Hi Chris,
You can check my last posts here [1]. It is an explanation how to import and
run tests in Intellij, but you could do this also in Flash Builder in the
same manner.
Additionally module:
automation_tests needs flexunit.swc
automation_core needs flexunit.swc
automation_apps needs flexuni
I confirmed that the CI server also timed out. IMO, I wouldn’t worry
about the arrow tests until you see them fail in a run where nothing
failed before it.
I ran the tests in the debugger and tons of trace statements were dumped
to the console. They look like:
[trace] 340 71.95 STR2ColBB auto
Oops, I meant ‘double-check Justin’s findings'
On 1/24/15, 12:50 PM, "Alex Harui" wrote:
>Ha, that's what I get for not double checking justin. I'll revert in a
>bit.
>Sent via the PANTECH Discover, an AT&T 4G LTE smartphone.
>
>Harbs wrote:
>
>
>Here’s the texts which are all clearly public d
Well, IMO, the default Spark skins are a middle ground where they support
certain styles, some of which are not CSS compliant. But a Spark Skin can
be completely non-styleable. If you create it via FXG, or just use MXML
without any code, you pretty much can’t get it to support styles.
I’m defini
Spark skinning also obeys CSS rules. In a spark skin, you can for example,
call getStyle("borderColor") and use that to draw a Rect with the given
color. Or call getStyle("themeColor") and apply that color to a
LinearGradient Rect.
In addition, on the SVG side, individual shapes can be fully sty
After talking with Alex a bit more, it seems there are two paths we can
take, both valid and both really worth exploring.
One is the MXML/designer UI path where skins are specific either in MXML,
drawn in some way, and/or created using a design tool. This skins follow
from Spark skins.
The other
Thanks Harbs!
That got me going a little further. I read in the FAQ about needing the
flexunit.swc in the libs folder of automation_core. I put the one I found at
C:\Program Files\Adobe\Adobe Flash Builder 4.7 (64
Bit)\eclipse\plugins\com.adobe.flexbuilder.flexunit_4.7.0.349722\flexunitframe
I think I have a clue. I am getting an error due to a script timeout.
Have you examined the test output on the build server to see if had
something like this?
I think the results are in
bin/apps/automation_apps/TEST-UnitTest.Tests.ContainerAttributeTest.xml
-Alex
On 1/28/15, 1:02 PM, "Harbs" w
run “ant all” on TLF.
On Jan 28, 2015, at 11:00 PM, Chris Martin wrote:
> How can I run these tests locally on my machine? :)
>
> Chris
>
>> From: harbs.li...@gmail.com
>> Subject: Re: Failing TLF Builds
>> Date: Wed, 28 Jan 2015 22:31:30 +0200
>> To: dev@flex.apache.org
>>
>> So maybe this i
How can I run these tests locally on my machine? :)
Chris
> From: harbs.li...@gmail.com
> Subject: Re: Failing TLF Builds
> Date: Wed, 28 Jan 2015 22:31:30 +0200
> To: dev@flex.apache.org
>
> So maybe this is something that needs to be fixed to make the tests more
> stable.
>
> On Jan 28, 20
So maybe this is something that needs to be fixed to make the tests more stable.
On Jan 28, 2015, at 10:02 PM, Alex Harui wrote:
> Actually, in the little I’ve looked at the test so far, I
> have concerns that a failing test could leave the suite in a bad state so
> the only important thing is t
They all pass for me:
[flexunit] Tests run: 428, Failures: 0, Errors: 0, Skipped: 22, Time elapsed:
340.071 sec
On Jan 28, 2015, at 10:02 PM, Alex Harui wrote:
> I’m also seeing that columnContainerBreakTest2000 failed in the last two
> runs. It and several others are failing for me too. Are
Hi,
>> bits are at the correct URL, and more importantly, if the bits are
>> propagating to the mirrors, then you are good to go.
>
> How do I check that?
By going here and looking at the age histogram at the bottom.
http://www.apache.org/mirrors/
Justin
I’m also seeing that columnContainerBreakTest2000 failed in the last two
runs. It and several others are failing for me too. Are they passing for
you? Maybe if I look to see why they fail it will shed light on the
Arrows test. Actually, in the little I’ve looked at the test so far, I
have conce
There are about 20 tests that are ignored because they had issues when they
were ported.
