00:00:02 <shachaf> fizzie: Did you press Alt?
00:00:21 <b_jonas> then later it became fake textareas that try to handle keypresses, and now it's password fields where the javascript tries to decide whether to reveal the password or not, instead of leaving that to the browser.
00:00:31 <fizzie> I think the text looks kind of blurry here.
00:00:55 <shachaf> fizzie: Hmm, as far as I know it's using MSDF trickery which should allow for pretty sharp text.
00:01:37 <shachaf> Oops, I shouldn't use acronyms. I mean multichannel signed distance field representation of typefaces.
00:02:00 <fizzie> I don't know how I could get a really comparable view. It's not exactly fair to compare it to a terminal window using a bitmap font, when it comes to crispness.
00:02:57 <shachaf> I wonder whether you're using a high-DPI screen and it's rendering at a lower resolution and then the browser is scaling it up, or something.
00:03:00 <b_jonas> and yes, custom checkboxes and option buttons too, although for those I can partially fault browsers, because there was no way to make the default check boxes and option buttons look reasonable on a dark background, and basically EVERY website forever have always overridden the background color, so it's something that browsers should have anticipated to allow
00:03:09 <ais523> b_jonas: after review of the BSD exit code standard, I've concluded that it's better than returning errno in almost every way
00:03:25 <b_jonas> ais523: what's this BSD exit code standard? can you point me to a description?
00:03:32 <ais523> because most of the exit code reasons it envisages aren't tied to errno failures at all, and for those that are, it returns what it was doing at the time of the error rather than the immediate cause
00:03:36 <ais523> http://www.manpagez.com/man/3/sysexits/
00:04:19 <ais523> even then, I find the BSD version is far from perfect, it's missing a lot of things that you'd want and doesn't differentiate well enough
00:04:48 <ais523> but, e.g., exit(ENOENT) is not much useful working out what went wrong, you'd definitely prefer the sysexits code in that case
00:06:26 <b_jonas> well, there'll always be specific programs or future needs that can't be covered by any of these standars
00:06:27 <fizzie> shachaf: I don't know. Probably not, it's not a huge-DPI screen; I do xrandr --dpi 104 as a compromise since there are two screens of different density. Actually, if I turn the browser scaling up to 150% the text actually looks pretty sharp (what?), although overly big. At 100% less so.
00:06:48 <ais523> say you're writing grep, what do you return if no matches were found?
00:07:03 <b_jonas> and in some cases, instead of exit code handling, you want command-line flags or other config to tell the program to consider certain errors as successes
00:07:16 <ais523> I think 1 should be added as an additional exit code with a similar meaning to "no"
00:07:32 <shachaf> fizzie: Hmm, I guess it wouldn't be shocking if this was an artifact of the MSDF rendering.
00:07:33 <ais523> which is how most programs use it already
00:07:59 <b_jonas> ais523: 1 if no matches, 2 if the input files aren't readable and you haven't ignored that with -s or something, 3 if the arguments are wrong or the regex has syntax errors
00:08:11 <ais523> https://man.openbsd.org/grep#EXIT_STATUS
00:08:20 <b_jonas> and yes, that's not a case of errno exits
00:08:33 <ais523> grep doesn't consider failure to find an error, but exits with nonzero status anyway
00:08:35 <b_jonas> the errno exist are useful only for quick and dirty scripts, they're not something you should use in serious programs
00:08:41 <b_jonas> but for typical one-liner programs, they work
00:08:54 <ais523> I guess we have a separate category of "failures"
00:09:05 <ais523> nothing went /wrong/ but the request turned out to be impossible
00:09:34 <b_jonas> ais523: yeah, that more or less agrees with what I said, 1 for normal false, 2 for actual error
00:09:39 <b_jonas> 2 or higher for actual error that is
00:09:48 <b_jonas> because sometimes you want to distinguish different errors
00:10:05 <ais523> hmm, does this mean we need a variant on false(1) that returns 69 rather than 1?
00:10:05 <fizzie> These days you'll probably also want the exit code equivalent of HTTP 451.
00:10:26 <b_jonas> and yes, I do understand that that clashes with errno exit, because 1 is EPERM, which is a reasonably common errno
00:10:28 <ais523> fizzie: 77 probably covers that in the sysexits system
00:10:30 <ais523> although it's a bit more general
00:10:58 <ais523> that said, I can't think of many executable programs that refuse to do something because it's illegal
00:11:07 <ais523> the only example that comes to mind is Photoshop refusing to open images of money
00:11:26 <fizzie> At the office, there's a monitor in one corridor, it's showing a video that shows HTTP error codes as illustrated scenarios at a restaurant, or something along those lines, as far as I have managed to figure out based on three-ish glances of a second each.
00:11:48 <ais523> ooh, also PDF readers that obey the "don't print this", etc., flags in the PDF file (which are just flags that can blatantly be ignored by a reader if it isn't programmed to respect them)
00:11:58 <b_jonas> ais523: some programs didn't have rar/mp3/gif encoder compiled into them because of patent or similar legal reasons
00:12:07 <b_jonas> does that count as "refusing" to do something?
00:12:11 <ais523> fizzie: just the 4xx codes, or all of them? :-D
00:12:25 <ais523> b_jonas: oh, that's a good point
00:12:29 <b_jonas> ais523: also some websites refuse to let you register if you're younger than a certain age limit
00:12:38 <ais523> the crypto export restrictions are another example of that
00:12:42 <fizzie> ais523: I've seen 404 not found, 300 multiple choices and one more I don't remember, which was probably a 4xx too.
00:13:03 <b_jonas> of these, only the rar encoder limitation remain
00:13:08 <ais523> this is totally the sort of thing that makes sense to have in a random corridor at Google :-D
00:13:17 <fizzie> The 404 not found was an empty table, for the 300 I didn't see the actual thing, just the "title card", and for the last one there were two people sitting around the table but I forget what they were doing.
00:13:35 <ais523> b_jonas: the crypto restrictions are a real bother because it's hard to figure out what counts as crypto sometimes
00:13:48 <b_jonas> and ebook readers and even browsers that refuse to save a video
00:13:53 <fizzie> Oh, I remember now! It was 418 I'm a teapot, and it wasn't two people, it was just one person, but there was a teapot on the table.
00:14:10 <ais523> 418 isn't the most useful status code to know :-D
00:14:22 <fizzie> I don't know how the scene continued, I've only been in a hurry to a meeting, should probably stop and actually watch them.
00:14:24 <b_jonas> ais523: yes, but didn't the legal situation in the US actually change there in a way that made most of that crypto problem disappear?
00:14:33 <ais523> I assume 204 would be people ordering something, being brought an empty plate, and they left happy because it's exactly what they ordered
00:14:46 <ais523> b_jonas: it's less insane than it was but the rules still exist
00:15:03 <kmc> it's now only restricted for countries on the naughty list
00:15:13 <ais523> and aren't specific to the US either, basically the old situation was insane rules in the US and slightly less insane rules elsewhere, the current situation is for the slightly less insane rules to be everywhere
00:15:18 <b_jonas> ais523: but are those rules that actually distinguish strong crypto software from software in general?
