00:05:27 -!- quintopia has quit (Ping timeout: 252 seconds).
00:20:03 -!- pikhq_ has quit (Ping timeout: 246 seconds).
00:20:08 -!- pikhq has joined.
00:49:57 -!- quintopia has joined.
00:55:23 <zzo38> What are you going to eat? Paper or plastic?^W^W^W
00:58:10 <Ilari> Heh this article I'm reading reminds me of what I regard as one of the most delusional programming-related comments I have ever heard, which basically said that if C# ever goes out of style, it is easy to auto-translate the program into something else. :-)
00:58:46 -!- iconmaster has quit (Quit: Goodbye, cruel world!).
01:08:54 <oerjan> zzo38: coarse bread slices, one with bacon/liver paté, one with blue cheese, and 1/2 with chicken/curry baguette spread. and a cup of orange juice.
01:12:34 * oerjan basks in saved daylight
01:14:10 <zzo38> I don't like daylight saving time. It doesn't save daylight; it just mixes up the time so that it doesn't match.
01:16:03 -!- quintopia has quit (Ping timeout: 246 seconds).
01:18:13 -!- lament has joined.
01:45:10 -!- quintopia has joined.
01:47:35 -!- zzo38 has quit (Remote host closed the connection).
01:58:46 -!- Wamanuz5 has joined.
02:02:14 -!- Wamanuz4 has quit (Ping timeout: 264 seconds).
03:16:13 -!- copumpkin has joined.
03:21:16 -!- Mathnerd314 has joined.
04:33:45 -!- elliott has joined.
04:33:50 <elliott> http://www.basis.netii.net/ursala/links.html
04:33:54 <elliott> ursala's links page links to esolang
04:33:58 -!- elliott has left ("Leaving").
04:35:20 -!- lediable has joined.
04:40:47 -!- lediable has quit (Quit: Leaving).
05:18:34 -!- zzo38 has joined.
05:43:48 <zzo38> I noticed RFC1 has some errors in it.
05:49:41 <myndzi> that's why they call it a request for comment
05:49:47 <myndzi> as opposed to a Stone Tablet
06:02:07 <zzo38> Yes. Some of the diagrams are mistyped.
06:02:44 <zzo38> Also, I downloaded it and it had no carriage returns, but I think RFC format is supposed to be printable text? It does have form feeds.
06:06:11 <zzo38> I wrote the program ANYTODVI it can print out this RFC by using a file with printer codes with Meta Printer Language. After inserting carriage returns it print correctly. (But it still has the errors in the ASCII diagrams)
06:10:10 -!- pikhq_ has joined.
06:10:41 -!- pikhq has quit (Ping timeout: 276 seconds).
06:19:10 -!- oerjan has quit (Quit: Good night).
06:50:06 -!- pikhq has joined.
06:50:22 -!- pikhq_ has quit (Ping timeout: 250 seconds).
07:04:11 -!- pikhq has quit (Read error: Connection reset by peer).
07:08:00 -!- lament has quit (Ping timeout: 276 seconds).
07:15:49 -!- lament has joined.
07:37:44 -!- asiekierka has joined.
07:50:53 -!- sftp has joined.
07:50:59 -!- sftp_ has quit (Read error: Connection reset by peer).
07:54:16 -!- zzo38 has quit (Remote host closed the connection).
09:28:47 -!- MigoMipo has joined.
09:42:36 -!- Phantom_Hoover has joined.
09:42:37 -!- wareya has quit (Read error: Connection reset by peer).
09:43:12 -!- wareya has joined.
09:45:42 -!- wareya has quit (Read error: Connection reset by peer).
09:53:16 -!- wareya has joined.
09:53:29 -!- Phantom_Hoover has quit (Ping timeout: 240 seconds).
10:15:42 -!- sebbu2 has joined.
10:15:43 -!- sebbu2 has quit (Changing host).
10:15:43 -!- sebbu2 has joined.
10:16:19 -!- FireFly has joined.
10:16:38 -!- Phantom_Hoover has joined.
10:19:06 -!- sebbu has quit (Ping timeout: 276 seconds).
10:19:07 -!- sebbu2 has changed nick to sebbu.
10:28:20 -!- copumpkin has quit (Ping timeout: 250 seconds).
10:28:46 -!- copumpkin has joined.
10:37:57 -!- asiekierka has quit (Read error: Operation timed out).
10:41:51 -!- ais523 has joined.
11:10:43 -!- asiekierka has joined.
11:20:46 -!- azaq23 has joined.
11:29:14 -!- cheater- has quit (Ping timeout: 246 seconds).
11:43:30 -!- cheater99 has joined.
11:46:08 -!- myndzi\ has joined.
11:49:25 -!- myndzi has quit (Ping timeout: 250 seconds).
12:22:03 -!- Sgeo has quit (Read error: Connection reset by peer).
12:22:34 -!- Sgeo has joined.
12:23:00 -!- variable has quit (Read error: Operation timed out).
12:24:17 -!- variable has joined.
12:42:02 -!- cheater99 has quit (Ping timeout: 246 seconds).
12:56:39 -!- cheater- has joined.
