Home freenode/#haskell: Logs Calendar

Logs: freenode/#haskell

←Prev  Next→
Page 1 .. 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 .. 5022
— 502,152 events total
2020-11-25 14:28:28 <honigkuchen> is there a reason that specifically fits to quicksort, why haskell is here "better" or whatsoever
2020-11-25 14:28:42 <tdammers> haskell isn't better at implementing quicksort
2020-11-25 14:28:50 → LKoen joins (~LKoen@169.244.88.92.rev.sfr.net)
2020-11-25 14:28:51 <maerwald> honigkuchen: https://augustss.blogspot.com/2007/08/quicksort-in-haskell-quicksort-is.html
2020-11-25 14:29:07 <tdammers> the famous "quicksort but not quite actually" example is used to demonstrate the elegance of the language
2020-11-25 14:29:16 × elliott__ quits (~elliott@pool-108-51-141-12.washdc.fios.verizon.net) (Ping timeout: 240 seconds)
2020-11-25 14:29:22 → star_cloud joins (~star_clou@ec2-34-220-44-120.us-west-2.compute.amazonaws.com)
2020-11-25 14:30:05 → supercoven joins (~Supercove@196.244.192.58)
2020-11-25 14:30:06 × supercoven quits (~Supercove@196.244.192.58) (Max SendQ exceeded)
2020-11-25 14:30:20 <n0042> merijn: The I/O difficulties I've been having are mostly just getting the hang of the syntax. I've been doing some assignments in Haskell that require using arrays, and the difference in syntax between ones that have to be treated like Monad-like objects and ones that don't is kind of a learning curve.
2020-11-25 14:30:20 → supercoven joins (~Supercove@196.244.192.58)
2020-11-25 14:30:21 × supercoven quits (~Supercove@196.244.192.58) (Max SendQ exceeded)
2020-11-25 14:30:35 → supercoven joins (~Supercove@196.244.192.58)
2020-11-25 14:30:36 × supercoven quits (~Supercove@196.244.192.58) (Max SendQ exceeded)
2020-11-25 14:30:50 → supercoven joins (~Supercove@196.244.192.58)
2020-11-25 14:30:51 × supercoven quits (~Supercove@196.244.192.58) (Max SendQ exceeded)
2020-11-25 14:30:52 <n0042> I've been messing with immutable arrays, mutable arrays, and vectors to make my stuff meet the speed requirements (which are contrived and strict)
2020-11-25 14:31:05 → supercoven joins (~Supercove@196.244.192.58)
2020-11-25 14:31:06 × supercoven quits (~Supercove@196.244.192.58) (Max SendQ exceeded)
2020-11-25 14:31:08 <merijn> n0042: ah, yeah, that can be a bit tricky
2020-11-25 14:31:11 <n0042> And it has been pretty fun. Haskell is a really fun language. Glad I took the plunge
2020-11-25 14:31:52 <honigkuchen> can it make sense to combine python and haskell in one program ?
2020-11-25 14:32:03 ← f-a parts (~f-a@151.68.209.164) ()
2020-11-25 14:32:32 <maerwald> python is used in GHC for the tests :p
2020-11-25 14:32:35 <maerwald> does that count?
2020-11-25 14:32:36 <n0042> Python is a great "glue" language (I use it for stuff instead of Bash all the time). So I could see myself writing a program in Haskell and then calling it from within a Python script.
2020-11-25 14:32:52 <merijn> n0042: tbh, I now consider Haskell a much better glue language :p
2020-11-25 14:33:05 <LKoen> honigkuchen: https://github.com/jacquev6/Polyglot
2020-11-25 14:33:09 <merijn> n0042: It's incredibly easy to call C code from Haskell
2020-11-25 14:33:35 <tdammers> merijn: until that C code isn't reentrant, or takes callbacks, or involves variadic arguments
2020-11-25 14:33:45 <LKoen> I once run into someone's code who used a makefile written in R, with code in C, C++, R and python
2020-11-25 14:33:46 <merijn> tdammers: Yeah, but that's difficult in C too :p
2020-11-25 14:33:48 → geekosaur joins (82659a09@host154-009.vpn.uakron.edu)
2020-11-25 14:33:58 <merijn> tdammers: tbh, even the callback one is easy
2020-11-25 14:34:06 × bitmapper quits (uid464869@gateway/web/irccloud.com/x-tripsbcbodayyxwv) (Quit: Connection closed for inactivity)
2020-11-25 14:34:25 <merijn> Lack of reentrancy is hard, but also problematic in C and variadic arguments can't be used portably anyway, even within C :)
2020-11-25 14:34:59 <merijn> tdammers: So at worst it's "no harder than using C from C" :p
2020-11-25 14:35:09 <merijn> (which can, admittedly, be pretty hard)
2020-11-25 14:35:45 × gproto23 quits (~gproto23@unaffiliated/gproto23) (Ping timeout: 240 seconds)
2020-11-25 14:35:47 <n0042> I love the way `readFile` works in Haskell. That's got to be the easiest way to do that ever.
