| --- Log opened Thu Aug 04 00:00:29 2016 |
| 00:15 | | Kindamoody is now known as Kindamoody|busy |
| 00:15 | | himi [sjjf@Nightstar-v37cpe.internode.on.net] has quit [Connection closed] |
| 00:22 | | Netsplit Kakrafoon.Nightstar.Net <-> Krikkit.Nightstar.Net quits: @PinkFreud |
| 00:27 | | Netsplit over, joins: @PinkFreud |
| 00:48 | | Derakon[AFK] is now known as Derakon |
| 00:50 | | catalyst [catalyst@Nightstar-bt5k4h.81.in-addr.arpa] has quit [[NS] Quit: Leaving] |
| 01:21 | | catadroid [catadroid@Nightstar-1mpovd.dab.02.net] has joined #code |
| 01:46 | | himi [sjjf@Nightstar-dm0.2ni.203.150.IP] has joined #code |
| 01:46 | | mode/#code [+o himi] by ChanServ |
| 01:47 | | Kindamoody|busy is now known as Kindamoody |
| 01:56 | | Kindamoody is now known as Kindamoody[zZz] |
| 02:21 | | * McMartin installs the Windows Subsystem for Linux |
| 02:21 | <&McMartin> | I can SSH over to irssi running in screen just fine, but it turns out that irssi can't quite handle windows console as a terminal type. |
| 03:01 | <@Alek> | Vorn, some shower thought math for you please. |
| 03:02 | <@Alek> | if a toilet paper roll increases in length by 10%, by how much does its radius increase? |
| 03:03 | <~Vornicus> | Very little. |
| 03:04 | <~Vornicus> | If the paper went all the way to the center, it'd be 4.8% radius increase; since it does not, it's even smaller. let me go measure a roll and give you a number |
| 03:10 | <~Vornicus> | okay. thing examined, 2.75" radius of paper, 0.875" radius of roll, gives 484/64 - 49/64 = 435/64 paper, increase by 10% you get 478.5 paper, plus the roll again = 527.5/64, square root of that is 22.96/8 = 2.871" final paper radius |
| 03:10 | <~Vornicus> | = 4.4% radius increase |
| 03:52 | | catadroid` [catadroid@Nightstar-53h.jtr.132.82.IP] has joined #code |
| 03:55 | | catadroid [catadroid@Nightstar-1mpovd.dab.02.net] has quit [Ping timeout: 121 seconds] |
| 04:50 | <@Alek> | Sweet. |
| 04:50 | <@Alek> | thanks muchly, vorn. |
| 04:51 | <@Alek> | I'm sure it still differs just a bit for various lengths, as well as various brands. :P |
| 04:54 | <~Vornicus> | The thing that makes it vary is the ratio of initial radius of the whole roll to the radius of the cardboard tube. |
| 04:56 | <~Vornicus> | I have a jumbo roll, so the tube has less of an effect. Get one of those giant institutional rolls and you're closer to the 4.8% figure which shows up for no tube; get a cheapo roll and the tube dominates. |
| 05:02 | | Derakon is now known as Derakon[AFK] |
| 05:04 | | starkruzr [quassel@Nightstar-m85rjf.fios.verizon.net] has quit [Ping timeout: 121 seconds] |
| 05:09 | <@Alek> | I was just looking at the 1000 sheet (1100) single-ply rolls in our bathrooms. XD |
| 05:10 | <@Alek> | would it be fair to say the % increase asymptotically approaches 4.8% from below as size increases? :P |
| 05:13 | <~Vornicus> | As the roll size increases relative to the tube, yeah |
| 05:23 | <&[R]> | Huzzah! npm search will do nothing other than hit the maximum memory the VM will allow it to take then abort. |
| 05:47 | <&McMartin> | science |
| 07:02 | | celticminstrel [celticminst@Nightstar-nhhr58.dsl.bell.ca] has quit [[NS] Quit: And lo! The computer falls into a deep sleep, to awake again some other day!] |
| 07:40 | <@gnolam> | Pandoc: great idea... in theory. Just doesn't work in practice. |
| 07:49 | < catadroid`> | I intended to write a blog post last night |
| 07:49 | < catadroid`> | But instead I generated a pathfinder character |
| 07:49 | | catadroid` is now known as catadroid |
