00:01:45 <esolangs> [[AEWNN]] N https://esolangs.org/w/index.php?oldid=83820 * ResU * (+1335) Created page with "'''AEWNN''' (or '''An esolang with no name''') is an esolang created by [[User:ResU]] in 2021. ==Variables== There are two types of variables: letter variables and VariablesWi..."
00:06:17 <esolangs> [[Talk:Malbolge]] https://esolangs.org/w/index.php?diff=83821&oldid=41030 * Monochromeninja * (+1824) /* Python interpreter */ new section
00:14:25 <esolangs> [[Talk:Emoji-gramming]] https://esolangs.org/w/index.php?diff=83822&oldid=57083 * Monochromeninja * (+445) /* Sign checking */ new section
00:18:33 <esolangs> [[AEWNN]] M https://esolangs.org/w/index.php?diff=83823&oldid=83820 * ResU * (+6)
00:22:04 -!- spirgel has joined.
00:23:53 <zzo38> A reason needing writing better web browser software is that existing software is like http://catb.org/jargon/html/U/user-friendly.html (It is one thing, not all of them, of course)
00:32:18 <esolangs> [[User:ResU]] N https://esolangs.org/w/index.php?oldid=83824 * ResU * (+68) Created page with "My esolangs: [[AEWNN]] (planned to make more esolangs in the future)"
00:34:36 -!- BlueFalconHD has joined.
00:37:20 <esolangs> [[User:ResU]] https://esolangs.org/w/index.php?diff=83825&oldid=83824 * ResU * (+12)
00:51:49 <fizzie> I agree, those should be unordered-list bullet points instead.
01:18:14 -!- spirgel has quit (Remote host closed the connection).
01:26:40 -!- BlueFalconHD has quit (Ping timeout: 250 seconds).
01:28:52 <nakilon> some trivial things you want to make for years but are lazy for no reason
01:34:21 <nakilon> $ ruby upload.rb empty_file
01:34:21 <nakilon> uploading as http://md5.nakilon.pro/d41d8cd98f00b204e9800998ecf8427e
01:34:21 <nakilon> $ ruby upload.rb empty_file
01:34:21 <nakilon> file exists as http://md5.nakilon.pro/d41d8cd98f00b204e9800998ecf8427e
01:35:04 <nakilon> finally I made a bucket to store arbitrary files under their md5
01:35:15 -!- spirgel has joined.
01:35:31 <nakilon> sure it loses the content-type but why would you download something without knowing what it is?..
01:40:02 <keegan> why would you use md5 for anything new in the year 2021?
01:41:40 <keegan> because it's extremely broken as far as cryptographic properties go, and if you don't need those properties there are better (mainly, faster) hash functions you can use
01:42:40 <keegan> it's easy to generate pairs of files with the same md5 sum
01:42:54 <keegan> this might be only a nuisance in your use case and not a security concern, but it's still easy to avoid
01:43:01 <nakilon> it's a public file storage -- what is the scenario of generating pairs of files?
01:43:38 <keegan> there is just no use case for which md5 is the best choice
01:44:09 <keegan> use sha256 and you won't need to worry about collisions (at least for the time being)
01:44:38 <nakilon> how many files you should upload to get a collision?
01:44:46 <keegan> only two if it's deliberate
01:44:56 <keegan> i think a generation of programmers was taught that "hash function" is synonymous with "md5" and they will never upgrade to something not broken
01:45:17 <keegan> there's no reason to use the broken thing when non-broken things exist that are better in every way
01:45:19 <zzo38> I agree that you shouldn't use MD5 hashes to identify files; there are better hashes such as SHA-3
01:45:25 <keegan> just having md5 code in your codebase is a risk
01:45:35 <keegan> "oh we aren't using these for cryptographic purposes, it's fine"
01:45:49 <nakilon> what's the reason to select custom hash function for a bucket that I will upload a thousand of files in my whole life to?
01:45:59 <keegan> in what way is SHA256 more "custom" than MD5
01:46:11 <keegan> they are both standards available in every modern hash function library
01:46:21 <nakilon> in what way md5 is more custom than sha256?
01:46:25 <keegan> you just got it in your brain that MD5 is "the" hash function
01:46:33 <keegan> and reveals you to be severely behind the times
01:46:39 <nakilon> you are angry on your own fantasies
01:46:40 <zzo38> Fossil uses MD5, but not to identify files. Rather, each deck ends with a Z card which has the MD5 hash of all preceding cards, and then the resulting deck is identified by the SHA-1 or SHA3-256 hash of the entire deck including the Z card.
01:46:54 <nakilon> fantasies that someone is stupid enough to not know about hash functions
01:47:09 <nakilon> angry without understanding the use case
01:47:18 <keegan> i've made my case and you choose not to understand it
01:47:22 <keegan> i'm going to eat pizza now so ttyl
01:47:42 <nakilon> you are just saying random trivia that is not applicable
01:48:18 <nakilon> wasting attention on inexisting problem
01:48:58 <zzo38> Some protocols require MD5, such as the HTTP auth
01:51:20 <nakilon> I must be using the HTTP auth because I'm from a "generation of programmers was taught that "hash function" is synonymous with "md5""
01:51:41 <nakilon> and probably my empty bucket will be used for a Mars mission
01:52:06 <zzo38> Even if you are, that doesn't mean that MD5 should also be used for other purposes too
01:52:11 <fizzie> I agree with the "there's no valid use case for MD5 except where it's required for interoperability" case. Libraries will eventually stop offering it as a choice; that's a good enough reason to not to pick it even when the choice would be otherwise arbitrary.
01:52:53 <nakilon> but where did I say that I should use md5 everywhere? it's just a way I index files in my bucket that I don't even immediately use
01:54:33 <keegan> but why'd you choose it
01:54:38 <keegan> when it's known to be broken
01:54:44 <keegan> and you could choose something not known to be broken, with zero downside?
01:54:54 <nakilon> you won't be even able to make me a problem because you don't have an Upload access to the bucket
01:55:13 <nakilon> it's broken only in your fantasies
01:55:31 <nakilon> in fact I'll never have an issue with this bucket and script
01:55:59 <zzo38> MD5 hashes are also shorter than SHA-1 hashes or SHA3-256 hashes
01:56:06 <keegan> that may be, but there is still no reason to choose the broken thing
01:56:28 <nakilon> there is no more sense in chosing another hash function for my bucket than in using Pi as 3.14 and a million of digits after .
01:57:15 <nakilon> zzo38 and it might be a reason to use it because it's an utl to a file -- it's good when it's short
01:57:32 <esolangs> [[Typeform]] N https://esolangs.org/w/index.php?oldid=83826 * S1(210) * (+273) creating page with language idea
01:57:46 <zzo38> Yes, but you could shorten it by using base64 instead of hex, or something like that
01:59:09 <keegan> using md5 is shoddy engineering that reveals you don't really care about doing things right, even if it doesn't cause any immediate practical issue in this particular application
01:59:17 <keegan> if you disagree that's fine
01:59:24 <keegan> there's no reason for me to keep stating this
02:00:41 <pikhq> i mean if being short is all that matters why not crc32?
02:00:54 <keegan> i didn't mean to shit on your accomplishment but i guess we both got confrontational about it
02:01:09 <zzo38> CRC-32 is probably way too short for that, collisions are too likely even if accidentally
02:01:10 <keegan> well, 32 bits is short enough to worry about even accidental collisions
02:01:19 <pikhq> sorry, context matters
02:01:24 <nakilon> zzo38 the "/" char would create a folders in bucket tree for no reason; also I like how hex looks
02:02:26 -!- spirgel has quit (Ping timeout: 245 seconds).
02:02:29 <fizzie> I'm surprised base62 is used so little. I guess it's because you can usually find at least two more safe characters, like the URL-safe variant's _- pair.
02:02:30 <zzo38> nakilon: You can use a variant of base64 with different characters in use
02:03:32 <nakilon> zzo38 adding some custom library dependency to my script...
02:04:20 <keegan> it's a lot more annoying to implement when the base is not a power of 2
02:04:22 <nakilon> maybe my stdlib even has base64, still it's overcomplicating things for no real purpose
02:06:40 <nakilon> and yes, right before starting making the script I googled the length of crc32 and md5
02:07:21 <nakilon> and make a choice between them
02:10:44 <fizzie> Hmm, maybe HackEso's `paste should have used a hash function for the file name for deduplication purposes... but then it'd have to rename the file after the fact, because it needs to (really *only*) work for non-seekable inputs.
02:10:51 <fizzie> I guess it's probably not worth it. There are a total of four pastes that have been pasted twice, and a single paste (unsurprisingly, the empty file) that has been pasted 7 times.
02:11:06 <fizzie> I think people have been cleaning up that directory every now and then though.
02:14:37 <zzo38> Can any newer hashing algorithms which have a secondary hash starting at a different offset and then add that hash at the end of the data when computing the primary hash? (e.g. similar to having H(X||H("0"||X)) although you can do other things such as having different parameters for the secondary hash, or adding the length differently, etc)
02:14:39 <nakilon> this reminds me how people love to play in an echo chamber about "regexes can't parse HTML/XML"
02:15:20 <nakilon> because it's easier and funnier to be in an echo chamber than learn that some regex engines support recursion and parse HTML/XLS with no issues
02:15:46 -!- spirgel has joined.
02:16:41 <nakilon> living in imaginary world where they've learned some "rules" of how things work and can't be assed to learn more
02:20:43 -!- spirgel has quit (Ping timeout: 272 seconds).
02:25:39 -!- moon has changed nick to cd.
02:29:20 <pikhq> I'd argue said regexes are not regexes
02:29:27 <pikhq> because regex refers to... regular expressions
02:29:41 <pikhq> instead they're regex-flavored programming languages because pain :)
02:32:11 -!- Noisytoot has quit (Quit: ZNC 1.8.2 - https://znc.in).
02:32:56 -!- Noisytoot has joined.
02:37:15 -!- spirgel has joined.
02:41:51 -!- spirgel has quit (Ping timeout: 268 seconds).
02:44:32 <keegan> it just depends on how you define the word "regex"
02:45:57 <keegan> there's also the question of whether you *should* try to parse HTML/XML with such grotesquely-extended regex engines even though you *can*
02:46:09 <keegan> but this is #esolangs so I guess the answer to that one is affirmative :)
02:46:42 <keegan> regardless, attempting to do anything tree-shaped with a regex is usually a sign that you're doing it wrong and should consider a different approach
02:47:16 <fizzie> I don't think I've ever come across a purported XML-munching regular expression that'd deal with <![CDATA[ ... ]]>, even though it's certainly possible.
