00:00:14 -!- iconmaster_ has quit (Ping timeout: 260 seconds).
00:03:39 -!- friendlyGoat has quit (Quit: Leaving).
00:03:56 -!- rodgort has quit (Quit: Leaving).
00:08:18 -!- rodgort has joined.
00:10:08 -!- wob_jonas has joined.
00:10:47 <wob_jonas> fizzie: I can't reproduce what you did. If I export to pdf from scribus, it seems it insists on recompressing the jpeg images, rather then embedding them as is.
00:11:10 <wob_jonas> fizzie: how exactly did you do that?
00:12:07 -!- wob_jonas has quit (Client Quit).
00:18:45 -!- doesthiswork has joined.
00:49:00 <fizzie> @tell wob_jonas I don't think I did anything special. But it seems to (unfortunately) have something to do with the image: I did it for two random .jpg files, and one gets embedded as-is but the other is re-encoded.
00:53:12 <fizzie> @tell wob_jonas I'm not sure what the crucial difference between the images might be. The re-encoded file has an embedded color profile while the embedded one doesn't, in case that matters.
00:59:30 <fizzie> @tell wob_jonas I don't think that's the reason, though, because removing all metadata (which does get rid of the color profile; at least Gimp no longer does the conversion prompt) doesn't help. It gets still stored re-encoded, actually as PPM unless I explicitly request JPEG (and if I do, it's re-JPEG'd).
01:06:14 <fizzie> @tell wob_jonas The non-working file is also a progressive JPEG (ImageMagick's "identify -verbose" reports "Interlace: JPEG"), while the working/embedded one is non-progressive ("Interlace: None"), in case *that* makes a difference.
01:08:11 <fizzie> @tell wob_jonas That seems to be it (at least for my pair of images): if I use 'convert' to re-encode it as a non-progressive JPEG file, that does get embedded as-is by Scribus.
01:10:08 -!- iconmaster__ has quit (Ping timeout: 276 seconds).
01:11:41 -!- user24 has quit (Remote host closed the connection).
01:15:09 <fizzie> @tell wob_jonas FWIW, some earlier PS/PDF spec versions seem to have supported baseline (non-progressive) JPEG only, though since PDF 1.3 progressive files should be supported. There might be some other limitations on JPEG images in PDF as well that are relevant in your case.
01:17:28 <fizzie> @tell wob_jonas If it *is* the same reason for you, jpegtran can do a lossless conversion from progressive to baseline (just call it on the file without the "-progressive" switch), and that was also enough to make my file get embedded as-is.
01:18:11 <fizzie> (I hope that's not too many messages.)
01:22:46 <oerjan> i recall lambdabot used to just lose them if there were too many, but vaguely think that was fixed.
02:56:31 -!- danieljabailey has joined.
03:00:33 <oerjan> apparently that does exist.
03:04:51 <ais523> are there any phonemes that technically exist but no langauge uses them?
03:05:09 <ais523> that can be made via the normal methods, that is
03:10:08 <erkin> Yes, they're documented in fact.
03:10:32 <erkin> They appear (almost exclusively) in speech impediments or cases of glossolalia.
03:10:49 <zzo38> What kind of phonemes are only possible for monster with multiple mouths?
03:11:04 <erkin> There's an IPA extension describing them. In fact there's a symbol for speaking like Donald Duck.
03:13:34 <doesthiswork> Yes I don't know of any languages that use buccal speach
03:13:38 -!- danieljabailey has quit (Ping timeout: 244 seconds).
03:14:50 <erkin> https://en.wikipedia.org/wiki/Voice_Quality_Symbols https://en.wikipedia.org/wiki/ExtIPA
03:15:30 -!- danieljabailey has joined.
03:26:26 -!- erkin has quit (Quit: Ouch! Got SIGIRL, dying...).
03:44:45 -!- aloril has quit (*.net *.split).
03:44:45 -!- Hoolootwo has quit (*.net *.split).
03:44:46 -!- Vorpal has quit (*.net *.split).
03:44:46 -!- dingbat has quit (*.net *.split).
03:44:47 -!- quintopia has quit (*.net *.split).
03:47:27 -!- danieljabailey has quit (Ping timeout: 240 seconds).
03:47:52 -!- dingbat has joined.
03:49:36 -!- danieljabailey has joined.