My concern is actually two specific tests:
GeneralFunctionsTest.arrowDown
GeneralFunctionsTest.arrowUp
These are passing for me locally, but are failing on the CI server.
The reason I suspect there’s a rea
Working backwards from JS/CSS seems to be the best route, no matter how it
is implemented. We know there are some least-common denominators such as
background color and border, so perhaps we start there to formulate a plan.
I do think we should try and use the core.graphics package which is
style-
I just ran the tests myself. I get:
[flexunit] FlexUnit test spanElementReplaceTextInvalidPos in suite
UnitTest.Tests.ElementOperationTest was ignored.
[flexunit] FlexUnit test testAllEnumProps in suite
UnitTest.Tests.AllParaAttributeTest was ignored.
[flexunit] FlexUnit test endTabLongStringTe
On 1/28/15, 9:45 AM, "Bertrand Delacretaz" wrote:
>Hi,
>
>On Wednesday, January 28, 2015, Alex Harui wrote:
>
>> My current thinking is that we keep the Apache Headers in the
>> source file [1] and give attribution in a NOTICE.test file...
>
>> Justin’s position is that he wants to remove the
On 1/28/15, 10:03 AM, "Erik de Bruin" wrote:
>> bits are at the correct URL, and more importantly, if the bits are
>> propagating to the mirrors, then you are good to go.
>
>How do I check that?
I think that we’ll eventually post this links like this on our website:
http://www.apache.org/dyn/
I don’t have well-formed thoughts on this topic, but the following
occurred to me.
On the JS side, the browser has a powerful CSS engine. Drawing skins in
AS/MXML won’t leverage that. So these kinds of themes may not be
styleable. I think that’s fine, but IMO, Spark skins were designed to
have
> bits are at the correct URL, and more importantly, if the bits are
> propagating to the mirrors, then you are good to go.
How do I check that?
EdB
--
Ix Multimedia Software
Jan Luykenstraat 27
3521 VB Utrecht
T. 06-51952295
I. www.ixsoftware.nl
I'm pretty sure that if it didn't happen on dev@, it didn't happen...
EdB
On Wed, Jan 28, 2015 at 6:59 PM, Alex Harui wrote:
> Hi Bertrand,
>
> I like this idea of tracking discussions in JIRA so we can reference it
> later. I’m not sure if you are aware, but JIRA activity goes to
> issues@fl
Hi Bertrand,
I like this idea of tracking discussions in JIRA so we can reference it
later. I’m not sure if you are aware, but JIRA activity goes to
issues@flex.a.o and not dev@ so we are effectively moving this discussion
off-list. I just wanted to make sure that’s ok with you and whether that
On 1/28/15, 8:00 AM, "Erik de Bruin" wrote:
>Does "dist@" send commit messages when committing to SVN? I don't see
>it in my list, but I see the 4.14 files here:
Well, I’m pretty sure that normally it does, but I saw an email on an
Apache-wide list yesterday where someone else was complaining
Hi,
On Wednesday, January 28, 2015, Alex Harui wrote:
> My current thinking is that we keep the Apache Headers in the
> source file [1] and give attribution in a NOTICE.test file...
> Justin’s position is that he wants to remove the Apache header and replace
> it with a header that calls out t
> My current thinking is that we keep the Apache Headers in the source file [1]
> and give attribution in a NOTICE.test file that would get merged with the
> main NOTICE file if we ever release these tests.
Just wanted to give my thoughts. I think the above should be fine. Right in
the Apache h
On Tue, Jan 27, 2015 at 5:48 PM, Erik de Bruin wrote:
>>> IIRC, the bits have to be moved to the dist.apache.org release area and
>>> the installer config needs to point to it. Or have you done that and the
>>> commit email just hasn’t shown up yet?
>>
>> No, I'm on hotel WiFi, so until the other
Arg!
Now the arrow tests are failing again.
I have a feeling that this is a legitimate failure, but it’s a sometimes/some
machine failure.
If someone is constantly getting these failures, I have something I’d like them
to try.
Thanks,
Harbs
On Jan 28, 2015, at 9:23 AM, Harbs wrote:
> At le
28 matches
Mail list logo