03:02:06 <salpynx> computational linguists are so pedantic
03:03:19 -!- spirgel has joined.
03:03:38 -!- cd has changed nick to moon.
03:05:18 <salpynx> thinking they can proscribe what regular expressions are - they should be studying how real groups of people _use_ regexes and simple describe
03:08:29 <salpynx> (that's (supposed) to be a joke, just in case anyone strongly disagrees, or worse, agrees)
03:08:38 <keegan> I wasn't totally sure ;)
03:08:52 <keegan> I think descriptive computer languages theory is a good idea though!
03:09:26 <keegan> it's definitely something people do
03:09:38 <keegan> take popular ad-hoc features and try to formalize them
03:10:07 <lambdabot> *** "proscribe" wn "WordNet (r) 3.0 (2006)"
03:10:11 <lambdabot> v 1: command against; "I forbid you to call me late at night";
03:10:13 <lambdabot> "Mother vetoed the trip to the chocolate store"; "Dad nixed
03:10:15 <lambdabot> our plans" [syn: {forbid}, {prohibit}, {interdict},
03:10:17 <lambdabot> {proscribe}, {veto}, {disallow}, {nix}] [ant: {allow},
03:11:34 <salpynx> ugh, I used the wrong word: prescribe I meant.
03:12:37 <int-e> it's "inflammable" all over again
03:13:03 <nakilon> can lambdabot adjust line width?
03:13:42 <salpynx> It feels like there could be a joke in there. I wish I'd done it deliberately. You can't proscribe what word I use to mean x ... or something
03:21:53 -!- spirgel has quit (Ping timeout: 264 seconds).
03:22:36 <nakilon> imagine that at the time you were learning programming people naming 10 random languages were not naming those that they would name today http://www.softpanorama.org/Bulletin/Humor/how_programmers_hunt_elephants.shtml
03:24:17 <nakilon> hm half of them sound like dbms though
03:25:28 -!- Lord_of_Life_ has joined.
03:25:36 -!- Lord_of_Life has quit (Ping timeout: 244 seconds).
03:26:33 -!- Lord_of_Life_ has changed nick to Lord_of_Life.
03:27:33 <nakilon> two other lists can be found on this page http://www.softpanorama.org/Lang/programming_languages_humor.shtml
03:31:38 <esolangs> [[Special:Log/newusers]] create * Goglesq * New user account
03:34:31 -!- spirgel has joined.
03:37:31 <nakilon> half of text on this page is written before my birth I guess
03:39:12 -!- spirgel has quit (Ping timeout: 272 seconds).
03:45:39 <esolangs> [[Esolang:Introduce yourself]] https://esolangs.org/w/index.php?diff=83827&oldid=83815 * Goglesq * (+247)
03:49:56 <esolangs> [[Language list]] https://esolangs.org/w/index.php?diff=83828&oldid=83765 * Goglesq * (+14)
03:52:12 -!- arseniiv has joined.
03:53:11 -!- spirgel has joined.
03:58:12 -!- spirgel has quit (Ping timeout: 272 seconds).
04:12:05 <esolangs> [[NScript]] N https://esolangs.org/w/index.php?oldid=83829 * Goglesq * (+1955) Created page with "NScript is a basic esoteric language designed to be at least somewhat unique. NScript has a couple nicknames by the creator: NS, NoahScript NS has plenty of flexibility and ha..."
04:17:14 -!- spirgel has joined.
04:21:36 -!- spirgel has quit (Ping timeout: 245 seconds).
04:34:08 -!- spirgel has joined.
04:35:04 <esolangs> [[NScript]] https://esolangs.org/w/index.php?diff=83830&oldid=83829 * Goglesq * (+2115)
04:37:07 <esolangs> [[NScript]] https://esolangs.org/w/index.php?diff=83831&oldid=83830 * Goglesq * (+0)
04:38:44 -!- spirgel has quit (Ping timeout: 272 seconds).
04:50:13 -!- spirgel has joined.
04:55:12 -!- spirgel has quit (Ping timeout: 272 seconds).
05:08:35 -!- spirgel has joined.
05:12:51 -!- spirgel has quit (Ping timeout: 245 seconds).
05:36:59 -!- spirgel has joined.
05:41:29 -!- spirgel has quit (Ping timeout: 244 seconds).
05:43:56 -!- tromp has joined.
05:51:02 <nakilon> is it German? can anyone read it? https://www.davidrumsey.com/luna/servlet/detail/RUMSEY~8~1~301174~90071678:Statistical-Table--Evangelisch---Lu?sort=pub_list_no_initialsort%2Cpub_list_no_initialsort%2Cpub_date%2Cpub_date&qvq=q:pub_title%3D%22Atlas%20der%20Evangelisch%20-%20Lutherischen%20Gemeinen%20in%20Russland.%20St.%20Petersburg.%20Buchdruckere
05:51:02 <nakilon> i%20der%20Kaiserlichen%20Akademie%20der%20Wissenschaften.%201855.%22;sort:pub_list_no_initialsort%2Cpub_list_no_initialsort%2Cpub_date%2Cpub_date;lc:RUMSEY~8~1&mi=12&trs=13
05:51:50 <nakilon> shorter link: https://www.davidrumsey.com/luna/servlet/detail/RUMSEY~8~1~301174~90071678:Statistical-Table--Evangelisch---Lu
05:54:12 -!- tromp has quit (Quit: My iMac has gone to sleep. ZZZzzz…).
05:56:15 <salpynx> I thought I recognised 'soul' in Seelenzabl -- which seems to be "soul count"
05:57:07 <int-e> 4394 souls per priest
05:58:39 <nakilon> that's stats appendix pages to maps, it's the last page of stats and others are pages per guberniya
05:59:15 <nakilon> on this page I wonder what the biggest titles say
05:59:35 <nakilon> "souls per priest" -- interesting
06:00:24 <salpynx> oh, 'zahl', as in zahlen ℤ
06:00:30 <int-e> the final thing is "summary of all consistorial districts in russia"
06:02:44 <int-e> The total number of souls in lutheran settlements in Russia for the time from 1853 to 1854, amounts to
06:04:20 -!- spirgel has joined.
06:06:10 -!- sprock has quit (Ping timeout: 245 seconds).
06:06:17 <salpynx> Евангелическо-лютеранская церковь в России, Украине, в Казахстане и Средней Азии is a thing it seems
06:08:14 <salpynx> reminds me -- does Google home exist in Russia? I was yelling at my Google home in frustration trying to listen to specific Russian soundtrack music that it played to me first and I wanted to hear again
06:08:17 -!- sprock has joined.
06:08:44 -!- spirgel has quit (Ping timeout: 252 seconds).
06:09:07 <nakilon> there should be a town Melitopol somewhere on the right page here, I can't find it: https://www.davidrumsey.com/luna/servlet/detail/RUMSEY~8~1~301173~90071677:Statistical-Table--Evangelisch---Lu
06:09:38 <salpynx> after ages of tweaking settings I got it to correctly recognise me saying 'поход' and it was spelled correctly on screen, then it's spin a bit and convert it to some silly English word and play me random stuff based on that
06:09:39 <nakilon> the right side is Tavricheskaya Gubernya that is this region https://ru.wikipedia.org/wiki/%D0%9C%D0%B5%D0%BB%D0%B8%D1%82%D0%BE%D0%BF%D0%BE%D0%BB%D1%8C%D1%81%D0%BA%D0%B8%D0%B9_%D1%83%D0%B5%D0%B7%D0%B4
06:11:10 <nakilon> salpynx not sure, but Russian analogue is from Yandex, called Yandex Alisa or something -- it's a name of the voice driven assistant
06:11:17 <int-e> hmm trying to match that with https://en.wikipedia.org/wiki/Courland
06:11:37 <nakilon> salpynx https://yandex.ru/alice/smart-home
06:12:42 <nakilon> int-e not sure what you mean, that's 1500km away
06:13:26 <nakilon> but if you mean the administrative division then the analogue would be https://en.wikipedia.org/wiki/Courland_Governorate
06:14:23 <nakilon> so yeah, this is the proper link to the right page : https://en.wikipedia.org/wiki/Taurida_Governorate -- you see there is "Melitopolsky Uyezd (Melitopol – Мелитополь)" in wikipedia article
06:17:50 <nakilon> it's here in the middle https://www.davidrumsey.com/luna/servlet/detail/RUMSEY~8~1~301169~90071673:Die-Kolonien-in-den-Gouvernements-J
06:20:29 -!- spirgel has joined.
06:21:44 <salpynx> interesting, I have one of those JBL speakers that work with the Yandex assisant. The vague non-committal Google messages make some sense now if the Russian market is already covered. It's like they'd started trying to support it years ago but not bothered getting over the line, unlike other languages(/markets)
06:21:48 <int-e> so it's outside of the red regions, so it's not part of the statistics
06:23:00 <nakilon> int-e that's weird, it's the center of Melitopolsky Uyezd
06:23:28 <nakilon> yeah those green and red regions don't make sense to me
06:23:49 <int-e> red = protestant/lutheran, yellow = catholic; your statistics were for lutheran areas.
06:24:56 -!- spirgel has quit (Ping timeout: 245 seconds).
06:25:11 <nakilon> oh you mean the map author only provided stats for regions that were relevant to him?
06:25:11 <int-e> your statistics are lutheran souls, so... yeah.. Melitopol would not be included
06:25:29 <nakilon> I thought those are just population charts and have to include all the biggest towns
06:26:04 <int-e> Nah I think it's just people associated with the church, not the whole population.
06:26:58 <int-e> Take the number for Moscow: 8251 people in 1853...
06:27:18 <int-e> Surely that was a fairly small part of the population.
06:28:25 <nakilon> salpynx from some points of view it's harder for Google to compete in Russia because Yandex was the first Web Search engine, a year before Google, and all the technologies then were built in parallel but Google has no clue in Russian language so the indexing, ranking and stuff is hard for them
06:28:59 <nakilon> or "the first morphological search", I don't remember the details
06:29:27 <nakilon> companies are technologically nearly equal, Google is just bigger
06:29:54 <nakilon> proportionally to US economical size
06:30:06 <salpynx> https://en.wikipedia.org/wiki/Alice_(virtual_assistant) has interesting comments about Russian speech recognition complexity too
06:30:48 -!- tromp has joined.
