47 replies

1. Zach Leatherman moments later

   Consider `accent-color` (Not Baseline) vs. CSS Nesting (Baseline 2023).

   Browser compatibility is only a small piece: it’s more important to consider the role it plays and how progressive-enhancement friendly it is.

   CSS Nesting requires cross-browser ubiquity: `accent-color` is much safer!

   1. Private reply 6 min later
      1. Zach Leatherman 15 min later

         where are you following 2025 status?

         1. Private reply 34 min later
            1. Private reply 35 min later
            2. Zach Leatherman 36 min later

               ah no need to apologize! just trying to understand the context — thank you!

            3. Mary Branscombe 1 hr later

               Interop isn't Baseline; things that are in Interop are by definition not ready for Baseline, but having them as priorities in Interop means they might be in Baseline in a year or so; I wrote a couple of articles about this a while back

               1. Private reply 1 hr later
                  1. Private reply 1 hr later
                  2. Mary Branscombe 2 hr later

                     Apple is a big part of why Interop happens (for both bad and good); it's at least moved things along from where they were before the annual Interop priority alignments - and no worries, I've been trying to explain how these fit together for a while and I'm about to have another go

                  3. Brian Kardell 2 hr later

                     Hi, I'd love to hear more (maybe not on this thread) about why you think interop 2025 is bad? Happy to use DMs or have a video chat or whatever you prefer

                     1. Mary Branscombe 3 hr later

                        equally curious, so if it does stay on the thread... 👀👀👀

                     2. Private reply 4 hr later
                        1. Private reply 4 hr later
                           1. Private reply 4 hr later
                        2. Brian Kardell 4 hr later

                           I can tell you, as someone on the committee, that it seems a so difficult to manage what people read into all of this. A lot of people think we should focus exclusively on the more mature, rough edges, and not new stuff - and lot of other people think it should include none of that.

                           1. Private reply 4 hr later
   2. Zach Leatherman 13 min later

      I might suggest modifications to the “Limited availability” and “Newly available” tags to include language about whether a feature is progressive enhancement friendly or not!

      1. Stuart Robson 21 min later

         renaming this side project to "Can I Baseline?"

         1. Zach Leatherman 35 min later

            I’M INTERESTED

      2. Patrick Brosset 39 min later

         We're definitely interested in talking more about this on the WebDX CG. Progressive enhancements, polyfills, and accessibility have come up multiple times in the past. So far, we've not been able to incorporate these aspects into Baseline calculation though. But I do believe we should.

         1. Patrick Brosset 44 min later

            One of the things I've tried to steer the group towards in the first years of the project is to avoid "hiding" things that are not yet supported. Anything we can do to make features easier to discover/use, not harder, is valuable.

            1. Zach Leatherman 1 hr later

               interesting! “hiding” in what view?

               1. Patrick Brosset 1 hr later

                  As in, Baseline comes with the risk of hiding too many features from the web: those that aren't supported across browsers even though they're perfectly fine to use.
                  Showing a "Limited" Baseline banner on a doc website about a given feature comes with the risk of steering people away from it.

                  1. Zach Leatherman 1 hr later

                     Ah, I see — yeah. That dovetails back to Baseline missing some pieces if you want to encourage or discourage usage on a feature, imo! “Limited but safe!”

                     1. Patrick Brosset 1 hr later

                        Filed github.com/web-platform... to put it on our radar. We'll likely discuss at our next CG meeting (which you're more than welcome to join, details here: www.w3.org/community/we...), but please feel free to comment on the issue too.
                        Pinging @miriam.codes who had made a similar point too.

                        1. Zach Leatherman 2 hr later

                           appreciate you Patrick! 🙌

      3. Zach Leatherman 1 day later

         Some additional discussion is happening on this GitHub Issue: github.com/web-platform...

         Thanks @patrickbrosset.com for filing!

   3. Stuart Langridge 49 min later

      "Baseline Newly Available" is useless. It means "someone somewhere thought of this. It's not everywhere. Don't use it." It is extremely annoying that it was invented, and afaict it was invented because people were sad that their new feature was not yet "baseline" and so people wouldn't use it.

      1. Stuart Langridge 50 min later

         this is harsher than I normally am about stuff. But I was super annoyed at "newly available". It was Really Useful to explain to teams I work with that "if a thing says 'baseline' it's probably OK for us, otherwise don't" and now there's a special new baseline which utterly invalidates that.

         1. Zach Leatherman 1 hr later

            I think that’s fair feedback. I’m not sure the arbitrary 30 month goalpost is a good finish line for “Widely available” either — hmm

            1. Stuart Langridge 1 hr later

               I think that's OK, myself (which is why I liked "baseline"). It's essentially a more formal definition for the oft-used and not-well-defined "this requires a modern browser", I think

               1. Zach Leatherman 1 hr later

                  ah, that phrase sets off red flags for me but I see your point 😅

   4. Nicholas C. Zakas 6 hr later

      While working on the ESLint CSS plugin’s require-baseline rule, we discovered that cursor is not baseline due to incomplete mobile implementation. 🤷‍♂️

   5. Josh Tumath 7 hr later

      This was something I fed back as well at the time. I still think it would make a huge difference. Same with more complex features like View Transitions.

2. Andy Davies 45 min later

   Got to admit I still tend to go to Can I Use

   Not quite sure why we needed another alternative

   1. Zach Leatherman 1 hr later

      I see the value in a simplified goalpost for features (for developers) but I think most of the benefit is in cross-browser planning — features categorized by aspirational support

   2. Mary Branscombe 1 hr later

      most granularity: going through the MDN pages individually for compat data on APIs and querying BCD
      lots of granularity: caniuse against your own logs to see what your audience can benefit from
      back of the envelope decisions about directions when you want to know what's mainstream: baseline

3. Ethan Marcotte 1 hr later

   oof, yeah. “baseline” doesn’t mean “universal”, and boy do i wish that informed we talk about these features

   1. Brian Kardell 2 hr later

      baseline widely available should mean something very close to that though - it's 30 months past the initial interoperability date, which we based on looking at gobs of data about rollouts. Of course, all of these require nuance and weighing of your own users needs - but as a rule of thumb, p good?

      1. Brian Kardell 2 hr later

         Idk, iirc @meyerweb.com and I attempted to argue for something even more granular and with stages as numbers rather than words that could be more easily misread. There are a handful of 'stages' that are useful to identify.

         1. Brian Kardell 2 hr later

            For example, we made bcd-watch.igalia.com/weekly-compl... which you can listen to a weekly rss feed of and it tells you when bcd records the final engine implementation. That can be before it's even available in releases, but it's a very strong "this is coming!" - so probably worth your time

4. Stephen Scaaaaaaaaaaaaaght 😱 2 hr later

   I thought of Baseline as a reference for us to tell stakeholders 'Yes, we can do this... but' in terms of enhanced clarity?

   Prior to Baseline, caniuse seems to be only alternative?

5. Henri Helvetica 🧑🏾‍🚀🇭🇹 8 hr later

   @rviscomi.dev welcome aboard!

6. Jeremy Keith 17 hr later

   At Clearleft, we use Baseline as a very handy starting point for our browser support policy:

   browsersupport.clearleft.com

   1. Zach Leatherman 20 hr later

      Unsurprisingly, this is a great way to frame it!

   2. Yann Bertrand 1 day later

      Do you have a source on what could be considered progressive enhancement? I think it’s not that easy to find out