by the boot loader
21:06:45 <ehird> You can read the initrd from a JFS partition, yes?
21:06:53 <ehird> lilo supports JFS, which is nice.
21:07:31 <AnMaster> ehird, about that "boot to other OS". grub-reboot maybe?
21:07:55 <ehird> Probably. Looks quite new, anyway.
21:08:04 <ehird> I believe lilo has had it for a decade or whatever.
21:08:19 <AnMaster> ehird, btw you know that karmic is using grub2 by default?
21:08:39 <ehird> Following Debian, presumably.
21:08:41 -!- fax has left (?).
21:08:56 <AnMaster> ehird, opinions on "Ubuntu One"?
21:09:12 <ehird> Ubuntu One isn't as good as Dropbox, and FUCK THEM for using the Ubuntu name.
21:09:33 <AnMaster> ehird, about from the dropbox bit (which I never tried): agreed
21:09:38 <ehird> Dropbox, for instance, has the feature "can access your files on obscure OSs like Windows and OS X".
21:10:00 <ehird> Ubuntu One is just a ripoff of Dropbox with less features and without supporting other OSs.
21:10:10 <AnMaster> ehird, how much space do you get with dropbox no cost version or such?
21:10:22 <ehird> Same as Ubuntu One.
21:10:31 <ehird> If you invite other people, set up other machines, blah blah blah, Dropbox rewards you with some more.
21:10:36 <ehird> Can get almost 3 GiB or something that way.
21:10:40 <ehird> You know, if you're a huge cheapskate.
21:10:43 <ais523> I hate that sort of referral scheme, it makes me avoid a company
21:10:49 <ais523> because, it means I can't trust peer reviews of the company
21:10:55 <ais523> if people are bribed to say good things about it
21:11:00 <ais523> then, I can't trust them as much
21:11:01 <ehird> Only if you think your friends are really, really scammy
21:11:18 <ehird> People who aren't massive assholes go "Hey, if you sign up for this I get some more blah"
21:11:36 <ehird> If it's on a website or whatever, just look to see if it's a referral link and if so ignoore the review.
21:11:56 <ehird> Anyway, it's not just referring (and it only lets you refer a few, not continuously).
21:12:10 <ehird> There's other things like "put our client on your other machines" blah blah. It's not worth the effort though.
21:12:27 <AnMaster> ehird, on other machines? Like all your computers?
21:12:34 <ehird> As many as you want.
21:12:50 <ehird> Incidentally, lilo has more eyecandy than GRUB 1, I think.
21:12:54 <AnMaster> well. If I used such a service I would likely want it on all of mine
21:13:03 <AnMaster> ehird, that feels sooo backwards XD
21:13:09 <ehird> You can set up a menu grid instead of a list so you can use the bitmap background to have a fancy graphical menu.
21:13:14 <AnMaster> so grub1 is the minimalist option?
21:13:19 <ehird> Position your own countdown timer. etc.
21:13:22 <ehird> AnMaster: No, lilo is smaller.
21:13:28 <ehird> It just does it all with less code.
21:13:39 <ehird> Besides, none of this stuff gets added to your MBR unless you enable it.
21:14:30 <ehird> Incidentally, lilo doesn't use the "boot: " interface by default; it uses a GRUB-style menu.
21:14:54 <ehird> I think GRUB 1 can do higher-resolution things, though; lilo only supports a paltry 640x480x8 bitmap!
21:14:54 <AnMaster> the "don't have to remember to rerun the lilo command after a kernel update" bit and "option to edit the commend line" bit are important to me
21:15:07 <AnMaster> ehird, what about password protection for booting recovery kernel?
21:15:14 <ehird> AnMaster: The former will be handled by my distro, the latter, I believe, you can just enter into the boot: prompt
21:15:15 <AnMaster> like the one with init=/bin/busybox
21:15:20 <ehird> Like boot: and enterr a lilo line. Whatever.
21:15:24 <ehird> Doesn't bother me, so I don't care.
21:15:31 <ehird> AnMaster: I think it has password protection.
21:15:55 <ehird> "menu-scheme=Wm intense white on magenta"
21:15:59 <ehird> I think I will pass on that colour scheme
21:15:59 <AnMaster> ehird, password required to edit the boot line is all I want.
21:16:20 <ehird> Upper case for intense.
21:16:52 <ehird> AnMaster: You can do password=
21:17:03 <AnMaster> ehird, I always wondered. Why does CMYK has Y for black
21:17:10 <ehird> It has K for black
21:17:16 <ehird> And because B is blue
21:17:21 <fizzie> "The “K” in CMYK stands for key since in four-color printing cyan, magenta, and yellow printing plates are carefully keyed or aligned with the key of the black key plate. Some sources suggest that the “K” in CMYK comes from the last letter in "black" and was chosen because B already means blue.[1][2] However, this explanation, though plausible and useful as a mnemonic, is likely inaccurate, the speculative invention of authors unfamiliar with traditio
21:17:21 <fizzie> nal printing technology."
21:17:41 <AnMaster> fizzie, what do they suggest instead?
21:17:51 <ehird> The “K” in CMYK stands for key since in four-color printing cyan, magenta, and yellow printing plates are carefully keyed or aligned with the key of the black key plate.
21:17:54 <ehird> Like it, you know, said.
21:17:54 <fizzie> Key, like it says there.
21:18:12 <ehird> I wonder why people like their init to spew out a ton of crap.
21:18:27 <ehird> Is it because they have so much crap that they like to see how far along their ages-long boot is?
21:18:38 <fizzie> ehird: "Watching shit scroll by for hours makes me a Linux expert overnight."
21:18:39 <ehird> Personally I'd only like to be bothered when something went wrong.
21:18:48 <ehird> fizzie: Yep, but every distro seems to do it
21:18:48 <AnMaster> ehird, when things go wrong you can easily see where?
21:18:51 <ehird> Apart from those newfangled graphical boots
21:19:06 <ehird> AnMaster: Have those kids never heard of `if ! blah; then echo OMG BLAH FAILED; fi`?
21:19:23 <AnMaster> ehird, what if the kernel freezes early on in some unexpected way
21:19:28 <ais523> it's probably so that if the kernel locks hard, you can see what the last successful operation was
21:19:31 <AnMaster> thus, what was the last line before that printed
21:19:33 <ais523> and guess what operation failed as a result
21:19:35 <ehird> AnMaster: I'm talking about in init
21:19:45 <ehird> By that point a fucked-up kernel lockup is probably quite unlikely
21:19:47 <AnMaster> ehird, oh it is just colourful messages here
21:19:52 <ehird> And if it does, good luck debugging it!
21:20:03 <ehird> It's pretty pointless
21:20:12 <AnMaster> ehird, the colours are pretty?
21:20:28 <ehird> Set a pretty image as your desktop so the 2-second boot gets to it :-P
21:20:37 <ehird> You could even animate it.
21:20:42 <AnMaster> ehird, with a 2 second boot it would be fairly pointless
21:20:51 <ehird> And as far as kernel lockups go...
21:20:59 <ehird> Just add it as a boot option.
21:21:15 <ehird> You could even add "Awesomedistro (debug kernel lockup)" or whatever as a lilo entry.
21:21:21 <AnMaster> ehird, is there anything wrong with those early kernel messages being printed though?
21:21:23 <fizzie> Make your init a Mandelbrot zoomer, advances a bit every time something happens. You can then tell from the shape where it hangs up if something goes wrong.
21:21:25 <ehird> It's just superfluous
21:21:38 <ehird> It takes up time and effort that could be spent speeding up the boot
21:21:47 <ehird> Literally time; those prints are so expensive. :P
21:21:56 <ehird> But seriously, it's just —
21:22:06 <ehird> I was just asking if there's a justifiable point apart from watching shit scroll by.
21:22:19 <AnMaster> ehird, you want just a single like like: "Loading <distroname>...."
21:22:21 <ehird> Since the only reason is an early kernel hangup, and you can easily just force the kernel to spew a lot, I guess I'll disable it.
21:22:24 <ehird> AnMaster: Maybe a few more.
21:22:42 <AnMaster> ehird, oh like "we go to init" "starting GUI"?