00:15:43 <b_jonas> I know there are still embargoes to export software to certain countries, such as North Korea or Cuba, but I don't think crypto vs non-crypto matters there
00:15:45 <ais523> b_jonas: AFAICT the rules are mostly based on the end user's use of the software
00:16:09 <ais523> so they wouldn't ban software with a large amount of dead code that implemented a strong crypto algorithm as long as users weren't meant to discover it
00:16:14 <b_jonas> ais523: the end users's use of the software? how the heck am I supposed to know that?
00:16:15 <ais523> I'm not sure on this, though, not being a lawyer
00:16:23 <ais523> b_jonas: I called them insane for a reason :-P
00:16:44 <ais523> I think they intend to make things like NetHack legal even if they use strong CSPRNGs
00:17:16 <b_jonas> ais523: do they still have rules where if you print the source code on paper, send it in mail, and scan and OCR it in Europe, it's no longer illega?
00:17:29 <ais523> I don't think so? besides, these rules apply to Europe too
00:17:42 <ais523> if they /only/ applied in the US I wouldn't care, I don't plan to go there
00:18:06 <b_jonas> ais523: you don't plan to go there, but you may want to use software that someone in the US writes, and you want them to be able to send the software to you
00:18:28 <fizzie> I didn't know about https://www.google.com/teapot
00:18:44 <fizzie> (It just returns a 418 with an appropriate body.)
00:19:09 <ais523> b_jonas: I think you're missing some of my inherent lawfulness; I am worried about /accidentally/ breaking crypto export laws in some piece of code I write
00:19:51 <b_jonas> ais523: but do you care about other people breaking the law when you ask them to send you software, even if you aren't breaking the law for receiving the software?
00:20:25 <ais523> fwiw, I believe that the current US laws on crypto require you to send a copy of any crypto code to a particular email address before you're allowed to export it
00:20:43 <b_jonas> ais523: hehe, that also sounds insane
00:20:53 <ais523> if it's closed-source you need to get permission before selling it, if it's open-source you don't need the permission before distributing it but still need to send a reference to it to the official address
00:22:14 <ais523> b_jonas: I don't like tricking people into breaking the law, but if they don't care about the law in question and are willing to risk any consequences, and I also don't personally consider the law to be useful/good, I find it hard to get worked up about their law breach happening to work to my benefit
00:23:41 -!- MDead_ has quit (Quit: Going offline, see ya! (www.adiirc.com)).
00:24:17 -!- MDead_ has joined.
00:24:17 <ais523> hmm, after thinking about this policy, it also probably needs adjusting for the capability for consent of the person in question
00:24:41 -!- MDead_ has quit (Remote host closed the connection).
00:24:42 <b_jonas> does Canada have such crypto-specific export restrictions too?
00:25:02 -!- MDude has joined.
00:25:45 <ais523> oh wow, the Wasenaar Arrangement has its own website: https://www.wassenaar.org
00:27:45 <b_jonas> ais523: that's written in management speak or politician speak or some such language. I don't understand a word.
00:27:52 <b_jonas> is there a language switcher somewhere?
00:28:43 <ais523> but if there is, the best you're likely to get is Hungarian politician-speak rather than English normal-person-speak
00:29:04 -!- xkapastel has quit (Quit: Connection closed for inactivity).
00:31:04 <b_jonas> nope: https://www.wassenaar.org/about-us/#faq says "The working language of the Wassenaar Arrangement is English and no official translations in other languages are provided by the Secretariat."
00:31:13 <b_jonas> but it's only what they call English
00:31:39 <ais523> hmm, page 92 of the Wassenaar Agreement "List of Dual-Use Goods and Technologies and Munitions List", the introduction to "Information Security", says "Not used since 2015"
00:31:44 <b_jonas> though with the sitemap I can get to other pages that are slightly more readable than the front page
00:31:52 <ais523> and the last time I checked the laws on this was from a source written before 2015
00:32:07 <ais523> this is potentially really good news, but I'm not sure exactly what it applies to
00:33:33 <b_jonas> for my vacation in may, I read some of the airplane luggage regulations, which tell things like that I'm not allowed to put billiard clubs or throwing stars into my hand luggage
00:33:53 <b_jonas> they do also tell more useful information, like about carrying batteries
00:35:34 <b_jonas> in particular, they basically say that all the lithium batteries must be either inside a device that uses it (so that the chasis of the device protects it mechanically) or in the hand baggage and its terminals covered by isolation tape (that's not the exact wording they used, but the sense)
00:36:15 <b_jonas> which sounds quite reasonable actually, provided you buy into the constraint that they really want to allow people to watch movies on their notebooks on the airplane
00:36:36 <b_jonas> the announcements on the plane also changed:
00:36:39 <ais523> what's a billiard club? the only English meaning I can think of for the noun phrase would be a club of people (or building housing such a club) that play billiards, either would be hard to fit into hand luggage
00:37:15 <ais523> is this a mistranslation from Hungarian, or something I was formerly unaware of?
00:38:25 <b_jonas> they talk a lot more about batteries, like warn people that you mustn't charge your lithium batteries during takeoff and landing, whereas they no longer care much about radio interference that your devices use (whether mobile phone protocols, wifi, bluetooth, etc; basically they recognized that with today's electronics it's no longer a practical problem, people won't use vintage equipment on the
00:38:31 <b_jonas> airplane, but they still reserve the right to ask you to turn off your devices in exceptional cases)
00:38:55 <b_jonas> ais523: the long stick that a player uses to hit the balls in the billiard pool game
00:39:36 <b_jonas> typically longer than a meter and rigid and a single peace, so it's practically impossible to get it into your hand luggage
00:39:41 <ais523> oh, that's called a "cue" in English
00:39:58 <ais523> normally "pool cue" or "snooker cue" because nobody plays billiards nowadays, but I guess "billiard cue" would be understandable
00:40:16 <b_jonas> they forbid it for any of the related games
00:41:05 <ais523> pool is probably more commonly played in both the US and UK because it's a common pub game, but snooker is more /popular/ in the UK, e.g. major snooker tournaments are televised on major channels and lots of people watch them
00:41:07 <b_jonas> you can carry nail clippers though
00:41:30 <b_jonas> ais523: yeah, pool is the sport people play, snooker is the one they watch in television
00:43:03 <b_jonas> like how in the weekend, I watched the Red Bull Air Race, where the main event is a airplane aerobatics obstacle course speed race, which is a sport that very few people practice, but much more want to look at it, so there was a crowd of spectators
00:43:22 <b_jonas> they even had a helicopter aerobatics demonstration as a side event, which is even crazier in this respect
00:45:35 <b_jonas> and as for luggage rules, on the public transport buses here, I often see (on particular parts of the city) people carrying adult hockey or fencing equipment or large musical instruments, which are technically against the rules of the transport company, but nobody actually cares about those particular rules or enforces them, so they didn't bother to update the rules to match the reality
00:46:55 <b_jonas> the rules also say that you mustn't eat on the buses, and that you can carry dogs either inside a box or (on a leash and with a muzzle) and only one unboxed dog per vehicle. these too are rules that people often break and they're basically never enforced.