| 07:55 | | Kindamoody[zZz] is now known as Kindamoody |
| 08:00 | | VirusJTG [VirusJTG@Nightstar-6i5vf7.sta.comporium.net] has quit [Connection closed] |
| 08:47 | | ion [Owner@Nightstar-6grqph.vs.shawcable.net] has quit [Ping timeout: 121 seconds] |
| 09:04 | <@gnolam> | wkhtmltopdf, however, looks like it can be tweaked to give an output that's good enough. |
| 09:06 | | Kindamoody is now known as Kindamoody|afk |
| 09:07 | | himi [sjjf@Nightstar-dm0.2ni.203.150.IP] has quit [Ping timeout: 121 seconds] |
| 09:11 | <~Vornicus> | w'rkncactner? |
| 09:11 | <@abudhabi> | Work cancer? |
| 09:11 | | * abudhabi breaks into the Daily Mail Song. |
| 09:13 | <@gnolam> | https://www.youtube.com/watch?v=5eBT6OSr1TI ? |
| 09:14 | <@abudhabi> | Ja. |
| 09:19 | <~Vornicus> | w'rkncactner: extremely powerful 'chaos-beings' from the marathon series |
| 09:23 | <@gnolam> | w'rknninetofive |
| 09:25 | < catadroid> | Ia ia cthulhu fhtagn |
| 09:31 | | ion [Owner@Nightstar-6grqph.vs.shawcable.net] has joined #code |
| 09:31 | | mode/#code [+o ion] by ChanServ |
| 09:36 | | Vornicus [Vorn@ServerAdministrator.Nightstar.Net] has quit [Connection reset by peer] |
| 09:59 | <@gnolam> | Ok, so this company apparently employs veterans. |
| 10:00 | <@gnolam> | Because if I'm not mistaken, that guy playing their on hold music is General MIDI. |
| 10:02 | < catadroid> | This code is impossible to understand without a lot of refactoring |
| 10:03 | < catadroid> | I am unwilling to care enough to fix it "properly" |
| 10:03 | | * catadroid sighs |
| 10:05 | < catadroid> | Now Rust feels like a half implemented lisp |
| 10:16 | <@abudhabi> | Just do it in Lisp. |
| 10:22 | | catadroid` [catadroid@Nightstar-1b191r.dab.02.net] has joined #code |
| 10:26 | | catadroid [catadroid@Nightstar-53h.jtr.132.82.IP] has quit [Ping timeout: 121 seconds] |
| 10:40 | | VirusJTG [VirusJTG@Nightstar-6i5vf7.sta.comporium.net] has joined #code |
| 10:40 | | mode/#code [+ao VirusJTG VirusJTG] by ChanServ |
| 11:06 | | ion [Owner@Nightstar-6grqph.vs.shawcable.net] has quit [Ping timeout: 121 seconds] |
| 11:25 | | ion [Owner@Nightstar-6grqph.vs.shawcable.net] has joined #code |
| 11:25 | | mode/#code [+o ion] by ChanServ |
| 12:19 | | himi [sjjf@Nightstar-v37cpe.internode.on.net] has joined #code |
| 12:19 | | mode/#code [+o himi] by ChanServ |
| 12:22 | < catadroid`> | Why can't anyone program well |
| 12:22 | < catadroid`> | (including me) |
| 12:38 | <@abudhabi> | Lack of death penalty for doing it poorly. |
| 12:39 | <@TheWatcher> | Well, that's be one way to have no programmers ever |
| 12:39 | <@TheWatcher> | *that'd |
| 12:40 | <@abudhabi> | Ionno about that. There used to be executions for fucking up at just about anything important, and people still did those things. |
| 12:41 | <@abudhabi> | Or are you saying every programmer is a bad programmer? :V |
| 12:41 | <@TheWatcher> | Yes, but y'see, there was a possibility that they could do it without fucking up |
| 12:41 | <@TheWatcher> | Yep |
| 12:41 | <@TheWatcher> | That's exactly what I'm saying |
| 12:41 | <@abudhabi> | Ha. |
| 12:42 | <@Tamber> | The only way to remove human error from the system, is to remove the source of the error. To wit, remove the humans. |
| 12:45 | <@TheWatcher> | Sure, most of the time they may be excellent, even godlike in their programming ability, but everyone - from a neophyte through to Don Knuth - writes crap code at some point, or comments poorly, or screws up in some fashion. |
| 12:48 | < catadroid`> | Value lifetime semantics and functional interface semantics are very different things |
