回覆: [FlexJS] JS Linting

[shih.weil...@gmail.com]

- Reply message -
寄件者。+ ： "Alex Harui" 
收件者： "dev@flex.apache.org" 
主旨： [FlexJS] JS Linting
日期： 週六, 11 月 23 日, 2013 年 05:05


For sure we should use it, the question is whether it should fail the
build if we get lint errors, and I'm not sure how to get it to use strict
mode from a script.

On 11/22/13 1:02 PM, "Peter Ent"  wrote:

>I recall the Erik recommended using it. I'll have to get a copy and see
>why types of things it is complaining about. My instinct is to suggest
>using it, but having experienced gjslint, I'm a little wary of over doing
>it.
>
>--peter
>
>On 11/22/13 3:46 PM, "Alex Harui"  wrote:
>
>>I'm finishing up build scripts for the flex-asjs repo.
>>
>>Should we/can we put linting into the build?  I tried running JSHint from
>>the command-line by feeding it a ‹config option, but we get tons of
>>"Missing "use strict" statement" errors apparently because JSHint says
>>that if you are going to use the strict-mode checking you need to have
>>the
>>'use strict" directive at the top of the file.
>>
>>I turned off strict mode and still got some 81 errors.
>>
>>Thoughts?
>>
>>-Alex
>>
>

回覆: Build failed in Jenkins: flex-sdk_mustella-mobile #382

[shih.weil...@gmail.com]

，; !.，+

- Reply message -
寄件者： "Alex 48" 
收件者： "dev@flex.apache.org" , "Erik de Bruin" 

副本： "Mark Kessler" , "jmcl...@apache.org" 

主旨： Build failed in Jenkins: flex-sdk_mustella-mobile #382
日期： 週六, 11 月 23 日, 2013 年 03:31
！丫＇+.53+4..OK, so we should see the mustella server pass as well.

Thanks for all your hard work on ScrollingStageText.  Good luck getting
all the last bugs out of it.

-Alex

On 11/22/13 11:28 AM, "Maurice Amsellem" 
wrote:

>I run the whole ActionBar suite, with AIR 3.9, and has not fails!
>
>Excellent.
>
>Thanks Alex for the hint
>
>Maurice
>
>-Message d'origine-
>De : Alex Harui [mailto:aha...@adobe.com]
>Envoyé : vendredi 22 novembre 2013 19:42
>À : dev@flex.apache.org; Erik de Bruin
>Cc : Mark Kessler; jmcl...@apache.org
>Objet : Re: Build failed in Jenkins: flex-sdk_mustella-mobile #382
>
>The png.xml files can be sensitive to what else runs before it.  Did you
>run the entire ActionBar folder?
>
>./mini_run.sh -mobile tests/mobile/components/ActionBar/
>
>If you still get a bad.png.xml, post it so I can see the differences.
>
>
>-Alex
>
>On 11/22/13 10:07 AM, "Maurice Amsellem" 
>wrote:
>
>>I run mustella tests with current SDK (develop) and AIR 3.9, using your
>>new XML DL file.
>>The tests failed as before,
>>But what is surprising, is that the XML files have a lot of differences
>>(except for the ScrollingStageTextInputSkin tags which are the same).
>>
>>See by yourself:
>>https://www.dropbox.com/sh/y1iw80qbe15pypd/cU4YCeiUSN
>>
>>Maurice
>>
>>-Message d'origine-
>>De : Alex Harui [mailto:aha...@adobe.com] Envoyé : vendredi 22 novembre
>>2013 18:28 À : dev@flex.apache.org; Erik de Bruin Cc : Mark Kessler;
>>jmcl...@apache.org Objet : Re: Build failed in Jenkins:
>>flex-sdk_mustella-mobile #382
>>
>>Sorry, you're right.  AIR 3.4.
>>
>>Anyway, I didn't realize there was already a png.xml for it.  I updated
>>the png.xml with the ScrollingStageTextInputSkin.  Let's see if it
>>passes on the next run.
>>
>>-Alex
>>
>>On 11/22/13 8:54 AM, "Maurice Amsellem" 
>>wrote:
>>
>>>Is FP also used in air mobile app?
>>>
>>>I thought it used airglobal.swc only?
>>>
>>>Maurice
>>>
>>>-Message d'origine-
>>>De : Alex Harui [mailto:aha...@adobe.com] Envoyé : vendredi 22
>>>novembre
>>>2013 17:47 À : dev@flex.apache.org; Erik de Bruin Cc : Mark Kessler;
>>>jmcl...@apache.org Objet : Re: Build failed in Jenkins:
>>>flex-sdk_mustella-mobile #382
>>>
>>>This test passes for me with FP 11.1.
>>>
>>>On 11/22/13 7:48 AM, "Maurice Amsellem" 
>>>wrote:
>>>
Eric,

Here is what  I did:

- running with the latest SDK a standalone mobile air app (not
mustella) that displays an action bar with embedded font and
underline => no artifact
- running current mustella test mobile/Action with current SDK =>
artifact is there
- running current mustella test mobile/Action with SKD 4.11 and 4.10
=> artifact is there Below is the procedure that I used to change the
SDK used in mustella.

IMO, the cause of this issue is  something has changed recently in
the mustella test itself, that caused the artifact.

The test that needs to be done yet is to revert mustella test recent
changes, and try again...


Procedure to change SDK used by  mustella:
1) In mustella,  local.properties, set sdk.dir= C:/Program Files
(x86)/Apache Flex/Apache_Flex_SDK_411_GA apollo_location=${sdk.dir}

then run again the same test:
$ ./mini_run.sh  -mobile
tests/mobile/components/ActionBar/styles/ActionBar_Styles.mxml

The log says:
[echo] sdk.dir: C:/Program Files (x86)/Apache
Flex/Apache_Flex_SDK_411_GA
...
[echo] apollo_location is C:/Program Files (x86)/Apache
Flex/Apache_Flex_SDK_411_GA

WDYT?

Maurice

-Message d'origine-
De : Erik de Bruin [mailto:e...@ixsoftware.nl] Envoyé : vendredi 22
novembre 2013 16:11 À : Maurice Amsellem Cc : dev@flex.apache.org;
Mark Kessler; jmcl...@apache.org Objet : Re: Build failed in Jenkins:
flex-sdk_mustella-mobile #382

Well, my records show this test started failing on Nov. 18th.

EdB



On Fri, Nov 22, 2013 at 3:44 PM, Maurice Amsellem
 wrote:
> Hi, I showed evidence this issue was already present in SDK 4.10
>and 4.11.
> So it's not caused by the fix on mobile TextInput.
>
> Maurice
>
> -Message d'origine-
> De : Erik de Bruin [mailto:e...@ixsoftware.nl] Envoyé : vendredi 22
> novembre 2013 15:36 À : dev@flex.apache.org Cc : Mark Kessler;
> jmcl...@apache.org; Maurice Amsellem Objet : Re: Build failed in
> Jenkins: flex-sdk_mustella-mobile #382
>
> Hi, since the Mustella VM seems to have settled down now, I think
>we can confidently say this failure needs to be addressed - it is
>not an artifact of the build machine.
>
> Anyone?
>
> EdB
>
>
>