06:30:55 <nakilon> int-e yeah but I would sudgest those numbers to be in thousands; then 8mln would be too much for that year, yeah
06:31:23 <nakilon> salpynx Alisa is one of the services that is a lot ahead of Google's Assistant
06:32:27 <nakilon> google Assistant can set a timer and tell a weather, Alisa at that time was doing relevant jokes, sarcasm, had a mood, was learning your conversation habits, etc.
06:33:24 <int-e> nakilon: nah, these are pretty surely raw numbers, single souls.
06:34:10 <int-e> Ah, I realize I mistranslated "Gemeinen" - these are communes.
06:36:51 <nakilon> oh now it makes sense, int-e here is the first page of the book: https://www.davidrumsey.com/luna/servlet/detail/RUMSEY~8~1~301163~90071667:Title-and-Index-Page--Atlas-der-Eva
06:37:31 <nakilon> interesting, the book in German was published in Peterburg
06:38:06 -!- arseniiv has quit (Ping timeout: 264 seconds).
06:41:07 -!- spirgel has joined.
06:45:21 -!- spirgel has quit (Ping timeout: 245 seconds).
06:48:59 <nakilon> salpynx basically because Yandex is much smaller it's easier to get things done; i.e. to make a prototype of such thing like Alisa it takes a team of guy with ideas, a mobile app dev, NLP guy, audio recognition guy -- they are all working within the same building and can gather for a meeting or a dinner in 5-10 minutes
07:05:17 -!- arseniiv has joined.
07:08:47 <esolangs> [[Talk:Polyglot]] https://esolangs.org/w/index.php?diff=83832&oldid=40598 * ColorfulGalaxy (disambiguation) * (+485)
07:08:59 <esolangs> [[Talk:Polyglot]] M https://esolangs.org/w/index.php?diff=83833&oldid=83832 * ColorfulGalaxy (disambiguation) * (+1)
07:10:04 -!- spirgel has joined.
07:14:42 -!- spirgel has quit (Ping timeout: 264 seconds).
07:26:08 -!- spirgel has joined.
07:28:14 -!- Sgeo has quit (Read error: Connection reset by peer).
07:30:21 -!- spirgel has quit (Ping timeout: 245 seconds).
07:41:59 -!- spirgel has joined.
07:46:50 -!- spirgel has quit (Ping timeout: 272 seconds).
07:51:45 <b_jonas> "<keegan> i don't play many games but when i do it's often because they are mentioned here" => yes, that helps, people here often give good recommendations
07:51:57 -!- spirgel has joined.
07:52:05 <b_jonas> "<arseniiv> I think now half my games are those which were mentioned here :D" => by people other than you?
07:52:21 -!- spirgel has quit (Client Quit).
07:53:07 <int-e> https://twitter.com/d_feldman/status/1399951777051598849
07:59:33 <river> why dont we get about/esolangs cloaks
07:59:48 <river> by registering as a community
08:02:22 <int-e> AIUI fizzie has mailed staff about it
08:02:59 <Taneb> river: we're in the queue
08:04:41 <int-e> See also https://esolangs.org/wiki/Esolang_talk:Community_portal#Freenode_and_the_future
08:06:32 <b_jonas> fizzie: one of the last steps to make the move to libera/#esolangs official will be to edit https://esolangs.org/wiki/Esolang:Community_portal to say "official home" again. I didn't dare to do that, but if you wish, you can now.
08:07:25 <b_jonas> putting it in the official password of the month was a good move too
08:07:56 <int-e> it seemed topical and simultaneously appropriate for the meme
08:08:24 <b_jonas> I was considering "dvd demagnetizer" as the password, but this is better
08:09:28 -!- hendursaga has quit (Ping timeout: 252 seconds).
08:09:52 <nakilon> the "normal" page of the map like this https://www.davidrumsey.com/luna/servlet/detail/RUMSEY~8~1~285423~90057692# has the link "View in Georeferencerer" but then I don't see a way to get from the georeferencerer back to normal page, that is vital because it has the Export button... I opened like 100 maps via Georeferencerer map navigation and see
08:09:52 <nakilon> no way to now export what I need...
08:12:57 -!- hendursaga has joined.
08:13:53 <HackEso> 188) <quintopia> vorpal: a lot of people in AK fly <Vorpal> quintopia, well getting a pilot cert is a lot more complex than a driving license :P <quintopia> being an AK resident is a lot more complex than a driver's license too
08:14:00 <HackEso> The password of the month is moving to Libera Chat.
09:04:21 <nakilon> (found it, there is a hyperlink "Website")
09:11:29 <myname> iirc, i learned about df here
09:11:58 <Taneb> Dwarf Fortress or the unix command
09:12:12 <HackEso> Filesystem Size Used Avail Use% Mounted on \ none 24G 16G 6.5G 72% /bin \ none 997M 0 997M 0% /dev \ none 24G 16G 6.5G 72% /lib \ tmpfs 124M 0 124M 0% /tmp \ none 24G 16G 6.5G 72% /usr \ none 24G 16G 6.5G 72% /sbin \ none 24G 16G 6.5G 72% /lib64 \ none 24G 16G 6.5G 72% /hackenv \ none 24G 16G 6.5G 72% /hac
09:12:15 -!- Noisytoot has quit (Quit: ZNC 1.8.2 - https://znc.in).
09:12:38 <salpynx> I'm struggling to find a direct treatment of the orientability of a disk -- seems like it should be pretty straightforward, but the fact I can't find any clear statements about it has me confused. Everyone talks endlessly about Mobius strips and more complex objeccts
09:13:00 <int-e> Taneb: the game would be my guess :)
09:13:08 <myname> best game i ever played
09:13:44 <fizzie> 10:13 [libera] -ChanServ(ChanServ@services.libera.chat)- The #esolangs namespace is registered to the esolangs project
09:13:47 <fizzie> 10:13 [libera] -ChanServ(ChanServ@services.libera.chat)- Public contacts: int-e, fizzie
09:13:55 <fizzie> That's just gone through.
09:14:06 -!- Noisytoot has joined.
09:14:31 <fizzie> (I don't know how to *actually* manage cloaks, though, even though we now have that namespace.)
09:16:20 <int-e> "Optional cloaks are available for members. These replace the hostname part with @about/yourcommunity/username, and can be requested on #libera-communities."
09:16:47 <int-e> that doesn't seem to be the way :P
09:17:18 <FireFly> I think you can just ask a friendly local staffer
09:17:47 <Taneb> I would like an esolangs cloak
09:23:46 <fizzie> I know some projects have some kind of a policy about cloaks; I don't know what ours should be. Maybe discuss on the wiki? I'll post an update on the talk page in a minute that it's now registered.
09:24:06 <fizzie> https://fedoraproject.org/wiki/LiberaCloaks "To avoid overwhelming libera.chat staff, we are only issuing cloaks in batches." Yeah, sounds like it's probably just an ask-a-staffer thing; and also makes sense to collect them into a batch rather than bother the staff one-by-one.
09:24:20 <fizzie> Not that I think our community is quite as populous, but still.
09:24:29 <int-e> ftr I don't want a cloak
09:25:17 <int-e> I like my hostname, and I don't want to give any of math, haskell, or esolangs any particular preference
09:25:28 <fizzie> I don't know about myself, but I think my bots could do with one. Or at least fungot. If it's not "about/esolangs/...", I don't know what is.
09:25:28 <fungot> fizzie: time to shove off! the name's bandeau. here to build the ocean palace! and if you wish! we shall hold this position to the last man! big fire where lavos fall from sky! we no can call that the chrono trigger. it is r66-y? cool? who knows what would become of my mystics? i must win!
09:26:25 <int-e> fungot and hackeso make sense to me. though hackeso wants a short hostmask so it's a tradeoff
09:26:25 <fungot> int-e: it's time you jumped off this mortal coil... well? yes no
09:27:12 <FireFly> oh in order to allow for as long message lines as possible?
09:27:15 <int-e> and the esolangs bot of course
09:35:23 <int-e> fizzie: I guess the policy could be something like having a wiki account with known corresponding IRC nick?
09:35:47 <fizzie> Yeah, that's approximately what I'm writing up in the next window over to that initial comment. :)
09:36:19 <int-e> We shoiuld get ais523 over here.
09:36:41 <esolangs> brctl: Ignore list: freenode/V freenode/shachaf libera/Sgeo libera/Soni libera/V libera/cd libera/int-e libera/shachaf
09:37:34 <int-e> Oh he's registered. I temporarily forgot that he's adopted a ninja style for entering the channel in the past year or so...
09:38:05 <int-e> ...so it's probably really just a matter of having logs :)
09:38:25 <fizzie> Yeah, I did in fact have a quick chat with him about this channel; he said he'll likely be continuing the thing of mostly reading via logs and then dropping by if there's anything particularly interesting.
09:47:36 <esolangs> [[Esolang talk:Community portal]] https://esolangs.org/w/index.php?diff=83834&oldid=83499 * Fizzie * (+2025) /* Libera.Chat community and cloaks */ new section
09:51:17 -!- arseniiv has quit (Ping timeout: 264 seconds).
09:55:43 <esolangs> [[Esolang:Community portal]] https://esolangs.org/w/index.php?diff=83835&oldid=83619 * Fizzie * (+115) /* IRC */ Insert the word "official".
10:14:21 -!- arseniiv has joined.
10:19:47 -!- wib_jonas has joined.
10:20:39 <wib_jonas> fizzie: be careful, I think the cloak would be longer than the bots' current hostname. if you want to request a cloak for them, at least consider that.
10:21:23 <wib_jonas> yeah, int-e pointed that out too, I see it in the losg
10:22:11 <fizzie> Yeah, it was brought up. I still think it's worth it (in terms of the length limit not being *usually* a problem), but it's a scow that you have to make such a tradeoff.
10:23:53 <wib_jonas> "<int-e> We shoiuld get ais523 over here." => he was here in the sense that he joined the channel on libera at least once; otherwise he generally logreads and joins only when needed
10:24:19 <nakilon> it should not be a problem for bot -- he doesn't see own messages
10:25:11 <wib_jonas> fizzie: if you want an actually shorter hostname, you may be able to get one, because unaffiliated cloaks are now shorter\
10:25:24 <wib_jonas> they start with user/ instead of unaffiliated/
10:25:32 <nakilon> I wonder if there is a bot that makes daily RSS from IRC logs
10:25:49 <HackEso> [U+0061 LATIN SMALL LETTER A] [U+0069 LATIN SMALL LETTER I] [U+FEFF ZERO WIDTH NO-BREAK SPACE] [U+0073 LATIN SMALL LETTER S]
10:26:06 <fizzie> Yeah, there's that. But these specific bots *are* so "about/esolangs", it'd be nice to keep it in.
