00:01:55 -!- FreeFull has quit (Ping timeout: 255 seconds).
00:04:42 -!- Lymia has quit (Quit: Hug~♪).
00:04:55 -!- Lymia has joined.
00:04:55 -!- Lymia has quit (Changing host).
00:04:55 -!- Lymia has joined.
00:05:58 -!- oren has quit (Ping timeout: 255 seconds).
00:09:53 -!- GeekDude has quit (Read error: Connection reset by peer).
00:10:15 -!- GeekDude has joined.
00:12:29 -!- not^v has joined.
00:16:46 -!- Lymia has quit (Ping timeout: 255 seconds).
00:23:23 -!- skj3gg has quit (Quit: ZZZzzz…).
00:26:27 `olist 973
00:26:28 olist 973: shachaf oerjan Sgeo FireFly boily nortti
00:27:26 -!- skj3gg has joined.
00:29:19 -!- Lymia has joined.
00:29:20 -!- Lymia has quit (Changing host).
00:29:20 -!- Lymia has joined.
00:29:49 -!- not^v has quit (Ping timeout: 245 seconds).
00:48:49 -!- Tritonio has joined.
00:53:53 -!- FreeFull has joined.
01:03:20 -!- mihow has quit (Quit: mihow).
01:22:06 -!- barryking has joined.
01:22:18 -!- barryking has changed nick to KrimHum.
01:22:51 -!- KrimHum has left ("Leaving").
01:24:16 -!- Phantom_Hoover has quit (Ping timeout: 255 seconds).
01:24:22 nice kink.com hostname
01:25:00 -!- dianne has joined.
01:28:06 -!- G33kDude has joined.
01:28:40 -!- GeekDude has quit (Read error: Connection reset by peer).
01:28:46 -!- G33kDude has changed nick to GeekDude.
01:34:18 -!- oerjan has joined.
01:36:46 ooh rjan
01:36:51 øøh, perhaps
01:38:14 shaxxxxaf
01:38:36 -!- Lymia has quit (Ping timeout: 272 seconds).
01:38:39 the joke is olist 973 hth
01:38:49 i noticed
01:40:31 i see simonpj has suggested simply making _every_ type Typeable as a way to solve the bug
01:41:04 and then someone suggested not requiring Typeable constraints
01:41:13 imo who needs 'em
01:41:20 yes, but i think that would be awkward to achieve...
01:41:55 in the dystopian ghc future, parametricity is but a distant memory
01:42:21 why did we ever let spj take over, we ask ourselves. but it is too late
01:42:34 * oerjan gets confused when people use nicks but still address each other with real names
01:42:58 but goldfire is richard eisenberg, right
01:43:04 yes
01:43:24 middle initial A
01:43:25 Would you be more confused by people addressing someone IRL with a screen name rather than with their real name?
01:43:34 good subject for a double dactyl
01:43:39 pikhq: let's ask Taneb hth
01:44:36 oerjan: due to great confusion of this sort i started keeping a file with nick<->name mappings for haskell people
01:44:36 pterodactyl meter
01:44:39 it has ~200 entries now
01:44:40 it's great
01:44:44 heh
01:46:45 pikhq: do people do that to you
01:47:11 Yes.
01:47:17 i would be confused how to pronounce it. but then Taneb doesn't put the stress where i would, either.
01:47:29 Fortunately I answer to it.
01:47:39 "pika pika"
01:47:43 * oerjan ducks and runs
01:47:45 Like "pikachu". "peek eitch kyuu"
01:48:10 well i really knew that
01:48:56 i don't pronounce pikachu like "peek eitch kyuu"
01:48:56 Darn stupid mouth feeling punched.
01:49:03 oerjan: :P
01:49:20 "peek a chuu" vs. "peek eitch kyuu". They sound similar.
01:49:33 if I said pikhq it'd be like p'k heich queue
01:49:33 yes but not identical
01:49:40 whereas pikachu would be peek a chew
01:50:25 i say banayna, you say banahna
01:51:08 * oerjan is inspired by the latest iwc poll
01:51:21 why're you all making 'pik' a single syllable?!
01:51:35 Cause I speak English.
01:51:52 oerhört
01:52:03 unerhört
01:52:19 When I speak Japanese I say "pikachu" as 4 morae (pi ka chu u) as is proper.
