00:01:37 <oren> たていすかんなにらせちとしはきくまのりつさそひこみも
00:02:27 -!- hppavilion[1] has quit (Ping timeout: 246 seconds).
00:04:07 <coppro> besides tateisukannanirasechitoshihakikumanoritsusasuhinimimo
00:04:35 <oren> That is apparently the result of pressing all the letter buttons qwertyuiop[asdfghjklzxcvbnm in kana input mode
00:04:47 -!- tromp has joined.
00:05:17 -!- GoToTell has quit (Read error: Connection reset by peer).
00:05:20 <oren> ゛ぬふあうえおやゆよわほたていすかんなにらせ゜むーちとしはきくまのりれゃつさそひこみもねるめ
00:06:10 <oren> they don't appear to be in any recognisable order
00:07:32 <coppro> so I guess those are kind of grouped, the rest not so much
00:07:38 <coppro> it could be done similar to qwerty?
00:07:50 <coppro> but... hmm, then you wouldn't want all the single vowels on the top row
00:08:48 <coppro> what's the vowel extender?
00:09:07 <coppro> and why does little ya have its own key, but little yo/yu do not?
00:09:34 -!- hppavilion[1] has joined.
00:10:15 <coppro> I like how mu is farther away than the handakuten
00:20:21 -!- hppavilion[1]_ has joined.
00:20:52 -!- hppavilion[1]_ has quit (Read error: Connection reset by peer).
00:22:24 -!- hppavilion[1]_ has joined.
00:26:41 -!- hppavilion[1]_ has quit (Ping timeout: 246 seconds).
00:27:51 <coppro> hppavilion[1]: threading in python is fucking disgusting, yes
00:28:09 <coppro> god help C++ when it gets a networking library (asio)
00:29:40 <pikhq> coppro: Lord, they'd actually manage to make something worse than BSD sockets.
00:29:53 <coppro> pikhq: I've not actually used asio
00:30:15 <pikhq> It's a boost lib. I'm already terrified.
00:30:29 <coppro> my understanding is it's high-power
00:30:40 <coppro> so if you just want to send some packets, look elsewhere
00:31:00 <pikhq> Well there's one thing it has that (sadly) standard BSD sockets doesn't.
00:31:01 <coppro> if you want to write a highly threaded webserver? probably like 6 lines, each 200 characters :P
00:31:16 <pikhq> Reasonably full-featured DNS querying.
00:31:44 <newsham> thread lightly and carry a big lock
00:32:28 <pikhq> The standard BSD socket lib only lets you *sanely* do A and AAAA queries.
00:32:52 <pikhq> Which, while by far the most common case and definitely something that should be supported, is not enough. :)
00:33:26 <coppro> it's going to be kind of nuts once they add coroutine support
00:37:19 -!- tromp has quit (Remote host closed the connection).
00:39:39 -!- tromp has joined.
00:41:29 -!- t3cora18 has changed nick to h0rsep0wer_znc.
00:42:43 -!- h0rsep0wer has quit (Quit: Leaving).
01:03:46 -!- Frooxius has quit (Read error: Connection reset by peer).
01:05:42 -!- Frooxius has joined.
01:10:16 -!- Phantom_Hoover has quit (Read error: Connection reset by peer).
01:25:38 -!- Frooxius has quit (Quit: *bubbles away*).
01:29:24 -!- Frooxius has joined.
01:29:33 <zzo38> How can I get a intercepting HTTPS proxy for Linux?
01:31:47 -!- MDude has changed nick to MDream.
01:37:26 <HackEgo> [wiki] [[Talk:Esoteric Operating System]] http://esolangs.org/w/index.php?diff=43598&oldid=32513 * Hppavilion1 * (+255) /* Move this to make room for a more accurate page */ new section
01:38:42 <hppavilion[1]> Does anybody have any objections to my proposal on that page?
01:38:49 -!- Wallacoloo has left.
01:46:04 <Jafet> zzo38: squid can do that http://wiki.squid-cache.org/Features/SslBump
01:47:44 <hppavilion[1]> Because this would be a (very simple) _graphical_ OS
01:48:25 <oerjan> aka it will never actually happen
01:48:56 <hppavilion[1]> Unless of course we slowly Esoterize a version of Linux.
01:49:31 <hppavilion[1]> The EsOS of my design would be more than just an OS
01:49:59 <hppavilion[1]> It would also include programs like the EsoBrowser which implements the EHTTP
01:50:59 <hppavilion[1]> What do you think of tthe idea of Esoterizing Ubuntu or something, oerjan?
01:51:28 <zzo38> Jafet: How can I tell it then to delete headers?
01:52:15 <Jafet> I don't know, you'll need to look at the documentation
01:52:37 <Jafet> A dove flies past oerjan, shedding a feather.
01:54:21 <Jafet> Note that the ubuntu project already contains an esoteric graphical operating system underneath an esoteric user-space operating system and an esoteric web browser that implements esoteric internet protocols; it's called ubuntu
01:55:27 <hppavilion[1]> Anyone want to design an esoteric internet protocol?
01:56:46 <oerjan> i think that if you can find enough people wanting to do such a thing, go ahead; also i don't think you will.
01:57:49 <Jafet> Apparently debian runs two versions of systemd: one as root and one as user...
01:59:41 <hppavilion[1]> Perhaps we could design an EsOS that need not be implemented?
02:00:12 -!- boily has joined.
02:00:19 * oerjan should stay out of this since he's obviously not going to participate
02:00:42 <lambdabot> fizzie said 8h 8m 34s ago: No, I'm too afraid.
02:01:14 <hppavilion[1]> <lambdabot> walrusman said 6 h 6 m 6 s ago: The password is "aaafkjeoaaefjaeljfaoeiofn"
02:01:25 <boily> @tell fizzie don't be afraid. it's good for you. it's for your own health. embrace. hth. fnord.
02:03:50 * boily should stay out of this :P
02:05:16 <hppavilion[1]> I'm currently taking ideas for an Esoteric Filesystem
02:05:55 <boily> how would you filesoterystem?
02:09:10 <hppavilion[1]> So should I start with the classic folder/file tree or should I scrap all existing logic of filesystems completely and start from scratch?
02:09:12 <oerjan> main verb unnecessary in these parts
02:09:20 <zzo38> Is there a HTTPS proxy that can work only when the browser is opened?
02:09:35 <zzo38> I want to strip out the "Strict-Transport-Security" header
02:10:39 <hppavilion[1]> Perhaps instead of a file tree I could do nested file tables?
02:10:55 <pikhq> I suspect Hacking Team has one.
02:11:40 <hppavilion[1]> So the simplest possible filesystem: One big block of ones and zeroes. I'll call this FS_0
02:12:16 <boily> you're pushing the everything-is-a-file concept a little bit too far there. hth.
02:12:27 <oren> FS1 is what DOS has
02:12:45 <hppavilion[1]> Files are in folders, folders can be in other folders
02:13:06 <hppavilion[1]> I'm going to rename that to FS infinity tree, actually
02:13:29 <pikhq> A directory is a file.
02:13:47 <pikhq> In my universe it is.
02:14:03 <pikhq> And my universe is POSIX.
02:15:27 <hppavilion[1]> Everything is in the same directory with different files: FS list
02:25:51 -!- GeekDude has quit (Quit: {{{}}{{{}}{{}}}{{}}} (www.adiirc.com)).
02:28:20 -!- boily has quit (Quit: RESPIRATORY CHICKEN).
02:49:01 <zzo38> I fixed my browser's HSTS support by changing the name of the header in libxul.so to contain a CRLF pair, so that it is impossible to match.
02:59:18 <zzo38> However, now HPKP doesn't work either, but the testing page also says that if it did work then you cannot bypass it; I do want to enable HPKP but make it bypassable, but to disable HSTS entirely.
03:00:38 <Sgeo> Why not bypassable HSTS (and not just "type in http:// or navigate there") rather than no HSTS?
03:00:39 -!- MDude has joined.
03:03:05 -!- MDream has quit (Ping timeout: 244 seconds).
03:08:35 <zzo38> HSTS does not solve anything.
03:10:56 <Sgeo> It solves MITMer directing you to http:// of a site you visited before then incercepting, when the site in fact supports HTTPS
03:12:59 <zzo38> If you can see it is http:// then you can see what it is; make user-defined redirections if you want otherwise. If there are problem with cookies or whatever, well, my opinion no cookies and login and so on should cross protocols; HTTP and HTTPS should each have their own set of cookies.
03:13:16 <zzo38> They should also use different cookies if the port number changes, too.
03:14:18 <Sgeo> But cookies do cross protocols, sadly it's too late to change that. And even if they didn't, you have stronger phishing risks
03:15:56 <Sgeo> (Although I suppose that would require a user savvy enough to check the domain name but unsavvy enough to trust http. Although those users might exist because there are sites out there that are actually http-only)
03:17:40 <zzo38> You can set cookies to be HTTPS only, although a client can nevertheless be designed to implement cookies in this more strict way; ignore the security flag on cookies and instead always use whatever protocol the cookie is sent with.
03:19:57 -!- oerjan has quit (Quit: leaving).
03:26:39 -!- Wright has quit (Read error: Connection reset by peer).
03:37:21 -!- hppavilion[1] has quit (Ping timeout: 246 seconds).
03:38:23 <zzo38> NoScript Lite doesn't work for me
03:45:29 -!- variable has quit (Ping timeout: 252 seconds).
03:47:58 -!- password2 has joined.
