You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@flex.apache.org by Tom Chiverton <tc...@extravision.com> on 2015/05/15 10:24:17 UTC
Mustella SDK tests
Can someone on a Windows machine try some of the following Mustella
tests ? They are consistently failing on the Jenkins job.
spark/core/fxg/properties/fxgtemplate_masks mask_clip_mask
spark/core/fxg/properties/fxgtemplate_tester blend_mode_radial_gradient_blends_blend_difference_radialgradient_fill
gumbo/components/DropDownList/Properties/DropDownList_Properties
And many more on
http://flex-mustella.cloudapp.net/job/flex-sdk_mustella/1471/console
--
*Tom Chiverton*
Lead Developer
e: tc@extravision.com <ma...@extravision.com>
p: 0161 817 2922
t: @extravision <http://www.twitter.com/extravision>
w: www.extravision.com <http://www.extravision.com/>
Extravision - email worth seeing <http://www.extravision.com/>
Registered in the UK at: 107 Timber Wharf, 33 Worsley Street,
Manchester, M15 4LD.
Company Reg No: 05017214 VAT: GB 824 5386 19
This e-mail is intended solely for the person to whom it is addressed
and may contain confidential or privileged information.
Any views or opinions presented in this e-mail are solely of the author
and do not necessarily represent those of Extravision Ltd.
RE: Mustella SDK tests
Posted by Kessler CTR Mark J <ma...@usmc.mil>.
+1 lol
-Mark
-----Original Message-----
From: omuppi1@gmail.com [mailto:omuppi1@gmail.com] On Behalf Of OmPrakash Muppirala
Sent: Sunday, May 24, 2015 12:53 PM
To: dev@flex.apache.org
Subject: Re: Mustella SDK tests
Great, I nominate you to fix all Mustella tests going forward ;-)
Re: Mustella SDK tests
Posted by Tom Chiverton <tc...@extravision.com>.
On 26/05/15 10:49, Tom Chiverton wrote:
> On 24/05/15 17:53, OmPrakash Muppirala wrote:
>> Great, I nominate you to fix all Mustella tests going forward
>
> He does have the magic touch :-)
....
> This might be player version related if it ran OK the run before ?
>
And the most recent run has passed fine !
Tom
Re: Mustella SDK tests
Posted by Alex Harui <ah...@adobe.com>.
On 5/27/15, 1:11 AM, "Tom Chiverton" <tc...@extravision.com> wrote:
>
>It seems happier overall now since you threatened it any way :-)
I know for sure there was at least one other time where, without any code
changes, things started working again for a while. So I’ll bet we start
seeing the failures again eventually. When things start failing, the
indication is that, despite the fact the SWF asks for a 600 pixel height,
the application ends up with 566 pixels in height. I’ve put some trace
statements in CompareBitmap to try to show us whether the SWF itself
didn’t get 600 pixels (which would indicate some player issue) or whether
the application is the only thing at 566, which would point to our code
somehow. I’m wondering if there is some player popup, status bar,
additional menu bar, or something like that, that would deny the SWF its
600 pixels to show some message about updating or being launched from a
service. Then when I go login and handle the player window from the user
application level, it thinks I’ve ack’d something and doesn’t take away
those 34 pixels.
-Alex
Re: Mustella SDK tests
Posted by Tom Chiverton <tc...@extravision.com>.
On 26/05/15 16:46, Alex Harui wrote:
>
> On 5/26/15, 2:49 AM, "Tom Chiverton" <tc...@extravision.com> wrote:
>
>
>> But the current run has already logged some failures:
> The only thing that really matters is the final result. Some tests always
> fail on the first run and pass on the second run when only failing tests
> are run. Usually that’s because some previously running test that passes
> affects the way a test runs. I’m not motivated to try to clean all that
> up.
>
It seems happier overall now since you threatened it any way :-)
Tom
Re: Mustella SDK tests
Posted by Alex Harui <ah...@adobe.com>.