| 12:48 | < catadroid`> | Hm |
| 12:48 | < catadroid`> | All good programmers are bad programmers |
| 12:49 | < catadroid`> | Because you should be constantly skeptical of code and know what aspects of the code you have just written are good, bad, and what compromises were made and why |
| 12:50 | < catadroid`> | Some of those are necessarily going to be 'bad' tradeoffs |
| 12:57 | < catadroid`> | I have too many words |
| 13:04 | <@abudhabi> | Sell excess words. |
| 13:56 | | You're now known as TheWatcher[d00m] |
| 13:58 | | catadroid` is now known as catadroid |
| 14:12 | | celticminstrel [celticminst@Nightstar-nhhr58.dsl.bell.ca] has joined #code |
| 14:12 | | mode/#code [+o celticminstrel] by ChanServ |
| 14:47 | < catadroid> | "your object system is too clasy" |
| 14:47 | < catadroid> | Classy |
| 14:47 | < catadroid> | Even |
| 14:49 | <@celticminstrel> | Speaking of classy, I wonder how I can do something prototype-like in C++. |
| 15:42 | < catadroid> | Hm. I'm interested in the results of that |
| 15:42 | < catadroid> | My intuition is that you can't get very close |
| 15:45 | | Vornicus [Vorn@ServerAdministrator.Nightstar.Net] has joined #code |
| 15:45 | | mode/#code [+qo Vornicus Vornicus] by ChanServ |
| 15:46 | <@celticminstrel> | I think I could probably manage it if it was a fixed-length prototype chain, at least. |
| 15:47 | <@celticminstrel> | Not entirely sure if that would be sufficient for my needs. |
| 16:09 | <@abudhabi> | Word processors: Expensive, crap, resource hog. Pick at least one. |
| 16:13 | | * catadroid processes one (1) word |
| 16:14 | <@abudhabi> | We have the Microsoft suite, which has the disadvantages of being expensive and not running (not *really*) on Linux. |
| 16:14 | <@abudhabi> | Then we have the Libre/Open suit, which is a gigantic resource hog and does not play nice with the Microsoft formats. |
| 16:14 | <@abudhabi> | And then are the third party contenders, who are mostly crap. |
| 16:16 | <@celticminstrel> | Does Corel still have one? (Though it probably shares problems you listed for Word x10.) |
| 16:17 | <@abudhabi> | I only ever used Corel Draw, and that was ages ago. |
| 16:18 | <@celticminstrel> | I feel like Lotus might've had one too... |
| 16:31 | < catadroid> | Hrngh, my brain is obsessed with type systems right now |
| 16:32 | < catadroid> | Remind me to actually write my ideas down this evening |
| 16:38 | <@celticminstrel> | I wonder if there's a pastebin site that can be accessed without a browser... >_> |
| 16:38 | | * Vornicus sets an alarm: "remind catadroid to write down her type system ideas" |
| 16:43 | < ToxicFrog> | celticminstrel: to the wget! |
| 16:43 | <@celticminstrel> | Oh, I can view a pastebin with curl no problem. |
| 16:43 | <@celticminstrel> | (At pastebin.com it requires setting the referrer, though.) |
| 16:44 | < ToxicFrog> | abudhabi: I've basically switched to googledocs for everything*. It's a resource hog, but less so than either word or libreoffice, and it doesn't confuse and terrify windows users. |
| 16:44 | < ToxicFrog> | celticminstrel: wget can also POST data |
| 16:44 | < ToxicFrog> | And if you look for [pastebin shell script] or similar you'll likely find a bunch of scripts using it or curl to upload to pastebin from the command line |
| 16:45 | <@celticminstrel> | I can't remember if I even have wget. It's certainly not installed by default on my system, but I might've installed it at some point. |
| 16:45 | <@celticminstrel> | I might search out some scripts later. |