01:52:32 (the incomprehensible joke is that oe is two syllables in that swedish word)
01:52:43 (which means "unheard of")
01:52:51 o-erjan?
01:53:05 oerjan: hey I accidently translated it into german correctly.
01:53:12 yay
01:53:41 it is fairly likely to have been half-borrowed from german in the first place, specifically the -er- part
01:53:50 pikhq: have you considered a legal name change to Pik Headquarters
01:53:56 No.
01:54:03 because that prefix isn't native nordic afair
01:54:56 (and the rest of the word was probably borrowed too, it's not like the prefix was borrowed in isolation)
01:55:36 hear hear!
01:56:54 although it may very well have expanded to native words afterwards.
01:59:39 some of the traditional difference in vocabulary in nynorsk vs. bokmål is because the nynorsk supporters had a certain repulsion against such obvious german loanwords
02:00:40 so e.g. the nynorsk word for "reality" is "røynd" rather than the german-borrowed bokmål "virkelighet"
02:02:01 nynorsk was envisioned as "going back to the roots"
02:03:07 however some of those words feel so archaic that most nynorsk users today use the german-derived loanwords anyway
02:04:34 (and of course being constantly exposed to the majority bokmål media probably doesn't help with keeping the differences)
02:04:56 -!- skj3gg has quit (Quit: ZZZzzz…).
02:11:39 -!- Lymia has joined.
02:11:39 -!- Lymia has quit (Changing host).
02:11:39 -!- Lymia has joined.
02:12:56 -!- skj3gg has joined.
02:22:58 every day, the Norwegian Question becomes confuser and confuser...
02:24:05 I had no idea Wikipedia had a main contents page.
02:24:06 http://en.wikipedia.org/wiki/Portal:Contents
02:24:39 -!- skj3gg has quit (Quit: ZZZzzz…).
02:28:15 boily: correct
02:31:52 -!- tswett has joined.
02:38:23 -!- tswett has quit (Quit: Colloquy for iPhone - http://colloquy.mobi).
02:41:33 -!- boily has quit (Quit: SWUNG CHICKEN).
03:03:28 Help me I'm excited about a Microsoft product
03:05:32 * oerjan hits Sgeo with the saucepan ===\__/
03:05:36 DID THAT HELP
03:06:18 help me oerjan i'm excited about a culinary product
03:06:28 No because the saucepan isn't 'holographic'
03:06:54 shachaf: sorry you're doomed hth
03:07:02 Ok we need a better word than holographic because iirc that means a specific means of projecting 3d images, and everything interesting is using... different ways of doing it
03:07:45 i'd say holographic implies that it can be correctly viewed from different angles, unlike e.g. 3d glasses
03:08:45 (at least "ordinary" 3d glasses with a flat cinema screen)
03:09:40 So Oculus Rift and HoloLens wouldn't count, since you see the 3d screen itself from one angle?
03:10:08 i haven't read enough about those products
03:10:54 i'd say, you'd ideally also want _more than one person_ to be able to watch, from different angles, at the same time.
03:11:27 of course even laser holograms are limited in the range of angles they allow, i think
03:11:47 Augmented Reality is the term used for stuff like HoloLens, I think.
03:12:36 For that, you could get an eyetap for both eyes.
03:12:51 3D effect would be fine if emulated via head tracking.
03:14:57 * Sgeo wonders what the SDK will be like
03:15:14 I assume it's more complicated than normal 3D programming, which I don't totally understand
03:24:06 https://www.reddit.com/r/gadgets/comments/2ta829/arstechnica_handson_microsofts_hololens_is/cnxfznw
03:29:33 -!- Lilax has joined.
03:35:52 Technically speaking the force of two planets smashing into eachother is considered an explosion
03:37:37 where's that thing about "objects are just a special case of functions" / "functions are just a special case of objects" <-- i think it's one of the "koans"
03:39:48 How do you keep apostrophe's in your latex text when you convert to pdf?
03:40:27 @tell elliott Google for "Qc na"
03:40:28 Consider it noted.
03:40:59 Lilax: i think you can drop at least one apostrophe hth
03:41:00 gazoogle
03:41:29 heh
03:41:44 * oerjan has no idea why apostrophes would be a problem.
03:42:18 /is doing this for someone
03:42:21 -!- Sprocklem has quit (Ping timeout: 256 seconds).
