Home freenode/#haskell: Logs Calendar

Logs: freenode/#haskell

←Prev  Next→
Page 1 .. 116 117 118 119 120 121 122 123 124 125 126 .. 5022
— 502,152 events total
2020-09-21 14:33:56 <merijn> tomsmeding: STM version of MVar
2020-09-21 14:34:10 <tomsmeding> isn't that TVar?
2020-09-21 14:34:12 <merijn> mastarija: What's the actual problem, though?
2020-09-21 14:34:13 <merijn> tomsmeding: No
2020-09-21 14:34:18 <merijn> tomsmeding: TVar can't be empty
2020-09-21 14:34:21 <tomsmeding> ah
2020-09-21 14:34:24 <mastarija> merijn, game loop for multiplayer game
2020-09-21 14:34:34 <mastarija> and I need to receive data form the server
2020-09-21 14:34:42 <merijn> mastarija: Eh, more detailed then that :p Why do you need to poll for data?
2020-09-21 14:34:43 <mastarija> so I can't have my loop block
2020-09-21 14:35:08 <merijn> You can do something super dumb
2020-09-21 14:35:22 <merijn> Like, use and MVar and have one thread "read from Chan, write to MVar"
2020-09-21 14:35:33 <merijn> and then your main loop can do "tryTakeMVar"
2020-09-21 14:35:38 <mastarija> merijn, that's what I suggested earlier
2020-09-21 14:35:43 <mastarija> I think I'll go with that :D
2020-09-21 14:35:44 <tomsmeding> that could work :p
2020-09-21 14:35:53 → John20 joins (~John@82.46.59.122)
2020-09-21 14:35:55 <tomsmeding> it won't even be that slow perhaps
2020-09-21 14:35:56 <merijn> It probably wouldn't even be that inefficient, tbh
2020-09-21 14:36:01 <tomsmeding> lol
2020-09-21 14:36:18 <merijn> sleeping on Chan/MVar is pretty cheap
2020-09-21 14:36:24 <mastarija> Ok, thanks for the brainstorm peeps!
2020-09-21 14:36:29 <Eduard_Munteanu> One other reason to refactor is to use a bounded channel instead of Chan to avoid leaks.
2020-09-21 14:36:50 <mastarija> Meh, I'll accept spaceleaks for now XD
2020-09-21 14:36:51 <John20> Hi all, Does anyone have any examples of LanguageExtensions that have made it into the haskell language 'proper'?
2020-09-21 14:37:16 <mastarija> Eduard_Munteanu, But yes, I will have to fix that eventually
2020-09-21 14:37:31 <merijn> John20: PatternGuards extension was added to the report in Haskell2010
2020-09-21 14:37:43 <merijn> as was EmptyDataDecls?
2020-09-21 14:38:11 <Eduard_Munteanu> John20, https://www.haskell.org/onlinereport/haskell2010/haskellli2.html#x3-5000
2020-09-21 14:38:24 → Kaiepi joins (~Kaiepi@nwcsnbsc03w-47-55-157-9.dhcp-dynamic.fibreop.nb.bellaliant.net)
2020-09-21 14:39:02 → heatsink joins (~heatsink@107-136-5-69.lightspeed.sntcca.sbcglobal.net)
2020-09-21 14:39:14 <Eduard_Munteanu> I'd count a few more as de facto into the language, such as FlexibleInstances, MPTCs...
2020-09-21 14:40:00 → ph88^ joins (~ph88@2a02:8109:9e40:2704:3dae:c0b9:eddf:4cf0)
2020-09-21 14:40:59 <Eduard_Munteanu> It's hard to imagine Haskell without some of those even if not in the report.
2020-09-21 14:42:05 → gmt joins (~gmt@pool-71-105-108-44.nycmny.fios.verizon.net)
2020-09-21 14:42:25 <John20> Thanks, that's really useful. I'm writing a blog post about some of the things that are surprising about Haskell for someone coming from a more mainstream language (Java in my case).