21:23:17 <AnMaster> Yay! time for init and we are still alive!
21:23:18 <ehird> I had written that as my first!
21:23:47 <AnMaster> Phew, we almost made it, just the GUI left. Pray for X working well for once!
21:23:56 <ehird> "Oh god, I'm bleeding, FUCK, fuck, get me to a doctor... my la-ast wish... is... to... start X11... farew-well... know me as...b-boot p-p... proc-cess..HYUAAAAAAAAAAAGhhgj"
21:24:26 <ehird> printk("Printing this message...\n");
21:24:35 <AnMaster> or "Starting X. Good luck. (You will need it!)"
21:25:15 <ehird> "You've started X11 successfully 10 times in a row. As a precaution, I will drop you to a single-user shell to fix the problems you will have this time."
21:25:46 <ehird> New device detected since last boot. Please remove it and reboot.
21:26:34 <ehird> fizzie: I like that mandelbrot init idea, by the way
21:26:47 <AnMaster> "Lucky it isn't Windows eh? They would would have had to re-activate"
21:27:04 <ehird> AnMaster: Hey! Just use Linux Genuine Advantage.
21:27:29 <AnMaster> ehird, btw. if you are going to do hardware detection it will slow you down
21:27:46 <ehird> I'll probably do some cursory detection, but nothing much.
21:27:50 * ehird wonders how he could do dependency-based asynchronous init with just an rc script
21:28:02 <ehird> Maybe I'll load all the start scripts at once
21:28:09 <ehird> and the depending scripts just check environment variables
21:28:11 <AnMaster> ehird, you need a service supervisor though
21:28:31 <AnMaster> and using a script for that would be bloated I suspect
21:28:35 <ehird> A vague term; can you clarify just so we're on the same page?
21:29:20 <AnMaster> ehird, something to try to restart critical services if they go down (and give up after some fixed number of retries in a short time period)
21:29:39 <AnMaster> ehird, like upstart does. Or sysvinit does
21:29:48 <ehird> I don't think BSD inits do that
21:29:58 <AnMaster> or even daemontools (I remember reading someone managed to use daemontools for init!)
21:30:01 <ehird> besides, it can't be very critical if you can run enough code to restart them
21:30:05 <ehird> AnMaster: daemontools is an init replacement...
21:30:08 <ehird> (among other things)
21:30:16 <AnMaster> ehird, usually it isn't used as that though
21:30:22 <ehird> It's meant to be,t hough
21:30:35 <ehird> Incidentally, here's my wonderful shutdown command:
21:30:55 <AnMaster> ehird, and it tells services to shut down cleanly then?
21:31:08 <ehird> Runs /etc/rc.stop and then halts
21:31:26 <AnMaster> ehird, and rc.stop sends SIGTERM and such to processes needing it?
21:31:35 <AnMaster> ehird, what about kexec? Oh wait I guess not
21:32:06 <ehird> Yeah; runs all the shutdown-service things, kills every process other than the ones needed to do this, waits until they all respond, kill -9s the rest, and turns ogg
21:32:23 <ehird> the nautilus integration does, though
21:32:28 <ehird> admittedly at the moment it is the only client
21:32:32 <ehird> the protocol is open though
21:32:32 <AnMaster> ehird, can't it just use fuse or such
21:32:40 <ehird> You can use the folder as-is
21:32:53 <AnMaster> because dropbox is like useless to me if I can't use it on my headless computers too
21:32:54 <ehird> (It doesn't use FUSE) it's synchronization
21:32:57 <ehird> so it needs to be on disk
21:33:13 <AnMaster> ehird, there goes the idea of *extra* storage space :P
21:33:25 <ehird> Just disable dropboxd and remove the directory after syncing :P
21:33:36 <ehird> AnMaster: Anyway, process 1 will be the thing that does the actual halting
21:33:51 <ehird> After /etc/rc.stop closes it'll have kill -9'd everything else so it'll just be process 1 and whatever the kernel's running chilling about
21:33:53 <AnMaster> ehird, yeah I believe that is already the case
21:34:14 <AnMaster> ehird, no kexec support I assume?
21:34:52 <ehird> AnMaster: what signal do you think I should use that instead of halting reboots?
21:35:07 <ehird> actually, I'm unsure whether shutting down should be QUIT or TERM
21:35:19 <AnMaster> ehird, what about suspend to ram/disk?
21:35:22 <ehird> I'm leaning towards quit because # kill 1 working without warning seems dangerous
21:35:30 -!- coppro has quit (Read error: 60 (Operation timed out)).
21:35:38 <ehird> AnMaster: Probably USR1 and USR2
21:35:50 <ehird> with a shell script that does that called suspend_ram or something
21:35:55 <ais523> ehird: SIGQUIT is supposed to create core dumps
21:35:57 <AnMaster> ehird, what about SIGVTALRM? Or SIGRTMAX
21:35:57 <ehird> (USR1 for RAM, USR2 for disk)
21:36:00 <ais523> or at least, does on most Unices
21:36:04 <ehird> ais523: ah, wasn't aware
21:36:16 <AnMaster> ehird, ooh idea for restart... SIGCONT
21:36:20 <ais523> you do it with C-\ if a program isn't responding to C-c
21:36:30 <AnMaster> ehird, "continue after reboot"
21:36:32 <ais523> or to interrupt a program to get a core dump
21:36:37 <ehird> AnMaster: that'd be suspend to disk + reboot
21:36:47 <AnMaster> ehird, however, normally SIGCONT is for the bg/fg commands
21:36:55 <ehird> I should let you do "fg 1"
21:36:59 <ehird> and press ^C to shut down
21:37:06 <AnMaster> I don't think you can intercept SIGCONT in fact
21:37:22 <ehird> I meant specialcase it
21:37:30 <ais523> you can handle SIGCONT
21:37:31 <AnMaster> fg is a shell thingy and related to shell job control
21:37:35 <ais523> the stuff happens after you're continued
21:37:43 <ais523> obviously, you have to be running in order to handle it
21:37:48 <AnMaster> ais523, when does SIGSTOP happen?
21:38:00 <ais523> you can handle SIGTSTP, though
21:38:05 <ais523> which is generated by C-z
21:38:12 <ehird> # kill -9 1 will force a hard shutdown, I think.
21:38:19 <ehird> without any process killing or anything
21:38:20 <AnMaster> ais523, sure SIGSTOP is NOT Ctrl-Z?
21:38:29 <ais523> AnMaster: Ctrl-Z generates SIGTSTP
21:38:31 <ehird> you know, when you're too lazy to hold down the power button
21:38:35 <ais523> which does the same as SIGSTOP by default
21:38:40 <ais523> but which /can/ be handled or interrupted
21:38:48 <ais523> (whereas SIGSTOP can't be)
21:38:55 <ais523> well, handled or masked
21:39:07 <ehird> SIGWINCH discard signal Window size change
21:39:13 <ehird> this is clearly for "change framebuffer console resolution"
21:39:19 <AnMaster> ehird, ... you only hold down power button when you can't do it the normal way :P
21:39:28 <ehird> AnMaster: Or if you're lazy.
21:39:40 <ehird> Admittedly it doesn't matter if a regular shutdown only takes, like, a second.
21:39:41 <AnMaster> ehird, yeah. But well that joke is rather twisted
21:40:00 <AnMaster> ehird, and holding down power button takes like 10 seconds on my computer
21:40:01 <ehird> Assume some bloated Ubuntu thingy that takes 30 seconds to shutdown :P
21:40:06 <ehird> 20 SECONDS ARE PRECIOUS
21:40:15 <ais523> SIGWINCH is for console apps
21:40:21 <ehird> ais523: I'm joking, man :|
21:40:21 <ais523> they get it if someone resizes the window they're running in
21:40:31 <ais523> it doesn't have an argument to say the new size
21:40:34 <ais523> it's just a notification
21:40:35 <ehird> doesn't kill refuse to send signals to process 1?
21:40:47 -!- ehird has quit (Remote closed the connection).
21:41:27 <ais523> <man kill> This command meets appropriate standards.
