←2017-10-12 2017-10-13 2017-10-14→ ↑2017 ↑all
00:24:04 -!- augur has quit (Remote host closed the connection).
00:30:12 <boily> shachaf: helloochaf. ysac has corrupted me; I whangjangle my food.
00:32:02 -!- augur has joined.
00:33:46 <imode> boily: do you suck at cooking.
00:34:58 -!- augur has quit (Remote host closed the connection).
00:35:34 -!- tromp has quit (Remote host closed the connection).
00:36:08 -!- augur has joined.
00:37:24 <boily> imode: according to a panel of reliable sources, apparently not.
00:37:32 <boily> (even if my dashi is weak.)
00:44:20 <imode> boily: then you cannot whangjangle. you aren't qualified.
00:53:03 -!- sebbu2 has changed nick to sebbu.
00:57:30 * imode wonders that, if multiset rewriting corresponds to working with register machines, what does regular set rewriting correspond to?
01:01:27 <shachaf> Machines with 1-bit registers?
01:02:18 <imode> that's kind of what I thought. loooooots of registers. :P
01:02:59 <imode> wonder if that's still TC, though. I imagine you could form rule 110 with that with some.. effort.
01:04:21 <imode> or build the semantics of a multiset rewriting system on top of that.
01:04:52 <imode> i.e treat a span of bits as one binary number and use that to represent one register. that'd be hellish.
01:05:16 <shachaf> A finite number of finite registers still isn't going to be enough on its own.
01:05:42 -!- Soni has quit (Ping timeout: 260 seconds).
01:05:43 <imode> yeah.
01:06:02 <imode> but then you just keep expanding the number of possible symbols.
01:06:09 <imode> so you end up with a really large alphabet.
01:24:39 -!- danieljabailey has changed nick to Danieljabailey.
01:24:54 -!- Danieljabailey has changed nick to danieljabailey.
01:29:51 -!- doesthiswork has joined.
01:38:44 <quintopia> helloily
01:42:19 -!- MrBismuth has joined.
01:42:26 -!- MrBusiness has quit (Ping timeout: 255 seconds).
01:42:35 <boily> QUINTHELLOPIA!
01:43:22 -!- oerjan has joined.
01:45:51 -!- boily has quit (Quit: STUMBLING CHICKEN).
01:46:43 <oerjan> byeli
01:53:07 * oerjan wonders how official boily's panel was
02:55:41 <zzo38> Why is ImageMagick not reading XPM format 1 and 2?
03:02:05 -!- ATMunn has quit (Quit: Read error: Connection reset by peer (just kidding)).
03:09:08 -!- xkapastel has quit (Quit: Connection closed for inactivity).
03:45:49 -!- imode has quit (Ping timeout: 255 seconds).
04:13:00 -!- tromp has joined.
04:17:46 -!- tromp has quit (Ping timeout: 255 seconds).
04:17:55 -!- erkin has quit (Quit: Ouch! Got SIGABRT, dying...).
05:10:00 -!- doesthiswork has quit (Quit: Leaving.).
05:39:55 -!- doesthiswork has joined.
05:51:07 -!- imode has joined.
05:53:48 <imode> I wonder why FRACTRAN hasn't gotten as crazy a following as something like brainfuck.
05:54:15 <oerjan> you have to understand prime numbers to use it hth
05:54:26 * oerjan grins evilly
05:54:47 <imode> not.. really. you just have to understand that each prime is a unique symbol and choose your alphabet mapping to each prime carefully
05:55:02 <imode> that's actually the really simple part. the odd part is dealing with pattern matching on registers and shifting values around.
05:55:14 <imode> it's not hard to design a compiler that will do that kind of thing for you.
05:56:24 <imode> I mean you can take any fractran program, look at each fraction, extract the prime factorization of it, convert that into a suitable symbol set (or a set of registers), and build a macro language on top of the resulting (somewhat more legible) rewrite rules.
05:56:48 <imode> I would've expected that a GoL implementation would have surfaced eventually. kind of surprised that I haven't found one yet.
06:03:13 <oerjan> obligatory accompaniment to today's freefall comic http://blogs.sciencemag.org/pipeline/archives/2008/02/26/sand_wont_save_you_this_time
06:05:18 <doesthiswork> that is a extremely famous article
06:06:16 <imode> "For dealing with this situation, I have always recommended a good pair of running shoes." good wisdom right there.