10:26:26 <int-e> (funny, xterm displays a box for that U+FEFF thing)
10:26:38 <fizzie> "about/esolangs/bot/botname" is 19+length("botname") characters; anything I run off the wiki server would otherwise be "techne.zem.fi" (unless I want to play with the reverse DNS) which is only 13. So that's a net loss of 13, for the obvious 7-character bot names.
10:26:53 <int-e> (overlayed on the i)
10:27:04 <wib_jonas> do we know who runs the tunes.org logs and whether they could join the bot here?
10:27:22 <wib_jonas> a backup set of logs might be nice
10:27:44 <wib_jonas> or if kspalaiologos wants to restart their logs, that could work too
10:28:22 <wib_jonas> fizzie: they're about esolangs, but I don't see why the irc hostmust has to say that, rather than saying that elsewhere
10:28:25 <nakilon> actually website hosted logs can easily have RSS format
10:29:03 <nakilon> for example, having 10 last days items (excluding current day that isn't yet ready)
10:29:09 <wib_jonas> I mean I can see some use of the hostmask for network services and IRCops for security reasons so that anyone can quickly verify that a message is from a real IRCop or service
10:30:17 <wib_jonas> but for just HackEso or esolangs or fungоt that's less important
10:30:34 <fizzie> At least fungot's current host"name" of "2a01:4b00:82bb:1341::a" (22 characters) is only moderately shorter than "about/esolangs/bot/fungot" (25). I haven't asked if my ISP would be interested in delegating the relevant ip6.arpa zone, but given that they're your typical end-user/consumer/residential ISP, I feel like probably not. I've only had two ISPs so far who've been willing to do that.
10:30:34 <fungot> fizzie: but cyrus! are you leaving!
10:31:35 <wib_jonas> oh, fungоt doesn't run on the same server as esolangs and HackEso or at least a neighboring one?
10:31:55 <fizzie> No, I run it at home. Less mission-critical, y'see.
10:32:24 <wib_jonas> I don't currently have an always-on machine, but perhaps I should still make logs on my often-on home machine
10:33:04 <wib_jonas> I'm not willing to run any evalbot, I decided that's a security problem that I'm not willing to tackle in the near future, but just making an IRC connection with logs can work
10:33:36 <wib_jonas> and it may still be useful even if it's off like one day every two weeks and possibly shorter interruptions from irc server disconnects or network outages
10:34:08 <fizzie> Also I don't want to abuse the donated resources esolangs.org has, our hosting provider's really sponsoring the wiki rather than "everything tangentially related to the wiki".
10:34:09 <wib_jonas> only I don't have a fixed IP address, so I'd have to set up something to point there if I want to make the logs public
10:34:25 <wib_jonas> fizzie: oh! I didn't know it was donated
10:34:35 <fizzie> Yeah, nobody pays for it.
10:34:58 <fizzie> I mean, I imagine a single tiny VPS is really a rounding error for a proper cloud provider, but still.
10:36:22 <wib_jonas> sure, I just didn't know it was donated
10:37:03 <wib_jonas> Is there like a banner thanking them on the esowiki main page or something?
10:37:20 <fizzie> The tiny little ":bytemark" one.
10:37:48 <fizzie> And also a brief mention at https://esolangs.org/wiki/Esolang:About actually.
10:37:54 <wib_jonas> and it also has the new Mediawiki icon rather than the old sunflower
10:38:12 <nakilon> my "evalbot" costs 0$, it's within free GCP tier
10:38:36 <wib_jonas> oh nice, the Bytemark banner leads to a 404 page
10:38:48 <fizzie> Yeah, looks like Bytemark removed those /r/... URLs, so our backlinks to their website are now 404s.
10:38:58 <fizzie> Well, I doubt we've been generating much traffic there either.
10:40:02 <wib_jonas> sure, it just, you know, makes an internet host provider unprofessional when they break their links
10:40:30 <fizzie> I'm not 100% sure it ever worked. ;)
10:41:38 <nakilon> reminds me some IBM technology that our bank used... forgot how it was called
10:42:33 <nakilon> it was about superfast and reliable server mirroring
10:43:24 <nakilon> their webpage sais it's super reliable and the link "read more" just leads to a page with server error
10:44:34 <wib_jonas> https://web.archive.org/web/20180119185751/https://www.bytemark.co.uk/r/esolangs says it did work at some point
10:44:42 -!- imode has quit (Ping timeout: 268 seconds).
10:45:49 <fizzie> Mhm. With a proper GA ?utm_source=... param too. I'd be interested in seeing the stats of that campaign. :)
10:49:05 <nakilon> the technology kind of really worked but servers were failing once in few months and it needed a day to relaunch/sync it after a crash and so it was really possible that two could fail at the same time
10:49:43 <nakilon> but we didn't buy the third server because they costed several mln $
10:52:51 <nakilon> it's funny how things there were at the same time ridiculously expensive and unreliable
10:53:24 <nakilon> especially considering that if bank stops operating for several hours it loses government license
10:57:50 <nakilon> oh I guess I found it: "MIMIX -- Availability protects your business from downtime and data loss. Data, applications and critical system information are replicated in real ." ... https://www.ibm.com/partnerworld/public/404-page-error?solution=11871
10:58:29 <nakilon> I mean https://www.ibm.com/partnerworld/gsd/solutiondetails.do?solution=11871 -- you'll redirect
11:05:12 <esolangs> [[Esolang:IRC cloaks]] N https://esolangs.org/w/index.php?oldid=83836 * Fizzie * (+2290) Actually, let's just create this one as a placeholder, it's not like it costs money.
11:08:51 <esolangs> [[Esolang talk:Community portal]] M https://esolangs.org/w/index.php?diff=83837&oldid=83834 * Fizzie * (-13) /* Libera.Chat community and cloaks */ Fix link.
11:13:25 -!- tromp has quit (Quit: My iMac has gone to sleep. ZZZzzz…).
11:33:58 <salpynx> https://en.wikipedia.org/wiki/Genus_g_surface says: "A non-orientable surface of genus zero is the disc." That sounded interesting, but I'm so far from finding a proof or even a clear statement of how, that I'm beginning to doubt it's true. Anyone here know or can justify it?
11:34:46 <river> i don't know this math
11:36:03 <river> > Let M be a connected topological n-manifold. There are several possible definitions of what it means for M to be orientable. Some of these definitions require that M has extra structure, like being differentiable. Occasionally, n = 0 must be made into a special case. When more than one of these definitions applies to M, then M is orientable under one definition if and only if it is orientable under the others.[2][3]
11:36:05 <lambdabot> <hint>:1:85: error: parse error on input ‘of’
11:39:50 -!- wib_jonas has quit (Ping timeout: 250 seconds).
11:42:00 <river> > Theorem 3.2. A closed and connected surface is non-orientable if and only if it contains a M¨obius strip.
11:42:02 <lambdabot> parse error (possibly incorrect indentation or mismatched brackets)
11:42:22 <river> it's a bit weird to think how a disk with boundry contains a mobius strip
11:42:49 <river> i guess that if you travel "off" the disk you bounce back, flipped
11:43:31 <river> I think the disk is equivalent to the 2d projective plane?
11:43:34 <salpynx> right, that's what I have been thinking, this n=o disc feels like it must be a special case, because I can prettty much follow any higher n argument, which are easy enough to find. If a disc/disk is a special case, I would have thought someone would deal with explicitly and set out what definitions and conditions apply
11:44:02 <river> https://math.stackexchange.com/questions/308804/homeomorphism-between-real-projective-plane-and-disc
11:44:35 <salpynx> and that's what I though must apply, a disk must be in some sense homeomorphic to a mobius strip and the projective plane minus a point, but I can't see that
11:44:36 <river> maybe it's easier to understand that RP^2 is non-orientable
11:44:38 -!- wib_jonas has joined.
11:45:12 <river> X = D/~ where ~ identifies antipodal points
11:45:45 <salpynx> mobius strip is homeomorphic to the projective plane with a hole
11:45:53 <river> X = D/~ where ~ identifies antipodal points *on the boundary of D*
11:46:10 <river> so, this aligns with the theorem posted
11:46:15 <river> "contains a mobius strip"
11:46:30 <river> http://math.uchicago.edu/~may/REU2016/REUPapers/Zhang,Y.pdf 3.2 on pg 4
11:46:52 <river> when these topologists are informally saying "a disk" I guess they mean something quite specific
11:47:09 <river> where the antipodal points of the boundary are identified?
11:47:27 <river> I can see why that is non-orientable
11:48:14 <river> but that detail is not stated on wikipedia
11:52:34 <salpynx> thanks, that paper looks interesting -- the se question has me confused with the different commentators talking about whether D is a ball of sphere, and the qn looks like it was edited at some point, but that's par for se
11:52:48 <river> yeah I think the original question was mistitled
11:52:50 <river> so they fixed that
11:54:05 <salpynx> Thanks for seeing the antipodal points thing, I was stuck in an unproductive search spiral, that looks like a detail that will help make some sense of what they meant
11:56:09 <salpynx> A mobius strip is described in various places as the most simple non-orientable surface -- a standard disk sounds more simple than that, so it didn't make sense that it was non-orientable unless there was something else going on
12:05:55 <salpynx> From the Zhang paper: "RP² is in essence a disk with boundary sewed diametrically."
12:06:25 -!- imode has joined.
12:06:59 <esolangs> [[NScript]] M https://esolangs.org/w/index.php?diff=83838&oldid=83831 * PythonshellDebugwindow * (+73) Categories
12:17:58 <salpynx> That se accepted answer does make sense now, and they are talking about a disk with a very specific boundary, which they are clear about, and that is homeomorphic to RP², and also non-orientable.
12:19:12 <salpynx> river: thanks! That was causing me undue mental anguish, all from taking wikipedia a bit too seriously.
12:22:08 <river> ideally someone should fix this on wikipedia
12:24:40 <salpynx> I was just trying to confirm what the correction should be "A non-orientable surface of genus zero is the projective plane." seems like the correct verison
12:26:12 <esolangs> [[AEWNN]] https://esolangs.org/w/index.php?diff=83839&oldid=83823 * ResU * (+102)
12:29:27 <salpynx> No, that's not right, RP² is genus 1
12:30:47 <esolangs> [[AEWNN]] https://esolangs.org/w/index.php?diff=83840&oldid=83839 * ResU * (+23)
12:37:42 -!- wib_jonas has quit (Quit: Client closed).
12:40:27 -!- wib_jonas has joined.
