Hello everyone, I have a question regarding the WebDriver standard there are 2 links, but what is the difference between w3c.github.io/webdriver (living document) and w3.org/TR/webdriver1 (CR level 1), is "Living Document" equal to "Working Draft"?

cuki: No.

cuki: The latter link is a snapshot taken 5 June 2018 when WebDriver went through the Recommendation track at W3C.

cuki: The first link is a living document and reflects the tip of the git repo.

cuki: Typically you will never use w3.org/TR/webdriver1 for anything but give it to lawyers.

cuki: The living document, on the other hand, represents the current reality as implemented by browser vendors.

ato: Thanks for that described explanation! In case I want to reference to the WebDriver standard, best would be then to use the living document, as it contains the latest changes which is most likely (to be) implemented by browser vendors, correct?

cuki: Absolutely!

ato: Great, thanks for clearing that up!

The changes in the living document are actually guaranteed to be implemented by at least two or more vendors.

We don’t allow changes to the specification without commitment from vendors to implement them, or without tests.

You might find wpt.fyi/results/webdriver/tests?label=stable&aligned useful for tracking which drivers implement what.

Of course there are bugs and so on, but generally what I said is true (-:

Actually we should probably link that in the READMe.

Wow, that is a nice approach to "approve" changes, there is already a markdown in the repository about vendor status: github.com/w3c/webdriver/blob/master/implementation-report.md

Edge is pretty far behind, looking at the tests :D

Yeah, but unfortunately that refers to the level 1 spec.

We should clarify that too…

I don’t know, but Edge will probably abandon their implementation seeing as they’re switching to Chromium.

)-:

What about reporting issues for Safaridriver? Where can I do this, for Geckodriver and Chromedriver there are some public repositories to report issues to.

For example if I look at wpt.fyi/results/webdriver/tests/set_window_rect?label=stable&aligned, it shows that Safari properly supports /window/rect (POST) entrypoint

developer.apple.com/bug-reporting maybe?

Or ping brrian in this room.

when in fact when specifying "0" or "1" as position value, it reads it as Boolean instead of Number, resulting in error: "Parameter 'x' has a boolean value, which is not allowed."

Oh that sounds like a bug.

also from what I can deduct from the standard, it should also be possible to specify negative values for x/y when setting window/rect, no?

Yes.

Depending on the window manager and system, you may or may not see the window be moved off screen.

But it will try its best.

Ok, for those cases I got a "no such window"

This is documented too: developer.mozilla.org/en-US/docs/Web/WebDriver/Commands/SetWindowRect

aha, but the window is open, anyway, just checking, as you mentioned there are some bugs, but at least this is moving into a unified direction, which is awesome

Will check out developer.apple.com/bug-reporting, as I have a few Safaridriver bugs to report

the weirdest one was that it could not send commands to a session browser instance when external monitors have the same name/specs

I’m sure brrian would be interested to hear about that.

lost a good 3-4 hours pinpointing that issue :D

will have to verify once more before reporting, but have the monitors at the office

All right, thanks ato for taking the time to help and clarify with any questions I had, will see later if I can find anywhere to report Safaridriver bugs, as the link to developer.apple.com seems to be for development issues, not actual product issues.

I have used it to successfully report bugs with Mail.app in the past.

But I don’t think I’ve used it for safaridriver.

Maybe you will find it listed under the product ‘Safari’.

ato: When you reported a bug, were you able to track it's progress or was it send into oblivion leaving you to hope and wonder?

I… can’t remember.

Ah well, for now I've described the /window/rect issue with Safaridriver partially in this issue webdriverio/webdriverio #3216 (under "Further comments") so brrian has something to read until I'll look into how to report a bug again tomorrow when I have both (identical) external displays to my disposal again.