06:30:12 * imode wonders if fractran's rules can be unordered.
06:30:58 <oerjan> imode: oh you didn't see my comment the other day ... they cannot, although i don't remember exactly how the proof went.
06:31:39 <imode> oerjan: interesting. I wonder what modifications would be required to make them unordered. probably something akin to a P system.
06:31:54 <imode> where the resulting rewrites determine which classes of rules get applied.
06:33:31 <oerjan> well you need a way to say that a rule won't be applied if it's divisable by something.
06:34:08 <imode> so rule dependencies.
06:34:11 -!- erkin has joined.
06:34:25 <imode> or rather, inverse dependencies.. I guess.
06:34:45 <imode> "do this if these primes are present, but if these are, don't."
06:35:00 <oerjan> in some sense, fractran's linear order is the _simplest_ option for that.
06:35:17 <imode> yeah, because you can prioritize rules.
06:35:38 <imode> https://en.wikipedia.org/wiki/P_system#Example_computation this is also an option.
06:51:21 <imode> parallel fractran. hmm.
06:52:12 -!- sleffy has quit (Ping timeout: 260 seconds).
06:54:03 -!- doesthiswork has quit (Quit: Leaving.).
06:59:53 -!- doesthiswork has joined.
07:00:29 -!- FreeFull has quit.
07:02:02 <shachaf> One thing I like about FRACTRAN is that it doesn't seem like a programming language at all.
07:02:15 <shachaf> If you make the rules slightly more contrived it loses that.
07:02:23 <oerjan> fizzie: hm i suspect the wiki bridge has croaked
07:02:33 <imode> well, it's pretty much just unordered string rewriting.
07:02:37 <shachaf> `? wewlcome.it
07:02:39 <shachaf> uh
07:02:42 <shachaf> `? welcome.it
07:02:45 <HackEgo> wewlcome.it? ¯\(°​_o)/¯
07:02:45 <HackEgo> welcome.it? ¯\(°​_o)/¯
07:02:53 <shachaf> Taneb!!
07:03:08 <oerjan> are you sure Taneb knows italian
07:03:35 <oerjan> isn't it a bit too sexy language
07:04:05 <shachaf> https://twitter.com/Ngevd/status/918012736721883137 hth
07:04:45 <shachaf> Did Taneb invent modal logic?
07:05:11 <oerjan> possibly
07:06:17 <oerjan> hm Taneb seems to have aged. shocking.
07:16:21 -!- erkin has quit (Quit: Ouch! Got SIGABRT, dying...).
07:28:36 -!- doesthiswork has quit (Quit: Leaving.).
07:41:45 -!- augur has quit (Quit: Leaving...).
09:05:56 -!- oerjan has quit (Quit: Nite).
09:27:37 -!- tromp has joined.
09:29:28 -!- imode has quit (Ping timeout: 258 seconds).
09:39:26 -!- AnotherTest has joined.
09:39:30 -!- augur has joined.
09:44:01 -!- tromp has quit (Remote host closed the connection).
09:50:22 -!- sebbu2 has joined.
09:53:28 -!- sebbu has quit (Ping timeout: 240 seconds).
10:07:45 -!- augur has quit (Remote host closed the connection).
10:08:21 -!- augur has joined.
10:12:35 -!- augur has quit (Ping timeout: 246 seconds).
10:46:40 -!- LKoen has joined.
10:59:15 -!- tromp has joined.
11:18:11 -!- tromp has quit (Remote host closed the connection).
11:34:47 -!- boily has joined.
12:00:04 -!- joast has quit (Remote host closed the connection).
12:03:11 -!- LKoen has quit (Remote host closed the connection).
12:03:27 -!- AnotherTest has quit (Ping timeout: 240 seconds).
12:08:13 <boily> `5 w
12:08:19 <HackEgo> 1/1:kallisti//kallisti is a former prophet swearing off his pastry deity. \ @//@ is an OS made out of only the finest vapour. \ cia//CIA sees it all. \ lens//A lens is just a store comonad coalgebra. \ fourth wisdom//.wisdoms other the all upon builds wisdom fourth The
12:24:14 -!- boily has quit (Quit: DOOR CHICKEN).