12:42:39 <esolangs> [[Special:Log/newusers]] create * Otesunki * New user account
12:45:44 <esolangs> [[Esolang:Introduce yourself]] https://esolangs.org/w/index.php?diff=83841&oldid=83827 * Otesunki * (+202)
12:45:59 <esolangs> [[NDBall]] M https://esolangs.org/w/index.php?diff=83842&oldid=83818 * Otesunki * (+2) The demonstration for vector lists is actually kinda wonk and incorrect
12:47:13 -!- Noisytoot has quit (Quit: ZNC 1.8.2 - https://znc.in).
12:47:29 -!- Noisytoot has joined.
12:48:20 <esolangs> [[AEWNN]] https://esolangs.org/w/index.php?diff=83843&oldid=83840 * ResU * (-29) /* Hello World */
12:59:39 <salpynx> I'm going to call it here: there is no "non-orientable surface of genus zero" -- non-orientable surfaces must contain a mobius strip (genus 1) , can't find a reference, but surely a lower genus surface can't contain a higher one (otherwise it would be classified with the higher n)
13:02:13 <salpynx> This started off as an esolang related investigation I swear -- looking into describing a fungeoid playfield that is a n-holed torus rather than just a torus like Befunge and RASEL
13:06:04 <esolangs> [[Hello world program in esoteric languages]] https://esolangs.org/w/index.php?diff=83844&oldid=82672 * ResU * (+90)
13:07:55 <esolangs> [[Hello world program in esoteric languages]] https://esolangs.org/w/index.php?diff=83845&oldid=83844 * ResU * (+14) /* AEWNN */
13:20:33 -!- salpynx has quit (Quit: Connection closed).
13:22:23 <esolangs> [[User:Batata]] https://esolangs.org/w/index.php?diff=83846&oldid=83721 * Batata * (+55)
13:22:39 <esolangs> [[User:Batata]] https://esolangs.org/w/index.php?diff=83847&oldid=83846 * Batata * (+20)
13:22:48 <esolangs> [[User:Batata]] https://esolangs.org/w/index.php?diff=83848&oldid=83847 * Batata * (+1)
13:27:39 -!- Guest28 has joined.
13:30:33 <Guest28> Just confirming the Libera official browser client is Kiwi chat. It is. Hope this hasn't messed up my existing kc stored settings.
13:30:49 -!- Guest28 has quit (Client Quit).
13:31:26 -!- salpynx has joined.
13:32:44 <wib_jonas> Guest28: the official browser client is https://web.libera.chat/ . I think you have to set your settings again, because it's a separate instance from the one in https://kiwiirc.com/ .
13:33:09 <fizzie> I imagine that was in fact the desired outcome here?
13:33:28 <fizzie> (As in, not affecting the settings of the non-branded Kiwi.)
13:33:45 <fizzie> If it's client-side settings, I rather imagine the web's origin-based security model is going to enforce that anyway.
13:33:46 <wib_jonas> well, you could still use the one at https://kiwiirc.com/ since it handles connections to multiple networks together
13:34:47 <salpynx> fizzie: yes, that was me. non branded kiwi settings are fine
13:35:04 <fizzie> The IP was a bit of a giveaway.
13:35:34 <wib_jonas> just saw an announcement on http://subversion.apache.org/ that their official irc presence (I don't think they are actually too much present there) is on libera now
13:35:40 <fizzie> (Especially since it lined up, what with length "salpynx" == length "Guest28" and consecutive quit/join.)
13:35:52 <salpynx> does the cloak thing help with that? Not that I'm that worries (it's too late anyway)
13:36:05 <wib_jonas> nice, this one is official for esolangs now
13:36:31 <fizzie> Android just made no channel or network official, by removing all mention of IRC from the support page.
13:36:57 <fizzie> (freenode's #android-dev used to be mentioned there, though really it didn't have any *other* sort of official status except for the mention.)
13:36:59 <salpynx> I was wondering is there any kind of semi-reasonable paranoia justified for using the kiwiiirc given all the freenet concerns about data.
13:37:30 <salpynx> freenode, it;s late, I need to stop and get some sleep soon
13:39:05 <salpynx> I am using https://kiwiirc.com/ , with the multinetwork settings etc
13:46:14 <esolangs> [[AEWNN]] M https://esolangs.org/w/index.php?diff=83849&oldid=83843 * PythonshellDebugwindow * (-10) Header level
13:49:58 <river> > once you're a group contact there's a channel you can get invited to and you request in there
13:50:00 <lambdabot> <hint>:1:86: error: parse error on input ‘in’
13:54:50 -!- salpynx has quit (Quit: Connection closed).
14:05:28 <fizzie> Yeah, I got the impression that's the -community channel that got already mentioned. But I'll look into that once there's some initial list of cloaks people want. https://esolangs.org/wiki/Esolang_talk:Community_portal#Libera.Chat_community_and_cloaks / https://esolangs.org/wiki/Esolang:IRC_cloaks for those who filter away the recent changes feed.
14:30:11 <esolangs> [[STACKOMP]] N https://esolangs.org/w/index.php?oldid=83850 * MartinAsdf * (+15366) Created page with "Example STACKOMP program: (logo.sk) v>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>v > rstack.....v >.p..v > p >OO d v >Smpv v ^ S p<<<<< A..."
14:31:39 <esolangs> [[User:MartinAsdf]] M https://esolangs.org/w/index.php?diff=83851&oldid=72772 * MartinAsdf * (+24) added stackomp
14:39:48 <esolangs> [[User:MartinAsdf]] https://esolangs.org/w/index.php?diff=83852&oldid=83851 * MartinAsdf * (+20)
14:41:43 -!- Sgeo has joined.
14:47:22 -!- imode has quit (Ping timeout: 272 seconds).
14:51:13 <esolangs> [[SIMPLE]] M https://esolangs.org/w/index.php?diff=83853&oldid=65700 * PythonshellDebugwindow * (+259) Acronym
15:30:22 -!- wib_jonas has quit (Quit: Client closed).
15:39:40 <esolangs> [[NScript]] https://esolangs.org/w/index.php?diff=83854&oldid=83838 * Goglesq * (+47) /* Returning data */
15:42:30 <esolangs> [[NScript]] https://esolangs.org/w/index.php?diff=83855&oldid=83854 * Goglesq * (+0) /* Variables */
15:42:59 <esolangs> [[NScript]] https://esolangs.org/w/index.php?diff=83856&oldid=83855 * Goglesq * (-151)
15:43:58 <esolangs> [[NScript]] https://esolangs.org/w/index.php?diff=83857&oldid=83856 * Goglesq * (+3) /* Variables */
15:59:01 -!- moon has changed nick to hd.
15:59:05 -!- hd has changed nick to cd.
16:57:35 <esolangs> [[Tech Support Scam]] N https://esolangs.org/w/index.php?oldid=83858 * CatCatDeluxe * (+3166) Created page with "Tech Support Scam is an esoteric programming language created by [[User:CatCatDeluxe]] where it sounds vaguely like you are on a phone call with a tech support scammer. Punct..."
16:58:59 <esolangs> [[User:CatCatDeluxe]] https://esolangs.org/w/index.php?diff=83859&oldid=83468 * CatCatDeluxe * (+28)
17:00:58 -!- tromp has joined.
17:20:43 -!- tromp has quit (Quit: My iMac has gone to sleep. ZZZzzz…).
17:22:53 -!- tromp has joined.
17:38:23 -!- tromp has quit (Quit: My iMac has gone to sleep. ZZZzzz…).
17:46:02 <esolangs> [[AEWNN]] https://esolangs.org/w/index.php?diff=83860&oldid=83849 * ResU * (+278)
17:48:21 <esolangs> [[AEWNN]] https://esolangs.org/w/index.php?diff=83861&oldid=83860 * ResU * (+2) /* Cat program */
17:48:39 -!- tromp has joined.
17:53:59 -!- imode has joined.
17:56:37 <esolangs> [[AEWNN]] M https://esolangs.org/w/index.php?diff=83862&oldid=83861 * PythonshellDebugwindow * (-14) /* Cat program */ Use pre tag
17:57:51 <esolangs> [[AEWNN]] M https://esolangs.org/w/index.php?diff=83863&oldid=83862 * PythonshellDebugwindow * (-3) /* Cat program */ /* Quine */ Isn't this really a quine?
18:06:55 <zzo38> I want to make nothing up my sleeve number by sufficiently old texts, such as "The quick brown fox jumps over the lazy dog". What other possibilities might there be?
18:15:27 <fizzie> Decimals extracted from famous constants (pi, e) are kind of the conventional ones, aren't they?
18:16:21 <river> i dont like nothing up my sleeve numbers
18:16:33 <fizzie> If you want to specifically make it from well-known pieces of text... hmm, the ones that come to mind first are kind of political or religious, which you might want to avoid for obvious reasons.
18:20:22 <cd> Alice in Wonderland?
18:20:29 <cd> and other classics?
18:21:50 <fizzie> How about: "Ash nazg durbatulûk, ash nazg gimbatul, ash nazg thrakatulûk agh burzum-ishi krimpatul."
18:23:02 <b_jonas> cd: not Alice I think, it doesn't have one canonical edition that you can get the text from I think
18:23:04 <fizzie> If you need a nothing-up-your-sleeve number for an Esolang article, something about the matrix of solidity would also work.
18:23:09 -!- dutch has joined.
18:23:15 <b_jonas> the same is the problem with Shakespeare or Vergilius probably
18:24:41 <b_jonas> mathematical constants are probably better in that it's less likely that someone a few decades from now will suddenly decide that your now completely innocious text is very offensive to them and all references to it should be purged from cryptographic protocols
18:26:48 <b_jonas> and it's not easy to find anything old enough and widely distributed while still having one canonical version, unless it's something sacred like the Torah
18:28:14 <fizzie> Another nothing-up-your-sleeve source might be something that's a matter of the public record, like past Dow Jones Industrial Average values used for geohashing.
18:36:07 <shachaf> The important thing is that people shouldn't have a lot of flexibility in these things.
18:36:20 <shachaf> See https://bada55.cr.yp.to/
18:39:56 <keegan> how about the dates of eclipses
18:40:42 <b_jonas> keegan: the problem is that for anything involving cryptography, you need to know an exact value, not an approximate real, and it's not clear that exact numerical values of eclipses are canonical
18:41:05 <b_jonas> or that you can always round them that way and don't get eclipses that happen close to where your rounding would flip a bit
18:41:26 <b_jonas> also they're sort of periodical which might be bad
18:41:53 <b_jonas> I still the use of like fractional parts of square roots of the first primes for crypto protocols
18:42:45 -!- tromp has quit (Quit: My iMac has gone to sleep. ZZZzzz…).
