|
01:04
frost joined
01:14
razetime joined
02:07
zacts joined
03:05
zacts left
03:16
Guest35 left
03:33
guifa joined
04:01
zacts joined
05:00
zacts left
|
|||
| deadmarshal | I've read that Raku is slower than Perl. why is that? Is it because it's based on jvm and not written in C? or what? | 11:13 | |
|
11:24
discord-raku-bot left,
discord-raku-bot joined
|
|||
| gfldex | deadmarshal: It is slower because it comes with more batteries and had less time to be optimised. Raku is catching up and is faster for some workloads already. | 11:37 | |
| deadmarshal | great | 11:42 | |
|
12:06
zacts joined
12:20
razetime left
12:33
razetime joined
|
|||
| Anton Antonov | <@788061414769164348> I think the techniques in Big Ball of Mud (BBM) apply to your problem (www.laputan.org/mud/). If you do not want to examine each file and its relation to others -- and that is totally fine! -- then you should come up with some strategy for automatic "name spacing" (and implement it in Raku.) I think "SHEARING LAYERS" from BBM applies the most to what you describe. | 12:39 | |
| I think you should make suitable Raku roles. A discussion on similar issue can be found here: stackoverflow.com/q/67564394 . | 12:41 | ||
| I think you should make suitable Raku roles. A discussion on a similar issue can be found here: stackoverflow.com/q/67564394 . | 12:43 | ||
| <@127299883806883840> For the data science / machine learning projects I have in Raku (recommenders, prefix trees, data wrangling) I find to Raku to be between 2 and 10 times slower than "established" players, like, Mathematica, Python, or R. | 12:45 | ||
| <@127299883806883840> For the data science / machine learning projects I have in Raku (recommenders, prefix trees, data wrangling) I find to Raku to be between 2 and 10 times slower than "established" players, like, Mathematica, Python, or R. (And sometimes 100 times slower...) | 12:46 | ||
| <@127299883806883840> For the data science / machine learning projects I have in Raku (recommenders, prefix trees, data wrangling) I find Raku to be between 2 and 10 times slower than "established" players, like, Mathematica, Python, or R. (And sometimes 100 times slower...) | 12:47 | ||
|
13:09
TempIRCLogger joined
13:59
zacts left
14:04
Guest35 joined
14:44
frost left
15:28
razetime left
15:38
razetime joined
|
|||
| guifa | deadmarshal as one example: perl has a much faster start up time. But as soon as you import the OO modules, the start up time is much more similar. It's just that Perl doesn't require OO, but it's a fundamental part of Raku | 15:47 | |
| deadmarshal | guifa: right good point | 16:01 | |
|
16:29
wingfold joined
16:44
wingfold left
16:45
razetime left
18:29
guifa left
|
|||
| MasterDuke | deadmarshal: fyi, lizmat has been doing a lot of work recently on some micro-optimizations for startup and compilation. startup probably won't be as fast as perl even at the end, but it's being worked on | 18:43 | |
| deadmarshal | MasterDuke: Great! can't wait for it | 18:49 | |
|
22:15
saint- joined
|
|||
| klebs | it wasn't so bad after all -- i restructured the repository a bit so that it could be easily reorganized. Basically, all that had to be done was to create separate directories for each language, treat every file within as a unit module, and update the use statements throughout. It was a nontrivial operation, but not all that invasive, ultimately. | 22:30 | |
| I do have another issue, though. It shows up as the following error: `===SORRY!===Object of type X in QAST::WVal, but not in SC` | 22:32 | ||
| any idea what this means? | |||
| it wasn't so bad after all -- i restructured the repository a bit so that it could be easily reorganized. Basically, all that had to be done was to create separate directories for each language, treat every file within as a unit module, and update the use statements throughout. It was a nontrivial operation, but not all that invasive, ultimately. flat hierarchies, as long as they remain a certain siz | 22:54 | ||
| it wasn't so bad after all -- i restructured the repository a bit so that it could be easily reorganized. Basically, all that had to be done was to create separate directories for each language, treat every file within as a unit module, and update the use statements throughout. It was a nontrivial operation, but not all that invasive, ultimately. flat hierarchies, as long as they remain a certain siz | 22:57 | ||
| saint- | Hi, I was wondering what was wrong with trying to use this grammar to parse into just separate lines? www.toptal.com/developers/hastebin...orefon.lua | ||
| Even something like this isn't working www.toptal.com/developers/hastebin...enopop.lua | 23:04 | ||
| Whoa the single quotes were screwing things up | 23:14 | ||