21:41:35 <ais523> umm, normally the man pages say /which/ standards
21:42:16 <AnMaster> ais523, the table in that man page is fucked up
21:42:16 <fizzie> The only signals that can be sent to process ID 1, the init process, are those for which init has explicitly installed signal handlers. This is done to assure the system is not brought down accidentally.
21:42:49 <AnMaster> 0 0 n/a exit code indicates if a signal may be sent
21:43:03 <AnMaster> there were multiple blank lines
21:43:06 -!- ehird has joined.
21:43:21 <ehird> launchd, to be precise
21:43:23 <AnMaster> <ehird> doesn't kill refuse to send signals to process 1?
21:43:23 <AnMaster> * ehird has quit (Remote closed the connection)
21:43:32 <AnMaster> <fizzie> The only signals that can be sent to process ID 1, the init process, are those for which init has explicitly installed signal handlers. This is done to assure the system is not brought down accidentally.
21:43:39 <AnMaster> ehird, there it is explained ^
21:43:40 <ehird> it overlaid the shutdown spinner (circling circle thing) over my desktop which froze
21:43:51 <ehird> had my same idea <3
21:44:13 <ais523> what signal did you kill init /with/?
21:44:42 <ais523> ehird: upstart isn't init either
21:44:46 <ehird> = init, cron, inetd, etc etc etc
21:44:47 <ais523> but it's called init in the process table
21:44:50 <ehird> yes, but upstart is distinct
21:44:53 <ehird> and launchd is called launchd in the table
21:44:58 <ehird> (upstart iis distinct from cron etc atm)
21:45:22 <ehird> tbh writing an init isn't hard
21:45:31 <ehird> everything's set up and cushy
21:45:34 <fizzie> It is distinct from cron so far, though:
21:45:36 <fizzie> "Will Upstart replace cron, atd or anacron?
21:45:36 <fizzie> Yes. A planned feature for Upstart is the ability to generate events at a particular scheduled time, regular scheduled time or particular timed intervals."
21:45:48 <fizzie> But they're not planning on adding inetd bits into it.
21:46:02 <fizzie> Well, it says "Maybe" there.
21:46:25 <AnMaster> and they suggest that dbus is trying to replace init
21:46:30 <ehird> who even uses cron without meaning anacron?
21:46:50 <ehird> "do this thing at this time but if you can't GIVE UP COMPLETELY"
21:46:57 <ehird> there aren't many operations that make sense there
21:47:39 <AnMaster> ehird, anyway why does ubuntu start both cron and anacron
21:47:41 <ehird> hmm http://en.wikipedia.org/wiki/Anacron#Drawbacks okay that's pretty bad
21:47:44 <ehird> AnMaster: see my link
21:47:50 <ehird> seems like anacron is very limited
21:48:22 <ehird> I wonder if it's the responsibility of the shutdown procedure to call sync(1)
21:48:30 <AnMaster> um I have tasks running every few hours
21:48:51 <ehird> "Anacron is not an attempt to make cron redundant. It cannot be used to schedule commands at intervals smaller than days"
21:48:53 <AnMaster> "anacron can only run tasks once a day (or less often such as weekly or monthly). In contrast, cron allows tasks to run as often as every minute (but does not guarantee their execution if the system goes down). In practice, this is not usually an issue, since it is rare to have a task that must be guaranteed to run more often than (at least) once a day"
21:48:57 <ehird> THAT'S WHAT I LINKED TO
21:49:05 <ehird> that's why ubuntu starts cron as well
21:49:15 <ehird> so why did you repeat my link
21:49:24 <AnMaster> and I was commenting upon a specific line
21:49:40 <pikhq> ... Anacron can only run tasks once a day?
21:49:54 <ehird> Every time Anacron is run, it reads a configuration file that specifies the jobs Anacron controls, and their periods in days. If a job wasn't executed in the last n days, where n is the period of that job, Anacron executes it. Anacron then records the date in a special timestamp file that it keeps for each job, so it can know when to run it again. When all the executed commands terminate, Anacron exits.
21:50:11 <ehird> i thought it just logged when it ran a regular cron command, and then if there are missing log entries, ran them then and logged it
21:50:17 <ehird> that design is stupid
21:51:17 <ehird> crontabs suck though
21:51:20 <pikhq> Vixie cron is nice in that it doesn't do anything particularly dumb.
21:51:30 <pikhq> Well, given the constraints of crontab format, that is.
21:51:59 <ehird> "Remember me from months ago? You've forgotten what I look like, don't I? BETTER GOOGLE FOR IT"
21:52:04 <AnMaster> <ehird> crontabs suck though <-- how so?
21:52:24 <pikhq> AnMaster: The specific file format is rather arcane and hard to remember.
21:52:35 <pikhq> A comment on top shouldn't be needed to remember the format.
21:52:56 <AnMaster> rather than some multi-line bind style config
21:53:00 <pikhq> The /etc/passwd file suffers from the same issue, but that at least isn't something that non-programs need to mess with.
21:53:11 <ehird> I can't really get rid of crontab, but I can get rid of /etc/fstab
21:53:30 <AnMaster> what the hell is wrong with fstab
21:53:33 <pikhq> By rewriting mount a bit, I presume.
21:53:38 <ehird> Exactly the same that's wrong with crontab
21:53:56 <pikhq> And being glad that very few things interact *directly* with fstab.
21:53:59 <AnMaster> device mountpoint fs options dump pass
21:54:07 <ehird> The main use of fstab is just to mount things on boot, so guess what that'll become?
21:54:17 <ehird> Lines in a shell script in /etc/rc.d calling mount(1).
21:54:30 <ehird> I have some thoughts for handling the "mount /dev/foo" case, but I have to mull on them.
21:54:35 <ehird> AnMaster: About the same, really.
21:54:39 <ehird> Maybe a few more characters. Doesn't matter.
21:54:41 <pikhq> ehird: Fuse, perhaps?
21:54:42 <ehird> It's significantly simpler.
21:54:55 <ehird> pikhq: How does FUSE help there?
21:55:06 <pikhq> You can mount as a normal user with FUSE.
21:55:13 <ehird> Oh, that's not what I meant
21:55:14 <AnMaster> ehird, you use fuse to create a virtual /media directory!
21:55:23 <AnMaster> with auto adding stuff according to an XML config
21:55:29 <pikhq> AnMaster: AAAAAGH.
21:55:29 <ehird> pikhq: I meant mount knowing where to mount devices
21:55:32 <ehird> and their FS type, etc
21:55:41 <pikhq> ehird: Oh, that. Yeah.
21:55:58 <ehird> Also, any support for FUSE I have will probably be via translation to 9P.
21:56:08 <AnMaster> pikhq, gnome vfs seems like that to me
21:56:12 <ehird> I'll reuse the FUSE library, and translate however it talks to the kernel module into 9P.
21:56:31 <pikhq> Probably already been done. If not, it should be easy to do.
21:56:44 <pikhq> The FUSE library supports multiple kernels by now.
21:56:51 <ehird> Hopefully the kernel has good 9P support, because the only other options use FUSE.
21:57:03 <AnMaster> ehird, btw I don't think llvm/clang supports a.out
21:57:11 <AnMaster> and gcc itself probably doesn't
21:57:17 <ehird> gcc did until a recent 4.x version
21:57:23 <pikhq> AnMaster: LLVM should at least be easy to retarget.
21:57:30 <pikhq> (being designed with that in mind and all.)
21:57:33 <ehird> wrt compiling gcc into an a.out
21:57:44 <AnMaster> ehird, not sure. Just sounds like something I heard someone mention once
21:58:04 <ehird> At least my Arch /etc/fstab is nicely simple
21:58:14 <ehird> The main line is just "/dev/sda1 / jfs defaults 0 1"
21:58:28 <AnMaster> ehird, ah but what if other disks are plugged in!
21:58:38 <ehird> Then let's just hope they become /etc/sdb.
21:58:59 <ehird> I'm not using udev, so hopefully I can find a quite stable devfs.
21:59:43 <AnMaster> ehird, devfs is no longer supported with recent kernels
22:00:00 <ehird> Then maybe I'll use static.
