03:00:55 -!- clog has joined.
03:00:55 -!- clog has joined.
04:13:58 -!- oerjan has quit ("Good blight").
05:06:52 -!- pikhq has joined.
05:20:26 -!- ab5tract has joined.
05:27:27 -!- ab5tract has quit.
05:40:17 -!- ab5tract has joined.
06:03:24 -!- ab5tract has quit.
06:17:29 -!- Sgeo_ has quit (Read error: 110 (Connection timed out)).
06:23:35 -!- psygnisfive has joined.
06:43:55 -!- FireFly has joined.
07:05:09 -!- Slereah has joined.
07:17:21 -!- Slereah_ has quit (Read error: 110 (Connection timed out)).
07:37:22 -!- olsner has joined.
07:57:16 -!- pikhq has quit ("leaving").
07:59:59 -!- clog has quit (ended).
08:00:00 -!- clog has joined.
08:02:54 -!- FireFly has quit ("Later").
08:07:48 <AnMaster> ehird, I got a reply on one of the mails to abuse@, automated one though, but saying "All notifications will be investigated within 1 business day"
08:27:42 <Ilari> AnMaster: BTW, sometimes when spamming the real sender isn't the one that has the srcIP (but srcIP still points to zombie under spammer control). As result system with 56k modem can appear to send at multi-megabit speeds...
08:28:28 <AnMaster> Ilari, I know about drones, but then the isp should make the user fix it
08:28:41 <AnMaster> and it was irc spam bots, so no forged senders.
08:29:30 <AnMaster> Ilari, anyway I'm pretty sure it was automated infected computers even.
08:29:34 <Ilari> AnMaster: OTOH, with IRC, one doesn't need lots of bandwidth for spamming.
08:30:00 <AnMaster> Ilari, indeed, and the auto reply was for the registrar of the domain that was spammed (which was used to host a trojan download)
08:31:11 <AnMaster> Ilari, so don't think the exact same rules apply there.
08:32:20 <Ilari> AnMaster: Guessing TCP/IP seqno's is already quite hard in modern systems.
08:33:19 <AnMaster> Ilari, that would prevent two way communication though?
08:33:43 <AnMaster> Ilari, and irc server is setup such that it needs a PONG reply from the client before accepting connection
08:34:07 <AnMaster> Ilari, oh btw one of the spambots connected from an ip which whois claimed belonged to Cogent's main office.
08:34:48 <Ilari> AnMaster: The same tricks that are sometimes used to send spam appearing to come from zombies presumably would work. But why go to that trouble as you can't use the bandwidth.
08:35:33 <AnMaster> Ilari, most likely infected computers were used. A botnet.
08:35:50 <AnMaster> considering the great number of different ips and isps
08:35:51 <Ilari> AnMaster: I say Infected computers definitely involved.
08:36:27 <AnMaster> just wonder what that signifies
08:39:45 <Ilari> For DOS attacks, some lecturer once (yeas ago) said that TCP/IP has no protection against fradulently sent ACKs. Send those and watch target use massive amounts of upstream...
08:40:45 <AnMaster> Ilari, and it wasn't DOS, just "<bot> Download <random well known windows software> at http://trojans.are.us/"
08:40:53 -!- M0ny has joined.
08:40:54 <AnMaster> well not that url, but lots of random ones
08:41:08 <Ilari> AnMaster: And that fake-ACK attack doesn't work against ircd.
08:41:25 <AnMaster> Ilari, it requires confirmation after all
08:41:48 <Ilari> AnMaster: Would probably work nicely against httpd.
08:42:08 <AnMaster> yes but it wouldn't result in the observed behaviour in this case either.
08:42:44 <Ilari> AnMaster: Yeah, in this case, its spim bots involved, sending messages.
08:45:25 <Ilari> The reason why it won't work agaist ircd is that link is in effect rate-limited by the server.
08:47:25 -!- ais523 has joined.
08:56:11 <ais523> note to #esotericers: I'm in the middle of a distro upgrade atm
08:56:19 <ais523> so if I suddenly disappear, you know what's happened
09:06:16 <ais523> Ubuntu from Intrepid to Jaunty beta-and-almost-release-candidate
09:06:18 <AnMaster> ais523, why would distro upgrade cause that?
09:06:33 <ais523> this is Ubuntu I'm distro upgrading, not Debian
09:06:44 <AnMaster> <insert rolling release is better comment here>
09:06:46 <ais523> it has a tendency to go wrong
09:06:55 <ais523> although each time, it's been a trivial fix
09:07:00 <ais523> and two out of three times has been my fauly
09:07:00 <AnMaster> ais523, I had debian upgrades go wrong on me too
09:07:16 <AnMaster> which is why I prefer rolling release
09:07:24 <AnMaster> because upgrades usually just work then
09:07:58 <ais523> well, Debian does a distro upgrade by temporarily switching to rolling release, in effect
09:10:59 <AnMaster> ais523, well I was trying to do old stable -> stable and it wanted to remove glibc, apt-get and aptitude
09:11:19 <ais523> well, this distro upgrade wants to install Flash
09:11:23 <ais523> presumably it's in a recommends somewhere
09:11:29 <ais523> but I'm just going to uninstall it again afterwards
09:11:59 <ais523> apt is like Gnome, IMO
09:12:11 -!- olsner has quit ("Leaving").
09:12:15 <ais523> it does a lot behind the scenes, but the actual UI is ridiculously simple and easy to use
09:12:17 <AnMaster> well apt or aptituide or dpkg or whatever
09:12:21 <ais523> I can understand why someone like you wouldn't like it
09:12:29 <AnMaster> ais523, it doesn't do what I want
09:12:38 <AnMaster> makes the user feel he/she isn't in control
09:12:54 <ais523> oh, I feel in control of it
09:13:08 <AnMaster> maybe it is just a question of how much you used it
09:13:17 <AnMaster> but I never felt like that with any other package manager
09:13:24 <AnMaster> not even rpm back on old old red hat
09:13:43 <AnMaster> never used yum, so that might be worse
09:14:04 <AnMaster> and of course there was rpm hell
09:14:22 <Ilari> Lots of behind the scenes? Like enough logic to solve a sudoku? :-)
09:14:26 <ais523> the general organisation of the debian package management structure is that dpkg works at packet level
09:14:36 <ais523> Ilari: has someone written a sudoku solver in apt?
09:14:42 <ais523> and that apt works at repo level
09:14:58 <ais523> I'm not all that surprised, I even worked out how to explain INTERCAL version numbering to it
09:15:13 <AnMaster> ais523, that is quite a feat yes
09:15:24 <ais523> including negative components and everything
09:15:47 <Ilari> From that, one can conclude that question wheither some package has satisfiable dependencies is NP-complete.... :-)
09:16:02 <AnMaster> ais523, hm didn't you use <number>:<real version> or something?
09:16:15 <ais523> that's only for major and minor releases
09:16:21 <ais523> for betas it's more fun
09:16:31 <AnMaster> ais523, how did you explain that then
09:16:51 <ais523> IIRC, it would be 29:0.~8.0.-2.0.29
09:16:55 <ais523> maybe with an extra char in there
09:17:00 <AnMaster> ais523, and what would that mean
09:17:09 <ais523> you have to convert the version number from sign-magnitude to ten's complement
09:17:09 <lifthrasiir> (iirc CLC-INTERCAL had a version of 1.-xx.-yy.-z...)
09:17:11 <ais523> for it to work properly
09:17:54 <AnMaster> what about versions with letters in them. Like: R12B-4
09:18:16 <AnMaster> since it is actually a simple format
09:18:30 <ais523> although it assumes that the letters are meant to go in alphabetical order unless you give it extra hints
09:18:35 <ais523> using tildes, normally, they have a special meaning
09:18:48 <AnMaster> ais523, actually R isn't ever changed I think.
09:19:16 <AnMaster> ais523, nor do I think B is changed except for pre-releases which are A, but it used to use C too
09:19:22 <ais523> "1:1.0~4pre1.-94.-2-2"
09:19:31 <ais523> CLC-INTERCAL's version number in Debian stable
09:19:32 <AnMaster> what do the ~ mean there though
09:19:53 <ais523> think of ~ as coming before the null string in alphabetical order
09:20:17 <ais523> as in, "ab~c" < "ab" < "ab.c"
09:30:07 <fizzie> Ooh, VirtualBox people seem to have released a 2.2.0 not long ago (well, a bit over a week) which adds, among other things, hw-accelerated OpenGL for Linux and Solaris guests (and not just Windows, like it used to be). Not that it probably works very well, but still.
09:30:32 <ais523> so now we can play our Linux games in Windows at decent speeds!
09:30:51 <AnMaster> fizzie, does that require special CPU support?
09:31:13 <ais523> "Tags: Software Development: Compiler, Interpreter, Games and Amusement: Toy or Gimmick, User Interface: Command Line, Role: Program, Scope: Utility, Works with: Source Code"
09:31:25 <fizzie> I don't see why it should; but I guess in general it's pretty slow without hardware virtualization supports.
09:31:53 <ais523> AnMaster: tags on clc-intercal in Debian stable
09:32:50 <AnMaster> ais523, anyway what about our Solaris games?
09:33:31 <AnMaster> err wait, those would be in java anyway so that would be pointless.
09:45:51 -!- psygnisfive has quit (Remote closed the connection).
10:10:16 <ais523> ok, dpkg just replaced itself, this is definitely really quite far past the point of no return
10:12:29 <Ilari> Hope some post-install script doesn't decide to error out, causing the whole process to error out, without setting up the packages... And that package can't be removed because apt-get refuses to work. And the repair tries to run that script and fails...
10:13:01 <Ilari> Had that happen to me (in rolling update)...
10:13:24 <ais523> Ilari: I've had (self-inflicted) similar problems in dist-upgrades before
10:13:42 <ais523> but dpkg is smart enough to be able to fix a mess like that all by itself, without needing any helper scripts or repo knowledge or anything
10:14:01 <ais523> having a borked dpkg might be problematic, I assume you'd have to download a new one as a tarball and run it manually
10:15:35 <Ilari> IIRC, the problem was broken package script in such place that it would cause dirty abort from both apt-get and repair process.
10:18:02 <Ilari> Don't remember what I did, but eventually I got that script to run so I could fix things.
10:26:50 -!- Gracenotes has quit (Remote closed the connection).
10:39:06 <AnMaster> Ilari, um wouldn't most distros allow a simple restore?
10:39:39 <Ilari> AnMaster: It tried to allow restore. But that restore ran the failing script and expected it to succeed.
10:39:50 <AnMaster> I mean, for the package manager on gentoo I usually do quickpkg before upgrading, resulting in a tarball of the old version which I could extract manually in worst case.
10:40:05 <AnMaster> sure it would need some cleaning up after, but it would work
10:40:13 <AnMaster> not that I ever needed to use that
10:40:36 <AnMaster> I'm surprised the same doesn't exist for other distros
10:41:44 <Ilari> Bit worse was when one kernel image was bad in way that made it unbootable. Luckily I had backup kernel image (earlier version) and could downgrade. Currently this comp has something like 5 kernel images available...
10:42:22 <ais523> this computer keeps something like the last 5 kernel images available too
10:42:29 <AnMaster> well I always build and install kernel manually and have a fallback kernel always
10:42:50 <AnMaster> but even for distro kernels I would expect one or two old versions available at all points of time
10:44:09 <AnMaster> personally I just keep two old kernels available, previous one and previous version. With that I mean for example 2.6.26.whatever and 2.6.27.2 when current kernel is 2.6.27.3
10:44:11 <Ilari> 2 23s, 2 26s and a 29.
10:44:32 <AnMaster> I save all the .configs though
10:45:02 * AnMaster has .config files dating back from 2.6.9 or something like that around still
10:46:56 <AnMaster> anyway: are you seriously suggesting some distros do not have facilities for creating a manually extractable tarball of the currently installed version of a package?
10:49:12 <ais523> I know debian has one, it's called cp
10:52:24 <AnMaster> ais523, much more work for packages with many files in many places
10:52:25 <ais523> a .deb is just a .tgz with an extra directory inside it, which you can ignore
10:53:15 <ais523> it's in /var/cache unless you've cleared it out
10:53:35 <AnMaster> well, wasted space most of the time
10:53:39 <fizzie> Normally aptitude keeps the .debs, and even "aptitude autoclean" only cleans out non-installed packages. Though "aptitude clean" clears the whole cache, I guess.
10:54:08 <AnMaster> Personally I only bother to run quickpkg before upgrading important packages such as package manager, glibc, coreutils, bash and so on
10:54:21 <AnMaster> on and python too of course, since the package manager is written in that on gentoo
10:54:42 <AnMaster> and yes python upgrades tend to work painlessly too
10:54:55 <fizzie> aptitude autoclean: "Freed 2644MB of disk space". Don't seem to run that one very often.
10:55:02 <AnMaster> in fact I never had issues with upgrades on any rolling release system
10:55:22 -!- GreaseMonkey has quit ("X-Chat -> http://xchat.org <- At least when I quit I don't look like a lamer").
11:09:24 -!- Dewi has joined.
11:25:51 <AnMaster> interesting, seems the freebsd bootloader is partly FORTH based...
11:25:59 <AnMaster> http://www.freebsd.org/cgi/man.cgi?query=loader&sektion=8
11:26:52 <ais523> I just don't have a comment on it
11:28:09 <AnMaster> fizzie, about sigsegv handling in jitfunge: how is it done? And/Or where are docs?
11:29:02 <AnMaster> I did look in /usr/src/linux/Documentation/ but couldn't find anything related
11:30:42 -!- neldoreth has quit ("Lost terminal").
11:34:13 -!- GiveMeMony has joined.
11:34:48 <fizzie> Uh... I'm not sure I've seen any docs. I did it with guesswork.
11:37:30 <fizzie> Basically I just register a SIGSEGV handler with sigaction + SA_SIGINFO in the sa_flags. Then the signal handler has the prototype "void handler(int sig, siginfo_t *info, void *context);" ... now here info->si_addr is the (virtual) address of the fault, and the context argument you can cast into a "ucontext_t*" type pointer from <ucontext.h>.
11:37:48 <fizzie> That's actually somewhat documented.
11:38:19 <fizzie> "If SA_SIGINFO is specified in sa_flags, -- receives the signal number as its first argument, a pointer to a siginfo_t as its second argument and a pointer to a ucontext_t (cast to void *) as its third argument." says my sigaction(2) man page.
11:39:20 <fizzie> On Linux the ucontext_t has a uc_mcontext member (of type mcontext_t) which has the machine registers and such.
11:39:26 <fizzie> At this point it goes very hardware-specific.
11:40:22 <AnMaster> fizzie, do you use setcontext() or such after fixing the values? or just do the jump manually?
