[00:35:39] <bmp> We do have a `HOOK:` keyword
[00:35:43] <bmp> s/keywod/keyword
[00:35:46] <bmp> But not really my point. :-p
[01:23:14] <imode> been working on "not garbage collection" for my lang.
[01:24:09] <imode> things don't get deleted until all references to them have been removed from the stack.
[01:25:22] <imode> references only apply to blocks, and all references to a block are linked by a doubly linked list. this means you can push large chunks of data around by reference, but modifying them always copies their data.
[01:43:10] * mwgkgk [mwgkgk!uid243904@gateway/web/irccloud.com/x-ozhpeuhthfjtaddj] has joined the channel.
[01:50:03] * Sgeo [Sgeo!~Sgeo@ool-18b982ad.dyn.optonline.net] has quit (Read error: Connection reset by peer).
[01:52:09] * Sgeo [Sgeo!~Sgeo@ool-18b982ad.dyn.optonline.net] has joined the channel.
[01:57:57] * zolk3ri [zolk3ri!~zolk3ri@gateway/tor-sasl/zolk3ri] has joined the channel.
[02:12:34] * ptrkriz [ptrkriz!~ptrkriz@81.4.110.225] has quit (Ping timeout: 260 seconds).
[04:01:07] * mwgkgk [mwgkgk!uid243904@gateway/web/irccloud.com/x-ozhpeuhthfjtaddj] has quit (Quit: Connection closed for inactivity).
[04:37:39] * Sgeo_ [Sgeo_!~Sgeo@ool-18b982ad.dyn.optonline.net] has joined the channel.
[04:40:16] * MDude [MDude!~MDude@71.50.47.112] has quit (Ping timeout: 240 seconds).
[04:40:36] * Sgeo [Sgeo!~Sgeo@ool-18b982ad.dyn.optonline.net] has quit (Ping timeout: 240 seconds).
[05:59:40] <the_cuckoo> imode, i think the standard term for that would be "copy on write" 
[06:06:03] <imode> yup!
[06:07:22] <the_cuckoo> sounds like fun - did the same thing recently for my video processing library - before that, it was a manual thing you had to before you modified an image
[06:08:31] <the_cuckoo> what is your language anyway?
[06:08:44] <imode> 's called feather.
[06:08:59] <imode> it's a scheme-y lisp-y concatenative-y thing with no distinction between compile and runtime.
[06:09:47] <the_cuckoo> interpreted or a hybrid thing?
[06:10:06] <imode> largely interpreted, but I'm working on a compiler.
[06:10:18] <imode> kinda has to be. 99% of the syntax is derived.
[06:10:32] <the_cuckoo> hmm - aiming for jit for that?
[06:10:34] <imode> there are 8 core primitives. 21 in total if you want the "full power" of the language.
[06:10:46] <imode> yeah, I'm hoping.
[06:10:57] <imode> in the worst case a tight dispatch loop won't hurt.
[06:11:23] <the_cuckoo> yeah - kinda looking that way in the long term for my stuff
[06:11:44] <imode> jitting can get complicated unless you have a really tight and minimal core.
[06:12:44] <the_cuckoo> in the short time, i try to 'optimise' my looped token streams down to their implied types (int, double etc) and function pointers
[06:13:21] <the_cuckoo> it's not like blindingly fast or anything, but it's not too bad either
[06:13:40] <the_cuckoo> which language are you implementing in?
[06:14:01] <imode> currently Python, but my intention is to bootstrap a small C interpreter, which can be used to generate native machine code.
[06:14:07] <imode> and also an implementation in Go.
[06:15:48] <imode> let's see if I can snag you a bit of syntax...
[06:16:01] <the_cuckoo> cool :) - my language doesn't really lend itself to multiple implementations :) - it's an embeddable c++ library
[06:16:15] <the_cuckoo> sounds good - am interested
[06:16:33] <imode> https://hastebin.com/rucuquluwo.txt 
[06:16:43] <imode> an example of defining if ... then ... else ... end blocks
[06:17:00] <imode> you can do something similar with while ... do ... end
[06:17:20] <imode> https://hastebin.com/yayosejifi.txt an example of what assembling a full block looks like in the REPL.
[06:18:10] <imode> https://hastebin.com/ihasuyedan.txt and here's the definitions required to build conditional blocks, recursion modifiers, and a while loop.
[06:18:56] <the_cuckoo> [[[:left: :right:] <- local vars?
[06:19:13] <imode> nope, stack modifiers.
[06:19:29] <imode> left takes something on the bottom of the stack and throws it on top. right takes something on the top and throws it on the bottom.
[06:19:41] <the_cuckoo> ah - ok
[06:19:47] <imode> the stack is technically a tape.
[06:22:01] <the_cuckoo> i see - my approach is quite different in that i implement my logical branches and loop parsing/execution in the 'host language' rather than try to define a parser which allows the implementation iyswim
[06:23:07] <imode> yeah, that makes sense. imo it's better to have that stuff be "built-in" or at least baked in at parse time. I'm fiddling with church numerals and shit.
