24 May 2024 |
tgodzik#4885 | But how would that influence detection of sbt | 19:21:37 |
tgodzik#4885 | I would really need some kind of reproduction because this is too weird | 19:21:55 |
tgodzik#4885 | It's probably some edge case we haven't thought about | 19:22:08 |
edmisml | sure let me poke around a bit more. I agree it does sound like it's verging on "superstition" ha | 19:26:03 |
edmisml | Managed to reproduce the issue I'm having (https://github.com/masonedmison/metals_gradle_proj_repro). It seems that when both a build.gradle and settings.gradle are present, metals is unable to find a build target. | 20:50:36 |
tgodzik#4885 | Cool! I can take a look. Do you mind opening an issue so that I don't forget? | 20:51:32 |
edmisml | Sure thing. Also if you could point me in a rough direction of where to start in the code base, I'm happy to take a look as well. | 20:52:20 |
tgodzik#4885 | Most likely GradleBuildTool.scala is a place to look but not 100 sure | 20:53:59 |
tgodzik#4885 | Btw. Is that a build with both SBT and gradle? | 20:54:15 |
tgodzik#4885 | Or just gradle? | 20:54:22 |
tgodzik#4885 | Och I see | 20:54:46 |
tgodzik#4885 | It's a mix | 20:54:49 |
edmisml | scala and gradle -- it's a bit of a weird one | 20:55:01 |
tgodzik#4885 | We didn't forsee that case | 20:55:04 |
edmisml | fair, I imagine this is a pretty niche use case, ha. | 20:55:27 |
tgodzik#4885 | I mean there is plenty of gradle scala projects but I haven't seen a mix | 20:55:32 |
25 May 2024 |
im_bharadwajj | Redacted or Malformed Event | 06:09:27 |
26 May 2024 |
edmisml | Arg, so I figured out what was the matter. I traced down that that metals should be prompting for user input when multiple build tools are selected -- here https://github.com/scalameta/metals/blob/main/metals/src/main/scala/scala/meta/internal/builds/BuildToolSelector.scala#L43. In my case choice was evaluating to null and this was because.... drumroll .... notifications were set to "do not disturb" for metals. | 18:49:25 |
28 May 2024 |
| younguyen joined the room. | 10:50:32 |
YeOldePorkchop | I'm looking to debug into how Metals goes about formatting. In my work code it takes a super long time, we have a fairly big project structure. Mainly I just want to see which functions are being called and from where. How would I go about debugging within the Metals code itself? | 23:45:40 |
29 May 2024 |
rohith_02_ | ![image.png](https://matrix.org/_matrix/media/r0/thumbnail/t2bot.io/7cb427bc3364ffe8142c67180717e2988d341e031795700291048308736?height=360&method=scale&width=360) Download image.png | 06:14:47 |
rohith_02_ | ![image.png](https://matrix.org/_matrix/media/r0/thumbnail/t2bot.io/48c64cf47b66c4e1de10bec3c18596f68354ec931795700297566257152?height=360&method=scale&width=360) Download image.png | 06:14:48 |
rohith_02_ | Hey. Why is the indexing keep on happening continously? It's not stopping forever. And it's consuming so much of battery. How to solve this? | 06:14:48 |
rohith_02_ | * Hey. Why is the indexing keep on happening continously? It's happening forever. And it's consuming so much of battery. How to solve this? | 06:16:05 |
tgodzik#4885 | I think there is something wrong there, but wasn't high on my priority list. CMD formatting is actually much faster. The best way to go around it is to try and create flame graphs based on our benchmarks. This should indicate if there is wasted time anywhere | 08:10:00 |
tgodzik#4885 | This looks like a bug, do you see any errors in the logs? Do you perhaps use spark? There was a bug that could cause infinite indexing, because they released broken source code | 08:11:21 |
tgodzik#4885 | This is fixed on metals side in the newest snapshot | 08:11:37 |
rohith_02_ | No, I don't see anything in Logs. I started to observe this after updating to Metals v1.34.0. I thought this was bug in the interface, but my battery is draining as hell since few days. And Yes, I am using spark version 3.0.1. | 10:14:36 |
tgodzik#4885 | och yeah, that seems to be the issue. Best to update to newest snpashot. There was a bug in our tokenizer for Java that would do infinite recursion. This was caused by one of the libraries publishing broken sources | 10:20:30 |
rohith_02_ | Is this the latest Snapshot version? 1.3.1+40-cd0f91ad-SNAPSHOT | 10:28:48 |