13:39:48 -!- copumpkin has quit (Ping timeout: 246 seconds).
13:40:17 -!- copumpkin has joined.
13:44:09 -!- iconmaster has joined.
13:57:46 -!- sebbu2 has joined.
14:00:38 -!- sebbu has quit (Ping timeout: 260 seconds).
14:09:59 -!- sebbu has joined.
14:09:59 -!- sebbu has quit (Changing host).
14:09:59 -!- sebbu has joined.
14:12:25 -!- sebbu2 has quit (Ping timeout: 250 seconds).
14:12:31 -!- cheater- has quit (Read error: Connection reset by peer).
14:14:51 -!- cheater- has joined.
14:38:58 <fungot> Phantom_Hoover: if the list is in random order, like poor ehird here
14:39:30 <Phantom_Hoover> @tell cpressey That data "scientist" somehow got an article into the Scientific American.
14:47:55 -!- sebbu2 has joined.
14:47:55 -!- sebbu2 has quit (Changing host).
14:47:55 -!- sebbu2 has joined.
14:50:33 -!- sebbu has quit (Ping timeout: 250 seconds).
14:52:05 <Vorpal> Phantom_Hoover, what "scientist"?
14:52:16 <Vorpal> `addquote <fungot> Phantom_Hoover: if the list is in random order, like poor ehird here
14:52:17 <fungot> Vorpal: " too much free time, fine :) i think that sicp is http://mitpress.mit.edu/ sicp/, the texinfo at http://www.neilvandyke.org/ sicp-texi/ ( texinfo) and http://twb.ath.cx/twb/ canon/ sicp/ ( html)
14:52:19 <HackEgo> 337) <fungot> Phantom_Hoover: if the list is in random order, like poor ehird here
14:52:36 <Vorpal> Phantom_Hoover, some fraud I guess, considering the quotes.
14:53:09 <Vorpal> Phantom_Hoover, no clue what that would even mean. Perhaps information theory?
14:54:21 <Vorpal> Phantom_Hoover, or bad translation. In Swedish for example, computer science is called datavetenskap (data here comes from dator, the Swedish word for computer)
14:57:43 <fizzie> "Tietojenkäsittelytiede" ("the science of processing information") in Finnish.
14:58:55 <Vorpal> fizzie, doesn't really work for the purpose of bad translation however
14:59:04 <fizzie> Well, no, but I doubt it's that.
14:59:14 <fizzie> "Data science" seems to be a somewhat hip neologism.
15:00:00 <fizzie> "Data science requires skills ranging from traditional computer science to mathematics to art. Describing the data science group he put together at Facebook (possibly the first data science group at a consumer-oriented web property), Jeff Hammerbacher said --" and so on, from google-hits.
15:00:13 -!- sebbu has joined.
15:00:13 -!- sebbu has quit (Changing host).
15:00:13 -!- sebbu has joined.
15:03:03 -!- sebbu2 has quit (Ping timeout: 260 seconds).
15:04:32 <Vorpal> fizzie, so what do they do?
15:05:56 <fizzie> I am not entirely sure, but it seems to be rather close to data mining, but also information visualization and such.
15:09:35 -!- azaq23 has quit (Ping timeout: 252 seconds).
15:10:49 -!- elliott has joined.
15:10:58 <elliott> (data scientists are like fizzie employed to do his #esoteric log stuff)
15:11:04 -!- elliott has left ("I really must stop it").
15:12:14 -!- asiekierka has quit (Ping timeout: 252 seconds).
15:12:25 <Vorpal> I agree with oerjan.... Worst implementation of channel avoidance ever
15:12:55 <fizzie> The lure of #esoteric, it is impossible to resist, it seems.
15:13:48 <Vorpal> fizzie, speaking of you doing your log stuff... Can you plot channel activity showing how much it declined since elliott decided to mostly leave it?
15:14:06 <Vorpal> I presume more than just his part
15:14:12 <Vorpal> since others would be less active
15:14:45 <Vorpal> iconmaster, you did. By several minutes ;P
15:16:10 -!- azaq23 has joined.
15:16:40 <fizzie> Our channel activity has quite a large variance, so I'm not sure how easy it is to see trends.
15:16:45 <fizzie> When did he leave, anyway?
15:16:59 <Vorpal> fizzie, some days or weeks ago
15:17:32 <fizzie> Well, here's messages/day from my own logs: http://p.zem.fi/donv
15:19:46 <Vorpal> fizzie, there is a clear longer term drop there for a while
15:21:22 <fizzie> Same numbers after filtering out elliott-messages: http://p.zem.fi/donv2
15:21:36 -!- oerjan has joined.
15:21:56 <Vorpal> fizzie, it doesn't surprise me that goes down as well.
15:22:31 <Vorpal> discussion and not monologue and so on
15:24:46 <oerjan> <Vorpal> Phantom_Hoover, or bad translation. In Swedish for example, computer science is called datavetenskap (data here comes from dator, the Swedish word for computer)
15:24:56 <oerjan> i am pretty sure dator comes from data
15:25:23 <oerjan> given that the first is a portmanteau and the second is a genuine latin word
15:26:05 <oerjan> hm well actually it may be a genuine latin word too, but i think that use is still a portmanteau.