2020-09-21 14:43:28 × heatsink quits (~heatsink@107-136-5-69.lightspeed.sntcca.sbcglobal.net) (Ping timeout: 256 seconds)
2020-09-21 14:44:06 <merijn> John20: The main problem is that "properly specifying" these extensions is tricky and that's what any new Haskell Report would have to do to include them. Which is why the past 2 (3?) attempts to produce a new report have stranded
2020-09-21 14:44:17 × ph88 quits (~ph88@2a02:8109:9e40:2704:b503:e755:2c19:955c) (Ping timeout: 272 seconds)
2020-09-21 14:44:27 hackage flink-statefulfun 0.1.0.0 - Flink stateful functions SDK https://hackage.haskell.org/package/flink-statefulfun-0.1.0.0 (tdbgamer)
2020-09-21 14:44:32 <cpressey> Hi. I have a list of Word8 and I'd like to render it as hexdigit pairs (e.g. "03 a9 1e" etc). I have written a function to do this, but it uses Text.Printf, which produces a String, but the consumer of my function now expects a Data.Text.Text. I could just DT.pack the string, but... any suggestions for something more elegant?
2020-09-21 14:44:38 <merijn> Everyone wants to include "the most common/important stuff into the report", but actually specifying them is hard and people give up
2020-09-21 14:45:06 <merijn> cpressey: fyi, the Numeric module has functions for doing that too
2020-09-21 14:45:15 <John20> Language Extensions certainly fit that bill. to my mind, they feel like an incubator for new language features however, they rarely seem to actually make it into the core language. Does anyone know why that is? Is my understanding roughly correct?
2020-09-21 14:45:19 <merijn> And those are less...questionable then printf :p
2020-09-21 14:45:31 <merijn> John20: See my above 2 comments ;)
2020-09-21 14:45:48 <merijn> > showHex 152 ""
2020-09-21 14:45:51 <lambdabot> "98"
2020-09-21 14:46:00 <merijn> > showHex 252 ""
2020-09-21 14:46:04 <lambdabot> "fc"
2020-09-21 14:46:07 <John20> thanks merijn. Missed them while typing out my response!
2020-09-21 14:46:45 <merijn> cpressey: So I'd probably go with '\n -> DT.pack . showHex n ""' or something
2020-09-21 14:46:52 × gmt quits (~gmt@pool-71-105-108-44.nycmny.fios.verizon.net) (Ping timeout: 256 seconds)
2020-09-21 14:47:01 × sleblanc quits (~sleblanc@unaffiliated/sebleblanc) (Ping timeout: 264 seconds)
2020-09-21 14:47:16 <merijn> cpressey: Alternatively, you can go via text's Builder: https://hackage.haskell.org/package/text-1.2.4.0/docs/Data-Text-Lazy-Builder-Int.html#v:hexadecimal
2020-09-21 14:47:49 <merijn> John20: The other factor is that, since GHC is really the only widely used compiler, there's less incentive to go through the hardwork to produce a new report
2020-09-21 14:48:06 <merijn> John20: There's like 2.5 actively maintained Haskell2010 compilers
2020-09-21 14:48:15 <merijn> And GHC has 99% market share
2020-09-21 14:48:37 <merijn> If there was a competitor to GHC there'd be more motivation to produce a new report to standardise the various compilers
2020-09-21 14:48:43 → Sanchayan joins (~Sanchayan@106.201.114.230)
2020-09-21 14:49:12 <glguy> cheater: what changed on windows recently that makes you ask? I haven't heard any vty/Windows news myself.
2020-09-21 14:49:16 <mmaruseacph2> what's the timeframe?