11:40:24 <fizzie> I just read the stuff from <sys/ucontext.h>... there's symbolic names for the general-purpose registers in uc_mcontext.gregs[] array.
11:41:10 <AnMaster> it seems POSIX.1-2008 removed setcontext() hm
11:41:23 <AnMaster> well getcontext() and so on too
11:41:28 <fizzie> I just call setcontext() on the provided cast-to-ucontext_t* context, after fixing some of the register values in there (like EIP to to instruction after the fault and so on).
11:41:30 <AnMaster> that whole family of functions in fact
11:41:58 <fizzie> Hm. Are there any substitutes? And what do they say of the sigaction SA_SIGINFO-handler's third-argument in posix-2008?
11:42:33 <AnMaster> you could download the pdf yourself but let me look
11:43:33 <fizzie> Oh, it's available? I've just been reading my 1003.1-2001 PDF-copy.
11:43:35 <AnMaster> um it mentions ucontext_t there
11:44:30 <AnMaster> fizzie, yes you need to sign up (free) to the austin group mailing list, then opt out of the mailing list part of it, then download it. Registrations take a day or two to be processed iirc.
11:44:50 <fizzie> Well, I'll maybe try that at some point.
11:47:27 <AnMaster> fizzie, ok this is strange: SA_SIGINFO mentions ucontext_t, but the ucontext stuff is otherwise missing as far as I can see
11:50:11 -!- Gracenotes has joined.
11:51:56 -!- M0ny has quit (Read error: 110 (Connection timed out)).
11:56:26 -!- bsmntbombdood_ has quit (Read error: 60 (Operation timed out)).
12:21:54 -!- Gracenotes has quit (Remote closed the connection).
12:29:46 -!- bsmntbombdood has joined.
12:40:19 -!- GiveMeMony has changed nick to M0ny.
12:41:47 -!- Gracenotes has joined.
12:49:22 -!- MizardX has quit ("What are you sinking about?").
13:01:21 -!- bsmntbombdood has quit (Read error: 110 (Connection timed out)).
13:20:42 <Gracenotes> I should be taking a shower .. but I sit watching Pokemon o:
13:22:35 <ais523> well, I'm kind-of stuck atm
13:22:44 <ais523> in a distro update, I can't really do anything but websurf and chat on IRC
13:25:30 -!- Gracenotes has quit (Remote closed the connection).
13:28:03 -!- oerjan has joined.
13:29:02 <ais523> incidentally, this channel seems rather empty atm...
13:29:26 <ais523> this distro update is almost complete, anyway
13:29:30 * oerjan hugs kerlo too to be fair
13:29:33 <ais523> so I'll be rebooting soon and hoping the computer still works
13:31:55 <oerjan> ais523: nice to have known you
13:32:40 <ais523> this has never gone smoothly before, I think, but two times out of three when it failed before it was my fault
13:32:42 <oerjan> also, http://xkcd.com/349/
13:32:44 <ais523> and in each case, I figured out how to fix it
13:33:08 <ais523> don't paste links while I'm updating, I have an urge to click on them and that doesn't work
13:33:14 <ais523> I have to copy-and-paste into the browser
13:33:48 <ais523> also, I don't get how installing BSD would cause you to be attacked by sharks
13:33:49 <oerjan> well, it's a highly relevant link
13:34:03 <oerjan> ais523: let's hope you don't find out
13:34:16 <ais523> good thing I'm not installing BSD, then, I suppose
13:35:25 <oerjan> also, http://blag.xkcd.com/2009/03/27/a-brief-pair-of-notes/
13:36:41 <ais523> rendering a system unbootable while trying to change a setting in a webform is pretty impressive...
13:39:04 <ais523> ah, there's an explanation
13:39:15 <ais523> and really, he rendered the system unbootable by messing up the repository lines
13:39:31 <ais523> that was just a delayed-action unbootable-rendering, which caught up with him when he tried to change the setting
13:40:35 <oerjan> your awesome reasoning powers won't save you when murphy's law hits!
13:43:41 <ais523> ok, it just hit the libc trigger
13:43:52 <ais523> things are going to get very interesting very soon
13:44:09 <ais523> (libc trigger = the last thing in most large package replacements, although this one was so big it had other triggers to process too...)
13:47:56 <ais523> ok, I'm into very likely breakage at any moment mode now
13:58:58 -!- neldoreth has joined.
13:59:01 -!- ais523 has quit ("rebooting after distro upgrade, I hope I come back...").
14:07:50 -!- ais523 has joined.
14:08:21 * oerjan searches ais523 for shark bites
14:35:47 -!- ais523_ has joined.
14:36:34 -!- ais523 has quit (Read error: 104 (Connection reset by peer)).
14:38:00 -!- ais523_ has changed nick to ais523.
14:55:41 -!- M0ny has quit ("PEW PEW").
15:00:11 <GregorR> breakage = the consumption of breakfast.
15:07:09 -!- neldoreth has quit ("leaving").
15:12:15 -!- neldoreth has joined.
15:27:53 <AnMaster> I just found something amusing in a manual
15:28:23 <AnMaster> the manual is for an electrical piano, and as usual there is a caution section at the start, with the normal "avoid liquids" and so on
15:28:30 <AnMaster> there is also some odd stuff in it
15:28:36 <AnMaster> like: "Do not stand on this device"
15:28:53 <ais523> oh, all sorts of manuals have that sort of thing in
15:29:12 <ais523> there was a rather silly TV show in the UK which tried to find things that the manual didn't disallow, but nevertheless destroyed the device
15:29:16 <ais523> such as dropping it 100 feet
15:29:41 <AnMaster> I think "Do not drop or subject to strong impact" covers that
15:30:14 <ais523> most of them involved explosions IIRC
15:30:39 <ais523> I think this was a british attempt to emulate it
15:30:52 <AnMaster> ais523, but seriously, would anyone outside a silly TV show even consider standing on an electrical piano?
15:31:08 <ais523> I can sort of imagine how it might happen
15:31:56 <ais523> they wanted to reach a higher place and needed something to stand on?
15:32:15 <ais523> not everyone, especially children, don't realise that not everything supports their weight
15:32:16 -!- bsmntbombdood has joined.
15:32:34 <AnMaster> ais523, that assumes a high stupidity level thoug
15:32:48 <ais523> AnMaster: yes, manuals generally do
15:33:00 <AnMaster> even more than standing on a office chair with wheels
15:33:03 <ais523> sometimes you seem to underestimate the levels of stupidity some people can reach...
15:33:27 <AnMaster> ais523, well, considering the rest of the manual seems to assume a rather non-stupid person...
15:33:29 <oerjan> oh i saw this link in a reddit comment yesterday...
15:34:43 <oerjan> http://wwwcsif.cs.ucdavis.edu/~leeey/stupidity/basic.htm
15:34:47 <AnMaster> ais523, actually some of those warnings are interesting, like "always turn down volume control to zero before turning the device on/off to avoid damaging the speakers"
15:35:03 <ais523> does anyone ever obey them?
15:35:19 <AnMaster> ais523, which are you referring to?
15:35:25 <ais523> AnMaster: your warnings
15:35:38 <ais523> I mean, for instance, would you store a laptop in a fridge at 40% battery charge?
15:36:07 <AnMaster> ais523, that would damage the harddrive. If you took the battery out though it would be a good way for long time storage as far as I understood
15:36:30 <AnMaster> but I usually use devices most of the time so that thing never came up
15:36:37 * oerjan recalls some devices where the volume control is also the off button, avoiding the problem
15:37:46 <AnMaster> oerjan, wouldn't work in this case. Since you can turn another switch to turn off internal speakers and use external output (either headphone connection or line-out connection thingy with separate left/right connectors)
15:38:31 -!- FireFly has joined.
15:39:02 <oerjan> hm that would be awkward, although you _could_ do it with one button - make the off setting midway between volumes for either case
15:39:13 <oerjan> won't work for more than two speaker options though
15:39:25 <ais523> just use a multidimensional switch, then
15:39:49 <ais523> imagine a switch which went left, right, and up
15:39:54 <AnMaster> well there is a balance switch for that
15:39:56 -!- ehird has left (?).
15:40:09 <ais523> bye ehird? I didn't even realise you were here...
15:44:51 <AnMaster> ais523, hm about the "do not stand" one, I wonder to what degree that applies to the pedal that was included with this electric piano...
15:45:37 -!- ais523 has quit ("lunch").
15:48:55 -!- ehird has joined.
15:49:00 <ehird> 01:12:29 <AnMaster> ais523, it doesn't do what I want
15:49:02 <ehird> 01:12:38 <AnMaster> makes the user feel he/she isn't in control
15:49:06 <ehird> dpkg is a very low level tool
15:50:02 -!- bsmntbombdood has quit (Read error: 60 (Operation timed out)).
15:50:11 <ehird> 01:30:32 <ais523> so now we can play our Linux games in Windows at decent speeds!
15:50:15 <ehird> 01:30:51 <AnMaster> fizzie, does that require special CPU support?
15:50:20 <ehird> No. GPU support. :P
15:50:41 <ehird> 01:31:53 <ais523> AnMaster: tags on clc-intercal in Debian stable
15:50:41 <ehird> 01:32:50 <AnMaster> ais523, anyway what about our Solaris games?
15:50:42 <ehird> 01:33:31 <AnMaster> err wait, those would be in java anyway so that would be pointless.
15:50:48 <ehird> ITT: stunning ignorance of solaris
15:51:01 <ehird> 02:13:01 <Ilari> Had that happen to me (in rolling update)...
15:51:14 <ehird> "(in rolling update)", aka anti-AnMaster-troll spray
15:51:44 <ehird> 02:39:50 <AnMaster> I mean, for the package manager on gentoo I usually do quickpkg before upgrading, resulting in a tarball of the old version which I could extract manually in worst case.
15:51:45 <ehird> 02:40:05 <AnMaster> sure it would need some cleaning up after, but it would work
15:51:47 <ehird> 02:40:13 <AnMaster> not that I ever needed to use that
15:53:02 <ehird> 02:54:55 <fizzie> aptitude autoclean: "Freed 2644MB of disk space". Don't seem to run that one very often.
15:53:11 -!- neldoret1 has joined.
15:53:12 -!- neldoreth has quit ("leaving").
15:53:17 <ehird> It occurs to me that that is probably bad for SSDs.
15:53:20 <ehird> storing them, I mean
15:54:27 <ehird> 03:41:10 <AnMaster> it seems POSIX.1-2008 removed setcontext() hm
15:54:27 <ehird> 03:41:23 <AnMaster> well getcontext() and so on too
15:55:14 <ehird> {get,set}context were *almost* continuations.
15:55:16 <AnMaster> ehird, there are still references to them in sigaction()
15:57:07 <ehird> http://xkcd.com/350/ ← I'd like to note that one day I will set this up.
15:58:03 <ehird> The current one is #569.
15:58:26 <AnMaster> how would you rate the last strip btw?
15:58:34 <ehird> It was not particularly funny.
16:00:00 <AnMaster> ehird, far from the worst though
16:00:12 <ehird> This is probably my favourite xkcd comic: http://xkcd.com/24/ just because it's so unlike all the others.
16:05:13 <ehird> 1) Dvorak typists: I’m sorry if the paper I relied on has some potential flaws. If you want to share your extensive rants on the merits of various keyboard layouts, send them to me at doctorow@boingboing.net and I’ll be sure to read them over carefully.
16:05:43 <ehird> (The worst part was, everyone kept saying “oh yeah — there’s a comic about that; have you read it?”)
16:05:47 <ehird> bet he gets that all the time
16:06:34 <ehird> http://blag.xkcd.com/2009/03/27/a-brief-pair-of-notes/comment-page-3/#comment-26267 ← Hey, it's coppro.
16:06:40 <ehird> Is it that coppro? Probably.
16:11:00 <ehird> "I’ll just add this anecdote (minus 16) to my list of reasons to never use Ubuntu. "
16:11:05 <ehird> How to miss the point, part 9999.
16:11:12 <ehird> "Ubuntu sucks because you can add debian repos and break your system"
16:12:29 <ehird> 07:28:24 <AnMaster> but
16:12:29 <ehird> 07:28:30 <AnMaster> there is also some odd stuff in it
16:12:30 <ehird> 07:28:36 <AnMaster> like: "Do not stand on this device"
16:12:38 <ehird> old ipod shuffle site had "Do not eat iPod touch."
16:12:42 <ehird> Do not eat iPod shuffle
16:12:50 <ehird> http://www.engadget.com/2005/01/12/the-ipod-shuffle-do-not-eat-in-the-us-or-chew-in-the-uk/
16:13:21 <ehird> 07:29:12 <ais523> there was a rather silly TV show in the UK which tried to find things that the manual didn't disallow, but nevertheless destroyed the device
16:13:33 -!- M0ny has joined.
16:13:39 <ehird> 07:30:52 <AnMaster> ais523, but seriously, would anyone outside a silly TV show even consider standing on an electrical piano?
16:14:01 <ehird> 07:32:15 <ais523> not everyone, especially children, don't realise that not everything supports their weight
16:14:02 <ehird> 07:32:34 <AnMaster> ais523, that assumes a high stupidity level thoug
16:14:05 <ehird> children aren't stupid
16:14:59 <ehird> "The First Basic Law prevents me from attributing a specific numerical value to the fraction of stupid people within the total population: any numerical estimate would turn out to be an underestimate. "
16:15:05 <ehird> *universe explodes*
16:15:42 <ehird> 19:01:13 --- join: clog (n=nef@bespin.org) joined #tunes
16:15:43 <ehird> 19:01:13 --- names: list (clog elias` slava ehird cpfr matthewf levitation[A] BrianRice)
16:15:44 <ehird> 19:20:12 --- join: Fare (n=Fare@c-98-216-111-110.hsd1.ma.comcast.net) joined #tunes
16:15:46 <ehird> 19:27:31 <BrianRice> thank you, Fare
16:15:48 <ehird> 19:27:53 <Fare> sorry for the disruption
16:15:50 <ehird> 19:27:57 <Fare> a problem with screen
16:15:52 <ehird> 19:28:00 <BrianRice> huh
16:15:54 <ehird> 19:28:25 <Fare> /var/run/screen had the wrong group, etc. http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=471763 causing clog to not start properly
16:17:36 <ehird> 07:36:07 <AnMaster> ais523, that would damage the harddrive. If you took the battery out though it would be a good way for long time storage as far as I understood
16:17:41 <ehird> I think the ice would fuck with it.
16:17:50 <ehird> Try and cryopreserve it :P
16:19:39 <ehird> 07:40:09 <ais523> bye ehird? I didn't even realise you were here...
