Caniuse: Can I use needs web scrapping/crawling

[LifeIsStrange]

Can I use is not actively maintained enough, the number of features where it has outdated informations, especially concerning chrome/blink feature supports, is astonishing.

But it happen that chromium features support has already an always up to date, complete knowledge base.
It's called https://chromestatus.com/features

We just needs to scrap each chromestatus feature
And get:
The feature title
The support status and the corresponding version.
Other data is bonus.
Then go to can I use matching feature title
And overwrite chrome data (and eventuality opera, new edge, etc)

The difficulty is the mapping between chromestatus feature title and can I use feature title.
So manually we should contribute
Association pairs:
{chrome status title, can I use title}
E.g :
{"CSS Scroll Snap Points", "CSS Scroll snap" }
So yes there is a quick manual effort but you do it once and then data will be updated automatically (just periodically re run the scrapper) which will be a huge improvement for showing what we can really use.

Bonus 1:
Scrap other browsers feature supports on can I use.
Disclaimer I don't know how much maintained it is for other browsers.

Bonus 2:
Many chromestatus features have no equivalent in can I use because can I use is deeply incomplete.
Then a contributor could just give an association key
{chromestatus title, NULL}
Which would then automatically create a new can I use feature using the same name, description and data as chromestatus.

Doing this would help make the world a better place.
I wonder if you could propose this as a Gsoc (Google summer of code)

BTW instead of scrapping, maybe they have an API?

[Fyrd]

Regarding maintenance of the caniuse data, it's partly because of this that recently the MDN Browser Compatibility Data was included in caniuse searches. This data is more actively maintained than the caniuse data and aims to cover all web technologies, where caniuse data prioritizes features that have the most developer interest.

Including the data provided by the Chrome Platform Status project is interesting and worth considering, however the biggest problem I see is that their data focuses primarily on the status in Chromium rather than across all of 19 browsers tracked by caniuse. Yes, they track if a feature is "shipped" in a few other browsers, but doesn't provide the version in which it was shipped or any other details. Most of the work in getting compatibility data is doing the testing required across browsers so that work would still need to be done.

Another concern I have is that because the Chrome Platform Status project is focused primarily on features being worked on in Chrome/Chromium I feel this would unfairly skew the data found on caniuse towards Chrome features compared to those being developed in other browsers.

I'll leave this issue open because I think it could still be of value to include CPS results in search results similar to what's happening with the BCD data, though as mentioned since the data's much more limited it would probably just link to the CPS site rather than provide a support table.

[jpmedley]

Chrome Status is reliable mostly because for the last several years I've hounded feature owners to keep it up to date. I would't build a system with a single human point of failure.

Regarding MDN, we (Google) are trying to make Chrome data more reliable. I can't say more than that now.

[Elchi3]

I was thinking about importing Chrome Status data into BCD, but never really looked into it more closely. If anyone is interested in doing that, let me know :)

I agree Chrome Status is Chrome biased. It even says "After a feature ships in Chrome, the values listed here are not guaranteed to be up to date." So, it happens to be wrong about other browser statuses in many entries, so I believe the site could just stop talking about non-Chrome implementations and refer to BCD/caniuse instead.