22:01:57 <ehird> When looking around the kernel config options I had a horrible dilemma.
22:02:17 <ehird> Heap randomisation
22:02:17 <ehird> Cons: Makes kernel bigger, BREAKS LIBC5 BINARIES
22:03:03 <pikhq> Libc5? Meh, who needs it?
22:03:19 <ehird> But, static binaries should work for decades!
22:03:41 <pikhq> On a somewhat more serious note, I do have at least one libc5 binary.
22:04:02 <ehird> Anyway, heap randomisation as a security feature made me think of OpenBSD almost instantly.
22:04:08 <ehird> It's the kind of thing they'd do, isn't it...
22:04:16 <pikhq> Only Playstation 1 emulator I could get to work used Libc5. And was a binary.
22:04:20 <ehird> Hmm, it also breaks position dependent code, doesn't it :P
22:04:21 <pikhq> ehird: Yeah, they do that.
22:04:38 <ehird> Instead of the first five fields, one of eight special strings may
22:04:40 <ehird> @reboot Run once, at startup.
22:04:49 <pikhq> Also, you can do position-independent staticly linked binaries. ;p
22:05:02 <ehird> HAY UNPRIVILEGED USER! I'M TOTALLY INIT
22:05:05 <ehird> Use meeeeeeeeeeeeeee
22:12:45 -!- coppro has joined.
22:13:33 <ehird> Does at(1) run the commands at the time you specify, or could it be some minutes late?
22:15:49 <ais523> not sure, at's been broken for months on ubuntu
22:16:46 <ehird> I'm wondering how to do a good clock in dwm; xsetroot -name foo changes the top-right text
22:17:04 <ehird> and I want to update it on the 0th second of every minute (give or take some seconds) without using CPU
22:18:00 <SimonRC> # A still more glorious dawn awaits: not sunrise, but a galaxyrise. A morning filled with 400 billion suns; the rising of the Milky Way. # -- http://www.youtube.com/watch?v=zSgiXGELjbc
22:18:22 <ehird> I linked that like a month ago
22:18:25 <AnMaster> <pikhq> On a somewhat more serious note, I do have at least one libc5 binary. <-- what huh?
22:18:31 <ehird> He explains a few lines down.
22:18:39 <ehird> I wonder what libc4 is like.
22:18:45 <AnMaster> ehird, I do your style of log reading :P
22:18:58 <ehird> It's remarkably rewarding!
22:19:03 <ehird> Like everyone has the exact same questions as you.
22:19:18 <AnMaster> <ehird> Hmm, it also breaks position dependent code, doesn't it :P <-- um no?
22:19:34 <AnMaster> that would be data/code segment randomisation
22:19:40 <SimonRC> hmm ... this evening; that video does not make me start crying
22:19:57 <ehird> I forgot just how hardcore the http://mastodon.biz/ author is
22:20:05 <ehird> "I'm trying to decide whether to roll to one of the super-bloated newer Linux kernels or write my own USB stack plus SATA and UDMA drivers for 2.0.28"
22:20:10 <ehird> you have to be pretty badass to even consider thtat
22:20:22 <AnMaster> <ehird> HAY UNPRIVILEGED USER! I'M TOTALLY INIT <-- the @reboot thing? Yes very useful. I use it to start some irc bots for example
22:20:39 <ehird> Why not just have ~/.rc.start or something
22:20:52 <AnMaster> ehird, yet another service to read that?
22:20:57 * ais523 wonders if fakeroot includes faking reboots, etc
22:21:03 <ehird> No, just have /etc/rc.start do it
22:21:06 <pikhq> Because Unix was designed and then crap got shoveled on.
22:21:15 <AnMaster> ehird, and suing to each user? Hm
22:21:19 <pikhq> Making an OS that works but doesn't know what consistency is.
22:21:25 <AnMaster> yet another code point that has to be audited for that
22:21:25 <ehird> It just doesn't seem like the kind of thing cron should do to me
22:21:34 <ehird> AnMaster: Dude, I trust my init scripts more than cron
22:21:36 <pikhq> (better than the Windows solution of "Fuck design". :P)
22:21:41 <ehird> Especially since it'd just be a few lines
22:21:43 <AnMaster> ehird, cron already done this for ages
22:21:53 <ehird> I don't care, it's still something you "have to audit"
22:22:01 <ehird> @reboot shouldn't be in cron, it's an init task
22:22:48 <AnMaster> <SimonRC> # A still more glorious dawn awaits: not sunrise, but a galaxyrise. A morning filled with 400 billion suns; the rising of the Milky Way. # -- http://www.youtube.com/watch?v=zSgiXGELjbc <-- argh now I have that tune on my head again
22:22:51 <SimonRC> ehird: I recall that Apple have a program that unifies the run-at-boot things and cron
22:23:04 <ehird> And it inexplicably does inetd
22:23:18 <ehird> plists are XML nowadays, unfortunately.
22:23:24 <ehird> Or the binary format.
22:23:26 <AnMaster> ehird, oh? they changed format?
22:23:36 <ehird> It used to be a wonderful JSON-type dealie.
22:24:17 <AnMaster> ehird, pretty sure they were binary blobs yeah
22:24:34 <ehird> That's one of the new formats
22:24:38 <ehird> They used to be almost identical to JSON
22:24:40 <ehird> and it was lovely.
22:24:51 <pikhq> ehird: Ah, the days of a sane format for plist.
22:25:03 <AnMaster> ehird, well, tiger is the only OS X version I used
22:25:06 <ehird> When men were men, women were men and plists weren't XML.
22:25:10 <SimonRC> The thing about XML is that everywhere has libraries that do all the parsing for you.
22:25:19 <ehird> SimonRC: It is very dark.
22:25:21 <ehird> You are in #esoteric.
22:25:25 <ehird> You are about to defend XML.
22:25:28 <ehird> You are likely to be eaten by a grue.
22:25:30 <AnMaster> SimonRC, are you trying to defend... XML?
22:25:39 <SimonRC> I am saying it can be practical
22:25:47 <ehird> You are eaten by a Grue.
22:25:58 <SimonRC> OTOH, almost everywhere has JSON parsers too
22:26:02 <pikhq> SimonRC: The thing about XML is that it has no advantages over JSON and many disadvantages.
22:26:04 <ehird> >backtrack slightly
22:26:10 <ehird> The Grue fails to reverse entropy and you stay eatetn.
22:26:15 <pikhq> (for generic object serialisation, that is)
22:26:25 <pikhq> And XML is not designed at *all* for what it's generally used for.
22:26:34 * SimonRC tries to recalls what teh JSON equivalent of xpath is
22:26:35 <pikhq> It's meant to just do markup.
22:26:52 <ehird> And it's pretty bad at markup too
22:26:58 <ehird> AnMaster: HTML for object serialisation?
22:27:07 <AnMaster> ehird, oh hah I didn't even think about that
22:27:08 <pikhq> ehird: But it at least does the job without making me want to go and kill everyone.
22:27:09 <SimonRC> why is it people confuse an explanation with a justification?
22:27:31 <AnMaster> ehird, you are the nutter who even thought I could have meant that
22:27:31 <ehird> SimonRC: AVOIDING MIND INFECTION
22:27:39 <ehird> It's good to be a nutter!
22:27:56 <ehird> Incidentally, a good rule of thumb:
22:27:58 <AnMaster> SimonRC, what is xpath good for?
22:28:27 <AnMaster> as in, what does it do, that is useful and harder with other ways
22:28:41 <ehird> Your program is to serve the user. Everything they make with it is theirs, not yours. Therefore, all your formats must either be well-known (if you must be compatible with other tools without translating) or minimalist and plain text (or binary if you really must).
22:28:42 <pikhq> http://en.wikipedia.org/wiki/Xpath This seems like a solution in search of a problem.
22:29:11 <ehird> Something like XML just takes the user's work away from them; it wraps their format in something they don't care about or want, and forces them to trawl through it to get their data out. In effect, the barrier to EXIT is high.
22:29:24 <ehird> So, don't use XML.
22:29:31 <ehird> AnMaster: It's not that, it's for traversing the tree.