03:54:15 <oren> My site is http-only because I don't store visitor's personal information anyway
03:54:49 <oren> the only sites that nees https are those that have login and store sensitive information
03:55:07 -!- MDude has quit (Ping timeout: 256 seconds).
04:00:48 -!- Wright has joined.
04:13:08 <zzo38> I don't store visitor's personal information either, and even if I did, HTTPS would not stop me from storing them
04:15:00 <ais523> zzo38: the idea's not to stop you storing them, but to stop other people taking copies of them / altering them while they're in transit to or from your site
04:19:17 <zzo38> Yes, I know, but still HTTPS won't stop the site you are connecting to from storing your information.
04:21:52 -!- MDude has joined.
04:25:18 <zzo38> It does stop intercepting, however it can be useful to intercept it yourself, so it should allow HTTPS over a plain HTTP proxy and the proxy would handle secure connections instead; you can run this proxy in the same LAN or even on the same computer for additional security, although a remote one might improve the speed if you have a slow computer.
04:38:49 <zzo38> I have made all sorts of hacks into Firefox in various places: the filesystem permissions, the SQLite database schemas, the about:config, the userChrome,js, and even hexediting the program binaries.
04:42:49 -!- llue has quit (Read error: Connection reset by peer).
04:43:13 -!- llue has joined.
04:45:32 -!- copumpkin has quit (Quit: My MacBook Pro has gone to sleep. ZZZzzz…).
04:47:02 -!- Wright has quit (Read error: Connection reset by peer).
04:47:09 -!- Wright_ has joined.
04:51:47 -!- Wright_ has quit (Ping timeout: 256 seconds).
04:52:48 -!- password2 has left ("Leaving").
04:54:30 -!- hppavilion[1] has joined.
04:57:19 <HackEgo> [wiki] [[Talk:Esoteric Operating System]] http://esolangs.org/w/index.php?diff=43599&oldid=43598 * Hppavilion1 * (+293) /* EsOS file system */ new section
05:17:28 <myname> zzo38: what hexediting?
05:27:46 <zzo38> I changed the string "strict-transport-security" in libxul.so by replacing two of the consecutive letters with a CRLF pair instead.
05:31:22 <zzo38> That seems the only way to disable HSTS
05:31:56 <ais523> zzo38: has HSTS caused problems for you? or do you just want to disable it out of principle?
05:33:42 <zzo38> I want to disable it because it is a bad idea.
05:34:10 <lifthrasiir> well, it is a bad idea made necessary for a bad existing design.
05:35:38 <zzo38> HPKP would make some sense if the user could override it. HSTS is just stupid though.
05:39:18 -!- GoToTell has joined.
06:03:00 <zzo38> hppavilion[1]: Files *must* be in folders, only one deep, is what TI-92 does
06:03:27 <zzo38> (Except for local variables)
06:04:03 <hppavilion[1]> Unfortunately, if there is any compilation of GUI project ideas, it's REALLY hard to find
06:05:04 <zzo38> I don't know. What kind of GUI are you using?
06:05:50 <zzo38> AmigaMML IDE, if you like that kind of stuff, is one kind of project I suppose
06:07:06 <zzo38> Yes, a GUI of a OS can be another idea
06:07:37 <hppavilion[1]> When I have a website, I'll format it properly and publish it
06:29:07 -!- Walpurgisnacht has joined.
06:45:15 <b_jonas> hppavilion[1]: do you mean an esoteric one?
06:46:16 <b_jonas> hppavilion[1]: ok, do you want an easy one or a hard one or a very hard one?
06:48:15 <b_jonas> hppavilion[1]: very hard one: a PDF viewer with interface that doesn't get in your way: controllable with keyboard and mouse such that you can hide all the toolbars and menus and such so they don't cover parts of the screen, preferably based on the xpdf backend as forked by okular, and most importantly: EASILY ADJUSTABLE GAMMA SETTING, per pdf, and possibly separately for text and images
06:48:39 <b_jonas> hppavilion[1]: possibly hack the missing parts of this into okular
06:50:34 <b_jonas> hard one: a raster image viewer with interface that doesn't get in your way, controllable with keyboard and mouse such that you can hide blah blah, adjustable gamma setting, and adjustable setting for how zoom is antialiased so you can get nearest neighbour or some more expensive interpolation,
06:51:16 <b_jonas> with a thumbnail view for all images, and if I specifically ask, it should be able to rotate photos by 90 degree and save the rotations in _a separate file_ rather than in the photos.
06:51:31 <b_jonas> image viewer with zooming should be a standard widget I think
06:52:09 <b_jonas> easy one? dunno, let me think. the problem with easy ones is that I just don't use GUIs for those.
06:52:38 <hppavilion[1]> It's a beginner's final project as a beginner sometimes
06:53:57 <b_jonas> oh btw, for both the image viewer and pdf viewer, I'd like the mouse panning speed customizable as a signed floating point number.
06:54:07 <b_jonas> because some people actually like reverse pan.
06:58:13 <b_jonas> I should write this down so I won't forget any feature when I write my letter to Father Christmas.
07:00:30 -!- Walpurgisnacht has quit (Quit: hhhh).
07:06:52 <hppavilion[1]> https://docs.google.com/document/d/1V7Ewci21-4dNp3hzORuzljOpTJeim8Vb1-qnHhLzugk/edit?usp=sharing
07:07:18 <hppavilion[1]> I trust the people here enough to share that link for anyone to submit ideas, doable and otherwise.
07:24:30 -!- hppavilion[1] has quit (Ping timeout: 246 seconds).
07:34:36 -!- GoToTell has quit (Quit: HydraIRC -> http://www.hydrairc.com <- In tests, 0x09 out of 0x0A l33t h4x0rz prefer it :)).
07:39:47 -!- hppavilion[1] has joined.
07:42:55 <zzo38> Why do you use Google Docs, it isn't very good
07:43:20 -!- hilquias has joined.
07:43:27 <hppavilion[1]> Is there a better free collaborative online document editor I should know about?
07:44:07 <zzo38> MediaWiki I suppose
07:44:25 <hppavilion[1]> But this is just a draft I'm going to put on my website
07:45:13 <zzo38> Or use a FTP server
07:46:38 <hppavilion[1]> Well FTP isn't a very good way to transmit text like this, IMHO
07:48:50 <hppavilion[1]> When it can be made on a website, make it a website
08:08:15 -!- hppavilion[1] has quit (Ping timeout: 246 seconds).
08:44:25 -!- white_bear has joined.
08:49:27 -!- Frooxius has quit (Read error: Connection reset by peer).
08:49:54 -!- Frooxius has joined.
08:54:05 -!- llue has quit (Quit: That's what she said).
08:54:17 <fowl> google docs is awesome
08:54:45 -!- white_bear has quit (Quit: leaving).
08:58:13 -!- lleu has joined.
09:00:41 <ais523> my main issue with Google Docs is that it plays rather fast and loose with the HTML it uses internally, especially if you copy/paste from Word
09:02:08 -!- white_bear has joined.
09:02:16 <fowl> Inter-suite compatibility is something I've never seen
09:02:42 <mroman> Can't we just use HTML for documents?
09:04:43 <ais523> HTML isn't restrictive enough
09:04:49 <ais523> and thus allows way too many ways to write the same thing
09:07:55 -!- TodPunk has quit (Ping timeout: 244 seconds).
09:08:59 -!- AnotherTest has joined.
09:09:40 <mroman> as long as it's rendered correctly why would that bother me?
09:10:02 -!- TodPunk has joined.
09:11:26 <ais523> mroman: because editing (especially collaborative editing) can easily make the formatting inconsistent in a way that's very difficult to fi
09:11:36 <ais523> (I know all this from personal experience)
09:11:39 <mroman> You mean WYISWYG editing?
09:11:53 <ais523> mroman: I mean editing via Google Docs specifically
09:12:10 <ais523> but that's due to the general issues of WYSIWYG editing of HTML
09:12:26 <ais523> especially if it allows inline styles but not separate CSS
09:22:28 -!- ais523 has quit.
09:24:18 <zzo38> That is why you can use MediaWiki or even just plain text, if HTML isn't restrictive enough
09:25:09 <zzo38> For many kind of documents plain text works just fine, although for collaborative work it still help to have automatic changelogs
09:26:09 <mroman> From an architectural standpoint.... I think pyongyang looks amazing.
09:27:18 <mroman> at least the parts that are shown in this video.
09:33:46 <mroman> zzo38: How can I do bold blinking text in plaintext?
09:33:56 <mroman> You know, to get people's attention?
09:34:59 <mroman> Micosoft Word totally has Text Effects!
09:35:04 <mroman> (Or had. I'm not sure if it still does.)
09:36:05 <mroman> They removed it in 2013.
09:37:20 <mroman> `? word (Microsoft Word) was a text-editor for animated texts but not anymore.
09:37:21 <HackEgo> word (Microsoft Word) was a text-editor for animated texts but not anymore.? ¯\(°_o)/¯
09:37:27 <mroman> `learn word (Microsoft Word) was a text-editor for animated texts but not anymore.
09:37:31 <HackEgo> Learned 'word': word (Microsoft Word) was a text-editor for animated texts but not anymore.
09:39:51 <b_jonas> mroman: shouldn't that mention "winword" too somewhere? that's what it was called by default for some time
09:40:08 <b_jonas> I renamed it, but only so that I can run two versions at the same time.
09:49:42 <zzo38> mroman: In some case you don't need such thing, anyways if you print it out it is hard to blink.