16:19:42 <ehird> my client parts on startup
16:19:47 <ehird> then joins when the bouncer stops logfucking it
16:23:54 <AnMaster> ehird, err, can't you tell your client not to part the channel even if client sends part, znc call this feature sticky channels, it is rather useful in case of misclicks/typos.
16:24:08 <AnMaster> but sounds like it could be useful for this too
16:24:10 <ehird> AnMaster: i don't really care, as I won't be using this client soon
16:24:22 <ehird> on account of it not running on linux
16:24:29 <ehird> (where I'll probably use xchat-gnome (!= xchat))
16:25:14 <ehird> (for the unknowing: xchat-gnome is an xchat frontend that uses gnome libs and generally sucks less)
16:30:45 -!- oerjan has quit ("snq!+ue").
16:34:40 <ehird> http://www.smbc-comics.com/comics/20090416.gif
16:35:27 -!- pikhq has joined.
16:42:05 <oklopol> wait the numerous statistics i've seen about how dvorak is better than qwerty are all urban legends?
16:42:24 <ehird> xkcd said okay, okay I was wrong a bit in his blahg.
16:42:50 <ehird> also the immortality thing is a crock of shit, you don't need immortality in 80 yrs
16:42:58 <ehird> you just need life extension in 80 years, so on until immortality
16:43:14 <ehird> and I think "we will be able to live longer within 80 years" as a prediction is almost certain
16:44:28 <oklopol> you mean xkcd is not only funny, it's become stupid too
16:44:34 -!- MizardX has joined.
16:44:53 <oklopol> please insert not, seems that was a bit too hard for me.
16:53:20 <AnMaster> <ehird> children aren't stupid <ehird> just naive <-- indeed, but the product was certainly not aimed at small children. As reading context would have shown
16:53:50 <ehird> I never said small children
16:56:02 <oklopol> my experience is they're pretty stupid too :|
16:56:26 <ehird> oklopol: I guess you have to have been one recently :P
16:58:09 <ehird> i hate trackballss..
16:58:15 -!- MizardX has quit ("What are you sinking about?").
17:03:31 <ehird> "The most common font reported for Unix family systems is the cursive font URW Chancery L at 95% frequency. Second position is held by the sans serif URW Gothic L, then DejaVu Sans Mono followed by serif font Century Schoolbook L. The monospace font Nimbus Mono L and complete the top 5 most common Linux fonts at 93% frequency."
17:03:35 <ehird> URW CHANCERY L FUCK YEAH
17:04:30 <ehird> http://www.codestyle.org/css/font-family/sampler-UnixResults.shtml
17:04:40 <Deewiant> Haven't even heard of those fonts, apart from DejaVu :-P
17:04:47 <Deewiant> But then I'm not much of a font person
17:04:50 <pikhq> I'm somewhat fond of DejaVu *, myself.
17:05:16 <ehird> I've used the Nimbus family; quite nice, e.g. Nimbus Sans makes a nice print font. DejaVu can be made to look nice.
17:05:22 <Deewiant> DejaVu being ahead of Bitstream Vera is a bit of a surprise, I thought it wasn't as widespread
17:05:23 <ehird> The rest are obscure rubbish :-P
17:05:35 <ehird> Deewiant: ubuntu ships with just dejavu, I think.
17:05:55 <Deewiant> I don't even have URW Chancery L installed
17:05:56 <ehird> Helvetica @ 50% is more than I would have thought
17:06:05 <AnMaster> ehird, about font, most systems have URW yes, but most systems actually use Bitstream Vera and/or Dejavu
17:06:06 <pikhq> DejaVu is libre, so just about every distro has it.
17:06:07 <Deewiant> Maybe it's another Ubuntu-default :-P
17:06:08 <ehird> that requires the dedication to pirate it from a windows version or convert a mac one :-P
17:06:38 <pikhq> And URW is installed because, well, it's been on Unix since the invention of X11. ;)
17:06:42 <ehird> It would be nice if Helvetica was libre. How dead is its maker, I wonder?
17:06:42 <Deewiant> I have URW Gothic but not URW Chancery
17:06:54 <ehird> Max Miedinger (1910-12-24 - 1980-03-08
17:07:00 <ehird> OK, couple of decades to go.
17:07:05 <AnMaster> pikhq, yes and it is like Bitstream Vera but supports more symbols
17:07:16 <ehird> Although, really, it's the font that's copyrighted, I imagine.
17:07:30 <ehird> 17:07 AnMaster: pikhq, yes and it is like Bitstream Vera but supports more symbols
17:07:33 <ehird> I am fairly sure pikhq knows this.
17:08:24 <ehird> AnMaster: You should be like zzo38 and talk like this then. Or not talk like zzo38 if you don't want to.
17:08:26 * AnMaster wonders if anyone got the reference
17:08:39 <ehird> HEY GUYS LOOK AT ME, I MADE AN OBSCURE REFERENCE
17:08:45 <ehird> HA, I AM MORE INDIE THAN THOU!
17:08:52 <Deewiant> Meh, I want teh URW cursive font
17:09:05 <ehird> http://research.cs.queensu.ca/Parallel/projects.html#Current ← Someone translate this for me please.
17:09:06 <AnMaster> ehird, it wasn't very obscure in a channel like this.
17:09:11 <Deewiant> ehird: Andale Mono is twice there, wtf?
17:09:27 <ehird> http://www.reddit.com/r/programming/comments/8cz51/alan_turing_was_wrong/c08wbdl
17:09:30 <AnMaster> I'm sure ais would get the reference
17:11:35 <ehird> As he walked into Columbine High School the morning of the massacre Eric Harris' T-shirt read: Natural Selection.[1] According to CNN, Harris wrote the following before the massacre: "Sometime in April [of 1999] me and V will get revenge and will kick natural selection up a few notches."[2] Eric Harris, who was a racist, had adulation for the evolutionary racist Adolph Hitler and on his website were four Hiel [sic] Hitlers.[3][4]
17:11:39 <ehird> — http://www.conservapedia.com/Evolution
17:11:43 <ehird> How I love Conservapedia.
17:12:08 <Slereah> Conservapedia on evolution isn't fun, ehird
17:12:21 <ehird> Slereah: But it's amusing.
17:12:44 <pikhq> Conservapedia was really fun when it had relativity deniers.
17:13:03 <ehird> pikhq: You mean quantum physicists? :P
17:13:36 <pikhq> Quantum physicists don't deny that it's a decent model at the scales Einstein could observe.
17:13:46 * ehird fiddles with MaxMSP
17:13:52 <pikhq> Conservapedia denied that.
17:14:13 <Slereah> What I love about Conservapedia is when it gets crazy on unrelated matters
17:16:15 -!- ais523 has joined.
17:16:47 <ehird> couldn't wait for the real thing? :P
17:17:06 <ais523> I was planning to get rc1, on the basis that it's likely to be the same as the finished version
17:17:15 <ais523> but the beta didn't have any known bugs that affected me
17:17:25 <ais523> which implies that any bugs in it that do affect me will be in rc1 too
17:17:58 <ehird> I'm waiting for the full release so I can get a nice CD of it.
17:19:02 <ehird> Some of the packages could not be retrieved from the server(s).
17:19:06 <ehird> Do you want to continue, ignoring these packages?
17:19:21 <ais523> network trouble? did you choose a mirror that didn't have all the packages/
17:20:00 <ehird> I'm installing it via Install/Remove
17:20:05 <ais523> anyway, if you choose no, nothing will be installed, if you choose yes, only the bits that can will be installed
17:20:13 <ais523> and my guess is it has dependencies on non-FLOSS stuff
17:20:26 <ais523> which therefore won't be installed unless you enable the non-FLOSS repos
17:20:37 <ehird> it could do with a better error
17:22:18 <ehird> ais523: ok, this is not goo
17:22:45 <ehird> Failed to fetch http://gb.archive.ubuntu.com/ubuntu/pool/main/f/ffmpeg-debian/libavdevice52_0.svn20090303-1ubuntu4_i386.deb
17:22:57 <ehird> then libavfilter and some ffmpeg stuff
17:23:08 <ais523> hmm... I know some mirrors don't have all packages
17:23:11 <ais523> but I'd expect that one to
17:23:21 <ehird> and ffmpeg isn't exactly OBSCURE.
17:24:11 <ehird> this is an issue with my system
17:24:30 <ehird> at least I hope so
17:24:30 <AnMaster> check if there is another version at http://gb.archive.ubuntu.com/ubuntu/pool/main/f/ffmpeg-debian/
17:24:43 <ehird> there there is but it doesn't depend on that version
17:24:58 <ehird> libavfilter0_0.svn20090303-1ubuntu6_i386.deb11-Apr-2009 01:04 43K
17:25:04 <ehird> ais523: what's the diff?
17:25:22 <ais523> that's the revision number of the Ubuntu-specific patches
17:25:25 <ais523> so basically, a different patchlevel
17:25:32 <ehird> ais523: so what should I do?
17:25:40 <AnMaster> ehird, try updating the package index or whatever ubuntu use
17:25:51 <ais523> my guess is that with all the rush to get the release out, they forgot to synch the index with the actual packages
17:25:59 <ehird> ais523: this is the beta, indeed.
17:26:28 * ehird $ sudo apt-get update
17:26:51 -!- MizardX has joined.
17:27:43 <ehird> Max has an actual event called "bang"
17:28:04 <ehird> I love this language
17:28:17 <ehird> ais523: you know how I said I wanted a graph language?
17:28:26 <ehird> max is one of them
17:28:32 <ehird> almost exactly as I imagined
17:28:57 <ehird> http://imgur.com/MRVIK.png
17:29:02 <ehird> that plays random notes on a piano
17:29:18 <ehird> the thing at the top is a toggle, it emits either on or off, pretty much
17:29:30 <ehird> metro 250 is a (silent) metronome with 250ms delay, I think
17:29:36 <ehird> random 128 is a random number from 0 to 128
17:29:45 <ehird> the three inputs of makenote are pitch velocity and duration
17:30:04 <ehird> and the first two inputs (and two outputs of makenote) of noteout are pitch and velocity
17:30:07 <ais523> reminds me of 2d a bit
17:30:19 <ehird> (the last input is lets you control the midi device it outputs to)
17:30:45 <ehird> I made that by double clicking, selecting object and typing what I want (with autocompletion) then it grew the inputs/outputs
17:30:52 <ehird> and I double clicked and selected toggle for the top one
17:30:59 <ehird> then locked the patch and clicked it on to start
17:31:18 <ehird> here's something a... little more involved: http://upload.wikimedia.org/wikipedia/en/7/72/Autechremax.jpg
17:31:27 <ehird> made by autechre, they use max in their music
17:32:49 <ehird> it costs $$$$$ unfortunately
17:38:14 <ehird> Does anyone know how to actually get the gpu to run something
17:39:25 <ehird> Deewiant: lower level.
17:39:33 <ehird> I mean actually writing ... whatever the equivalent of asm is.
17:39:37 -!- Gracenotes has joined.
17:39:54 <ehird> Then, somehow, telling the CPU to tell the GPU to get on that.
17:40:28 <ehird> Deewiant: isn't that limited to like actually doing stuff on screen though
17:40:34 <ehird> as opposed to calculating with boring numbers and whatnot
17:47:09 -!- bsmntbombdood has joined.
17:53:59 <GregorR> It tastes like matrix math, but it's as healthy as algebra!
17:58:00 <ehird> http://imgur.com/MPGBU.gif
18:00:11 <ehird> No! The answer is no!
18:00:44 <ehird> [[ Let me throw out an argument for a greater spirit of equity here. [If this
18:00:44 <ehird> fails, it is probably the last time I will ever believe that equity was
18:00:46 <ehird> possible here or pitch for its use]: ]]
18:00:48 <ehird> Just like the last time.
18:03:22 <Gracenotes> oh, goodness, Pokemon writers and their innuendo. "We went at it back in the day"
18:04:35 <Gracenotes> and the 'randomized' content pairings, they also pair up the most expected people for last
18:04:56 <ais523> Gracenotes: do you play pokemon, or just watch the cartoon?
18:05:07 <GregorR> It's almost as if ... the cartoon was written by humans!
18:05:08 <Deewiant> I don't think they even pretend that those are random :-P
18:05:11 <ais523> unfortunately I seriously doubt it's TC
18:05:33 <Gracenotes> the cartoon is entertaining.. I watched all 10 seconds last summer in a few weeks
18:06:03 <Gracenotes> since then I've been waiting a few months and watching the episodes I missed, repeating this. I don't watch it every week.
18:06:06 <GregorR> Then proceeding to watch all of another 10 seconds, then another 10 seconds, then another three seasons ...
18:06:29 <Gracenotes> I have played the games and beaten many of them
18:06:50 * GregorR tries to decide whether Pokelust is /ignore-worthy :P
18:06:57 * ehird attempts to run MaxMSP on Linux
18:07:06 <ais523> well, "beaten them" is rather unclear
18:07:25 <ehird> ais523: he means the gameboy games, probably
18:07:29 <GregorR> Oh, I beat them. I beat them with a baseball bat. Consider them thoroughly beaten.
18:07:47 <ais523> ehird: it's not clear what constitutes completing the gameboy games
18:07:59 <Gracenotes> by 'get everything' I mean capturing everything
18:08:10 <ehird> ais523: They're open-ended? I wasn't aware.
18:08:14 <ais523> in later versions, there are two sets of everythings to capture
18:08:27 <ais523> ehird: how could you not be aware of that?
18:08:34 <ais523> generally, completing the elite four unlocks a bit more of the map
18:08:36 <Gracenotes> oh, that's the later version, which has a backwards-compatible Pokemon-land.
18:08:39 <ehird> ais523: Not good enough to get far enough to realise that.
18:08:42 <ais523> so you can go and complete all the openendedness
18:08:42 <GregorR> I wasn't aware either, but then I've never played them :P
18:08:51 <ais523> Gracenotes: there have been at least two versions since then
18:09:22 <Gracenotes> are you talking about the Southern Islands in Firered/Leafgreen?
18:09:50 <Gracenotes> but, as I said, I don't play it that much, not competitively, just when bored.
18:09:56 <GregorR> ehird: Is this running something on wine?
18:10:12 <ehird> GregorR: Is that a joke?
18:10:44 <GregorR> ehird: Idonno, I haven't been following, I just saw you saying you were trying to get something called MaxMSP running on Linux, then that it didn't work.
18:10:54 <ehird> GregorR: Yes. That is correct.
18:11:10 <ehird> It's a driver issue.
18:11:14 <GregorR> Then no, that wasn't a joke :P
18:11:32 <ais523> Gracenotes: oh, I thought you were referring to the copy of the red and blue map in gold and silver
18:11:38 <ehird> This software requires installation of device driver TPkd and a reboot before running. Please reboot or reinstall the software.”
