Below is a of the discussion from this week’s JavaScript chat (agenda, Slack Transcript)
Have a topic for discussion for the next meeting? Leave a suggested edit on next week’s agenda.
Editor-specific .gitignore entries in Gutenberg
- Context: https://github.com/WordPress/gutenberg/pull/18257#discussion_r342255694
- Proposing to add a gitignore entry for VSCode’s “.vscode” local folder
- Historically we’ve avoided editor-specific entries in the project’s gitignore to avoid opening pandora’s box (proliferation of ignore entries), but is this too rigid a guideline causing friction in new contributor onboarding?
- Question: Should we allow more of these entries, and if so, which?
@gziolo said that he does not have a strong opinion on this item, but WordPress as a project is strict about it.
@mkaz also does not feels strong either way but said he would probably merge the PR because vscode its perhaps the largest and most popular editor.
@chancethedev shared his perspective as an instructor:
I often include vscode settings in my repos to make things easier for students when running workshops, so ignoring it globally won’t work for me. That said, I can just remember to remove it before I commit so it’s not a HUGE deal, I can just imagine folks slipping up often enough to where it’ll end up in commits from time-to-time and we’ll just have the conversation again.
@aduth and @adamsilverstein asked if the fact that WordPress os strict is written down somewhere?
@gziolo shared the following comment from https://core.trac.wordpress.org/ticket/34345:
IDE-specific .gitignore rules should not be present in a project. They should be managed in the user’s ~/.gitignore_global.
@gziolo added that it does not look like it is a written rule, as he was not able to find it in the docs. So, we can go both ways; we should document it to make it easy for contributors to update patches.
@nerrad said :
I lean towards adding the rule and just evaluating on a case-by-case basis.
It seems to be more dev friendly and avoids unnecessary noise in reviews.
@mkaz asked if there is any downside to adding the rule and @aduth answered:
I don’t think there’s a downside in-and-of-itself, more the slippery slope of how we make these decisions in the future without a well-defined approach.
@adamsilverstein suggested as action item working on some language for HOW we decide.
@mkaz and @gziolo suggested we add 3-5 top popular IDEs. @gziolo said we could also link to docs how to add an IDE to ~/.gitignore_global.
@aduth suggested, and @adamsilverstein supported that we can base it on “X% or greater” of some well-defined data, e.g., https://insights.stackoverflow.com/survey/2019#technology-_-most-popular-development-environments.
@chancethedev said he is joining the docs team specifically for Gutenberg docs, so he is happy to bring this up in the docs meeting today as well.
If you have any thoughts on this, please share them.
Bundling and loaders – extend core to allow SVG components
As part of the open floor, @mkaz introduced the topic by saying:
For the open floor, I’m interested in the topic above regarding bundling and loaders – trying to extend core to allow SVG components.
Parcel sounds like an interesting option but looks like v2 is not out yet, so would need to wait
@aduth asked clarification on the background of this task and what is trying to be achieved by importing an SVG.
@mkas linked to the ticket https://github.com/WordPress/gutenberg/issues/14628 and said:
Right now the Social Links duplicates the SVG in CSS and PHP to work – we want to figure out some way to get it to a single .svg file
@aduth asked if it’d be feasible to manually shove the SVG into JSX. And asked if not an option to use something like <img src="icon.svg">
.
@mkaz answered that the problem for the publishing view is KSES, and linking to images, one loses the ability to color them.
@aduth shared the main questions he thinks we should answer are “Should we do it, and if so, how?” and shared some thoughts :
- I am conscious of the size inflation, though I think we need to be solving this problem anyways. Could be a way to prompt us to investigate better handling of “large” blocks (opening path for more complex blocks too, like reintroducing Code+SyntaxHighlighting).
- I guess there is some issue with our build that the Webpack loader might not work? The Babel transform seems less ideal, but also workable.
@gziolo shared the following analysis of what creat react app did:
CRA experiments with Babel macros, but the blocker at the moment is cache invalidation. They had a very solid proposal to inline many types of files, including SVG with Babel macros https://github.com/facebook/create-react-app/issues/3722#issuecomment-458124126.
Macros struggle with the cache invalidation for the dependent files. See the comment in create-react-app https://github.com/facebook/create-react-app/commit/11737bc78676868e76ce5d0c04ddbb5c758d3908#diff-b1524f1e1ccabdb327a12f23b432b20bR13.
Related Babel issue which might resolve this issue in the future: https://github.com/babel/babel/issues/8497.
@mkaz said Parcel 2 sounds like a potential option for the future. Then, asked if we want to add SVGR support to wp-scripts without it in core Gutenberg, relevant PR https://github.com/WordPress/gutenberg/pull/18243. @gziolo answered that we use a custom config in Gutenberg, so it should be fine.
#core-js, #javascript, #meeting-notes