| 16:45 | <@celticminstrel> | For now I just killed and relaunched Firefox instead. >_> |
| 16:46 | < ToxicFrog> | *everything I'm not using LaTeX for |
| 16:46 | <~Vornicus> | celmin: you might be able to use gist via git but I'm not sure |
| 16:49 | <@abudhabi> | ToxicFrog: I have now, per [R]'s suggestion, switched to using the universal language: HTML. |
| 16:49 | < ToxicFrog> | abudhabi: I'm so sorry |
| 16:50 | <~Vornicus> | As a professional web developer: you fool |
| 16:50 | <@abudhabi> | I *am* a professional web dev myself. |
| 16:50 | <@celticminstrel> | I don't get it. |
| 16:51 | <@abudhabi> | And I find HTML to be greatly better in terms of portability and fine-tunability than fucking around with DOC/ODT/what-have-you. |
| 16:51 | <~Vornicus> | get something like tinymce or github-flavored markdown + mathjax and let *it* generate your html |
| 16:51 | <@abudhabi> | I'm using KompoZer for the drudgery. |
| 16:52 | <@abudhabi> | Editing the source where I need to. |
| 16:52 | <~Vornicus> | guess that works |
| 16:52 | <@celticminstrel> | Wait, using HTML for word-processing? |
| 16:52 | | catadroid` [catadroid@Nightstar-isd.jtr.132.82.IP] has joined #code |
| 16:53 | <@abudhabi> | I need to make a manual document. I might as well write a HTML manual document. |
| 16:53 | <@abudhabi> | Screw page-based formatting and printability. Internet access is a given. |
| 16:54 | <@abudhabi> | (Manual pertains a procedure that cannot be properly done without internets. So it doesn't matter.) |
| 16:54 | | catadroid [catadroid@Nightstar-1b191r.dab.02.net] has quit [Ping timeout: 121 seconds] |
| 17:09 | < catadroid`> | That moment when you trust a colleague you mentored to do well enough designing a feature that you can give then a couple lines if description instead of a full design |
| 17:10 | < catadroid`> | Of* |
| 17:11 | <~Vornicus> | it's a good moment |
| 17:11 | < catadroid`> | Hm. I think my terminology is settling on descriptive typing vs prescriptive typing |
| 17:20 | | gizmore [kvirc@Nightstar-0jchfs.dip0.t-ipconnect.de] has joined #code |
| 18:57 | <@Pi> | catadroid`: Woo :) |
| 19:13 | | gnolam [lenin@Nightstar-t1tbf0.cust.bahnhof.se] has quit [[NS] Quit: Galileo! Galileo! Galileo Figaro!] |
| 19:36 | | catadroid` [catadroid@Nightstar-isd.jtr.132.82.IP] has quit [The TLS connection was non-properly terminated.] |
| 19:36 | | catadroid [catadroid@Nightstar-bt5k4h.81.in-addr.arpa] has joined #code |
| 19:40 | | gizmore [kvirc@Nightstar-0jchfs.dip0.t-ipconnect.de] has quit [Operation timed out] |
| 19:43 | | gizmore [kvirc@Nightstar-2i1s2t.dip0.t-ipconnect.de] has joined #code |
| 20:00 | | * Vornicus 's alarm goes off |
| 20:00 | <~Vornicus> | Catadroid! Write down your type system ideas! |
| 20:06 | < catadroid> | Thank you! I will when I no longer need to keep my leg raised due to injury D: |
| 20:07 | <~Vornicus> | D: |
| 20:10 | <&[R]> | D: |
| 20:11 | <&[R]> | http://www.antipsychiatry.org. <-- For laughs |
| 20:39 | <~Vornicus> | that spinning e |
| 20:39 | <@abudhabi> | Web 1.0. |
| 20:41 | < catadroid> | Web 0.9 |
| 20:53 | | gizmore [kvirc@Nightstar-2i1s2t.dip0.t-ipconnect.de] has quit [A TLS packet with unexpected length was received.] |
| 21:36 | | Kindamoody|afk is now known as Kindamoody |
| 21:50 | | gnolam [lenin@Nightstar-t1tbf0.cust.bahnhof.se] has joined #code |
| 21:51 | | mode/#code [+o gnolam] by ChanServ |
| 21:52 | <&McMartin> | Speaking of injury, my pulled muscle from yesterday is now merely a stiff back as opposed to actively impeding my ability to move |
