00:00:04 <uorygl> I mean, they take floppy disks. But apart from that, they are not obsolete!
00:00:12 <augur> sounds like obsolescence to me!
00:00:54 -!- osaunders has quit.
00:01:20 <uorygl> I guess one of those player pianos was actually an electronic piano.
00:02:04 <augur> ive been tempted on numerous occasions to build an electromechanical computer
00:02:07 <uorygl> Which is pretty much the same thing as a keyboard, except it's actually fashioned to look like a piano and doesn't have as many features.
00:02:25 <AnMaster> uorygl, my piano takes an usb cable
00:02:28 <uorygl> Ah yes, electromechanical computers. Have you ever designed one?
00:02:34 <augur> anmaster: "an usb cable"?
00:02:45 <AnMaster> augur, it's called MIDI over usb
00:02:48 <augur> AnMaster: its "a usb cable"
00:02:54 <uorygl> AnMaster: is it a player piano, an electronic piano, or a keyboard? :-D
00:02:58 <augur> because it doesnt start with a vowel.
00:03:05 <AnMaster> augur, you pronounce it "U S B"
00:03:20 <augur> the name for the letter "u" does not start with a vowel.
00:03:27 <AnMaster> augur, what about pronouncing it "usb" though
00:04:01 <AnMaster> augur, well then I pronounce it that way ;P
00:04:05 <augur> uorygl: partially. my idea was to try and imagine what a computer wouldve looked like if you took a morse-code like device and extended that model
00:04:21 <uorygl> My ideas have involved creating a computer using only relays.
00:04:22 -!- augur has quit (Read error: 104 (Connection reset by peer)).
00:04:33 -!- augur has joined.
00:04:36 <uorygl> My ideas have involved creating a computer using only relays.
00:04:48 <AnMaster> uorygl, also, no one uses a player piano these days
00:04:50 <augur> it always seemed odd to me that we didnt have electromechanical computers 150 years ago given that we had morse code back then
00:04:59 <AnMaster> and hardware midi to be specific
00:05:01 <uorygl> AnMaster: I have heard of MIDI player pianos.
00:05:03 <augur> AnMaster: noone uses midi these days. long live mp3.
00:05:32 <uorygl> MP3 is inferior to MIDI when it comes to things MP3 is incapable of doing and MIDI is capable of doing.
00:05:38 <AnMaster> augur, no one uses mp3, Long live flac
00:05:53 <augur> the converse is true too ofcourse.
00:06:04 <augur> i think we should build an electromechanical computer.
00:06:05 <uorygl> I agree that we ought to have had electromechanical computers since 1835.
00:06:26 <augur> you know that the morse code machines were originally like type writers
00:06:32 <madbrain> midi is the shit you use to make mp3
00:06:34 <uorygl> I think I had some idea.
00:06:40 <augur> but some douchebag convinced morse to use a push-level machine instead
00:07:00 <uorygl> I think we should build an electromechanical router. Then we could make an Internet using only 19th century technology.
00:07:01 <augur> were it not for that, we'd probably have had the internet in 1850.
00:07:04 <lament> midi is a transport protocol
00:07:21 <augur> think of it, an electromechanical type writer that could connect to a tape store remotely
00:07:29 <madbrain> unless you do music with straight musicians or something
00:07:42 <lament> i only do music with gay musicians
00:07:55 <augur> the tape would store bits directly rather than as letters so it'd just feed right into the line
00:08:16 <uorygl> So, augur, let's figure out an error correction scheme that can be implemented using relays.
00:08:32 <augur> well you wouldnt have error correction initially, right
00:08:35 <augur> you'd just have hard connections
00:08:52 <augur> and it wouldnt matter much either, because the signals are pretty strong, comparatively speaking
00:08:59 <uorygl> Isn't error correction pretty necessary?
00:09:00 <augur> AnMaster: whats that
00:09:14 <AnMaster> augur, http://en.wikipedia.org/wiki/Reed–Solomon_error_correction
00:09:16 <augur> uorygl: only when youre signals are weak
00:09:16 <madbrain> uorygl: depends on the noise levels and such
00:09:31 <uorygl> Well, the signals are going to get corrupted every time you do something to them.
00:09:47 <augur> sure but an electromechanical computer is going to emply pretty strong signals anyway
00:09:48 <AnMaster> augur, works by oversampling a polynomial, thus being able to reconstruct the missing data points if some are gone
00:09:55 <augur> so the signal degredation isnt going to be significant
00:10:02 <uorygl> If a signal goes 1,000 miles and passes through 10 routers, I think there's going to be quite a bit of signal degradation.
00:10:16 <augur> maybe maybe not, uorygle
00:10:22 <madbrain> ah, but that's transmission, not computation
00:10:26 <uorygl> Even worse if a signal is circulated for an indefinite length of time.
00:10:35 <uorygl> Besides, error correction is easy to implement, is it not?
00:10:47 -!- osaunders has joined.
00:11:14 <augur> i mean, they had routers back then for these signals
00:11:23 <augur> they already had transatlantic telegraphs
00:11:31 <augur> by like 1850 or 1880 or whatever
00:11:37 <augur> so i dont think thats an issue
00:14:20 <uorygl> Okay, so we don't need error correction.
00:15:05 <uorygl> So, what sort of signals do we want to support? Packet switching? Circuit switching with in-band signaling?
00:15:19 <uorygl> How would packets be delimited? Time? Number of bits?
00:16:39 <uorygl> In-band is good! It means you only need one band.
00:16:50 <uorygl> The phone guys came up with filters that prevented the whistling stuff, no?
00:17:18 <AnMaster> uorygl, so some freqs were forbidden? Would break said data
00:17:25 <AnMaster> you need to *escape* it instead then
00:17:39 <uorygl> Now, let me read a bit about how relays work.
00:18:01 <AnMaster> uorygl, well if you were going to build a computer with that you could do it high level anyway
00:24:20 <uorygl> It looks like in an ordinary relay, the switch is thrown by passing current in either direction.
00:25:17 <uorygl> So you could say that a packet ends whenever the voltage drops below the threshold.
00:25:25 * AnMaster locates a boot cd with gparted
00:26:16 <uorygl> Or you could use a latching relay, and say that the packet ends whenever the voltage becomes negative; then the data can include both positive and zero voltages.
00:26:20 <augur> uorygle, lets not try to rebuild TCP/IP on aethernet just yet
00:26:27 <augur> TEEHEE AETHERNET 8D
00:26:32 -!- anmaster_l has joined.
00:26:49 <uorygl> Well, the thing about routers is that they route data. I'm just pondering how that data could be delimited.
00:27:19 <augur> special bit patterns, obviously
00:27:37 <uorygl> That won't do if you're transmitting analog data.
00:27:53 <augur> special tone patterns, obviously
00:28:10 <uorygl> That requires something other than relays. :-P
00:28:26 <augur> analog data would in general!
00:28:28 <AnMaster> uorygl, what about a fixed packet size?
00:28:48 <AnMaster> x milliseconds before switching
00:29:04 <uorygl> I don't know if that's a good idea.
00:29:13 <augur> i think we should get our system up and running locally first before trying to get a transatlantic system up
00:29:22 <uorygl> I think the packet overhead could potentially be several seconds.
00:29:39 <AnMaster> uorygl, also what about bouncing with relays
00:29:48 <AnMaster> wouldn't it be a rather severe issue
00:29:57 <uorygl> We can compensate for bouncing by using error correction. >.>
00:30:24 <AnMaster> which is a rather large downside
00:30:30 <augur> guys we should really build one of these
00:30:36 <augur> we could make it all steampunkish
00:30:45 <augur> and show it off on one of the steampunk blogs
00:30:47 <AnMaster> augur, it will eat more power than my old p4, and it will also be slower
00:30:50 <uorygl> Let's each build our own and then figure out how to connect them. :-P
00:30:55 <augur> anmaster: EXACTLY! :D
00:31:06 <AnMaster> congrats, you get promoted to Intel Chief Engineer some years ago
00:31:30 <AnMaster> augur, then you were fired when they decided to produce core 2
00:32:05 <AnMaster> btw, going to go offline with this connection for a while
00:32:53 <augur> IF YOU KNOW WHAT I MEAN
00:33:07 <uorygl> Hmm. Packet switching and circuit switching aren't really that different when packets can go on for long periods of time.
