00:18:18 <zzo38> CEA-708 has text tags to specify the purpose of the caption (dialog, sound effects, song lyrics, etc). There is even invisible, which I think is good, since sometimes you might want invisible captions, such as if the text is already part of the picture (e.g. the clues in Jeopardy!). However, tag 12 and 13 and 14 is undefined, according to Wikipedia.
00:19:02 <zzo38> I can make the suggestion of the use of these tags: video translation, described video (invisible by default, like tag 15), and miscellaneous (in case none of the other tags are suitable).
00:23:50 -!- oerjan has quit (Quit: Nite).
00:30:02 -!- imode has quit (Ping timeout: 240 seconds).
00:35:19 <esowiki> [[Brainfuck constants]] M https://esolangs.org/w/index.php?diff=66802&oldid=58043 * A * (+38) /* 97 */
01:11:31 <pikhq> Whoa, more Homestuck
01:11:52 * pikhq should read herself the epilogue
01:40:33 <zzo38> I saw the note by wib_jonas about "vanishing three card blind" M:tG game. I think that perhaps there should be a restriction to Vintage or pseudo-Vintage cards, rather than everything they ever released.
01:41:19 <zzo38> (And if you do play it often enough that there will be only draws, then you will either have to wait for new cards, or start to use some unofficial cards too.)
01:46:06 <zzo38> And I disagree to delete stuff from the quotation file a lot; keep them until you are wasting too much disk space.
01:46:21 <zzo38> Since they are not pictures, they don't normally waste too much disk space.
02:46:19 -!- imode has joined.
05:08:28 <shachaf> lifthrasiir: 13 0s? That's a lot.
05:08:39 <shachaf> Imagine all this hashing going toward breaking SafeHaskell!
05:09:17 <shachaf> I guess you need thousands times more hashing.
05:10:33 <lifthrasiir> AFAIK the previous winner (with 12 zeroes) had spent 16 hours on GTX 1060
05:11:25 <lifthrasiir> unless the current winner used FPGA or similar (unlikely), the current winner should have spent about a week or two
05:13:47 <lifthrasiir> I have submitted that link about a day ago, but it says "3 hours ago" in the HN front page
06:03:49 <b_jonas> zzo38: it's probably too late now, with possibly several people having submitted decks using Blacker Lotus
06:11:07 -!- b_jonas has quit (Remote host closed the connection).
06:35:56 -!- imode has quit (Ping timeout: 265 seconds).
07:05:42 -!- atslash has quit (Quit: This computer has gone to sleep).
07:11:09 -!- Phantom_Hoover has joined.
07:16:03 -!- Phantom_Hoover has quit (Ping timeout: 265 seconds).
07:29:59 -!- atslash has joined.
08:59:43 -!- atslash has quit (Quit: Leaving).
09:03:50 -!- atslash has joined.
09:07:49 -!- sebbu has quit (Quit: reboot).
09:20:16 -!- atslash has quit (Ping timeout: 240 seconds).
09:20:40 -!- atslash has joined.
09:31:04 -!- kspalaiologos has joined.
09:31:23 <kspalaiologos> has someone already worked on esoteric operating system draft?
09:32:42 <kspalaiologos> so I could in my very own kernel implement a module loader, a few drivers and a brainfuck interpreter
09:37:16 -!- LKoen has joined.
09:40:14 -!- LKoen has quit (Remote host closed the connection).
09:44:33 -!- atslash has quit (Read error: Connection reset by peer).
09:45:28 -!- arseniiv has joined.
10:10:43 -!- LKoen has joined.
10:37:53 -!- sebbu has joined.
10:52:51 -!- b_jonas has joined.
10:54:54 <arseniiv> I did it! I defined cross-ratio in normal human-readable terms involving exterior product only
10:56:12 <arseniiv> though it looks like black magic; I don’t know how to derive it in a straightforward manner, it was constrained guessing
10:58:45 <arseniiv> the final result is this: where “metric-dependent” (though independent) definition goes (A, B; C, D) = |AC| |BD| / |BC| |AD|, the invariant one goes (a ∧ c) (b ∧ d) / (b ∧ c) (a ∧ d), and I’m not joking it’s really that similar. As I said, black magic
11:00:28 <arseniiv> where a, b, c, d are nonzero vectors from lines A, B, C, D
11:04:50 <arseniiv> I specifically was looking for an expression invariant under scaling these a, b, c, d separately, and one that is invariant under any linear maps. This would mean it’s invariant in projective space, and this also enforced fractions and the form where each vector occurs equally often in denominators and numerators. So the simplest expression to satisfy that is the one I got here, upto a permutation of (a, b, c, d)
11:08:25 <arseniiv> oh, by (a ∧ c) (b ∧ d) / (b ∧ c) (a ∧ d) I meant (a ∧ c / b ∧ c) (b ∧ d / a ∧ d) (or another rewriting), division is allowed because bivectors are all from one-dimensional space here, as all vectors line in the same plane (as all projective points need to be collinear for cross-ratio to be defined)
11:26:58 <arseniiv> eh, I wrote about this in three places and still no reaction :D though I have no patience. I hadn’t seen the expression anywhere to date, so I’m agitated at least because that shouldn’t be the case, it’s simple and it’s definitely useful when teaching. Or learning, for that matter
11:30:22 <arseniiv> though contemporarily it’s not a consensus yet that exterior products make many things more transparent. Exterior algebra captures linear dependence, in a way
11:30:41 <b_jonas> arseniiv: may I recommend you the book https://esolangs.org/logs/2019-08.html#lN5c ?