00:48:02 <b_jonas> I think it's actually one dog per carriage for trams and metros
00:48:04 <ais523> on a few occasions I've seen stray (or at least no-accompanying-owner) dogs use the UK public transport on their own
00:48:38 <ais523> nobody attempts to charge them or check their tickets, so they get a discount compared to humans
00:48:42 <b_jonas> ais523: I haven't seen those on buses here, though I have seen dogs that follow an owner but don't much obey them
00:49:13 <b_jonas> ais523: does anyone attempt to get them off the vehicles? that sometimes happens with homeless people here
00:49:21 <ais523> I haven't seen it often enough to know whether the dogs had any idea what they were doing, but have seen anecdotes from other people that some dogs do the same journey repeatedly and thus have worked out some method to know what stop they want
00:49:36 <ais523> b_jonas: not really, I don't think
00:49:42 <b_jonas> I've never seen it happen with dogs, because the kind of people who carry those big out of control dogs are intimidating and the controllers wouldn't want to deal with them
00:50:02 <ais523> in London, if someone starts begging on an Underground train, they change the electronic signs to tell people not to give them monry, that's one method of trying to get them off the vehicle
00:51:05 <b_jonas> ais523: that doesn't sound surprising, there are usually lots of clues that let you figure out where to get off if you have travelled a lot in the same location,
00:51:23 <b_jonas> except sometimes in underground tunnels and a few confusing places
00:52:13 <b_jonas> which is why I can usually get off at the right place even if I'm reading a book, unless the book is VERY interesting, or if it's cold and the windows are completely opaque from dew, and blind people can get off fine too
00:52:51 <ais523> <ais523> hmm, page 92 of the Wassenaar Agreement "List of Dual-Use Goods and Technologies and Munitions List", the introduction to "Information Security", says "Not used since 2015" ← ugh, apparently it's just the specific line in question that's not used since 2015, not the entire section
00:52:52 <b_jonas> of course it's getting easier now that buses have gps-based trackers connected to the automatic voice announcement that *usually* get the stations right without the driver having to intervene
00:53:06 <ais523> like, they deleted the line entirely and that's a way of saying "this line intentionally left blank" to not throw off the numbering
00:53:48 <b_jonas> changing the sign when people are begging => wow, I've never seen that
00:55:21 <b_jonas> there's a person who's regularly in the Kálvin tér metro station, usually either in the M3 platform or in the tunnel between the M3 or M4 platforms (he can't be on the M4 platform itself because that always has guards on it because the M4 is driverless)
00:55:34 <b_jonas> he's making terribly annoying noises with a fiddle
00:55:49 <b_jonas> I wish someone got him out of there or silenced him, but no
00:56:12 <fizzie> I don't think they really do that super-consistently, at least on the stretch I'm at.
00:56:19 <b_jonas> and no, he doesn't seem to be asking for money to stop making the noises, because I do often see people giving him money and he doesn't stop the noise when he gets them
00:56:38 <fizzie> They do have a recorded announcement, though. "There are beggars and buskers operating on this train. Please do not encourage them by giving them money."
00:57:08 <b_jonas> fizzie: well sure, because the people who beg know that the transport company doesn't want them to be their, so they're stealthy, they don't attract attention of the driver or controllers
00:57:39 <fizzie> A lot of the times the people who do the "play one folk song during the tmie between the stations" thing go from car to car using the "danger, do not use these doors" doors.
00:57:46 <fizzie> Which I presume is for avoiding attention.
00:58:04 <b_jonas> and yes, begging for money or selling stuff without explicit permission is also explicitly against the transport company rules, even in the metro platforms, and that is sometimes enforced
00:58:25 <b_jonas> but it doesn't seem to be enforced in the case of this particular man in Kálvin tér
00:58:39 <fizzie> TfL has a registered busker program, though not on the trains, just on some specifically designated locations at stations.
00:58:42 <ais523> in the London Underground busking is allowed at specific locations in *stations*, presumably as long as you're good enough
00:58:47 <ais523> they're circles marked on the ground
00:58:51 <ais523> oh, I see fizzie said that first
00:59:01 <fizzie> ais523: They do auditions and all.
00:59:06 <fizzie> https://tfl.gov.uk/corporate/about-tfl/culture-and-heritage/busking
00:59:26 <b_jonas> I don't think that happens here in areas that belong to the platform company,
00:59:50 -!- atslash has quit (Quit: This computer has gone to sleep).
01:00:09 <b_jonas> but performers are allowed in some public areas like streets or underground passages that don't technically belong to the transport company but are still hubs that are very hard to avoid for many people going that way
01:00:21 <b_jonas> I think even they're not allowed to be very distracting, but they are
01:01:07 <ais523> in at least one station in Birmingham there's a permanently installed piano, with instructions only to play it if you're sufficiently good at the piano
01:01:17 <b_jonas> one example is a certain group of people who do music with siging and shouting with some christian evangelism theme in the evenings in a part of the Keleti pu underground passage
01:01:23 <ais523> it turned out to be a pretty simple and cheap (in a relative scale) way to give a nice variety of music for the station for free
01:01:28 <fizzie> There's been a lot written about the privatization of public space in London recently, a lot of the places with the new developments have "public" squares and the like which aren't actually public land, it's just a deal they've done to get planning permission to allow the public some level of access.
01:01:54 <ais523> (I've never seen anyone ask for money for playing it, and it tends to get played by random members of the pubilc waiting for a train)
01:01:56 <fizzie> But the landowners still reserve the right to move "undesirable" elements out, and of course they don't list their own rules and regulations anywhere.
01:02:20 <fizzie> I think the Pancras Square where our office is is one of those private-public places.
01:02:44 <b_jonas> there was also an american style country music singer, with the stereotypically appropriate hat, in the Móricz underpass. I found that odd.
01:04:19 <b_jonas> but I guess it makes some amount of sense: if we can have jazz and blues bands, why not country music singers?
01:05:03 <fizzie> TfL also has a program called "Platform 88" where they've put some public pianos at stations, and anyone can sit down and play.
01:05:22 <b_jonas> fizzie: oh, we've had some public pianos on streets. not in stations or platforms.
01:05:22 <fizzie> Usually the people that do tend to know how to play. But not always.
01:05:42 <b_jonas> fizzie: well no, those pianos do attract children too to some amount
01:06:33 <b_jonas> I think we should also have children-centric buttons in public places and vehicles, where those buttons do visibly control something, such as change the pattern of some signal lights, but don't control anything actually important
01:06:49 <fizzie> There's one in the King's Cross underground station ticket hall area, and another at the railway station, somewhere in the general vicinity of platform 9¾.
01:07:47 <b_jonas> and the adults should have a conspiracy where they mostly pretend in front of the children that the buttons do something important and that the children shouldn't press the button unless the adult asks them, but in reality they also direct the children to press those buttons and enjoy doing something forbidden and pressing buttons,
01:07:51 <fizzie> At Granary Square, you used to be able to play Snake on the fountains with a phone app. Not sure if it's still operational.
01:07:57 <b_jonas> rather than pressing the actually important buttons
01:08:07 <fizzie> (The fountain is essentially a largeish grid.)