2020-11-25 14:35:48 <honigkuchen> lol
2020-11-25 14:35:52 <merijn> tdammers: Anyway, I was mostly comparing to, say, the insanity of the FFI of python, Java, etc. which are much more convoluted and painful to use
2020-11-25 14:36:01 <tdammers> no argument there :D
2020-11-25 14:36:09 <merijn> n0042: Using the one from Prelude that returns String?
2020-11-25 14:36:15 <n0042> Yeah
2020-11-25 14:36:38 ski . o O ( "Modelling Large Datasets Using Algebraic Datatypes: A Case Study of the CONFMAN Database" by Markus Mottl in 2002-05-15 at <http://www.ofai.at/cgi-bin/get-tr?paper=oefai-tr-2002-27.pdf> )
2020-11-25 14:37:01 <merijn> n0042: Might wanna use Text for reading textual data: https://hackage.haskell.org/package/text-1.2.4.0/docs/Data-Text-IO.html#v:readFile
2020-11-25 14:37:04 <gentauro> geekosaur: do you have a multiscreen setup with XMonad?
2020-11-25 14:37:12 <merijn> (since you said performance matters)
2020-11-25 14:37:19 <gentauro> more specifically, daisy chaining?
2020-11-25 14:37:49 ski . o O ( "Using Algebraic Datatypes as Uniform Representation for Structured Data" by Markus Mottl in 2003-03-07 at <http://www.ofai.at/cgi-bin/get-tr?paper=oefai-tr-2003-07.pdf>,<http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.13.5712> )
2020-11-25 14:37:50 <n0042> merijn: Thank you very much
2020-11-25 14:38:10 <tdammers> also, there's the caveat that readFile does lazy I/O, and that means the actual reading may happen at surprising times
2020-11-25 14:39:07 <tdammers> e.g.: do { str <- readFile "foo"; unlink "foo"; putStrLn str } -- this may crash due to trying to read from a deleted file
2020-11-25 14:39:09 → Sgeo joins (~Sgeo@ool-18b982ad.dyn.optonline.net)
2020-11-25 14:39:13 × darjeeling_ quits (~darjeelin@122.245.219.209) (Ping timeout: 256 seconds)
2020-11-25 14:39:21 <Boomerang_> I hear Data.Text.IO is not good if you care about locales
2020-11-25 14:39:26 <Boomerang_> https://www.snoyman.com/series/haskell-bad-parts
2020-11-25 14:42:27 <gentauro> tdammers: where do you delete the file? `unlink`?
2020-11-25 14:42:32 <tdammers> yeah
2020-11-25 14:42:53 <tdammers> although I believe on a typical (Linux) filesystem, you have to actually truncate it for that to crash
2020-11-25 14:43:46 × chrpape quits (~user@2a01:4f9:c010:632d::1) (Quit: ERC (IRC client for Emacs 25.2.2))
2020-11-25 14:43:56 × Nahra quits (~Nahra@unaffiliated/nahra) (Ping timeout: 240 seconds)
2020-11-25 14:44:09 <n0042> Thanks Boomerang. Lots of information in that article.
2020-11-25 14:44:32 <n0042> I wasn't aware of some of the pitfalls of `readFile`
2020-11-25 14:44:59 <gentauro> tdammers: I recall me doing this -> `readFileStrict f = LBS.readFile f >>= \bs -> LBS.length bs `seq` pure bs` :'(
2020-11-25 14:45:20 <gentauro> a strict version cos Linux complained about having to many files open
2020-11-25 14:46:03 <merijn> tdammers: Text one doesn't do lazy IO
2020-11-25 14:46:05 <gentauro> tdammers: how do you write `seq` in do-nation?
