Open zcorpan opened 1 year ago
Someone asked in the meeting if there's a list of all prefixed attributes that exist, which I had actually researched a few years ago: https://github.com/whatwg/html/issues/2271#issuecomment-273716582
Thank you @zcorpan for researching this and letting us know; this is great news. I am sure the Adapt Task Force will be filing an issue over on the HTML repo in order to begin dialogue with the WHATWG in due course.
I think that may've been me asking the question about the list of prefixed attributes. I'm also quite interested in searching the archive as part of our further development of the Adapt spec, and in relation to a separate personal project relating to landmark regions, so thanks for the pointers to those results, the docs, and the queries you used.
As per our discussion in today's Adapt weekly call, I'm adding the link to the filed issue requesting adapt- as our new prefix - https://github.com/whatwg/html/issues/9084.
At the APA & WHATWG joint meeting at TPAC today we discussed the usage of
data-*
attributes. The reason for not usingdata-*
is that those attributes are intended store information that is private to the page. The HTML standard says:There seemed to be a preference in the meeting for the
adapt-
prefix. I promised to research usage in HTTP Archive for such attributes, which I've now done. I also looked forssml-
while at it.The result is that out of the 15,000,000 pages in the dataset, there are 19 pages which use an attribute
adapt-container-width="true"
. There were nossml-
prefixed attributes.Conclusion: it should be web compatible to use
adapt-
andssml-
as prefixes.query and detailed result
``` SELECT * FROM ( SELECT page, ARRAY_TO_STRING(REGEXP_EXTRACT_ALL(body, r'(?i)(<[a-z][a-z0-9_-]*(?:\s+[^>]+)*\s+(?:adapt|ssml)-[a-z0-9_-]*[^>]*>)'), "\n") AS match FROM `httparchive.response_bodies.2022_08_01_mobile` WHERE page = url ) WHERE match != "" ``` Result: https://docs.google.com/spreadsheets/d/1cFW6vGH5u3qgFyP2a3AvNkDCkKobWlhc_lI6_djCYkY/edit?usp=sharing Note that the results contain false positives with e.g. `adapt-*` in `class=""`.