01:08:17 <b_jonas> this should also happen in museum exhibitions and other places that children may visit
01:10:43 <b_jonas> I think I've also seen a public space piano in the netherlands on my vacation
01:11:31 <b_jonas> but there's one here too near Ferenciek, and I think there are a few more somewhere elce in the center
01:11:33 <fizzie> I keep waiting for somebody to play video game music on the piano, but it never happens.
01:11:42 <ais523> OK, so almost anything crypto-related that's easily user-modifiable can't be exported from either the US or the UK, unless it uses only weak crypto (56 bytes symmetric, 512 bytes when using old-fashioned common asymmetric problems), or it falls under the General Software Note
01:12:01 <ais523> I need to find another source for the General Software Note, though, because the version I'm currently using is missing indentation that makes a huge difference to the associativity of the requirements in it
01:12:08 <fizzie> Back in Finland there was a busker playing the Monkey Island theme tune in the Leppävaara train station underpass, on an unlikely instrument, I forget exactly which.
01:13:07 <ais523> got it, this is a fairly favourable associativity
01:14:03 <b_jonas> is silence technically video game music?
01:14:25 <fizzie> I guess? There's games with no music, or sound.
01:14:30 <fizzie> Not quite in the spirit of the thing though.
01:15:02 -!- sprocklem has quit (Ping timeout: 245 seconds).
01:15:16 <ais523> any software is legal, at least based on crypto-export-related requirements, if it's either sold unrestricted to all-comers (i.e. anyone can just buy it without being vetted, etc.) and "Designed for installation by the user without further substantial support by the supplier"; or else "in the public domain"
01:15:39 <ais523> but the "in the public domain" here is using a weird restriction that allows it to be copyrighted as long as there are no restrictions on its further dissemination
01:15:56 <ais523> so I guess CC0 is OK, and BSD and GPL are in a grey area
01:16:05 <b_jonas> ais523: hmm, that's interesting
01:16:12 <ais523> *using a weird definition
01:17:54 <b_jonas> that sounds like an excuse that the companies that sell industry software can use about their abysmal lack of designing anything with security in mind, since they sell software that doesn't match those conditions
01:18:44 <b_jonas> ais523: does this apply if you export software from th UK to within the EU?
01:19:20 <b_jonas> maybe they can't use that excuse then
01:19:20 <ais523> or well, I think there may be more lenience, up to automatic export-is-allowed, for within-EU exports
01:20:12 <b_jonas> they have to argue that they can't make sure that customers don't use the EU-only secure version of their software in asia
01:20:31 <b_jonas> I guess that's still a reasonable excuse
01:20:49 <b_jonas> I mean as much as the whole thing is
01:21:16 <b_jonas> it's just an excuse, their design does actually suck
01:21:30 <ais523> so it looks like within-EU follows different rules which are much more liberal without being completely free; for exports that would otherwise be controlled, you can get a license valid for export to Australia, Canada, Japan, New Zealand, Norway, Switzerland, and the US fairly eaily; for other countries you need specific authorisation
01:22:43 <b_jonas> wekk we do reexport the industry software to korea, so that doesn't sound enough
01:22:58 <b_jonas> (as well as to EU countries)
01:26:17 <ais523> actually, one potential issue I noticed was that there's an explicit list of countries where you can't apply for the specific authorisation, it includes all the EU countries listed individually (presumably because you already have a blanket authorisation)
01:26:57 <ais523> so when Brexit hits, if they forget to adjust the law and the EU blanket authorisation stops applying to the UK, then the rest of the EU will end up in the same status as Iran and North Korea (i.e. you can't even attempt to apply for a license)
01:27:50 <b_jonas> ais523: sure, that sort of thing always happens whenever some autonomous region in the balkan or the caucasus gradually converts into a country, they get into country lists too late
01:28:47 <ais523> you're reminding me of the running joke that Berwick-upon-Tweed was at war with Russia until 1966
01:28:56 <ais523> (when they explicitly signed a peace treaty, just in case)
01:29:11 <b_jonas> admittedly those are generally places not recommended for tourists, so if an insurance company doesn't add it to the list of countries with Europe rate, it could be a feature
01:30:01 <ais523> the issue being that it was in disputed territory between England and Scotland and thus occasionally listed explicitly when announcements were made, and the announcement that started the Crimean War did so but the announcement that ended it didn't
01:30:47 <ais523> (there are several reasons why this can't possibly be true, but it made for a fun publicity stunt)
01:31:37 <b_jonas> things that definitely aren't a country on their own, but are disupted between countries, and the people far away don't want to make it seem like they politically support any country's claim on the area
01:31:57 <ais523> <Wikipedia> Ancient Rome and Ancient Carthage never signed a peace treaty after the Romans seized and completely destroyed the city of Carthage in 146 BC and enslaved its entire surviving population, leaving no entity with which to make peace. In 1985 the mayors of modern Rome and Carthage municipality signed a peace treaty and accompanying pact of friendship.
01:32:12 <ais523> …that's a really interesting legal problem
01:32:16 <b_jonas> I think there's several of those between Israel and Palestine, possibly nested
01:34:10 <b_jonas> ais523: so basically Carthage was disbanded without a legal successor?
01:34:19 <ais523> b_jonas: right, it was completely destroyed by the Romans
01:34:42 <ais523> so you have this permanent war of the Roman Empire against, effectively, a null set
01:35:04 <ais523> presumably the establishment of a new city in the location of Carthage would be a separate nation altogether, unable to do anything about the previous war
01:35:13 <ais523> (and of course, modern-day Rome is not part of the Roman Empire either)
01:36:01 <b_jonas> yeah it can happen that there are multiple different groups of people who claim that they're the worthy successors, such as basically everyone in europe for the roman empire, and you can't regard any of them as the successor without insulting all the others
01:36:30 <b_jonas> that could happen with a country if they were in war with Atlantis or Numenor when it got destroyed
01:36:53 <b_jonas> it gets particularly ugly with churches
01:36:58 <ais523> I thought it was fairly well established that the Roman Empire split into two parts, the western part was entirely overrun and ceased to exist as an entity, the eastern part held out longer but was eventually conquered by the Ottomans
01:37:15 <ais523> (but the Eastern Roman Empire didn't control Rome)
01:37:25 <b_jonas> ais523: yes, but who is the worthy successor of the western empire?
01:37:31 <fizzie> There's that one place on a border which is anti-disputed in the sense that neither side wants it.
01:37:33 <b_jonas> it also split to two, then to three, then to much more
01:37:38 <fizzie> There's two potential borders, and there's a bigger and better chunk of land that's in actual dispute, but the way the borders go, if you claim that good bit you can't claim the other bit.
01:37:43 <fizzie> https://en.wikipedia.org/wiki/Bir_Tawil
01:37:50 <b_jonas> and several parts in europe claimed to be the one true roman emperor
01:37:54 <ais523> b_jonas: I don't think "successor" makes sense when it comes to countries
01:38:00 <ais523> especially if the boundaries are different
01:38:08 <b_jonas> ais523: how about for thrones?