15:26:41 <oerjan> (in norwegian the word is "datamaskin")
15:27:17 <fizzie> Heh, that's a silly ad. It says (in Finnish by geolocation, I guess, but translated) "Congratulations! Your IP address has been chosen in your city!! You have 2 minutes to claim your reward! T:TT [Click here]", and T:TT is a countdown from 2:00 down, with the last ten seconds in red; then when it reaches 0:00, it stays there for ~10 seconds, then restarts again from 2:00.
15:27:27 <fizzie> I must be very lucky to win over and over again.
15:27:28 <oerjan> (the genuine latin word "dator" would mean one who gives)
15:27:41 <Zwaarddijk> of data and various latinish words ending in -or
15:28:15 <Zwaarddijk> in Swedish, computer science is also called informationsbehandling in some universities
15:28:16 <oerjan> Zwaarddijk: well that's what i'm saying
15:28:33 <oerjan> our university called it informatikk
15:28:48 <Vorpal> <oerjan> i am pretty sure dator comes from data <-- yes indeed
15:29:20 <Vorpal> oerjan, but I'm saying datavetenskap comes from dator, not from data. Likely.
15:29:23 <Zwaarddijk> datavetenskap can easily be construed without involving any dator in how it' been formed
15:29:43 <Vorpal> Zwaarddijk, yes it could. But since datavetenskap is about computers...
15:29:44 <oerjan> Vorpal: unlikely. since norwegian does similar things and doesn't _have_ the word dator.
15:30:01 <Zwaarddijk> computers is to computer science what telescopes is to astronomy
15:30:33 <Vorpal> Zwaarddijk, good point
15:31:17 <Zwaarddijk> datavetenskap is about how to do things with data, essentially.
15:31:51 -!- elliott has joined.
15:32:03 <elliott> has anyone ever fucked with gcc spec files
15:32:10 <Vorpal> elliott, I looked at them
15:32:19 <Vorpal> and then I went and hid under the be
15:32:36 <elliott> the thing is, I can dump them, modify them, and get gcc to use them
15:32:45 <elliott> but i have no idea what it's _generated_ from initially
15:32:50 <elliott> and I want to change /that/
15:32:52 <elliott> (to change default libc path)
15:32:57 <elliott> (and make -static default)
15:33:20 <Vorpal> elliott, I suspect ais523 is the right person to ask
15:33:36 <elliott> just because he wrote a backend doesn't mean he knows all the other crap, gcc has like 50000000000000000 lines of code
15:33:57 <Vorpal> I think you got a few too many zeros there, but yeah
15:34:08 <elliott> relatedly, gcc 3.4.6 has no dependencies (well apart from libc) and builds in a minute or two. compare with 4.5.
15:34:27 <elliott> (bootstrapping non-glibc libc. the most painless way is to bootstrap via gcc 3.)
15:34:32 <Vorpal> elliott, you have a nice computer then. I'd say 5-7 minutes for 3.4.6
15:34:47 <elliott> well, point is I immediately expected it had errored out when it finished
15:34:54 <elliott> after having spent all day building gcc 4.5.2 over and over again
15:35:24 <Vorpal> elliott, 4.3 is probably the best in the 4.x series. Either that or 4.4
15:36:05 <elliott> Vorpal: i would just stick with gcc 3 if i could maintain the delusion that it'd compile anything.
15:36:16 <Vorpal> elliott, are the spec files generated during compilation? Sure they aren't just included?
15:36:32 <elliott> Vorpal: i have find(1)'d the whole source tree for any file with name *spec*
15:36:39 <Vorpal> elliott, hey it can compile cfunge!
15:36:40 <elliott> checked them all, no generator
15:36:59 <Vorpal> elliott, grep -R <some static looking string from the spec file, such as a header or whatever> .
15:37:34 <elliott> elliott@elliott-MacBookAir:~/stage2/gcc-3.4.6$ grep -r '*version:' .
15:37:34 <elliott> elliott@elliott-MacBookAir:~/stage2/gcc-3.4.6$
15:37:46 <elliott> if it is generated, it's in a file without "spec" in the name, and in a totally different format
15:38:00 <elliott> i think it is instead cobbled together by code. this is just a hunch though.
15:38:02 <Vorpal> elliott, it could be that the : is added separately from the string :
15:38:17 <Vorpal> from the string version
15:38:18 <ais523> I know a bit about gcc backends, but not frontends
15:38:22 <elliott> it just sets the variables internally based on configuration and the like, I suspect
15:38:26 <elliott> and only parses if you give it a file
15:38:42 <Vorpal> elliott, somewhere it must do the dumping?
15:38:53 <elliott> but the dumping isn't what i want to change :)
15:39:03 <Vorpal> elliott, and what about where it can load dumped spec files?
15:39:14 <elliott> dumper + parser for data structure.
15:39:19 <elliott> data structure initialised manually w/ code.
15:39:29 <Vorpal> elliott, wait, I remember gentoo used to patch the spec file... Don't know if they still do...