11:55:54 -!- atslash has joined.
12:08:47 <arseniiv> oh there is a nice remark also that cross-ratio values naturally lie in RP! I should fix my definition so that be the case
12:10:34 <arseniiv> and also that the distances in the classical definition are directed. Yes, this makes more sense, and yes it’s nice that the invariant definition was taking that into account all along
12:27:10 <arseniiv> hm constructing a definition such that cross-ratio lies in RP isn’t trivial!
12:27:40 <arseniiv> we should divide bivectors or we get nonsense, but we should also take away all divisions
12:28:04 <arseniiv> I think there is an invariant way though
12:37:30 -!- xkapastel has joined.
13:07:08 -!- LKoen has quit (Remote host closed the connection).
13:20:31 -!- LKoen has joined.
13:20:35 <esowiki> [[Dilemma]] N https://esolangs.org/w/index.php?oldid=66803 * A * (+635) Created page with "== Drafts == Maze language based on a searching algorithm that does not remember the way it goes. <pre> X@. ^|^ S/. </pre> The direction priority is clockwise; i.e. it tries t..."
13:29:28 <esowiki> [[Dilemma]] https://esolangs.org/w/index.php?diff=66804&oldid=66803 * A * (+650)
13:39:09 <arseniiv> hm xy/zw = x′y′/z′w′ for x, y, z, w ∈ (one-dimensional vector space) can be defined as x ⊗ y ⊗ z′ ⊗ w′ = x′ ⊗ y′ ⊗ z ⊗ w. That doesn’t solve everything yet, we need to define what does it mean for a quadruple (x, y, z, w) to equal 1
13:40:24 <arseniiv> but we still need to show these quadruples upto that equality are iso with RP
13:40:31 -!- LKoen has quit (Remote host closed the connection).
13:44:33 -!- LKoen has joined.
13:51:29 <arseniiv> using ⊗ we even don’t need to use quadruples, pairs (x ⊗ y, z ⊗ w) will suffice
13:54:53 -!- FreeFull has quit.
13:55:34 <arseniiv> hm it seems to map those to RP we need to explicitly demand [(x, x)] ↦ 1, [(x, −x)] ↦ −1, [(0, x)] ↦ 0, [(x, 0)] ↦ ∞, or else we could pick a wrong projective transformation
14:03:52 <esowiki> [[Dilemma]] https://esolangs.org/w/index.php?diff=66805&oldid=66804 * A * (+235)
14:05:59 -!- LKoen has quit (Remote host closed the connection).
14:10:04 -!- LKoen has joined.
14:18:20 -!- Phantom_Hoover has joined.
14:25:20 <arseniiv> seems fairly improvable? => hm nope, we could just restate it but I don’t see a way clearer restatement
14:39:38 -!- Phantom_Hoover has quit (Ping timeout: 240 seconds).
14:55:18 -!- atslash has quit (Quit: This computer has gone to sleep).
14:57:11 -!- xkapastel has quit (Quit: Connection closed for inactivity).
16:58:21 -!- xkapastel has joined.
17:44:44 <kspalaiologos> is there some kind of penality for writing android apps in Flash or C++ Builder?
17:45:10 <kspalaiologos> I wish, has someone taken on programming in Brainfuck for mobile
17:46:02 <kspalaiologos> ^^^ like, is there a chance a desperate Java programmer will hunt me down?
18:26:57 -!- imode has joined.
18:31:17 -!- imode has quit (Ping timeout: 240 seconds).
18:36:11 -!- imode has joined.
18:41:20 -!- LKoen has quit (Remote host closed the connection).
19:28:42 <fizzie> The "penalty" is that all the platform APIs, including the standard UI framework, are really accessible only from the "Java" (I mean, it's not really all that Java) runtime.
19:28:56 -!- joast has quit (Ping timeout: 240 seconds).
19:32:12 <zzo38> Doesn't Android use Dalvik?
19:32:44 <fizzie> It used to, then they switched to ART.
19:33:07 <fizzie> Both use the Dex bytecode format, they're just different implementations.
19:33:37 -!- imode has quit (Ping timeout: 265 seconds).
19:34:44 -!- imode has joined.
19:43:57 <zzo38> What is the difference?
19:45:12 <zzo38> I read the documentation for Dalvik and do not know how I/O works.
19:46:08 <fizzie> AIUI, the biggest difference is that ART does an AOT compilation step (around install time?) to native ELF executables.
19:48:11 <zzo38> Is the file format the same or is there some difference?
19:49:34 <fizzie> The .dex bytecode file format that applications contain is the same, the internal files it generates (.odex and maybe some others) are different.
19:55:31 -!- Cale has quit (Ping timeout: 250 seconds).
