profile
viewpoint
If you are wondering where the data of this site comes from, please visit https://api.github.com/users/salutis/events. GitMemory does not store any data, but only uses NGINX to cache data for a period of time. The idea behind GitMemory is simply to give users a better reading experience.
Rudolf Adamkovič salutis Internets Perfectionist.

eggheadgames/apple-receipt 2

Swift receipt framework to retrieve iOS app purchase date

eggheadgames/storyboard-lint 2

Scan Xcode iOS storyboard files for potential problems.

salutis/privacy-policy 2

https://salutis.github.io/privacy-policy/

salutis/swift-array-safe-subscripting 2

Safely retrieve array values by index

eggheadgames/aws-sdk-ios-mobileanalytics 0

AWS iOS SDK for Mobile Analytics, and dependancies - Just the files that are needed for Analytics

eggheadgames/RMStore 0

A lightweight iOS library for In-App Purchases

issue commentzevlg/telega.el

Remove triangles from mode names

@zevlg I see! Thank you so much for the explanation.

salutis

comment created time in 4 days

issue commentzevlg/telega.el

Remove triangles from mode names

Today, I found telega-symbol-mode described as "*String used for telega modes." It equals "◁" by default. Perhaps the lighters should use that? For completion, I also looked for all the places with a hard-coded "◁" symbol:

elpa/telega-20210914.1321/telega-chat.el:(define-derived-mode telega-chat-mode nil "◁Chat"
elpa/telega-20210914.1321/telega-root.el:  `("◁Root" (telega-root-auto-fill-mode
elpa/telega-20210914.1321/contrib/telega-mnz.el:  :lighter " ◁Mnz"
elpa/telega-20210914.1321/telega-modes.el:  :lighter " ◁Squash"
elpa/telega-20210914.1321/telega-modes.el:        (concat "◁Image" (when image-type (format "[%s]" image-type))))
elpa/telega-20210914.1321/telega-modes.el:  :lighter " ◁Edit"
elpa/telega-20210914.1321/telega-modes.el:  :lighter " ◁Highlight"
elpa/telega-20210914.1321/telega-customize.el:(defcustom telega-symbol-mode "◁"
salutis

comment created time in 4 days

issue commentminad/corfu

Suggestion: "soft" highlight first implicit candidate when none are selected

@jdtsmith I apologize for unclear communication. When I hit RET, I expect Emacs to insert a newline into the buffer. To put it differently, Corfu should either (1) highlight the first candidate to communicate its pre-selection to the user or (2) avoid binding the first candidate to RET by default. Personally, I would go for (1), as seen virtually everywhere else, including in all mainstream IDEs.

jdtsmith

comment created time in 4 days

pull request commentbdarcus/bibtex-actions

open file refinements

That test.pdf file should only be listed once, and when chosen, should open either internally or externally depending on how the bibtex-actions-file-extensions* vars are set. So I can't just do delete-dups on the list returned by bibtex-actions-file--possible-names.

If pdf figures in both bibtex-actions-file-extensions and bibtex-actions-file-extensions-external, I would expect the file show twice, e.g. as /tmp/test.pdf and once /tmp/test.pdf (external).

bdarcus

comment created time in 6 days

issue openedemacs-grammarly/flymake-grammarly

Allow the user to ignore some Grammarly warnings

I would like to ignore the following warning:

"Provides synonyms for repeated words"

Perhaps

created time in 8 days

issue commentemacs-grammarly/flymake-grammarly

Add "flymake-grammarly-unload"

@jcs090218 Thank you!

salutis

comment created time in 8 days

startedmagit/magit

started time in 10 days

issue commentminad/vertico

vertico-directory feature: restart a path

For the future readers, here is how to enable path "restarting" à la Ido/Ivy:

Unfortunately, this breaks savehist-mode, i.e. the recently opened files do not show in the list. For instance, when I visit ~/org/ in C-x C-f, I see the last n files I opened in the past. For ~/org/~/org/, I do not. Anyone knows how to solve the problem?

jdtsmith

comment created time in 11 days

issue commentbdarcus/bibtex-actions

PDF file always open externally, even with pdf-tools installed.

Ok, I'll modify the behavior of open to match open-library-file, and maybe set external to nil by default.

Thank you. I also have the same problem as @pRot0ta1p.

pRot0ta1p

comment created time in 13 days

issue commentminad/corfu

Fix slow auto-completions and respect user-preferred maximum width

@minad Thank you! I am looking forward to using Corfu 0.12 with a sufficiently wide completion window! As for flyspell, I replaced it with Flymake + Vale + Hunspell. Even better. Again, thank you so much for helping me out.

salutis

comment created time in 13 days

startederrata-ai/vale

started time in 13 days

issue openedemacs-grammarly/flymake-grammarly

Add "flymake-grammarly-unload"

We have flymake-grammarly-load but no flymake-grammarly-unload.

created time in 14 days

starteddalanicolai/sketch-mode

started time in 14 days

issue commentbdarcus/bibtex-actions

How should new "open" commands work?

If I remove those items from the list, the downside is you can't get them back, without exiting and opening a different command.

