!CENmKDnMngVwJJrTry:matrix.org

Inko (lang)

23 Members
General purpose chat room about the Inko programming language.1 Servers

Load older messages


SenderMessageTime
19 Jun 2020
@mathnerd314:matrix.orgmathnerd314 joined the room.23:10:57
20 Jun 2020
@hellonomad:matrix.orgkiti_nomadIf you have learned powershell, you can pack a copy for windows15:24:57
@hellonomad:matrix.orgkiti_nomadNow, many package management tools have appeared on the windows platform, such as choco, scoop, winget15:25:58
@yorickpeterse:matrix.orgyorickpeterseMy knowledge of Windows is very limited, so I'd rather leave that to those with more experience :)22:53:15
29 Jun 2020
@yorickpeterse:matrix.orgyorickpeterseGetting pretty close to the end of type inference/checking19:34:27
@yorickpeterse:matrix.orgyorickpeterseClosures, method calls, and throwing values remain19:34:37
@yorickpeterse:matrix.orgyorickpeterseof course those all happen to be difficult ones, but at least the end is in sight19:34:50
@yorickpeterse:matrix.orgyorickpeterse random example: the Ruby compiler lets you run blocks by sending call to them. But this isn't an actual method though, it just looks like one; defining one for every closure would be too expensive 19:42:09
@yorickpeterse:matrix.orgyorickpeterse So now I'm thinking of removing that, in favour of using () (so e.g. callback() opposed to callback.call) 19:42:45
@yorickpeterse:matrix.orgyorickpeterse Since Inko is now statically typed, it doesn't need to support dynamic calls to call either, so we can drop the whole idea 19:43:05
@yorickpeterse:matrix.orgyorickpeterse downside is that when you look at foo() you don't know if it's a method call, or runs a closure 19:43:32
@yorickpeterse:matrix.orgyorickpeterse also downside is that if you do just foo it wouldn't run the closure, so you'd have to use foo(), which is different from methods 19:46:46
@yorickpeterse:matrix.orgyorickpeterse * also downside is that if you do just foo it wouldn't run the closure, so you'd have to use foo(), which is different from methods 19:46:58
@yorickpeterse:matrix.orgyorickpeterse oh and ugh, we'd have to deal with things like @some_attribute(). I think I will just stick with call for now, but treat it as syntax sugar 19:51:35
1 Jul 2020
@hellonomad:matrix.orgkiti_nomadSome simple example functions, you always named foo or bar18:56:34
@hellonomad:matrix.orgkiti_nomadIs there any allusion?18:56:50
@hellonomad:matrix.orgkiti_nomadIn Chinese universities, the textbooks used to learn programming usually name functions as a(), b(), c()...18:58:12
@hellonomad:matrix.orgkiti_nomadThis led some people to think of learning programming as a math problem18:59:23
2 Jul 2020
@yorickpeterse:matrix.orgyorickpetersefoo, bar, baz, etc are just commonly used example names; they don't mean anything12:57:45
@hellonomad:matrix.orgkiti_nomadI known it17:45:37
5 Jul 2020
@hellonomad:matrix.orgkiti_nomadI wrote a short promotion for inko, the largest IT information platform in China15:13:58
@yorickpeterse:matrix.orgyorickpeterse kiti_nomad: Thanks! :) 17:57:17
7 Jul 2020
@briley899:matrix.org@briley899:matrix.org joined the room.17:36:17
@briley899:matrix.org@briley899:matrix.org left the room.22:19:20
8 Jul 2020
@alani367:matrix.org@alani367:matrix.org joined the room.11:25:53
@alani367:matrix.org@alani367:matrix.org left the room.11:40:06
13 Jul 2020
@yorickpeterse:matrix.orgyorickpeterseSpent a week or so re-working parts of the interpreter loop. Specifically, instructions are now fixed-width (16 bytes) and no longer rely on separate heap-allocated vectors for arguments19:57:50
@yorickpeterse:matrix.orgyorickpetersePrior to this they were 32 bytes, plus the vector space (2 bytes per argument)19:58:11
@yorickpeterse:matrix.orgyorickpeterseNow they are always 16 bytes, with a maximum of 6 arguments19:58:17
@yorickpeterse:matrix.orgyorickpeterseThis required a bunch of changes in the compiler so its register allocator is no longer as braindead as it used to be19:58:37

There are no newer messages yet.


Back to Room List