18:11:40 <ais523> but you have pal park and the battle park in the latest versions
18:11:44 <AnMaster> ehird, one of the abuse@ mails had success :)
18:11:49 <GregorR> ehird: Well that's just obnoxious.
18:11:51 <ehird> If I run the driver installer, I get an error saying the archive is malformed.
18:12:07 <ehird> Also, I got a minor error thing from the .msi but it seemed t owork.
18:12:28 <Gracenotes> ais523: oh. Yeah. By backwards-compatible I meant "catching Pokemon from previous generations"
18:12:39 <GregorR> If that's actually a device driver, I would not anticipate it working.
18:12:40 <ehird> The Runtime starts.
18:12:45 <ehird> And gives a lot of worrying errors, but it starts.
18:12:53 <AnMaster> ais523, you may be interested in that too
18:12:56 <ehird> But the runtime is useles for making patches
18:13:02 <ais523> AnMaster: well, what was the reply?
18:13:32 <GregorR> Dear AnMaster@...: SCREW YOU HA HA HA WE SPAM LAWL
18:14:05 <GregorR> Getting any sort of response at all from an abuse@ address is pretty darn successful :P
18:14:37 <AnMaster> http://rafb.net/p/522Q4u44.html
18:15:21 <GregorR> What was the site you got blorped?
18:16:28 <AnMaster> GregorR, irc bot spamming download to trojan, reported the domain used for forwarding to a generic file hosting site used for hosting the trojan.
18:17:03 <ehird> Ok. So the runtime works.
18:17:07 <ehird> Just not the editor thing.
18:17:19 <AnMaster> reported to abuse@ found in whois for ip of spambot, ip of server hosting forwarder and so on.
18:18:03 <AnMaster> the file hosting site found out quite fast and removed it even before I could send a mail notifying them of it (or someone else already did that)
18:18:43 <AnMaster> anyway I got sick of lots of spambots spamming similar messages, sure spamfilters now catch them pretty well but even so...
18:19:09 <Deewiant> fact a Free Domain registrant, therefore
18:19:09 <Deewiant> because of the violation, the domain has
18:19:17 <Deewiant> So if he had paid for it it would be alright :-P
18:19:35 <ehird> ah. it is "PACE" protected.
18:19:41 <ehird> which is some kind of piracy protection I guess
18:19:47 <ehird> that's why the runtime works, it's not protected
18:20:28 <ais523> antipiracy tends not to work very well on genuine Windows either
18:20:55 <ais523> I really admire Bioware for patching out the antipiracy checks in Neverwinter Nights 1 in a relatively early update
18:21:11 <ehird> ais523: Don't copy that floppy.
18:21:28 <AnMaster> ehird, since I'm trying to solve an issue with swap trashing I won't google it, links/lynx/w3m are just a pain to use.
18:21:35 <ais523> don't worry, I've bought all the copies of that I need
18:21:37 <AnMaster> so maybe I'll look it up later
18:21:40 <ehird> w3m is a pain to use??
18:21:44 <ais523> as in, I bought a Windows CD to get a licence for the Linux version
18:22:31 <ehird> it occurs to me that linux satisfies the needs of Regular People(TM) and the ubergeeks, but not the people inbetwee
18:23:05 <ais523> especially if the inbetween people grew up on Windows
18:23:17 <ais523> changing Linux <-> OS X isn't as hard as from Windows to either, I don't think
18:23:29 <ais523> because although the UIs are different, they both have POSIXish internals
18:24:05 <pikhq> And if you *really* want it to, OS X can look almost identical to Linux.
18:24:06 <ehird> ais523: well, max/msp is portable
18:24:12 <ehird> it's available for both windows and linux
18:24:15 <ehird> and is identical on both
18:24:16 <pikhq> ... (granted, the same is true of Windows)
18:24:25 <ehird> they just haven't bothered to do linux...
18:24:26 <ais523> Linux can look almost identical to Windows, too
18:24:33 <ehird> pikhq: you can run X11 on os x
18:24:37 <ehird> and get it looking identical to unix
18:24:42 <ehird> although that's rather missing the point
18:24:48 <ais523> presumably you can run gdm on OSX, you just wouldn't
18:24:54 <pikhq> That's what I was referring to.
18:25:17 <ehird> there is an advantage to doing that over just using linux
18:25:25 <pikhq> ais523: It's possible to get gdm or kdm running instead of the normal OS X stuff rather easily.
18:25:32 <ehird> in that quartz is more stable than X11's gfx drivers
18:25:36 <ehird> (and apple's X11 is quartz-backed)
18:25:42 <pikhq> Just a slight modification to OS X's inittab equivalent.
18:26:07 <pikhq> ehird, you also have the option to use X.org...
18:26:24 <ehird> pikhq: yes, but then there's 0 advantage over using linux :P
18:26:27 <pikhq> Though, honestly, at that point, you should just install GNU/Darwin; it's the same damned thing.
18:26:27 <ehird> negative advantage, even
18:26:30 <ehird> since it's not tuned for doing that
18:26:52 * ais523 wonders what a Windows port of getty would be like
18:27:04 <pikhq> Ask Cygwin, I bet they have one.
18:27:34 <ehird> cygwin is reversed wine
18:28:01 <ais523> is it possible to get cygwin to run at the binary level?
18:28:04 <ais523> or only with a recompile
18:28:13 <ehird> only with a recompile
18:28:26 <ehird> you could write an ELF "emulator" for windows
18:28:29 <ehird> and hook it into cygwin
18:28:55 <ais523> like userspace qemu, for instance?
18:29:11 * ais523 was using qemu-arm to test C-INTERCAL cross-compilation
18:29:13 <ehird> not emulate the actual instructions
18:29:21 <ehird> it'd just load an ELF and run it
18:29:26 <ehird> with some LD_PRELOAD-esque magic for cygwin
18:29:52 <pikhq> I don't think Windows would handle it too readily.
18:30:00 <pikhq> Windows doesn't much like alternate executable formats.
18:30:13 <ehird> pikhq: I meant, you'd do all the parsing and whatnot yourself
18:30:23 <ehird> then run the actual code section as whatever window's using this week
18:32:55 <GregorR> I wrote an ELF loader for Windows.
18:33:17 <GregorR> It doesn't hook up to Cygwin or anything. The other major problem you'd run in to are binaries that call syscalls directly.
18:33:46 <ehird> Just grep for syscalls
18:33:49 <ehird> and replace with stub function calls
18:33:54 <ehird> that emulate the linux ones
18:34:12 <Deewiant> Might break self-modifying code since the instructions aren't the same size!
18:34:21 <ehird> Deewiant: Pad them out
18:34:36 <Deewiant> ehird: As in, they might be longer
18:34:36 <ehird> If they're too big, fix that.
18:34:43 <ehird> Deewiant: Right. So don't do that :-P
18:34:54 <GregorR> ehird: Have fun. Feel free to use WinELF :P
18:35:01 <Deewiant> Well, what're ye gonna do when you have such a binary :-P
18:35:16 <ehird> Deewiant: there are a limited number of syscalls that exist
18:35:24 <ehird> so, there's a maximum size a syscall instruction can be
18:35:30 <ehird> so, make sure your replacements are always shorter
18:35:41 <Deewiant> 'syscall' is always the same size
18:35:47 <ehird> There you go then.
18:36:57 <GregorR> ehird: http://svn.berlios.de/svnroot/repos/crosslibc/other/
18:37:27 <ehird> elfload.exe - no source?
18:38:01 <pikhq> That'd work rather well on that somewhat screwy hybrid of WINE and Linux that supports both Windows and Linux system calls...
18:38:04 <ehird> Deewiant: Oh brother.
18:38:14 <GregorR> Oh, sorry, the source to elfload.exe is rtload, it's in the trunk of that project
18:38:29 <ehird> someone should make a system that's a combination of both WINE and cygwin
18:38:43 <ehird> it supports both windows AND linux, both inferior to the real thing!
18:38:50 <GregorR> http://svn.berlios.de/svnroot/repos/crosslibc/trunk/rtload/
18:38:58 <AnMaster> pikhq, something based on ReactOS maybe?
18:39:04 <pikhq> That'd be ReactOS/
18:39:43 <ehird> CARCASS OF MAGNIMOUSLY
18:40:26 <AnMaster> reactos doesn't work very well, that was kind of my point.
18:40:55 <pikhq> ehird: Said hybrid is called E/OS, IIRC. The idea is to support programs for every OS.
18:40:55 <GregorR> Last I tried it, it was both better and worse than I expected.
18:41:10 <Deewiant> Stuff tends to work less well under virtualization than otherwise
18:41:23 <AnMaster> GregorR, how do you mean both better and worse
18:41:40 <pikhq> Sorry, make that Linux Unified Kernel, which only aims to support Windows & Linux.
18:41:43 <GregorR> AnMaster: Some features were much better than I anticipated, others were much worse.
18:41:50 <pikhq> Including drivers.
18:42:15 <AnMaster> I mean what about the security model of windows...
18:42:35 <Deewiant> You can improve your kernel all you want but you can't deal with crap software
18:42:48 <pikhq> Well, it *is* a Win32 implementation...
18:42:50 <AnMaster> too many things would depend on admin rights on windows
18:42:58 <pikhq> Deewiant: The only way to deal with it, I guess, would be SELinux.
18:43:13 <Deewiant> Well yeah, you can "deal" with crap software by not letting it run :-P
18:43:14 <AnMaster> so thus it would be just as insecure as real windows, or not be able to run most apps
18:43:25 <ehird> AnMaster: false dichotomy, wayoh!
18:44:00 <ehird> Incorrect premise: "All Windows security problems are due to requiring Administrator access and the only way to fix this is to not use the admin account."
18:44:19 <AnMaster> pikhq, selinux: Did you mean: having to spend half an hour to write a policy for a program that takes 10 seconds to run.
18:44:56 <AnMaster> well I was being optimistic I guess
18:45:07 <AnMaster> my experience is that it takes way longer
18:46:20 <ehird> SELinux works fine ime.
18:46:39 -!- Judofyr has joined.
18:46:56 <AnMaster> but the fact that so many windows programs won't work without admin rights, and users are lazy, tends to be one major problem
18:49:46 -!- Gracenotes has quit (Connection timed out).
18:49:53 -!- GregorR-L has joined.
18:50:14 * GregorR-L spams http://codu.org/wiki/index.php?title=Earth at the channel again.
18:50:47 -!- GregorR-L has quit (Remote closed the connection).
18:51:01 -!- GregorR-L has joined.
18:51:06 <GregorR-L> But first I'm going to make xchat crash :P
18:51:48 <ehird> GregorR-L: I have an idea.
18:52:04 <ehird> The conlang for it should be hopelessly unsuited to the task of this documentary.
18:52:10 <ehird> For instance, "Earth" should be 3 lines of text.
18:52:27 <ehird> üxu¨¥©¨¥˙©gauyhdƒ©˝ʼøðőa üxu¨¥©¨¥˙©gauyhdƒ©˝üxu¨¥©¨¥˙©gauyhdƒ©˝ʼøðőa
18:52:28 <ehird> üxu¨¥©¨¥˙©gauyhdƒ©˝ʼøðőaüxu¨¥©¨¥˙©gauyhdƒ©˝ʼøðőaüxu¨¥©¨¥˙©gauyhdƒ©˝ʼøðőa
18:52:31 <ehird> üxu¨¥©¨¥˙©gauyhdƒ©˝ʼøðőaüxu¨¥©¨¥˙©gauyhdƒ©˝ʼøðőaüxu¨¥©¨¥˙©gauyhdƒ©˝ʼøðőa
18:52:37 <GregorR-L> Put ideas on wiki, not #esoteric :P
18:52:46 <ehird> I'll put my idea wherever I want tyvm :|
18:53:05 <ehird> GregorR-L: it should have ẍ as a letter.
18:53:14 <Deewiant> "As we know, the well-accepted signs of intelligence are communication, [no translation], [no translation], civilization, [no translation], and tool use."
18:53:25 <ehird> Being the blend of "x" as in "ch" in "Bach", the English "x" and the sound of "six"
18:53:51 <AnMaster> bbiab, rebooting due to kernel recompile (change could not be done as module this time)
18:54:09 -!- Gracenotes has joined.
19:02:05 <ais523> Yet Another Enigma Level: http://pastebin.ca/raw/1394124
19:02:19 <ais523> AnMaster might actually like that one, but I don't think that ehird will necessarily hate it
19:02:55 <ehird> do I _have_ to save it manually from there?
19:02:59 <ehird> you didn't even give a filename!
19:03:16 <ais523> and filebin.ca has stopped working for me, I'm not sure why
19:03:29 <ehird> ais523: does it say it can't be saved?
19:03:32 <ehird> and the link appears
19:03:56 <ehird> i love enigma's menu music
19:04:11 <ais523> the menu music has copyright issues, IIRC they're replacing it
19:04:27 <ehird> What're the issues?
19:04:34 <ehird> I'm talking about the one in v1.01 fwiw
19:05:08 <ehird> It's a fuckin' mod file
19:06:38 <ais523> it should be pretty simple to figure the rules
19:06:45 <ais523> especially as it's a WYSIWYG level
19:06:51 <ehird> it's just frustrating :D
19:07:23 <ehird> I got to the other side a second time
19:07:24 -!- AnMaster has quit (Connection reset by peer).
19:07:30 <ehird> and bashed one of the no-go-through blocks
19:07:37 <ehird> and onto an exposed death
19:08:37 <ehird> I bounced to the other side
19:08:41 <ehird> onto one of the no-go blocks
19:08:44 <ehird> and bounced BACK WHERE I STARTED
19:08:52 <ais523> I do that on occasion too
19:09:57 * ehird takes out a full row of death protectors :x
19:10:20 -!- Gracenotes has quit (Success).
19:11:17 -!- Gracenotes has joined.
19:12:00 <ais523> http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=321669 explains the copyright issue
19:12:58 <ehird> We have a special permission from Andrew Sega (aka Necros) to distribute
19:12:58 <ehird> menu.s3m with Engima. He is also listed in the manual
19:13:16 -!- BeholdMyGlory_ has joined.
19:13:34 <ehird> "I think it will be at least problematic." Fuck
19:13:42 <ais523> so you can get the music from the official site, with Enigma
19:13:47 <ais523> but none of the mirrors have it for legal reasons
19:13:56 <ehird> what? Is that the situation now?
19:14:01 <ehird> When I used Enigma on Ubuntu it had the music
19:14:04 <ehird> well, that was years ago
19:14:21 <ais523> yes, nowadays you can download the music and put it in the right place by hand and it works
19:14:24 <ais523> but it isn't in the repos
19:14:27 <ehird> "Erich, applying the GPL to a documentation is ok, but don't you think you are
19:14:27 <ehird> pushing things a bit hard by applying it to a music file too ? "
19:14:30 <ehird> ais523: that's bullshit
19:15:00 <ehird> ais523: got two out of four on your leve;
19:15:30 <ais523> yep, it's mostly an AnMastery level, I think
19:15:41 <ehird> ais523: not much exploration or finding out.