15:39:39 <elliott> I could just do that, but it seems awfully ugly.
15:39:44 <elliott> since it looks distinctly generated
15:40:11 -!- azaq23 has quit (Ping timeout: 246 seconds).
15:40:15 <Vorpal> elliott, well patch the internal variables being set then?
15:40:17 <oerjan> <Zwaarddijk> datavetenskap can easily be construed without involving any dator in how it' been formed <-- although it would probably be as accurate as astronomy before telescopes too
15:40:29 <elliott> i don't know where they're set. also that sounds even uglier :)
15:40:53 <Vorpal> elliott, well if your suspicion is right I doubt there is much choice
15:42:03 <elliott> i'm a pessimist, i'm hoping my suspicion is wrong.
15:42:29 <Vorpal> elliott, don't you mean optimist?
15:42:48 <elliott> an optimist would predict that it is perfect.
15:45:50 * oerjan sets fire to elliott's optimist strawman
15:48:08 <oerjan> <Vorpal> I agree with oerjan.... Worst implementation of channel avoidance ever
15:48:20 <oerjan> i've been trying to resist suggesting a betting pool
15:51:32 -!- asiekierka has joined.
15:51:54 <oerjan> (mostly because that might make him try harder)
15:55:00 <Vorpal> oerjan, I would think he is back for a while now
15:56:23 * oerjan gives elliott a welcome back swat -----###
15:56:25 <Vorpal> elliott, there is one thing to try. It involves gdb however. Assuming that gcc uses a generated spec file, it would have to load it during normal operation right? If it doesn't, then the loader won't be called. Set a breakpoint in some important part of the loader and check if it is ever called.
15:56:36 <elliott> i just haven't devoted the energy to typing /part yet, also i look lagged
15:56:53 <elliott> really i'm waiting for someone who knows gcc to magically prance in
15:56:58 <elliott> i hear this is where all the gcc experts hang out
15:57:04 <oerjan> argh, will we be thrown back into despair again!
15:57:12 <elliott> Vorpal: gdb on gcc, yes, that sounds fun, i bet gcc doesn't run 3985734958347958345 lines just to dump out spec file
15:57:31 <elliott> Phantom_Hoover: modifying spec files is unsupported for some incomprehensible reason :)
15:57:33 <Vorpal> elliott, okay do it with printf debugging
15:57:55 <elliott> Vorpal: or bug people in here until it gets so annoying that someone figures out the solution just to make me shut up
15:57:56 <Vorpal> elliott, and gdb works fine in chroots as long as /proc is mounted
16:01:24 -!- azaq23 has joined.
16:12:29 -!- asiekierka has quit (Ping timeout: 248 seconds).
16:18:12 -!- asiekierka has joined.
16:23:29 -!- variable has quit (Quit: Daemon escaped from pentagram).
16:27:01 -!- azaq23 has quit (Ping timeout: 276 seconds).
16:31:20 -!- augur has quit (Remote host closed the connection).
16:31:37 <fizzie> Again couldn't help myself, had to use Google chart thing to plot those previous numbers: http://p.zem.fi/donv3
16:32:22 <Vorpal> fizzie, why not graphviz?
16:32:36 <fizzie> Because graphviz plots graphs, not charts?
16:33:02 <Vorpal> fizzie, I meant gnuplot
16:33:10 <Vorpal> fizzie, I confuse their names all the time -_-
16:34:09 <fizzie> No reason, really; mangling the data into a Google chart API URL was probably approximately as bothersome as writing the corresponding gnuplot datafile + command would've been.
16:34:12 <elliott> http://tunes.org/~nef/logs/esoteric/?C=M;O=D size column is distinctive enough
16:34:25 <elliott> fizzie: also gcharts are prettier :P
16:34:54 <fizzie> Oh, I could probably get quite similar output from gnuplot, it just always requires fiddling.
16:40:28 -!- FireFly has quit (Read error: Connection reset by peer).
16:43:39 -!- FireFly has joined.
16:52:57 -!- augur has joined.
16:53:22 -!- oerjan has quit (Quit: leaving).
16:58:09 -!- azaq23 has joined.
16:58:21 <fizzie> Compare http://p.zem.fi/donv3 vs. http://zem.fi/~fis/donv.png -- lines are a bit less thick, and to get anti-aliased fonts I'd need to give a path to a .ttf file in some complicated manner (or use the "pngcairo" terminal, but I don't know how to do custom colors there), but other than that they're quite close.
17:01:44 <Vorpal> fizzie, the gnuplot antialiasing is better
17:02:20 <fizzie> That's because it doesn't have any.
17:06:46 -!- sftp has quit (Remote host closed the connection).
17:14:21 -!- asiekierka has quit (Ping timeout: 276 seconds).
17:17:13 -!- azaq23 has quit (Ping timeout: 260 seconds).
17:20:01 -!- azaq23 has joined.
17:24:11 -!- elliott has left ("Leaving").
17:25:26 -!- sftp has joined.
17:44:41 -!- zzo38 has joined.