09:50:35 <b_jonas> zzo38: What? but users do expect the printed versions to be animated too.
09:50:59 <b_jonas> at least if the internet stories about typical stupid users are true.
09:51:15 <zzo38> I am aware of that.
09:51:52 <Taneb> zzo38, what is the URL for your gopher server?
10:01:09 <mroman> well it's not microsoft's fault printers suck at printing animations.
10:01:31 <mroman> Word was just decades ahead of printers.
10:02:05 <mroman> I'm certain if printers were advanced enough Microsoft had done everything to make animations work on prints as well.
10:02:18 <mroman> You could have spinning word-arts!
10:03:39 <mroman> You could have had power-point effects on OHP!
10:03:54 -!- x10A94 has joined.
10:03:56 <mroman> with plastic foils that support animations.
10:04:17 <mroman> How do you call that thing in english?
10:05:16 <b_jonas> overhead projector transparent celluloid foil slide or whatever
10:05:52 -!- x1365C has joined.
10:06:02 <b_jonas> apparently "transparencies"
10:06:21 <b_jonas> https://en.wikipedia.org/wiki/Transparency_%28projection%29
10:07:09 <fizzie> I think they're called "cels" in the context of animation.
10:07:26 <fizzie> "A cel, short for celluloid, is a transparent sheet on which objects are drawn or painted for traditional, hand-drawn animation. Actual celluloid (consisting of cellulose nitrate and camphor) was used during the first half of the 20th century, but since it was flammable and dimensionally unstable it was largely replaced by cellulose acetate. With the advent of computer-assisted animation ...
10:07:32 <fizzie> ... production, the use of cels has been all but abandoned in major productions. Disney studios stopped using cels in 1990 when Computer Animation Production System (CAPS) replaced this element in their animation process."
10:08:31 <mroman> overlay transparencies
10:08:34 <fizzie> @tell boily It's not that, I'm afraid it'd alter fungot's PERSONALITY, because it's a whole different thing I'm using now to make them models.
10:08:35 <fungot> fizzie: did you *just* implement call/ cc in terms of lambda abstraction and application, too
10:08:38 <mroman> and move the ones that need moving
10:08:44 <mroman> there's a nice documentary about that on youtube
10:09:16 <fungot> fizzie: at least that's how i find it curious that this seemingly nonstandard extension is the default size?
10:10:00 -!- x10A94 has quit (Ping timeout: 264 seconds).
10:12:02 -!- x1365C has quit (Read error: Connection reset by peer).
10:13:06 -!- ais523 has joined.
10:13:51 -!- x10A94 has joined.
10:14:31 -!- Frooxius has quit (Quit: *bubbles away*).
10:21:08 -!- boily has joined.
10:21:56 -!- lambdabot has quit (Ping timeout: 240 seconds).
10:26:36 -!- lambdabot has joined.
10:30:09 <mroman> fungot: Can you play the harp?
10:30:09 <fungot> mroman: really cool would be some defined database types... e.g. id like to mind.
10:30:32 <mroman> Meanwhile germany wants 48h weeks .
10:32:36 <Taneb> That's like only two days
10:32:45 <Taneb> What are they going to do with the other five!?
10:33:07 <ais523> Taneb: sleeping, eating, commute, leisure, I guess
10:33:27 <ais523> and sleeping again to make five
10:33:34 <Taneb> Oh, a 48h work week?
10:34:01 <Taneb> That's 9 to 5 six days a week!
10:34:31 <ais523> AIUI it's a common scam to try to persuade your workers that doing that's normal
10:35:07 <ais523> the EU has maximum work hours for this sort of reason (but unfortunately there are too many loopholes involved, e.g. in the UK it's possible to opt out, and businesses put you under a lot of pressure to do so)
10:35:09 <fizzie> A six-day work week *was* normal, right?
10:37:32 <ais523> although then it also tended to be limited by available daylight
10:37:43 <ais523> so eight hours every non-Sunday every month would be unusual
10:50:06 -!- FreeFull has quit (Quit: BBL).
10:51:09 -!- Hoolootwo has joined.
10:51:39 -!- J_A_Work has joined.
10:57:54 <mroman> "Inappropriate key specification"
11:01:40 -!- ais523 has quit (Quit: lunch).
11:05:01 -!- zadock has joined.
11:06:38 <mroman> java.security.spec.InvalidKeySpecException: Inappropriate key specification: IOException: ObjectIdentifier() -- data isn't an object ID (tag = -96)
11:06:43 <mroman> well.. that's a little bit more detail :D
11:07:20 <boily> mrelloman. you're doing nasty java stuff. this is inappropriate.
11:07:36 -!- FreeFull has joined.
11:08:08 <mroman> java.security.spec.InvalidKeySpecException: Inappropriate key specification: IOException: DerInputStream.getLength(): lengthTag=127, too big.
11:12:01 <lambdabot> fizzie said 1h 3m 26s ago: It's not that, I'm afraid it'd alter fungot's PERSONALITY, because it's a whole different thing I'm using now to make them models.
11:12:14 <boily> @tell fizzie *gasp*!
11:14:37 -!- zadock has quit (Quit: Leaving).
11:14:43 <mroman> I'm doing a p2p message networking system
11:14:50 <mroman> where you have to exchange public keys and stuff
11:14:55 <mroman> to verify each others identity
11:15:00 <mroman> every message is signed ;)
11:15:11 <mroman> right now pulbic keys can't be loaded :D
11:16:49 -!- TieSoul has joined.
11:21:18 -!- TieSoul has quit (Ping timeout: 260 seconds).
11:24:32 -!- boily has quit (Quit: WHETSTONE CHICKEN).
11:25:27 <mroman> although this probably consumes a lot of CPU
11:25:31 <mroman> since it signs every single package :D
11:34:14 <mroman> java.security.SignatureException: invalid encoding for signature
11:34:49 -!- ais523 has joined.
11:36:48 -!- TieSoul has joined.
11:43:15 <mroman> ok. finally everything works
11:52:44 -!- Phantom_Hoover has joined.
11:53:23 <fizzie> mroman: So I can close all the bugs in our company bug tracker as fixed, then?
11:55:57 -!- Phantom__Hoover has joined.
11:56:06 <ais523> I had to pay for something over the phone recently, because there were technical problems with paying over the Internet (which I suspect were caused by the connection dropping at exactly the wrong moment)
11:56:22 <ais523> I asked them to send a receipt by email so that I had a record of the transaction
11:56:51 <ais523> they did so: it's a scanned-in copy of the physical paper receipt that was produced by their credit card reader
11:56:56 <ais523> somehow I find this hilarious
11:58:05 <fizzie> No wooden table, though?
11:58:30 <nortti> Sgeo: no idea, if you mean http://theevilliouschronicles.wikia.com/wiki/Seven_Crimes_and_Punishments_(story)
11:58:38 -!- Phantom_Hoover has quit (Ping timeout: 246 seconds).
11:58:46 <fizzie> I paid something with a debit card on a ferry, and they used one of those old-fashioned things where you put the card in and drag the handle back and worth and it copies the embossed numbers.
11:58:56 -!- white_bear has quit (Ping timeout: 244 seconds).
11:59:03 <fizzie> Then it took 2-3 weeks to show up on the account.
11:59:18 <fizzie> I'm a bit surprised they accepted a debit card for that.
12:00:25 <ais523> fizzie: because it might not have had money in?
12:00:44 <fizzie> Well, that, and just in general, I associate those things with credit cards explicitly.
12:00:46 <ais523> I think most financial transactions rely on the principle of "most people are honest, and most people who aren't can be caught and sued for the money"
12:00:57 <olsner> I think debit cards are typically also credit cards
12:01:03 <ais523> in the UK I've known them to be used for cheque guarantee cards as well
12:01:25 -!- ais523 has quit (Read error: No route to host).
12:01:45 <fizzie> And the precursor to a debit card, a "Visa Electron" I used to have in Finland, was unusable in many places, because it was (at least supposedly) strictly online-only.
12:01:46 -!- ais523 has joined.
12:02:01 <fizzie> [Repeating:] And the precursor to a debit card, a "Visa Electron" I used to have in Finland, was unusable in many places, because it was (at least supposedly) strictly online-only.
12:02:10 <ais523> olsner: at least in the UK, debit cards typically can't be used for purchases on credit unless you have an arranged overdraft, which you usually don't
12:02:36 <ais523> fizzie: I used to have a Visa Electron; it was mostly only intended to be used in cash machines
12:02:47 <ais523> although some websites did accept it
12:02:54 <fizzie> Well. Also at POS terminals in shops.
12:02:58 <ais523> shops didn't (at least the only one I tried to use it at didn't), so I had to find a cash machine nearby
12:03:16 <fizzie> At least in Finland "normal" shops did/do accept it.
12:03:30 <fizzie> Although it's been somewhat phased out in favor of Visa Debit.
12:03:38 <ais523> Visa Debit seems to be accepted everywhere
12:03:52 <fizzie> I think they still have Electron cards for young people etc.
12:04:25 <fizzie> I have a debit/credit card too, and sometimes the people don't ask, just assume debit or credit.
12:04:45 <fizzie> And/or ignore what you tell them.
12:05:50 <mroman> just make sure you add a "won't fix" label
12:05:57 <mroman> or "works for me you dipshit"
12:06:30 <fizzie> We have a "won't fix, intended behavior" status.
12:06:43 <mroman> good, then abuse the hell out of that.