18:55:58 -!- tromp has joined.
19:01:23 -!- arseniiv has quit (Ping timeout: 272 seconds).
19:12:06 -!- tromp has quit (Quit: My iMac has gone to sleep. ZZZzzz…).
19:23:51 -!- tromp has joined.
19:34:20 -!- dutch has quit (Quit: WeeChat 3.1).
19:37:32 -!- tromp has quit (Quit: My iMac has gone to sleep. ZZZzzz…).
19:37:58 -!- dutch has joined.
19:44:58 <esolangs> [[Pxem]] https://esolangs.org/w/index.php?diff=83864&oldid=77847 * Jedgrei * (-30)
19:49:16 <esolangs> [[Pxem]] https://esolangs.org/w/index.php?diff=83865&oldid=83864 * Jedgrei * (-169)
19:56:25 -!- tromp has joined.
20:03:20 <oren> the old testament can be used as a number and has an established text-to-numer conversion
20:04:56 <fizzie> And of course doesn't fall into the "kind of political or religious" category at all.
20:06:29 <oren> hmm... I guess Greek also has established sufficiently-old numerical values for letters, so you could use e.g. the iliad
20:06:43 -!- tromp has quit (Quit: My iMac has gone to sleep. ZZZzzz…).
20:08:08 <keegan> tfw your cryptographic algorithm accidentally generates the true name of god
20:10:27 <b_jonas> oren: yes, and the greek vs hebrew numbers are mostly the same for corresponding letters too
20:10:27 <fizzie> Is that an Unsong reference or something?
20:16:10 <nakilon> is there any French on this channel?
20:19:06 <nakilon> n-holed torus funge field sounds crazy ..D
20:27:54 -!- tromp has joined.
20:30:20 <keegan> fizzie: no it's vaguely a reference to https://en.wikipedia.org/wiki/Pi_(film)
20:32:36 <b_jonas> hey #esoteric community. someone recommended me the "uMatrix" plugin for Firefox browser. apparently this is similar in purpose to the NoScript plugin which I'm already using, in that it lets me configure with whatlists what scripts or other fancy webshit content webpages are allowed to load. do you have any experience with this "uMatrix" plugin, and can tell me about it, especially that one non-obvious
20:32:42 <b_jonas> hint that will save me days of suffering if I know it in advance?
20:38:41 -!- tromp has quit (Quit: My iMac has gone to sleep. ZZZzzz…).
20:42:12 -!- perlbot has quit (Changing host).
20:42:12 -!- perlbot has joined.
20:46:50 <b_jonas> looks good so far. there'll be some setup period, but that's normal for this kind of thing.
20:47:43 <zzo38> I have not had experience with it. Does it allow scripts and stuff to be substituted, or only blocked? What about headers? (I do have an extension to deal with headers, but making it a core part of the program dealing with sending requests and retrieving responses might be better; it can extend to any protocol and allow more to be added too.)
20:50:02 -!- tromp has joined.
20:50:53 <b_jonas> it seems it only lets you allow or block stuff, not substituted, except possibly for the Referer, but how it differs from noscript seems to be that you can allow/block based on the pair of the domain of the main webpage and the domain of the requested resource, rather than only based on the requested resource like noscript does. which means I can tell it to forbid the giantitp.com forum to block loading
20:50:59 <b_jonas> any off-site images, because it's a stupid forum that lets any poster embed externally hosted images to their posts. and then I can override that for images loaded from particular domains.
20:52:03 <b_jonas> and on the forum, those images will still be in the document tree and show up an empty rectangle, so I can load any one image by clicking on "view image", which makes that the main webpage, and images are allowed in general for most webpages in the default config
20:52:30 <b_jonas> zzo38: as for headers, it can specifically control cookies, and has something about Referer, I don't know about other headers
20:53:48 <b_jonas> zzo38: the cookies are allowed or blocked based on combination of the domain of main website loaded on the tab and the domain of the nested request where the cookies would be sent
20:54:44 <b_jonas> I like it so far, but I'll of course I may encounter problems weeks into browsing with it when visiting some particular website
20:57:04 <zzo38> Adding the user rules of headers into the core system can also define a uniform interface for setting language, do-not-track, cookies, etc.
20:57:58 <b_jonas> there are still a few things for which I'd like a plugin (or an option in this plugin). namely firefox has a global option where it can forbid webpages to set font faces except to the four user-controllable ones (default, serif, sans-serif, monospaced), which is nice and I use it (it doesn't *quite* work as it should, if my choice of font lacks characters it will use whatever the webpage chose as a
20:58:02 <zzo38> For setting cookies you might also want cookies to be allowed or blocked based on their name and/or value as well, and possibly on path, not only by the domain. You might also want to override the expiry of cookies, or possibly also the values of cookies.
20:58:04 <b_jonas> fallback font, but this is rarely relevant), but I'd like to be able to whitelist font faces per domain too
20:58:22 <nakilon> "giantitp.com forum" sounds like giant tit
20:58:31 <zzo38> I would like to be able to whitelist font faces per MIME type of the document that uses the fonts.
20:58:37 <b_jonas> and I'd like something like that for line-spacing too rather than font-face. a lot of webpages could be made better if the browser just ignored line-spacing declared in CSS everywhere.
20:59:48 <zzo38> Yes, I also; it is one of the things that my idea of "meta-CSS" would allow to do.
21:00:13 <b_jonas> and I know I can override the individual CSS rules, and I sometimes do that, but it would be easier to just ignore the website's idea about the font-face or line-spacing (or whatever those are actually called) property no matter how it's set
21:00:41 <b_jonas> now I'll visit some of my more frequently visited websites so I can set up settings for them
21:02:39 <zzo38> (Meta-CSS would allow to change the definition of CSS properties, including conditionally based on the selectors and other properties.)
21:03:20 <esolangs> [[Nevermind]] N https://esolangs.org/w/index.php?oldid=83866 * OfficialCraftCGame * (+1393) Created page with "'''Nevermind''' is a simplistic programming language made by [[User:OfficialCraftCGame|CraftCGame]] in 2021. It uses commas ('','') to separate arguments. There is no escape c..."
21:04:55 <b_jonas> one difficulty is that, just like noscript, if a webpage immediately redirects to another domain (usually via javascript but the principle is the same if it's a HTTP redirect), it's not easy to see what was allowed/blocked in the page that did the redirect, even though that's where you would have to change the permissions
21:04:57 <zzo38> Another of my idea is a nwe !priority CSS command that can only be specified by the end user (not by documents), which is another way to glboally override CSS even if !important is specified; you have to write !priority(0) or !priority(255) or whatever to set it.
21:05:24 <esolangs> [[Language list]] https://esolangs.org/w/index.php?diff=83867&oldid=83828 * OfficialCraftCGame * (+16) Added language "Nevermind" into the list
21:05:50 <zzo38> b_jonas: Yes, I know of that, and a solution (when it is a HTTP redirect) could be to record the chain of redirects and other requests so that they can be viewed later.
21:06:35 <zzo38> (Firefox will record the requests but only if the appropriate web developer window is open.)
21:06:57 <b_jonas> zzo38: that's not really enough, !important already does most of what you want in that !important in user styles are higher priority than !important in page styles,
21:07:16 <esolangs> [[Nevermind]] https://esolangs.org/w/index.php?diff=83868&oldid=83866 * OfficialCraftCGame * (-8) Changed the caption text in the table at the "Functions" section
21:09:09 <b_jonas> hmm wait. specificity of css selectors have a total order, right? is there a way to write a selector that fakes to be more specific than almost any other selector, including id selectors, but still can match whatever I want, including all elements if I choose so? if that's possible with css, I could use that to override font-face or line-spacing on a whole webpage
21:09:15 <zzo38> b_jonas: That is true, although there is still the priority by how specific the selector is. Also, in some cases you might want to be lower priority than page styles instead.
21:09:29 <b_jonas> well, except I'm not sure if it could override the style attribute of the element too
21:10:55 <esolangs> [[Nevermind]] M https://esolangs.org/w/index.php?diff=83869&oldid=83868 * PythonshellDebugwindow * (+70) /* Truth Machine */ Cats
21:16:42 <nakilon> use Stylebot extension to override styles
21:18:14 <b_jonas> what I'd also like is to link multiple website domains together so they're controlled by the same rules
21:28:37 <zzo38> You might want to match more of the URL than just the domain, too.
21:31:52 <b_jonas> zzo38: yes, there are some rare cases like taht
21:31:55 -!- tromp has quit (Quit: My iMac has gone to sleep. ZZZzzz…).
21:32:24 -!- Thelie has joined.
21:37:29 -!- tromp has joined.
21:38:40 <fizzie> Hmm, I wonder if it's worth it to enable the `chghost` capability for the logger bot, and format those messages appropriately. Guess the question is, is the simulated "Changing hostname" QUIT-and-JOIN the canonical truth or not.
21:47:29 <dutch> b_jonas: (didn't read entire backlog) uMatrix is great, but unfortunately gorhill has stopped development. See https://github.com/gorhill/uMatrix
21:48:55 -!- tromp has quit (Quit: My iMac has gone to sleep. ZZZzzz…).
21:53:26 -!- river has quit (Read error: Connection reset by peer).
21:56:48 -!- river has joined.
21:58:51 <nakilon> https://www.reddit.com/r/uBlockOrigin/comments/i240ds/request_for_a_stable_umatrix_release_for_cname/g048wyk/
22:04:06 -!- ais523 has joined.
22:05:30 <ais523> a nable is conceptually an inverted delta, not an inverted increment
22:05:37 <ais523> so why isn't there a decrement operator in Unicode?
22:08:13 <ais523> int-e: I mostly follow the channel by logreading rather than idling
22:08:49 <int-e> ais523: Yes, I should've known as I pointed out a bit later.
22:08:55 <b_jonas> ais523: I find it funny that you are mostly logreading but join to use HackEso's unidecode command of all things
22:09:53 <b_jonas> but I understand you only did it to make a point to us
22:11:08 <ais523> the other way round, actually, I was using it to *discover* that point
22:11:20 <ais523> after going through two character map applications and discovering them to both be useless
22:11:26 <ais523> then I started complaining, once I learned what the reality was
22:11:41 <ais523> <wib_jonas> do we know who runs the tunes.org logs and whether they could join the bot here? ← I believe that their tunes username is part of the URL
22:12:37 <int-e> "INCREMENT" just sounds wrong, everyone calls it a delta.