[06:23:07] <the_cuckoo> (i think i probably could implement some of those constructs in the language too, but would definitely execute slower)
[06:24:01] <the_cuckoo> yeah - think it's mostly about compromise and where you want to put the complexity - in the parser or in the parsed code
[06:24:26] <imode> my parser is splitting on whitespace. :p
[06:24:28] <the_cuckoo> in the latter, it may simplify reimplimentations of another parser
[06:24:51] <imode> as a consequence, I have no string literals. so if you want that... implement an evaluator. :P
[06:24:54] <the_cuckoo> yes :) - that's the great thing about concanetive languages - that's all they need to do:)
[06:25:49] <the_cuckoo> heh - i have a 'parse' type of thing which consumes the next token in the input string and puts it on the stack as a string
[06:26:13] <the_cuckoo> so basically, a use of that is the simplest: $ "hello world"
[06:26:53] <imode> your parser must know how to differentiate between quoted string literals, though.
[06:27:05] <imode> i.e knowing that the space between those two words doesn't count as a token separator.
[06:28:12] <the_cuckoo> yes - if i see a " as the first char, i keep scanning/concatanating until the closing " followed by whitespace - also [should] allow nested \" combinations
[06:28:34] <the_cuckoo> if the closing quote isn't followed by whitespace, it's an error
[06:29:04] <imode> huh. I treat that as a separate token.
[06:29:10] <imode> ...that's a neat corner case.
[06:29:49] <the_cuckoo> my first use is from the command line :) - command line parsing is pretty much as i describe it there
[06:29:57] <imode> neat. a"b"c yields [97 98] c
[06:30:24] <the_cuckoo> i would see that as an error :) - unless a token of a"b"c exists :)
[06:30:25] <imode> that's just from my first pass junk parser. really the language only supports whitespace delimited tokens.
[06:31:52] <imode> the intent behind feather was initially a fully sandboxed language for users to run code in.
[06:33:00] <the_cuckoo> similar goal with mine :) - the base implementation only understands numerics, strings, arrays and stacks :) - it can't even reference files :)
[06:33:26] * ephe_meral [ephe_meral!~amnesia@2a02:8109:8880:2004:f098:4989:c1f3:8ee] has joined the channel.
[06:34:18] <the_cuckoo> (the exception to that are loading script files, but only from very specific directories)
[06:34:34] <the_cuckoo> https://gitlab.com/lilo_booter/rpany
[06:35:00] <imode> oh hey I've actually seen this.
[06:35:04] <imode> neat shit.
[06:35:46] <the_cuckoo> i think it's pretty :p
[06:43:12] * ephe_meral [ephe_meral!~amnesia@2a02:8109:8880:2004:f098:4989:c1f3:8ee] has quit (Ping timeout: 260 seconds).
[06:46:57] <the_cuckoo> imode: i actually wrote it to replace a testing tool which i wrote for the day job - it's all about video processing, and it implements a similar language, but in a really embarassing hacky way :)
[06:47:27] <the_cuckoo> rpany allowed me to replace it with https://gitlab.com/lilo_booter/vml-batch 
[06:47:45] <imode> siiiick.
[06:48:58] <the_cuckoo> there's a ton of stuff that's not mentioned there :) - filters to do most things you need to manipulate video for example (barring 3d stuff, but it's on the way :))
[06:49:57] <imode> damn.
[07:02:02] <the_cuckoo> (added a little more info the usage summary on that vml-batch link there - indicates where the rpn kicks in for building video graphs)
[08:01:13] * zolk3ri [zolk3ri!~zolk3ri@gateway/tor-sasl/zolk3ri] has quit (Remote host closed the connection).
[08:32:39] * Sgeo_ [Sgeo_!~Sgeo@ool-18b982ad.dyn.optonline.net] has quit (Read error: Connection reset by peer).
[09:35:03] * tme5 [tme5!~tme5@185.214.220.196] has joined the channel.
[11:52:54] * imode [imode!~linear@unaffiliated/imode] has quit (Ping timeout: 272 seconds).
[13:08:28] * tmew5 [tmew5!~tme5@185.214.220.209] has joined the channel.
[13:09:04] * tme5 [tme5!~tme5@185.214.220.196] has quit (Read error: Connection reset by peer).
[13:09:23] * tmewett [tmewett!~tme5@185.214.220.197] has joined the channel.
[13:12:45] * tmew5 [tmew5!~tme5@185.214.220.209] has quit (Ping timeout: 240 seconds).
[14:57:36] * Sgeo [Sgeo!~Sgeo@ool-18b982ad.dyn.optonline.net] has joined the channel.
[15:47:15] * MDude [MDude!~MDude@71.50.47.112] has joined the channel.
[17:51:05] * tmewett [tmewett!~tme5@185.214.220.197] has quit (Quit: Leaving).
[19:00:58] * imode [imode!~linear@unaffiliated/imode] has joined the channel.
[21:41:36] * [1]MrMobius [[1]MrMobius!~MrMobius@208.58.206.154] has joined the channel.
[21:43:32] * MrMobius [MrMobius!~MrMobius@208.58.206.154] has quit (Ping timeout: 256 seconds).
[21:43:33] * [1]MrMobius is now known as MrMobius