12:07:15 <mroman> I have a "not a bug, feature" label
12:07:24 <mroman> and "not a bug, marketing reasons"
12:07:45 <mroman> "crashes randomly every week" ==> "not a bug, marketing reasons"
12:08:06 <mroman> (aka otherwise nobody would buy the next version coming out next month)
12:12:30 -!- white_bear has joined.
12:36:08 -!- TodPunk has quit (Ping timeout: 244 seconds).
12:38:12 -!- TodPunk has joined.
13:00:42 -!- mauris has joined.
13:01:21 -!- copumpkin has joined.
13:01:36 <oren> I have a debit card and a credit card, which are both green, with a Visa symbol, a +Plus and an Interac sysmbol on them. it is confusing, because they are only distinguished by a tiny word that says "credit" on one and "debit" on the other.
13:08:39 <b_jonas> oren: yeah. it's even worse about virtual cards.
13:09:12 <mroman> It's more fun if you have a reddit card too
13:24:13 -!- copumpkin has quit (Quit: My MacBook Pro has gone to sleep. ZZZzzz…).
13:28:11 <mroman> fungot: Do you have a credit card?
13:28:11 <fungot> mroman: it was in here a few weeks while waiting for the ehlo answer, whether he supports fnord and so i do
13:30:19 <mroman> Mail servers with fnord support?
13:31:57 <fizzie> fungot: What's the RFC for that?
13:31:57 <fungot> fizzie: short-term versus long-term " currently in php but i could install run?' using fnord.
13:40:33 -!- J_A_Work has quit (Quit: J_A_Work).
13:57:10 -!- Sgeo_ has joined.
13:57:26 -!- `^_^v has joined.
13:59:02 <mroman> If I were a malware developper
13:59:07 <mroman> I'd use images to transport malware
13:59:19 <mroman> and then extract the code dynamically through javascript from the image
13:59:21 -!- Sgeo has quit (Ping timeout: 255 seconds).
14:00:52 -!- Patashu has quit (Ping timeout: 246 seconds).
14:00:54 <Hoolootwo> the problem is that javascript is supposed to not run malware
14:00:58 -!- nycs has joined.
14:02:55 <mroman> Any suffiecently advanced malware is indistinguishable from non-malware".
14:03:52 <ais523> mroman: it might be distingushable from obvious non-malware, though
14:03:54 -!- `^_^v has quit (Ping timeout: 265 seconds).
14:04:39 <mroman> Any sufficiently advanced program is malware.
14:04:42 <mroman> that's probably better.
14:05:01 -!- zadock has joined.
14:05:40 <Hoolootwo> yeah that's truer than I'm willing to admit
14:06:42 -!- oerjan has joined.
14:09:26 <b_jonas> Q: What is perl useful for? A: Writing a syntax highliter for perl code using regular expressions.
14:12:11 <ais523> Perl is really hard to syntax-highlight
14:12:25 <ais523> so far Kate holds the record out of programs I've tried, but even it screws up sometims
14:12:57 <ais523> (or, well, Perl is TC to syntax-highlight because you can write code that runs the instant it's parsed and can change the parsing of the rest of the file as an effect)
14:13:35 <b_jonas> yes, or even code that's read twice and parsed differently in the two parses
14:14:00 <int-e> ooh, polyglot hilighting.
14:14:08 <b_jonas> I don't care, because I don't syntax highlight anything, and I think syntax highlighting is a bad idea in genreal.
14:14:09 <APic> mroman: Check out the „American Fuzzy Lops“
14:14:20 <APic> mroman: Semiautomatic JPeg-Fuzzing
14:14:21 <b_jonas> If you can't read the code without help from the computer, then the code is written unclear.
14:14:42 <mroman> b_jonas: I disagree with that.
14:14:45 <b_jonas> It's the unclear code that should be fixed, not the highlighting.
14:15:02 <mroman> visual cues help you read code faster
14:15:11 <b_jonas> I've heared some counter-arguments, like how it helps catch errors when you write the code, or stuff.
14:15:19 <mroman> for example highlighting return statements makes you see them much better
14:15:34 <mroman> and yes, it shows you syntax errors before you hit "compile"
14:15:51 <mroman> also auto-complete really increases productivity
14:16:02 <mroman> but that's not really that much related to syntax highlighting itself
14:16:24 <b_jonas> Auto-complete is just horrible. It makes people type a word once, with a typo, and then copy the typo everywhere in the code, and they never read it.
14:16:25 <mroman> but I like editors that can highlight and format code automatically
14:16:35 <b_jonas> Then I have to replace the typo globally.
14:16:36 <mroman> b_jonas: That does happen, yes :)
14:16:46 <mroman> that's why eclipse has a "rename"
14:16:50 <mroman> which renames all references as well.
14:17:13 <b_jonas> As for "before you compile", I think that's distracting, when I write code I want to concentrate on the code I write, I can fix typos in a later pass.
14:17:26 <b_jonas> I don't want the editor interrupting me with the typos that I'll find when compiling.
14:17:31 <mroman> but otherwise you'd need to have the documentation of all the things open somewhere
14:17:37 <mroman> and then switch between eclipse and the documentation
14:17:46 <mroman> and that is just annoying.
14:18:19 <b_jonas> Well sure, requesting documentation (or matching parenthesis) with a keystroke is fine, just don't distract me with that info while I'm writing stuff.
14:18:43 <mroman> matching parenthesis sucks :)
14:18:48 <mroman> I hate when eclipse does that
14:19:12 <mroman> I'd prefer if it had a command that "closes open parentheses"
14:19:31 <mroman> so you can write things like Math.abs(Math.min(a,b and then hit that command
14:19:49 <mroman> when you use auto-complete sometimes eclipse will automatically insert arguments
14:20:00 <mroman> you wrote Math.abs(foo);
14:20:07 <mroman> then you want to change that to Math.sqrt
14:20:12 <mroman> and it will result in Math.sqrt(p)foo;
14:20:20 <mroman> that's somewhat annoying sometimes.
14:20:42 <b_jonas> I need a command that jumps to the matching parenthesis to the one under the cursor, bound to a keystroke. Luckily, most sane editors have this already.
14:24:29 <mroman> ToDO before I die: Write a LISP compiler
14:24:33 <mroman> I still haven't done that
14:27:32 <mroman> Alright, I'll do it now.
14:29:26 -!- nycs has quit (Ping timeout: 240 seconds).
14:29:40 -!- nycs has joined.
14:29:57 <HackEgo> poland/Połąńd is a European country. Its population consists of two main ethnicities, the North Połes and the South Połes.
14:30:08 <b_jonas> mroman: can you tell a bit more about the goals of what lisp compiler you want?
14:32:08 <mroman> I don't know. I just wanted to write a compiler for some restricted LISP-Dialect
14:36:50 -!- SopaXT has joined.
14:47:23 <mroman> http://codepad.org/RDtFNClF
14:52:31 <mroman> do I need a garbage collector?
14:52:44 <fizzie> No, just let it pile up.
14:52:51 <mroman> No I meant something else
14:53:10 <fizzie> Require the programmer to call 'free' for every 'cons'.
14:53:38 <mroman> (cons (foo) (list 1 2)) you can free the (list 1 2) after the cons call
14:54:08 <shachaf> Well, I guess you can free whatever you want whenever you want.
14:54:37 <mroman> oh wait that's not cons
14:55:07 <mroman> Reference counting is usually easiest
14:55:11 <mroman> but it doesn't work in all cases
14:55:24 <mroman> the question is: Do you need a case where it doesn't work?
14:55:42 <Taneb> Cyclic data structures, right?
14:56:05 <shachaf> Declare them undefined behavior. Done.
14:56:06 <mroman> That's what I'm pointing at, yes.
14:56:28 <fizzie> If it's good enough for Perl, it would be good enough for you.
14:56:30 <mroman> they are not required for turing completeness?
14:56:41 <mroman> and most data structures arent cyclig
14:56:46 <shachaf> Turing completeness is scow.
14:57:01 <shachaf> You should make your lisp thing not be Turing complete.
14:57:02 <Taneb> mroman, they are not required for turing completeness
14:57:06 <shachaf> Turing complete things aren't reasonable.
14:58:05 <mroman> lisp doesn't have references wimre?
14:58:13 <mroman> mutating a list always results in a new one
14:58:23 <mroman> if you always copy everything you don't need a gc anyway
14:58:48 <Taneb> If you always copy everything you need a gc all the more
14:58:55 <Taneb> Because you build up garbage very quickly
14:59:01 <mroman> you copy and free immediately
14:59:11 <mroman> append would just create a new list
14:59:20 <mroman> copy the elements of the two lists into that new list
14:59:25 <mroman> and then frees the two original lists
14:59:29 <mroman> then you don't need no gc
14:59:36 <mroman> it'll be horribly slow though
14:59:40 <mroman> I don't want horribly slow
14:59:45 <mroman> alright. Reference counting it is.
15:01:02 <mroman> Lisp is usually copy-on-write
15:01:46 <shachaf> It'll be correct and simple.
15:01:48 <mroman> i'll guess I make append and append!
15:02:12 <mroman> that's even more horrible
15:02:14 <shachaf> You were going to make stop-the-world reference counting.
15:02:36 <shachaf> GC and reference counting are the same thing.
15:02:48 <shachaf> http://researcher.watson.ibm.com/researcher/files/us-bacon/Bacon04Unified.pdf
15:03:11 <mroman> use reference counting
15:03:17 <mroman> and then from time to time the GC with cycle detection kicks in
15:03:21 <shachaf> Except they find the greatest and least fixed points.