03:42:25 I have no idea either
03:58:33 @tell elliott I suddenly realised I'm not being listened to. <-- you should have seen yesterday's tensor conversation.
03:58:33 Consider it noted.
04:00:29 @tell elliott everything he said was technically correct yet missing the point of our discussion.
04:00:29 Consider it noted.
04:01:51 maybe i should have said so outright at the time.
04:13:32 -!- Jafet has quit (Ping timeout: 246 seconds).
04:14:37 -!- Jafet has joined.
04:25:35 -!- nys has quit (Quit: quit).
04:34:12 -!- skj3gg has joined.
04:45:08 -!- Jafet has left.
05:20:06 I have a feeling HoloLens UI as a productivity tool is more good for graphic design and less good for other stuff
05:20:21 It's the whole 'do you really want to wave your arms in the air for extended periods of time' thing
05:20:42 Although mouse+keyboard+no monitor could be cool
05:25:28 -!- GeekDude has quit (Quit: {{{}}{{{}}{{}}}{{}}} (www.adiirc.com)).
05:27:20 -!- skj3gg has quit (Quit: ZZZzzz…).
05:38:52 -!- skj3gg has joined.
05:48:36 -!- skj3gg has quit (Quit: ZZZzzz…).
05:51:57 -!- skj3gg has joined.
06:02:01 -!- skj3gg has quit (Quit: ZZZzzz…).
06:03:52 -!- skj3gg has joined.
06:13:51 -!- skj3gg has quit (Quit: ZZZzzz…).
06:25:22 -!- Lilax has quit (Quit: Connection closed for inactivity).
06:41:12 -!- Jafet has joined.
06:49:53 -!- skj3gg has joined.
06:53:53 -!- MDude has changed nick to MDream.
07:00:10 hm today's girl genius is the last of a double spread, but seems to be missing the usual elegant and finely crafted link to a big version.
07:03:51 -!- skj3gg has quit (Quit: ZZZzzz…).
07:06:28 This is a hyperlink. All craftsdevship is of the highest quality.
07:09:36 -!- skj3gg has joined.
07:10:20 no no "elegant and finely-crafted link" is an official term, e.g. http://www.girlgeniusonline.com/comic.php?date=20061011
07:18:10 be two days' worth of story, but this time we thought it
07:18:11 best to use the whole thing for one day. Simply scroll down to read the second page. You can also view the whole thing by following this
07:18:11 elegant and finely-crafted link.
07:18:31 Avant-garde, daring in its asymmetry. I like it
07:19:38 -!- skj3gg has quit (Quit: ZZZzzz…).
07:23:52 -!- skj3gg has joined.
07:25:32 oerjan: Do you think I'd be better off not avoiding coördinates?
07:25:54 You did say that that was how you learned it.
07:26:09 well learned *of* it
07:26:49 pretty sure it makes more sense if you also know the categorical viewpoint
07:27:44 in fact i'm not sure there is much more to the coordinates than "pairs of basis elements for each space form a basis for their tensor product"
07:29:57 i still don't have a good motivation for a lot of the things
07:31:17 well if you understand matrices as the most obvious way of writing an element of V* (x) W in form of basis elements, and that that is canonically isomorphic to V -o W...
07:31:50 -!- Patashu has joined.
07:32:47 mind you i don't know either how to actually _calculate_ with tensors whenever derivatives are involved...
07:33:55 -!- skj3gg has quit (Quit: ZZZzzz…).
07:33:57 the basic motivation for tensors afaiu is "how to describe physics properties in terms that are independent of your choice of coordinate systems"
07:34:49 -!- skj3gg has joined.
07:34:59 and the non-coordinate way of defining things makes it bloody obvious that what you've done actually _is_ independent of them.
07:35:21 even if you can do things with coordinates and show that they behave nicely under transformations.
07:36:10 -!- Patashu_ has joined.
07:36:10 -!- Patashu has quit (Disconnected by services).
07:44:25 -!- skj3gg has quit (Quit: ZZZzzz…).
07:45:16 * oerjan still doesn't actually see how to show that V* (x) W is isomorphic to V -o W without using coordinates
07:46:02 and also, V must be finite-dimensional
07:46:28 oh well
07:46:41 doesn't it just work given that V ~ V**?