12:26:26 <fizzie> @tell oerjan It might've been because I had to restart HackEgo. Did the same for the bridge as well, maybe it works now.
12:26:26 <lambdabot> Consider it noted.
12:40:50 -!- joast has joined.
13:08:38 -!- augur has joined.
13:12:35 -!- augur has quit (Ping timeout: 240 seconds).
13:13:27 -!- jaboja has joined.
13:26:17 -!- HackEgo has quit (Remote host closed the connection).
13:26:40 -!- HackEgo has joined.
13:49:35 -!- jaboja has quit (Ping timeout: 248 seconds).
13:56:28 <b_jonas> fungot, in Greek mythology, is there such a thing as artifact weapon or armor forged by mortals, or do all of them have to be forged by a deity, usually Hēphaistos or Pallas Athēnē?
13:56:29 <fungot> b_jonas: mr president, on a point of order, and i regret not having been used to ban tv advertising and indirect advertising. it has not yet put forward any reasoned objections within the 30-day period; secondly, certainty as to the future, within the meaning defined in the treaty.
13:59:10 <int-e> the longwinded europarl style makes it abundantly clear that fungot cannot form any coherent thought.
13:59:11 <fungot> int-e: i do not know for how long it will take shape exactly is still to be done to help us determine the real responsibilities of each of these points is new to us on the issue of alcohol, but also for all those who have flouted the rules in relation to the completion of the internal market.
13:59:39 <int-e> (though arguably that is also true for many politicians)
14:00:44 -!- doesthiswork has joined.
14:03:19 -!- tromp has joined.
14:20:01 -!- ATMunn has joined.
14:31:58 -!- sebbu has joined.
14:32:57 -!- sebbu2 has quit (Ping timeout: 240 seconds).
14:40:13 -!- `^_^v has joined.
14:43:22 -!- tromp has quit (Remote host closed the connection).
14:46:14 -!- hkgit03 has joined.
14:47:26 <hkgit03> a
14:55:34 <int-e> b
15:04:24 <hkgit03> The next logical letter would be v of course.
15:05:16 <int-e> abvgd?
15:05:38 <hkgit03> http://oeis.org/A279619
15:06:21 -!- doesthiswork has quit (Quit: Leaving.).
15:08:23 <int-e> https://www.youtube.com/watch?v=jWDlaXrJOI4 hmmmm
15:08:48 <hkgit03> Ah ^^
15:08:51 <hkgit03> logical.
15:08:54 <int-e> strange stuff, but indeed it does include the russian alphabet.
15:09:35 <int-e> `unidecode абвгд
15:09:36 <HackEgo> ​[U+0430 CYRILLIC SMALL LETTER A] [U+0431 CYRILLIC SMALL LETTER BE] [U+0432 CYRILLIC SMALL LETTER VE] [U+0433 CYRILLIC SMALL LETTER GHE] [U+0434 CYRILLIC SMALL LETTER DE]
15:10:04 <b_jonas> int-e: use the full alphabet. abvgdeëžzijklmnoprstufhcčšŝʺyʹèûâ
15:10:58 <hkgit03> That is not the russian alphabet, is it?
15:11:42 <b_jonas> hkgit03: it is, in ISO 9 transcription
15:11:47 <int-e> I can see the resemblence
15:12:59 <b_jonas> or maybe you want the serbian alphabet, a b v g d đ e ž z i j k l lj m n nj o p r s t ć u f h c č dž š
15:13:07 <b_jonas> or one of the many other alphabets that start with a b v g d
15:13:08 <hkgit03> Uhuh okay
15:13:33 <hkgit03> Of course. That was exactly what I was going for.
15:15:48 <int-e> abgde is a logical order for an alphabet
15:16:18 <int-e> the abcde order should have a different name ;-)
15:17:33 <b_jonas> int-e: the order is not really logical, more like traditional, inherited from the Phoenicians, and now various variants of it are used by tons of alphabets, sometimes multiple different orders for the same language
15:18:52 <b_jonas> `? ameraval
15:18:53 <HackEgo> ameraval? ¯\(°​_o)/¯
15:22:58 <b_jonas> the battery of my mobile phone not only drains faster but also gets warm if I listen to music all day
15:23:29 <b_jonas> still not too hot, just warm
15:25:53 -!- hkgit03 has quit (Quit: http://www.kiwiirc.com/ - A hand crafted IRC client).