00:33:19 <AnMaster> GNU Parted - a partition manipulation program
00:33:30 <augur> AnMaster: i didnt care :D
00:33:35 <augur> uorygl, have you seen those little calculators
00:33:38 * coppro goes to play more engineer of the people
00:33:40 <augur> theyre like little black drums with a crank?
00:33:50 <uorygl> Now I'm thinking that everything should be digital and packet-switched, and then we can just use special bit patterns.
00:33:54 <uorygl> I don't think I have seen those.
00:34:40 <augur> http://www.vintagecalculators.com/html/curta_i.html
00:35:11 <uorygl> What do you call that code where every 0 is encoded as 01 and every 1 is encoded as 10?
00:36:38 <uorygl> There's a name for it.
00:40:30 <uorygl> Aha. The Manchester code.
00:42:48 <Sgeo> What's the poin.. oh, error detection
00:42:49 <augur> can we not use crazy coding schemes for this? :|
00:43:02 <coppro> I wonder what a binary version of the L&S sequence would look like
00:43:15 <uorygl> I guess crazy coding schemes aren't strictly necessary.
00:43:37 <coppro> 1, 11, 101, 111011, 11110101, 100110111011
00:43:37 <uorygl> We could just transmit at a constant rate.
00:44:07 <uorygl> coppro: that's kind of an irreversible sequence.
00:44:31 <coppro> I wonder at which base it becomes reversible
00:44:38 <uorygl> Try this: 1, 011, 010101, 010011010011010011, . . .
00:46:05 <uorygl> I guess the thing about not using any code is that timing errors can happen.
00:46:20 <uorygl> Suppose they transmit 010101010101010000000000000000000000000000000000000000000000001010101010101010.
00:46:33 <uorygl> You're going to have to time all those zeros in the middle in order to know how many there are.
00:46:48 <Ilari> 0 -> 01, 1 -> 10... NRZ?
00:46:51 <uorygl> Then again, we probably could use an error correction scheme wherein that can't happen.
00:47:00 <uorygl> Ilari: congratulations, you've invented the Manchester code.
00:47:37 <oerjan> now if you iterate it, you can invent the thue-morse sequence too :)
00:48:13 <Ilari> Well, the point of coding that mangles symbols is usually to avoid long runs of same symbol, as that tends to mess up timing...
00:49:00 <uorygl> Well, Manchester coding is one way to do things. Using it, I think we can pretty much correct for every sort of error we need to correct for.
00:49:11 <oerjan> i recall someone (ais523?) saying something about using 01 and 10 to ensure there were equal number of 0s as 1s to have no net charge
00:51:03 <oerjan> uorygl: um you cannot correct for an actual switching of neighbor bits :D
00:51:23 <uorygl> Bits don't just spontaneously get switched.
00:51:40 <oerjan> no, but you could have two neighboring errors
00:52:15 <AnMaster> oerjan, in a bit, once this CD is burned, you could say I parted to use parted
00:52:20 <uorygl> One of the routers could also spontaneously explode, causing a very large number of errors. Though that's less likely.
00:52:49 <uorygl> The thing is, Manchester at least has some error correction capability for every likely type of error.
00:53:08 <AnMaster> oerjan, hey it's supposed to be a bad pun
00:53:29 <AnMaster> oerjan, you know what parted is?
00:53:32 <Ilari> There are sequences over four symbols that never have any subsequence followed by any permutation (including identity) of that subsequence. Keränen's sequence is one of those (has recursive structure with generator of length 85).
00:53:54 <oerjan> i have heard the name, and can guess it's a partition editor
00:53:58 <uorygl> Ilari: I assume that means "immediately followed".
00:54:07 <Ilari> uorygl: Oops, right.
00:55:05 -!- AnMaster has quit ("ZNC - http://znc.sourceforge.net").
00:55:23 <oerjan> Ilari: um subsequences must have at least length 2, i take
00:55:31 <Ilari> Any sequence with recursive structure has to start with "identity" symbol.
00:56:13 <oerjan> oh wait fours symbols duh
00:58:37 <Ilari> Let G be some group. Finite sequence a1 = e, a2, ... an can be expanded into infinite one by a1, a2, .. an, a2 + a1, a2 + a2, ... a2 + an, a3 + a1, ... an + an, a2 + a1 + a1, ...
00:59:04 <coppro> Ilari: Keranen's sequence has "abc" at the start, and "bac" later on... or do I misunderstand?
00:59:15 <Ilari> coppro: Immediately followed
00:59:31 <Ilari> (that was mistakenly left out).
00:59:57 <Ilari> There is no sequence over finite such that no subsequence eventually repeats.
01:00:36 <coppro> I realized that about 2 seconds after I said that
01:00:57 <Ilari> Also *infinite sequence
01:01:08 <oerjan> uorygl: hah wrong, epsilon is both at start and end
01:03:33 <Ilari> In Z2, using generator 011 would yield 011100100100011011100011011100011011011100100011100100100011011011100100011100100... or something like that.
01:04:14 <uorygl> Here, let me generate an infinite sequence.
01:06:02 <Ilari> Using elliptic-curve-type group with large amount of points could probably make some whacky sequences.
01:06:03 <uorygl> Using Thue, an excellent infinite sequence generator.
01:06:41 <Ilari> And of course there is: Monster group!
01:06:57 <uorygl> Aww, fudge. This Thue interpreter doesn't seem to be working.
01:07:23 <Ilari> Except that monster group isn't abelian, and thus one would need to define order to do the additions in.
01:08:17 <uorygl> Okay, here is an awesome sequence: 11011011101101101110110111011011011101101110110111011011011101101110110110111011011101101101#
01:09:08 <Ilari> Put all generators of monster group as generator of sequence and eventually the infinite sequence resulting would contain all elements of monster group.
01:09:22 <uorygl> That sequence consists of strings of 1s separated by 0s. Those strings have the following lengths: 2, 2, 3, 2, 2, 3, 2, 3, 2, 2, 3, 2, 3, 2, 3, 2, 2, 3, 2, 3, 2, 2, 3, 2, 3, 2, 2, 1
01:10:08 <uorygl> Ignoring the 1 on the end, that sequence consists of strings of 2s separated by 3s. Those strings have the following lengths: 2, 2, 1, 2, 1, 1, 2, 1, 2, 1, 2
01:11:42 <uorygl> Unfortunately, my Thue program simply produces what is effectively an arbitrary finite piece of a random infinite sequence. It cannot generate a continuation of that sequence.
01:11:44 <Ilari> Of course, monster group has 808017424794512875886459904961710757005754368000000000 elements and duplicates probably exist before last element is found. Not to mention that computing group addition for moster group is quite slow.
01:12:24 -!- soupdragon has joined.
01:12:52 <uorygl> At least it can generate sequences of arbitrary length.
01:13:57 <Ilari> Monster group has generator of size 2. And interestingly monster group is isomorphic with subset of 196882x196882 matrices over Z2 with matrix multiplication as addition operator.
01:15:08 <oerjan> isn't every finite group isomorphic to such a subset, really
01:16:05 <oerjan> i suppose the interesting part is that 196882 is rather small
01:16:41 <Ilari> Of course, 38 762 521 924 elements total...
01:17:38 <oerjan> although the full matrix group is probably enormous compared to the monster
01:18:31 <Ilari> Well, the full group of that size would be all matrices of that size over Z2 with determinant 1.
01:18:50 <oerjan> so, unless there is only a tiny minority of determinant 1 elements
01:19:58 <oerjan> hm you can do it by choosing independent vectors
01:20:26 <Ilari> Well, there are already 2^19381162521 upper triangular (and same amount of lower triangular) such determinant 1 matrices.
01:20:28 <oerjan> first vector anything non-zero, so 2^196882-1 elements. that's already anormous
01:24:02 <Ilari> Hmm... Is Z3 isomorphic to some multiplicative matrix subset over Z2? Its equivalent to question if there is matrix over Z2 (and if there is, what's the smallest one) that has nontrivial cube root of identity matrix.
01:24:04 -!- coppro has quit (Remote closed the connection).
01:24:31 -!- coppro has joined.