On 5/26/15, 2:49 AM, "Tom Chiverton" <tc...@extravision.com> wrote:
>But the current run has already logged some failures:
The only thing that really matters is the final result. Some tests always
fail on the first run and pass on the second run when only failing tests
are run. Usually that’s because some previously running test that passes
affects the way a test runs. I’m not motivated to try to clean all that
up.
-Alex
Re: Mustella SDK tests
Posted by Tom Chiverton <tc...@extravision.com>.
On 24/05/15 17:53, OmPrakash Muppirala wrote:
> Great, I nominate you to fix all Mustella tests going forward
He does have the magic touch :-)
But the current run has already logged some failures:
[java] Wrote file: c:/jenkins_slave/workspace/flex-sdk_mustella/mustella/tests/gumbo/components/Image/swfs/../styles/baselines/Image_Scale_stretch_smoothingQuality_high.png.bad.png.xml length: 9769
[java] Wrote file: c:/jenkins_slave/workspace/flex-sdk_mustella/mustella/tests/gumbo/components/Image/swfs/../styles/baselines/Image_Scale_stretch_smoothingQuality_high.png.bad.png length: 10698
[java] FAIL: gumbo/components/Image/styles/Image_styles_smoothingQuality Image_Scale_stretch_smoothingQuality_high
[java] SCRIPTDONE! 03:07:58.210
[java] GET /ScriptComplete?7 HTTP/1.1
This might be player version related if it ran OK the run before ?
Tom
Re: Mustella SDK tests
Posted by OmPrakash Muppirala <bi...@gmail.com>.
Great, I nominate you to fix all Mustella tests going forward ;-)
On May 23, 2015 9:52 PM, "Alex Harui" <ah...@adobe.com> wrote:
> Well, I must have magical powers because I did some digging last night and
> upon running a test that failed in the main SDK run in FDB, it passed.
> Then I restarted the main SDK run and it started passing!
>
> I’ve put some diagnostic trace statements in CompareBitmap so when it
> starts failing again we will have a little more information as to why the
> bitmap is the wrong size.
>
> Last time I went looking the tests also started passing for a while. Not
> sure why that is.
>
> I also removed the index.lock on the air workspace to see if that gets
> that test run to work better.
>
> -Alex
>
> On 5/22/15, 9:56 PM, "Alex Harui" <ah...@adobe.com> wrote:
>
> >I have shut down jenkins slave agent so no mustella runs for a while.
> >Please don’t restart jenkins as I may not have time to finish the
> >investigation tonight and will continue over the weekend. I’ll restart it
> >myself when I’m done digging.
> >
> >Thanks,
> >-Alex
> >
> >On 5/22/15, 5:47 AM, "Alex Harui" <ah...@adobe.com> wrote:
> >
> >>OK, go shut down Jenkins. I hope to get this FlexJS fix in today and
> >>then
> >>I’ll look at Mustella.
> >>
> >>-Alex
> >>
> >>On 5/22/15, 5:31 AM, "Tom Chiverton" <tc...@extravision.com> wrote:
> >>
> >>>
> >>>On 21/05/15 16:13, Alex Harui wrote:
> >>>> On 5/21/15, 8:01 AM, "Tom Chiverton" <tc...@extravision.com> wrote:
> >>>>
> >>>>> I've pressed 'keep forever' on
> >>>>> http://flex-mustella.cloudapp.net/job/flex-sdk_mustella/1484/
> >>>> OK, not sure what that button does, but I need the workspace not the
> >>>> artifacts.
> >>>>
> >>>>
> >>>
> >>>Well the most recent build appears to have hung, and there are some
> >>>bad.png's in the console log up to that point, so it should be good to
> >>>poke at any time now.
> >>>
> >>>Tom
> >>
> >
>
>
Re: Mustella SDK tests
Posted by Alex Harui <ah...@adobe.com>.
Well, I must have magical powers because I did some digging last night and
upon running a test that failed in the main SDK run in FDB, it passed.
Then I restarted the main SDK run and it started passing!
I’ve put some diagnostic trace statements in CompareBitmap so when it
starts failing again we will have a little more information as to why the
bitmap is the wrong size.