15:03:23 <shachaf> http://www.pvk.ca/Blog/2012/02/19/fixed-points-and-strike-mandates/
15:04:02 <shachaf> A moment ago you were talking about copying the entire heap with every instruction.
15:04:23 <shachaf> (Which is pretty much what a GC would do, except less frequently.)
15:09:28 <oerjan> <mroman> lisp doesn't have references wimre? <-- some lisps have mutable cons cells
15:11:14 <fizzie> I guess making it really unpronouncable is one way of warning people away from it.
15:11:39 <fizzie> Scheme would call that set-car! with the scary exclamation point.
15:12:39 <ais523> does it create references or copies of its argument into the cons cell in question?
15:12:51 <ais523> suddenly, it starts mattering
15:13:01 <ais523> (with references you could place a cons cell inside itself)
15:13:14 <shachaf> <mroman> i'll guess I make append and append!
15:13:22 <shachaf> it sounds like this is already the plan
15:13:39 <fizzie> I'm not sure, the plan seemed to be changing every second.
15:13:40 <Phantom__Hoover> append! is for when you're really enthusiastic about your appending
15:13:44 <shachaf> RPLACA and NCONC are such good names
15:14:31 <shachaf> every common lisp name is great
15:15:05 <fizzie> ais523: http://sprunge.us/eGfb
15:15:29 <fizzie> I think some implementations would pretty-print that a little better.
15:16:40 <ais523> for some reason, it never crossed my mind that it'd give you an infinite string of open parentheses
15:16:50 <ais523> it's like the opposite of Lisp in a way
15:17:18 <lambdabot> "\"\\\"\\\\\\\"\\\\\\\\\\\\\\\"\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\"\\\\\\\\\\\\\...
15:17:40 <fizzie> It makes sense. You start at t; it's a cons cell, so you print '(' and then look at the car; that's t, a cons cell, so you print '(' and then look at the car; ...
15:18:25 <ais523> fizzie: oh, it makes perfect sense
15:18:26 -!- Frooxius has joined.
15:18:33 <ais523> I just hadn't realised that that's what would have happened until I saw it
15:19:38 <fizzie> I don't have a mzscheme (what was it now, racket?) implementation here, I think it should've printed something like #0=(#0# 2).
15:19:52 <fizzie> Or something like that, I forget the exact syntax.
15:21:49 -!- zadock has quit (Quit: Leaving).
15:21:55 <Taneb> I am not sure that Racket has set-cons!
15:22:17 <fizzie> It's one of those Schemes with immutable cons cells.
15:22:30 <fizzie> I think it had something like 'mcons' for mutable cells, and some pair of setters.
15:22:49 <fizzie> mcons, mcar, mcdr, set-mcar! and set-mcdr!, apparently.
15:22:55 <fizzie> And mpair? as the predicate.
15:23:19 <fizzie> That's not how it was when it was still called MzScheme, though.
15:24:09 <shachaf> I think it switched to immutable cons by default before renaming to Racket.
15:24:40 <fizzie> It also means you have to jump through hoops to get cyclic lists out of the immutable pairs.
15:25:08 <Taneb> > (define t (mcons 1 2))
15:25:25 <fizzie> Huh, I got it (almost) right.
15:25:47 <shachaf> Taneb: That was an interesting error for the first thing.
15:25:58 <shachaf> Looks like a lambdabot bug.
15:26:03 <lambdabot> Not in scope: ‘define’ Not in scope: ‘mcons’
15:26:03 <lambdabot> ‘cons’ (imported from Control.Lens),
15:26:08 <shachaf> I think I've seen it before.
15:26:16 <fizzie> I think it would've been #0=(#0# . 2) before the mcons thing, actually; forgot it was a pair, not a list.
15:28:15 <oerjan> <fizzie> That's not how it was when it was still called MzScheme, though. <-- i thought racket was PLT scheme?
15:29:24 <shachaf> I thought it was called Dr. Scheme.
15:29:24 -!- rodgort has quit (K-Lined).
15:29:35 <fizzie> oerjan: It was MzScheme before that.
15:29:45 <fizzie> MzScheme for the engine, and Dr. Scheme for the GUI, or some-such.
15:29:54 <fizzie> Then it went PLT, then it went Racket. Or that's at least my recollection.
15:30:40 <fizzie> Or maybe it's a branding thing.
15:30:47 <fizzie> The company was PLT from the get-go, anyway.
15:30:57 <fizzie> But "mzscheme" was the command-line thing you ran.
15:30:58 <b_jonas> fizzie: I thought it was PLT for the fancy system, MzScheme for the important scheme interpreter part,
15:31:03 <shachaf> I think MzScheme was part of PLT Scheme.
15:31:05 <b_jonas> and then it became Racket as the new branding
15:31:22 <shachaf> In the end it's all the same software.
15:32:01 <fizzie> DrScheme was maybe the teaching-oriented programming environment.
15:32:32 <b_jonas> anyway, the important part is MzScheme
15:32:33 <fizzie> The old PLT pages say "MzScheme is the name of the core virtual machine for PLT Scheme" and "DrScheme is an interactive, integrated, graphical programming environment that is included with PLT Scheme".
15:32:39 <oerjan> <shachaf> what now <-- based on my view from another channel, someone possibly just klined ro*
15:33:02 <fizzie> And MrEd is the thing DrScheme is built on, right.
15:33:06 <oerjan> i'm sure you can confirm or deny that
15:33:19 <shachaf> oerjan: i was talking about the many names of plt scheme
15:33:45 <fizzie> http://plt-scheme.org/software/ lists the components.
15:33:47 <shachaf> But I see several ro* who aren't klined.
15:33:52 <int-e> 17:29:24 --- rodgort has quit [K-Lined]
15:33:54 <fizzie> (If you dismiss the Racket ads, anyway.)
15:34:16 <int-e> (is there a reason not to mention that name?)
15:34:35 <shachaf> int-e: I think oerjan is saying that everyone matching /^ro.*/ was klined.
15:35:25 <b_jonas> maybe it's a botnet connecting with all ro* names at first
15:35:31 <shachaf> also i didn't know oerjan was in any other channels
15:35:55 <fizzie> Yes, let's focus on what's important. Such a lack of channel loyalty.
15:36:02 <oerjan> shachaf: i'm in ##nomic, where Roujo just got klined at the same time
15:36:15 <b_jonas> I've seen stuff like that, they connect twenty bots all with similar names, they start to spam, people ban them with a pattern and rejoice thinking they won, bot owner modifies pattern and connects twenty new bots, repeat
15:36:16 <fowl> I see an ro* in a channel who isnt klined
15:36:28 <ais523> definitely ro* people who weren't k-lined
15:36:33 <oerjan> and my nomic loyalty is far older than my esolang loyalty, so there. although it's somewhat lapsed.
15:36:34 -!- x1365C has joined.
15:37:19 <b_jonas> I agree, I see ro* nicks connected
15:37:24 <b_jonas> so not all of them are k-lined
15:37:31 <fowl> The one i see is not logged in either
15:37:43 <fowl> Er authenticated
15:39:50 -!- rodgort has joined.
15:40:26 <b_jonas> Let me note that there's at least one freenode staffer logged in right now whose nick matches ro*
15:40:36 -!- x10A94 has quit (Ping timeout: 264 seconds).
15:41:22 <oerjan> y'all people cannot convince me with your stupid evidence
15:42:01 -!- roejan has joined.
15:42:44 <shachaf> now i can't tell which overlord is evil
15:43:02 <oerjan> shachaf: i may have to do a ban_jos if this escalates
15:43:05 <ais523> shachaf: roejan has a suspiciously similar hostname to b_jonas
15:43:15 <oerjan> ais523: all ahead of you
15:43:28 <shachaf> oerjan's pun was better, though.
15:43:36 <ais523> oerjan: that's not fair, I had to type my sentence
15:43:44 <ais523> also I wasn't making a pun
15:43:52 <ais523> and oerjan is one of the few people who's allowed to make /good/ puns
15:43:56 <ais523> normally they're supposed to be bad
15:44:11 <shachaf> oerjan's pun didn't need to be very good to be better than no pun at all.
15:44:29 <shachaf> b_jonas: you'll rue the month you decided to impersonate oerjan
15:44:45 <shachaf> (that's an example of a worse pun)
15:45:24 <ais523> well, if you get a reputation for everything you say being a pun
15:45:32 <ais523> eventually you can just say a non sequitur
15:45:37 <ais523> and everyone assumes it's a pun but it's so subtle they can't spot it
15:45:53 <roejan> shachaf: sorry, "roen" was already in use, I didn't have much cohice
15:46:09 <shachaf> ais523: oerjan's puns may be good, but are they fun?
15:46:22 <b_jonas> ais523: http://www.xkcd.com/559/
15:46:38 <ais523> shachaf: that's the first pun in rot-13 I've seen
15:47:56 <oerjan> <ais523> normally they're supposed to be bad <-- wait, what, this changes everything!
15:48:14 -!- GoToTell has joined.
15:48:57 <oerjan> if i say enough non sequiturs ais523 will sure quit
15:49:20 <ais523> that isn't a non sequitur though
15:49:24 <ais523> also I'm planning to quit soon anyway
15:50:47 <oerjan> shachaf: you are so easily overwhelmed
15:50:57 <shachaf> oerjan: i wouldn't say i'm overwhelmed
15:50:59 <oerjan> finally Roujo returned
15:51:07 <shachaf> http://gatherer.wizards.com/Pages/Card/Details.aspx?multiverseid=378429
15:51:49 -!- x1365C has quit (Read error: Connection reset by peer).