01:24:33 <oerjan> every permutation group Sn is trivially embeddable into n x n matrix
01:25:42 <oerjan> a permutation s is mapped to the matrix with M_ij = 1 is s_j = i, 0 otherwise
01:26:34 -!- lament has quit.
01:26:54 -!- osaunders has quit.
01:27:54 <oerjan> well they're even called permutation matrices iirc
01:28:35 <oerjan> http://en.wikipedia.org/wiki/Permutation_matrix
01:28:38 <Ilari> Actually, any permutation cycle of length n can be used to find subset of matrices over Z2 that are isomorphic to Zn.
01:30:36 <oerjan> ah my confusion of whether to do s_i = j or s_j = i seems to have confused other before, because the standard definition is _wrong_ :D
01:32:27 <Ilari> Actually, it seems that for any Zn, there is isomorphic nxn matrix subset over Z2 with multiplication.
01:32:58 <soupdragon> every group is a subset of a permutation
01:33:07 <soupdragon> and if you can matrixify permutations then you can matrixify any group
01:33:41 -!- MizardX has quit ("zzz").
01:34:56 <oerjan> i think when n is a prime and using permutation matrices, that may be minimal
01:35:23 <oerjan> since p does not divide (p-1)!
01:36:05 <oerjan> may ask if it is still minimal if not using permutations...
01:36:47 <oerjan> (order of S_n is n!, and subgroup orders always divide total group order)
01:41:47 <Ilari> And if isn't minimal, fun question is what is smallest n where it isn't and what would be generator matrices for that?
01:41:52 <oerjan> if it is not a prime power then it is not minimal, because n = n1*n2 with n1, n2 coprime means Zn = Zn1 x Zn2 which can be embedded in S_(n1+n2) by using blocks
01:42:18 <Ilari> That would give n=6...
01:43:09 -!- adam_d has quit (Read error: 60 (Operation timed out)).
01:43:37 <oerjan> i'm not sure S5 is minimal for Z6, though
01:44:01 <Ilari> [[0,1,0,0,0][1,0,0,0,0][0,0,0,0,1][0,0,1,0,0][0,0,0,1,0]]?
01:44:16 <oerjan> S3 is too small, only 6 elements that _don't_ commute
01:44:18 -!- lament has joined.
01:44:32 <oerjan> i meant i was not sure S4 doesn't work
01:45:24 <oerjan> come to think of it there's probably a sequence in that sequence encyclopedia for this :D
01:45:37 <Ilari> If you have program that can calculate that, try calculating 2nd, 3rd and 6th powers of matrix I gave (over Z2)...
01:46:57 <oerjan> um i can see perfectly well it's using the block method i suggested, no need to use a program
01:51:11 <Ilari> Current methods would give 2, 3, 4, 5, 5, 7, 8, 9, 7, 11, 7, 13, 9, 8, 16, ...
01:52:09 <oerjan> hm wait of course the generator must be a _single_ permutation...
01:52:10 <Ilari> Oops, that should be 11, 8, 13...
01:52:41 <oerjan> hm that means this actually is optimal, i think
01:53:21 <oerjan> dividing that permutation into cycles is the same as dividing the matrix into blocks, and the resulting order is lcm of cycle length
01:53:59 <Ilari> Oops, 11, 7, 13...
01:54:06 <oerjan> so indeed dividing n into prime powers and adding them is the best you can do
01:55:05 <Ilari> "a(n) is the minimal number m such that the symmetric group S_m has an element of order n - Ahmed Fares (ahmedfares(AT)my-deja.com), Jun 26 2001".
01:58:03 <Ilari> And indeed element of order n means subset is isomorphic to Zn.
02:00:07 <Ilari> Hmm... I guess I should write my own raytracer to properly trace this fractal pattern. Pov-Ray can't do it properly because of 255 reflections limit.
02:01:55 <Ilari> Nah. Fairly simple to write raytracer that can just deal with it.
02:02:16 <soupdragon> can you write a 'ray tracer' based on quantum physics?
02:05:04 <soupdragon> so that it renders refraction and stuff
02:05:50 <oerjan> http://en.wikipedia.org/wiki/Field_with_one_element O_O
02:11:30 <Sgeo> When I walked around at school, I couldn't help thinking that the tiles on some of the walls looked ray-traced
02:12:55 <Sgeo> Why can't the additive identity and the multiplicitive identity be the same?
02:14:21 -!- lament has quit.
02:50:33 <uorygl> It it particularly useful to say that there is no field with one element?
02:55:52 <oerjan> "For technical reasons, the additive identity and the multiplicative identity are required to be distinct."
02:56:10 * oerjan looks for actual reasons...
03:00:40 <Sgeo> Once ReactOS is released and fully stable, would there be any real advantage to using it? I mean, it will be prone to Windows viruses, presumably
03:01:28 <Sgeo> Although it would be nice that in 10 years after release, if ReactOS is still alive, it would still be considered modern, without having to pay money, unlike with Windows
03:02:31 * oerjan didn't really find any
03:02:54 <uorygl> Well, ReactOS will be free. We could extend it in ways that we can't extend Windows.
03:03:53 <Sgeo> oerjan, because then it's not what the mathematical community decided is a "field"?
03:04:49 <oerjan> that is _not_ an improvement to "for technical reasons"
03:07:08 <soupdragon> it's probably something to do with homomorphism
03:08:54 <oerjan> well it would be a terminal object... while other fields only have homomorphisms at all if their characteristics match. or wait, are there homomorphisms from Q to F_n?
03:09:28 <oerjan> *if at minimum their characteristics match
03:10:00 <oerjan> um wait no, obviously not
03:10:11 <oerjan> the characteristics must match
03:10:53 <oerjan> smallest integer n with n*1 equal to 0, or 0 otherwise
03:11:57 <oerjan> well i just wanted to point out that n is not a field element
03:12:34 <oerjan> it's "intuitively obvious"
03:13:05 <oerjan> and in characteristic 0, there is no harm in identifying them, since the rationals are always embedded
03:15:58 <oerjan> n is always a prime or 0, btw
03:17:58 <oerjan> and for each characteristic n, there is a unique prime field, those are Z_n and the rationals
03:18:33 <oerjan> and the prime field would be an initial object in that subcategory, since it embeds uniquely in every other
03:19:10 <oerjan> the field of size 1 ruins all that
03:19:45 <oerjan> it would be of characteristic 1, i guess, but still have a homomorphism to it from everything else
04:13:49 <uorygl> Hey, now. I'm not following all that well, but it seems like you have certain objects associated with p^n where p is prime, and letting n be 0 ruins stuff because p^0 is a factor of everything.
04:19:36 <oerjan> on the other hand i think that may actually fit _because_ you always have a homomorphism to the 1 element field
04:20:18 <oerjan> and you may in general have homomorphisms from F_p^n to F_p_m when n divides m, or something like that (let me check)
04:21:31 <oerjan> well you would _need_ n dividing m, at least, since then F_p^m is a vector space over F_p^n
04:22:00 <oerjan> if of dimension d, then p^m = p^(nd) => m = nd
04:27:29 * oerjan doesn't find an easy reference
04:37:11 -!- Asztal has quit (Read error: 110 (Connection timed out)).
04:55:02 -!- zzo38 has joined.
04:55:18 <zzo38> What should I name this class: http://www.dandwiki.com/wiki/User:Zzo38/untitled_class_1_(3.5e_Class)
04:57:00 <oerjan> ceci_n_est_pas_un_classe
04:59:22 <zzo38> I like this joke but don't want to call it that.
04:59:43 <oerjan> i take it the improbability drive is the important aspect
05:00:19 <zzo38> It is one of the important aspects, but probably not the most important one
05:00:57 <zzo38> That's the domain name, I didn't write it
05:01:10 <zzo38> "it" is the domain name.
05:01:38 <uorygl> I guess it's better than "dandd".
05:01:38 <oerjan> well rephrase that: the name "improbability drive" was the only part i got any gist from
05:01:47 <uorygl> But is it better than "dnd"?
05:02:16 <zzo38> OK. But can you understand how all the class features works, however?
05:07:05 <oerjan> there is something called complex abilities. however i do not know d&d sufficiently well to understand how this is any different from what any spellcasting/psionics character can usually do
05:07:59 <coppro> D&D 3.5 is not meant to be understood
05:08:32 <oerjan> well let me leave it to the experts then
05:11:20 <zzo38> The "Complex Abilities" collectively refers to the spells, powers, and potentially others.
