Sign in

romainmenke

@romainmenke.bsky.social
151 followers 61 following 241 posts

Gender is a spectrum. Fixes things, mostly with code. www.romainmenke.com

PostsRepliesMedia
Reposted by romainmenke
Lea Verou, PhD @lea.verou.me · 12/10/2026
🔥 New blog post: The best way to use icons in Web UIs What if I told you, you could use CSS to *insert* and *tweak* icons without touching any HTML? And that you can manage them alongside your other design tokens? lea.verou.me/blog/2026/ic...
lea.verou.me
The best way to use icons in Web UIs • Lea Verou
What if I told you, you could use CSS to insert and tweak icons without touching any HTML? And that it could be more flexible and maintainable than what you’ve been using?
78920
romainmenke @romainmenke.bsky.social · 12/10/2026
Awesome!
000
romainmenke @romainmenke.bsky.social · 09/10/2026
Unfortunately no time for a chat. But maybe there is a github repo? You can always ping me there on issues :)
110
romainmenke @romainmenke.bsky.social · 09/10/2026
Very cool! I would change the syntax for defining objects. A more idiomatic CSS approach would be something like: ``` at-object --lamp { /* descriptors for an object */ } at-scene --scene { /* descriptors for a scene */ objects: --lamp, ...; } ```
110
romainmenke @romainmenke.bsky.social · 07/10/2026
.card-a, .card-b, .card-c We've been doing that for years. It works 🤷 Doesn't have to be sequential or logical either. We have projects with -a, -b, -c, -f, -y, -z
110
romainmenke @romainmenke.bsky.social · 03/10/2026
postcss-preset-env now includes the mixins and custom functions #postcss plugins: preset-env.cssdb.org/blog/postcss... Turns out that it is also possible to implement one of the gamut mapping algorithms as a custom function!
preset-env.cssdb.org
PostCSS Mixins And PostCSS Custom Functions | CSS Tools
Use @mixin and @function in CSS
1102
romainmenke @romainmenke.bsky.social · 03/10/2026
But also style queries in `if` query the element itself.
010
romainmenke @romainmenke.bsky.social · 03/10/2026
> Open issue on what color-scheme() returns when used in color-scheme property github.com/w3c/csswg-dr...
110
romainmenke @romainmenke.bsky.social · 03/10/2026
`if` queries the element itself. That is the interesting feature of `if` :)
110
romainmenke @romainmenke.bsky.social · 02/10/2026
We mostly use html forms and just let them submit. Spam prevention is quite painful, not sure what the platform can do about this :)
010
romainmenke @romainmenke.bsky.social · 30/09/2026
example.com was updated!
static.klipy.com
Shocked Cat Reaction
Alt: A very surprised cat
061
Reposted by romainmenke
Andrey Sitnik @en.sitnik.es · 25/09/2026
I made a new PostCSS plugin for the new CSS `corner-shape: squircle`. It creates smooth corners like in iOS (and nature). The plugin adjusts `border-radius` for squircle geometry and can add `corner-shape` automatically. github.com/postcss/post...
2235
romainmenke @romainmenke.bsky.social · 23/09/2026
macos27 has so many visual glitches. Feels very LLM generated. An actual person would never have overlooked these.
001
romainmenke @romainmenke.bsky.social · 20/09/2026
Mistakes can happen at any time, even in 1991 ;) Anyway, seems much more prudent to carve out a specific namespace for author defined custom things.
010
romainmenke @romainmenke.bsky.social · 20/09/2026
ugh, please no. Much better to carve out a namespace for custom attributes specifically so that any attribute can still have hyphens. Hyphens really help with readability (I think also screen readers). Accessibility is not only for users of a website, also important for authors.
110
romainmenke @romainmenke.bsky.social · 20/09/2026
So can't make any changes that make the primary case slower or worse. But maybe there are small affordance possible? Changes that aren't observable otherwise but do make it easier to run in a browser.
100
romainmenke @romainmenke.bsky.social · 20/09/2026
Yeah, might be best to explain your use case there and list any new pain points that haven't been mentioned yet. Maybe some issues have already been resolved on the v18 branch? Stylelint will obviously never be browser first. The main use case is local/ci where we can assume a node env
100
Reposted by romainmenke
Jane Ori 💚 CSS Runner 👩‍💻 @janeori.propjockey.io · 19/09/2026
This is as close to a cry for help that you'd see from me. I need work. I need money for shelter and food in the meantime. I've run out.