| 21:54 | <@TheWatcher[d00m]> | This is good, hopefully the stiffness clears soon. |
| 21:54 | | You're now known as TheWatcher |
| 22:02 | | catadroid [catadroid@Nightstar-bt5k4h.81.in-addr.arpa] has quit [[NS] Quit: Bye] |
| 22:11 | <@Azash> | McSmartin' |
| 22:13 | | catadroid [catadroid@Nightstar-atnkn7.dab.02.net] has joined #code |
| 22:17 | < catadroid> | rawr |
| 22:17 | < catadroid> | What words did I use earlier |
| 22:17 | < catadroid> | Oh yeah, descriptive and prescriptive |
| 22:17 | <&McMartin> | ZIENCE |
| 22:21 | | Kindamoody is now known as Kindamoody[zZz] |
| 22:31 | < catadroid> | Brain c'mon |
| 22:31 | < catadroid> | You aren't allowed to spend all day thinking about this then leave me stuck here late at night |
| 22:32 | <~Vornicus> | fuckin brain |
| 22:32 | <~Vornicus> | also fuckin brian |
| 22:33 | <~Vornicus> | (he knows what he did) |
| 22:33 | < catadroid> | Oh, wow, Windows 10 us getting clever with the windows key + arrow placement of things |
| 22:34 | <&McMartin> | I do like the new half-fullscreen with a bar that shifts both windows |
| 22:34 | <~Vornicus> | I like that, and the quarter screen ones |
| 22:57 | | Derakon[AFK] is now known as Derakon |
| 23:24 | < catadroid> | Yeah, having played with Go interfaces a bit, I think I like them a lot |
| 23:24 | < catadroid> | And I think they've accidentally stumbled on something quite interesting |
| 23:25 | < catadroid> | Also, I actually quite like Go itself |
| 23:25 | < catadroid> | Which I was not expecting |
| 23:34 | <&McMartin> | Go rubbed me super-hard the wrong way, for various reasons |
| 23:35 | <&McMartin> | Go interfaces are pretty OK but they're also not IMO a full replacement for OO, and it really felt like they thought it was |
| 23:35 | <&McMartin> | Channels are super-fun though |
| 23:36 | <~Vornicus> | I need to at some point learn another language here |
| 23:36 | | * McMartin has only written one actual proper program in Go. |
| 23:36 | <~Vornicus> | I've been all python js php lua for way too long |
| 23:36 | <&McMartin> | You should pick up C, just for the maximum distance :D |
| 23:37 | <&McMartin> | And then not use it for anything, but being able to kind of map it to core silicon concepts is probably pretty OK |
| 23:37 | <~Vornicus> | C is a language I know |
| 23:37 | <~Vornicus> | I just don't have use for it. |
| 23:39 | <~Vornicus> | Which I guess is mostly because I haven't done anything that actually pegs the processor in forever. |
| 23:39 | < catadroid> | I honestly believe that Go's interfaces are an idea that could enhance something like C++ |
| 23:40 | < catadroid> | Because they describe abstraction at the point of use rather than prescribing it at both the point of use and point of definition of type |
| 23:40 | < catadroid> | They act like what concepts are trying to achieve |
| 23:40 | <~Vornicus> | which actually is a big barrier. I don't have any idea what I'd use any of these things for |
| 23:40 | < catadroid> | I suppose concepts are probably powerful enough |
| 23:40 | < catadroid> | Not to require the idea in C++ |
| 23:41 | <&McMartin> | I think so |
| 23:41 | <&McMartin> | Especially since interface{} is almost literally void * |
| 23:41 | <&McMartin> | Semantically, anyway |
| 23:41 | < catadroid> | Anyway, I think interfaces match human intuition in terms of abstraction better than any OO system I've seen |
| 23:41 | <&McMartin> | The closest thing I've seen for this is the ObjC notion of Protocols |
| 23:41 | < catadroid> | This is why I need to formulate these ideas into a paper or long form blog post or something |