2020-09-21 14:49:22 <mmaruseacph2> bad window, sorry
2020-09-21 14:49:30 → pingiun joins (~pingiun@ip-213-124-184-182.ip.prioritytelecom.net)
2020-09-21 14:49:33 <tdammers> problem with coming up with a competitor is the economics of it
2020-09-21 14:49:55 <tdammers> in order to compete, the other compiler would have to be able to do most of what GHC can do, and keep up with GHC, and then deliver some additional benefits
2020-09-21 14:50:13 <tdammers> but at this point, with GHC being a thing and open-source, it is cheaper and more efficient for most people to just contribute to GHC instead
2020-09-21 14:50:19 <merijn> tomsmeding: I hope csaba's work and massively overambitious plan to separate out the Core backend would allow alternate Haskell implementation that compile to Core and share code gen with GHC
2020-09-21 14:50:26 <merijn> eh
2020-09-21 14:50:33 <merijn> s/tomsmeding/tdammers
2020-09-21 14:50:33 <tdammers> maybe what we need to do is make the GHC proposals procedure more difficult and cumbersome
2020-09-21 14:50:54 <tdammers> hmm, that does sound like a good plan
2020-09-21 14:51:18 <tdammers> massively ambitious indeed, and also subject to problems with Core itself not being properly version-managed
2020-09-21 14:51:21 <dolio> There were multiple compilers for many years without new reports, though.
2020-09-21 14:51:24 <cpressey> merijn: Thanks. I think I picked printf over showHex originally to get leading 0's. I guess my deeper question is, is there actually much gain performance-wise in switching from Strings to DT.Texts, if I need to call DT.pack in lots of places? Is it still justifiable to use DT.Text over String in the name of simply "being modern"?
2020-09-21 14:51:26 <merijn> tdammers: https://www.patreon.com/posts/introducing-ghc-38173710
2020-09-21 14:51:45 <merijn> cpressey: "it depends"
2020-09-21 14:51:57 <merijn> cpressey: Like, what are you outputting and when and how much
2020-09-21 14:52:09 <cpressey> I guess another way to say it is, do people laugh derisively when they see String in your code these day? :)
2020-09-21 14:52:16 <merijn> cpressey: If you're just visualising a few 100 Word8 for users it probably doesn't matter much
2020-09-21 14:52:17 <dolio> And the one new report didn't really incorporate much.
2020-09-21 14:52:48 <merijn> cpressey: plenty of stuff is still String based, depends what you use it for
2020-09-21 14:52:59 × nbloomf quits (~nbloomf@2600:1700:83e0:1f40:dceb:16d3:9c64:27d9) (Quit: My MacBook has gone to sleep. ZZZzzz…)
2020-09-21 14:54:46 <oats> for anyone who knows java/oopystuffs, is an AbstractVisitor just a shittier Traverse?
2020-09-21 14:54:52 <cpressey> merijn: Makes sense, thanks. This is definitely not very performance-sensitive, I'm more wondering about norms and "nice" ways to write it. I'll probably just leave it as is.
2020-09-21 14:55:42 <merijn> cpressey: You could write a small wrapper that goes straight to Text which can be convenient when you need to combine with lots of other Text
2020-09-21 14:56:46 → stef204 joins (~stef204@unaffiliated/stef-204/x-384198)
2020-09-21 14:57:14 → nbloomf joins (~nbloomf@2600:1700:83e0:1f40:c87d:6f52:4c16:4130)
2020-09-21 14:58:02 <dolio> oats: I think a visitor is the function you pass to e.g. traverse.
2020-09-21 14:59:08 <dolio> `T -> ()` basically.
2020-09-21 14:59:22 <oats> right, thanks
2020-09-21 14:59:26 <dolio> Or `T -> M ()`, since the former is useless.
2020-09-21 14:59:47 <oats> trying to read some java that's making me want to never venture near java ever again
2020-09-21 14:59:53 → heatsink joins (~heatsink@107-136-5-69.lightspeed.sntcca.sbcglobal.net)
2020-09-21 15:00:01 × tronical quits (~tronical@185.244.214.216) ()
2020-09-21 15:00:07 <dolio> There are multiple patterns that could just be functions, I think.
2020-09-21 15:01:24 → bahamas joins (~lucian@unaffiliated/bahamas)
2020-09-21 15:02:29 × dale quits (dale@unaffiliated/dale) (Remote host closed the connection)
2020-09-21 15:03:17 → dale joins (dale@unaffiliated/dale)
2020-09-21 15:04:12 × heatsink quits (~heatsink@107-136-5-69.lightspeed.sntcca.sbcglobal.net) (Remote host closed the connection)
←Prev  Next→
Page 1 .. 116 117 118 119 120 121 122 123 124 125 126 .. 5022

All times are in UTC.