15:44:14 -!- tromp has joined.
15:50:02 -!- tromp has quit (Ping timeout: 260 seconds).
16:04:13 -!- LKoen has joined.
16:08:46 -!- LKoen has quit (Ping timeout: 264 seconds).
16:08:53 -!- sleffy has joined.
16:24:42 -!- AnotherTest has joined.
16:55:45 -!- imode has joined.
17:03:30 -!- FreeFull has joined.
17:18:58 -!- `^_^v has quit (Quit: This computer has gone to sleep).
17:26:26 -!- tromp has joined.
17:41:21 -!- AnotherTest has quit (Ping timeout: 240 seconds).
17:48:41 -!- AnotherTest has joined.
18:05:31 -!- `^_^v has joined.
18:14:45 -!- tromp has quit (Remote host closed the connection).
18:15:04 -!- tromp has joined.
18:21:57 <zzo38> I do not have the file "/usr/lib/X11/rgb.txt" in my computer. Is there another directory it may be under instead?
18:23:15 <b_jonas> zzo38: try /usr/share/X11/rgb.txt
18:23:42 <b_jonas> zzo38: these days unix distros put everything in usr that is shared among all machines of all architectures in /usr/share
18:23:59 <b_jonas> and architecture-specific stuff only in /usr/lib and /usr/libexec
18:24:17 <zzo38> OK
18:25:03 -!- tromp has quit (Remote host closed the connection).
18:25:16 <b_jonas> I think it started back when this was used to spare disk space by mounting /usr/share and /usr through NFS, but mounting the same /usr/share on multiple machines, or something, but these days more and more stuff is split to multiple directories according to architectures, mostly so that you can have x86_32 and x86_64 libraries ran on the same machine or something,
18:25:27 <b_jonas> so the share split turns out to be very useful.
18:26:06 <b_jonas> Note that /usr has read-only data that can be shared among multiple machines of the same architecture in first place
18:26:24 <b_jonas> data specific to one machine goes to /var or /etc
18:28:00 <b_jonas> debian often patches install directories from upstream packages to conform to these directory structures
18:28:20 <b_jonas> although most software these days already use the right directories anyway
18:29:02 <b_jonas> and most allow overrides via make variables or configure options or something, because GNU encouraged that in the guidelines they sneakily put into the gnu make reference docs
18:29:44 <b_jonas> but of course in most free software it's easy to modify paths in the source
18:30:36 <zzo38> The program I am writing doesn't actually need to find the file anyways, since if it is wanted the user will specify the filename explicitly anyways, but it is helpful to know where it is installed.
18:31:18 -!- xkapastel has joined.
18:33:07 <zzo38> (I am writing a program to read XPM pictures. I already wrote one to write XPM pictures.)
18:33:28 <zzo38> (In both cases, all three formats are implemented.)
18:34:54 <zzo38> ImageMagick supports only format 3, while Netpbm supports only format 1 and 3, it seems, and neither support monochrome mode.
18:37:42 <b_jonas> do people even still use xbm and xpm pictures for anything? I thought those were formats for icons on machines with little memory and low bitdepth monitors.
18:38:11 <zzo38> XPM2 seems a reasonable format for using with a text editor, at least.
18:41:12 <zzo38> They can also be used including in C programs (except XPM2).
18:44:09 -!- imode has quit (Ping timeout: 258 seconds).
18:57:44 <HackEgo> [wiki] [[Arrows]] M https://esolangs.org/w/index.php?diff=53202&oldid=52782 * HereToAnnoy * (+31) wip
18:58:15 <fizzie> Looks like it's working again.
19:18:17 <pikhq> Once upon a time they were very useful as a lowest-common-denominator file format.
19:18:27 <pikhq> I'm not sure who uses it much right now, though.
19:21:09 <\oren\> pikhq: yeah the standard "output an image ad-hoc" format is PBM
19:21:13 <b_jonas> still no word of a newIOCCC
19:21:50 <pikhq> PBM is a bit of an irritating format, though.
19:21:54 <zzo38> pikhq: As I mentioned, for including in C codes and to be suitable for editing with a text editor.