05:13:45 <zzo38> "Complex Abilities" does refer to maneuvers, but whether or not I add that to this class is different
05:23:16 <zzo38> One of the types of complex abilities that I might add, though, is warlock invocations, if I can figure out the cost that should be applied to them
05:41:33 <zzo38> But what C.A.P cost? I want it to be not too low, and not zero like a actual warlock's invocations
05:41:49 <zzo38> And anyways, what should I name this class
05:50:38 -!- Pthing has joined.
06:02:41 <coppro> zzo38: no, add all complex abilities
06:02:55 <coppro> invocations, spells, powers, maneuvers, infusions, and anything else you can think of
06:03:06 <coppro> except you randomly get certain kinds
06:03:22 <coppro> so you never have access to all of them
06:04:23 <coppro> hmm... I want a blank white cards bot
06:04:34 <zzo38> I don't think I will put the "randomly get certain kinds" but I could figure it out if (and only if) I can assign reasonable C.A.P costs to each of them, of course, you have a limited number of complex abilities known and a limited number of C.A.P/day so you can't possibly use all of them in one day
06:06:07 <zzo38> I don't know if I really want to add all of them though, if some kinds are more powerful it might require a feat to provide access to them
06:06:28 <coppro> then again, I don't do much else with 3.5
06:06:35 <coppro> except add templates to random creatures
06:07:03 * coppro still contemplates purchasing GR's book that's 100% templates, just for the lulz
06:07:45 -!- oerjan has quit ("Good night").
06:19:08 <coppro> zzo38: ever played 1000 blank white cards?
06:19:17 <zzo38> No, I never have done so
06:19:29 <zzo38> But I have seen descriptions and rules
06:27:00 <coppro> a tad late, but otherwise well played
06:36:04 <Sgeo> "Brad Cox and Tom Love create Objective-C, announcing "this language has all the memory safety of C combined with all the blazing speed of Smalltalk." Modern historians suspect the two were dyslexic."
06:36:17 <Sgeo> I may have pasted that in here verbatim before
06:38:10 <coppro> I haven't seen it before
06:38:44 <Sgeo> coppro, http://james-iry.blogspot.com/2009/05/brief-incomplete-and-mostly-wrong.html
06:39:08 <zzo38> Well, what I know, is, my ideas for improvement of C is not Objective C or C++, but is differently
06:39:54 <zzo38> I would have a few new preprocessor directives is one thing: #xmacro #calc #string
06:41:25 -!- HackEgo has quit (Read error: 60 (Operation timed out)).
06:41:38 <coppro> Sgeo: I like the line about Java and lambdas
06:42:05 <coppro> actually, I want explanations of all of those
06:42:31 <zzo38> And also, when using #include to be able to specify after the filename any number of names, which will be defined as blank macros while including the file and reverted afterward.
06:42:52 <zzo38> Example: #include <extra.h> EXTRA_DOS_PROGRAM EXTRA_1
06:43:03 <coppro> zzo38: write those in a patch to the GNU preprocessor, propose to the C++ committee and the C committee, in that order
06:43:19 <coppro> and I still want to know what your 5 macros do
06:43:51 <zzo38> #xmacro creates a macro that does an include. Example: #xmacro Xmacro1(_1,_2) "extra.h" __Xmacro1
06:44:43 <zzo38> #calc does like #define but calculates all values when the #calc line is evaluated, instead of afterward. Example: #calc FooBar FooBar+1
06:45:35 <coppro> I'm confused about xmacro...
06:45:44 <zzo38> #string acts like #calc but does an unstringize of the result
06:45:46 <coppro> calc is sort of not really needed
06:46:08 <coppro> okay, that I could see as more useful
06:46:09 <zzo38> Example of #string: #string CHAR(x) "'" #x "'"
06:46:52 <coppro> *I don't understand what Xmacro adds, and I really don't understand what #calc adds
06:47:07 <zzo38> Still? I thought I explained it.
06:47:22 <coppro> I don't know what they add
06:48:00 <zzo38> With the example of #xmacro given: When Xmacro(a,b) is found, it works like #define _1 a #define _2 b #define __Xmacro1 #include "extra.h" and then it can revert the _1 _2 __Xmacro1 because those are local to the macro
06:49:28 <coppro> why not just use #include
06:49:42 <coppro> why does it need to be a macro?
06:49:50 <coppro> also, a macro expanding to a directive is truly horrible
06:49:55 <coppro> and should never be allowed to happen
06:51:02 <coppro> anyone know where I can find some news about hormones? Ideally on a science- or medicine-oriented site
06:51:12 <coppro> needs to be sort of recent
06:54:02 <Sgeo> Does anyone actually use Eiffel?
06:54:11 -!- EgoBot has quit (Read error: 110 (Connection timed out)).
06:54:27 <coppro> Sgeo: not as far as I know, which is a shame
06:55:23 <Sgeo> Hold on, isn't there overloaded .NET stuff? And isn't there a .NET version of Eiffel? And doesn't Eiffel not allow overloading?
06:57:47 <coppro> Most .NET languages do not support all of its capabilities
06:58:21 <coppro> and adding overloading to Eiffel doesn't seem too difficult, particularly when you have additional limitations inherited from .NET
07:06:02 <Sgeo> Are there a lot of jobs for C# programmers?
07:06:18 <Sgeo> So, a good language to learn, then
07:06:25 <coppro> if you're in it for employment, yes
07:07:30 <Sgeo> What's wrong with it, other than the .NET legal issues and the whole type thing
07:13:22 <zzo38> And also #trap #mark #unmark
07:15:45 <zzo38> #trap is used to trap compiler errors. If it is trapped, it will stop, and preprocess again with a different macro set or something, and then recompile. If ? is used it traps within the marked area. Example: #trap error ?section1 define section1_error
07:16:17 <zzo38> You have to indicate the types of errors or other stuff too, possibly with parameters in parentheses, if you put "error" it means any error
07:19:17 <zzo38> Is it really scary?
07:27:51 <Sgeo> When will MS come up with ORG?
07:27:55 -!- madbrain has quit ("Radiateur").
07:28:13 <zzo38> I don't know, maybe never
07:28:27 <Sgeo> It's just that with "COM", then ".NET"
07:28:36 <Sgeo> </explaining-the-joke>
07:28:58 <zzo38> I know why you asked
07:29:49 <zzo38> Although it can show time of received messages, it won't currently show the time of sent messages, therefore I ought to fix that
07:30:52 <zzo38> Since the message is already typed, it has to show the time *after* the sent message, instead of *before*, even though it is slightly inconvenient
07:34:47 <zzo38> Sometimes I try to play pinball and watch the IRC both at the same time.
07:35:13 -!- zzo38 has quit (Remote closed the connection).
07:45:50 <Sgeo> I must admit, the whole explicit typing thing makes autocomplete actually work, so C# has that going for it
07:46:32 <coppro> clang will bring awesome autocomplete to C++
07:46:41 <Sgeo> As opposed to PythonWin's "Oh, I see you typed .x sometime in the past, and you just typed a .. Do you want me to put .x?"
07:46:58 <coppro> kate's default code complete is like that :(
07:57:08 -!- Sgeo has quit ("Leaving").
07:59:59 -!- clog has quit (ended).
08:00:00 -!- clog has joined.
08:14:06 -!- Warriphone has quit (Read error: 110 (Connection timed out)).
08:26:47 -!- ehirdiphone has joined.
08:27:14 <ehirdiphone> Bah, no interesting people are here right now.
08:27:29 -!- ehirdiphone has quit (Client Quit).
08:53:50 -!- Slereah_ has joined.
09:04:31 -!- Slereah has quit (Read error: 110 (Connection timed out)).
09:12:27 -!- Warriphone has joined.
09:23:30 -!- jpc has quit (Read error: 110 (Connection timed out)).
09:27:28 -!- ehirdiphone has joined.
09:29:48 <soupdragon> I might read Three Worlds Collide today
09:50:24 -!- lament has joined.
09:59:51 -!- ehirdiphone has quit ("Get Colloquy for iPhone! http://mobile.colloquy.info").
10:09:41 -!- ais523 has joined.