15:52:19 <oerjan> `` rot13 rue # i don't think i'm getting this
15:52:24 -!- x10A94 has joined.
15:52:44 <shachaf> oerjan: now i can't tell whether that's a hidden pun
15:53:25 <ais523> oerjan: if it is in fact us trolling you rather than vice versa, you might be looking at the wrong line
15:54:06 -!- roejan has quit (Ping timeout: 240 seconds).
15:54:51 <oerjan> now i'm relieved, i thought shachaf actually meant that literally.
15:55:46 <oerjan> or to quote the ancients, furrfu
15:58:30 -!- heroux has quit (Ping timeout: 240 seconds).
15:58:34 <GoToTell> no, but it shows up in certain places. sheesh.
15:58:49 * mauris notices scary concentrations of belgians
15:59:21 <oerjan> shachaf: well but "asheesh" didn't ring any bell...
15:59:35 <mauris> (hallo|bonjour|guten tag), GoToTell!
15:59:40 <shachaf> oerjan: it's a p. common name
15:59:52 <shachaf> https://en.wikipedia.org/wiki/Ashish -- apparently this variant is mentioned in India's national anthem
16:00:04 <shachaf> also that page says "it's" instead of "its"
16:00:13 <shachaf> but i'm trying not to tell you about those things anymore
16:00:33 <mauris> yeah. it's good looking at people's IRC hostnames and deducing that they live a 2 hour drive away, tops
16:00:33 -!- heroux has joined.
16:00:36 <oerjan> shachaf: ic, it's not common where i've been previously acquainted
16:00:54 <ais523> mauris: how long does it take to drive to the UK from where you are?
16:00:54 <shachaf> I don't know how common it is.
16:01:01 <ais523> I'm actually confused trying to work it out
16:01:11 <ais523> partly because I suck at north European geography
16:01:18 <ais523> and partly because I'm not quite sure how to count the ferry/tunnel
16:03:45 <mauris> shorter than i thought: https://goo.gl/maps/khNJf
16:04:01 <shachaf> ais523: Did you publish your thesis?
16:04:23 <ais523> shachaf: I actually sent the finished manuscript to the printers today
16:04:35 <ais523> they'll have the printed versions finished on Friday
16:05:03 <ais523> I'll submit the electronic versions then; I don't know when they'll be printed
16:05:08 -!- heroux has quit (Ping timeout: 246 seconds).
16:05:25 <fizzie> They'll print them, scan them, and put the scans online.
16:05:36 <ais523> fizzie: nah, they were sensible and asked me for a PDF
16:05:45 <mauris> wow, cool stuff, what's the thesis about?
16:05:51 <ais523> I have one with hyperlinks and clickable internal cross-references and accessibility tagging and the like
16:05:55 <ais523> mauris: finite-state type systems
16:06:07 -!- heroux has joined.
16:06:28 <fizzie> ais523: I don't have one like that, because our thesis series template didn't want to play nice with hyperref. But maybe I can fix that later.
16:06:33 <ais523> i.e. languages in which programs provably use only finite memory
16:06:49 <ais523> fizzie: I made my own thesis series template; our regulations just say "such and such text X/Y distance from the margins"
16:07:14 <ais523> took me a while of messing around with minipages to get that right
16:07:26 <fizzie> We have a fully-fledged thing with lots of crufty stuff.
16:07:53 <fizzie> I'm sure it's not mandatory, but it seemed like the sensible choice.
16:07:54 <ais523> pretty much our entire process assumes you're using Word, without actively stating it anywhere
16:08:00 <ais523> and being open for other editors in theory
16:08:10 <ais523> when I do use a different editor (LyX, in my case), it gets a little confusing
16:08:41 <fizzie> There are only four begin/end minipage pairs in the .cls file.
16:09:57 <ais523> wow this is a major bug: https://github.com/dotnet/coreclr/issues/1296
16:10:04 <ais523> broken TCO in .NET 4.6
16:10:13 <ais523> meaning that parameter values can change silently, and only in release builds
16:10:27 <ais523> I'm not sure I understand the theory of doing all debugging without optimization, then using optimization only for a release build
16:10:53 <ais523> because bugs that are only exposed by optimization then won't be caught by your tests
16:11:12 <mauris> that does sound, not ideal
16:13:06 <shachaf> mauris: 3.5 hours is not all that far
16:13:41 <MDude> I would have thought debugging would be done last, and optimization only saved for after all other non-debugging changes.
16:14:07 <ais523> I actually like to do test builds at high optimization levels if I can
16:14:24 <ais523> because you get better warnings, and UB is more likely to break in a visible way
16:15:05 <fizzie> I would have thought it'd be obvious that the build you test (at least for some largeish fraction of testing) is the build you'll be shipping.
16:15:41 <ais523> so many people test at -O0 and ship at -O2 or -O3
16:15:49 <Jafet> Well, there's debugging, there's testing, and then there's end-user testing.
16:15:57 <ais523> in NH4 I test at -Og on the Linux side, and I test at release settings on the Windows side
16:16:04 <MDude> Yeah. I just meant that if you're going to do optimization later, it'd make sense to still have debugging later.
16:16:04 <Jafet> I suppose people might confuse those things.
16:16:12 <ais523> (my release settings include -g, to increase the chance I get useful bug reports)
16:16:30 <MDude> But I don't think every software development team does everything in the same kind sof stages.
16:17:40 <MDude> Also, what does Uwe Boll have to do with this and how can code make him visibly break?
16:19:24 <MDude> Yeah I just have no idea what I'm talking about then.
16:20:12 <MDude> And alaos ocnfusing manual and automatic optimization.
16:20:41 -!- ais523 has quit.
16:21:04 <Jafet> I suppose UB could summon Uwe Boll, but only if you're a really bad person.
16:21:11 -!- ais523 has joined.
16:22:41 -!- ais523 has quit (Client Quit).
16:22:54 -!- ais523 has joined.
16:23:19 <oerjan> <MDude> Also, what does Uwe Boll have to do with this and how can code make him visibly break? <-- i'm sorry i don't get the pun twh
16:23:24 -!- atrapado has joined.
16:23:40 <MDude> I just wans't sure what UB stood for.
16:23:56 <ais523> oh, UB = undefined behaviour
16:24:07 <ais523> it's a pretty commonly seen phrase among C programmers, also in #esoteric
16:24:50 <Jafet> How many other languages use the term (or an equivalent term)?
16:25:42 <ais523> all the languages which have the concept use it, I think
16:25:50 <ais523> but many languages intentionally avoid the concept
16:25:57 <ais523> http://esolangs.org/wiki/Undefined_behavior
16:26:12 <shachaf> UB could also be blue-black.
16:26:39 <oerjan> istr there are some people who wouldn't mind Uwe Boll visibly breaking
16:27:34 <ais523> anyway, I'm surprised that Uwe Boll comes up in conversation sufficiently often to be worth abbreviating
16:28:23 <shachaf> ais523: Do you know if there are statistics about color choices in limited M:tG games?
16:28:41 <shachaf> In particular the variance of different color combinations in drafts, and that sort of thing.
16:28:43 <oerjan> hm people with only two words in their name probably are less likely to get an abbreviation.
16:28:49 <shachaf> Or maybe data to measure these things from?
16:28:56 <Jafet> Well, UB is special in some sense because once UB ever appears (even in your future, not just present, I think), you can no longer rely on anything working ever again. The same could be said about undefined behaviour, too.
16:29:21 <ais523> shachaf: I know that the playtesters do them internally
16:29:28 <ais523> however, color balance in drafts is not normally a useful measure
16:29:38 <ais523> because it's exactly balanced because every card has to be picked by someone
16:30:06 <ais523> you could check color balance of decks that are actually played, but even then draft's self-correcting because worse colours tend to be more open
16:32:45 -!- Wright has joined.
16:32:47 <oerjan> `le/rn uwe boll/Uwe Boll is the undefined behavior of cinematography.
16:34:24 <shachaf> ais523: I'd like a large data set of real mtg limited games to look at, but I doubt a good one is easily obtainable.
16:35:07 <ais523> shachaf: apparently Wizards have asked people with access to that sort of data (like StarCityGames) to not make it public, in order to slow down the rate at which formats are broken
16:36:15 <shachaf> ais523: The situation with sealed is pretty different from draft, and would be interesting to know too.
16:36:55 <ais523> shachaf: this is actually why most Wizards internal playtests are sealed
16:37:08 <ais523> I'm not sure if anyone collects that data, though; most large sealed tournaments are casual
16:40:13 <shachaf> ais523: I'm sure mtgo collects it.
16:40:50 <ais523> mtgo's incompetent enough that maybe it doesn't
16:41:29 <shachaf> All that data would be very helpful to them in designing new sets, though, I imagine.
16:42:33 -!- oerjan has quit (Quit: leaving).
16:44:44 <ais523> shachaf: bear in mind that this is the game that failed during its most important tournament of the year, and lost all its data about the state of the tournament at that point in the process
16:45:33 -!- white_bear has quit (Quit: leaving).
16:50:14 -!- ais523 has quit.
16:53:28 -!- atrapa has joined.
16:54:56 -!- atrapado has quit (Ping timeout: 272 seconds).
16:58:24 -!- atrapa has changed nick to atrapado.
17:01:09 -!- x10A94 has quit (Read error: Connection reset by peer).
17:04:58 -!- SopaXT has quit (Ping timeout: 246 seconds).
