Intent to Implement and ship: getElementsByTagName accepts qualified names

[TAMURA, Kent]

Contact emails

tk...@chromium.org 

Spec

https://dom.spec.whatwg.org/#concept-getelementsbytagname

Summary

getElementsByTagName() accepts qualified names as well as local names.

Is this feature supported on all six Blink platforms (Windows, Mac, Linux, Chrome OS, Android, and Android WebView)?

Yes 

Interoperability and Compatibility Risk

Very low. This is a small enhancement of an existing function.

Two other browsers already shipped this feature.

Edge: No signals.

Firefox: Shipped

Safari: Shipped

Web developers: No signals?

Is this feature fully tested by web-platform-tests?

Yes. There are some tests in web-platform-tests/dom/.

OWP launch tracking bug

http://crbug.com/645165

Entry on the feature dashboard

https://www.chromestatus.com/features/5673616809132032

--

TAMURA Kent
Software Engineer, Google

[Philip Jägenstedt]

LGTM1. Can you file an issue at https://developer.microsoft.com/en-us/microsoft-edge/platform/issues/ if Edge is failing some of the tests? Once they are the only ones to fail some tests, they may want to prioritize it.

[Dimitri Glazkov]

LGTM2

[Chris Harrelson]

LGTM2

--
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 blink-dev+unsubscribe@chromium.org.

[Christian Biesinger]

That's a weird function, why is this a good addition to the web
platform? Per the spec, this will return nodes from (potentially) a
mixture of namespaces. That seems confusing. Why not just have people
use getElementsByTagNameNS?

Christian

[Johnny]

I agree it's a weird one. If memory serves this was done in DOM Level 2 way back in 2002 or some such. It was known to basically be redundant with getElementsByTagNameNS(), but considered a convenience function for those who used namespaces in XML a lot etc. Given how old it is and the fact that Gecko has had this implementation for a decade and a half, and those who are used to DOM trees outside of the web might have ran into it in other contexts, I'd still be in favor of adding this even though blink has managed w/o this thus far.

My 2c.

[Christian Biesinger]

Fair enough -- I didn't realize this was such an old API!

Christian

[TAMURA, Kent]

Filed a microsoft issue: https://developer.microsoft.com/en-us/microsoft-edge/platform/issues/12211978/

[John Jansen]

Thanks! I’ll make sure it gets assigned out.

[PhistucK]

A bit of a fallout -

crbug.com/736430

crbug.com/753370

> getElementsByTagName() accepts qualified names as well as local names.

That summary tells me it actually is a superset of the previous behavior, rather than a different-set (is there a word for that? :)), but it turns out, if I understand the issues correctly, that <a:b> would no longer be found when you call getElementsByTagName("b") whereas before it did. That makes the interoperability risk higher, because it breaks code, even if that code only worked in Chrome (and formerly Safari perhaps). Too late now, but, still, it was not clear from the intent that this is a breaking change or a removal of some kind.

☆PhistucK

--
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 blink-dev+unsubscribe@chromium.org.

To view this discussion on the web visit https://groups.google.com/a/chromium.org/d/msgid/blink-dev/SN1PR0301MB2112A08D3F76033D20A6A179F1F70%40SN1PR0301MB2112.namprd03.prod.outlook.com.

[Dave Tapuska]

I triaged this to the DOM team today as well....

https://bugs.chromium.org/p/chromium/issues/detail?id=752150

This one might have larger impact.

dave.

To view this discussion on the web visit https://groups.google.com/a/chromium.org/d/msgid/blink-dev/CABc02_KQKK2jaU6tevJ%3Dao9yV3-TrFy%3Dc8%3DzgGHx%3D385%3Dx6Haw%40mail.gmail.com.

[TAMURA, Kent]

On Wed, Aug 9, 2017 at 5:14 AM, Dave Tapuska <dtap...@google.com> wrote:

I triaged this to the DOM team today as well....

https://bugs.chromium.org/p/chromium/issues/detail?id=752150

This one might have larger impact.

dave.

On Tue, Aug 8, 2017 at 4:03 PM, PhistucK <phis...@gmail.com> wrote:

A bit of a fallout -

crbug.com/736430

crbug.com/753370

> getElementsByTagName() accepts qualified names as well as local names.

That summary tells me it actually is a superset of the previous behavior, rather than a different-set (is there a word for that? :)), but it turns out, if I understand the issues correctly, that <a:b> would no longer be found when you call getElementsByTagName("b") whereas before it did. That makes the interoperability risk higher, because it breaks code, even if that code only worked in Chrome (and formerly Safari perhaps). Too late now, but, still, it was not clear from the intent that this is a breaking change or a removal of some kind.

Actually, I wasn't aware of this incompatibility when I posted this intent. I'm sorry for the lack of the important incompatibility information.

Even though the intent had this incompatibility information, I guess the ship was approved anyway because Firefox and Safari already shipped it. However, we could have provided a warning in the Beta blog post.