963299 - Bump marionette_client version to 0.7.3 and release to PyPI

Status: VERIFIED FIXED

(Remote Protocol :: Marionette, defect)

Description

[Dave Hunt [:davehunt] [he/him] ⌚BST]

Here's a summary of the changes since the last release (so far):

Bug 947205 - fix missing test durations from the HTML report
Bug 867600 - Change hardcoded sleeps into waits
Bug 925688 - Marionette unit test intermittent failure, temporarily disable the test
Bug 950641 - Automatically use the appropriate GeckoInstance class
Bug 957162 - Marionette Wait's should indicate timeout duration
Bug 957248 - Use marionette.timeout if no explicit timeout is given for Wait
Bug 956803 - Fix busted autogen docs
Bug 959730 - Move cleanup logic out of __del__
Bug 927285 - Fix for intermittent on test_submit.py
Bug 960375 - Marionette runner should pass mozinfo.info into manifestparser
Bug 941102 - Fix closeWindow not matching WebDriver command close
Bug 956655 - Part 1: Move MobileConnection related files to dom/mobileconnection
Bug 941129 - Fix getAllCookies not matching WebDriver command
Bug 961874 - Python client should delete marionette session in cleanup()
Bug 958024 - Wrap last exception from Marionette's Wait in TimeoutException
Bug 941136 - getUrl not matching webdriver command getCurrentUrl
Bug 941132 - getElementPosition not matching webdriver command

Generated (and then edited) using:
$ hg log -M -r ba35f88d8104: --template "{desc}\n" testing/marionette/client

Comment 1

[Dave Hunt [:davehunt] [he/him] ⌚BST]

If we can release with the new expected conditions too, that would be great.

Comment 2

[Dave Hunt [:davehunt] [he/him] ⌚BST]

Attachment: Bump marionette_client version to 0.7.3

Comment 3

[Andreas Tolfsen ❲:ato❳]

dhunt, thanks for landing bug 948075 on m-i!

Comment 4

[Andreas Tolfsen ❲:ato❳]

Added bug 963166 and 963162 as dependencies for this.  With them included, the next version of marionette_client will be using the full WebDriver command set in the background for communicating with Marionette.

Comment 5

[Dave Hunt [:davehunt] [he/him] ⌚BST]

(In reply to Andreas Tolfsen (:ato) from comment #4)
> Added bug 963166 and 963162 as dependencies for this.  With them included,
> the next version of marionette_client will be using the full WebDriver
> command set in the background for communicating with Marionette.

Does that deserve a bigger version bump? 0.8?

Comment 6

[Malini Das [:mdas] - Away, not checking bugmail]

(In reply to Dave Hunt (:davehunt) from comment #5)
> (In reply to Andreas Tolfsen (:ato) from comment #4)
> > Added bug 963166 and 963162 as dependencies for this.  With them included,
> > the next version of marionette_client will be using the full WebDriver
> > command set in the background for communicating with Marionette.
> 
> Does that deserve a bigger version bump? 0.8?

Just what I was thinking! +1.

Why not finally put it to 1.0? I think marionette's usable enough for an 'official' release.

Comment 7

[Malini Das [:mdas] - Away, not checking bugmail]

well "official sounding"

Comment 8

[Andreas Tolfsen ❲:ato❳]

Does that mean we can't make backwards incompatible changes anymore?

Comment 9

[Malini Das [:mdas] - Away, not checking bugmail]

(In reply to Andreas Tolfsen (:ato) from comment #8)
> Does that mean we can't make backwards incompatible changes anymore?

I'd think that if we make a backward incompatible release, we'll just change the x.0 number?

Comment 10

[Dave Hunt [:davehunt] [he/him] ⌚BST]

I thought there were some things that were preventing us from releasing 1.0, or we were holding off until the changes in server side for '1.0' reach a release version of Firefox. Or maybe I'm not sure what I'm talking about. :P

Comment 11

[David Burns :automatedtester]

My Personal preference is that when we have parity with webdriver we hit 1.0. I am fully expecting there are things that may break backwards compatibility to get there.

As I say, this is my personal perference since I like that we are level with webdriver for version 1 but I can easily be swayed

Comment 12

[Jonathan Griffin (:jgriffin)]

Should we wait for better desktop compatibility before we declare 1.0?  I'm particularly thinking of lack of modal dialog support, better tab support, etc.

Comment 13

[Andreas Tolfsen ❲:ato❳]

(In reply to David Burns :automatedtester from comment #11)
> My Personal preference is that when we have parity with webdriver we hit
> 1.0. I am fully expecting there are things that may break backwards
> compatibility to get there.

After some consideration, I agree with dburns here.  There will be more
breakage in the near future, especially considering that the WebDriver spec
isn't finished.

That said, Marionette's version numbering doesn't necessarily have to be
a paired with the spec's.

Comment 14

[Malini Das [:mdas] - Away, not checking bugmail]

I find dburns suggestion sounds good, at least our version numbers will have some relevance and we'll know which version of webdriver is being supported by which version of marionette, and we can just update the 1.0.x to reflect our own updates.

Comment 15

[Dave Hunt [:davehunt] [he/him] ⌚BST]

Okay, let's stick with 0.7.3 then. As all blockers are resolved, I will land this on inbound.

Comment 16

[Dave Hunt [:davehunt] [he/him] ⌚BST]

Landed on inbound:
https://hg.mozilla.org/integration/mozilla-inbound/rev/38cb01168c32

Comment 17

[Ryan VanderMeulen [:RyanVM]]

https://hg.mozilla.org/mozilla-central/rev/38cb01168c32

Comment 18

[Dave Hunt [:davehunt] [he/him] ⌚BST]

Released:
https://pypi.python.org/pypi/marionette_client/0.7.3

Comment 19

[Stephen Donner [:stephend] Not actively reading bugmail]

Great; thanks, all!