10:28:02 <soupdragon> I have done two euler problems in IRP!!
10:30:47 <lament> how many lines did your programs have?
10:31:23 <soupdragon> although there's usually a few extra lines buttering them up
10:31:45 <soupdragon> since if you just come out with the algoorithm they tell you to piss off :(
10:37:20 -!- ehirdiphone has joined.
10:42:13 -!- MizardX has joined.
10:42:41 <soupdragon> my lexicon thinks fission is a root morpheme
10:48:08 -!- MigoMipo has joined.
10:56:33 <ehirdiphone> It would be cool if you could depend on the result of the continuation
11:02:31 -!- |MigoMipo| has joined.
11:03:29 <ehirdiphone> (if foo is 1 then it's 1+1; since 2>1 foo must be 1. So actually it's 2 in that case)
11:03:35 -!- lament has quit.
11:04:09 <ehirdiphone> (if foo is 0 then it's 0+1; the conditition should be >= 1. So it's 1)
11:06:43 <ehirdiphone> if foo = 1 then it's 1-1 = 0; foo is only 1 when $ >= 1 so contradiction
11:10:41 <ehirdiphone> soupdragon: are there any languages with bigo notation in the types?
11:13:07 -!- MigoMipo has quit (Nick collision from services.).
11:13:18 -!- |MigoMipo| has changed nick to MigoMipo.
11:18:41 -!- ehirdiphone has quit (Remote closed the connection).
11:18:56 -!- ehirdiphone has joined.
11:29:02 <soupdragon> of course there's languages whre everything takes polytime or whatever
11:29:55 <ehirdiphone> the type inferrer would actually work out the big os
11:30:30 <ehirdiphone> so you could enforce the time complexity of functions; write a function and see what complexity it has; etc.
11:31:37 <ais523> ehirdiphone: I like that idea
11:32:08 <ehirdiphone> I think it's impossible in the general case; take:
11:32:45 <ehirdiphone> but it could just require you to specify the type there
11:33:59 <ehirdiphone> fact n = if n<0 then error "argh" else ...
11:40:13 <soupdragon> to be honest, O-analysis is so difficult I can't imagine programming a computer to do it
11:51:55 <augur> ehirdiphone: what did you think of End of Time?
11:52:50 <ehirdiphone> I've only seen the first; don't spoil the second please
11:53:42 <ehirdiphone> augur: I'm psyched that Moffat is taking over
11:54:44 <soupdragon> I don't even watch it, but it's trendy enough to be complete shit
11:56:37 <soupdragon> im talking about hte new series obviously
12:03:38 <soupdragon> waters not trendly, smirnoff frosty frootz is trendy
12:04:41 <ehirdiphone> ok then being alive is trendy. if you ignore all the suicide
12:09:08 <augur> speaking of sex, ehird
12:11:20 <soupdragon> there seems to be this idea like "a million people watch it onec a week it must be brilliant" about whatever new sitcom replaced friends or lost, but in reality it's dumbed down to the LCM so people have a common language to say nothing in -- like talking about the weather except 'cool'
12:12:08 <soupdragon> apparently in hot parts of US they talk about the traffic because the weather doesn't fluctuate enough
12:17:13 <soupdragon> yeah I know I'm too cynical for my own good
12:39:04 <ehirdiphone> augur: I do not want to know what follows.
12:39:27 <augur> nothing follows i just wanted to say that to make you think that ;D
12:52:25 -!- osaunders has joined.
13:04:22 -!- Asztal has joined.
13:13:29 -!- soupdragon has quit ("Leaving").
13:25:15 -!- soupdragon has joined.
13:36:30 <soupdragon> just the 4chan and 'internet is for porn' references make me baulk
13:37:02 <soupdragon> yeah I just said it's stupid, I have nothing against stupid
13:37:24 <ais523> just think, all the amazing resources we have on the Internet are a byproduct of people wanting porn
13:49:05 -!- osaunders has quit.
14:09:04 -!- ehirdiphone has quit ("Get Colloquy for iPhone! http://mobile.colloquy.info").
14:14:44 -!- uorygl has quit (Read error: 54 (Connection reset by peer)).
14:14:48 -!- uorygl has joined.
14:16:05 -!- osaunders has joined.
14:43:26 -!- oerjan has joined.
14:45:34 <soupdragon> question: I have an idea for a website, but I have no experience in any of the necessary skills to build it myself. What do I do?
14:45:41 <soupdragon> answer: I have an idea for a faster-than-light spacecraft which would accelerate space exploration exponentially. I have no idea how to build it. Suggestions?
14:47:10 <oerjan> indeed, with a faster-than-light spacecraft you could travel backwards in time with a copy of your website
14:49:14 <oerjan> there is however a significant danger things will get messed up and you have to be saved by an anthropomorphic/stuffed tiger
14:51:48 <oerjan> yes (not that i've read them all)
15:01:50 <oerjan> <ehirdiphone> So if fuse:fusion, I guess fiss:fission.
15:02:07 <oerjan> apparently fission comes from the latin verb "findo"
15:02:32 <oerjan> which probably wasn't borrowed because it resembles "find"...
15:03:40 <oerjan> latin 3rd conjugation verbs to all sort of consonant merging and stuff
15:32:07 -!- osaunders has quit ("Bye").
16:14:32 -!- oerjan has quit ("leaving").
16:43:36 <soupdragon> Vinge has said [1] that the "important" sequel to Bookworm would have featured the first human with amplified intelligence; however, when he attempted to sell such a story to John W. Campbell, Campbell rejected it with the explanation "You can't write this story. Neither can anyone else."
16:43:46 <soupdragon> I don't get this, why can't anyone write this story?
16:46:46 <mycroftiv> soupdragon: no clue if this is what campbell was thinking, but you can claim that you cant accurately simulate an intelligence greater than your own and that would be necessary for such a story
16:47:02 <mycroftiv> i dont think thats a very good argument though personally
16:47:30 <mycroftiv> given that we generally assume that art can meaningfully portrary/refer/represent things even without actually possessing those qualities
16:52:18 -!- |MigoMipo| has joined.
16:52:26 -!- ehirdiphone has joined.
16:54:51 -!- |MigoMipo| has quit (Client Quit).
16:56:10 <ehirdiphone> C opinion poll: typedef struct _Foo Foo; struct _Foo {...};
16:56:58 <ehirdiphone> I prefer the former; it doesn't have the strange dangling name and lets you use the alias in the strict itself
16:57:19 <ehirdiphone> Deewiant: Inconsistent when you also define recursive structures.
16:57:41 <Deewiant> ehirdiphone: No, self-documenting when I do.
16:58:11 <Deewiant> "This struct is self-recursive."
16:58:26 <ehirdiphone> Anyway, it still has the freaky- dangling name.
16:58:55 <ehirdiphone> Anyway that is self evident from the definition.
16:58:57 <Deewiant> As in, not mutually recursive with something else.
16:59:11 <ehirdiphone> I admit though struct type names are a c wart
16:59:43 <Deewiant> Underscores followed by a capital letter are reserved identifiers, you shouldn't be using them :-P
16:59:50 <Deewiant> Can't remember if that was only POSIX though.
17:01:28 <ehirdiphone> No need to adorn names in the struct namespace.
17:05:33 <ehirdiphone> I wish Plan 9 C's struct inheritance was widely supported.
17:06:15 <ehirdiphone> Sure, the standard lets you do struct foo { struct bar *parent; ... }
17:06:20 <fizzie> Deewiant: C99 7.1.3 Reserved identifiers: "All identifiers that begin with an underscore and either an uppercase letter or another underscore are always reserved for any use. All identifiers that begin with an underscore are always reserved for use as identifiers with file scope in both the ordinary and tag name spaces."
17:07:47 <ehirdiphone> I wonder; does C99 let you use Unicode in identifiers? I guess not.
17:09:25 <Deewiant> ehirdiphone: Not arbitrary Unicode.
17:09:35 <Deewiant> And it's not even all letters.
17:09:51 -!- |MigoMipo| has joined.
17:10:05 <ehirdiphone> Feckless is my new favourite autocorrection of feck.
17:10:19 <Deewiant> You'll find the list in ISO/IEC 9899:1999(E) Annex D.