| 23:42 | <&McMartin> | But yeah, unlike ObjC, Go seems to be specialized into non-global method calls |
| 23:42 | < catadroid> | Because it's hard to show you what's currently in my mind |
| 23:43 | < catadroid> | I should look at ObjC and Scala structured typing in more detail as well |
| 23:43 | < catadroid> | Go seems alright, the authors feel very... Authoritarian though |
| 23:44 | <&McMartin> | If I cared about drama, which I do, a little, Go is in my "run away" bin |
| 23:44 | <&McMartin> | It's the Rob Pike Full Employment Act, but if I allow that a bunch of minor niggles about the language suddenly become incredibly dumb |
| 23:44 | <&McMartin> | Like the automatic semicolons |
| 23:44 | <&McMartin> | Or the result, err tuple returns instead of a proper variant type |
| 23:45 | < catadroid> | Ah, protocols seem like what I'm thinking of me precisely |
| 23:45 | <&McMartin> | It's like he got the memo on OO not being that great, and exceptions being more trouble than they're worth, but then didn't realize that you still have to actually provide the power they provided somehow |
| 23:46 | <@Pi> | Go is a "Greatest Hits of the 70s" compilation album :) |
| 23:46 | | * Pi ducks |
| 23:46 | <@celticminstrel> | Speaking of concepts, does C++ have those in 14 or 17? |
| 23:46 | < catadroid> | No |
| 23:46 | <&McMartin> | Some of that is probably just fixing one's idioms - it's actively a bad idea to return an object of unknown type but matching an interface in Go, last I checked, and that's just basic hygiene in Java - but it's still an obvious crash in power |
| 23:46 | < catadroid> | They nearly made it into 17 |
| 23:46 | <&McMartin> | Pi: Not even that. Pretty sure Delphi is that. :) |
| 23:47 | <@celticminstrel> | Well, they nearly made it into 11 too. |
| 23:47 | <&McMartin> | A number of fixed concepts are hard-coded in |
| 23:47 | <&McMartin> | And a number more aren't but have defined meanings |
| 23:47 | < catadroid> | I'm still in a world where Lisp is irreducibly the best idea |
| 23:47 | <@celticminstrel> | Hard-coded in? |
| 23:47 | < catadroid> | But I'm currently doing a grand tour of type systems for fun and profit |
| 23:47 | <&McMartin> | std::is_std_layout, for instance. |
| 23:48 | < catadroid> | Go feels like a Google domain language |
| 23:48 | <&McMartin> | Lisp is definitely one of the most powerful weapons in the dynamic typing camp |
| 23:48 | <@celticminstrel> | I wasn't think of that kind of thing as a concept, but okay. |
| 23:48 | <&McMartin> | catadroid: I'm 99.9999% positive it is, and that it's specifically the end result of them getting fed up with Python's Global Interpreter Lock |
| 23:49 | < catadroid> | celticminstrel: they're much closer to existing than 11's |
| 23:49 | <~Vornicus> | I don't even really know what I'd do with type systems, it feels like stuff I just... don't care about, because most of the time I can keep enough state in my head to know what everything is myself |
| 23:50 | < catadroid> | I think Lisp is just the natural end point of templating and has immense power because of that |
| 23:50 | <&McMartin> | The concept that I want to push on more thoroughly came in at the end of my OO exploration |
| 23:50 | <&McMartin> | "data objects really have two componentsâa shared component that is the same as all other instances of its type, and an instanced component for data that is unique to that one object" |
| 23:50 | <@Pi> | Vornicus: Imagine how much you could do if you can have the computer free your mind from the low-level accounting. |