19:22:02 <pikhq> It *looks* simpler than it is.
19:22:16 <pikhq> zzo38: Well, yes, I know why you'd want to use it. :) I just don't know if anyone does.
19:22:19 <zzo38> For "output an image ad-hoc" I prefer farbfeld.
19:22:56 <pikhq> Though for outputting only, PBM is similarly simple.
19:23:14 <zzo38> (Although I wrote dvipbm before farbfeld was invented, and anyways the printer driver needs PBM.)
19:23:14 <pikhq> The file format just has some surprising complexity when reading it.
19:23:19 <pikhq> (comments, for instance)
19:23:38 <zzo38> pikhq: Yes, that is true, for outputting only PBM is simple enough (and supports multi-page, which is needed for using with printing).
19:24:48 <b_jonas> I find PNM a bit limiting, so I prefer to write raw images, and send the metadata separately, such as in the command line of ffmpeg or imagemagick
19:25:04 <b_jonas> and the same for reading it
19:26:00 <zzo38> b_jonas: Why won't you use farbfeld instead then? (ImageMagick does not implement farbfeld, but it ought to! Write a module for it if you want it)
19:26:06 <b_jonas> PNM only supports one endianness and it's the wrong one, only supports grayscale or RGB, not other channel configurations, etc
19:26:54 <zzo38> Nevertheless I intend these Farbfeld Utilities program will including conversion to/from any format, whether common (such as PNG) or obscure (such as XZIP Picture Library format).
19:28:05 <pikhq> b_jonas: PAM fixes the channel configuration bit.
19:28:23 <b_jonas> zzo38: besides the bit depth (which PNM can hvae multiple), I want to use the exact channel configuration (eg. possibly bgr, bgr0, bgra, rgb, argb, alpha only, yuv444) and interlacing (channels of a pixel together, channels for whole image one after another) that
19:28:29 <pikhq> As for the endianness... Bi-endian file formats can fuck off.
19:28:45 <b_jonas> are the most convenient for the particular use. a program like ffmpeg or imagemagick can use whatever configuration I want, and handle the conversions itself.
19:28:54 <b_jonas> why bother with the conversion myself when they can do it already?
19:29:06 <zzo38> Yes, OK, you can do that if you wish, too.
19:29:39 <b_jonas> oh, and also padding at the end of lines.
19:30:44 <b_jonas> also, ffmpeg and imagemagick (imagemagick in any way, as in, command line, high-level C api, C++ api) can do some simple filtering steps that I often want to do right before reading or writing the image, such as cropping or resampling to smaller before write, rotating or flipping before read.
19:31:16 <zzo38> (I could also provide a program in Farbfeld Utilities that will support raw non-indexed pictures; I already have a file for raw indexed pictures. Although there is also ff-swizzle. And in the case just of "bgr" or "rgb" simple 8-bits-per-channel, together, you could probably use ff-scanf to do this too (although I have not tried).)
19:31:48 <pikhq> It amuses me that, if you don't mind having a somewhat silly and large output, you can reasonably do uncompressed PNG output pretty easily.
19:31:55 <b_jonas> I actually have a simple C++ wrapper that spawns an ffmpeg process to read or write an image or video, and allows a variety of these options,
19:32:00 <pikhq> (it's more complex than Farbfeld or PNM, naturally, but it's not that hard)
19:32:12 <b_jonas> but only those options that I have needed so far, because ffmpeg has an infinity of useful options, so the simple wrapper won't ever cover it all.
19:32:26 <zzo38> Yes, I suppose you can do that if you like to do (although then, I think, you should need the checksum?)
19:32:37 <pikhq> zzo38: Yes, you do need the checksums.
19:32:42 <pikhq> (there's two)
19:32:52 <pikhq> It's CRC-32 and Adler32 both.
19:33:04 <pikhq> They're both pretty easy to write, though.
19:33:14 <pikhq> I will freely grant this is not the most useful thing in the world.
19:33:25 <pikhq> But it's pretty quick, and easier than using libpng.
19:34:51 <zzo38> Farbfeld Utilities "ffpng" program can do that though, with a command such as "ffpng b0 c2 e-1 f0" or such.
19:35:12 <pikhq> *nod*
19:35:14 <b_jonas> pikhq: crc-32? what the heck?