17:10:29 * anmaster_l wonders why there is no package for znc in arch
17:10:33 -!- MigoMipo has quit (Read error: 110 (Connection timed out)).
17:12:39 <ehirdiphone> #define if(x) if(__builtin_constant_p(x) ? (x) : !(x))
17:12:53 <fizzie> ehirdiphone: http://pastebin.com/m4406d890
17:12:56 <ehirdiphone> I am become WTF, destroyer of programmers' minds.
17:13:31 <ehirdiphone> fizzie: Thank you for that entirely useless list. :P
17:13:38 <fizzie> ehirdiphone: You're welcome!
17:13:46 <ehirdiphone> After a while you don't see the codepoints.
17:14:18 <fizzie> Blonde, brunette, redhead, bopomofo.
17:15:29 <anmaster_l> ehirdiphone, so I'm dual booting gentoo and arch atm. In the process of switching over
17:15:37 <anmaster_l> may take a bit before I drop gentoo completely
17:18:18 <ehirdiphone> anmaster_l: Abandoning source distros? You? Thought I'd never see the day
17:20:08 <soupdragon> I didn't know you could paste like that
17:20:46 <soupdragon> apply3(##apply2(_1,func)##, someval) ?
17:21:08 <ehirdiphone> I know cpp has a specific rule forbidding , interpolation but surely ## overrides it
17:21:42 <soupdragon> error: macro "apply2" requires 3 arguments, but only 2 given
17:21:42 <anmaster_l> <ehirdiphone> anmaster_l: Abandoning source distros? You? Thought I'd never see the day <-- actually I use freebsd on another system
17:24:52 <mycroftiv> this idea seems vaguely relevant at the moment, ehird might like it: in a purely source based environment, why not build static binaries with no use of libraries at all, just a preprocessing step where every function needed (and no others) is inlined
17:27:46 <ehirdiphone> cpp is repeatedly executed on its output until there is no change. there is a special define as if #define NL (newline)
17:28:27 <ehirdiphone> #define foo(x) #include #x NL #define included_##x
17:29:35 <anmaster_l> one could argue freebsd is a freebsd distro
17:31:27 <ehirdiphone> anmaster_l: in that case *bsd are just 386bsd (aka jolix) distros
17:37:24 <ehirdiphone> btw, the kernel is essentially BSD-on-Mach
17:39:42 <ehirdiphone> Linux pre-2.4 isn't even updated for security, is it?
17:44:35 -!- madbr has joined.
17:55:26 <ehirdiphone> http://www.iro.umontreal.ca/~felipe/IFT2030-Automne2002/Complements/tinyc.c wow this code is tight
17:59:55 <Deewiant> That doesn't look like a compiler
18:00:17 <Deewiant> Looks like an interpreter that prints out every variable's value at termination
18:01:59 <ehirdiphone> Regardless, it's under 300 lines of code and very readable.
18:02:58 <Deewiant> Well, line count is easily reduced by having 5 statements per line :-P
18:07:32 <ehirdiphone> Whitespace, lines with just } and similar aren't really active lines of code.
18:11:40 -!- MigoMipo has joined.
18:17:19 -!- |MigoMipo| has quit (Read error: 60 (Operation timed out)).
18:32:35 -!- AnMaster has joined.
18:36:18 -!- soupdragon has quit ("Leaving").
18:49:35 -!- ehirdiphone has quit ("Get Colloquy for iPhone! http://mobile.colloquy.info").
19:31:17 -!- lament has joined.
19:41:51 -!- harriman has joined.
19:43:03 <coppro> hmm... doesn't have the command I need
19:43:05 <coppro> time to add it I guess
19:46:09 <coppro> note that none of those work
19:47:48 <AnMaster> coppro, how is it esolang related?
19:47:55 <coppro> AnMaster: It really isn't
19:48:02 <coppro> I'm trying to add a command to move it out of here
19:48:35 -!- harriman has left (?).
19:49:12 <AnMaster> Many parallel port chipsets provide hardware that can speed up
19:49:12 <AnMaster> printing. Say Y here if you want to take advantage of that."
19:49:18 <coppro> the bot's actually modular; the games module is the only one with any real development though
19:49:38 <AnMaster> since when would you need DMA for a parallel printer XD
19:49:58 <AnMaster> well maybe a better question would be "since when wouldn't you"
19:50:18 -!- augur has quit (Read error: 110 (Connection timed out)).
19:52:52 <coppro> something to do with my bot
19:52:57 <coppro> just the way it parses some commands
20:01:11 <madbr> man, looking at chip-8
20:01:16 <madbr> interesting design
20:01:42 <madbr> sorely lacking in "easy shit that looks pretty" stuff tho :D
20:05:42 <coppro> hmm... pretty sure that isn't supposed to happen
20:05:45 <coppro> oh well... homework to do
20:30:50 -!- osaunders has joined.
20:44:00 -!- Azstal has joined.
20:44:50 -!- Aszstal has joined.
20:51:25 -!- nooga has joined.
20:59:54 -!- Asztal has quit (Read error: 110 (Connection timed out)).
21:00:40 -!- Azstal has quit (Read error: 110 (Connection timed out)).
21:14:24 -!- osaunders has quit.
21:15:24 -!- lament has quit.
21:16:01 -!- ehirdiphone has joined.
21:17:03 <ehirdiphone> I love how R5RS seems pretty benign and then you get to call-with-current-continuation. A few minutes later, it bludgeons you with dynamic-wind.
21:17:57 <madbr> hmm, how could something like chip-8 be modernized
21:18:07 <madbr> http://en.wikipedia.org/wiki/CHIP-8
21:18:16 <AnMaster> while configuring a 2.6.32 kernel:
21:18:18 <AnMaster> "This option will be removed in 2.6.29."
21:19:10 <ehirdiphone> Bah, I'm turning into a grumpy old bastard who hates everything modern.
21:19:20 <nooga> and uses an iphone
21:19:53 <AnMaster> you should say "14 years young"
21:20:41 <ehirdiphone> Gimme my libc4, my Linux 2.0, my a.out, my XFree86, my BSD userland. My X terminals!
21:21:41 <AnMaster> ehirdiphone, know that Linux 2.0 definitely didn't have any journaling fs
21:21:44 <ehirdiphone> Um. I don't know. How about ext's little-known and nonexistent predecessor, whose name is the null string.
21:21:53 <AnMaster> also what the crap is up with the spell checking on here
21:22:17 <AnMaster> it is abysmal, in fact it doesn't even know the word "abysmal"
21:22:42 <AnMaster> on gentoo I had one that knew everything, even stuff no one used any more
21:22:50 <AnMaster> ehirdiphone, well it had no suggestions for it either
21:23:11 <AnMaster> ehirdiphone, it doesn't accept Absymal either
21:23:28 <AnMaster> I'm not sure what this one uses...
21:23:41 <AnMaster> since that isn't even installed !?
21:24:42 <ehirdiphone> I forget that every so often then look at the spec again
21:24:43 <AnMaster> ehirdiphone, about "blog", won't make a difference here. It doesn't know it either.
21:25:03 <AnMaster> also it thinks that "doesn't" are two words, doesn and t, t being a valid word, doesn is not
21:27:00 <ehirdiphone> Continuation-passing style transformation and Cheney on the MTA garbage collection are good for you.
21:27:30 <nooga> how does it look in practise?
21:27:59 <ehirdiphone> R5RS is the last true Scheme, if you don't know it by that moniker.
21:28:26 * AnMaster wonders why xchat has --enable-mmx
21:28:41 <AnMaster> it seems so out of place for an irc client
21:28:46 <coppro> (actually, I don't know enough about Scheme to know why people hate R6RS and I don't particularly care)
21:28:46 <ehirdiphone> (R6RS, the latest report, defines a language superficially similar but a complete miscarriage of Scheme's philosophy in actuality.)
21:29:04 <coppro> ehirdiphone: is there a good link on this?
21:30:06 <nooga> isn't Scheme just grotesque Lisp dialect?
21:30:10 <ehirdiphone> coppro: The R6RS votes page. Note the lack of reasoning for almost all yes votes. Note the in depth objections from experienced Schemers on the no votes. Note how it only passed by a small majority.
21:31:23 <ehirdiphone> coppro: Also note that the vast majority of implementers said they would not adopt R6RS and indeed haven't.