19:15:49 <ais523> yes, but lots of patience
19:15:49 <ehird> it's too fun for that
19:16:54 <ehird> ais523: this is from 2005
19:16:58 <ehird> has it not been resolved yet?!
19:17:00 <ais523> 11 is finished, by the way, but I'm not sure if I'm evil enough to create it
19:17:11 <ehird> I had Enigma music after 2005 on linux
19:17:15 <ehird> ais523: create it?
19:18:01 <ais523> the idea was that I wanted to make levels that pushed the limits of difficulty in the 5 categories, int/dex/spd/kno/pat
19:18:06 <ais523> it's designed as a maximum-dex level
19:18:13 <ehird> what're the ratings
19:18:14 <ais523> I've only done it on easy, as a result
19:18:38 <ais523> int 2 / dex 6 (5's the maximum) / spd 1 / kno 4 / pat 3
19:18:50 <ehird> that's not very maxed out
19:18:51 <ais523> (pat is linked to difficulty, if a level's so hard you have to keep restarting it the pat goes up)
19:19:01 <ais523> no, I mean, 5 different levels
19:19:05 <ais523> each of which maxed a different stat
19:19:11 <ehird> you should make a 6 of all
19:19:20 <ais523> I'd certainly like to try a max-kno level
19:19:51 -!- BeholdMyGlory_ has quit (Remote closed the connection).
19:20:06 <ehird> a level requiring great puzzle-solving skills and very fine movement while under constant danger if you don't go fast, using incredibly obscure interactions, objects and *bugs* -- and where you can die at any corner.
19:20:08 <ehird> ais523: MAKE IT :D
19:20:25 -!- BeholdMyGlory has joined.
19:20:43 <ais523> I've already found one bug that I had to work around in Tourism
19:20:47 <ais523> but it's fixed in 1.10, I think
19:21:07 <ehird> I say bugs, I mean obscure corners that might not be totally intentional
19:21:24 <ehird> int 6 comes from the very complex puzzle intwined in every part of the level
19:21:44 <ehird> dex 6 comes from every move you make affecting the puzzle: one wrong move and you've lost
19:21:48 <ehird> spd 6 is from the constant danger from things chasing you and whatnot
19:21:53 <ehird> kno 6 is from the incredibly obscure interactions
19:22:02 <ehird> and pat 6 is because there's so many oppertunities to lose
19:22:33 * ais523 just tried to use shift-f3 to reload a Web page
19:22:58 <ais523> restart level key in Enigma
19:23:02 <ais523> you can tell what I've been doing recently
19:24:21 <ais523> anyway, if you're curious: http://pastebin.ca/raw/1394185 is ais52311_1.xml
19:24:27 <ais523> I've only solved it on easy
19:24:35 <ais523> but removing all my mistakes, I could have solved it on hard
19:24:49 <ais523> (easy and hard are the same except easy gives you infinite extra lives, even then the level is still quite hard)
19:26:13 <ehird> what's default mouse speed again
19:26:30 <ais523> according to the forums most people have it from about 5 to 7
19:26:34 <ais523> and set it to 15 on a few levels
19:26:56 <ehird> ais523: how do you get past the chasm of water
19:27:03 <ais523> those blocks are movable
19:27:40 -!- AnMaster has joined.
19:27:47 <ehird> ais523: this doesn't look that hard
19:27:54 <ehird> ... hey AnMaster ;;;;;;)
19:28:05 <ais523> the bit in the middle is extremely hard
19:28:17 <ais523> push a chess stone when a laser denies you access to all but a tiny bit
19:28:51 <ehird> my issue is the bounce-back from the blocks you move
19:29:00 -!- Gracenotes has quit (Connection timed out).
19:29:12 <ais523> AnMaster: ais52311_1.xml = http://pastebin.ca/raw/1394185 ; ais52312_1.xml = http://pastebin.ca/raw/1394124
19:29:30 <ehird> AnMaster: because tedious boringness is exactly your level style.
19:29:31 <ais523> also, as it's an evil dex level, I deliberately tried to maximise the problems caused by block bounceback
19:29:36 <ais523> as that's one of the thing that sends dex sky-high
19:30:08 <ehird> ais523: not quite infinite lives, is it
19:30:14 <ais523> ehird: it is on easy, it isn't on hard
19:30:19 -!- Gracenotes has joined.
19:30:24 <ehird> ais523: yes but it's just a large amount of finite lives
19:30:24 <ais523> 11 was me trying to create the highest-dex level I could
19:30:38 <ais523> it generates extra lives on the square you get respawned on
19:30:53 <ais523> you'll notice that your inventory starts full of it-extralife, and never gets smaller no matter how often you die
19:31:30 <ehird> ais523: I literally cannot bounce them without bouncing back into something
19:31:31 <ehird> is it even possible
19:31:48 <ais523> I've done the entire level on easy with only two deaths which alter the gamestate
19:32:02 <ais523> therefore, taking all my best attempts together and cutting out the mistakes I could have done it on hard
19:32:48 <ehird> a chess piece surrounded diagonally by water.
19:33:04 <ais523> the block to its right fills in the space to its right
19:33:13 <ehird> i didn't get that one
19:34:32 <ehird> ais523: wait, what?
19:34:39 <ais523> what's the wait, what? for?
19:34:53 <ehird> screenshot of a position where you can hit the first chess piece?
19:35:10 <ehird> http://pics.livejournal.com/koalafrog/pic/002g4c16/s640x480
19:37:20 <ais523> strangely, imagebin.ca isn't working either
19:38:18 <ais523> grr, it's setting foreground colour but not background colour for the text box you put the filename in
19:38:37 <ais523> which makes it white on light grey
19:38:52 <ehird> ais523: aw, cute little ais523, using a browser without a file selection widget
19:39:01 <ais523> it has a file selection widget
19:39:07 <ais523> which is the only reason it isn't tolerable
19:39:10 <ais523> http://imgur.com/2A6RO anyway
19:39:18 <ehird> ais523: wrong link
19:39:20 <ehird> you need to append .png
19:39:24 <ehird> http://imgur.com/2A6RO.png
19:39:31 <ehird> how do you diagonally hit that
19:39:40 <ais523> chess stones aren't hit diagonally
19:39:43 <ais523> they're hit orthogonally
19:39:46 <ais523> just whilst moving diagonally
19:39:51 <ais523> as in, to move a chess stone south
19:40:34 <ais523> you hit it on its north face whilst moving south-west
19:41:40 <ais523> I deliberately set up situations where the chess stones were hard to move
19:42:19 <ehird> this is like pat 9001
19:43:31 <ehird> ais523: i'm going down-left and bashing into it
19:43:42 <ais523> well, I just pushed it via that method
19:43:48 -!- pikhq has quit ("leaving").
19:43:54 <ais523> oh, you have to hit it with a set minimum force too
19:44:20 <ehird> i have mouse speed set to 1
19:44:22 <ehird> so this may be difficult
19:45:00 <ais523> can you move chess stones on other levels?
19:45:09 <ehird> although with less carefulness
19:45:13 <ehird> and more trial and error
19:45:41 <ais523> I just moved the stone at mouseforce 1, anyway
19:45:47 <ais523> although I drowned immediately afterwards
19:45:49 -!- oerjan has joined.
19:46:00 <ais523> because I couldn't slow down fast enough from the rebound because the mouseforce was too low
19:46:06 <ehird> i wish this had some respawney point things
19:46:45 <ais523> but it's only one screen, and linear
19:47:12 <ehird> ais523: yes, but I keep dying
19:47:15 <ehird> and retracing my steps over and over
19:47:24 <oerjan> <ehird> and I think "we will be able to live longer within 80 years" as a prediction is almost certain
19:47:54 <ais523> really? I'd assume that our life expectancy would be 80 shorter
19:47:57 <ehird> 19:46 oerjan: <ehird> and I think "we will be able to live longer within 80 years" as a prediction is almost certain
19:48:08 <oerjan> it is conceivable that life extension requires preventing some damage from birth, so that it is too late for us
19:48:14 <ehird> ais523: i meant, "within 80 years we will be able to extend our life spans"
19:48:25 <ehird> oerjan: conceivable, but perhaps not likely.
19:48:28 -!- Gracenotes has quit (Connection timed out).
19:48:41 <ehird> i would be incredibly surprised if it were completely impossible to extend existing lifespans.
19:48:44 <oerjan> i.e. we already know _some_ ways of extending life spans, which most of us are too lazy to do
19:49:09 <oerjan> (well maybe not final lifespan, but average)
19:50:51 -!- Gracenotes has joined.
19:50:55 <lament> i expand my lifespan every time i fail to kill myself
19:51:52 <ehird> lament: conversation killer extraordinaire
19:51:56 <oerjan> obviously, immortality requires us to be able to extend lifespan by more than 1 second per second, eventually
19:52:32 <ehird> oerjan: i invented a time machine
19:52:38 <oerjan> yes most conversations stop when participants kill themselves
19:52:42 <ehird> it can only go forward, and it goes at 1 second per second
19:52:43 <ais523> AnMaster: had a look at my levels yet?
19:53:26 <AnMaster> ais523, notice lack of plural above.
19:53:39 <ais523> but surely you formed some opinion?
19:53:43 <AnMaster> actually one of them doesn't seem too bad.
19:53:47 <ais523> or you just glanced at them and mentally put them off too later?
19:53:59 <ais523> also, which one doesn't seem too bad?
19:54:02 <AnMaster> ais523, well mostly that because I'm busy
19:54:02 -!- MigoMipo has joined.
19:54:06 <oerjan> ehird: i think there is some prior art there
19:54:09 <ais523> and fine, I guessed you were
19:54:43 -!- GregorR-L has quit (Read error: 110 (Connection timed out)).
19:54:48 <AnMaster> Yay! Got electric piano to work with USB in rosegarden in linux :D
19:54:57 <AnMaster> actually I'm not surprised these days
19:55:23 <AnMaster> just need to fiddle a bit with midi mixer in rosegarden
19:55:33 <ais523> AnMaster: you use rosegarden too?
19:55:51 <ehird> yay! i tried to do something with linux and IT WORKED with only a little bit of mindless configuration twiddling!
19:55:52 <AnMaster> ais523, yes, I don't know any other good app like it, and I have no reason to change
19:55:59 -!- Gracenotes has quit (Operation timed out).
19:56:02 <ehird> by wasting my time i am expressing my superiority to other operating systems!
19:56:08 <ehird> ais523: you'll have to ask AnMaster
19:56:12 <AnMaster> ehird, I get same config twiddling under cubasis under windows too
19:56:24 <ais523> well, I got rosegarden working without any config twiddling
19:56:27 <AnMaster> ehird, thing is midi will always need a bit of setup in the mixer
19:56:32 <ehird> AnMaster: not on os x
19:56:33 <ais523> just by installing timidity
19:56:38 <AnMaster> ehird, had to select the right output and input unit
19:56:46 <AnMaster> so it can't auto detect in any way
19:56:52 <ais523> AnMaster: I've needed to fiddle with that on Windows, but not on Linux
19:56:54 <ehird> err selecting an input isn't what you implied
19:57:02 <ehird> also, it can easily DTRT
19:57:10 <ehird> if you have a keyboard you probably want to input via that, for instance
19:57:19 <AnMaster> ehird, well it can't know from usb that I do
19:57:27 <ais523> ehird: in theory the midi can't tell whether it's a keyboard or something else
19:57:32 <AnMaster> it can only know it is a unit implementing midi
19:57:37 <ais523> although IMO, the obvious thing to do is just to accept via /every/ midi input
19:57:47 <AnMaster> ehird, also I wanted output to keyboard too
19:58:02 <ehird> that's what garageband does
19:58:02 <AnMaster> instead of hardware midi on the sound card
19:58:30 <ais523> well, I use software MIDI via my sound card here
19:58:41 <AnMaster> automatically handling that would be something like "read mind of user"
19:58:54 <AnMaster> because sometimes you may want output to one unit, and sometimes to another
19:59:07 <AnMaster> getting output to all units is probably NOT what you want in fact
19:59:18 <ehird> in that case it can prompt
19:59:22 <ehird> but for inputs just accept everything
19:59:49 <AnMaster> ehird, there are reasons to avoid that too sometimes. When you only want to capture some of the output from an unit
20:00:04 <ehird> http://conservativedialysis.com/~mnick/wp/wp-content/themes/dkubrick/images/Oval_CD_V2_01.png
20:00:08 <ehird> CONSERVATIVE DIALYSIS
20:00:10 <ehird> REMOVING LIBERAL WASTE FORM
20:00:13 <ehird> THE AMERICAN BLOODSTREAM
20:00:15 <AnMaster> which was the fact here, since this keyboard seems to send odd system messages on channel 16
20:00:23 <ehird> worst blog name/tagline ever y/n
20:00:42 <ehird> AnMaster: FROM. :P
20:00:45 <ehird> the first two sidebar entries are Never Forget followed by a picture of 911
20:00:57 <ehird> and then The Religion of Peace and some bullshit statistics about ISLAMIC TERRISTS killing people
20:01:29 <ehird> ooh, it even has "Anti-Idiotarian" on the page
20:01:41 <AnMaster> now to figure out midi-thru on the unit
20:01:43 <ehird> "It is not my intention to be fair or balanced."
20:01:57 * AnMaster waits for ehird to claim this is a linux issue too.
20:02:12 <ehird> AnMaster: when did you last edit your Xorg.conf?
20:02:14 <AnMaster> (which it isn't, it is just an issue with MIDI)
20:02:25 <AnMaster> ehird, hm, a year ago or so probably
20:02:26 <ais523> ehird: I last edited an Xorg.conf two days afo
20:02:37 <ais523> but I was undoing borkage that I introfuced editing it months ago
20:02:41 <ehird> go and edit it so I can troll you
20:02:53 <AnMaster> ehird, anyway you just changed topic there
20:03:10 <ehird> oh boy, this conservative dialysis guy supported .xxx
20:03:50 <AnMaster> ehird, why do you 1) bother reading it 2) bother writing about it here
20:04:12 <ehird> AnMaster: 1) idiots are funny 2) i've only said about 7 lines.
20:04:29 <AnMaster> oh? you tend to get irritated when I reach the second line
20:04:48 <ehird> i wish that made any sense so I could tell you how stupid it is
20:06:00 <AnMaster> it did make sense. You just failed at understanding it
20:06:17 <ehird> right. if anyone fails to understand someone it's their fault
20:06:26 <ehird> it's not like communication is attempting to inject meaning into other's minds or anything
20:06:37 <Deewiant> http://pics.livejournal.com/koalafrog/pic/002g4c16/s640x480
20:06:55 <ehird> Deewiant: i just linked that.