14436
romainmenke @romainmenke.bsky.social · 16/09/2026
If you work(ed) at a web agency, have you ever convinced a client to spend resources on implementing user preference media queries? Things like dark mode. Specifically asking about the agency aspect as there you need to convince the client to spend money on extra design and dev work.
000
romainmenke @romainmenke.bsky.social · 13/09/2026
Thank you for writing this! Might be interesting to turn this into an open letter asking Microsoft/GitHub/npm to implement this?
010
Reposted by romainmenke
Laurie Voss @seldo.com · 13/09/2026
I have been writing this post about the economics of open source, on and off, for 13 years. I wasn't expecting to end up at "Microsoft should send corporations a huge bill" but nothing else we've tried has worked. seldo.com/posts/nobody-pays-for-ope…
seldo.com
Nobody pays for open source. We can force them to. | Seldo.com
Thirty years of voluntary funding schemes haven't paid open source maintainers because asking doesn't scale. The real money already flows — to JFrog, Docker, Sonatype — at the registry layer. The fix: registries charge companies for supply (they already do), then route a slice automatically to the maintainers who make that supply worth having.
2016857
romainmenke @romainmenke.bsky.social · 09/09/2026
You could try with the v18 branch (if you have time). Will be a while until we start planning a v19 so now is a good window to request such changes :)
110
romainmenke @romainmenke.bsky.social · 09/09/2026
Feel free to open an issue if there were specific dependencies that made it hard to run Stylelint in the browser :)
110
romainmenke @romainmenke.bsky.social · 04/09/2026
Traffic to sites we host, I mean, not internet traffic in general. This has also been trending in the wrong direction since LLM's took off 🤷 Used to be 70-80 percent that was somewhat modern.
000
romainmenke @romainmenke.bsky.social · 04/09/2026
By our metrics 40-60 percent of all traffic (including bots we don't block) is coming from user agents that were released since fall 2025. Everything else is running older versions.
100
romainmenke @romainmenke.bsky.social · 04/09/2026
It is worth considering that those stats aren't really correct 😅 Mobile for example is wildly incorrect. It counts all Chrome Android usage as if coming from the latest Chrome version. Thereby inflating adoption. Currently 44% of usage data is Chrome for Android. So basically half is bogus.
200
romainmenke @romainmenke.bsky.social · 04/09/2026
Mainly wondering how this can be avoided. But maybe this is unavoidable despite the best intentions from everyone involved?
000
romainmenke @romainmenke.bsky.social · 04/09/2026
Interesting! Was that proposal accepted or rejected (and despite being rejected it still signaled demand)? To be clear, I am not personally concerned about any of the current policies or way of working. Only commenting because there was some negativity in the past editions about some rejections.
100
romainmenke @romainmenke.bsky.social · 04/09/2026
Thank you 🙇
010
romainmenke @romainmenke.bsky.social · 04/09/2026
Simply stating up front that positive signals from all vendors is a requirement would help I think.
100
romainmenke @romainmenke.bsky.social · 04/09/2026
Indeed, I saw that too. But this is more about managing expectations. In previous years some were frustrated because they felt they had made a strong case for X or Y and it still got rejected. The bot reply leaves it up to the reader to interpret the various data points.
100
romainmenke @romainmenke.bsky.social · 04/09/2026
Might not hold true for all past proposals but it also helps if there is at least one solid implementation already. A feature not having any support isn't really an interop problem. For example: github.com/web-platform... I might have missed it but also didn't read much about this in the docs.
github.com
CSS @mixin · Issue #1359 · web-platform-tests/interop
Description The @mixin CSS at-rule enables defining reusable and parametrizable block of CSS. Once defined, a mixin can be called using the syntax @apply --mixin-name(<parameters>);. This feature b...
100
romainmenke @romainmenke.bsky.social · 04/09/2026
For example the very first submission this year 😅 github.com/web-platform...
github.com
WebUSB · Issue #1331 · web-platform-tests/interop
Description The WebUSB API :) Specification https://wicg.github.io/webusb/ web-feature https://web-platform-dx.github.io/web-features-explorer/features/webusb/ Test Links https://wpt.fyi/results/we...