This confuses me. Why would I want to get them back? If I say "open link," why would I want to see the items with no links? Would I want to press RET on them to see the error message, or what?

bdarcus

comment created time in 24 days

issue commentbdarcus/bibtex-actions

How should new "open" commands work?

But you're missing something fundamental in my question: how do you open multiple entries in the same file?

Oh, I forgot that not everyone stores one BibTeX entry per file, like I do. Still, Emacs does not couple files and windows together, like modern editors do, so we can open multiple windows at the correct positions, right?

I could remove ability to select multiple keys on the open commands, but have a feeling many people would think that less than ideal, probably me included.

Personally, I would prefer single-item selection by default and multi-item selection on C-u or something along those lines. I can live with pressing RET twice everywhere, but it gets old.

bdarcus

comment created time in 24 days

issue commentbdarcus/bibtex-actions

How should new "open" commands work?

@bdarcus

The affixation function in Emacs 28 shows you which items have links etc, but that's not helpful if you're using an earlier version.

In my opinion, if RET results in an error ("No link found for ..."), the item should be either (1) disabled for selection with RET or (2) hidden from the list altogether. As it stands, RET does nothing but resets the context for the user, which does nothing but annoys them, no?

Opening multiple files or links makes sense, but not multiple BibTeX entries.

How would you deal with that?

This is a false proposition. For instance, I have never opened multiple files or links, but I regularly need to open multiple BibTeX entries to improve their publisher, links, or title field. Further, I think all "open" APIs should behave in exactly the same way.

bdarcus

comment created time in 24 days

issue commentbdarcus/bibtex-actions

How should new "open" commands work?

@bdarcus

Also, the title of the issue was a question: how should this all work? The code reflects a first stab, upon which to iterate. Since this piece is all internal now, we have quite a bit of flexibility.

Yup, that is why I dumped all my "WTF moments" into the issue.

P.S. -open could perhaps also include the BibTeX entry itself. I would expect it to show everything that could be opened: links, notes, files, and the entry itself.

bdarcus

comment created time in 24 days

issue commentbdarcus/bibtex-actions

How should new "open" commands work?

@bdarcus

You're going to need to be more specific, particularly on the "fails to do so". What does that mean?

For instance, -open-link says "No link found for ..." That is correct, as there is no link. But then, why would it offer to open a link if there is none? In other words, the list should include only the BibTeX entries that actually have some link to open.

bdarcus

comment created time in 24 days

issue commentbdarcus/bibtex-actions

How should new "open" commands work?

Here are my random observations:

-open-library-files, worded in plural ("files"), offers to open one file, while -open-link, worded in singular ("link"), offers to open more links. Say what?

-open does not show any library files. Say what?

-open-library-files offers to open library files even for BibTeX entries with no library files, and then it fails to do so. Say what?

-open-link offers to open links even for BibTeX entries with no links, and then it fails to do so. Say what?

bdarcus

comment created time in 24 days

issue commenterrata-ai/vale

TokenIgnores not working

@jdkato Thank you so much for your help!

salutis

comment created time in a month

issue openederrata-ai/vale

TokenIgnores not working

I am trying to make TokenIgnores work on

bash-5.1$ vale -v
vale version 2.10.5

In the documentation, I found a TokenIgnores regexp that matches inline math$...$.

I added the regexp to my .vale.ini:

bash-5.1$ vale ls-config
{
<snip>
  "TokenIgnores": {
    "*": [
      "($+[^\\n$]+$+)"
    ]
  },
<snip>
}

Then, I tested:

bash-5.1$ echo 'This $is very interesting$!' | vale --ext=.md

 stdin.md
 1:6  warning  'very' is a weasel word!   write-good.Weasel 

✖ 0 errors, 1 warning and 0 suggestions in stdin.

Why is vale not ignoring the text between $...$?

created time in a month

issue commenterrata-ai/vale

Feat: Org mode support

+1 for Org mode support.

tecosaur

comment created time in a month

startedSerenityOS/serenity

started time in a month

startednotmuch/notmuch

started time in a month

issue commentminad/corfu

Suggestion: "soft" highlight first implicit candidate when none are selected

Chiming in from #56. This issue is a big problem for me, as Corfu intrusively changes the text I type to nonsense. I find it helpful to see the available completions at all times, but I want Corfu to never change the text I typed, except when I give it a command to do so, such as C-n and RET.

jdtsmith

comment created time in a month

issue openedminad/corfu

Disable "smart" completion on RET

I am trying to figure out how Corfu works. There is no selection in the screenshot below, and yet when I press RET Corfu completes let. Is it possible to disable this behavior? I want Corfu to complete the selected item, if I wish to select one with C-n, with zero (invisible) "smarts."

Screen Shot 2021-08-23 at 16 53 59

Perhaps related to #45.

created time in a month

startedjoaotavora/eglot

started time in a month

startedswift-emacs/swift-mode

started time in a month

startedemacs-grammarly/flymake-grammarly

started time in a month

issue commentdanielmartin/swift-helpful

Is there a plan to support eglot?

+1 for decoupling from lsp and supporting eglot.

FYI: I am using eglot with Swift (SourceKit-LSP) with no additional Emacs packages, and it has been working great.

galeo

comment created time in a month