22:12:39 <fizzie> I think those two triangles are a bit weird, anyway. Because ∆ is also ∇². So they're kind of related.
22:12:49 <ais523> int-e: but a delta is a different codepoint entirely
22:12:51 <int-e> Language is weird.
22:13:06 <int-e> ais523: *a* delta, distinct from the letter delta
22:13:09 <HackEso> [U+0394 GREEK CAPITAL LETTER DELTA]
22:13:23 <int-e> also distinct from the river thing
22:13:25 <ais523> I think of a nabla as being an upside-down Greek delta
22:14:03 <fizzie> Just to add to the confusion, ∇ is also called "del".
22:14:31 <ais523> anyway, the question is, I have an esolang which could plausibly use ∆ as one of the spellings of the increment operator
22:14:43 <int-e> distinguishing between ∆ and Δ may be the real mistake here
22:14:46 <ais523> should it have a corresponding spellng for the decrement operator? and if so, should it be a nabla?
22:15:01 <int-e> Unicode is weird in its attempt to capture some semantics in code points.
22:15:16 <ais523> it is remarkably difficult to name an increment operator unambiguously
22:15:19 <fizzie> I think *if* you're going to use a pair, ∇∆ is that pair, and don't get delta-the-letter involved.
22:15:27 <ais523> I may have to resort to "+=1" and "increment"
22:16:13 <fizzie> gucharmap's notes say "U+2207 NABLA = backward difference; gradient, del * used for Laplacian operator (written with superscript 2)" and "U+2206 INCREMENT = Laplace operator; forward difference; symmetric difference (in set theory) * other symbols may also be used for symmetric difference".
22:16:20 <fizzie> I've always wondered where those extra notes are from.
22:16:21 -!- tromp has joined.
22:16:56 <ais523> oh, I actually have a defined policy about these things, and if it also means "symmetric difference" (which is also a real function) it gets disallowed
22:17:14 <fizzie> (As in, are the notes also part of the Unicode standard / code charts or not.)
22:17:27 <ais523> that makes things much easier
22:17:43 <river> <int-e> also distinct from the river thing
22:17:44 <fizzie> Apparently they are, judging from https://www.unicode.org/charts/PDF/U2200.pdf
22:17:57 <fizzie> 1. (1) delta -- (a low triangular area of alluvial deposits where a river divides before entering a larger body of water; "the Mississippi River delta"; "the Nile delta")
22:18:11 <ais523> also, thank you for telling me what gucharmap is called, now I have an actually viable character map program :-)
22:18:13 <ais523> (it's still in the repos)
22:18:31 <int-e> Oh, ∆ and Δ even look subtly different in the `fixed` font.
22:19:01 <HackEso> [U+25B3 WHITE UP-POINTING TRIANGLE] [U+25BD WHITE DOWN-POINTING TRIANGLE]
22:19:04 <ais523> they look subtly different in my font too, but it's very subtle
22:19:10 <int-e> (That seems a bit unnecessary.)
22:19:11 <ais523> the second one is a pixel taller, I think
22:19:13 <b_jonas> ais523: I thought the Delta as used for ∇² was just a greek Delta used in this meaning. I don't really understand why there's a separate ∆ character for it, and probably wouldn't use it. but apparently I do have a glyph for the ∆ character in my font, which is odd because it still has too few of these math glyphs, I should add some more.
22:19:25 <fizzie> You could go with the WHITE UP-POINTING TRIANGLE and the WHITE DOWN-POINTING TRIANGLE, those don't seem to have any semantic baggage.
22:19:48 <ais523> well, one of my aims is for the language to be fast to program in and easy to read
22:20:04 <ais523> so I support a lot of possible names for the builtins, then normalize them into something that looks better
22:20:27 <int-e> ais523: In fixed it's not a pixel taller but the sloped lines are moved ...1/3 pixel to the sides each.
22:20:31 <ais523> but it seems like increment/decrement don't have any more readable and unambiguous way to write them than "increment" and "decrement"
22:21:06 <ais523> "+=1", "-=1" are also clear, but you have issues like "++" meaning append in some languages, or "--" being a comment marker
22:21:07 <HackEso> [U+0394 GREEK CAPITAL LETTER DELTA] [U+2206 INCREMENT] [U+25B3 WHITE UP-POINTING TRIANGLE] [U+1F6C6 TRIANGLE WITH ROUNDED CORNERS] [U+1F702 ALCHEMICAL SYMBOL FOR FIRE]
22:21:13 <ais523> and "+1" and "-1" look like integers
22:21:40 <ais523> maybe I should just go for "inc" and "dec", although even "dec" looks like "decimal"
22:21:59 <int-e> https://int-e.eu/~bf3/tmp/nabla.png
22:22:22 <fizzie> All those five look pretty alike in my browser: https://zem.fi/tmp/tri.png
22:22:48 <ais523> well, they're obviously all different, but all triangles in a similar orientation
22:22:50 <int-e> (no nabla there, it's named from the final line)
22:23:00 <nakilon> https://i.imgur.com/Gh3jd4K.png
22:23:55 <b_jonas> ais523: "+=1" and "-=1" look good to me, people are already used to it because that's how you write in-place increment in python and some other languages
22:24:24 <b_jonas> though I'd usually write " += 1" but the spaces clearly have no semantic value there
22:25:43 <fizzie> Hmm, interesting. When I use `scrot` to take a screenshot, it renders pixel-perfect in a "dumb" image viewer (say, sxiv), but it's very messily scaled when the same file is opened in a browser, even when the zoom level is set to 100%.
22:25:49 <fizzie> I'm guessing maybe scrot writes the display DPI into the PNG metadata somewhere, and then the browser tries to "correct" it for the physical size, and they disagree about that.
22:26:19 -!- tromp has quit (Quit: My iMac has gone to sleep. ZZZzzz…).
22:26:24 <ais523> b_jonas: the spacing is important in this language, it tokenizes on spaces
22:26:36 <ais523> and it's semi-important for a user to know where the token boundaries are
22:26:47 <b_jonas> ais523: yes, but here it's between tokens
22:26:56 <b_jonas> spaces inside tokens are important of course, that happens in many languages
22:27:00 <ais523> well, increment is a single token in this language, even if it's two in Python
22:27:03 <fizzie> (Is there a `pdfinfo` equivalent for PNG files?)
22:27:48 <ais523> fizzie: `less` works for basic info, as does `file`
22:28:10 <ais523> apparently there's a pnginfo in the Ubuntu repositories
22:28:12 <fizzie> imagemagick's `identify` command says "PNG 158x41 158x41+0+0 8-bit sRGB 1389B 0.000u 0:00.000".
22:28:50 <ais523> huh, is sRGB there a color space or specification of which color channels exist or both?
22:29:14 <b_jonas> fizzie: maybe exiftool? I don't know what pdfinfo does. or ImageMagick's identify with specific parameters such as -verbose to print everything.
22:29:19 <b_jonas> fizzie: is there something specific you want to extract?
22:30:04 -!- cd has quit (Quit: leaving).
22:30:17 <fizzie> Whether there's something encoded about the physical size / resolution. But the `pnginfo` tool from the `pngtools` package says "Resolution: 0, 0 (unit unknown)".
22:30:18 -!- iovoid has quit (Quit: iovoid has quit!).
22:30:18 -!- Bowserinator has quit (Quit: Blame iczero something happened).
22:30:27 <fizzie> So I'm not sure why Chrome would display it scaled.
22:30:41 <ais523> maybe because the resolution info is invalid?
22:30:43 <b_jonas> if you want to see everything, I recommend (identify -verbose "$filename"; exiftool -G2 "$filename";) to see everything,
22:30:56 <nakilon> which file are you talking about?
22:30:58 <b_jonas> then once you find out what you want, there's a way to restrict either of those tools to print only that one thing
22:30:59 <fizzie> I guess maybe it's going "hmm, this image has no specified resolution, I'm going to guess 72 dpi" and then "hmm, this resolution doesn't match the screen DPI, I must scale it to present it in the right physical size".
22:31:30 <fizzie> Well, https://zem.fi/tmp/tri2.png in this case.
22:31:54 <ais523> hmm, both my Chromium and my Firefox sem to handle it correctly
22:32:24 <nakilon> pixelWidth: 158 pixelHeight: 41 typeIdentifier: public.png format: png formatOptions: default dpiWidth: 72.000 dpiHeight: 72.000 samplesPerPixel: 3 bitsPerSample: 8 hasAlpha: no space: RGB
22:32:25 <fizzie> Maybe it's a new Chrome thing. I don't think I've seen it before at home, but my work Chrome has been persistently doing something weird around those lines.
22:33:10 <fizzie> b_jonas: Thanks for `identify -verbose`. I tried `-h` and `--help`, and when neither provided useful output, stopped guessing.
22:33:27 <fizzie> Seems like I should've gone with `-help`.
22:33:32 <ais523> sometimes programs give useful information if you run them with no args at all
22:33:35 <fizzie> But single-dash long options just feel a little weird.
22:33:47 <fizzie> ...oh, that would have worked too.
22:33:55 <ais523> and there will often (but not always) be a man page, too
22:34:10 <b_jonas> fizzie: also in general for media files, you can also try (ffprobe -of flat -show_format -show_streams "$file";) but for a png file it's unlikely to tell anything that identify and exiftool doesn't show.
22:34:23 <ais523> I think my usual convention for usage information is "provide a man page, give usage information on --help or if there's an invalid argument, also give usage information if run with no arguments and that doesn't otherwise make sense for the program"
22:34:59 <b_jonas> fizzie: (identify;) without arguments works for the help, but you may have to look at HTML manuals too and even then you won't easily find all info about ImageMagick
22:35:33 <ais523> hmm, i initially interpreted "HTML manuals" as manuals about HTML, rather than manuals formatted using HTML
22:35:47 <b_jonas> fizzie: wait, I forgot an important one
22:36:00 <ais523> English can be so ambiguous sometimes, that's one of the reasons to create esolangs
22:36:26 <nakilon> width: 158 height: 41 bands: 3 format: uchar coding: none interpretation: srgb xoffset: 0 yoffset: 0 xres: 2.83465 yres: 2.83465
22:36:32 <b_jonas> fizzie: there was a specific tool for pngs that you can use to redo the compression or losslessly crop at block boundaries without redoing the fourier transform and quantization
22:36:42 <fizzie> Eyeballing the sizes (between sxiv and Chrome), I don't think it's quite "assume 72 dpi, scale to adjust to correct physical size given the 104 dpi screen density" level weird. But it's doing *something* odd.