07:46:53 well that's of course pretty darn equivalent
07:47:05 to being finite dimensional
07:47:33 although maybe that means it also works for infinite dimensional hilbert spaces if interpreted right
07:48:04 -!- Froox has joined.
07:48:22 but anyway i don't see how it follows from V ~ V** either
07:49:03 Morning
07:49:14 morning
07:49:20 -!- Frooxius has quit (Read error: Connection reset by peer).
07:49:27 oerjan: wait, what was wrong with the thing i did yesterday?
07:49:37 i don't remember what you did
07:50:22 (V -o W) (V -o ((W -o F) -o F)) ((V (x) (W -o F)) -o F)
07:50:29 oh, right, and i wasn't sure about the last step
07:50:45 um it's V that needs to be finite dimensional, not W
07:50:45 but surely (V (x) W*)* = V* (x) W
07:51:02 we were talking about the case where V=W anyway
07:51:06 since it's a tensor
07:51:29 i guess now your question is more general
07:51:31 -!- chaosagent has quit (Remote host closed the connection).
07:52:47 hm
07:53:11 i'm pretty sure (V (x) W*)* is only ~ V* (x) W if the whole thing is finite dimensional
07:53:18 http://planetmath.org/tensorproductofdualspacesisadualspaceoftensorproduct
07:53:47 sounds like a German compound word
07:55:28 I sorproduct pretty infrequently.
07:55:38 shachaf: it sounds like they're constructing something extremely non-canonical
07:56:04 yes
07:57:01 and implying in a side note that it's better if one of the spaces is FD
08:17:11 http://blog.jle.im/entry/io-monad-considered-harmful
08:29:22 -!- arjanb has quit (Quit: bbl).
09:08:36 -!- CADD has quit (Quit: Lost terminal).
09:32:57 -!- skj3gg has joined.
09:42:09 -!- vanila has joined.
09:42:14 https://github.com/dramforever/kspl
09:42:18 https://github.com/dramforever/kspl/blob/master/sample.kspl
09:42:55 -!- skj3gg has quit (Quit: ZZZzzz…).
09:56:16 wow
09:56:17 ok
09:56:18 never
09:56:52 NEVER watch guiness world records top 100
09:57:06 unless you want to be disgusted
09:57:11 then go for it.
10:06:51 World's most disgusting world records program
10:20:21 vanila: you should be able to implement that with TeX...
10:21:21 yes (I didnt make this)
10:22:02 Ah
10:25:36 -!- _meta has joined.
10:31:10 -!- _meta has quit (Quit: Leaving...).
10:48:24 -!- hjulle has joined.
11:07:52 -!- aretecode has quit (Read error: Connection reset by peer).
11:10:54 -!- aretecode has joined.
11:16:52 I got my Lego BF interpreter working[0] last night. [0] It needs some attention around the rack/pinion interface, but apart from that it works :-)
11:17:48 congrats !
11:18:30 ty - I was very please when it managed to multiply 4*3 last night.
11:18:56 Unfortunately a colleague has told me the tape drive sounds like a sarcastic duck - and he's right !
11:19:22 -!- boily has joined.
11:20:32 haha
11:21:00 https://plus.google.com/photos/113373535180413523278/albums/6105388692419944513?authkey=CL2uopH9jZOkEQ
11:21:09 Vid of it at work. Quack, quack, quack ...
11:29:08 haha it really does
11:29:50 It doesn't sound as bad IRL, but maybe this is programmed in QF, for QuackF*ck ;-)
11:30:13 Hm, I can see a GIF preview but not any video.
11:30:25 click on the gif I believe
11:33:37 Are the coloured squares the program or the tape? (Unfortunately I can't find any video.)
11:34:40 yes
11:35:00 Hmm. Wonder why you've got no video.
11:35:14 https://plus.google.com/113373535180413523278/posts/D1rFySseXQ2
11:35:32 Try that - public post, so you should be able to click the gif to get the video
11:35:41 Red/Green = < >
11:35:48 Black / White = - +
11:35:58 Yellow / Brown = , .
11:36:02 Blue / Grey = [ ]
11:41:04 Ah well, I found a video URL in the HTML text.
11:46:17 -!- Patashu_ has quit (Ping timeout: 256 seconds).
11:50:37 That's interesting, the program is on the tape and I suppose the tape is in digital memory
11:50:55 I don't store anything that's on the tape.
11:51:06 The memory is internal - 256 bytes of it.