21:31:31 * AnMaster recompiles xchat to use gtkspell so he can get that language selection pop-up menu.
21:31:53 <AnMaster> also what the crap is up with not knowing "recompiles" but knowing "recompile"
21:31:57 <AnMaster> that is just so very very broken
21:32:00 <coppro> I cannot find this page you speak of
21:32:04 <coppro> link or it didn't happen
21:33:17 -!- Pthing has quit (Remote closed the connection).
21:33:41 <ehirdiphone> Re my earlier remarks about implementing R5RS being good for you: ...but figuring out how DYNAMIC-WIND interacts with everything else is like shooting 5,000 bags of heroin a day for 1,000 years, except instead of getting high your face is stomped on by a burning poker.
21:34:36 <nooga> ehirdiphone: is it enough ungrotesque to make you want to design hardware r6rs scheme processor and write very sophisticated OS for that in r6rs scheme?
21:35:46 <AnMaster> ehirdiphone, what is DYNAMIC-WIND?
21:35:55 <coppro> AnMsater: http://www.schemers.org/Documents/Standards/R5RS/HTML/r5rs-Z-H-9.html#%_idx_576
21:35:56 <ehirdiphone> coppro: Ugh, they removed the ratification results.
21:36:02 <AnMaster> I don't remember that from r5rs? or would it be some r6rs thing?
21:36:08 <coppro> looks pretty straightforward, actually
21:36:34 <nooga> ehirdiphone: that's not an answer
21:36:38 <coppro> ehirdiphone: yes, I see
21:37:07 <AnMaster> nooga, what has r6rs got to do with being ungrotesque?
21:37:20 <AnMaster> ehirdiphone, I think he must have misinterpreted you
21:37:29 <nooga> ehirdiphone said that it's VERY ungrotesque
21:37:35 <AnMaster> nooga, ehirdiphone meant that r5rs was ungrotesque... Not r6rs
21:38:11 <AnMaster> ehirdiphone, maybe he will someday turn less idiotic by the osmosis?
21:38:15 <ehirdiphone> He's just making fun of my OS/hardware tendencies
21:38:43 <AnMaster> test -z "/etc/gconf/schemas" || /bin/mkdir -p "/home/arvid/src/system/xchat/pkg/etc/gconf/schemas"
21:38:43 <AnMaster> ../../../0 -m 644 'apps_xchat_url_handler.schemas' '/home/arvid/src/system/xchat/pkg/etc/gconf/schemas/apps_xchat_url_handler.schemas'
21:38:43 <AnMaster> /bin/sh: line 4: ../../../0: No such file or directory
21:39:15 <nooga> ehirdiphone: I was not making fun
21:41:08 <ehirdiphone> "urgh" immediately otherwise and you referring to it as r6rs when it was r5rs I praised, plus historical evidence, suggests otherwise. But whatever
21:41:32 <AnMaster> ehirdiphone, I believe he must simply have misunderstood you
21:41:35 <nooga> naaah, that urgh was about my weird grammar
21:41:53 <nooga> and i thought we were talking about r6rs
21:42:00 <nooga> nvm, i'm a fool and troll
21:42:10 <AnMaster> ehirdiphone, maybe it worked after all ^
21:42:33 <nooga> i'll better shut up
21:42:50 <AnMaster> ehirdiphone, isn't it said that realising your own faults is the first step towards getting rid of those?
21:42:57 -!- dbc has quit (Read error: 110 (Connection timed out)).
21:43:08 <AnMaster> (no offence meant to anyone here)
21:43:50 <nooga> because I LIKE watching how my behaviour infuriates ehird, it's amusing :
21:44:34 <nooga> but it's over, I promise
21:44:59 <AnMaster> UNROLL drivers/md/raid6altivec1.c
21:44:59 <AnMaster> CC [M] drivers/md/raid6altivec1.o
21:44:59 <AnMaster> UNROLL drivers/md/raid6altivec2.c
21:45:06 <AnMaster> I wonder why it is compiling that
21:45:29 <AnMaster> also how the heck does it even succeed in compiling it
21:45:47 -!- adam_d has joined.
21:47:23 <ehirdiphone> http://sisc-scheme.org/r5rs_pitfall.php I love the MAP caveat
21:50:05 -!- ehirdiphone has quit ("Get Colloquy for iPhone! http://mobile.colloquy.info").
21:50:29 -!- ehirdiphone has joined.
21:51:14 <nooga> ehirdiphone: ever tried xmonad?
21:51:42 <ehirdiphone> Configuration system sucks, generally not as good as dwm or wmii.
21:52:49 <nooga> how about awesome?
21:53:11 <ehirdiphone> awesome is dwm with a lot of stuff, mostly superfluous, added to it.
21:54:42 <nooga> I asked about xmonad because Wadler inspired me to play with haskell again
21:54:59 <nooga> last month I went to Edinburgh to visit my friend from UoE and accidentally met Phil Wadler after his lecture
21:55:10 -!- dbc has joined.
21:56:02 <AnMaster> shiretoko, hard to remember name for arch linux's firefox
21:56:37 <ehirdiphone> Blame their fucking idiotic trademark policies.
21:56:39 <AnMaster> ehirdiphone, you mean the branding thing, well yes
21:56:47 <AnMaster> but why call it "shiretoko" instead
21:57:24 <coppro> At least Debian is consistent about there
21:57:37 <AnMaster> ehirdiphone, gentoo can actually work around it, since you can compile it for personal use with official logo iirc
21:57:41 <AnMaster> so they offer a useflag for it
21:57:58 <AnMaster> ehirdiphone, source based distros are better at some stuff :P
21:58:09 <AnMaster> of course it is stupid this is required
21:58:23 <ehirdiphone> I read the debian thread that kicked it all off
21:58:51 <ehirdiphone> Debian were like "Fuck you guys, we can't call firefox firefox now"
21:59:13 <coppro> I can see the rationale behind blocking one user from using a trademark
21:59:15 <coppro> if they're bad about it
21:59:19 <coppro> but blanket policies = :(
22:00:06 <coppro> most applications are shit
22:00:33 <ehirdiphone> Firefox is a bad browser, though. There are better.
22:00:41 <coppro> Depends on the purpose
22:00:59 <AnMaster> ehirdiphone, "works on most web pages"? I think firefox manages very well there
22:01:08 <ehirdiphone> If your purpose isn't "experience hell", firefox is probably the wrong choice
22:01:21 <coppro> Firefox is pretty usuable
22:01:23 -!- ais523 has quit (Remote closed the connection).
22:01:26 <ehirdiphone> AnMaster: Webkit + presto (operas engine) do that perfectly well
22:01:40 <AnMaster> ehirdiphone, presto is open source?
22:01:54 <AnMaster> ehirdiphone, not relevant to me then
22:01:58 <coppro> ehirdiphone: how, in your mind, is it unusable?
22:02:19 <AnMaster> ehirdiphone, webkit might be worth a try
22:02:47 <ehirdiphone> coppro: Crufty, slow ui; slow performance; requires extensions to just be not retarded
22:03:00 <nooga> webkit? don't be riddiculous, just look at Slowfari
22:03:19 <AnMaster> ehirdiphone, if it supports features I use. Like adblock, noscript, firebug (script debugger and live editor for css/html plus more). And I'm not interested in your opinion on those add-ons
22:03:34 <coppro> ehirdiphone: It's not slow to the point of unusability unless you're doing something stupid; the bit about extensions is part of its appeal
22:03:58 <ehirdiphone> "ehird:" should preferably be relevant to me...
22:04:17 <coppro> (not that the complaints about speed/memory are not valid, because they are)
22:04:33 <ehirdiphone> coppro: You have to download third party stuff to make it not terrible == it is shit
22:05:04 <coppro> ehirdiphone: what exactly?
22:05:06 <AnMaster> ehirdiphone, well since you were recommending them to me
22:05:26 <AnMaster> ehirdiphone, I thought you were suggesting ones that would fit me
22:05:29 <coppro> before or after you back up your assertions?
22:05:42 <AnMaster> coppro, before, wouldn't be ehird otherwise
22:05:45 <ehirdiphone> coppro: Its all a conspiracy to avoid answering you!
22:05:59 -!- ehirdiphone has quit ("Get Colloquy for iPhone! http://mobile.colloquy.info").