19:35:42 <zzo38> ("b0" disables compression, "c2" specifies RGB, "e-1" forces it to use a single block, and "f0" disables filtering.)
19:36:03 <pikhq> b_jonas: It's a standard checksum, and PNG requires it for each chunk.
19:36:13 <b_jonas> I didn't know png had that
19:36:40 <pikhq> Adler32 comes into it because the contents of the image chunk(s) are just zlib streams.
19:37:58 <b_jonas> yes, I know that part
19:38:07 <b_jonas> I just didn't know there was an extra checksum
19:38:14 <b_jonas> do reader programs actually check that?
19:38:19 <pikhq> Yes.
19:38:43 <pikhq> Also, every *single* chunk in PNG is CRC32'd.
19:38:45 <b_jonas> I mean, a browser can start displaying the image before having downloaded the whole thing, and usually most of the image is one chunk
19:38:58 <b_jonas> one big chunk
19:38:58 <pikhq> (there's more chunks than just image data)
19:39:02 <b_jonas> pikhq: sure
19:39:11 <pikhq> Yeah, but they barf if the checksum turns out to have been wrong, generally.
19:39:37 <pikhq> And there's reader programs that aren't browsers.
19:40:00 <b_jonas> sure
19:40:46 <b_jonas> mind you, I mostly just see truncated jpegs on the web, I don't remember having seen a truncated png
19:50:40 <zzo38> A few of my programs can work with multi-frame pictures, and generally expect the frames to be arranged in a vertical strip. (Although also a program is provided to convert this if needed.) For formats such as Famicom pattern tables, the reading is automatically like a vertical strip, and can write using a vertical strip as input too even if the height is unknown, because it is the order of the data already.
19:51:24 <zzo38> (Some other programs use a horizontal strip instead, although I think vertical strip format is more a more natural format.)
19:53:17 <zzo38> One feature of XPM that others usually don't have is customizable colours. Also, it can include a monochrome as well as a colour version in the same picture.
19:54:34 <zzo38> (As far as I know, other formats do not do this, except that the YZIP Picture Library format (not the XZIP format) allows palettes to include optional stipple patterns for use with monochrome display.)
19:55:51 <b_jonas> zzo38: for videos, the important constraints in my program is that (a) the video read and write is streaming, so if I write a long video, then only data for a few dozen frames is held in memory at a time,
19:56:07 <b_jonas> and (b) I don't spawn a program for each video frame, because that's slow.
19:56:22 <zzo38> b_jonas: Yes, that is correct, it make sense.
19:56:43 <b_jonas> by the way, my code does two more conveniences, although both are because ffmpeg can already easily do that:
19:57:03 <b_jonas> firstly, I can read or write multiple numbered image files with the same shape instead of a video,
19:57:09 <zzo38> (And the draft for the TAVERN picture format allows a picture to have any number of associated palettes, so you could have a different version of the picture (or just a different palette) for composite output, RGB output, monochrome, low resolution vs high resolution, etc.)
19:57:42 <b_jonas> and I can display image or video in a window instead of writing it to a file (useful because sometimes I want one, sometimes the other in the same program, for debugging)
19:57:56 <zzo38> OK
19:58:45 <b_jonas> the C++ wrapper is needed because the ffmpeg interface is a bit... strange, and there's a lot of stupid workarounds needed, so the wrapper does the ones I've encountered so far
19:59:09 <zzo38> OK
19:59:12 <b_jonas> but I got the ffmpeg devs to fix a bug or two I encountered in this, which is nice
20:06:03 <zzo38> You can tell me of other formats that you or someone else might want in Farbfeld Utilities, and/or to possibly contribute the codes for such formats. Currently it does not have many, and some can read only, and some can write only. (Some subformats also are currently not implemented, such as, cannot read bmp with RLE, and cannot use JPEG with 12-bits or arithmetic coding.
20:09:30 -!- jaboja has joined.
20:09:50 <b_jonas> `? farbfeld
20:09:51 <HackEgo> farbfeld? ¯\(°​_o)/¯
20:10:03 <b_jonas> do we have a wisdom about this?
20:10:36 <zzo38> Add one if you know what to write about it
20:13:00 <int-e> I rather suspect that farbfeld is not vaporware.