20:07:13 <Deewiant> Sorry if I missed it before that
20:07:19 <ehird> BE TELEPATHIC DAMMIT
20:07:38 <ais523> why did you both just link it?
20:07:41 <Deewiant> (Those two pageups go up 30 mins)
20:07:48 <ehird> ais523: because we think it was amusing, presumably.
20:07:57 <Deewiant> (So your "few minutes" is a bit of a stretch)
20:08:02 <ais523> but where did you find it?
20:08:07 <ehird> ais523: reddit, probably.
20:08:34 <ehird> /nick PersonalRedditQualityFilteringSystemShareAndEnjoy
20:10:05 <ehird> Historically, locate only stored characters between 32 and 127. The cur-
20:10:06 <ehird> rent implementation store any character except newline (`\n') and NUL
20:10:07 <ehird> (`\0'). The 8-bit character support does not waste extra space for plain
20:10:09 <ehird> ASCII file names. Characters less than 32 or greater than 127 are stored
20:10:19 <ehird> it's like... ASCII-7
20:11:20 <AnMaster> ehird, which locate... it isn't mentioned in the one I have
20:11:54 -!- Gracenotes has joined.
20:11:54 <Deewiant> ehird: I wonder how it stores characters outside the BMP in two bytes
20:12:09 <ehird> Deewiant: it just does 0-256
20:12:18 <ehird> just like utf-8 does 0-ridiculous in 0-256
20:12:21 <Deewiant> I wonder why it uses two bytes for that :-P
20:12:33 <ehird> Deewiant: beats me! compatibility?
20:13:16 <ehird> " The locate utility does not recognize multibyte characters. "
20:13:27 <fizzie> Heh, my two pageups go approximately 8 minutes now; I updated some drivers (more-or-less-beta-maybe-I-guess-or-should-it-be-stable proprietary nvidia 180.44, xorg 7.4, 2.6.29.1 kernel) and now (a) cursor droppings are left on the other screen when I move the cursor from one screen to the other, and (b) X crashes after approximately 5 minutes. That's very suboptimal.
20:13:48 <Deewiant> I don't even need to press pageup to go back 13 minutes :-P
20:14:15 <fizzie> This is the 80x25 text console, since I had some problems with the nvidiafb framebuffer console, too.
20:14:32 <Deewiant> I haven't bothered even trying to run a framebuffer on this AMD card
20:14:52 <Deewiant> Don't really feel a need for it anyway
20:15:15 <fizzie> The hilarious part is that by default the card is set to mirror the 80x25 screen on both connected monitors, and it doesn't know about the rotation, so I get a horribly big and stretched 80x25 console, and a 90 degrees tilted copy of it to the left.
20:18:06 -!- Gracenotes has quit (SendQ exceeded).
20:20:50 -!- Gracenotes has joined.
20:22:45 <AnMaster> fizzie, never had such issues with nvidia drivers. Other issues sometimes but not those.
20:23:43 <fizzie> Actually I don't think it (well, except the cursor thing maybe) is related to the nvidia drivers.
20:23:56 <fizzie> Some 180.2x version used to work just fine.
20:24:00 <AnMaster> <sfdsfdfds> hy,scuze me for disturbing you, colud you help me an check the lines if they are corect writen ? , in my "ppp.conf" file in here http://paste2.org/p/185018 , Best Regards
20:24:00 <AnMaster> * [sfdsfdfds] (n=l0renzo@79.116.94.107): anonym
20:24:00 <AnMaster> * [sfdsfdfds] ##bsd #bsdhelp #freenode #openbsd
20:24:25 <AnMaster> I should try to figure out what setting to use to turn off /msg from non-registered users
20:25:09 <oerjan> well something like that
20:25:18 <fizzie> The X backtrace is: XkbProcessKeyboardEvent - XkbHandleActions - ProcessOtherEvent - CheckMotion - [something] - [something] - [something] - [something in libc.so] - xf86SigHandler.
20:25:22 <oerjan> strange thing is, /mode oerjan doesn't show it for me
20:25:40 <oerjan> anyway it was in the freenode faq somewhere
20:25:44 <fizzie> And the most reliable way of getting it to crash seems to be to write "ls" in a terminal, and then hold enter down long enough that it starts to scroll.
20:26:19 <AnMaster> fizzie, downgrade to old version
20:26:25 <AnMaster> should be easy with most sane package manager
20:26:45 <AnMaster> on gentoo it would be emerge =x11-drivers/nvidia-drivers-180.29 or such to force a specific version of those available
20:26:51 <AnMaster> and considering how many are available heh
20:27:01 <oerjan> huh the faq link doesn't work
20:27:19 <AnMaster> and then report it to freenode staff
20:27:26 <fizzie> Really, I'm not that sure it's nvidia-driver-related.
20:27:38 <AnMaster> fizzie, try downgrading and see it it solves the issue?
20:27:42 <ehird> downgrading on debian is unsupported.
20:27:53 <oerjan> the link _from_ the main faq list
20:28:04 <ehird> otherwise you get a combinatorial explosion of dependency combinations you have to test
20:28:06 <AnMaster> but for some stuff it is useful
20:28:06 -!- Gracenotes has quit (SendQ exceeded).
20:28:12 <ehird> which is roughly on the scale of impossible.
20:28:19 <AnMaster> ehird, not an issue on source based distros :)
20:28:57 <AnMaster> ehird, much less, source level compatibility breaks less often than binary compatibility
20:29:06 <ehird> .....................
20:29:16 <ehird> point, your head: no collision.
20:29:48 <oerjan> AnMaster: http://freenode.net/faq.shtml#spambots works
20:30:30 <ehird> AnMaster: sure it did.
20:30:36 <ehird> the point did not collide with your head; it went over it.
20:30:41 <ehird> Insert whoosh sound.
20:30:46 <oerjan> although i _think_ blocking is the default...
20:30:55 <AnMaster> ehird, That may be because there was no point
20:31:04 <AnMaster> oerjan, it used to do that. It doesn't any longer.
20:32:04 <oerjan> AnMaster: some of your problem could be that sfdsfdfds actually _is_ a registered user...
20:32:06 <AnMaster> ehird, Not under normal conditions for dihydrogenmonoxide based water
20:32:29 <AnMaster> oerjan, well most of the time they aren't
20:33:27 <oerjan> actually maybe i _don't_ have that blocking. i have +i and am in only one channel so probably the spammers just don't find me
20:34:03 -!- jix has joined.
20:34:08 <AnMaster> oerjan, I seem to have the modes +eEiuw atm.
20:34:21 <AnMaster> and you would have least have +ei
20:34:22 <ehird> oerjan: huh, aren't you usually in #haskell?
20:34:32 <oerjan> yeah that's what i have
20:34:48 <oerjan> ehird: haven't been regularly in #haskell for a year
20:34:59 <ehird> oh right you stopped programming :P
20:35:20 <ehird> AnMaster: what do you mean why
20:35:31 <AnMaster> why would anyone stop programming by their free will?!
20:35:44 <ehird> cuz he's concentrating on his maths work and other stuff?
20:35:52 <ehird> or is bored with programming?
20:39:34 <oerjan> /help mode only lists channel modes
20:41:27 <fizzie> Heh, how typical; if I attach a gdb to that X process (have nothing better to do than to poke at it) and tell it to continue, it seems to change timing or whatever enough so that it no longer crashes.
20:42:12 <fizzie> An elegant workaround: a cronjob that makes sure any X servers always have a gdb attached to them.
20:42:19 <AnMaster> oerjan, +u is: can join more channels than the default limit of 20
20:43:13 <oerjan> i probably don't need those then :D
20:43:22 <AnMaster> you get it by hanging around for a long time, muttering in various channels with staff in about that you need to part channel because you hit the 20 limit and so on and finally asking some staffer nicely about it
20:43:38 <AnMaster> thankfully I'm not close to that on freenode yet
20:44:09 <ehird> 20:41 fizzie: An elegant workaround: a cronjob that makes sure any X servers always have a gdb attached to them.
20:44:26 <fizzie> Unfortunately it just made it harder to crash. :p
20:44:28 <ehird> AnMaster: I think you could just ask instead of being passive-agressive
20:44:38 <fizzie> But at least I have a prettier backtrace now.
20:44:41 <ehird> fizzie: why'm I switching to linux again?
20:45:54 <fizzie> ehird: Some sort of masochism, maybe?
20:47:16 <fizzie> This particular crash seems to be a null pointer dereference ("mov %rax,0x20(%rbx)" with %rbx zero) in ../../mi/mipointer.c:209, but unfortunately I don't happen to have X sources hanging around. Maybe I should.
20:47:54 <fizzie> As a disclaimer: this is, after all, Debian's "unstable" distribution; I certainly appreciate getting these nice puzzles right on my face when I least expect them.
20:48:38 <ehird> The best option for breaking your computer completely.
20:49:39 <fizzie> Well, so far it hasn't actually done anything to the hardware.
20:49:49 <ehird> So, guys, I found a time machine. Watch:
20:49:57 <ehird> who has the lowest ICQ number?
20:50:14 <ehird> It only goes back to the early 2000s, I'm afraid.
20:50:44 -!- Gracenotes has joined.
20:50:49 * oerjan may or may not have had an ICQ number at one time, he's not quite sure
20:51:06 <ehird> Deewiant: well, the low-ICQ craze wasn't until later
20:51:11 <ehird> let's say late nineties to early 2000s
20:51:25 <ehird> Yes, among the geocities type of folk.
20:51:38 <ehird> 5 digits one sell for sth like $60 a pop
20:52:08 <fizzie> My number seems to be 9372355; that is a rather uninteresting number. Very close to being 8 digits, even.
20:52:10 <Deewiant> Mine is 9 digits, my brother has a 7-digit one
20:52:10 <oerjan> are these people from dubai? :D
20:52:26 <Deewiant> I only got mine in 2006 or so, haven't used it much
20:52:37 <ehird> my icq number is 2i
20:53:16 <ehird> for your imaginary internet friends
20:53:31 <oerjan> what's the q for anyway
20:53:33 <fizzie> I don't know when I got it; the backed-up ~/.licq config files I used to find out that number seem to have timestamps around 2001.
20:53:52 <ehird> oerjan: I See Qdeadpeople
20:54:02 <fizzie> I wonder if CU-seeme's still alive.
20:54:43 <fizzie> "The is still a small but active community" says wikipedia.
20:54:59 <fizzie> No official releases since 2000, so..
20:55:05 -!- Slereah_ has joined.
20:55:19 <ehird> "My husband and I both have 6 digit UINS, both under 300000. I didn't think the low number was a big deal until my husband had his "stolen". He was silly enough to enter his password in an email "from icq". The idiot who took it was even less clever than my husband because he forgot his password and had it emailed to the email address on the account which belongs to my husband. I took the number back for him. His is a good number and easy to remember bec
20:55:22 <ehird> ause 5 of the 6 numbers are consecutive. "
20:55:27 <ehird> SUBTLE BRAGGING OH I DIDN'T KNOW IT WAS A BIG THING
20:55:30 <ehird> EVEN THOUGH I THINK ABOUT IT
20:56:32 <ehird> "I'm 151029. Although admittedly I bought it."
20:57:39 <ehird> brb →→→→→→→→→→→→→→→→→→
20:57:54 -!- Gracenotes has quit (Operation timed out).
20:59:47 <oerjan> now if it were 314159...
21:01:50 -!- pikhq has joined.
21:05:02 <fizzie> Since I'm sure you're all interested in xorg internals: the root cause of the problem is that MIPOINTER(dev) (a rather hairy macro) is returning NULL there, even though I guess it shouldn't be possible, based on the fact that no MIPOINTER use checks the return value.
21:05:53 <fizzie> At least digging that deep let me know some good words for googling and finding the git commit fixing this issue.
21:05:56 -!- Slereah has quit (Read error: 110 (Connection timed out)).
21:06:14 <Deewiant> + * Add 169_mipointer_nullptr_checks.patch:
21:07:10 <fizzie> That sounds like the workaround.
21:07:35 <Deewiant> +With -nvidia, when using Xinerama, holding down a key in a text field
21:07:36 <Deewiant> +on a non-primary screen can cause an X crash. This is caused because
21:07:36 <Deewiant> +the MIPOINTER(pDev) can return a NULL pointer for a non-null pDev in
21:07:36 <Deewiant> +some cases, and the mipointer.c code lacks checks for this condition.
21:07:59 <Deewiant> I find these kind of edge cases amusing
21:08:13 <fizzie> Yes, it sounds a bit nvidia-related after all. (Though not sure what the - there in "-nvidia" means in this case.)
21:08:24 <Deewiant> 1) Only on nVidia. 2) Only with Xinerama. 3) Holding down a key. 4) Only in a text field. 5) Only on a non-primary screen.
21:08:45 <Deewiant> fizzie: Presumably the driver.
21:08:53 <Deewiant> Anyhoo, that was from http://www.nabble.com/xorg-server:-Changes-to-%27ubuntu%27-td22743305.html
21:09:06 <fizzie> Yes, it is what I googled into too.
21:09:25 <fizzie> Though I get a crash by holding down a key in a rxvt-unicode on a primary screen. At least I *think* it's the primary screen.
21:09:42 <fizzie> I had some really good reason for using Xinerama instead of nvidia's own TwinView trickery.
21:10:00 <fizzie> I think TwinView did not support "dualhead with only one head rotated" or something.
21:10:41 <fizzie> TwinView being essentially what I guess the radeon driver calls MergedFB; not sure about fglrx.
21:13:02 <fizzie> Oh, it's called BigDesktop there? Heh, that's about the silliest of the various names.
21:13:13 <fizzie> Very descriptive, though.
21:13:25 -!- Gracenotes has joined.
21:14:25 -!- oerjan has quit ("Reboot").
21:29:18 -!- oerjan has joined.
21:31:23 -!- MigoMipo has quit (Read error: 110 (Connection timed out)).
21:39:12 <AnMaster> <ehird> The best option for breaking your computer completely.
21:39:31 <AnMaster> even arch linux is more stable, And arch linux tends to be very bleeding edge
21:41:05 * ais523 wonders if anyone runs sid+experimental with all packages installed
21:41:08 <ais523> that must break even more...
21:45:35 -!- KingOfKarlsruhe has joined.
21:50:25 <kerlo> I wonder why Conservapedia even mentions mathematics.
21:51:53 -!- M0ny has quit ("PEW PEW").
21:52:05 <oerjan> sheesh, don't you know that the liberal cancer _started_ with those ancient greek mathematicians?