01:38:27 <b_jonas> (and even more claimed to be liege to one of those successors)
01:40:13 <ais523> the laws of succession for the Roman Emperor weren't really well-defined, I don't think
01:40:16 <ais523> given how often they got assassinated
01:40:30 <b_jonas> ais523: oh, as for that, there's country top level domain names. .cs and .yu were both eventually deleted some time after the country it was assigned to got split up.
01:40:40 <ais523> generally speaking the job went to anyone who was capable of convincing sufficiently many important people that they held or should hold it
01:41:05 -!- sprocklem has joined.
01:41:05 <ais523> so from a philosophical sense, either nobody is the rightful Roman Emperor, or everyone is
01:41:09 <b_jonas> deleting a country top level domain that is in use seems totally stupid to me from a technical standpoint
01:41:15 <b_jonas> because there can be lots of references to it
01:41:46 <b_jonas> and this actually suggest that if you register domains, it's usually better to register them under a generic domain, because that won't go away for stupid political reasons
01:41:51 <fizzie> If you're currently on terra nullius, do you get a NullPointerException from Locale.getDefault().toString()?
01:42:12 <b_jonas> obviously for some country-level domains like .hu this is obvious anyway because the country top level domain has stupid registration rules anyway
01:42:16 <b_jonas> but for some it might not be
01:42:48 <b_jonas> though of course you can register domains under country top level domains to redirect to your generic domain, for findability
01:43:01 <b_jonas> like how google and amazon buy google.* and amazon.* for every top level domain name
01:43:02 <ais523> some domain registrars have warnings when buying .tv domains because global warming is threatening the entire territory of Tuvalu
01:43:21 <ais523> that could cause major disruption because it's one of the most popular ccTLDs
01:43:26 <b_jonas> ais523: "buying"? haven't they been giving them away for free? ah no wait, that was .tk domains.
01:43:49 <ais523> really, the situation with TLDs nowadays is complete chaos
01:44:26 <b_jonas> well sure, being roman emperor didn't even start out to be hereditary, even in pretense. that came only later.
01:45:10 <b_jonas> and what makes it worse is that all the shortest top level domain names are country ones
01:45:33 <ais523> emperor wasn't originally even an official position, really, just a de facto description for someone who held all the power
01:46:08 <ais523> now the Roman Empire has no power, thus everyone individually holds all of it, thus everyone is the Roman Emperor
01:46:24 <b_jonas> yet the position and legitimacy recursively derived from it (by liege and successor relations) later became pretty important
01:46:59 <ais523> <Wikipedia> Communications in Tuvalu rely on satellite dishes for telephone and internet access. The available bandwidth is only 512 kbit/s uplink, and 1.5 Mbit/s downlink. ← that's a good argument for them selling their domain name, it's not like they could host much locally with that sort of Internet connection
01:48:12 <b_jonas> ais523: the roman emperor itself doesn't have power, but being european royalty does have some power, in the sense that royalty can marry other royalty much easier without getting excluded from their dynasty, and I have the impression that being royalty is a claim that's basically derived from being a worthy successor of the roman emperor or a close liege of one
01:48:48 <b_jonas> ais523: but they don't have to host the domain name servers in the island, do they? they can buy a server abroad
01:49:04 <b_jonas> multiple servers obviously, for redundancy
01:49:56 <b_jonas> Australia used to be infamous of having really bad internet connectivity to the largest component of Earth, but I don't know if that's still true
01:50:59 <ais523> right, it'd make sense for them to host their domains abroad using their own TLD
01:51:31 <b_jonas> and I hope that eventually people get to live outside earth on other planets, in which case we'll have a problem of huge latency in the connections, which won't be fixed even when throughput is increased to good enough, and that will require to improve a lot of internet protocols and infrastructure in difficult ways but often in directions that I like
01:52:44 <b_jonas> the good part is that webpages won't be able to use all that stupid interactive javascript that tries to phone the server home whenever you scroll to the next page;
01:53:06 <b_jonas> the bad part is that it needs a lot of caching on separate planets, which is hard to combine with effective end-to-end cryptography
01:53:39 <b_jonas> also we'll need an incompatible change to the semantics of IRC that allows non-tree topologies
01:53:43 <ais523> even without separate planets this is still a useful infrastructure to think of, for use with things like IPoAC
01:54:25 <b_jonas> ais523: technically it can matter even for the moon, but I don't think we'll have people live there before they live on other planets
01:54:25 <ais523> I don't think IRC would be a good communication medium for a situation where no realtime communication was possible, something like a mailing list would work better
01:54:51 <b_jonas> ais523: I think something close to IRC but without the tree requirement could work
01:54:59 <ais523> the Moon is much easier to reach than Mars is and doesn't seem that much harder to inhabit?
01:55:16 <ais523> so I'd expect it to be the first natural extraterrestrial body to be inhabited
01:55:25 <ais523> (the International Space Station is inhabited and extraterrestrial, but not natural)
01:55:35 <b_jonas> the relay part makes sense, because it allows you to transport each message only once between planets as long as there's a server on each planet, even if multiple people on the planet are going to read the message, and you might not even be able to tell who'll read it when it's sent
01:55:49 <ais523> you're basically describing Usenet at this point
01:55:58 <ais523> which, to be honest, is probably a pretty good technology for something like this
01:56:15 <b_jonas> ais523: it does seem harder to inhabit for longer term. it might be easier for just a very few people to spend a few months in a research station.
01:56:16 <ais523> back when it was invented, latencies were often very large
01:57:23 <b_jonas> ais523: sure, I'm just more familiar with IRC than usenet. I heared that people used email before the internet and continuous connectivity, with servers only occasionally calling each other, and I've no clue how that could possibly have worked, it sounds like throwing a message to the sea in a bottle and hoping it arrives
01:57:53 <b_jonas> though I admit I also don't understand how routing on the current internet could possibly work, it also sounds about as impossible to me, yet by experience it works most of the time.
01:58:07 <ais523> email addresses weren't a single, fixed thing at the time
01:58:16 <ais523> instead of a domain name, you wrote a list of servers to relay the message via
01:58:39 <ais523> (and the names were much simpler, often a few letters, because they were local hostname lookups on the server before rather than being global domain names)
01:59:06 <ais523> many published email addresses from the time started vax! because "vax" was a fairly famous server that most people knew how to reach
01:59:49 <ais523> nowadays, of course, servers that relay any email they see are incredibly rare because they tend to get exploited by spammers, but back then it was very common to relay any email you came across in order to get it closer to its destination, few people were abusing it
01:59:57 <b_jonas> on the plus side, this has left the good heritage to email, where even though now email practically always goes in at most three hops, from client to sender server to receiving server to client, but it still works if the clients are only occasionally connected, and the servers are only connected most of the time (so that there is often a time when both are connected at the same time)
02:01:26 <b_jonas> ais523: right, so now the "direct" tcp connection between the two servers all happens magically behind the scenes by routing, which I still don't understand
02:01:56 -!- b_jonas has quit (Quit: leaving).
02:06:50 -!- ais523 has quit (Quit: quit).
02:11:12 <tswett[m]> So I'm playing with PCjs some more. Turns out that getting the chipset all set up correctly without the BIOS is a big challenge.