Last time I went looking the tests also started passing for a while. Not
sure why that is.
I also removed the index.lock on the air workspace to see if that gets
that test run to work better.
-Alex
On 5/22/15, 9:56 PM, "Alex Harui" <ah...@adobe.com> wrote:
>I have shut down jenkins slave agent so no mustella runs for a while.
>Please don’t restart jenkins as I may not have time to finish the
>investigation tonight and will continue over the weekend. I’ll restart it
>myself when I’m done digging.
>
>Thanks,
>-Alex
>
>On 5/22/15, 5:47 AM, "Alex Harui" <ah...@adobe.com> wrote:
>
>>OK, go shut down Jenkins. I hope to get this FlexJS fix in today and
>>then
>>I’ll look at Mustella.
>>
>>-Alex
>>
>>On 5/22/15, 5:31 AM, "Tom Chiverton" <tc...@extravision.com> wrote:
>>
>>>
>>>On 21/05/15 16:13, Alex Harui wrote:
>>>> On 5/21/15, 8:01 AM, "Tom Chiverton" <tc...@extravision.com> wrote:
>>>>
>>>>> I've pressed 'keep forever' on
>>>>> http://flex-mustella.cloudapp.net/job/flex-sdk_mustella/1484/
>>>> OK, not sure what that button does, but I need the workspace not the
>>>> artifacts.
>>>>
>>>>
>>>
>>>Well the most recent build appears to have hung, and there are some
>>>bad.png's in the console log up to that point, so it should be good to
>>>poke at any time now.
>>>
>>>Tom
>>
>
Re: Mustella SDK tests
Posted by Alex Harui <ah...@adobe.com>.
I have shut down jenkins slave agent so no mustella runs for a while.
Please don’t restart jenkins as I may not have time to finish the
investigation tonight and will continue over the weekend. I’ll restart it
myself when I’m done digging.
Thanks,
-Alex
On 5/22/15, 5:47 AM, "Alex Harui" <ah...@adobe.com> wrote:
>OK, go shut down Jenkins. I hope to get this FlexJS fix in today and then
>I’ll look at Mustella.
>
>-Alex
>
>On 5/22/15, 5:31 AM, "Tom Chiverton" <tc...@extravision.com> wrote:
>
>>
>>On 21/05/15 16:13, Alex Harui wrote:
>>> On 5/21/15, 8:01 AM, "Tom Chiverton" <tc...@extravision.com> wrote:
>>>
>>>> I've pressed 'keep forever' on
>>>> http://flex-mustella.cloudapp.net/job/flex-sdk_mustella/1484/
>>> OK, not sure what that button does, but I need the workspace not the
>>> artifacts.
>>>
>>>
>>
>>Well the most recent build appears to have hung, and there are some
>>bad.png's in the console log up to that point, so it should be good to
>>poke at any time now.
>>
>>Tom
>
Re: Mustella SDK tests
Posted by Alex Harui <ah...@adobe.com>.
OK, go shut down Jenkins. I hope to get this FlexJS fix in today and then
I’ll look at Mustella.
-Alex
On 5/22/15, 5:31 AM, "Tom Chiverton" <tc...@extravision.com> wrote:
>
>On 21/05/15 16:13, Alex Harui wrote:
>> On 5/21/15, 8:01 AM, "Tom Chiverton" <tc...@extravision.com> wrote:
>>
>>> I've pressed 'keep forever' on
>>> http://flex-mustella.cloudapp.net/job/flex-sdk_mustella/1484/
>> OK, not sure what that button does, but I need the workspace not the
>> artifacts.
>>
>>
>
>Well the most recent build appears to have hung, and there are some
>bad.png's in the console log up to that point, so it should be good to
>poke at any time now.
>
>Tom
Re: Mustella SDK tests
Posted by Tom Chiverton <tc...@extravision.com>.
On 21/05/15 16:13, Alex Harui wrote:
> On 5/21/15, 8:01 AM, "Tom Chiverton" <tc...@extravision.com> wrote:
>
>> I've pressed 'keep forever' on
>> http://flex-mustella.cloudapp.net/job/flex-sdk_mustella/1484/
> OK, not sure what that button does, but I need the workspace not the
> artifacts.