22:29:39 <ehird> Think of it like CSS selectors
22:29:41 <ehird> but with worse syntacx
22:29:55 <ehird> It's not all that bad
22:29:56 <AnMaster> no simple code samples there on wikipedia
22:30:01 <SimonRC> there is internal data, which isn't for users, and there is external data, which the users requested be XML
22:30:14 <ehird> http://www.w3schools.com/XPath/xpath_syntax.asp
22:30:22 <ehird> but those examples are good
22:30:29 -!- coppro has quit (Remote closed the connection).
22:30:32 <ehird> SimonRC: By user, I mean people
22:30:40 <AnMaster> ehird, yeah because it is used as a db there?
22:30:53 <AnMaster> /bookstore/book[price>35.00] Selects all the book elements of the bookstore element that have a price element with a value greater than 35.00
22:30:58 <ehird> Want as in want as humans, not as in want for compatibility, or because of corporate bear-ocracy
22:30:59 <SimonRC> xpath is good for processing heirarchical data (the details of XML don't leak in much) the same way regexes are good for processing text
22:31:00 -!- coppro has joined.
22:31:16 <AnMaster> SELECT * from bootstore.book WHERE price > 35.00 ?
22:31:21 <AnMaster> something similar to that at least
22:31:28 <ehird> AnMaster: You could make filesystem paths into SQL too
22:31:30 <ehird> doesn't prove anythiing
22:31:58 <SimonRC> AnMaster: yeah, that is about it
22:32:07 <ehird> ah, awesome is indeed a fork of dwm
22:32:11 <AnMaster> /bookstore/book[price>35.00]/title <-- SELECT title FROM bootstore.book where price > 35.00 ?
22:32:33 <ehird> And its own multihead
22:32:42 <AnMaster> xft is nice. I mean, non-bitmapped fonts
22:32:44 <ehird> So basically they took dwm and fucked with it until it was bloated
22:32:52 <AnMaster> ehird, emwh I have no clue what it is
22:33:01 <ehird> AnMaster: For the list of workspaces (numbers), the []= diagarm of the current layout, the title bar, and status bar?
22:33:04 <AnMaster> and xcb is well, something we probably can't avoid even if we want
22:33:07 <ehird> It's just a few vetical pixels at the top.
22:33:17 <ehird> More like an ASCII icon tbh
22:33:23 <ehird> (title bar is global at top of screen)
22:33:35 <ehird> xcb is quite avoidable, use the xlib wrapper over it :-P
22:33:49 <fizzie> I like the "bloat" in the multihead sense; though admittedly dwm page says "NEW dwm creates a view for each Xinerama screen".
22:33:51 <ehird> AnMaster: XFCE seems to try and be Gnome without... well, Gnome.
22:34:03 <zzo38> Do you have a copy of this game on a VHS tape?
22:34:04 <AnMaster> ehird, xlib is quite eww unless I misremember. But xcb might be worse.
22:34:13 <ehird> AnMaster: xcb is lower-level but slightly more sane
22:34:14 <pikhq> Xcb does seem like an improvement over Xlib for what X is commonly used for these days (hardly even touched by anyone other than toolkit authors).
22:34:18 <ehird> but xlib code is easier to understand
22:34:31 <ehird> AnMaster: I don't think it (Xfce) works; for instance, I quickly get agitated by the lack of configurators for the system
22:34:31 <zzo38> I mean, the game, Super ASCII MZX Town Part II
22:34:38 <pikhq> Harder to use, but less crazy bizarre bugs.
22:34:43 <zzo38> Someone in the game asks for that copy
22:34:45 <ehird> It's inconsistent to have a GUI to configure the GUI but nothing else, UI-wise
22:34:49 <ehird> (understandable implementation-wise)
22:35:08 <ehird> And it seems to basically come down to GNOME with more settings, whereas GNOME's philosophy leads to less settings
22:35:11 <ehird> So the end result is quite odd
22:35:16 <SimonRC> anyone using Debain unstable at the moment?
22:35:20 <AnMaster> <ehird> AnMaster: I don't think it (Xfce) works; for instance, I quickly get agitated by the lack of configurators for the system <-- and you are minimalist!
22:35:32 <ehird> AnMaster: Confusion is not minimalist.
22:35:34 <pikhq> ehird: XFCE is also more lightweight while having more settings.
22:35:51 <SimonRC> 'cause they have a hosed keyboard config app and the upstream fix hasn't come down yet, and I can't downgrade
22:35:51 <ehird> There's no reason to have GUI configuration and system configuration separate; they're facets of the same thing.
22:36:02 <AnMaster> worth trying as replacement for KDE 3.5.10 on my gentoo box
22:36:03 <ehird> Anyway, dwm uber alles :P
22:36:25 <AnMaster> ehird, tried tiling and decided I didn't like it. Probably could get used to it
22:36:33 <ehird> Most tiling managers suck balls
22:36:35 <SimonRC> ehird: not a very unixy attitude
22:36:35 <ehird> Which did you try?
22:36:46 <ehird> SimonRC: It is, but I cba to explain
22:36:51 <SimonRC> depending on what you actually mean
22:36:53 <AnMaster> ehird, hm awsome, xmonad and dwm iirc
22:37:12 <AnMaster> maybe slightly less than a year
22:37:13 <ehird> dwm is the only good one of those, because its layout is the right thing
22:37:21 <ehird> it has a main window, to the left
22:37:25 <AnMaster> ehird, iirc I found dwm most annoying of them
22:37:27 <ehird> and a stack of secondary windows to the right
22:37:35 <ehird> AnMaster: The adjustment period is just a few minutes, just stop trying to fight the WM
22:37:36 <fizzie> Uh, that's awesome's default layout too.
22:37:45 <ehird> awesome sucks in other ways
22:37:53 <pikhq> I found the most irritating thing about xmonad was it's freaking crazy configuration scheme.
22:37:56 <AnMaster> ehird, oh yeah I tried ratpoison too.
22:38:03 <pikhq> And after that was its retarded default bindings.
22:38:05 <ehird> AnMaster: ratpoison is unusable
22:38:12 <ehird> it's tiling but it doesn't manage the windows for you!
22:38:17 <pikhq> (that conflict with *everything else*)
22:38:17 <ehird> so it's just a pointless waste of time
22:38:27 <ehird> AnMaster: One unique one to try might be wmii
22:38:36 <ehird> It's basically Plan 9's acme for window managers
22:38:57 <ehird> there's the little knob on the window you can drag to move across columns, resize, etc; you can stack windows so it's just title bars that collapse, it's also controllable by the keyboard
22:39:03 <AnMaster> ehird, hm. interesting. Me and acme never agreed with each other though
22:39:13 <ehird> AnMaster: It's not acme as far as mouse-only goes
22:39:17 <ehird> just as far as layout goes
22:39:25 <ehird> http://wmii.suckless.org/
22:39:30 -!- Jaykul has changed nick to Jaykul[AFK].
22:39:32 <ehird> and it uses dmenu like dwm
22:39:41 <ehird> so spawning applications, etc is nice
22:40:00 <AnMaster> ehird, I like a desktop filled with icons though
22:40:19 <SimonRC> wait, people have visible desktop past all their windows?
22:40:21 <ehird> It's hard to stop liking gratuitous eye-candy
22:40:22 <AnMaster> stuff like C99.pdf POSIX.1-2008.pdf
22:40:33 <AnMaster> SimonRC, atm I can see a bit of that
22:40:34 <ehird> AnMaster: you can still use a graphical file manager
22:40:50 <AnMaster> ehird, I don't use a file manager *except* for desktop
22:41:07 <ehird> I never use my desktop
22:41:07 <pikhq> Isn't that what a shell's for?