02:11:38 <tswett[m]> So, the problem is pretty much the perennial problem: I'm not getting any interrupts from the interrupt controller.
02:13:34 <tswett[m]> I've determined that this is because the interrupt controller's interrupt request register contains 6, indicating that it's aware of interrupt requests numbered 1 and 2; and because of this, any *new* interrupt requests numbered 1 and 2 are being ignored.
02:15:51 <tswett[m]> In real systems, what happens is that since the interrupt controller has interrupt requests, it signals an interrupt to the CPU; then the CPU signals an acknowledgement to the interrupt controller; then this causes the interrupt controller to clear the corresponding bit in the interrupt request register.
02:21:38 <tswett[m]> And it looks like the way that the emulator simulates this is, the interrupt controller calls the updateINTR method on the CPU, which sets the CPU's INTR flag. But the interrupt controller isn't doing that since it's not receiving any *new* interrupt requests.
02:28:09 <tswett[m]> So, here's what I knew a minute ago. In order to acknowledge an interrupt, you have to write a 20h to port 20h.
02:28:19 <tswett[m]> The assembly I was using for that was:
02:28:39 <tswett[m]> In words: "Load the byte 20h into register AH, then output the contents of register AL to port 20h."
02:51:49 <tswett[m]> I was accidentally loading the immediate byte into the wrong register.
03:31:56 <tswett[m]> All right, sweet, I've successfully made something that takes keyboard input and writes corresponding output to the screen.
03:32:33 <tswett[m]> Now for the big challenge: making it so that the characters typed *are* the characters displayed on the screen rather than merely having some vague relationship.
03:40:37 <esowiki> [[User talk:A]] M https://esolangs.org/w/index.php?diff=64537&oldid=64530 * A * (-4164) /* Char language */ Now that Char has a separate page, I will add my recent ideas here.
03:48:50 <esowiki> [[User talk:A]] https://esolangs.org/w/index.php?diff=64538&oldid=64537 * A * (+889) /* Hey stack! */ Work in progress. +Example
03:51:50 <tswett[m]> Also... make it so that stuff shows up on the screen immediately instead of not showing up until you hit the "Halt" button on the emulator.
03:51:57 <tswett[m]> I feel like I had that problem before and I figured out how to fix it.
03:53:17 <esowiki> [[User talk:A]] M https://esolangs.org/w/index.php?diff=64539&oldid=64538 * A * (+360) /* Hey stack! */ + a lot of syntactic sugar
03:54:24 <esowiki> [[User talk:A]] M https://esolangs.org/w/index.php?diff=64540&oldid=64539 * A * (+126) /* Hey stack! */
04:00:06 -!- FreeFull has quit.
04:01:16 <esowiki> [[User talk:A]] M https://esolangs.org/w/index.php?diff=64541&oldid=64540 * A * (+426) /* Hey stack! */ MTC
04:05:40 <esowiki> [[User talk:A]] M https://esolangs.org/w/index.php?diff=64542&oldid=64541 * A * (+57) /* Description of Conway's Cellular Automaton in Hey stack! */
04:06:21 <esowiki> [[User talk:A]] M https://esolangs.org/w/index.php?diff=64543&oldid=64542 * A * (+65) /* Hey stack! */
04:09:29 <esowiki> [[User talk:A]] M https://esolangs.org/w/index.php?diff=64544&oldid=64543 * A * (-105) /* Hey stack! */
04:10:09 <esowiki> [[User talk:A]] M https://esolangs.org/w/index.php?diff=64545&oldid=64544 * A * (+36) /* Example: Greeting an array-based stack(simply an array+cursor) with length 5 with a finite length before you use it */
04:17:17 <esowiki> [[User talk:A]] M https://esolangs.org/w/index.php?diff=64546&oldid=64545 * A * (+66) /* Hey stack! */
04:21:38 <esowiki> [[User talk:A]] M https://esolangs.org/w/index.php?diff=64547&oldid=64546 * A * (+261) /* Hey stack! */
04:27:30 <esowiki> [[User talk:A]] M https://esolangs.org/w/index.php?diff=64548&oldid=64547 * A * (+148) /* Quick reference of a nice little golfing language */
04:30:11 <esowiki> [[User talk:A]] M https://esolangs.org/w/index.php?diff=64549&oldid=64548 * A * (+147) /* DO()/UNTIL()/END() */
04:47:36 -!- sprocklem has quit (Ping timeout: 244 seconds).
05:09:45 -!- sprocklem has joined.
05:51:36 -!- Sgeo__ has joined.
05:55:16 -!- Sgeo_ has quit (Ping timeout: 268 seconds).
05:58:43 <esowiki> [[User talk:A]] M https://esolangs.org/w/index.php?diff=64550&oldid=64549 * A * (-2015) /* Quick reference of a nice little golfing language */ Terrible Solutions for Code Golf problems
06:06:42 -!- atslash has joined.
06:59:42 <esowiki> [[User talk:A]] https://esolangs.org/w/index.php?diff=64551&oldid=64550 * A * (+129) /* Quick reference of a nice little golfing language */
07:20:59 -!- Lord_of_Life has quit (Ping timeout: 268 seconds).
07:23:48 -!- Lord_of_Life has joined.
07:38:14 -!- cpressey has joined.
07:38:35 -!- xkapastel has joined.
07:50:47 <HackEso> cpressey: Welcome to the international hub for esoteric programming language design and deployment! For more information, check out our wiki: <https://esolangs.org/>. (For the other kind of esoterica, try #esoteric on EFnet or DALnet.)
07:51:51 <cpressey> Yes, I'm familiar with this channel, thanks though.
07:52:36 <shachaf> Are you familiar with the tradition of misusing `welcome?
07:53:08 <shachaf> (It's not all that great a tradition now that I think of it.)
07:55:14 <cpressey> No, I'm not familiar with that tradition. Perhaps a few words about it should be added to the `welcome message, to help introduce it to newcomers.
07:56:23 <HackEso> cpressey invented the esolang, the pipe cleaner and the electrical mousse.
07:56:39 <shachaf> Second only to Taneb in inventiveness.
08:00:46 <cpressey> Not all that inventive anymore. Actually, I dunno. It's mostly that I don't have as much time as I'd like, to actually build out the ideas I have.
08:03:10 <cpressey> I bring good news, though: apparently it's recently been established that, if P != NP, then all instances of CSP are either in P, or are NP-complete.
08:03:45 <cpressey> ( https://mathoverflow.net/a/322335 )
08:07:11 -!- b_jonas has joined.
08:07:27 <HackEso> The CIs are a secret society led by David Morgan-Mar, bent on conquering the world from Sydney with webcomics and unsolvable puzzles. They invented Taneb.
08:07:31 <b_jonas> ^ this one is no longer true
08:08:00 <Taneb> They're no longer secret?
08:08:16 <b_jonas> no, they are no longer bent to use puzzles: http://www.irregularwebcomic.net/draakslair/viewtopic.php?p=173108#173108
08:09:13 <b_jonas> ``` perl -pi -e 's/ and unsolvable puzzles//; warn$_' wisdom/ci # http://www.irregularwebcomic.net/draakslair/viewtopic.php?p=173108#173108
08:09:15 <HackEso> The CIs are a secret society led by David Morgan-Mar, bent on conquering the world from Sydney with webcomics. They invented Taneb.