18:00:19 <Gregor> `addquote <Phantom_Hoover> elliott, incidentally, I started my explorations again after getting bored of the Himalayas.
18:00:22 <HackEgo> 338) <Phantom_Hoover> elliott, incidentally, I started my explorations again after getting bored of the Himalayas.
18:00:27 -!- Phantom__Hoover has joined.
18:01:14 -!- Phantom_Hoover has quit (Ping timeout: 246 seconds).
18:14:34 <zzo38> Instead of daylight saving time, make something else: At sunrise the time is "I /". At sunset the time is "/ I". Is OK with you?
18:19:04 <ais523> I'm not convinced I understand how that works
18:19:45 <lament> i have no idea what you're talking about, but in the past it was common to measure time by dividing the interval between sunrise and sunset into a constant number of hours
18:19:54 <zzo38> How it works is you do not need to keep changing the day for when daylight saving time is, because it is always at sunrise/sunset, no exceptions.
18:20:28 <lament> what if there's a big mountain to the east of you, so sunrise comes later?
18:20:51 <zzo38> You would still have the 24-hour clock too, going from "00:00J" to "23:59J" (or Z if you want the same time everywhere)
18:21:15 <lament> does J stand for Jupiter
18:21:25 <zzo38> No. J stands for local time.
18:21:59 <zzo38> No. Z stands for GMT/UTC/Zulu time.
18:22:43 <zzo38> lament: And I am not sure what happen if there is the mountain. Don't they count sunrise by the horizon though? I don't know?
18:24:04 <Gregor> What I really want is for the value of a second to vary depending on the season. That helps a lot with organizing timetables. "We'll have a one-hour meeting. No, wait, that'll be in February, let's make it a two-hour meeting, that should be about an hour."
18:24:50 <lament> it will also vary depending on the latitude
18:25:12 <zzo38> Gregor: Why? That doesn't seem very good.....
18:25:18 <Gregor> "We'll be having a one-hour meeting (1.5 hours in Canada)"
18:25:21 <Gregor> zzo38: This is /your idea/.
18:26:27 <zzo38> Gregor: It isn't my idea for hours or seconds or anything like that to vary at any time.
18:26:59 <lament> i think time should be absolute, everywhere
18:27:13 <lament> in order to achieve that we might have to accelerate the entire universe to the speed of light
18:27:21 <lament> it seems worth it, no?
18:27:36 <zzo38> Time cannot be absolute, spacetime is relative.
18:27:55 <Gregor> zzo38: Then your idea is to have a rather arbitrary symbol meaning "sunrise" and another one meaning "sunset". But we already have these symbols, they're "sunrise" and "sunset".
18:28:01 <zzo38> But that doesn't mean we cannot use the same units everywhere.
18:28:58 <zzo38> Gregor: These symbols mean something different. Where during the daylight hours, we have a roman numeral followed by a fraction (in normal digits, with / for zero and 1/2 for one half, 3/4 for three quarter), in night hour is other way around.
18:34:29 <Gregor> And the roman numeral indicates what?
18:34:36 <Gregor> And the fraction indicates what, for that matter?
18:35:22 <zzo38> I do not know the units yet.
18:38:31 <zzo38> I do not know if they should be measured in hours, or in something else.
18:38:50 <coppro> Gregor: we already have that
18:38:54 <coppro> it's called Newfoundland
18:41:59 <zzo38> coppro: What does Newfoundland have to do with it?
18:49:31 <Gregor> Anybody with a medium-profile website familiar with those persistent semi-spam link exchange emails?
18:49:40 <Gregor> I just got one saying he'd send me a hat in return for a link :P
18:49:44 <Gregor> I'm actually tempted :P
18:54:36 <quintopia> or make it nonclickable. just plaintext
18:54:51 <quintopia> "copy and paste this if yoj want to go look at some shit"
18:55:36 <Gregor> Alternatively I could just put it there, wait for my hat, then remove it :P
18:57:13 <ais523> just make it really clear, "someone sent me a hat for this link"
18:57:27 <ais523> (the underline's going to be stripped by the +c mode, I assume?)
18:58:09 <zzo38> ais523: Yes, I agree, write that on there.
19:02:33 <Gregor> OH BTW GUYS: libc.so auction starts in five hours; if you haven't donated, now's a great time!
19:06:34 <zzo38> As a part of something I was testing, I had successfully converted RFC1 to DVI format without using TeX. Although when I downloaded it the carriage returns were missing and it came out with only the first line correct and the others too far to the right. When I added the carriage returns then it came out correct.
19:06:49 * Phantom__Hoover realises that he'd never pay £10 for an email address under normal circumstances.
19:07:19 <zzo38> I do not need or want any email address.
19:07:33 <Lymia> Are you sure zzo38 isn't a Markov chain bot?
19:08:09 <ais523> unless it was fed with zzo38ese as the seed information
19:08:22 <ais523> I assume zzo38 has his own competitive protocol with email
19:08:52 <Lymia> It must be something like glados then.
19:08:55 <Lymia> Faulty, and insane.
19:09:17 <zzo38> ais523: Actually I simply do not use email. Knuth stopped using email first, but I did not learn that until after I stopped using email.