11:51:50 I have a program counter and a stack internally - used to manage the [ ]
11:52:04 -!- Phantom_Hoover has joined.
11:52:12 What are the tank treads for, though?
11:52:33 they are used to move the sensor/flipper arm in the Y-axis
11:52:44 The rack/pinion moves it in the X-axis.
11:54:34 Yes, I don't quite see what they are for
11:55:40 the arm underneath the bits is attached to them. The move that arm up/down to select which bit to test/flip
11:55:56 The rack is used to select which nibble and which direction to test/flip
11:56:17 I take it that it's not clear how they are used to move the arm in the video ?
11:56:53 That is clear, but the eight "bits" at the top of the frame aren't
11:57:00 yes
11:57:07 held above a sensor/flipper
11:58:03 i'll be back in a while - I'll post some more vids that show the flipper in operation,
11:58:10 Can you explain, how they affect the interpretation of the program (or the other way round)?
12:00:54 Oh nevermind, clearly those are the input and output ports.
12:01:48 i want to build a physical computer too
12:03:31 In principle you don't need a photosensor, you could control the tape with a stepper motor (but it will probably use an optical encoder internally)
12:18:16 -!- Phantom_Hoover has quit (Ping timeout: 272 seconds).
12:24:09 -!- boily has quit (Quit: THEOLOGICAL CHICKEN).
12:32:30 I like the sound effects of that bf lego thing
12:36:25 -!- GeekDude has joined.
12:43:22 -!- anamaria1 has joined.
12:44:52 -!- GeekDude has quit (Quit: {{{}}{{{}}{{}}}{{}}} (www.adiirc.com)).
12:46:00 -!- SopaXorzTaker has joined.
12:54:31 @let data DQ a = DQ (DQ (a, a))
12:54:33 Defined.
12:55:36 Parametricity
12:55:55 isn't that isomorphic to data X a = X X
12:56:56 oerjan: thanks, but I don't think I need the link now that I had that realisation :P
12:57:00 *I've
12:58:36 -!- ais523 has joined.
12:58:43 The Kaplan-Tarjan deque has something like DQ a = DQ0 | DQ [a] (DQ (a, a)) [a]
12:59:17 It just occurred to me that this is a very strange recursive type
12:59:28 nested data types, I think they're called
12:59:31 there's some paper about them
12:59:44 data BinTree a = Leaf a | Branch (BinTree (a,a))
13:01:05 “Mycroft calls such schemes polymorphic recursions. We prefer the term nested datatypes” are they serious
13:02:15 http://www.google.com/safebrowsing/diagnostic?site=google.com
13:03:19 polymorphic recursion is often used for the value-level version I guess
13:03:24 but yeah nested data types is a weird term
13:05:48 Please provide a regular expression that can parse JSON. Go ahead, I'll wait. ← http://www.reddit.com/r/programming/comments/2svijo/commandline_tools_can_be_235x_faster_than_your/cnvdsak
13:05:56 I did it because I could, not because it's a good idea
13:07:22 ais523: nice "test JSON"
13:08:17 I haven't tried it on any really big JSON
13:08:22 luckily, JSON has a really simple spec
13:08:31 the main thing that took me problems was when it said "except control characters"
13:08:36 *that gave me problems
13:08:39 when it works in terms of Unicode
13:08:41 but JSON is complex compared to s-expressin isnt it?
13:08:50 vanila: yes, although not that much more complex
13:09:05 s-expressions still need to distinguish between strings and numbers and atoms, normally
13:09:18 so the only difference is that JSON has dictionaries and lists, and s-expressions just have lists
13:09:20 you know the code you posted is not an regular expresison...
13:09:48 Jafet, I still need the photo sensor as that's what reads the program
13:10:01 vanila: indeed, it's a Perl regex
13:10:04 which is a superset of regular expressions
13:10:20 I knew that that was the sort of "regular expression" people wanted based on context, though
13:10:53 JSON can't be parsed by a finite state machine (proof: 1, [1], [[1]], [[[1]]], etc.)