20:13:26 <HackEgo> [wiki] [[TwoFiftyFive]] N https://esolangs.org/w/index.php?oldid=53203 * HereToAnnoy * (+1584) Created page with "TwoFiftyFive is an [[esoteric programming language]] by [[User:HereToAnnoy]] with only two instructions and limited memory. ==Syntax== ===Instructions=== * Move instruction (<..."
20:14:36 <pikhq> It isn't, it's just niche.
20:14:38 <zzo38> One thing I can think of, is if there is some way to make the JPEG decoder so that (by user option) it can try to recover some of the quality of the picture if it was encoded with low quality, to make it less "blocky" output.
20:14:41 <HackEgo> [wiki] [[TwoFiftyFive]] M https://esolangs.org/w/index.php?diff=53204&oldid=53203 * HereToAnnoy * (+0) minor formatting edit
20:14:45 <b_jonas> `slashlearn farbfeld//zzo38's Farbfeld utilities is a package of command-line programs (with rather strange command-line syntax) for manipulating images in a specific raw format called farbfeld. It's somewhat underdocumented. http://zzo38computer.org/fossil/farbfeld.ui/home
20:14:48 <HackEgo> Learned 'farbfeld': zzo38's Farbfeld utilities is a package of command-line programs (with rather strange command-line syntax) for manipulating images in a specific raw format called farbfeld. It's somewhat underdocumented. http://zzo38computer.org/fossil/farbfeld.ui/home
20:15:25 <HackEgo> [wiki] [[TwoFiftyFive]] M https://esolangs.org/w/index.php?diff=53205&oldid=53204 * HereToAnnoy * (+0) another minor formatting edit
20:15:59 <zzo38> Yes, those things is true, that is OK.
20:16:19 <HackEgo> [wiki] [[Focus]] M https://esolangs.org/w/index.php?diff=53206&oldid=53200 * HereToAnnoy * (+0) minor formatting edit
20:17:42 <b_jonas> zzo38: how do you use ffpng to write 8-bit per channel depth images instead of 16-bit channel deep?
20:17:53 <HackEgo> [wiki] [[User:HereToAnnoy]] M https://esolangs.org/w/index.php?diff=53207&oldid=53196 * HereToAnnoy * (+67) added TwoFiftyFive to contributions list
20:18:52 <zzo38> b_jonas: It will automatically do that if it can be done losslessly. (You can also explicitly specify by the use of the "c" option.)
20:19:19 <b_jonas> zzo38: also, nice, you've developped farbfeld.ui and its documentation a lot since I've last seen it
20:19:51 <b_jonas> zzo38: sure, but I don't know how to reduce an image to 8 bit per channle depth easily first
20:20:02 <b_jonas> ah, the c option
20:20:39 <b_jonas> c sets the color space and then the bit depth. nice
20:20:50 <zzo38> To reduce the image first, use "ff-poster 8", if you need to do that.
20:21:00 <b_jonas> that reminds me
20:21:27 <b_jonas> since farbfeld format doesn't use premultiplied alpha, is there a program to turn all fully transparent pixels to black?
20:21:33 <b_jonas> to transparent black I mean
20:21:53 <b_jonas> so as to save space when writing in an image format that also doesn't use premultiplied alpha, such as png
20:22:08 <zzo38> No, although you are right, it can be a good idea.
20:22:35 <b_jonas> this could be an option in ffpng too of course
20:23:10 <zzo38> I will just make it a separate program I think
20:24:02 <b_jonas> zzo38: ok, but note that you may have to do this after reducing bit depth for a true color (non-indexed) image
20:24:14 <b_jonas> so that's two extra steps for a 8-bit per channel deep png
20:24:36 <b_jonas> which is why it would be useful in ffpng
20:24:53 <b_jonas> if you make it a separate program, then offer a transparency value cutoff parameter too
20:25:10 <zzo38> OK, I will add a cutoff parameter (zero by default).
20:25:24 <b_jonas> thanks
20:26:11 <b_jonas> and if you don't add an option to ffpng, then at least write about the correct pipeline for 8-bit per channel rgb with alpha in the docs of ffpng
20:26:44 <zzo38> Yes, OK I will do that (the wiki is editable by everyone though, so you can also write about it if you want to do so)
20:30:34 <zzo38> OK, I added "ff-transopt" program which does that.