19:09:36 <ais523> his secretary uses email, and relays important announcements to him
19:10:31 <zzo38> I have written a letter to him using paper mail instead. As it turned out I could get it delivered by someone I know who happened to be going to that area for business purposes, so I did not need a stamp.
19:11:14 <Phantom__Hoover> Does zzo38 have... wait, he *does* have his own version of logic.
19:11:45 <Lymia> And it's not a consistent system either.
19:14:16 * Phantom__Hoover wonders, futilely, why he calls modus ponens "Rule of Detachment".
19:14:46 <zzo38> Phantom__Hoover: Because Hofstadter called it that.
19:20:55 <Gregor> After looking at all the VPS providers, I'm tempted to buy in to one of the super-cheap ones (like $20/yr) just to see how bad it is :P
19:24:26 * Lymia injects zzo38 with estrogen
19:24:35 <Lymia> Let's use zzo38 as a test subject!
19:26:14 <zzo38> You cannot inject me with anything, because I am too far away
19:30:25 -!- sebbu2 has joined.
19:30:26 -!- sebbu2 has quit (Changing host).
19:30:26 -!- sebbu2 has joined.
19:33:24 -!- sebbu has quit (Ping timeout: 240 seconds).
19:38:19 -!- impomatic has joined.
19:40:51 -!- sebbu2 has quit (Read error: Connection reset by peer).
19:43:05 -!- sebbu has joined.
19:43:05 -!- sebbu has quit (Changing host).
19:43:05 -!- sebbu has joined.
19:45:39 <zzo38> Would it be a useful machine command to have the INTERCAL select operator as well as its opposite command (the unselect operator)?
19:46:48 <ais523> is unselect unambiguous?
19:47:05 <ais523> select would be useful, though, I think
19:47:12 <zzo38> ais523: Yes, if all the extra bits are set to zero.
19:47:27 <ais523> I'm not sure that'd be so useful
19:47:34 <ais523> perhaps a command to do arbitrary permutations on bits would be better
19:48:03 <zzo38> ais523: Yes, maybe some command can be made something like that.
19:50:32 <zzo38> How do you think it would be done, what way of making such things would be best?
19:54:18 <zzo38> How much space and how much time needed for hardware implementations of different thing compares, such as addition, subtraction, multiplication, division, LFSR, NAND, XOR, etc?
19:56:01 <Vorpal> and NAND is probably the simplest one in CMOS
19:56:04 <Deewiant> Linear feedback shift register
19:57:50 <zzo38> Vorpal: Perhaps NAND is the simplest one in CMOS, but how to the others compare? (Both in space and in time) And is there other implementation other than CMOS depend which ones are more energy efficient or whatever are used in modern computers (which I don't know)?
19:58:15 <Vorpal> CMOS is the most energy efficient I know of
19:58:58 <Vorpal> and uh, sub is basically add but negating value first
19:59:00 <zzo38> Vorpal: OK. But still, how to the other things I listed compares? Like, how does LFSR compare with arithmetic? How does AND compare with OR and XOR? etc
19:59:35 <Vorpal> zzo38, too complicated to explain since I'm going to bed shortly. I suggest consulting a text book on digital logic.
19:59:49 <zzo38> Vorpal: And yes I know how subtract/add works like that, I have written programs in INTERCAL to do addition and subtraction, so I would have figured out these things.
19:59:53 <ais523> zzo38: arithmetic's tricky, e.g. there are several versions of addition, some better in time and some better in space
20:00:21 <Vorpal> ais523, generally you use full adders with some carry forwarding, no?
20:01:08 <ais523> as for logic, in CMOS NAND/NOR/NOT are equally easy, AND/OR have twice the space and time cost, XOR has something like three times the space cost and twice the time cost but I can't remember exactly
20:01:50 <ais523> with different synthesis methods it might work differently, e.g. I think there are TTL versions where NAND is easier than NOR with typical logic levels (unlike CMOS, TTL has asymmetrical levels)
20:02:13 <zzo38> But how does addition compare to multiplication, and how does addition compare to LFSR, and how does addition and LFSR compare to an increment or decrement counter?
20:02:19 <Vorpal> TTL wastes power though afaik
20:02:37 <Vorpal> zzo38, multiplication is way worse
20:03:07 <zzo38> Vorpal: By how much?
20:03:14 <Vorpal> not sure about LFSR. probably easier than addition.
20:03:37 <Vorpal> as no clue how inc/dec compare to addition. Slightly easier I'd guess
20:06:15 <zzo38> How much efficient would a computer be if LFSR was used for instruction counter (and also for a few other things) instead of increment? And if addition was not normally used for indexing arrays and stuff, but only sometimes when necessary? And also if there was other operations, such as INTERCAL select? And so on?
20:06:58 <ais523> it'd be less efficient with LFSR for increment, because even though it's an easier operation, CPU time is completely irrelevant in modern computers
20:07:05 <ais523> because memory bandwidth is a much larger issue
20:07:22 <ais523> and the memory wouldn't easily be able to pick up the necessary instructions in LFSR order
20:07:30 <ais523> you're trying to optimise the wrong thing
20:07:55 <Vorpal> and besides, there are slower components in a CPU than the IP.