100
romainmenke @romainmenke.bsky.social · 04/09/2026
Nice! Might be good to also link that in the interop docs/blog posts. Being upfront and clear about this will avoid some frustration :)
100
romainmenke @romainmenke.bsky.social · 04/09/2026
A list of features that browsers won't implement anyway 😅 An unspoken "rule" with the interop proposal is that it only seems to influence priority and not whether something will be implemented.
220
romainmenke @romainmenke.bsky.social · 02/09/2026
Ugh, browserstack seems to have dropped support for a bunch of versions for real devices. I don't think the simulator can be used to test this. Would be nice to figure this out so people feel safe remove this :)
010
romainmenke @romainmenke.bsky.social · 02/09/2026
Do you have a hosted test? I can try it through browserstack.
110
romainmenke @romainmenke.bsky.social · 02/09/2026
Do you have any details on the exact version of iOS when this became redundant? Might not be dead for everyone 😅
130
romainmenke @romainmenke.bsky.social · 02/09/2026
Personally I would prefer a zero tolerance for anything coming out the orange blob and the associated movement. Would be neat if everyone just went "nope"
010
romainmenke @romainmenke.bsky.social · 02/09/2026
Very sure that not all will be "reverted" after. Only that which makes sense for the bottom line 🤷 So protesting now is still important. For those issues that actually matter (immigration, trans rights, health care, ...) Maybe not so much for names on maps...
110
romainmenke @romainmenke.bsky.social · 02/09/2026
These institutions probably don't have devrel teams ;) So you see this anger and some of it might even be directed at you (hopefully not) or maybe people try to reach out and voice concerns through you or other devrel members. I understand both you and those who protest in this.
110
romainmenke @romainmenke.bsky.social · 02/09/2026
I also think it's fine for people to be critical of companies. Even if they shouldn't be surprised if companies act as companies do. As long as they don't blame or focus individual employees of that company.
100
romainmenke @romainmenke.bsky.social · 02/09/2026
I think you missed my point :) Being that people shouldn't expect companies to be some force for or against a particular "-ism" (except capitalism) Google is also an easy target for criticism. Everyone knows them, uses their services. They also claim(ed?) to do no evil...
100
romainmenke @romainmenke.bsky.social · 02/09/2026
Same is true for consumers. Very easy to be vocal about issues when that doesn't require you to do anything. But how many really switch service providers over such issues? Especially when alternatives are more costly or when it involves migrating (lots of) data.
110
romainmenke @romainmenke.bsky.social · 02/09/2026
Much easier to take a stand when it doesn't cost you anything. For Google it would be very costly to resist. I would obviously prefer them having a spine and resisting fascist orange blobs but what are you gonna do 🤷
100
romainmenke @romainmenke.bsky.social · 02/09/2026
Kidding aside, if this is about maps then I think people should be much less surprised that Google is following the law. It is a company that makes a lot of money both directly from and through lobbying the U.S. gov. It is not going to bite the hand that feeds it.
110
romainmenke @romainmenke.bsky.social · 02/09/2026
Maybe a popup? "Opt-in to fascism to see local lake names" With the "yes" button barely legibly.
110
romainmenke @romainmenke.bsky.social · 02/09/2026
Is this about maps and names of lakes? I don't think most large companies can resist such changes. But it is telling how fast they comply. Google is very creative when it comes to compliance with GDPR and making products that are GDPR compliant. They could have employed some of that skill set :D
140
Reposted by romainmenke
Lea Verou, PhD @lea.verou.me · 31/08/2026
❓Fun CSS trivia! What do you get if you use contrast-color() on an out of gamut color? 😀 😀 😀 😀 😀 1. Chrome: uses the clipped color, so the result is at least legible 2. Safari/Firefox: uses the original, so the result is ILLEGIBLE ❌ 3. Firefox w/ gamut mapping enabled: uses the mapped color ✅
1175
romainmenke @romainmenke.bsky.social · 30/08/2026
Private Custom Properties: github.com/csstools/pos... #postcss Not yet added to postcss-preset-env as it is an early spec. But has some interesting properties :)
github.com
postcss-plugins/plugins/postcss-private-rule at main · csstools/postcss-plugins
PostCSS Tools and Plugins. Contribute to csstools/postcss-plugins development by creating an account on GitHub.
020