22:41:18 <AnMaster> "A Source Mage spell for the 20070516 wmii snapshot is available. As usual, just type
22:41:19 <ehird> I'd like a hybrid shell/file manager
22:41:32 <ehird> a graphical file viewer above it, I can double click folders to cd
22:41:38 <ehird> but keyboard always enters commands
22:41:45 <ehird> so I can stop doing ls all the time
22:41:47 <AnMaster> ehird, try sourcemage. Just for fun in a vm
22:41:54 <ehird> AnMaster: nooo thank-you
22:42:06 <SimonRC> ehird: how about keyboard to search the file list
22:42:19 <ehird> SimonRC: maybe tab completion will be done with the graphical list
22:42:41 <ehird> instead of below the command entry
22:42:47 <AnMaster> ehird, it is like LFS in a way. A great experience, but not something you would want to use every day. Quite like going to some far away country as a tourist I expect
22:42:49 <ehird> AnMaster: because the terminology is enough to put me off
22:42:51 <ehird> and it's source-based
22:43:02 <AnMaster> you wouldn't want to live in Egypt. But seeing the pyramids once. Fun
22:43:08 <SimonRC> the feature I would like in bash is the ability to type half a command, do something else, and come back, like you can in irssi
22:43:17 <AnMaster> (not that I have been there. Only extrapolating from going to other countries)
22:43:25 <ehird> SimonRC: I usually hit enter and ctrl-c in quick succession for that
22:43:33 <ehird> it goes in the history
22:43:44 <AnMaster> ehird, that is why I said VM :P
22:43:51 <ehird> Ctrl-a ;# ctrl-a command enter
22:44:17 <AnMaster> the enter and ctrl-c is just too risky
22:44:18 <ehird> gah, why did people stop using bluecurve
22:44:30 <ehird> you should be concentrating when doing risky things.
22:44:31 <AnMaster> ehird, bluecurve is a theme for what?
22:44:41 <ehird> metacity, gnome, kde, ...........
22:44:44 <ehird> everything, basically
22:44:46 <ehird> think old redhat theme
22:44:56 <ehird> http://www.ensode.net/images/tiger_bluecurve.png
22:44:57 <AnMaster> ehird, well not risky. But "quite annoying to have to fix it"
22:45:02 <ehird> ignore the window contents
22:45:06 <ehird> it's java swing crap
22:45:07 <AnMaster> like ./configure long line here
22:45:23 <ehird> http://sqladmin.sourceforge.net/images/bluecurve.png
22:45:32 <ehird> older version of iwndow title, butt window contents is the same
22:46:16 <ehird> up until like 2005 i think
22:46:28 <AnMaster> ehird, pretty sure redhat 5 didn't have it. But could be wrong
22:46:40 <AnMaster> or at least. it wasn't default there
22:46:55 <ehird> Fedora Core 4 dropped bluecurve window border, 5 dropped theme
22:47:02 <ehird> So add a few years onto that and that's when redhat dropped it :P
22:47:16 <ehird> 4 was released june 05
22:47:31 <ehird> you know... I'm pretty sure a regular install of my distro will have no GNU software at all
22:48:04 <AnMaster> ehird, you are going to write an a.out backend for llvm then?
22:48:13 <ehird> SimonRC: good point, I'm not sure which to use
22:48:19 <ehird> AnMaster: not sure I'll use rc for command interpreter
22:48:22 <ehird> it's lacking in several areas
22:48:29 <AnMaster> ehird, hm. zsh is way to bloated for you
22:48:31 <ehird> for scripts around the system, though, definitely
22:48:37 <ehird> zsh is as bloated as bash
22:48:43 <pikhq> SimonRC: ... You mean bash doesn't have a kill buffer?
22:48:45 <ehird> or whatever it's called
22:48:53 <AnMaster> ehird, like, it has mmap module and what not
22:48:56 <ehird> the korn shell derivative
22:49:26 <ehird> Its weak points are that there are still a few differences from ksh88 (the major one is that `echo hi | read x' does not set x in the current shell - the read is done in a separate process).
22:49:28 <SimonRC> not quite as easy as hitting down in irssi though
22:49:37 <ehird> although really, use rc for scripts like that
22:49:57 <ehird> if pdksh has decent globbing, filename completion, and good variable expansion kind of things...
22:49:59 <SimonRC> how about a concatenative shell? RPN and all that
22:50:01 <ehird> then i might consider it
22:50:04 <AnMaster> huh. I never messed much with readline stuff
22:50:09 <ehird> rc is definitely the thing for scripts though!
22:50:32 <SimonRC> the bash man page explains all the keys you can hit
22:50:34 <pikhq> AnMaster: Readline implements most of the Emacs bindings.
22:50:52 <ehird> You are about to say "csh".
22:50:55 <ehird> You may: cower in fear
22:51:02 <ehird> AnMaster: you didn't mean that did you
22:51:06 <ehird> you did not just say yay tcsh
22:51:08 <ehird> you were sarcastic
22:51:16 <SimonRC> I was naming random shells
22:51:23 <AnMaster> <ehird> AnMaster: you didn't mean that did you <-- no
22:51:37 <AnMaster> <ehird> you did not just say yay tcsh <-- yes I did, but in a sarcastic way
22:52:03 <AnMaster> ehird, what about that original shell
22:52:05 <SimonRC> oh, wait that's not publically available
22:52:10 <ehird> http://cygwin.com/packages/posh/posh-0.6-1
22:52:11 <fizzie> I used to use tcsh; I think I switched to bash mostly out of laziness; too lazy to deviate from the norm.
22:52:32 <fizzie> I have a feeling tcsh is/was the default shell at some of the university systems.
22:52:58 <ehird> "not publically available"
22:52:59 <SimonRC> ehird: um, a shell. Don't want to say much more
22:53:01 <ehird> He probably can't say
22:53:14 <AnMaster> SimonRC, something used at work? You have your own?
22:53:14 <fizzie> [23:52:55] htkallas@kosh ~> ^D
22:53:14 <fizzie> Use "exit" to leave tcsh.
22:53:14 <fizzie> Yes, seems to be there.
22:53:17 <ehird> SimonRC -(soul)-> The Man
22:53:24 <ehird> SimonRC <-($$$)- The Man
22:53:31 <SimonRC> nah, I just made it up to sound more mysterious
22:53:36 <ehird> SimonRC: suuuuuuuuuuuuuuuuure
22:53:47 <ehird> I wonder if there's a compatible alternative to ncurses; not that I really care, since it's only barely a GNU project and MIT-licensed
22:53:56 <ehird> but it would be fun to have a viable linux distro with no gnu software by default
22:54:02 <Gregor> <TheMan> NOM NOM NOM SOUL
22:54:27 <ehird> Optimal Fracturing File System.
22:54:33 <Gregor> ehird: pdcurses only does X11 on Linux IIRC, but maybe that's changeable.
22:54:35 <ehird> It intentionally introduces beneficial fragmentation.
22:54:52 <ehird> OTOH, ncurses apps are sufficiently fucked-up that maybe making them use X11 is good :P
22:55:39 <ehird> the command line is for command UIs, ncurses programs are point-and-click WIMP UIs
22:55:45 <ehird> or rather, tap-and-click
22:55:50 <ehird> certainly not command-line, anyway
22:55:56 <AnMaster> ehird, in X they support mouse
22:56:06 <ehird> They support mouse with xterm too
22:56:12 <AnMaster> ehird, yes that is what I meant
22:56:27 <ehird> Correcting corrections FTW
22:56:34 <Gregor> And pdcurses+X actually works pretty well w/ mouse :P
22:56:45 <ehird> pdcurses seems pretty cool then
22:57:02 <ehird> does it come with an example program? arch might have a package
22:57:05 <AnMaster> ehird, I always keep reading "ftw" as "fuck the what"
22:57:07 <Gregor> Looks like it supports SDL too, and SDL supports FB consoles, so if you don't want X you could be sneaky that way :P
22:57:19 <ehird> PDCurses is a public domain curses library for DOS, OS/2, Win32, X11 and SDL, implementing most of the functions available in X/Open and System V R4 curses. It supports many compilers for these platforms. The X11 port lets you recompile existing text-mode curses programs to produce native X11 applications.
22:57:25 <ehird> Gregor: I want X :P
22:57:29 <ehird> AnMaster: I thought it meant that for years
22:57:36 <Gregor> Well, then, problem solved.
22:57:52 <AnMaster> ooh lament, lament is speaking.