04:10:41 <zzo38> I also think 24 characters maximum password length is too short. The minimum should be four and the maximum should at least one hundred, preferably more. All bytes other than null bytes should be allowed in the password.
04:25:05 <zzo38> Did you know that? You can use bsdtar to read truncated ZIP archives; 7-Zip cannot read truncated ZIP archives, but bsdtar can. I learned that a week ago, so that is good in case I need to load a truncated ZIP archive in my computer.
04:46:50 -!- Melvar has quit (Quit: thunderstorm).
04:48:01 <pikhq> Oh yeah, because libarchive reads from the start, and thus has some minorly quirky behavior with ZIP files.
04:53:57 <izabera> that will come in handy the next time that you'll have to deal with a damaged tape drive
05:00:22 <pikhq> Hey, you never know.
05:02:24 -!- ais523 has quit (Quit: quit).
05:07:46 <pikhq> Good lord, is the Spaceworld '97 demo of Pokemon Gold weird in spots.
05:09:17 <pikhq> Like, there's a Ditto evolution and it's freaking disturbing.
05:09:55 <pikhq> Sorry, wrong channel.
05:10:04 <zzo38> My brother mentioned it earlier today.
05:11:44 <oerjan> in freefall, dvorak may have to learn some boundaries, quickly.
05:12:00 <pikhq> You say that, but it's Dvorak.
05:12:36 <oerjan> unless he wants to be forcefully thrown out of the house...
05:13:57 <pikhq> Which you know will happen and he will just be confused about it. :P
05:23:02 <oerjan> ok, somewhere in the foglio family there must be a pair of bickering siblings that gil and tarvek are patterned after.
05:40:32 -!- SopaXorzTaker has joined.
06:40:42 -!- imode has joined.
06:44:25 -!- xkapastel has quit (Quit: Connection closed for inactivity).
06:52:26 -!- SopaXorzTaker has quit (Remote host closed the connection).
06:53:45 -!- SopaXorzTaker has joined.
07:06:48 -!- Sgeo has joined.
07:08:54 -!- Sgeo_ has quit (Ping timeout: 265 seconds).
07:34:25 -!- arseniiv has joined.
07:47:26 -!- aloril has joined.
07:47:26 -!- Hoolootwo has joined.
07:47:26 -!- Vorpal has joined.
07:47:26 -!- quintopia has joined.
07:53:26 -!- imode has quit (Ping timeout: 260 seconds).
08:09:02 -!- wob_jonas has joined.
08:09:43 <wob_jonas> Oh! So maybe I have to change the PDF version in the export dialog.
08:10:32 <wob_jonas> I wouldn't like to embed non-progressive jpegs, because those would increase the file size.
08:11:20 <wob_jonas> Thanks for the hint, I'll experiment with this.
08:13:18 <wob_jonas> "<zzo38> What kind of phonemes are only possible for monster with multiple mouths?" => https://stickman.qntm.org/comics.php?n=178 mentions something that the protagonists can't say because they don't have a second tongue
08:19:06 <wob_jonas> "<doesthiswork> I have a new favorite consonant. Labialised m" => wtf
08:30:36 -!- Sgeo_ has joined.
08:32:00 -!- Sgeo has quit (Ping timeout: 260 seconds).
09:13:27 <fizzie> wob_jonas: I tried all the PDF versions supported by my version of Scribus (1.3, 1.4, 1.5), but couldn't make it embed a progressive JPEG. It might be a limitation in the software. I don't think it's generally that big a size difference, though. (At least my test image only goes from 87902 bytes to 88946 with "jpegtran -optimize progressive.jpg >baseline.jpg".)
09:14:37 <wob_jonas> fizzie: yes, but this is for large resolution high quality images of which I've already reduced the size
09:15:00 <wob_jonas> and the total size of images is like 260 megabytes or something, so the size difference matters
09:15:20 <wob_jonas> I'll have to look at some other software though, it's possible that pdftex can help here
09:19:09 -!- oerjan has quit (Quit: Nite).
09:19:50 <fizzie> If the layout you need is really simple, it might also be possible just to handcraft a .ps program (the jpeg files would just go in as-is as the image data, with filter /DCTDecode) and try any of the N tools that do PS-to-PDF conversion.
09:20:33 <fizzie> (In theory you could even export the postscript from somewhere and just replace the images in it.)