08:15:29 <b_jonas> that does indeed sound fancy
08:15:35 -!- AnotherTest has joined.
08:18:34 -!- b_jonas has quit (Quit: leaving).
08:21:46 -!- tromp has quit (Ping timeout: 276 seconds).
08:26:07 -!- rodgort has quit (Quit: Leaving).
08:27:34 -!- Sgeo_ has joined.
08:29:52 <shachaf> i,i this is also true if P = NP
08:30:11 -!- tromp has joined.
08:30:21 -!- rodgort has joined.
08:30:58 -!- Sgeo__ has quit (Ping timeout: 248 seconds).
08:44:58 <cpressey> shachaf: I thought that when I read it too but now I'm not so sure. The proof relies on the assumption that P!=NP (as does Ladner's theorem that NPI exists at all).
08:46:11 <cpressey> i.e., maybe P=NP but there are instances of CSP that are not even in NP.
08:46:12 <Taneb> I think it's false if P = NP and there's a CSP not in P or NP
08:46:52 <shachaf> cpressey: Hmm, probably I don't actually know what CSP is then.
08:46:53 <cpressey> :) I don't know CSP well enough to know if an instance of CSP not even in NP ought to be surprising or not.
08:47:07 <Taneb> Well, it's at least as surprising as P=NP
08:48:58 <Taneb> (if P=NP, is everything in P "NP-complete"?)
08:49:31 <cpressey> A pretty good intro to CSP that I also found recently: https://web.archive.org/web/20180214194125/https://www.doc.ic.ac.uk/~sgc/teaching/pre2012/v231/lecture15.html
08:49:47 <cpressey> (and was apparently taken down even more recently, thus the wayback machine)
08:51:27 <cpressey> Taneb: if P=NP then P-complete=NP-complete but... I'm sure there are problems in P that aren't P-complete?
08:51:48 <shachaf> cpressey: Based on 15.1 I'm confusil on how this isn't in NP.
08:52:08 <Taneb> cpressey: given an oracle to a problem in P one can solve any other problem in P in polynomial time
08:52:17 <cpressey> (I am using the word "sure" very loosely, above)
08:52:28 <Taneb> Which was my understanding of "-complete"
08:53:08 <shachaf> I would imagine P-complete would use a different reduction.
09:10:33 <cpressey> Okay, yes, from what I see (from 15.1 and from some web searching), CSP is NP-complete. So, on the face of it, yes, it looks like you could make a stronger claim than in that mathoverflow answer: that all instances of CSP are in P or are NP-complete.
09:13:38 <shachaf> Maybe they were using "or" in the colloquial exclusive sense.
09:15:12 <shachaf> I started reading a book about SAT.
09:29:15 -!- wob_jonas has joined.
09:30:32 <wob_jonas> cpressey: no, if P=NP then every P problem is NP-complete, even the most trivial one, because we define NP-completeness using a P powered reduction
09:30:52 <wob_jonas> P-completeness is trickier, that one is defined with more restricted reductions
09:48:57 <cpressey> wob_jonas: Taneb: I think I follow you and I can't disagree. I find it hard to swallow the implications of that, but I already find it hard to swallow the implications of P=NP I guess, so I guess that doesn't change anything.
09:49:22 -!- xkapastel has quit (Quit: Connection closed for inactivity).
09:49:43 <cpressey> Or maybe it's more like, if P=NP, then P-reductions seem almost meaningless
10:33:33 <cpressey> To relate this all back to esolang (kind of), CSP without bounds on the variables values is Turing-complete. I don't remember where I read that though, I thought it was in that tutorial but can't find it now.
10:37:48 <cpressey> Would make sense though, because it's basically just constraint programming at that point.
10:40:28 -!- arseniiv has joined.
10:43:34 -!- Sgeo__ has joined.
10:47:02 -!- Sgeo_ has quit (Ping timeout: 272 seconds).
10:54:19 <wob_jonas> cpressey: can it simulate diophantine equations?
10:58:39 <cpressey> wob_jonas: Yes, it can. Yes, I thought of that too :)
11:00:07 -!- atslash has quit (Ping timeout: 244 seconds).
11:02:00 -!- atslash has joined.
11:08:04 <shachaf> Phase transition in random 3SAT: Empirically, there seems to be a constant k such that random 3SAT instances are satisfiable iff their clause/variable ratio is smaller than k, and k seems to be around 4.27.
11:09:45 <wob_jonas> of course you have to define the distribution of that "random 3SAT" precisely before you can claim that
11:11:06 <shachaf> Yes. There are a few ways to define it but it seems pretty straightforward.
11:12:52 <cpressey> It seems a bit surprising to me, the `iff` part especially
11:13:05 <wob_jonas> well, maybe it is partly surprising depending on what the distribution is like, because if you do it differently, you'll need an asymptotic more like clause**2/variable or something
11:13:24 <myname> cpressey: welcome to randomness
11:13:25 <wob_jonas> but it's not surprising that there's a phase transition if you count the clauses in the right way
11:13:39 <wob_jonas> it's like all those random percolation things
11:13:55 <myname> i mean, obviously you can construct cases that have more clauses and are satisfiable
11:14:03 <myname> just repeat some indefinitely
11:14:14 <shachaf> cpressey: Oh, I didn't mean "iff" in a logical sense.
11:14:41 <shachaf> Only that the probability is low below the threshold and high above it, and the phase transition becomes sharper as the number of variables increases.
11:15:10 <wob_jonas> shachaf: the iff is in a logical sense, but you need a limit to large number of variables and an almost always in there
11:17:07 <wob_jonas> but I agree with cpressey, it's hard to be surprised by phase transitions like this after I've grown up among the crazy probability theorists here that keep studying problems like that
11:18:56 <shachaf> There is also such a constant for 2SAT, with k=1
11:19:40 <shachaf> And I'm now remembering that someone told me about a similar phase transition in random graphs which is probably related.
11:20:34 -!- atslash has quit (Read error: Connection reset by peer).
11:20:35 <wob_jonas> shachaf: yes, there are many such theorems. for random graphs too, but also for various regular grids with each edge deleted independently with a probability
11:20:55 -!- atslash has joined.
11:21:11 -!- mniip has joined.
11:21:19 <wob_jonas> in some cases the phase transfer parameter is a nice round number because of some symmetry argument, in other cases it's just some odd constant that you can't connect to other numbers
13:21:40 <arseniiv> (I wanted to say something but I don’t have anything interesting)
13:33:14 <Taneb> It'll get here when it gets here
13:33:49 <wob_jonas> before or after the CALESYTA results?
13:34:15 -!- john_metcalf has joined.
13:46:13 <HackEso> The IOCCC is the Industrial Ordovician COBOL Conference Circuit. Not to be confused with OIC. See also ioccclist.
13:46:34 -!- Sgeo_ has joined.
13:49:14 -!- arseniiv_ has joined.
13:49:14 -!- arseniiv has quit (Read error: Connection reset by peer).