>
>
Well the most recent build appears to have hung, and there are some
bad.png's in the console log up to that point, so it should be good to
poke at any time now.
Tom
Re: Mustella SDK tests
Posted by Alex Harui <ah...@adobe.com>.
On 5/21/15, 8:01 AM, "Tom Chiverton" <tc...@extravision.com> wrote:
>I've pressed 'keep forever' on
>http://flex-mustella.cloudapp.net/job/flex-sdk_mustella/1484/
OK, not sure what that button does, but I need the workspace not the
artifacts.
-Alex
Re: Mustella SDK tests
Posted by Tom Chiverton <tc...@extravision.com>.
I've pressed 'keep forever' on
http://flex-mustella.cloudapp.net/job/flex-sdk_mustella/1484/
Tom
On 21/05/15 13:43, Alex Harui wrote:
> OK, I want to fix one more FlexJS bug first. Then we have to stop Jenkins
> after an SDK run completes with failures so the SWFs are still sitting
> around on the machine.
Re: Mustella SDK tests
Posted by Alex Harui <ah...@adobe.com>.
On 5/21/15, 4:11 AM, "Tom Chiverton" <tc...@extravision.com> wrote:
>Feel free to poke at it ;-)
OK, I want to fix one more FlexJS bug first. Then we have to stop Jenkins
after an SDK run completes with failures so the SWFs are still sitting
around on the machine.
>
>Would it maybe also be possible to run it in parallel on the other build
>machine ?
>That would rule out the issues we sometimes have where it's just
>flex-mustella that has an issue and it fixes it self when it cycles
>round to the same Player version again.
>I think apacheflexbuild has enough spare daily run time to do it's
>current jobs plus Mustella.
IIRC, the SDK mustella run takes 9 hours and ties up the machine so that
folks would have to wait much longer to find out if they broke other
builds or not. Having another Azure VM might be useful.
-Alex
Re: Mustella SDK tests
Posted by Tom Chiverton <tc...@extravision.com>.
Feel free to poke at it ;-)
Would it maybe also be possible to run it in parallel on the other build
machine ?
That would rule out the issues we sometimes have where it's just
flex-mustella that has an issue and it fixes it self when it cycles
round to the same Player version again.
I think apacheflexbuild has enough spare daily run time to do it's
current jobs plus Mustella.
Tom
On 20/05/15 15:40, Alex Harui wrote:
>
> On 5/20/15, 3:52 AM, "Tom Chiverton" <tc...@extravision.com> wrote:
>> Is it hard to kinda destroy the checkout and start again ?
> Don’t know. I’m not sure the bitmap failures have anything to do with the
> contents of the repo, but I could be wrong about that.
>
>> We're going to struggle to do more releases if we have a broken
>> Mustella...
> Yeah. If these runs have the bitmap failures, let’s shutdown Mustella and
> I will login and step through a test and see if I see anything.
>
> -Alex
>
>
> ______________________________________________________________________
> This email has been scanned by the Symantec Email Security.cloud service.
> For more information please visit http://www.symanteccloud.com
> ______________________________________________________________________
>
Re: Mustella SDK tests
Posted by Alex Harui <ah...@adobe.com>.
On 5/20/15, 3:52 AM, "Tom Chiverton" <tc...@extravision.com> wrote:
>Is it hard to kinda destroy the checkout and start again ?
Don’t know. I’m not sure the bitmap failures have anything to do with the
contents of the repo, but I could be wrong about that.
>
>We're going to struggle to do more releases if we have a broken
>Mustella...
Yeah. If these runs have the bitmap failures, let’s shutdown Mustella and
I will login and step through a test and see if I see anything.
-Alex
Re: Mustella SDK tests
Posted by Tom Chiverton <tc...@extravision.com>.
I've moved the player window to the top left corner, and Jenkins to the
lower right.
Is it hard to kinda destroy the checkout and start again ?
We're going to struggle to do more releases if we have a broken Mustella...