09:21:03 <wob_jonas> the layout is simple, except for three pages where I'm putting text, but I can make those text with a separate tool and export to ps and slice them together
09:21:12 <wob_jonas> um, four pages where I'm putting text actually
09:21:41 <wob_jonas> and ideally the table of contents page should have links to other pages, and a nice pdf table of contents, and I think pdftex can help there
09:21:50 <wob_jonas> or I can try libreoffice or ms office, maybe they can do this
09:22:42 <wob_jonas> the layout is basically just one image per page, with a variable but precise amount of white space on their four corners, that is. all pages are the same size, but I need the images placed exactly at the right spot.
09:57:12 -!- iconmaster__ has joined.
10:05:33 -!- iconmaster_ has joined.
10:09:00 -!- iconmaster__ has quit (Ping timeout: 245 seconds).
10:24:47 -!- mniip has quit (Ping timeout: 600 seconds).
10:29:42 -!- mniip has joined.
10:57:41 -!- doesthiswork has quit (Quit: Leaving.).
11:13:16 <esowiki> [[Special:Log/newusers]] create * MostlyHarmless * New user account
11:20:06 <esowiki> [[Esolang:Introduce yourself]] M https://esolangs.org/w/index.php?diff=55490&oldid=55483 * MostlyHarmless * (+194) /* Introductions */
11:27:45 -!- iconmaster_ has quit (Ping timeout: 245 seconds).
11:33:16 -!- tromp has quit (Remote host closed the connection).
11:42:08 -!- tromp has joined.
12:15:48 -!- Melvar has joined.
12:48:23 -!- doesthiswork has joined.
12:59:56 <wob_jonas> but... this worked just a few hours ago. and I changed nothing, except for most of the code.
13:38:29 -!- xkapastel has joined.
13:53:51 -!- MDead has joined.
13:55:11 -!- MDude has quit (Ping timeout: 268 seconds).
13:55:16 -!- MDead has changed nick to MDude.
14:00:15 -!- wob_jonas has quit (Quit: http://www.kiwiirc.com/ - A hand crafted IRC client).
15:17:48 -!- variable has quit (Quit: Found 1 in /dev/zero).
15:31:13 -!- zzo38 has quit (Ping timeout: 248 seconds).
16:45:02 -!- arseniiv_ has joined.
16:46:24 -!- arseniiv has quit (Read error: Connection reset by peer).
16:46:34 -!- arseniiv has joined.
16:50:47 -!- arseniiv_ has quit (Ping timeout: 256 seconds).
17:36:10 -!- zzo38 has joined.
17:51:25 -!- plokmijnuhby has joined.
18:07:48 -!- plokmijnuhby has quit (Quit: Page closed).
18:25:59 -!- imode has joined.
18:45:49 -!- Phantom_Hoover has joined.
18:45:49 -!- Phantom_Hoover has quit (Changing host).
18:45:49 -!- Phantom_Hoover has joined.
19:07:49 -!- wob_jonas has joined.
19:08:37 -!- MDead has joined.
19:09:27 <wob_jonas> I'd like to ask about a certain computational model that has both real-world and esoteric uses. I don't have any specific questions except what its name is, but I'd like any general information.
19:10:35 -!- tromp has quit (Remote host closed the connection).
19:10:42 -!- MDude has quit (Read error: Connection reset by peer).
19:10:52 -!- MDead has changed nick to MDude.
19:12:04 <zzo38> I do not know how to answer if you do not give specific questions or at least to mention what kind of computational model it is.
19:13:07 <wob_jonas> The basic idea is that you have a computer with both a cpu+ram and a disk (which could be magnetic or solid state), but the computer could lose power at any moment unexpectedly, after which all the data in the ram and cpu is lost, and the program counter and other important registers are reset to some clean state from which you can boot easily.
19:14:35 <wob_jonas> The disk is non-volatile, but the sector you were writing at the moment when the power got lost can be corrupted, such that it will read with arbitrary contents and you can't directly tell that it is corrupt. (This is sort of a worst case model, a malicious corruption.)
19:15:52 <wob_jonas> In the simple case, you have just one machine like this, and you write software for this. There is a well-known way how you can do reliable operations on this, with cow and a one-bit transaction boolean choosing between two roots, but there are nontrivial costs so optimizing for this model can be interesting.
