Given that we are disabling this for 3rd parties, would it make sense to already start warning 1st-party usage (saying that this doesn't work at all for 3rd parties and that it will be removed in the future)?
I don't know what's the guideline regarding having a deprecation warning without a clear milestone and also if it's reasonable to do a warning for a feature that has 1% usage <https://www.chromestatus.com/metrics/feature/timeline/popularity/10>. But I suggest this because: 1. the current behavior (only 1st party) may be a bit surprising, if not contextualized with "this feature is being deprecated". 2. early warning is better? On Wed, Oct 27, 2021 at 5:33 PM Ari Chivukula <[email protected]> wrote: > Sorry for the delay, the data for actually reading/writing a database > <https://www.chromestatus.com/metrics/feature/timeline/popularity/3985> > (as opposed to opening it > <https://www.chromestatus.com/metrics/feature/timeline/popularity/3865>) > in a third party context ended up being nearly identical (0.03% of loads > and about 10% of top-sites). > > The code I spot checked still seems to gate usage of WebSQL on > availability. That said, the change in M97 is linked to a feature flag so > can be undone if there are issues in dev/beta. Also, there's an enterprise > policy available to enable it (planned to exist until M102, when it'll be > removed entirely and not simply disabled by default). > > ~ Ari Chivukula (Their/There/They're) > > > On Thu, Aug 5, 2021 at 12:21 PM Alex Russell <[email protected]> > wrote: > >> LGTM3 with the caveats Yoav outlines. Would like to hear back about the >> new usecounter data on this thread before 97 branch point. >> >> Thank you for taking so much care with this deprecation. >> >> On Thursday, July 29, 2021 at 3:50:14 PM UTC-7 Chris Harrelson wrote: >> >>> LGTM2 >>> >>> On Wed, Jul 28, 2021 at 9:50 AM Yoav Weiss <[email protected]> >>> wrote: >>> >>>> Deprecating in M94 while adding a usecounter sounds like a solid >>>> strategy. Please update this thread with the results from the usecounter >>>> once they are available. >>>> >>>> *LGTM1* to deprecate in M94 and remove support in M97, barring >>>> usecounter surprises. >>>> >>>> On Wed, Jul 28, 2021 at 6:38 PM Ari Chivukula <[email protected]> >>>> wrote: >>>> >>>>> I'm saying something slightly different, that the usage I'm seeing is >>>>> (1) checking for the existence of the function (instead of simply looking >>>>> for if the engine is Blink) and (2) wrapping the subsequent call to >>>>> opendatabase in a try statement (as per the spec which allows errors to be >>>>> raised https://www.w3.org/TR/webdatabase/#databases). I think we're >>>>> safe to (1) deprecate the function in third party contexts to alert >>>>> developers to the change in M94 and then (2) raise SECURITY_ERR in a later >>>>> milestone once we've completed outreach around M97 or later. >>>>> >>>>> When I do (1) I could also add a counter into M94 which looks for >>>>> actual usage of a created database beyond the feature detection we seem to >>>>> be catching now. >>>>> >>>>> ~ Ari Chivukula (Their/There/They're) >>>>> >>>>> >>>>> On Wed, Jul 28, 2021 at 6:21 AM Yoav Weiss <[email protected]> >>>>> wrote: >>>>> >>>>>> So, you're saying that the usage we see is simply feature detection >>>>>> for the API and not real usage? Any way to add use counters that would >>>>>> confirm this? >>>>>> >>>>>> On Monday, July 26, 2021 at 6:38:19 PM UTC+2 Ari Chivukula wrote: >>>>>> >>>>>>> @manuel: Thanks for the corrected code pointer! >>>>>>> @yoav: The sample on the platform status page ( >>>>>>> https://www.chromestatus.com/metrics/feature/timeline/popularity/3865) >>>>>>> highlighted a bunch of Hyundai dealerships, but when loading locally I'm >>>>>>> unable to reproduce the creation of any WebSQL databases. I did find >>>>>>> usage >>>>>>> on https://accounts.mint.com/index.html, but when searching the >>>>>>> scripts which created it I found that all checked if window.openDatabase >>>>>>> even existed. The lack of support in Gecko and WebKit seems to have >>>>>>> broken >>>>>>> in the deprecation/removal codepath for us. >>>>>>> >>>>>>> ~ Ari Chivukula (Their/There/They're) >>>>>>> >>>>>>> >>>>>>> On Mon, Jul 26, 2021 at 7:46 AM Manuel Rego Casasnovas < >>>>>>> [email protected]> wrote: >>>>>>> >>>>>>>> >>>>>>>> >>>>>>>> On 26/07/2021 12:32, Yoav Weiss wrote: >>>>>>>> > WebKit: Deprecated in version 608 (Safari 13) >>>>>>>> > < >>>>>>>> https://github.com/WebKit/WebKit/commit/761bce943c0696a6bb93116eb0576ed07dbfdc65 >>>>>>>> > >>>>>>>> >>>>>>>> That's a change related to WebKitGTK+ and WPE, where they marked >>>>>>>> the API >>>>>>>> as deprecated. >>>>>>>> >>>>>>>> Anyway it seems this was removed in Safari (at least on WebKit2, it >>>>>>>> might be still available in WebKit1 though). See this thread: >>>>>>>> >>>>>>>> https://lists.webkit.org/pipermail/webkit-dev/2019-November/030970.html >>>>>>>> And also this commit: >>>>>>>> https://trac.webkit.org/changeset/277564/webkit >>>>>>>> >>>>>>>> Cheers, >>>>>>>> Rego >>>>>>>> >>>>>>> -- >>>> >>> You received this message because you are subscribed to the Google >>>> Groups "blink-dev" group. >>>> To unsubscribe from this group and stop receiving emails from it, send >>>> an email to [email protected]. >>>> >>> To view this discussion on the web visit >>>> https://groups.google.com/a/chromium.org/d/msgid/blink-dev/CAL5BFfUCNvRNGQN7Oj6y7R2682KTv%3DgHis8nOSYs7Ne5q%2BHxBg%40mail.gmail.com >>>> <https://groups.google.com/a/chromium.org/d/msgid/blink-dev/CAL5BFfUCNvRNGQN7Oj6y7R2682KTv%3DgHis8nOSYs7Ne5q%2BHxBg%40mail.gmail.com?utm_medium=email&utm_source=footer> >>>> . >>>> >>> -- > You received this message because you are subscribed to the Google Groups > "blink-dev" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to [email protected]. > To view this discussion on the web visit > https://groups.google.com/a/chromium.org/d/msgid/blink-dev/CAGpy5D%2BaQtspAhFrR1LUrvu2XZi58q4OhXV6G1kTjPO-P%3DuVmg%40mail.gmail.com > <https://groups.google.com/a/chromium.org/d/msgid/blink-dev/CAGpy5D%2BaQtspAhFrR1LUrvu2XZi58q4OhXV6G1kTjPO-P%3DuVmg%40mail.gmail.com?utm_medium=email&utm_source=footer> > . > -- You received this message because you are subscribed to the Google Groups "blink-dev" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion on the web visit https://groups.google.com/a/chromium.org/d/msgid/blink-dev/CADp2-T9xtkMhoX4c5HM496qBkcRD2gA_k5y9qu-iW0WH_Bj1bA%40mail.gmail.com.