| 23:50 | <@Pi> | Vornicus: I good type system is like power armour for your brain. |
| 23:50 | <@Pi> | (A good type system, even) |
| 23:51 | <&McMartin> | Pi: The whole point is that the answer to that is "nothing" in most cases |
| 23:51 | <@Pi> | It amplifies whatever you can do without it. |
| 23:51 | <~Vornicus> | Pi: that's the thing, I don't see it helping because at some level I *must* know what the thing is anyway, in order to operate on it at all |
| 23:51 | < catadroid> | I think types have great power in allowing you to condense your thinking by reducing the number of concepts your mind needs to hold at once |
| 23:51 | <&McMartin> | If static typing actually was necessary to build anything at scale, the static vs. dynamic typing wars would have ended in victory for static decades ago |
| 23:51 | <&McMartin> | This very pointedly did not happen. |
| 23:52 | <~Vornicus> | Without knowing what the thing I'm working with is I literally can't program |
| 23:52 | < catadroid> | I honestly think that the static vs dynamic typing argument is missing the point |
| 23:52 | <&McMartin> | I think it's bound up with the notion of uniform reference |
| 23:53 | <&McMartin> | Which, if you're in an ML-like land, is forced on you by the type system because otherwise a type signature like 'a list -> 'a can't make any sense at all |
| 23:53 | < catadroid> | Vornicus: think of types as a way to give your data meta information that lets tooling automate some of the things you do in your head |
| 23:53 | < catadroid> | If that helps any |
| 23:54 | <&McMartin> | Look at Vorn's list, though |
| 23:54 | <&McMartin> | The parts that he's running into are things like "the doc says that this function takes a file-like object" |
| 23:55 | <&McMartin> | Everything else is arguably jumping through hoops to try to teach a machine what "file like object" means well enough to check your code for you without unduly restricting you. |
| 23:55 | <&McMartin> | And the idea that 'without unduly restricting you' was a valuable goal at all seems like a relatively recent development. |
| 23:56 | <~Vornicus> | Yeah. When I learned old-school C++, auto didn't exist, for instance. |
| 23:56 | < catadroid> | Oh god C++ is awful and will always be awful |
| 23:56 | < catadroid> | But it's very powerful |
| 23:56 | <&McMartin> | One of my objections to Go is that it relies on a deduction mechanism that is actually a bit weaker than auto and nevertheless thinks it can use that to have polymorphism |
| 23:57 | <&McMartin> | Like, my Go program was an image processor |
| 23:57 | <&McMartin> | In any other language in that vein, I would be returning an Image |
| 23:57 | <&McMartin> | But here I am returning a specific RGBA * struct that I have to name precisely |
| 23:57 | <&McMartin> | Code that *uses* my library is supposed to use auto and then treat that like an Image |
| 23:58 | <~Vornicus> | and the other languages I technically know are, like, Java, for fuck's sake, which is terrible and I never saw any advantage to telling it what I'm up to |
| 23:58 | < catadroid> | That's because Go's interfaces are descriptive but not prescriptive |
| 23:58 | <&McMartin> | And if I need to alter the implementation class later, everything works as long as everybody used auto everywhere |
| 23:58 | < catadroid> | Hm. I'm not sure my terms work in that context very well |
| 23:59 | < catadroid> | But yes, Go interfaces lack the ability to explicitly instance them |
| --- Log closed Fri Aug 05 00:00:05 2016 |