20:08:10 <Ilari> AFAIK NAND and NOR are both equally simple in CMOS. Both are 4 transistors one deep.
20:08:12 <ais523> because you have so much happening on one cycle
20:08:23 <ais523> Ilari: indeed, that's what I said isn't it?
20:08:52 <Ilari> Actually, NOT is bit simpler than NAND/NOR (2 transistors one deep).
20:09:04 <ais523> oh, right, I missed that
20:09:16 <Vorpal> and that's what I said to. I just took them from zzo3's list. NOT and NOR weren't listed there
20:09:25 <Vorpal> so NAND was the simplest on the list
20:09:38 <zzo38> An idea would be if you have your program loops stored to the microcode so less memory access is needed?
20:10:37 <ais523> CPUs try to do that sort of thing automatically nowadays, it's what cache is for
20:10:54 <ais523> if you want to do it by hand rather than automatically, it's better to use an entirely different architecture
20:11:15 <zzo38> What kind of entirely different architecture?
20:14:30 <ais523> something with many more ALUs
20:14:40 <ais523> so you can take advantage of how much faster they are than memory
20:14:51 <ais523> then it also has to do arithmetic and memory activity at the same time
20:16:17 <Ilari> Hmm... How many classical CMOS gates with n inputs (all significant) there are? For 1 input, there is 1 (NOT). For 2 there are 2 (NAND and NOR), For 3 there are at least 8.
20:17:32 -!- TLUL has joined.
20:21:42 <Vorpal> Ilari, define significant
20:21:51 <Vorpal> Ilari, and afaik NAND is universal, no?
20:22:54 <ais523> not just that, it's constant-time universal
20:23:05 <Vorpal> ais523, hm that means?
20:23:27 <ais523> in that you can do any finite state combinatorial circuit with NAND gates stacked just 2 deep
20:23:31 <Ilari> That is, no inputs that are completely ignored.
20:24:28 <Vorpal> Ilari, what about AND and OR then for 2 inputs?
20:24:57 <Ilari> There is no AND nor OR CMOS gate (implementing those would require 2 gates).
20:24:57 <Vorpal> or wait, are those not classical?
20:25:38 <Vorpal> Ilari, there also the identity one for one input
20:26:37 <Ilari> Well, there's no buffer gate. CMOS tends to invert the value in each step.
20:27:32 <Vorpal> Ilari, there is the straight wire
20:28:09 <Ilari> Abstract interpratation. What is number of trees with each node labeled using one of n labels, each label used at least once and trees equivalent if for every subset of labels, there is path from root to leaf using only those labels in both or in neither.
20:29:59 <Ilari> E.g. NAND corresponds to two edges (labeled A and B) one after another. NOR would be two edges (A and B) in parallel.
20:31:34 <Ilari> The 3-label ones are ABC, {A,B,C}, A{B,C}, B{A,C}, C{A,B}, {AB,C}, {AC,B} and {BC,A}.
20:32:25 -!- oerjan has joined.
20:32:45 -!- lament has quit (Ping timeout: 248 seconds).
20:35:15 <Ilari> It is also equal to number of boolean functions with n variables present, only using AND and OR as logical operations.
20:35:33 <zzo38> In INTERCAL, although there is no command for arithmetic, some things still use counting such as RESUME and FORGET, and STASH stacks. Could a variant of INTERCAL be made that works so that an implementation does not need to have any arithmetic or counting in it?
20:59:16 -!- Sgeo has quit (Read error: Connection reset by peer).
20:59:41 -!- iconmaster has quit (Quit: ISCOM W00T).
20:59:45 -!- Sgeo has joined.
21:00:59 -!- Phantom__Hoover has quit (Remote host closed the connection).
21:16:56 -!- MigoMipo has quit (Read error: Connection reset by peer).
21:41:27 <Gregor> Surely there must be some package that provides a /usr/sbin/sendmail-compatible interface without being a whole fekking MTA.
21:41:47 <Gregor> All I want is "I know how to look up MX records, connect to an SMTP server, and send mail"
21:44:17 <zzo38> Gregor: Write one if such thing does not yet exist.
21:44:47 <Gregor> I will do that with my infinite free time.
21:50:02 <fizzie> Gregor: There is "connect to a fixed SMTP server" sendmail-compatibles, if you have a suitable (ISP/provider) forwarding host handy.
21:51:01 <fizzie> (And not all the "real" mail servers are so over-complicated, I don't think, and most can be configured into a no-local-mail mode.)
21:51:39 <Gregor> "Connect to a fixed SMTP server" is so stupid, it's so much less additional effort for them to look up the proper MX host X_X
21:52:59 <fizzie> Yes, but the assumption is that the fixed SMTP server is "nearby" and properly maintained so that it can just fail if it can't reach it.
21:53:24 <fizzie> For arbitrary-destination mail you almost-need a queue for delayed-delivery attempts.
21:57:22 <olsner> plus it's common for all outgoing smtp traffic to be blocked except for a designated proxy
22:06:11 <Gregor> Maybe that's common in a business environment, but not on an arbitrary ISP.