Tom
On 15/05/15 16:52, Alex Harui wrote:
> These pass for me.
>
> We’ve had this problem several times in the past on that machine. For
> some reason the player gets in a funky state. Once I was able to fix
> it by moving the player window, but I can’t reproduce the same problem
> on my computer.
>
> -Alex
>
> From: Tom Chiverton <tc@extravision.com <ma...@extravision.com>>
> Reply-To: "dev@flex.apache.org <ma...@flex.apache.org>"
> <dev@flex.apache.org <ma...@flex.apache.org>>
> Date: Friday, May 15, 2015 at 1:24 AM
> To: "dev@flex.apache.org <ma...@flex.apache.org>"
> <dev@flex.apache.org <ma...@flex.apache.org>>
> Subject: Mustella SDK tests
>
> Can someone on a Windows machine try some of the following Mustella
> tests ? They are consistently failing on the Jenkins job.
> spark/core/fxg/properties/fxgtemplate_masks mask_clip_mask
> spark/core/fxg/properties/fxgtemplate_tester blend_mode_radial_gradient_blends_blend_difference_radialgradient_fill
> gumbo/components/DropDownList/Properties/DropDownList_Properties
> And many more on
> http://flex-mustella.cloudapp.net/job/flex-sdk_mustella/1471/console
>
> --
> *Tom Chiverton*
> Lead Developer
> e: tc@extravision.com <ma...@extravision.com>
> p: 0161 817 2922
> t: @extravision <http://www.twitter.com/extravision>
> w: www.extravision.com <http://www.extravision.com/>
>
> Extravision - email worth seeing <http://www.extravision.com/>
> Registered in the UK at: 107 Timber Wharf, 33 Worsley Street,
> Manchester, M15 4LD.
> Company Reg No: 05017214 VAT: GB 824 5386 19
>
> This e-mail is intended solely for the person to whom it is addressed
> and may contain confidential or privileged information.
> Any views or opinions presented in this e-mail are solely of the
> author and do not necessarily represent those of Extravision Ltd.
>
>
> ______________________________________________________________________
> This email has been scanned by the Symantec Email Security.cloud service.
> For more information please visit http://www.symanteccloud.com
> ______________________________________________________________________
Re: Mustella SDK tests
Posted by Alex Harui <ah...@adobe.com>.
These pass for me.
We’ve had this problem several times in the past on that machine. For some reason the player gets in a funky state. Once I was able to fix it by moving the player window, but I can’t reproduce the same problem on my computer.
-Alex
From: Tom Chiverton <tc...@extravision.com>>
Reply-To: "dev@flex.apache.org<ma...@flex.apache.org>" <de...@flex.apache.org>>
Date: Friday, May 15, 2015 at 1:24 AM
To: "dev@flex.apache.org<ma...@flex.apache.org>" <de...@flex.apache.org>>
Subject: Mustella SDK tests
Can someone on a Windows machine try some of the following Mustella tests ? They are consistently failing on the Jenkins job.
spark/core/fxg/properties/fxgtemplate_masks mask_clip_mask
spark/core/fxg/properties/fxgtemplate_tester blend_mode_radial_gradient_blends_blend_difference_radialgradient_fill
gumbo/components/DropDownList/Properties/DropDownList_Properties
And many more on
http://flex-mustella.cloudapp.net/job/flex-sdk_mustella/1471/console
--
Tom Chiverton
Lead Developer
e: tc@extravision.com<ma...@extravision.com>
p: 0161 817 2922
t: @extravision<http://www.twitter.com/extravision>
w: www.extravision.com<http://www.extravision.com/>
[Extravision - email worth seeing]<http://www.extravision.com/>
Registered in the UK at: 107 Timber Wharf, 33 Worsley Street, Manchester, M15 4LD.
Company Reg No: 05017214 VAT: GB 824 5386 19
This e-mail is intended solely for the person to whom it is addressed and may contain confidential or privileged information.
Any views or opinions presented in this e-mail are solely of the author and do not necessarily represent those of Extravision Ltd.