13:11:39 [0] lol backreferences, lol sed is Turing-complete
13:11:42 im not sure I undersatnd this ^
13:12:02 or "I've just come to realize a sad fact though: processing raw text streams through mostly-regular languages is really weak"
13:12:26 sed is TC, but it has nothing to do with backreferences
13:12:31 deterministic-thue is trivial to implement in sed
13:12:44 err, to compile into sed
13:12:45 but baically what the fuck is he on about
13:12:48 is what im wondering
13:13:44 well, it's a random reddit comment thread
13:13:52 so it might just be a random sequence of words that are vaguely related to the subject
13:14:08 Or vaguely unrelated, in this case
13:14:28 does he know about parsing
13:14:43 The title post reminded me of a recent talk I attended, the speaker talked about challenges processing "several gigabytes" of medical records.
13:15:56 ic ant make any sense of it
13:16:00 it's interesting to see just how big "Big Data" is at any given point in time
13:16:19 he seems to be saying regular languages (and by this he includes non regular languages too) isnt good for everything
13:16:23 last time the question came up (a few weeks ago), someone defined it quite specifically as "the indexes of your database take up more than 224 gigabytes"
13:16:50 because with today's technology, that's the point at which you need specialized hardware to be able to fit them all in memory on a single computer at the same time
13:17:11 Which chipset has this 224 GiB limit? (Presumably AWS uses it.)
13:17:42 Jafet: that was actually the largest buyable AWS instance at the time of the comment
13:17:48 but yes, it seems like a reasonable chipset limit
13:17:50 It still is, apparently
13:18:22 (and costs $3 per hour. Tempting...)
13:20:25 I think most chipsets have rounder numbers as limits
13:21:30 224 is moderately round
13:21:44 $ factor 224
13:21:45 224: 2 2 2 2 2 7
13:22:27 !blsq 224fc
13:22:27 | {1 2 4 7 8 14 16 28 32 56 112 224}
13:22:29 !blsq 224fC
13:22:29 | {2 2 2 2 2 7}
13:23:17 using a factorisation primitive is almost as cheating as using a factorisation program ;-)
13:23:32 now I want to write factor in underlambda
13:23:38 the problem being I haven't figured out how arithmetic works yet
13:24:00 anyone here have advice for writing pure-functional factoring algos?
13:24:01 -!- TieSleep has changed nick to TieSoul.
13:24:31 I can implement loops as folds but it's a pain, and I can write fixpoint operators to get arbitrary loops but that's even more of a pain
13:24:36 ais523: there are definitely plenty of servers with more RAM than 224 gigabytes
13:24:46 I'm not sure that's even big data (tm)
13:24:56 elliott: yes, but those are going to be specialized for the purpose of having a lot of RAM
13:25:05 I mean, it's big, but I think Big means something... a bit more than "you need TWO servers!!"?
13:25:12 it's the "indexes" bit that sold me, though
13:25:18 I guess so, fair enough.
13:25:23 like, how much disk space would you need to store the raw data for a database with 224GB of indexes
13:25:28 actually I guess it depends on how many indexes you have
13:25:44 I don't feel like database rows are usually that huge.
13:25:47 Does underlambda have integer division
13:26:27 it doesn't even have integer addition yet
13:26:31 or, well, it's easy enough to build it by hand
13:26:41 the same way as in underload
13:26:52 but it took me months just to figure out the I/O primitives
13:27:28 I think addition, subtraction, multiplication, division/modulus are going to be high-tier primitives, though
13:27:28 not sure if div and mod should be one operation or two
13:27:45 You could just split the search range recursively
13:27:53 I'm saying all this without knowing what underlambda is
13:28:33 Jafet: underlambda is a project of mine to cross-interpret as many esolangs as possible to and from
13:28:46 basically, it's a range of languages, each of which is a superset of the ones below
13:28:55 and higher tier languages compile into lower tier languages
13:29:11 the higher tier languages are intended to be reasonably pleasant to write in, if esoteric (think Funge-98)
13:29:16 the lower tier languages are intended to be easy to implement
13:29:27 Oh, the r3.8x instance now offers 244GiB.
13:29:41 !blsq 244fC
13:29:41 | {2 2 61}
13:30:48 maybe Amazon have some reason to be a little under 256GiB
13:31:57 -!- anamaria1 has left.
13:31:57 ais523: because the systems are virtualised, say?
13:32:00 -!- callforjudgement has joined.
13:32:05 ais523: because the systems are virtualised, say?
13:32:09 that leaves 12 gigabytes of RAM for the hypervisor
13:32:21 that would make sense
13:32:28 -!- ais523 has quit (Disconnected by services).