22:06:25 <AnMaster> also, firefox 3.5 seems quite snappy to me, older versions less so
22:07:05 <AnMaster> nooga, codewise it is horrible
22:07:20 <nooga> and firebug is irreplaceable
22:07:24 <coppro> AnMaster: try running a ton of JavaScript
22:07:29 <coppro> my 3 complaints for Firefox 3.5:
22:07:35 <AnMaster> coppro, well I do use noscript anyway
22:07:49 <AnMaster> coppro, I seldom visit js heavy sites
22:07:53 <coppro> AnMaster: pretty irrelevant; regular page script isn't what does it
22:07:55 <AnMaster> of course firefox itself is js-heavy
22:07:58 -!- augur has joined.
22:08:09 <coppro> - JavaScript garbage collection shuts the whole thing down
22:08:16 <coppro> AnMaster: running a JS application (like ChatZilla)
22:08:36 <coppro> ChatZilla isn't terrible, but Wave just nukes it
22:08:40 <AnMaster> coppro, for irc I tend to prefer a real client
22:08:50 <AnMaster> as for wave, well google want people to use chrome, no?
22:09:03 <coppro> CZ is a real client, if I ever get it running in XULRunner
22:09:07 <AnMaster> coppro, why not replace firefox's js engine
22:09:16 <coppro> AnMaster: because they just got a knew one?
22:09:27 <coppro> at least it doesn't leak memory any more
22:09:47 <coppro> now the memory leakage is left up to Xorg
22:10:29 <coppro> nooga: I think it's other applications leaking X resources or something, but it's only cured by restarting X
22:10:38 <AnMaster> coppro, also there is a solution that google won't be able to do anything about: start using v8d in firefox
22:10:47 <AnMaster> then it will be exactly as fast as chrome
22:10:59 <AnMaster> but there are pretty large downsides to that
22:11:12 <AnMaster> coppro, whatever the one google used
22:11:24 <coppro> oh, you mean the script engine?
22:11:45 <coppro> I don't know; I think Firefox's script engine is pretty married to the rest of it
22:11:48 <AnMaster> coppro, v8d sounds like an irc network I was on years ago. long before I was on freenode
22:12:00 * AnMaster wonders if that irc network still exists
22:12:55 <AnMaster> coppro, anyway, why so married you meant
22:12:55 <coppro> oh, I also hate the abysmal set of Linux plugins, but that's not really FF's fault
22:13:07 <coppro> AnMaster: difficult to separate
22:13:10 <AnMaster> abysmal set of Linux plugins? for what?
22:13:19 <AnMaster> coppro, bad design, should be made modular
22:13:27 <coppro> AnMaster: Linux plugins cause freezes, leaks, etc. especially Flash
22:13:47 <AnMaster> coppro, oh hah, I don't use any plugins. Especially not closed source ones
22:13:50 <coppro> AnMaster: yes, I agree it's bad design; I'm not 100% sure that's the case though and it's not my problem either way
22:14:01 <coppro> AnMaster: you really care that much about open source?
22:14:12 <AnMaster> coppro, well there is one limit: nvidia drivers
22:14:33 <AnMaster> coppro, apart from that and BIOS, plus possible some firmware. I think I'm clean
22:14:34 <coppro> isn't the new open-source driver supposed to be better than the closed-source ones?
22:14:53 <coppro> AnMaster: No... what was it called... it was on /.
22:15:14 <AnMaster> bin boot dev etc home lib lib64 lost+found media mnt opt proc root sbin srv sys tmp usr var
22:15:22 <AnMaster> can't see anything about nvidia there
22:15:49 <AnMaster> "3D support is worked on using Gallium3D and can (depending on the Chip generation and the applications) be quite usable. Breakage in the 3D-drivers can (and will) however occur, they are known and are not needed to be reported. When that status change this page will be updated. "
22:15:55 <AnMaster> sounds like out of question for me
22:15:59 <coppro> According to the /. article, there's a preloader that's part of the closed-source stuff, but otherwise it's open-source
22:16:02 <AnMaster> coppro, I need flight sim and such to work
22:16:22 <coppro> still, I'm surprised you care that much about open source
22:16:24 <AnMaster> 01:00.0 VGA compatible controller: nVidia Corporation G73 [GeForce 7600 GS] (rev a2)
22:16:36 <coppro> I mean, I'm a big fan of it, but I'm not stupid about it
22:16:38 <AnMaster> coppro, can you figure out which on http://nouveau.freedesktop.org/wiki/FeatureMatrix that is?
22:17:35 -!- MigoMipo has quit ("When two people dream the same dream, it ceases to be an illusion. KVIrc 3.4.2 Shiny http://www.kvirc.net").
22:18:55 <AnMaster> coppro, probably it might be useful around the time nvidia drops driver support for my card
22:19:14 <coppro> AnMaster: why do you hate closed-source stuff so much
22:19:26 <coppro> (this is a real question, not an accusation or the like)
22:20:03 <AnMaster> coppro, binary blob, you can't fix it if it breaks. for libraries and drivers: you can't easily debug crashes in your own programs if they happen in binary blobs
22:20:40 <coppro> AnMaster: those are all reasons against using it when an alternative exists, but if there's no alternative (like with Flash)?
22:20:59 <AnMaster> coppro, the security aspect (harder to sneak in malicious code) is another part. Sure, I won't review everything myself for open source, but the fact that any user could means it is much more risky to try it in open source
22:21:24 <AnMaster> coppro, with flash there is, only flash I care about is youtube videos. Works with youtube-dl + vlc
22:21:55 <coppro> but you seem to have a rather RMSan aversion from proprietary stuff
22:21:57 <AnMaster> coppro, also, even buggy but non-malicious code tends to be more rare in open source in my experience
22:22:44 <AnMaster> for any open source project with a sufficiently large user base, there will be someone who does fix bugs he/she encounters
22:22:59 <coppro> not counting things I do at work, the only proprietary stuff I use regularly is Flash and various web applications
22:23:11 <coppro> s/proprietary/closed source/
22:23:31 <AnMaster> coppro, well, there is java mostly at university web portal system thingy
22:23:48 <AnMaster> I don't like that either but not as bad as flash
22:23:55 <coppro> Java's open source now :)
22:24:19 <AnMaster> coppro, yes but it is still buggy, open source haven't yet had full effect on it
22:24:36 <coppro> but so it's more a usability thing than a proprietary thing?
22:24:37 <AnMaster> considering how long it took for firefox to get reasonable...
22:24:46 <AnMaster> (and it still is fairly bad in part)
22:25:07 <AnMaster> coppro, also rolling release distros tend to be least buggy, and most up-to-date
22:25:21 <AnMaster> least buggy I can explain with "no deadlines".
22:25:25 <coppro> don't quite understand that
22:25:40 <AnMaster> but "most up-to-date" would require deadlines. so well quite a paradox
22:25:54 <AnMaster> coppro, consider arch linux vs. ubuntu
22:26:29 <AnMaster> ubuntu is buggier than arch I would say. More well integrated, but bugs exist and are often fixed in a "not really fixed" way
22:26:34 <AnMaster> ais could tell you more about that
22:27:05 <coppro> as an Ubuntu user, I agree about the 'not really fixed' bit
22:27:36 <AnMaster> coppro, arch on the other hand tends to be 1) more bleeding edge (sometimes uncomfortably so) 2) less buggy
22:27:59 <AnMaster> however, not as well integrated, things won't work out of box. But it won't beep at you a lot during shutdown
22:30:11 -!- jpc has joined.
22:41:45 <anmaster_l> here we go *compiles nvidia for custom kernel*
22:44:02 -!- AnMaster has quit (Nick collision from services.).
22:49:32 <anmaster_l> coppro, slightly more than 25 seconds boot time
22:50:20 <anmaster_l> to be specific, moving stuff forwards and letting them start concurrently
22:50:58 <anmaster_l> coppro, almost 5 seconds of that is mostly idling while waiting for dhcp reply
23:12:07 <coppro> anyone have a good logic game I can play quickly (like *gasp* Flash?)
23:17:13 <anmaster_l> smartd actually makes it slower before, due to more disk activity making dhclient take longer to load
23:49:50 -!- AnMaster has joined.