by

Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Repo info
Activity
    Siddhant Sanyam
    @siddhant3s
    @gabfssilva , https://scalameta.org/metals/docs/editors/overview.html says yes.
    We should call them "sheetmetal" :P
    Gabriel Volpe
    @gvolpe

    hmmm, intereresting the -Dmetals.client=coc-metals was actually taken out a while ago, so that must not be the newest version?

    old version of what, coc.nvim? I can have a look at this

    because coc-metals updates itself, right? @ckipp01
    Gabriel Francisco
    @gabfssilva
    @siddhant3s they do work, but it seems that I'm unable to find how to run them. I mean, for now metals only knows how to compile them?
    Siddhant Sanyam
    @siddhant3s
    image.png
    @gabfssilva Are you saving them as "worksheet.sc" files? I have to hit save to run them.
    Chris
    @ckipp01
    @gvolpe scratch that, I have no sense of "oldness" since I'm always on a snapshot. It happened last version, so it's not a big deal. And regarding automatic update, that depends on how you install it actually
    recently coc.nvim changed this, and the used to update daily, but that's not the case anymore I believe and if you do an install with CocInstall coc-metals it doesn't do it automatically anymore
    which reminds me that I need to update the docs to change this
    you now need to manually do a CocUpdate
    but if you install it with vim-plug for example, then you'd need to update it with that like any other plugin
    Gabriel Volpe
    @gvolpe
    okay, thanks for letting me know! It used to get automatically updated because I got upgraded to 0.9.0
    Chris
    @ckipp01
    o really? That was released this morning, so I guess it is still working!
    Gabriel Volpe
    @gvolpe
    but it didn't upgrade to 0.9.1 yet so I'm not sure how that works
    Chris
    @ckipp01
    do you mean server version or coc-metals version?
    Gabriel Volpe
    @gvolpe
    emm Metals version 0.9.0
    not sure what the coc-metals version is, I probably ned to update that via CocUpdate is what you're saying?
    Chris
    @ckipp01
    yea, if you update that, it should put you at 0.9.0 of coc-metals, then that will prompt you to update the server to 0.9.1
    they are close right now which is annoying since they are completely independent versions
    Gabriel Volpe
    @gvolpe
    oh nice
    yeah that got me confused, I thought they were the same
    Chris
    @ckipp01
    :grimacing: yes, you're not the first one
    Gabriel Francisco
    @gabfssilva
    @siddhant3s, oh, ok, I was using only .sc instead of .worksheet.sc
    It worked now, thanks. :)
    Gabriel Volpe
    @gvolpe
    what is the latest version @ckipp01 ? the latest on nixpkgs is https://github.com/NixOS/nixpkgs/blob/master/pkgs/misc/vim-plugins/generated.nix#L522
    maybe it needs an update
    Chris
    @ckipp01
    https://github.com/scalameta/coc-metals/releases
    as of this morning it's 0.9.0
    how does that list get updated?
    Gabriel Volpe
    @gvolpe
    via PRs, not that fast
    oh it's under https://github.com/scalameta/coc-metals now
    definitely needs an update :D
    anyway I can get onto the latest one by overriding that version in my nixpkgs
    I will try updating next week (going away now), thanks for your help :)
    Chris
    @ckipp01
    :thumbsup:
    James Cosford
    @jamescosford

    @ckipp01 I mean that previously, metals/vscode would do auto imports for me, for example in one file it would look like this:

    package com.geomoby.mozambique

import com.geomoby.mozambique.model.Common.Point

    Where ...Common.Point is also in com.geomoby.mozambique but maybe in another sub-project, but then suddenly I found myself with lots of errors, saying that Common.Point is not in package com.geomoby.mozambique.com.geomoby.mozambique, and the imports had to be changed to

    package com.geomoby.mozambique

import _root_.com.geomoby.mozambique.model.Common.Point

    Ultimately that's not a huge problem, but it was a strange phenomenon.

    According to my git log I hadn't changed any project settings, hadn't messed with imports at all.
    DavidGoodenough
    @DavidGoodenough
    How do I get the message from the unused import check to tell me which of a {} list if imports is missing?
    Chris
    @ckipp01
    AFAIK, that's currently not possible and one good reason to not used grouped imports. The diagnostic coming from the compiler when using -Ywarn-unused just gives the line of the import without specifying which in the group it is.
    again I think that's the case
    DavidGoodenough
    @DavidGoodenough
    @ckipp01 thanks, I will stop using grouped imports. Sad though, it made the code neater.

    Why is the compiler complaining that id in the second line is unused,
    when it is used in the third line?

        for(party <- da.participantParty;
        id <- party.party.partyIdentification; 
        if(id.id.schemeId.getOrElse("?") == ui.App.app.company.get.schemeId)) yield party

    participantParty and partyIdentification are both lists

    Ólafur Páll Geirsson
    @olafurpg
    @DavidGoodenough @ckipp01 compiler diagnostics actually have three offsets: start, point and end. The “point” offset correctly points to the unused identifier in imports but the diagnostics we get from bloop only include start and end.
    It would be possible to adjust the positions of unused import diagnostics to become offsets that only use “point”. The SemanticDB compiler plugin has this logic built in to support RemoveUnused in Scalafix.
    Chris
    @ckipp01
    ah, my mistake!
    DavidGoodenough
    @DavidGoodenough
    @olafurpg adjusting the position would seem the best route to me.
    Raitis Veinbahs
    @siers
    If a dependency provides a configuration file project/src/main/resources/applicaion.conf in its .jar and a config file in my repo is including it, then would it be metals' responsibility to find it?
    Eric Peters
    @er1c
    FWIW, I can run metals perfectly in VS Code on a Chromebook laptop
    other than the fact the chromebook has barely any ram :) next step is the remote-ssh testing :)