13:49:35 -!- Sgeo__ has quit (Ping timeout: 244 seconds).
14:03:27 -!- arseniiv_ has changed nick to arseniiv.
14:03:47 <HackEso> ‘Life,’ said Marvin, ‘don't talk to me about life.’
14:05:38 -!- user24 has joined.
14:05:59 <wob_jonas> well, we have https://esolangs.org/wiki/Game_of_Life at least. no specific article for high life.
14:07:48 -!- arseniiv_ has joined.
14:10:03 -!- arseniiv has quit (Ping timeout: 245 seconds).
14:40:10 -!- arseniiv has joined.
14:42:40 -!- arseniiv_ has quit (Ping timeout: 246 seconds).
14:53:28 -!- atslash has quit (Quit: This computer has gone to sleep).
14:54:38 -!- atslash has joined.
15:11:54 <cpressey> fwiw, I read over a gentle(ish) description of Ladner's NP-intermediate construction, and it's totally reasonable to me that you couldn't write it as a CSP.
15:12:29 -!- atslash has quit (Quit: This computer has gone to sleep).
15:18:06 -!- Sgeo__ has joined.
15:21:54 -!- Sgeo_ has quit (Ping timeout: 272 seconds).
15:26:26 <int-e> cpressey: The CSP result feels a bit like the n-SAT dichotomy, in that you have a parameterized family of problem classes, and you can separate them into P-solvable and NP-complete based on the choice of parameter. It's far more interesting and much harder than the n-SAT case though. (For n-SAT I mean: 0-SAT, 1-SAT, and 2-SAT are in P; 3-SAT and beyond are NP-complete.)
15:29:32 -!- wob_jonas has quit (Remote host closed the connection).
15:29:39 <int-e> cpressey: do you have a pointer to that "gentle(ish) description"?
15:30:44 <cpressey> https://iccl.inf.tu-dresden.de/w/images/a/ab/CT2017-Lecture-14-overlay.pdf
15:30:46 <int-e> (I think I like the notion of 0-SAT. It's not even P-complete.)
15:31:12 <cpressey> I'm afraid I have to take off now but I'll try to pop in tomorrow too.
15:31:15 -!- cpressey has quit (Quit: WeeChat 1.4).
15:37:34 -!- Sgeo_ has joined.
15:40:40 -!- Sgeo__ has quit (Ping timeout: 244 seconds).
15:48:51 -!- xkapastel has joined.
16:05:04 -!- Sgeo__ has joined.
16:07:57 -!- Sgeo_ has quit (Ping timeout: 245 seconds).
16:11:21 -!- user24 has quit (Quit: Leaving).
16:49:38 -!- Sgeo__ has quit (Read error: Connection reset by peer).
16:50:02 -!- Sgeo__ has joined.
17:09:11 -!- b_jonas has joined.
17:09:27 <HackEso> damnation//The Damnation was an evil empire of yore, until the dam no longer held and they got flooded. \ 279) <oklopol> i actually do like sucking
17:20:00 <HackEso> luftballon//A Luftballon is an experimental weapon first developed by the German military in 1983 designed to scramble fighter jets, causing chaos and starting wars between their enemies. \ 662) <fizzie> oerjan: Hey, what's your country code for telephonistic dialling from the outside world? <oerjan> fizzie: +47 <fizzie> oerjan: Ooh, you're, like, right next to Sweden there. <fizzie> I... guess you are geographically, too.
17:21:50 -!- Phantom_Hoover has joined.
17:26:58 -!- FreeFull has joined.
17:46:06 -!- mich181189 has quit (Ping timeout: 257 seconds).
17:47:07 <HackEso> outstanding//Outstanding is when someone is upright in an outdoors position, as in "That scarecrow is outstanding in his field". \ 386) * Sgeo is risking massive forest fires <Sgeo> The bacon is worth it
17:47:12 <HackEso> After the London terrorist attacks of 2005, the Underground was completely evacuated. Without air resistance, the trains would go at blazingly fast speeds between the terminals. This is called a vacuum tube. Sadly, current technology doesn't let passengers travel that way.
17:48:27 -!- mich181189 has joined.
18:02:06 -!- atslash has joined.
18:07:37 -!- Sgeo_ has joined.
18:08:18 -!- atslash has quit (Quit: This computer has gone to sleep).
18:10:30 -!- Sgeo__ has quit (Ping timeout: 244 seconds).
18:25:33 -!- Sgeo__ has joined.
18:28:46 -!- Sgeo_ has quit (Ping timeout: 246 seconds).
18:47:05 -!- Sgeo_ has joined.
18:50:26 -!- Sgeo__ has quit (Ping timeout: 258 seconds).
18:51:00 -!- xkapastel has quit (Quit: Connection closed for inactivity).
18:59:52 <HackEso> rm: cannot remove 'wisdom/386': No such file or directory
19:18:25 -!- Lord_of_Life_ has joined.
19:19:56 <HackEso> *poof* * Sgeo is risking massive forest fires <Sgeo> The bacon is worth it
19:21:18 -!- Lord_of_Life has quit (Ping timeout: 245 seconds).
19:21:29 -!- Lord_of_Life_ has changed nick to Lord_of_Life.
19:22:37 -!- \oren\ has joined.
19:23:19 <\oren\> if oiseau is "wazo" then is oieau "wao"?
20:19:26 <b_jonas> \oren\: you haven't been in here for a while. is everything all right?
21:18:05 <quintopia> rebecca black's 22nd birthday was on a Friday last month and i missed the opportunity to tweet about it. y'know, to make people feel old
21:18:54 <rain1> why would one be old if rebecca black is 22
21:24:42 <int-e> I've now read the wikipedia page. I'd like to reiterate my question.
21:32:49 -!- AnotherTest has quit (Ping timeout: 276 seconds).
21:33:54 -!- atslash has joined.
21:34:30 -!- xkapastel has joined.
21:37:35 <quintopia> one would not be old but one might feel old if 2011 feels like yesterday
21:40:18 <int-e> quintopia: you missed the point. you need to pick some event or person that people can actually relate to. :P
21:50:40 <arseniiv> a cubehelix article publication date?
21:51:11 <arseniiv> though I’m not sure if it was 80s, 90s or 00s
21:53:05 <arseniiv> though again who am I to know it, I didn’t see what it was then in the internets, I started crawling them in ≈2007
22:00:07 -!- Sgeo__ has joined.
22:03:14 -!- Sgeo_ has quit (Ping timeout: 248 seconds).
22:12:48 -!- Sgeo has joined.
22:15:13 -!- Sgeo__ has quit (Ping timeout: 246 seconds).
22:48:33 -!- Sgeo_ has joined.
22:51:34 -!- Sgeo has quit (Ping timeout: 244 seconds).
22:54:07 -!- Sgeo has joined.
22:54:58 -!- arseniiv has quit (Ping timeout: 248 seconds).
22:56:18 -!- Sgeo_ has quit (Ping timeout: 245 seconds).
23:39:09 -!- Phantom_Hoover has quit (Read error: Connection reset by peer).
23:58:33 -!- Sgeo_ has joined.