22:57:53 <olsner> ooh, curses library that has a native X11 backend :>
22:58:02 <ehird> that's what we're discussing
22:58:34 <olsner> oh, I thought you were discussing the feasability of making a fully non-gnu linux distribution
22:58:41 <ehird> olsner: well, that too; mine is almost there
22:58:51 <AnMaster> ehird, where can I get the ISO?
22:58:52 <ehird> apart from ncurses and, uh, that's about it
22:58:56 <ehird> (without intending to do it)
22:58:57 <ehird> AnMaster: nowhere yet.
22:59:03 <ehird> but I've almost completely designed it
22:59:34 <AnMaster> don't* think you will ever implement it
22:59:44 <ehird> Yes I will, considering I'm switching to it
23:00:04 <Gregor> I would like to see no-GNU/Linux.
23:00:04 <ehird> I don't want to stay on OS X, but I cannot bring myself to use any of the current distros except maybe Arch
23:00:30 <ehird> Yeaah. vs the current crop of distros it's compelling enough to stay.
23:00:44 <ehird> But having everything work without fiddling is just boring.
23:00:51 <ehird> AnMaster: Because it can't even do USB afaik?
23:00:52 <AnMaster> <ehird> But having everything work without fiddling is just boring.
23:01:02 <ehird> It's the nerd in me.
23:01:10 <ehird> Making a distro is a nice way to solve that.
23:01:19 <AnMaster> ehird, this is not ehird. Please bring him back. What have you done to him.
23:01:20 <ehird> Tinkering goes in the area where you have to tinker anyway (maintanence).
23:01:26 <ehird> AnMaster: I killed him.
23:01:28 <Gregor> It's also a nice way to get rid of that pesky sanity.
23:01:35 <AnMaster> ehird, it might be a good idea to do LFS first to learn some internals.
23:01:36 <ehird> Gregor: Wow, I have sanity?
23:01:54 <ehird> anyway, back to pdcurses. no more will we have to write our terminals based on obscure corners of vt100, crippling our text editing capabilities for those vagrant applications! In X we can manage the real windows again, instead of our morass of terminals! LIBERATE THE UNKNOWLEDGABLE PROGRAMS! VIVA LA REVOLUCION!
23:02:07 <ehird> when (not if) X11 breaks...
23:02:11 <ehird> and i'm dumped to a console...
23:02:16 <ehird> Can't connect to display
23:02:19 <ehird> Can't connect to display
23:02:31 <ehird> OKAY NCURSES IT IS
23:02:32 <AnMaster> ehird, you forgot emacs in that list
23:02:40 <ehird> AnMaster: No GNU stuff, remember? :P
23:02:41 -!- FireFly has joined.
23:02:42 <AnMaster> ehird, also bash links to ncurses
23:02:59 <pikhq> ehird: He never said GNU Emacs. :P
23:03:15 <ehird> "GNUmacs? FUCK THAT SHIT! GOSLING EMACS!"
23:03:16 <AnMaster> ehird, or rather, readline links to ncurses. and that causes bash to also
23:03:24 <ehird> *GOS! LING! EMACS!
23:03:31 <pikhq> Or that Emacs-alike that Linus uses.
23:03:37 <pikhq> What AnMaster said.
23:03:50 <ehird> Gosling Emacs circa 1981/
23:04:00 <ehird> First Lisp-like extension language.
23:04:13 <AnMaster> ehird, what did it run on before
23:04:14 <ehird> Some code used in initial GNU Emacs.
23:04:22 <ehird> And ITS and stuff.
23:04:28 <AnMaster> ehird, ... I know but what did TECO run on I meant
23:04:31 <ehird> Basically, everyone who says rms invented Emacs?
23:04:39 <ehird> Has never heard of Gosling Emacs.
23:04:42 <fizzie> "Its extension language, Mocklisp, has a syntax that appears similar to Lisp, but Mocklisp has no lists or other structured datatypes." Heh, Lisp with no lists sounds like a winner.
23:04:50 <ehird> and Bill Joy, another java guy, made vi...
23:04:56 <ehird> THAT'S why Java sucks
23:04:57 <pikhq> ehird: I could've sworn rms wrote TECO Emacs.
23:05:04 <ehird> the vi inventor and the Gosling Emacs inventor
23:05:07 * SimonRC wonders where Eclipse came from then
23:05:10 <ehird> Why didn't I think of this before?!
23:05:19 <ehird> pikhq: the point is that gosling emacs pioneered what we know as emacs
23:05:30 <ehird> and TECO emacs wasn't anything like that at all
23:05:38 <pikhq> Rms saw Gosling Emacs and felt that it had a lot of good ideas.
23:05:41 <ehird> I'll just assume nobody got my reference, incidentally
23:06:05 <ehird> "GNUmacs? FUCK THAT SHIT! GOS! LING! EMACS!"
23:09:29 <ehird> http://www.youtube.com/watch?v=snhiofL2Rh4
23:11:05 <ehird> guess who else didn't get references? Hitler!
23:11:29 -!- Cerise has changed nick to Jerry.
23:11:56 <ehird> Guess who else invoked Godwin's law just by existing?
23:12:45 <AnMaster> wait. Is this some law about mentioning godwin's law?
23:12:45 * SimonRC recalls blog posts comparing Hitler to BO.
23:13:02 <ehird> http://obamaisliterallyhitler.tumblr.com/
23:13:11 <ehird> Barack Obama is literally Hitler.
23:13:51 <ehird> http://16.media.tumblr.com/tumblr_kqebfvzCxB1qzniowo1_500.jpg
23:15:35 <AnMaster> <ehird> http://obamaisliterallyhitler.tumblr.com/ <-- is that conservative or is it joking with conservative
23:15:54 <ehird> This person totally believes that Barack Obama is literally the same person as Hitler. Yep.
23:16:07 <ehird> And also believes that that comic is actual proof of it.
23:16:16 * SimonRC liked the Onion video about how Obama plans to deal with a raging wildfire
23:16:20 <AnMaster> ehird, a nutcase then? Or you being sarcastic :P
23:16:25 <AnMaster> ehird, anyway I didn't say that
23:16:29 <ehird> I'm absolutely not sarcastic, how can you even suggest that
23:16:53 <SimonRC> I am way left by american standards, yet I found that vidoe funny. Odd.
23:17:15 <ehird> SimonRC: Pretty sure the Onion staff is too.
23:18:06 <ehird> Anyone who can't laugh at themselves is an idiot, so anyone who can't laugh at someone else, no matter who they are, is too.
23:18:36 <AnMaster> ehird, I didn't hear what they said clearly in that video.
23:18:54 <AnMaster> <ehird> http://www.youtube.com/watch?v=snhiofL2Rh4
23:19:32 <ehird> "What kind of beer do you like?" "(muffled)" "HEINEKEN? Fuck that shit! PABST! BLUE! RIBBON!"
23:20:47 <ehird> What kind of bear do you like?
23:24:09 <ehird> I wish writing an init had some kind of hidden complexity instead of just being sundry C code.
23:24:42 <ehird> /bin/kernel calls /bin/init calls /bin/login, la la la
23:25:03 <ehird> Wonder what login(1) usually is. Doesn't look like GNU from the manpage.
23:26:17 <ehird> AnMaster: Incidentally, a cool thing about dwm and wmii:
23:26:22 <SimonRC> is this an attempt to create a linux distro that is not "GNU/LINUX"?
23:26:23 <ehird> The different desktops are actually tags
23:26:26 <ehird> A window can be on more than one tag
23:26:39 <ehird> SimonRC: No, it's an attempt at creating a minimalist distro focused on simplicity and usability
23:26:42 <ehird> To a radical degree
23:26:54 <ehird> Static binaries only, a.out, no kernel modules, non-glibc libc...
23:26:59 <AnMaster> ehird, guess what. They can in kde3 too
23:27:10 <SimonRC> won't the binaries be quite ... big?
23:27:16 <ehird> AnMaster: Ah, but is it easy to apply them? In dwm it's just Mod+Shift+number.
23:27:35 <ehird> SimonRC: Nope. Most of the binary size when using static binaries is caused by the huge glibc.
23:27:37 <SimonRC> I suppose busybox will cover a lot of stuff
23:27:44 <ehird> SimonRC: And static binaries only include the parts they use.