20:31:56 <zzo38> OK, and now I added that note to the documentation of ffpng, too.
20:37:13 <b_jonas> nice
20:38:36 -!- tromp has joined.
20:49:11 -!- tromp has quit (Remote host closed the connection).
20:51:59 -!- jaboja has quit (Ping timeout: 248 seconds).
21:11:58 <\oren\> lol I decreased the memory usage of a program by a factor of ten thousand
21:12:27 <\oren\> previously used 49 gigabytes, now uses 5.4 kilobytes
21:12:35 <\oren\> er, megabytes
21:12:54 <\oren\> stupid ps outputting
21:13:37 -!- tswett has joined.
21:13:41 <tswett> Yo.
21:13:57 -!- augur has joined.
21:14:31 <tswett> I just realized that it's possible to write the traditional call-with-current-continuation function in "fake Lua".
21:14:45 <zzo38> tswett: OK, how is that?
21:15:00 <tswett> That is, it's almost valid Lua code, but it does something illegal.
21:15:06 <tswett> Lemme see if I can write it out here.
21:16:26 <tswett> function call_cc(callback, arg) local result; function curr_cont(r) result = r; goto finish end; result = callback(curr_cont, arg); ::finish:: return result end
21:16:56 <tswett> I guess the "arg" bit is unnecessary, but kinda convenient.
21:17:58 <tswett> It's not legal Lua because you can't jump from an inner function into a label defined in the outer function.
21:19:00 <zzo38> Yes, and actually I think once I have suggested to add "goto" into JavaScript and to allow it to work like that in order to make up such thing, by making a kind of "evasive exceptions".
21:24:16 <tswett> Yeah, I guess Lua has two forms of "stack-subverting" control flow.
21:24:26 <tswett> Exceptions and coroutines.
21:31:21 <zzo38> Do you know what is the syntax for HSV colour codes in XPM file? The document just says that it is not implemented.
21:32:57 <tswett> No clue.
21:39:33 -!- brandonson has joined.
21:40:14 -!- brandonson has quit (Client Quit).
21:40:22 -!- brandonson has joined.
21:50:01 -!- tromp has joined.
21:54:35 -!- tromp has quit (Ping timeout: 240 seconds).
22:29:10 -!- AnotherTest has quit (Ping timeout: 264 seconds).
22:29:57 -!- `^_^v has quit (Quit: This computer has gone to sleep).
22:33:30 -!- jaboja has joined.
22:34:14 -!- shikhin has changed nick to shikhin_shikhin_.
22:37:05 -!- shikhin_shikhin_ has changed nick to shikhin__nihkihs.
22:40:25 -!- shikhin__nihkihs has changed nick to shikhin.
23:02:55 -!- augur has quit (Remote host closed the connection).
23:04:01 -!- wob_jonas has joined.
23:04:11 -!- boily has joined.
23:04:49 <wob_jonas> I'm reading https://esolangs.org/wiki/List_of_ideas and lol on "four loop"
23:10:48 -!- augur has joined.
23:11:49 -!- augur has quit (Remote host closed the connection).
23:18:22 -!- augur has joined.
23:21:26 -!- tromp has joined.
23:26:12 -!- tromp has quit (Ping timeout: 260 seconds).
23:26:57 <wob_jonas> What is the name of the theoerm that it's possible to write comments in any sane programming language that's expressive enough?
23:31:12 -!- tromp has joined.
23:44:19 <Guest95245> What if it's a language that only accepts non-alphabetic characters
23:49:21 <boily> the language, or the comment language?
23:51:38 -!- oerjan has joined.
23:53:28 <oerjan> @messages-fold
23:53:28 <lambdabot> fizzie said 11h 27m 2s ago: It might've been because I had to restart HackEgo. Did the same for the bridge as well, maybe it works now.
23:53:53 <shachaf> is that the command that reduces your messages
23:58:26 <oerjan> sheesh it's been a long time since i looked at my stackoverflow tab, there's a favorite edit from jan '16 i hadn't seen
23:59:13 <boily> hellørjan, helloochaf.
23:59:52 <wob_jonas> Are there programming languages where a team of programmers have to enter the program quickly on a guitar hero controller?
←2017-10-12 2017-10-13 2017-10-14→ ↑2017 ↑all