13:32:31 -!- callforjudgement has changed nick to ais523.
13:32:33 I think when you get an instance that large you get the entirety of one physical box
13:32:37 (but you still run in Xen on it)
13:32:52 I'm not sure whether the amount of RAM hypervisors needs scales with the size of the OSes they're managing
13:33:02 12 gigabytes seems like maybe a lot, but I guess it's probably needed for keeping track of all that RAM in internal structures and stuff
13:33:13 yeah
13:33:26 they probably have fancy stuff on the boxes that needs RAM, anyway
13:37:59 -!- h0rsep0wer has joined.
13:41:25 !blsq 256 20 2**.*4./
13:41:25 | 25600
13:41:33 !blsq 256 2 20**.*4./
13:41:33 | 67108864
13:43:12 !blsq %s=1024 "~ -> ~"|[%s? %s? 1024 ?/]|f~
13:43:12 | "1024 -> 1"
13:56:55 -!- h0rsep0wer has quit (Quit: Leaving).
13:57:41 -!- h0rsep0wer has joined.
14:13:03 -!- h0rsep0wer has quit (Ping timeout: 256 seconds).
14:16:47 -!- h0rsep0wer has joined.
14:38:43 -!- copumpkin has quit (Quit: My MacBook Pro has gone to sleep. ZZZzzz…).
14:44:49 -!- h0rsep0wer has quit (Ping timeout: 255 seconds).
14:58:20 -!- ProofTechnique has joined.
15:16:08 -!- nycs has joined.
15:16:11 -!- nycs has changed nick to `^_^v.
15:17:23 -!- shikhin has joined.
15:23:27 -!- oren has joined.
15:30:59 -!- adu has quit (Quit: adu).
15:37:11 So it should be possible to translate some languages to otherslosslessly
15:38:45 ??
15:38:51 programming languages or natural languages?
15:39:01 programming languages
15:39:14 yeah without interpretive overhead
15:40:09 if the set of primitive commands is the same but their syntax is different
15:40:54 but the definition of "primitive commands" can be made more general
15:42:18 But the key is, "losslessly" includes both loss of speed, and also loss of comprehensibility by humans
15:42:34 i dont think you should factor in "comprehensibility by humans"
15:43:02 Why not?
15:43:04 you loose variable names when translating to brainfuck
15:43:22 you will generally lose efificency too
15:43:25 Exactly, but not when translating to, say, C
15:43:40 well yeah
15:43:51 So that is a loss of human-readablity
15:44:09 So which languages are compatible in this manner?
15:44:39 I think, for starters, C# and VB.net are
15:44:54 ruby, python, perl , php
15:44:57 all equivalent
15:45:13 there are differences in object models there
15:45:19 that you'd have to model in a kind of ugly way
15:45:24 php loses the array-hash distinction
15:45:46 almost no two real languages are going to be exactly compatible in this way
15:45:53 but plenty of them have large, practical subsets which are
15:46:43 obviously brainfuck and ook
15:46:56 obviously?
15:47:00 and all the other "Hey I'm a new esolang but it's just brainfuck using different symbols"
15:47:15 int-e: Yes, I think that's the word for it.
15:47:22 [-]Ook! is a brainfuck program
15:47:49 it's probably also an Ook program
15:47:58 Ook [-] polyglots should be really easy
15:48:03 OOk Brainfuck
15:48:06 OOOOOKK!K!KK!K!
15:49:33 I guess you can encode the non-brainfuck operators as whitespace.
15:50:51 How human-readable.
15:50:52 Ook! doesn't have much leeway in its syntax.
15:50:54 I guess translating C to C++ is pretty loss-lessy
15:50:56 and readable
15:51:41 the other way around, though...
15:51:51 C++ written by me is translatable to C
15:51:52 (yeah, I know that's how C++ started out.)
15:52:01 https://en.wikipedia.org/wiki/Compatibility_of_C_and_C%2B%2B#Constructs_valid_in_C_but_not_in_C.2B.2B
15:52:07 in particular it's pretty much exactly impossible for C++ -> C
15:52:15 because you have to erase templates
15:52:18 because the only features of C++ I use are vector and string
15:52:24 so it's no bijective
15:52:26 C just isn't powerful enough for the translation
15:52:30 but oren didn't ask for bijections!
15:52:35