17:07:48 <HackEgo> U+270A RAISED FIST \ UTF-8: e2 9c 8a UTF-16BE: 270a Decimal: ✊ \ ✊ \ Category: So (Symbol, Other) \ Bidi: ON (Other Neutrals) \ \ U+1F44A FISTED HAND SIGN \ UTF-8: f0 9f 91 8a UTF-16BE: d83ddc4a Decimal: 👊 \ 👊 \ Category: So (Symbol, Other) \ Bidi: ON (Other Neutrals)
17:08:11 <HackEgo> tar/The command you're looking for is probably either tar -xavkf or tar -cavf
17:10:22 <fowl> I didnt know wisdoms had to be true
17:11:18 -!- SopaXT has joined.
17:14:09 <HackEgo> nethack/<Jessicatz> you play too much nethack when: you look down both sides of the corridor, start to sweat and then realize you're looking at your email address
17:19:06 <zzo38> The command is "tar c", "tar t", or "tar x", although there are other options too
17:21:48 <zzo38> (Those three are the only options I have used though)
17:23:26 <b_jonas> some useful ones are: f = filename of tarball; v = print files as extracted or compressed, or print more than filenames for listing; z = pack to gzip; j = pack to bzip2; J = pack to xz; plus some more complicated ones for when you want to control list of files, I think they're -I listfile --null --no-recurse
17:25:00 <b_jonas> these days I try to copy files with rsync -tve "sudo -u" someuser:source dest instead of sudo -u someuser tar c source | tar xvC dest
17:25:27 <b_jonas> oh yeah, another important switch I use is: C = filenames except for name of tarball is relative to this directory
17:25:53 <zzo38> You can use pipes and do not need most of the commands it provides.
17:26:29 <zzo38> If I want the files inside of the archive in a different directory I can just switch to that directory.
17:26:58 <b_jonas> zzo38: sure, and you can (cd foo && make) instead of make -C foo
17:27:03 <b_jonas> but it's still convenient enough that I use it a lot
17:27:50 <b_jonas> also, if foo is a directory you can't even execute to, compare (sudo -u someuser tar cC foo .) to (sudo -u someuser sh -c 'cd foo && tar c .')
17:28:12 <b_jonas> switch is there for a good reason
17:28:46 <b_jonas> basically I think cd should be a convenience thing, and you can use absolute paths if you want most of the time. this doesn't work with tar:
17:28:55 <mauris> you play too much nethack when: that nethack joke becomes profoundly unfunny for you :(
17:29:37 <b_jonas> because it doesn't (usually) rewrite filenames between the file system and the tarball, so if you want a file to be named ./bar in the tarball, you have to cd to its directory or else use the -C option (which does that)
17:29:51 <zzo38> Some programs do not need any command-line arguments at all; AmigaMML currently ignores any it may be given.
17:31:02 <shachaf> zzo38: If a program doesn't need command-line arguments at all, I prefer that it gives an error if it gets any.
17:31:18 <zzo38> shachaf: Yes, that might be better, but the current version does not.
17:31:19 <b_jonas> shachaf: except for true and false and : which shouldn't do that
17:31:30 <b_jonas> shachaf: there's the contentious question on what true --help should do
17:31:48 <shachaf> zzo38: I've done a lot of "program input" instead of "cat input | program" where I wish the program had failed.
17:32:04 <zzo38> b_jonas: As far as I am concerned, it should do nothing
17:32:05 <b_jonas> and also what false --help
17:32:07 -!- SopaXT has quit (Remote host closed the connection).
17:32:20 <zzo38> b_jonas: And false --help should fail.
17:33:04 <shachaf> I don't care about true --help and false --help
17:33:10 <shachaf> I slightly care about echo --help
17:33:24 <b_jonas> there's an underlying problem that people aren't even sure what foo --help should do in general: should it exit with success or failure status, and should it print help to stdout or stderr (or depending on isatty results)
17:33:26 <mauris> bash is a bit of a hack, huh
17:33:50 -!- SopaXT has joined.
17:34:07 <zzo38> For help you can just use a man page
17:34:17 <b_jonas> in the end, implementations agree in that true --help should exit success, and false --help should exit failure, but they differ in what they print
17:34:48 <zzo38> b_jonas: Well, yes that is one way too I suppose
17:35:02 <b_jonas> zzo38: man page? ok, how about true --version
17:35:12 <b_jonas> or does that give no useful information because who cares about the version of true?
17:35:27 <zzo38> b_jonas: That program doesn't need a version number
17:36:08 <b_jonas> zzo38: note that --version customarily prints the license information for programs where just invoking it without argument doesn't,
17:36:24 <b_jonas> and some licenses may require that the binary should be able to give a notice about license informatoin
17:36:33 <b_jonas> so that you know that it uses code by the reagents of whatever
17:37:00 <b_jonas> but there's still a case for saying that screw it, true and false can ignore it as long as the other programs in coreutils or other builtins in bash do tell you that info
17:37:14 <b_jonas> or you can just make /bin/true and /bin/false public domain
17:38:05 <b_jonas> echo is different because it already must handle at least _some_ switches
17:38:18 -!- sebbu has quit (Ping timeout: 250 seconds).
17:38:32 <zzo38> I suppose yes that it can be useful for AmigaMML to display error in case any unrecognized switches are given, since even though there are no recognized one now, in future it is possible that later version might add switches to output debugging information for use with a IDE, or whatever
17:38:36 <b_jonas> plus I think echo genuinely has to be different between the shell version and the /bin/echo or something? I don't follow, it's complicated
17:39:01 <b_jonas> zzo38: it could display a non-fatal warning
17:39:50 <zzo38> That could be a possible too I suppose
17:41:16 <zzo38> Currently though what it does is, it reads MML code from stdin, writes out a MOD or XM to stdout, and displays the number of rows emitted per channel on stderr. In case of error it will display an error message to stderr, write nothing to stdout, and a nonzero exit code.
17:43:34 <zzo38> To create a IDE for AmigaMML, I am not exactly sure which options would be useful to add
17:53:57 -!- `^_^ has joined.
17:56:24 -!- nycs has quit (Ping timeout: 244 seconds).
18:11:10 -!- SopaXT has quit (Read error: Connection reset by peer).
18:15:08 -!- FreeFull has quit.
18:16:13 -!- `^_^v has joined.
18:17:39 -!- `^_^ has quit (Ping timeout: 255 seconds).
18:20:35 -!- FreeFull has joined.
18:20:43 -!- copumpkin has joined.
18:41:54 -!- `^_^v has quit (Quit: This computer has gone to sleep).
18:42:56 -!- `^_^v has joined.
18:53:32 -!- zzo38 has quit (Remote host closed the connection).
19:02:50 -!- Phantom__Hoover has quit (Ping timeout: 246 seconds).
19:06:07 -!- Phantom_Hoover has joined.
19:10:21 -!- Phantom_Hoover has quit (Ping timeout: 244 seconds).
19:12:41 <mauris> the sams book too! are those the ones that promise to teach you C++ exponentially faster over time, and we're down to 10 minutes
19:13:02 <mauris> http://www.amazon.com/Sams-Teach-Yourself-Minutes-Edition/dp/0672324253 mmm
19:14:32 <Slereah__> Why would I want to learn C++ in any amount of time, though
19:15:25 -!- hppavilion[1] has joined.
19:16:02 -!- Phantom_Hoover has joined.
19:26:28 -!- henriqueleng has joined.
19:55:15 <HackEgo> [wiki] [[LOLCODE]] http://esolangs.org/w/index.php?diff=43600&oldid=42122 * 70.185.102.142 * (-101) /* Keywords */
19:58:31 <HackEgo> [wiki] [[Doorspace]] http://esolangs.org/w/index.php?diff=43601&oldid=40970 * 174.236.84.154 * (-1) /* Introduction */
20:03:52 -!- TieSoul has quit (Read error: Connection reset by peer).
20:10:10 -!- GoToTell has quit (Quit: HydraIRC -> http://www.hydrairc.com <- In tests, 0x09 out of 0x0A l33t h4x0rz prefer it :)).
20:22:08 <Melvar> Today’s result: In Idris, every Monad is also a monadic value!
20:30:34 <Melvar> ( do x <- List; y <- Stream; return $ Either x y
20:30:34 <idris-bot> \i => Either (List i) (Stream i) : Type -> Type
20:39:42 -!- hppavilion[1] has quit (Ping timeout: 246 seconds).
20:49:28 <mauris> @tell fizzie what does finnish "liemme" mean?
20:50:10 <fizzie> mauris: It's kind of hard. I think it might be a first-person plural potential case of the verb "be".
20:50:31 <nortti> mauris: something like "we probably are"
20:50:45 <fizzie> Yes, or "we might be".
20:51:31 <mauris> aha. google translate was being useless about it for some reason, but that makes sense!
20:51:42 <fizzie> The "be" verb has a rather nonstandard potential form.
20:51:46 <mauris> how does anyone handle finnish though.
20:52:11 <fizzie> I think "lienemme" would also be a valid way to say it, and in fact that's what's the first-person plural potential in the table at https://en.wiktionary.org/wiki/olla#Conjugation
20:52:35 <fizzie> (But there's a French wiktionary page that uses "liemme" for the same thing.)
20:53:30 <fizzie> The potential mood for verbs is overall pretty rarely used.
20:53:55 <fizzie> Unless you're being deliberately fancy, or just naturally so.