21:52:20 <GregorR> So they can use nonsensical mathematical "arguments" to prove God.
21:52:52 <ehird> 21:08 fizzie: Though I get a crash by holding down a key in a rxvt-unicode on a primary screen. At least I *think* it's the primary screen.
21:54:39 <ehird> I have a file called /W
21:54:43 <ehird> it is 0 bytes and owned by me
21:55:10 <GregorR> You were using DOS-style flags one time :P
21:57:28 <ais523> ehird: /W isn't a legal filename...
21:57:33 <ehird> ais523: sure it is
21:57:37 <ehird> it's the file W in /
21:57:53 <ehird> http://www.met.police.uk/campaigns/counter_terrorism/
21:57:57 <ehird> And the UK takes one more step to 1984.
21:58:12 <ehird> looking vaguely "suspicious" for whatever reason is highly illegal! be afraid of it!
21:58:22 <fizzie> I urxvt at home, at work and on the laptop (even though the laptop runs OS X); it is my favouritey terminal.
21:58:33 <ehird> let us arrest them for committing the crime of looking suspicious.
21:58:58 <oerjan> but but ... all brits look suspicious to me
21:59:05 <ehird> fizzie: trying to get urxvt working w/ x11 on here caused my x11-continually-starts-and-restarts-even-if-I-kill-everything issue.
21:59:11 <GregorR> Better than arresting college students for using "prompt commands"
21:59:12 <oerjan> _especially_ the politicians
21:59:29 <ehird> fizzie: gave up anyway since X11 windows look like tiger, which sux
21:59:45 <ehird> http://www.google.com/search?client=safari&rls=en-us&q=x11&ie=UTF-8&oe=UTF-8
21:59:48 <ehird> first result is apple's x11 site
22:00:12 <fizzie> I just use X11.app in the full-screen-mode with evilwm; so it's just this black screen. (In Boston they'd already have arrested me.)
22:00:15 <ais523> what happens if you remove the ?client=safari bit
22:00:32 <ehird> fizzie: wait, you use OS X without aqua?
22:00:47 <ehird> and just earlier today I was talking about that in hypotheticals, saying it was possible but stupid :D
22:00:54 -!- pikhq has quit ("leaving").
22:00:59 <GregorR> So do I, only swap X11.app with VMware running Debian :P
22:01:16 <fizzie> Just for this full-screen terminal usage. I keep flipping out of it for web-browsing and things like that.
22:01:24 <fizzie> It is a silly setup, I guess.
22:02:18 <ehird> who needs more than 80x24
22:02:34 <ais523> Befunge-93 programmers
22:02:39 <ehird> GregorR: you're bonkers.
22:03:11 <fizzie> Hm. Configurated the desktop to power on from the keyboard, but only if you type "power" (because otherwise the cat would be booting it up all the time); then configured it to support suspend-to-ram mode so I can make it go to sleep without having to spend dozens of seconds in booting; but now it turns out that it wakes up from suspend-to-ram (acpi S3 thing) on any keypress, so the cat will wake it up anyway.
22:07:31 -!- olsner has joined.
22:11:29 <oklopol> http://www.met.police.uk/campaigns/counter_terrorism/ <<< the posters, are they about actual terrorist threats, or possible ones?
22:11:53 <oklopol> also sounds like a bad time to look foreign
22:13:58 <oklopol> http://xkcd.com/532/ okay i give up, explain it to me
22:14:22 <ehird> oklopol: small piano, penis, I think
22:14:49 <ais523> oklopol: been logreading?
22:15:14 <ais523> I've been mostly discussing Enigma...
22:15:33 <oklopol> ehird: err the genie gave him a smaller penis?
22:15:50 <ehird> oklopol: a player for his small penis thus the last line; I may be totally misunderstanding
22:15:50 <oerjan> oklopol: it's the reversal of a well-known genie joke
22:15:53 <ehird> it is a pretty lame joke
22:16:08 <oklopol> oerjan: okay that's language i understand, now what's that joke
22:16:42 <oklopol> ehird: well i got that but assumed, and still do, that it's not that
22:17:04 <oklopol> but well i guess it could be that. i just assumed something was actually being misheard.
22:17:05 <GregorR> Usu xkcd isn't that stupid.
22:17:10 <oerjan> oklopol: http://www.afunnystuff.com/jokes/Bar-jokes/Genie-pianist.html
22:17:49 <oklopol> i thought it was one of those "if i had a dollar..." things, but without ever having an ending.
22:18:34 <GregorR> I guess it's supposed to be an inversion of that joke.
22:18:40 <GregorR> And that's why it's supposed to be funny.
22:19:17 <oklopol> i just never heard that joke
22:22:09 <oerjan> that link may not be a very good version though, it was just the first google hit on genie pianist
22:23:13 <oklopol> i don't care about the details really
22:25:09 <GregorR> http://xkcd.com/563/ Better xkcd sex joke :P
22:32:33 <ehird> does xfree86 still exist? ais523?
22:32:51 <ais523> ehird: I'm not sure I've ever seen it
22:32:57 <ais523> presumably the source is still hanging around somewhere
22:33:02 <ais523> although no idea if it still compiles
22:33:19 <ehird> ais523: they're still releasing
22:33:24 <ehird> http://www.xfree86.org/releases/rel480.html
22:33:51 <ehird> http://www.xfree86.org/xnews/ latest news from 2005 ^_^
22:34:01 <ehird> also, I'm totally pissed off that the x.org folks got "x.org"
22:34:03 <fizzie> The xfree86.org main page is rather retro.
22:34:05 <ehird> you can't do that, dammit!
22:34:10 <ehird> ADD MORE LETTERS OR GIVE THE REST OF US SOME
22:35:01 <ehird> ais523: wow, x.org's only existed since 2004
22:35:06 <ehird> that's when xfree86 changed license
22:35:35 <ehird> x.org are switching to ...
22:37:19 -!- Judofyr has quit (Remote closed the connection).
22:38:15 <fizzie> Funny that Apple hasn't registered www.xn--u65c.com -- or in other words, www..com, where the character in there is that private-use character that contains the Apple "apple-with-a-bite-taken-out" logo in OS X fontsies.
22:38:31 <ehird> fizzie: It's not exactly worth the money.
22:38:45 <ehird> fizzie: Imagine writing a proposal to do that.
22:38:55 <fizzie> Well, I just tried to go there! If they had registered it, I might've purchased something from there. Maybe. Unlikely.
22:39:04 <ehird> "Let's waste $5 a year on a domain that only makes sense to OS X users, isn't widely supported, and that nobody goes to."
22:43:10 <fizzie> More strange: λ.org does not seem to be registered, though λ.com goes to some sort of portal-looking Λ.com.
22:43:16 <ehird> Deewiant: yeah, yeah, ofc :P
22:43:34 -!- ais523 has quit (Remote closed the connection).
22:43:45 <Deewiant> I don't know that by heart :-P
22:43:45 <ehird> fizzie: it's a squatted domain
22:43:54 <ehird> why you would squat an IDN I have no fuckin idea
22:45:07 -!- jix has quit ("...").
22:46:55 <ehird> Agh, this elliotth guy is freaky.
22:47:02 <fizzie> InterNIC's whois query returns an old record for λ.org (well, xn--wxa.org) which shows that it was registered already in 2001 by some Greek guy, then expired in 2007. Heh.
22:47:04 <ehird> Stop having my same first name -- down to the spelling -- and second initial, dammit.
22:47:26 <ehird> fizzie: that's like registering l.org (lowercase L)
22:47:53 <ehird> guys what do you do to do 'sudo foo>x'
22:47:58 <ehird> I do sudo zsh -c 'foo>x'
22:48:01 <ehird> but that's annoying
22:48:11 <Deewiant> Type sh instead of zsh and save a character
22:48:14 <fizzie> I usually do "sudo sh -c 'foo>x'" which is not really an improvement.
22:48:22 <ehird> sudo x | cp /dev/stdin foo
22:48:29 <ehird> Deewiant: I use zsh features. Sometimes. :P
22:48:57 <Deewiant> I typically don't when sudoing
22:49:14 <ehird> True, me neither. Still./
22:49:15 -!- pikhq has joined.
22:49:36 <fizzie> Actually I more often do "sudo bash" followed by "foo > x". But still.
22:49:53 <ehird> maybe I should make gt
22:50:00 <ehird> which is "cp /dev/stdin"
22:50:05 <Deewiant> I guess it's shorter to do 'su<CR>foo>x<C-D>'
22:50:25 <ehird> alias sgt='sudo cp /dev/stdin'
22:50:33 <ehird> echo foo | sgt foo
22:50:43 <ehird> Deewiant: sudo -s, bitch
22:51:05 <ehird> "...able to work with Bazaar, BitKeeper, CVS, Mercurial, and Subversion repositories."
22:51:15 <ehird> I smell a certain thing is missing.
22:51:24 <ehird> (And also a certain one thing that nobody uses.)
22:51:31 <ehird> ((And another one that only idiots use.))
22:51:41 <ehird> GregorR: AnMaster does. Also Canonical.
22:51:43 <ehird> Apart from that, uh.
22:51:44 -!- BeholdMyGlory has quit (Remote closed the connection).
22:51:56 <ehird> Deewiant: Missing: git (duh); nobody uses: BitKeeper; idiots: CVS.
22:52:09 <ehird> they're so weird,.
22:52:25 <AnMaster> <Deewiant> Git, Darcs, Monotone, Arch, ? <ehird> Deewiant: Missing: git (duh); nobody uses: BitKeeper; idiots: CVS.
22:52:44 <ehird> I think you're misinterpreting either Deewiant or me.
22:53:03 <AnMaster> Deewiant, anyway some people use git, Linux kernel, ehird. That's all I can think of.
22:53:18 <ehird> Barely anyone uses bzr.
22:53:23 <ehird> Deewiant: he's mocking:
22:53:30 <ehird> 22:51 GregorR: Do people use Bazaar?
22:53:30 <ehird> 22:51 ehird: GregorR: AnMaster does. Also Canonical.
22:53:45 <ehird> except that what I said about bazaar was true, and AnMaster's was false.
22:53:50 <AnMaster> just go look at launchpad and count non-Canonica projects
22:53:54 <fizzie> You could do something similar than env, which lets you "sudo env FOO=bar BAZ=quux command args"; maybe something like "into file command args" which will pretty much do "command args > file"; then you could "sudo into x foo".
22:53:59 <ehird> AnMaster: they only use it because launchpad does.
22:54:13 <ehird> also, most launchpad projects are tied to the canonical/ubuntu ecosystem.
22:54:18 <AnMaster> ehird, and why do they choose launchpad? Because it uses bzr
22:54:27 <AnMaster> so your argument made no sense.
22:54:29 <Deewiant> I think the first DVCS I tried was Monotone
22:54:41 <Deewiant> I didn't really get it and grudgingly set up a local SVN
22:54:45 <ehird> please, I can't deal with talking to someone with such a vague connection to reality
22:54:47 <Deewiant> Which I ended up not using much
22:54:54 <ehird> Deewiant: monotone is really weird. Arch is weirder though.
22:55:06 <AnMaster> ehird, also I find this flame war irritating. Go troll in #bzr instead.
22:55:09 <Deewiant> I think I may have looked at Darcs back then as well
22:55:15 <GregorR> Nothing can out-weird Arch/Bazaar.
22:55:15 <ehird> I went cvs → hg → darcs or something → wrote my own → git.
22:55:20 <ehird> I went svn → hg → darcs or something → wrote my own → git.
22:55:27 * AnMaster ignores ehird until he gets less irritating
22:55:31 <Deewiant> (Where 'back then' is something like 2003-2004?)
22:55:34 <ehird> AnMaster: you're seriously insecure
22:55:46 <Deewiant> Can't be 2003 actually, that'd be too early
22:55:47 <ehird> you don't need five. fucking. lines. to tell us all how much of a troll I am and that you're ignoring me
22:56:02 <AnMaster> Gracenotes, bazaar isn't very weird. It is like a mix of svn and hg IMO.
22:56:05 <GregorR> cvs -> svn -> [past this point I still use svn for certain things] -> darcs -> hg -> git -> hg
22:56:10 <ehird> AnMaster: You've never used non-NG bzr?
22:56:17 <ehird> Bazaar was an arch fork.
22:56:30 <AnMaster> GregorR, indeed hg is better than git :)
22:56:45 <Deewiant> Monotone (WTF?) -> SVN (meh) -> zip files, most of which got lost -> hg -> git
22:56:46 <ehird> It's embarrassing to see someone defend a tool they don't know shit about.
22:57:16 <ehird> My own was quite fun.
22:57:27 <Deewiant> I think the only thing I put into that SVN was a sudoku solver I wrote in D prior to CCBI
22:57:29 <ehird> It never really worked, but I actually needed to write my own, or use git.
22:57:34 <ehird> I wanted to import the whole Qt and KDE trees.
22:57:37 <ehird> Into a single repo.
22:57:40 <ehird> And have operations still go fast.
22:57:52 <Deewiant> Each commit was a full release, I imported it into git and ran diffstat a few days ago and the changes are 1000s of lines each
22:57:56 <ehird> GregorR: svn is dog slow
22:58:00 <ehird> and I didn't like git
22:58:02 <ehird> so I wrote my own.
22:58:03 <GregorR> ORLY? I thought svn scaled well.
22:58:09 <ehird> but it's still slow.
22:58:14 <Deewiant> I didn't really get version control back then but used it anyway because consensus seemed to be that it's a good idea
22:58:19 <AnMaster> svn does scale pretty well, cvs tends to be worse
22:58:24 <AnMaster> but it doesn't scale that well
22:58:32 <GregorR> ehird: If you're importing something huge, scalability is more important than raw speed (duh)
22:58:43 <ehird> lol, I just realised what this looks like to AnMaster
22:58:47 <ehird> 22:57 Deewiant: I think the only thing I put into that SVN was a sudoku solver I wrote in D prior to CCBI
22:58:48 <ehird> 22:57 GregorR: For that you need git.
22:58:48 <GregorR> Idonno, I never write anything unwiedly enough to care :P
22:58:50 <ehird> 22:57 GregorR: Or svn.
22:58:52 <Deewiant> When I started with CCBI I had got sick of SVN already and regressed to just versioned zip files
22:58:53 <AnMaster> GregorR, actually I have seen projects using svn for source code and cvs for data files
22:59:04 <ehird> for sudoku solvers settle for nothing other than git or svn!
22:59:11 <AnMaster> where the overhead of .svn (double size) was too large for data
22:59:18 <AnMaster> as in data was over 1 GB already
22:59:29 <Deewiant> Which somewhat sucks since I've lost old versions of CCBI due to that
22:59:34 <GregorR> My kitty is lying in the sun!