19:55:33 <fizzie> Apparently there's a few minor differences, like ART does a more strict verification of the bytecode. And I don't know about I/O, never looked too closely into the actual bytecode. Maybe it's all done through the usual invoke-* instructions, just some methods are magical. (Just a guess.)
20:08:32 -!- Cale has joined.
20:15:12 -!- imode has quit (Ping timeout: 268 seconds).
20:17:05 -!- joast has joined.
20:22:54 <b_jonas> fizzie: aren't there already ready-made wrappers for C++ though?
20:28:05 <kspalaiologos> and Starling/Feathers UI for flash looking awesome
20:35:35 -!- oerjan has joined.
20:38:11 <fizzie> I'm not sure which ready-made wrappers you mean. I mean, the standard NDK has a bunch of APIs, sure. Not for the native UI, but for OpenGL/EGL/Vulkan and all kinds of hardware. And of course there's a truckload of different cross-platform development things that just draw their own stuff (Flutter, React Native, Xamarin, that Feathers thing...).
20:47:32 <b_jonas> fizzie: I mean for the native android services
20:48:15 <zzo38> Can you run a X server on Android?
20:53:38 <fizzie> I don't see why you couldn't in theory, but I haven't really heard of anyone doing it.
20:54:02 <kmc> i think people do
20:54:13 <kmc> because they want to use GUI programs in a GNU/Linux chroot
20:54:20 <kmc> you can use Xvnc or something
20:55:02 <fizzie> I use Termux on my phone, which is basically an Android terminal emulator app, plus a Linux userland installed inside the app's data directory, and an APT repository for convenient packages.
20:55:10 <kmc> oxample: https://www.reddit.com/r/amateurradio/comments/84jbgy/wsjtx_on_android/
20:55:13 <kmc> fizzie: that sounds dope
20:55:17 <kmc> I should use that
20:55:21 <kmc> is it less buggy than juicessh
20:55:25 <kmc> can you run the linux mosh client in it
20:56:06 <fizzie> So far I like it better than JuiceSSH, which is where I switched from. At least you can run the standard OpenSSH client and ssh-agent.
20:56:16 <fizzie> I assume mosh should be fine as well, I've just never set that up.
20:56:19 <kmc> you can probably run mosh too
20:56:26 <kmc> it doesn't have much in the way of system requirements
20:56:47 <fizzie> (The nice bit with Termux is that it requires no root.)
20:58:07 <fizzie> Looks like there's also some support for running X11, through the VNC trick: https://wiki.termux.com/wiki/Graphical_Environment
21:06:24 -!- LKoen has joined.
21:07:00 -!- kspalaiologos has quit (Quit: Leaving).
21:07:39 <myname> i didn't like termux that much for some reason i fail to remember
21:07:53 <myname> maybe because of the softkeys or the colors or something like that
21:08:07 <myname> but the url handling of juicessh is the worst i've ever seen
21:08:14 <myname> even connectbot did it better
21:12:18 <fizzie> The default colors had that thing where the "dark blue" is very light.
21:13:19 <fizzie> But that was just a matter of copying my normal set of colors into ~/.termux/colors.properties on the phone.
21:15:56 <fizzie> The softkeys are also configurable, I set them to a multi-row layout pretty similar to what I had in JuiceSSH (esc, /, -, home, up, end, pageup on top row; tab, ctrl, alt, left, down, right, pagedown on bottom).
21:17:02 <fizzie> One thing that seems a bit wonky is that when you swipe the softkeys to switch to the "input a line of text using the normal Android text input" mode, it also seems to always treat that as a keypress of whatever key you start the swipe from.
21:17:50 <fizzie> Also the URL handling isn't that great either, there's a long-press "select URL" option but it doesn't consider # a URL character or some-such.
21:18:44 <fizzie> I like the volume key shortcuts. (Holding volume down acts like holding control down, and holding volume up turns wasd into arrows.)
21:18:52 -!- imode has joined.
21:23:43 <zzo38> Can you configure the URL characters and schemes?
21:26:09 <fizzie> Not via a configuration file, as far as I can tell. You could always just edit the source and build the app, though, it's open source.
21:26:25 <fizzie> The relevant logic would seem to be here: https://github.com/termux/termux-app/blob/master/app/src/main/java/com/termux/app/TermuxActivity.java#L676
21:31:51 -!- imode has quit (Ping timeout: 264 seconds).
21:34:37 <zzo38> We can see that some schemes are missing, such as "news" and "nntp" and "ssh" and "magnet". Also, the "news" and "magnet" schemes do not use // after the colon. And, yes it does looks like it doesn't include # as a URL character, but it ought to do; even that URL linking to the source code has it, and in many other HTTP(S) URLs it will be common, so it should be included.
21:38:05 -!- xkapastel has quit (Quit: Connection closed for inactivity).
22:25:51 -!- arseniiv has quit (Ping timeout: 264 seconds).
23:03:40 -!- LKoen has quit (Quit: “It’s only logical. First you learn to talk, then you learn to think. Too bad it’s not the other way round.”).
23:09:30 -!- oerjan has quit (Quit: Nite).
23:21:47 -!- imode has joined.