19:16:57 <wob_jonas> In more practical but harder to define real world cases, you have multiple computers each of which can lose power independently, and you want them to communicate with each other and solve tasks reliably, which is the difficult model that banks have to solve on their backend mainframes.
19:20:29 <wob_jonas> I was also thinking of a related computing model, in which you only have a RAM and cpu, the cpu has very few registers, the RAM is written one word or one byte at a time, but at least writing a word is atomic, and the program counter could be reset to the entry point of the program at any time between any two memory writes.
19:22:06 <wob_jonas> Obviously in all cases you need a little bit of progress guarantee, namely that at least sometimes the program gets ran for some nontrivial time, like for a day in the first case or 100000 instructions in the second case, but you still can't tell in advance when that will happen, because sometimes the program will get reset very early.
19:24:52 -!- atslash has joined.
19:33:25 -!- atslash has quit (Ping timeout: 256 seconds).
19:38:37 <int-e> wob_jonas: "the cpu has very few registers" - do you intend those to be preserved when the CPU resets? (otherwise I don't quite see how this is relevant)
19:39:24 <wob_jonas> int-e: the program counter definitely have to be reset. for the rest of the registers, it doesn't matter, you can just imagine they're memory-mapped registers and that's why they're preserved.
19:39:42 <int-e> you mean you can't just do some big processing in the registers and commit with a few writes
19:41:23 <int-e> Anyway, this is easy to program for in principle (make a virtual machine with idempotent operations; update the virtual program counter last when emulating an instruction), and probably hard to optimize.
19:41:30 <wob_jonas> or you can also have a cpu that has no registers apart from the program counter, except for internal ones used within an instruction, and the single memory write happens at the end of the instruction, and the internal registers are forgotten at that point anyway, in which case you can't observe whether they're preserved at the reset or not.
19:41:58 <wob_jonas> int-e: yes, exactly, easy to program in principle but annoying in practice
19:44:46 <int-e> Well I would've expected the whole CPU state to be reset. (There are hot-pluggable CPUs, right? Though I suppose that in reality you'll want to shut down the chip in an orderly fashion, power down the socket, then change the CPU and then boot the new one up. But I don't know.)
19:45:18 <wob_jonas> int-e: you can choose the whole cpu to be reset too, sure
19:45:44 <wob_jonas> I'm thinking more of a software emulation of this rather than an actual cpu
19:47:28 -!- SopaXorzTaker has quit (Remote host closed the connection).
19:47:57 -!- arseniiv has quit (Ping timeout: 240 seconds).
19:56:14 -!- tromp has joined.
21:02:30 -!- iconmaster_ has joined.
21:25:54 -!- doesthiswork has quit (Ping timeout: 260 seconds).
22:09:09 -!- choochter has quit (Read error: Connection reset by peer).
22:09:35 -!- choochter has joined.
22:12:22 -!- boily has joined.
22:26:19 <fungot> boily: ( ( go fishing yeah)) the one that's at home now i've got grandkids of about a month
22:26:44 <HackEso> Nostril is a common Québécois greeting.
22:26:45 <boily> fungot: no fishing, I'm staying inside with AC.
22:26:45 <fungot> boily: ( ( it's back at))
22:27:09 <HackEso> 11569:2018-05-29 <oerjän> learn Nostril is a common Qu\xc3\xa9b\xc3\xa9cois greeting.
22:28:40 <lambdabot> CYUL 012200Z 28008KT 15SM FEW025TCU FEW035 SCT150 OVC160 28/21 A2968 RMK TCU1CU1AC2AS5 TCU TR SLP053 DENSITY ALT 2000FT
22:57:05 -!- wob_jonas has quit (Quit: http://www.kiwiirc.com/ - A hand crafted IRC client).
23:30:53 -!- tromp has quit (Remote host closed the connection).
23:31:54 -!- variable has joined.
23:37:35 <esowiki> [[Micro]] https://esolangs.org/w/index.php?diff=55491&oldid=52156 * Raddish0 * (+199) I added a link to the completely new specification for Micro Version 2
23:40:28 <esowiki> [[User:Raddish0]] M https://esolangs.org/w/index.php?diff=55492&oldid=52093 * Raddish0 * (+14) I am now 14 (way back in December, but I'm lazy)