2020-11-25 14:46:09 <tdammers> gentauro: why not just LBS.fromStrict <$> BS.readFile f
2020-11-25 14:46:21 <tdammers> tdammers: lazy text one does I think
2020-11-25 14:46:44 <merijn> Boomerang_: eh, that's wrong
2020-11-25 14:46:51 → chrpape joins (~user@2a01:4f9:c010:632d::1)
2020-11-25 14:46:55 <merijn> Boomerang_: readFile works fine if your locale isn't broken
2020-11-25 14:47:23 <merijn> Boomerang_: readFileUtf8, etc. are dumb hacks that are broken on machines that are setup correctly
2020-11-25 14:47:31 <tdammers> readFile does the correct thing: decode raw bytes into Unicode as per the current locale
2020-11-25 14:47:42 <gentauro> tdammers: I don't think I have a `strict` ByteString
2020-11-25 14:47:46 <gentauro> LBS is `lazy`
2020-11-25 14:47:50 <merijn> Boomerang_: They're basically "ignore the local specified encoding and always use utf8", which is bad and anyone doing that should feel bad >.>
2020-11-25 14:48:00 <tdammers> gentauro: yes, but you can read a strict bytestring and convert it to lazy
2020-11-25 14:48:02 × Fractalis quits (~Fractalis@2601:987:280:8d40:eda9:f9e1:2072:cea7) (Ping timeout: 264 seconds)
2020-11-25 14:48:04 <gentauro> «fromStrict :: ByteString -> ByteStringO(1) Convert a strict ByteString into a lazy ByteString.»
2020-11-25 14:48:09 <merijn> n0042: ^^ see above :)
2020-11-25 14:48:21 <gentauro> tdammers: oh, got it
2020-11-25 14:48:32 <merijn> gentauro: lazy bytestring is, effectively, just [Strict.ByteString] :p
2020-11-25 14:49:01 <merijn> (i.e. a lazy list of strict chunks)
2020-11-25 14:49:06 <tdammers> it's slightly more clever than [ByteString], but yeah, essentially that
2020-11-25 14:49:13 <gentauro> merijn: yeah, I got that
2020-11-25 14:49:23 <n0042> Luckily I'm only using it to read predictable input right now, so it works pretty well.
2020-11-25 14:49:50 <tdammers> anyway, the problem with readFile and locales is when people assume that any file they might possibly want to read from is encoded according to the current locale
2020-11-25 14:50:05 <tdammers> which is of course often wrong, but just assuming UTF8 isn't any better
2020-11-25 14:50:11 → machinedgod joins (~machinedg@135-23-192-217.cpe.pppoe.ca)
2020-11-25 14:50:30 hackage haskell-xmpp 2.0.0 - Haskell XMPP (eXtensible Message Passing Protocol, a.k.a. Jabber) library https://hackage.haskell.org/package/haskell-xmpp-2.0.0 (Jappie)
2020-11-25 14:50:33 <gentauro> tdammers: the only way to be 100% is to read the hole file
2020-11-25 14:50:35 <merijn> tdammers: Also, people who have badly configured machines with broken locales complaining that they're broken and "fixing" it by breaking it for everyone who *does* have a properly configured locale >.>
2020-11-25 14:51:01 <Boomerang_> merijn: I guess it depends on the use case but I think more often I would rather it reported an error if it's not UTF-8 than do something weird instead
2020-11-25 14:51:18 <merijn> Boomerang_: It doesn't do anything weird, though
2020-11-25 14:51:36 <merijn> Boomerang_: It literallys asks the environment "what is the environment encoding?" and then uses that
2020-11-25 14:52:03 <merijn> Boomerang_: If you want explicit control over decoding you can either 1) override the encoding for a handle or 2) read a ByteString and explicitly decode it
2020-11-25 14:52:19 → darjeeling_ joins (~darjeelin@122.245.217.23)
←Prev  Next→
Page 1 .. 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 .. 5022

All times are in UTC.