22:06:38 <fizzie> It's common for arbitrary ISPs around these parts.
22:07:01 <Gregor> So you can only send email by using the ISP's SMTP server? Not your own, your school's, whatever?
22:07:23 -!- FireFly has quit (Quit: swatted to death).
22:07:37 <fizzie> You can usually connect to the TLS-wrapped SMTP port.
22:08:20 <zzo38> My ISP used to block SMTP, but we complained and they changed it.
22:08:27 <fizzie> I suppose that's what most authenticated SMTP uses nowadays, anyway.
22:08:31 <fizzie> They already need to provide the relay in any case, since "regular" applications like Thunderbird need an outgoing-email SMTP server.
22:08:53 <Gregor> Does your ISP provide you with an email address?
22:09:15 <Gregor> Mine doesn't, and to my knowledge doesn't provide a relay server, since they expect whatever email service you use to have one.
22:09:19 <Ilari> Ah, there are 9 3-input classic gates, not 8. And the amount of gates for n inputs is apparently A006126 in OEIS.
22:09:39 <fizzie> That's rather weird, from my viewpoint anyway.
22:09:46 <zzo38> Gregor: My ISP does, but I no longer use it and now it is full of junk and if you try to send a message to it, it will be blocked due to full mailbox.
22:09:46 <fizzie> Must be cultural differences.
22:10:16 <Gregor> fizzie: Why would you want your email address to be attached to your ISP? If you move or switch you're stuck.
22:10:58 <fizzie> Gregor: Of course you don't need to *use* it, but they all provide something like five mailboxes per connection.
22:11:54 <fizzie> Gregor: I don't think web-based email services used to provide a SMTP server anyway, "way back then". Does something like Hotmail do nowadays?
22:12:52 <Gregor> Purdue's email does, since IMAP/SMTP is the conventional way of using it.
22:13:01 <Gregor> I would imagine most school or corporate email servers do.
22:13:48 <fizzie> I'm not sure if our school does, since the "isp provides an outgoing email server" is so widespread thing here.
22:15:03 <Gregor> Y'know, "origin SMTP server != MX record for host" is one of the more common spam-identifying heuristics :P
22:15:05 <fizzie> All of hotmail/yahoo/gmail have their outgoing-mail servers in SSL/TLS non-25 ports so the no-outgoing-plain-SMTP block doesn't affect them anyway.
22:15:50 <fizzie> And that's a pretty sucky heuristic.
22:16:20 <Gregor> Since most people use webmail, it's a pretty accurate heuristic.
22:16:40 <Gregor> Of course you can lie and say you're relaying, but then that's another heuristic :P
22:18:29 <fizzie> Wouldn't that heuristic catch your sendmail-replacement-with-direct-SMTP sent emails too?
22:18:57 <Gregor> Do you understand the meaning of the word "heuristic"? :P
22:22:00 <fizzie> It just sounds like a not-so-good one to me. Though I guess it manages to not-match both your "random webmail user" and the thing I think of when someone says "email user", which is "someone with Outlook Express from the ISP's installation CD preconfigured to use the ISP's mailbox + outgoing server", which might be a bit dated view.
22:22:53 <fizzie> At least I've managed to substitute Outlook Express with that... what was it called? Not Pegasus Mail, the other one.
22:25:48 -!- elliott has joined.
22:25:53 <elliott> relevant: toad.com why do i keep
22:25:57 -!- elliott has left ("doing this").
22:27:11 <fizzie> Ah, Mr ellio "hopeless" tt strikes again.
22:28:01 <Gregor> I bought a $20 year's service at a VPS host I shall name "Retarded VPS Hosts Inc."
22:28:14 <Gregor> I was just too curious to see how awful $20/y VPS hosting is :P
22:32:19 -!- zzo38 has quit (Quit: Confusing yourself is a way to stay honest.).
22:35:04 <ais523> hmm, mysql.com was just hacked
22:37:51 <oerjan> i guess oracle couldn't predict that
22:38:44 -!- impomatic has quit (Quit: ChatZilla 0.9.86.1 [Firefox 3.5.18/20110319140258]).
22:39:13 -!- ais523 has quit (Remote host closed the connection).
22:41:42 -!- wareya has quit (Read error: Connection reset by peer).
22:42:38 -!- wareya has joined.
22:48:54 <olsner> elaborate marketing prank: look at the extra security afforded by our enterprise database!
22:56:51 * oerjan suddenly realizes "marketing" means "worm eating" in norwegian
23:14:34 <Gregor> Funny, same in English.
23:24:19 -!- TLUL has changed nick to TLUL|afk.
23:29:16 <Gregor> http://www.youtube.com/watch?v=1GaKaGwch0U Hyuk
23:29:39 <Gregor> I've never seen the original, so I can easily convince myself that this is the "correct" version :P
23:38:45 -!- TLUL|afk has changed nick to TLUL.
23:43:39 -!- TLUL has changed nick to TLUL|afk.
23:45:10 <olsner> Gregor: that's probably the best version of it