22:36:42 <b_jonas> and that same library had an info tool
22:36:59 <fizzie> https://zem.fi/tmp/tri3.png shows both side-by-side (sxiv left, Chrome right).
22:37:35 <fizzie> (The left side should have no antialiasing/scaling going on.)
22:37:50 <b_jonas> fizzie: pngtools debian package, (pnginfo "*.png";)
22:38:05 -!- Bowserinator has joined.
22:38:09 <nakilon> the above outputs were from sips and vips
22:38:29 <b_jonas> probably also won't tell you anything that exiftool doesn't, but the pngtools package is worth knowing in general
22:39:00 <ais523> b_jonas: why do you use "(…;)" to quote shell commands on IRC?
22:39:08 <b_jonas> ais523: no, I usually omit the semicolon
22:39:14 <ais523> I don't think real shells need the semicolon
22:39:19 <b_jonas> but I do use parenthesis to quote shell commands often, even on irc
22:39:40 <b_jonas> parenthesis are nice because they work this way in both shell and cmd, so I can quote either kind of command with them
22:39:45 <esolangs> [[Talk:Turing Machine But Way Worse]] https://esolangs.org/w/index.php?diff=83870&oldid=79860 * MilkyWay90 * (+199)
22:39:58 <ais523> although in sh-alikes it's more than just a quote, it runs the command in a new subshell
22:40:02 <b_jonas> and shell commands often have other kinds of quotes or weird characters in them
22:40:25 <b_jonas> ais523: yes, you have to put a function header before them like f() if you want to quote the contents as a shell command
22:40:33 <ais523> I've taken to using «» for quoting code (including shell commands)
22:40:42 <ais523> although even that doesn't work for some languages, like Perl 6 and Jelly
22:41:33 <b_jonas> that might work, but the problem with them is that there are both french-style «..» quotes and german-style »..« quotes which can make any use confusing
22:41:55 <b_jonas> (and yes, the french-style would technically be « .. » because they put spaces inside
22:41:57 -!- iovoid has joined.
22:42:11 <nakilon> »..« looks like a confusion sign
22:42:14 <b_jonas> which makes it even harder to tell if it's a starting or ending quote character, there's a space on both sides)
22:43:11 <b_jonas> there's probably no nice way to quote any programming language, but you can often find a natural style for any particular language
22:43:27 <fizzie> Weird. It's just doing the wrong thing throughout. Even in the devtools inspector thing, if I look at a preview of the response, it does the bad scaling. (Okay, stands to reason it'd use the same code to render images. But still.)
22:43:47 <fizzie> (Chromium does the same.)
22:44:04 <b_jonas> fizzie: you could try to extract the image with some other tool and reencode it
22:44:15 <b_jonas> or is this something you want to debug to fix?
22:44:55 <nakilon> when I need to do a screenshot with details I drag the window to retina display
22:44:57 -!- cd has joined.
22:45:08 <nakilon> if I do screenshot on a usual one it's blurred
22:45:20 <nakilon> I mean, it has lower resolution
22:45:24 <fizzie> Well, preferably I'd like it fixed. But I'm not sure how much effort I want to dedicate to it.
22:49:48 <fizzie> Looking at an entirely different image that does have a set resolution (a pretty arbitrary one, 38.98x38.58 PixelsPerCentimeter), it's being scaled too. So maybe it *is* just assuming a resolution for the first PNG (just not 72; maybe it's been upgraded to default to 96 or something) and then trying to present it with the "physical" size. But I'd really rather it just render the pixels at 1:1 size.
22:50:41 <b_jonas> fizzie: I think Gimp has an option to do either for its zoom levels
22:51:01 <b_jonas> fizzie: is this a standalone PNG in the tab, or inside a HTML?
22:51:27 <fizzie> A standalone PNG in a tab. But it does the same if I just do a minimal HTML wrapper, <body><img src="tri3.png" /></body>.
22:51:46 <esolangs> [[User:OfficialCraftCGame]] https://esolangs.org/w/index.php?diff=83871&oldid=83814 * OfficialCraftCGame * (+14)
22:52:31 <b_jonas> ok. I imagine the two can differ.
22:52:49 <fizzie> If I slap a style="image-rendering: pixelated;" on it, it still maintains the scaling, just does the nearest-neighbor thing.
22:54:20 <b_jonas> because HTML has its own odd rules about images that are partly inherited from multiple decade long tradition of netscape/mozilla/firefox emulating msie, then chrome or opera or safari emulating mozilla, then edge emulating everything else again etc
22:55:49 <b_jonas> and then a growing set of extensions, including this new one that lets you give multiple image sources of which the browser chooses one depending on what resolution the image will be displayed in which can depend on whether you're using a high resolution monitor that has two pixels for every traditional pixel measured in a webpage, which is a nice extension but its syntax is seriously messed up in a way
22:55:55 <b_jonas> you think people would have learned not to do anymore when adding features to HTML
22:57:39 <fizzie> Even if I render it as the page background with style="background-image: url('tri3.png');", it's still scaled. And not in Firefox. Odd.
23:00:35 <fizzie> And the exact same thing happens to something like https://upload.wikimedia.org/wikipedia/commons/e/e6/Png_pixel_cube.png -- the image as viewed in Firefox is rendered "naturally", but the one in Chrome is scaled by... 69/64.
23:00:53 <b_jonas> fizzie: make sure this isn't just caused by some unusual setting or plugin that you set in those browsers
23:01:51 <fizzie> If it is, I don't know what setting it could be. The Chromium one definitely should be pretty vanilla.
23:02:01 <zzo38> Also some servers will try to force you to download the picture instead of displaying it, although I can usually force it to display it in the browser by entering a data: URI
23:02:11 <fizzie> My "desktop environment", on the other hand, might be more uncommon, though.
23:02:30 -!- salpynx has joined.
23:02:50 <zzo38> What desktop environment is that?
23:04:33 <fizzie> Well, it's no specific one. It's just Xmonad as a window manager, and some bits and pieces of GTK+ thing.
23:04:36 <zzo38> (Also, I don't have the problem with wrong scaling on the browser)
23:04:50 <nakilon> what do you mean scaled by 69/64? distorted the aspect ratio?
23:05:15 <fizzie> No, just uniformly scaled.
23:05:26 <fizzie> The image itself is 64x64 pixels, but it's rendered as 69x69.
23:06:57 <nakilon> renders as 64 and 128 for me
23:07:13 <fizzie> I did a chrome://flags + "reset all" just to be sure, so it shouldn't be a manually enabled weird setting either.
23:07:41 <nakilon> make sure to do ctrl +/- until it says 100%
23:08:24 <nakilon> ctrl-0 resets to your settings default that can be different
23:08:36 <fizzie> It's been 100% the whole time.
23:08:58 <zzo38> I use a program I wrote by myself to take screenshots. (They don't have as many options as scrot, because other programs can be used for many of the things. However, one option that perhaps should be added is the option to control if the mouse cursor is captured; scrot does have that option.)
23:13:30 <fizzie> Just verified that it's not a PNG-specific problem. Opening a (data URI that encodes a) 222x227 pixel JPEG file renders a rectangle that's exactly 222x227 pixels in Firefox, but 239x245 in Chromium.
23:13:31 <int-e> fizzie: so that's in line with assuming 96dpi and adjusting that to 104dpi. eww.
23:14:52 <int-e> or maybe 103.5 rather than 104.
23:15:14 <fizzie> I can't believe it's intentional, I think it must be some kind of a bug. But not an obvious one to track down.
23:15:41 <esolangs> [[Language list]] https://esolangs.org/w/index.php?diff=83872&oldid=83867 * ResU * (+12) Added AEWNN
23:16:23 <int-e> well it could be a feature intended for 200+ dpi displays
23:16:51 <fizzie> I started chromium with --force-device-scale-factor=1 and it stopped scaling.
23:16:59 <fizzie> So I guess you might be right in that.
23:17:02 <nakilon> basically retina is just doubling the resolution
23:17:15 <nakilon> and disables font antialiasing
23:17:18 <int-e> and quadrupling required bandwidth
23:17:47 <fizzie> It's just that, I don't think it makes sense to do scaling like that when the display depth is so close.
23:18:31 <fizzie> Well, for now I'll just stick that flag onto my list.
23:18:45 <int-e> Hmm, my display reports 94x95 dpi, that could be fun. But maybe chrome refuses to downscale at least.
23:19:08 <fizzie> Oh, I guess another alternative would be to just make the display resolution lie.
23:19:23 <fizzie> But maybe it's convenient to be actually able to approximate physical sizes in Gimp or whatnot.
23:20:00 <nakilon> btw, recently Chrome started showing everything smaller for me both on Windows and macOS
23:20:14 <int-e> (But there's no chrome or chromium currently installed here, so I won't test)
23:22:26 <esolangs> [[Language list]] M https://esolangs.org/w/index.php?diff=83873&oldid=83872 * CatCatDeluxe * (+24)
23:38:38 -!- Thelie has quit (Remote host closed the connection).
23:43:50 -!- Bowserinator has quit (Remote host closed the connection).
23:43:52 -!- iovoid has quit (Remote host closed the connection).
23:53:50 <esolangs> [[Truth Machine (esolang)]] N https://esolangs.org/w/index.php?oldid=83874 * ResU * (+587) Created page with "'''Truth Machine''' is an esolang created by [[User:ResU]] in 2021. [[Category:2021]][[Category:Languages]] ==Commands== {| class=wikitable !Cmd !Description |- |<code>input(a..."
23:55:01 <esolangs> [[Special:Log/move]] move * ResU * moved [[Truth Machine (esolang)]] to [[Truth Machine (language)]]
23:55:11 <fizzie> Yeah, the GTK UI does `resolution / 96` as the scale factor, and then rounds it to the nearest 1/64th, meaning my 104/96 rounds to exactly 69/64. Or something along those lines. https://source.chromium.org/chromium/chromium/src/+/main:ui/gtk/gtk_ui.cc;l=1085;drc=71ce45f2f742ca6449749a906ce5bb9948c8b555
23:55:17 <fizzie> It doesn't seem to have changed recently, but maybe what has changed is how universally fractional scaling factors are applied to images, or something. Because while --force-device-scale-factor=1 does solve the ugly image scaling, everything *else* does look a little smaller now than I'm used to. But maybe it's not really logically consistent to scale text and other scalable things while keeping images
23:55:38 <fizzie> Maybe I'll just get used to the smaller size.
23:57:12 <int-e> fizzie: not scaling pixel images a form of hinting, which has always been a compromise.
23:57:30 -!- Bowserinator has joined.
23:59:41 -!- iovoid has joined.