23:27:53 <AnMaster> ehird, if you want "all desktops" or "one desktop" yes. if you want "some desktops" it is incredibly complex
23:28:22 <ehird> SimonRC: Plus, it's faster to execute statically linked binaries, and they're completely immune to ABI changes. Besides... look at the dependencies of some package in your distro.
23:28:28 <ehird> See all those lib* dependencies?
23:28:36 <pikhq> SimonRC: No, but it's not like it's *hard* to create a non-GNU Linux distro.
23:28:37 <ehird> On the whole, a program using a lot of libraries may actually end up smaller.
23:28:47 <pikhq> A busybox/uclibc based system would just about do it.
23:29:03 <ehird> Oh, I forgot one: init system just based on simple rc shell scripts
23:29:39 <ehird> Anyway, non-GNU isn't specifically a goal, but it turns out that GNU software is basically the opposite of simplicity, minimalism, usability and Unix.
23:30:01 <ehird> So I end up simply not wanting to add much GNU software at all
23:31:59 <ehird> I'm not sure I'll be using busybox
23:32:18 <ehird> I like smallness, but I don't like how everything is in one binary, and the tools often seem overly barebones just to eke out the last bytes.
23:32:37 <pikhq> Busybox is most useful if you want a really absurdly barebones base install.
23:33:08 <pikhq> (like, say, /bin/kernel, /bin/init, /bin/rc, /bin/busybox...)
23:33:34 <ehird> I should totally do a hack so that you can execute /bin/kernel from inside it :P
23:33:45 <ehird> pikhq: Is that /bin/rc the startup binary?
23:33:47 <ehird> As opposed to a script
23:33:54 <pikhq> ehird: The rc shell.
23:34:01 <ehird> Busybox has a shell :P
23:34:17 <pikhq> True. I only put that in because you wanted init to be in rc. :P
23:34:53 <ehird> Well, not /bin/init, but that'll be tiny.
23:35:05 <ehird> Most everything will be handled by /etc/rc.{stop,start}.
23:35:09 <ehird> which /bin/init will call
23:35:21 <SimonRC> is it possible to make init a shell script?
23:35:25 <ehird> No inittab or run levels or anything
23:35:29 <ehird> SimonRC: I'm not sure
23:35:34 <pikhq> Sounds a lot like Gentoo's init setup (except that it uses a bog-standard sysvinit)...
23:35:37 <ehird> SimonRC: I think so; the kernel just uses a regular exec, I believe
23:35:55 <ehird> pikhq: Well, sysvinit is a lot more complicated than just handling signals and calling shell scripts.
23:36:22 <pikhq> ehird: All the inittab does in Gentoo is start /sbin/rc for the runlevels.
23:36:25 <ehird> init will be a binary because you need to be able to kill it to shut down and the like
23:36:30 <ehird> and I want to make sure it won't die
23:36:41 <ehird> it'll be like 1 KiB though
23:36:51 <pikhq> (which are "shutdown", "single", "nonetwork", "default", and "reboot")
23:36:57 <ehird> Incidentally, you could make /etc/rc.{start,stop} in Python or whateverr
23:37:03 <ehird> Just change the shebang
23:38:22 <ehird> OTOH, busybox gets me a lot of tools that work with Linux like mount
23:38:29 <ehird> whereas, say, porting a BSD's tools would be more work
23:39:00 <ehird> pikhq: No runlevels in my system! >:)
23:39:02 <SimonRC> brainfuck lacks the power to spawn new processes alas
23:39:23 <ehird> If you want an X login manager make /bin/login a shell script that starts x or whatever
23:39:29 <pikhq> ehird: So, more like Busybox init.
23:39:39 <ehird> Otherwise just run startx after logging in, or even condition on
23:39:43 <ehird> "are we in a console?"
23:39:50 <ehird> ofc that breaks consoles :P
23:39:52 <pikhq> (though that does handle an inittab. Just no runlevels.)
23:40:10 <ehird> You could make a /bin/login shell script that starts x and then runs the old /bin/login, of course
23:40:27 <ehird> I'll probably have a ready-made file to do that
23:40:31 <ehird> so you can just do
23:40:33 <SimonRC> Why couldn't init just be a shell script that calls a list of stuff to start in sequence?
23:40:39 <SimonRC> easy to maintain, very simple
23:40:44 <ehird> SimonRC: that's what it is!
23:40:58 <ehird> it's just that init has to run constantly, and my shutdown procedure involves sending signals to init
23:40:59 <SimonRC> so you can disable things just by commenting out lines?
23:41:30 <ehird> my /bin/init will basically set up any infrastructure needed, run /bin/rc.start, and then sit there waiting for the right signals (shutdown, restart, suspend to RAM, etc)
23:41:40 <ehird> in which case it'll act appropriately; for shutdowns and reboots, it'll run /bin/rc.stop first
23:41:42 <ehird> before halting or rebooting
23:41:52 <ehird> /etc/rc.{start,stop{
23:42:17 <SimonRC> every time you add a new thing to run in startup, you can just re-create init, e.g. using "tsort"
23:42:24 <ehird> No need to recreate init
23:42:26 <SimonRC> that way /etc/rc.* are not needed
23:42:59 <SimonRC> to try out something no other distro does?
23:43:13 <ehird> Seems like it'd just make changing the init stuff fussy for no benefit.
23:43:19 <SimonRC> as a simpler way to get startup dependancies?
23:43:25 <ehird> But no distro has an init system as simple as mine, without any inittab or anything
23:43:31 <ehird> you're just proposing embedding the script, right?
23:43:58 <ehird> You have not even adequately explained your proposal...
23:44:02 <ehird> So I don't understand.
23:44:13 <SimonRC> I haven't really thought about it...
23:44:30 <SimonRC> so, there is a list of things that must be run at startup?
23:44:32 <ehird> tsort is a good idea, though
23:44:35 <ehird> SimonRC: and shutdown
23:44:37 <SimonRC> and this list changes occasionally?
23:44:50 <ehird> Note that the scripts may have logic
23:44:56 <ehird> Since starting stuff can be non-trivial
23:45:14 <ehird> if (file exists /dev/foo)
23:46:36 <SimonRC> But init need not re-determine what order to run things in every time the system starts. It might be nicer to "compile" the list of things to call in the right order with the maximum parallelism into a new /bin/init every time the list of things present changes
23:46:46 <SimonRC> or that might just turn into a nightmare
23:47:08 <ehird> Yes, the dependency parallelism thing is a good idea and a good justification for it
23:47:14 <ehird> But a shell script is a lot simpler
23:47:23 <SimonRC> but it can be a shell script
23:47:27 <ehird> And my distro is minimalist enough that boot to X will take, what, 2 seconds?
23:47:39 <ehird> So it seems like a lot of complexity for a slightly faster boot is a waste.
23:47:52 <ehird> Especially as this way is much easier to maintain; just slot it in the right place.
23:48:28 <SimonRC> of course, init would need to be recreated every time you added or removed a demon, which might be tricky
23:48:38 <ehird> Not tricky, just a pain.
23:51:18 <ehird> SimonRC: Incidentally, know any kernel patches for good suspend/hibernate/resume support?
23:51:30 <ehird> LinuxOnIce looked promising but its suspend-to-disk mode is unbelievably slow from the looks of it
23:52:34 <ehird> Come to think of it, you don't need to edit /bin/login to get an X login manager, duh
23:52:39 <ehird> Just edit /etc/rc.start and change
23:53:06 <ehird> but you get the idea
23:56:00 <ehird> PATH=~/bin:/local/bin:/bin will be pretty cool to have.
23:56:20 <ehird> Although I think it should just be PATH=/bin by default.
23:58:31 <SimonRC> using those overlaid dirs, PLan9-style?
23:58:43 <ehird> Nope, although I might consider having something like that.
23:59:11 <ehird> It's just that I know I'll have no /usr (pointless directory), /local will be uncommon due to there being, you know, packages, and ~/bin isn't that common either (besides, it's easy to add them)
23:59:20 <ehird> All distro binaries and packages you make's binaries go into /bin