20:54:47 <mauris> oh, is it like an archaic/poetic thing? this is from värttinä lyrics, so that'd make sense
20:55:10 <fizzie> Well, a little bit. It's not *that* archaic, you can still see it in a newspaper and it doesn't necessarily feel weird.
20:57:29 <fizzie> But you'd certainly see it written more often than spoken.
20:57:46 <fizzie> Er, I guess you wouldn't normally *see* it spoken ever, but...
20:59:03 <fizzie> In related news, the bus this morning had a sign that said approximately: "Do not speak to or obstruct the driver's vision while --", and I couldn't help wondering how you'd speak to the driver's vision.
21:03:20 <Hoolootwo> perhaps the driver has synesthesia
21:09:02 <coppro> Melvar: you've made me see things I can't unsee
21:10:35 -!- sebbu has joined.
21:10:59 <Melvar> ( do x <- List; y <- Either x; return (x, y)
21:10:59 <idris-bot> \i => (List i, Either (List i) i) : Type -> (Type, Type)
21:11:08 -!- sebbu has quit (Changing host).
21:11:08 -!- sebbu has joined.
21:11:36 <Melvar> ( do x <- List; y <- Either x; return $ the Type (x, y)
21:11:36 <idris-bot> \i => (List i, Either (List i) i) : Type -> Type
21:12:33 <int-e> > map (do x <- (+1); y <- (+2); return (x*y)) [0..]
21:12:34 <lambdabot> [2,6,12,20,30,42,56,72,90,110,132,156,182,210,240,272,306,342,380,420,462,50...
21:15:33 <int-e> coppro: is the reader monad really so much worse on the type level than on the value level?
21:23:24 -!- J_A_Work has joined.
21:23:24 -!- J_A_Work has quit (Client Quit).
21:26:10 -!- Alcest has quit (Ping timeout: 256 seconds).
21:27:44 -!- Phantom__Hoover has joined.
21:29:41 -!- Phantom_Hoover has quit (Ping timeout: 256 seconds).
21:31:04 <HackEgo> [wiki] [[Talk:Tangle bracket language]] http://esolangs.org/w/index.php?diff=43602&oldid=43565 * Rottytooth * (+235) added question
21:33:35 <int-e> oh will you look at that, there was an actual GG comic last Friday...
21:35:05 -!- Alcest has joined.
21:46:36 -!- boily has joined.
21:50:59 -!- Frooxius has quit (Read error: Connection reset by peer).
21:51:56 -!- Frooxius has joined.
21:54:32 -!- `^_^v has quit (Quit: This computer has gone to sleep).
21:56:05 <lambdabot> CYUL 272130Z 16013KT 15SM BKN040CB BKN240 27/21 A2991 RMK CB5CI2 WSHFT 2115 CB E SLP126 DENSITY ALT 1500FT
21:57:25 <boily> rain rain rain, it shall raaaaain ♪
21:57:58 <lambdabot> LOWI 272150Z AUTO VRB02KT 9999 FEW090 16/13 Q1012
21:59:01 <int-e> great summer weather. (max. 25 today, somewhat rainy)
22:00:16 <boily> int-ello. is it generally humid in Austria?
22:00:46 -!- mauris_ has joined.
22:00:58 <int-e> I don't know about Austria in general. Innsbruck is actually fairly dry overall. I welcome the rain.
22:01:53 -!- mauris has quit (Ping timeout: 244 seconds).
22:02:55 <int-e> (I don't like heat very much.)
22:03:11 -!- oerjan has joined.
22:03:39 <boily> same about not supporting heat.
22:04:49 <oerjan> we only support 0 K, for best efficiency
22:05:12 -!- AnotherTest has quit (Ping timeout: 264 seconds).
22:06:01 -!- Frooxius has quit (Read error: Connection reset by peer).
22:06:45 -!- oren has quit (Ping timeout: 265 seconds).
22:07:30 <int-e> hmm, 904mm rain/year ... actually that's more than I thought.
22:07:35 <Melvar> oerjan: Today’s result: In Idris, every Monad is also a monadic value!
22:08:21 <Melvar> ( do x <- List; y <- Either x; return $ the Type (x, y)
22:08:21 <idris-bot> \i => (List i, Either (List i) i) : Type -> Type
22:08:29 -!- Frooxius has joined.
22:09:15 <Melvar> Unwrapped reader monad, because they must have type (Type -> Type).
22:10:34 <Melvar> Sadly, coppro now hates me for showing this.
22:10:47 <oerjan> ( do x <- List; y <- Either x; return $ Left []
22:11:29 <oerjan> ( do x <- List; y <- Either x; return $ the y (Left [])
22:11:29 <idris-bot> (input):1:43:When checking argument value to function Prelude.Basics.the:
22:11:29 <idris-bot> Either (List elem) b (Type of Left [])
22:12:02 -!- odor has joined.
22:12:05 <oerjan> ( do x <- List; y <- Either x; return $ the (y x) (Left [])
22:12:05 <idris-bot> Can't disambiguate since no alternative is valid:
22:12:05 <idris-bot> Effects.>>=, BotPrelude.LiftEq.>>=, Prelude.Monad.>>=
22:12:37 <Melvar> y is a lambda arg of type Type there.
22:13:02 <oerjan> wait why didn't the y (Left []) work then
22:13:21 <Melvar> You can’t tell what it might be. It’s just a lambda arg that gets something passed in later.
22:13:29 <int-e> oerjan: not the (y (Left [])) ?
22:13:48 <oerjan> int-e: um no? the takes two arguments
22:15:08 <oerjan> ( :t do x <- List; y <- Either x; return $ Left []
22:16:01 <Melvar> I still don’t know what that one is on about, but apparently it’s talking about an implicit argument that gets quantified wrong somewhere.
22:16:41 <Melvar> ( :t do x <- List; y <- Either x; return $ Left {b=Type} []
22:17:17 <Melvar> ( :t do x <- List; y <- Either x; return $ Left {b=Type} (the (List Int) [])
22:17:18 <idris-bot> return (Left (the (List Int) [])) : Type -> Either (List Int) Type
22:17:55 <Melvar> ( do x <- List; y <- Either x; return $ Left {b=Type} (the (List Int) [])
22:17:55 <idris-bot> \i => Left [] : Type -> Either (List Int) Type
22:18:56 <oerjan> ( (do x <- List; y <- Either x; return $ the y (Left [])) Bool
22:18:56 <idris-bot> (input):1:57:Can't disambiguate since no alternative is valid:
22:18:56 <idris-bot> Effects.>>=, BotPrelude.LiftEq.>>=, Prelude.Monad.>>=
22:19:41 * oerjan ambles back to haskell
22:19:47 <Melvar> I told you there’s nothing known about y because it’s just a lambda arg.
22:20:28 <oerjan> well i thought maybe giving it the argument would solve that
22:20:49 <Melvar> No, it’s just not well-typed in the first place.
22:21:32 -!- Frooxius has quit (Read error: Connection reset by peer).
22:22:26 <Taneb> oerjan, that is exactly the same type
22:22:29 <Taneb> Just more colourful
22:22:57 <Melvar> > :t \y => return $ the y (Left [])
22:22:58 <lambdabot> <hint>:1:1: parse error on input ‘:’
22:23:02 <Melvar> ( :t \y => return $ the y (Left [])
22:23:02 <idris-bot> Can't disambiguate name: Effects.return, Prelude.Monad.return
22:23:03 <oerjan> nevertheless, it answered my question after a little consideration
22:23:38 <Melvar> ( :t \y => Monad.return {m = (\a => Type -> a)} $ the y (Left [])
22:23:57 -!- Frooxius has joined.
22:23:57 -!- Frooxius has quit (Client Quit).
22:24:22 -!- Frooxius has joined.
22:26:55 -!- idris-bot has quit (Read error: Connection reset by peer).
22:27:10 <Melvar> No clue what that fail was about.
22:27:20 -!- idris-bot has joined.
22:27:39 <Melvar> As though it lost the network but neither side noticed.
22:28:41 <Melvar> ( :t \y => Monad.return {m = (\a => Type -> a)} $ the y (Left [])
22:28:41 <idris-bot> (input):1:53:When checking argument value to function Prelude.Basics.the:
22:28:41 <idris-bot> Either (List elem) b (Type of Left [])
22:32:11 -!- hilquias has quit (Remote host closed the connection).
22:43:20 -!- mauris_ has quit (Ping timeout: 246 seconds).
22:48:45 <lambdabot> CYUL 272200Z 17009KT 15SM BKN040TCU BKN240 27/20 A2990 RMK TCU6CI1 SLP125 DENSITY ALT 1500FT
22:49:09 * boily rubs his hands together and cackles evily
22:49:17 -!- atrapado has quit (Quit: Leaving).
22:49:36 <lambdabot> EGLL 272220Z 24014KT 9999 SCT022 SCT024 15/11 Q1004
22:49:55 <fizzie> This is starting to feel like a Finnish summer.
22:52:31 <boily> EG is not Finland. did you illegally bring weather across states lines?
22:58:09 -!- boily has quit (Quit: TOWERING CHICKEN).
22:58:31 -!- idris-bot has quit (Ping timeout: 260 seconds).
23:12:55 <lambdabot> ENVA 272250Z 22005KT 190V250 9999 FEW030 09/07 Q1001 RMK WIND 670FT 24009KT
23:30:06 -!- Patashu has joined.
23:48:19 -!- Phantom_Hoover has joined.
23:49:15 -!- idris-bot has joined.
23:49:46 -!- Phantom__Hoover has quit (Ping timeout: 240 seconds).