22:59:45 <Deewiant> I made a point of keeping a detailed changelog to mitigate that
23:00:08 <ehird> The mark of an idiot's mind: your only relation to one word is a pop culture reference.
23:00:15 <Deewiant> With the idea that if I ever need to find a change I can just use that
23:00:16 <ehird> And you must, must, must point it out. Every time.
23:00:27 <Deewiant> And the idea that I'd never need an old version anyway
23:00:34 <Deewiant> (Which has, in fact, proven true thus far)
23:00:53 <GregorR> Let's all make needlessly inflammatory statements.
23:01:14 <Deewiant> I tried to build my Sudoku solver, but the damn thing was last tried on a DMD 0.16x
23:01:24 <GregorR> <Deewiant> And the idea that I'd never need an old version anyway <Deewiant> (Which has, in fact, proven true thus far) // sounds like sheer luck.
23:01:26 <ehird> GregorR: he's /ignoring me. It's fun.
23:02:01 <Deewiant> GregorR: Seriously, what are they needed for?
23:02:50 <GregorR> Deewiant: You f***'d something up and introduced a bug and/or deleted code that it turned out you needed and/or made a change that had some subtle interactions you didn't realize, and need to go back to examine exactly what happened.
23:03:33 <Deewiant> I did, but I think my statement applies to any old version
23:04:03 <Deewiant> GregorR: I'm too good a programmer to do something like that ;-)
23:05:06 <ehird> I'm a shit programmer :P
23:05:08 <Deewiant> On a more serious note, changelogs have been good enough for me in that regard
23:05:42 <Deewiant> I don't think I've ever deleted large amounts of code that I'd need later
23:05:48 <Deewiant> For small amounts, I just rewrite it
23:05:57 <Deewiant> Probably ends up cleaner that way anyway
23:07:13 <Deewiant> Hmm, actually I think I used bisect once with CCBI to figure out a TRDS-related bug/weirdness
23:07:38 <Deewiant> I distinctly remember having done 'hg bisect' once, anyway :-P
23:09:07 <Deewiant> Hmm, I think GDC's latest release could compile my Sudoku solver out of the box, unless it's too new
23:10:48 <Deewiant> Haha, an entry from the changelog for the "next version to be released" in August 2006
23:10:51 <Deewiant> - Internal: following the fixing of Issue 314 in the D Bugzilla, finished
23:10:53 <Deewiant> making most imports selective. ******
23:11:08 <Deewiant> That issue currently has the most votes in the Bugzilla and still isn't fixed
23:11:50 <Deewiant> But yeah, anyway, GDC 0.19 could be worth a try
23:12:06 <ehird> someone should make a vcs in prolog
23:12:10 <ehird> slow as fuck but always right!
23:12:51 <ehird> "The reason many people have different opinions about the way to Heaven is because they have never heard the truth that is found in the Word of God, also called the Holy Bible. "
23:13:01 <ehird> http://heaventruth.org/
23:13:23 <Deewiant> I think I'll go to bed instead of commenting on that, or worse, reading it ->
23:13:43 <ehird> "7. People that have been redeemed by the Lord Jesus Christ can never lose their salvation."
23:13:51 <ehird> I accept the Lord Jesus Christ as my saviour.
23:17:35 <pikhq> Now, if you just did, bravo. Welcome to Christianity, brother.
23:18:20 <ehird> pikhq: is it seriously widely accepted that once you've done that, nothing you can do can keep you away from heaven?
23:18:23 <ehird> I find that hard to believe
23:19:41 <pikhq> ehird: Somewhat common amongst Protestants.
23:19:56 -!- GreaseMonkey has joined.
23:20:02 <pikhq> Fairly certain that's not believed in Catholicism.
23:20:10 <ehird> (23:17 pikhq: Now, if you just did, bravo. Welcome to Christianity, brother. ← this was sarcasm right? you never know on the interwebs :P)
23:20:51 <pikhq> GregorR: Sweetness.
23:20:59 <ehird> how can that be partially sarcastic o.O
23:21:03 <pikhq> GregorR: I'll fetch the latest Plofiness once I'm done upgrading my Debian box.
23:21:19 <ehird> That thing still being made?
23:21:45 <pikhq> Slowly but surely.
23:22:43 <GregorR> ehird: I get back to it whenever I can.
23:22:56 <GregorR> ehird: http://codu.org/plof/ , and CNFI is the C Native Function Interface.
23:23:03 <ehird> You mean... an FFI.
23:23:44 <ehird> CNFI is not a very nice name.
23:23:49 <ehird> May I suggest "C FFI".
23:24:41 <ehird> GregorR: Did my suggestions ever get implemented? :P
23:24:50 <GregorR> You may. But I don't see a major benefit between 'native' and 'foreign' ... I guess the one little nit to it is that the interpreter which supports C{N,F}FI may be written in (e.g.) Java, making the C{N,F}FI not really a /native/ function interface in some sense.
23:25:10 <ehird> GregorR: I made about a gazillion syntax suggestions about a year ago
23:25:17 <ehird> You said you liked some of them, iirc :P
23:25:26 <GregorR> Well, the syntax has changed quite a bit since then, so maybe :P
23:25:35 <GregorR> Although some things I'm sure you still won't like.
23:26:03 <ehird> GregorR: They mainly revolved around making "if" look like aproper control structure, with less () and ; cruft, which in turn made the whole of the syntax quite a lot more light-weight and, IMO, neither.
23:26:05 <GregorR> if((a), (bleh)) is now if (a) (bleh), so it still uses () instead of {} but allows them not to look quite so nasty.
23:26:22 <ehird> GregorR: this was before all that, I think.
23:26:38 <ehird> You still had {}s. :P
23:26:59 <GregorR> I still do have {}s, but they /only/ represent full, real functions.
23:27:22 <ehird> Hmph, you should have implemented my suggestions, they were nice :P
23:28:02 <ehird> GregorR: Well, I still see a lot of semicolon infestation :-P
23:28:38 <GregorR> The semicolons I'm not sure I want to change. They do eliminate a lot of grammatic ambiguity quickly.
23:29:04 <ehird> GregorR: My edits basically made a semicolon be implied if the current expression is complete.
23:29:36 <GregorR> That's super, except that function calls are functional-style, so the only way to tell that an expression is complete is if a semicolon is used or the containing block is closed.
23:29:37 <ehird> i.e., if it's waiting for something it carries over to the next line, but if we're not waiting for something (That is, we could accept something next if it came, but we don't need any more to be valid) then we plonk a semicolon in.
23:29:59 <ehird> we need more to be valid
23:30:07 <ehird> GregorR: duh, but that's not how it's resolved
23:30:12 <GregorR> Ahh, so it's if more is /needed/ to be valid.
23:30:40 <ehird> Lines so long you have to break them up are probably a red flag anyway, and you can always one of thems line continuation thingymabobs,.
23:31:23 <GregorR> The change is possible, but requires a vote of the Plof board.
23:31:38 <GregorR> The vote is in, we're deadlocked.
23:31:42 <ehird> GregorR: You and pikhq, I assume.
23:32:00 <ehird> GregorR: how many patches make up one vote?
23:32:06 <GregorR> Lemme do a quick grep of the copyright lines, actually IIRC pikhq is in ...
23:32:24 <GregorR> (But multiple patches does not multiply vote power)
23:32:30 <GregorR> Oh yeah, pikhq still has living code.
23:32:34 <GregorR> OK, waiting for one more vote :P
23:32:42 <ehird> GregorR: So if I made a patch implementing it, that'd count as a "FOR" vote, and it gets in. Unless pikhq votes against, in which case I guess you're the tiebreaker.
23:32:56 <ehird> GregorR: how's the parser thingy set up
23:33:00 <GregorR> It has to be a patch that actually gets accepted, that patch would be pending awaiting vote ;)
23:33:01 <oklopol> then why isn't he the tiebreaker now
23:33:14 <ehird> GregorR: aieeeeeeeee
23:33:17 <ehird> "photographing anything to do with transport is strictly forbidden." — london police
23:34:02 <ehird> i want a revolution plox
23:35:20 <ehird> GregorR: So yeah, how's that parser thang set up?
23:36:39 <GregorR> core/pul contains the entire core library. The order (which is relevant because the grammar is built incrementally) is in Makefile.pslcode, but suffice to say that pul.plof comes first, then object.plof, then the rest. pul.plof comes in with a /very/ basic grammar and builds things around some basic boxing, implemented by pul_eval, pul_funcwrap and a few other wrapping functions.
23:36:58 <ehird> GregorR: Parser. The actual parser infrastructure/
23:37:05 <ehird> I am wondering how easy it would be to mod to do this.
23:37:31 <GregorR> Well, the mod would be entirely to pul.plof, which is why I started there. Remember, the grammar of Plof is built at runtime by the core library :P
23:37:48 <ehird> Oh yeah, that crazy shit.
23:38:01 <ehird> (Metalogic does it better.)
23:38:08 <GregorR> Basically, you'd need to change how it uses "white" (the whitespace nonterminal), adding two different whites for "expression not terminated whitespace" and "expression terminated whitespace"
23:38:38 <GregorR> Then just change them as appropriate in the rest of the grammar (probably only need to make the standard-white -> nonterminated-white in a few places, actually)
23:38:39 <ehird> wouldn't it be to base.psl
23:38:44 -!- FireFly has quit ("Later").
23:38:58 <GregorR> Nah, that implements a particular definition of whitespace, but there's no reason to use it if it's not what you need.
23:39:09 <ehird> GregorR: oh god, it's a parser and a compiler in one
23:39:09 <GregorR> Touching base.apsl has been known to be hazardous to health :P
23:39:31 <ehird> .apsl looks like forth
23:39:36 <ehird> you should replace PSL with forth :P
23:39:41 <ehird> designed for plof.
23:39:42 <ehird> that would be cool.
23:39:50 <GregorR> That would be cool *shrugs*
23:40:00 <ehird> I wonder if there's a git-hg.
23:40:18 <GregorR> In fact, I could probably start with a very basic forth, add the PSLish bits that are necessary, and make very minimal changes to the rest for everything to work.
23:40:41 <ehird> GregorR: It's just that forth would be about as efficient and vastly more human-mungable than PSL
23:40:54 <ehird> And, well, it's Forth. Forth implementation techniques are pretty.
23:42:07 <ehird> GregorR: I might write that, or something. It doesn't look too hard.
23:42:14 <ehird> Is cplof maintained?
23:43:13 <GregorR> cplof is going to be the primary Plof "very soon"
23:43:31 <GregorR> But right now, cplof doesn't have a parser. I may or may not decide to keep it that way, depending on whether I tie Plof into a knot and put the parsing in Plof.
23:43:40 <ehird> GregorR: So, in this hypothetical Forth version, I could just ignore dplof.
23:43:46 <ehird> That's nice; I'm not too good at D.
23:44:46 <GregorR> dplof was nice for making an implementation quickly, but cplof will be better for portability, etc.
23:45:26 <ehird> GregorR: If I say the word "hsplof", how much will you kill me?
23:45:52 <GregorR> I'd love to see it, but I hardly thing that that's the best implementation strategy for the very-imperative PSL or Forth :P
23:47:13 * ehird decides that a very useful feature for a forth base language is <N and >N, for values of N from 0 to, say, 16.
23:47:26 <ehird> They're "stackpointer -= N" and "stackpointer += N", pretty much.
23:47:36 <ehird> So <1 is drop, and >1 regains that value.
23:47:54 <ehird> 1 2 3 <2 . >1 is "1 2 . 3" in final effect.
23:49:50 * ehird has some sort of compulsion to download high-quality, huge (3 digit gigs) source files of just about anything when presented with the oppertunity
23:50:57 <GregorR> Gay_group_porn_source_footage.raw 221G
23:53:09 <GregorR> I wonder how efficiently 3D "raster" video could be implemented using only simpl(ish) changes to (e.g.) MPEG-4
23:53:24 <ehird> GregorR: what do you mean?
23:54:02 <GregorR> Take a 3D scene, assign a color (and importantly, alpha) value to every cubic millimeter cell, and encode the whole thing in some way derived from MPEG-4.
23:54:05 <ehird> GregorR: So, Plof function calls are now 'f a b c'?
23:54:25 <ehird> GregorR: then why 20 False.ifTrue(psl { "BAD" print });
23:54:35 <ehird> surely False.ifTrue (psl { "BAD" print }); would make more sense
23:54:52 <GregorR> Probably just old *shrugs*
23:54:59 <ehird> GregorR: what would f(a,b,c) do?
23:55:03 <ehird> call f with the tuple (a,b,c)?
23:55:18 <GregorR> There are no tuples (yet?), and f(a, b, c) is an alias for f a b c
23:55:21 <ehird> GregorR: also, why can't it be:
23:55:30 <ehird> while taking a function seems reasonable
23:55:36 <ehird> it'd just be while (condition) { func(); }
23:55:44 <ehird> GregorR: final question -
23:55:48 <ehird> so "f()" is special cased?
23:55:51 <GregorR> Because when you return from a function, only that function's scope vanishes.
23:55:54 <ehird> consider f being a variable with a function
23:55:59 <ehird> "f" should reference f
23:56:09 <ehird> GregorR: what about the scope?
23:56:18 <ehird> while (foo) { return 2; }
23:56:21 <GregorR> while (a) { return 3; } <- returns 3 from the inner functiopn
23:56:25 <ehird> GregorR: Aight, first question.
23:56:36 <ehird> GregorR: It seems to me like (...) vs {...} is thin vs thick in a new skin.
23:57:07 <GregorR> () isn't a function, it's an expression that just may or may not be evaluated due to the laziness of the language.
23:57:15 <GregorR> Which is in some ways similar I suppose.
23:57:21 <GregorR> Internally it's all the same.
23:57:30 <GregorR> But semantically it's supposed to have different implications.
23:57:32 <ehird> GregorR: So, what you're saying is that Plof is Haskell.
23:57:42 <ehird> Except not as good at being Haskell as Haskell is :P
23:57:54 <GregorR> Because it's not Haskell :P
23:58:05 <ehird> GregorR: I dunno, passing imperative stuff around and having it silently evaluated or not seems pretty insidious to me
23:58:28 <ehird> GregorR: Thought regarding thick vs thin:
23:58:37 <ehird> The *caller* could decide.
23:58:41 <ehird> While would just do f(), being thin.
23:58:49 <ehird> f() would be thick.
23:58:54 <ehird> But you could do f.thin() or whatever.
23:59:08 <GregorR> I don't recall, it wasn't good :P
23:59:17 <GregorR> You could override at either site I beileve.
23:59:36 <ehird> Well, having f() = thick and f.thin() = thin lets you do {} for control structures, and they just have to do .thin()
23:59:42 <ehird> while still having {} be functiony functions