Re: ADG and cellIndex.toString()

According to [1], array is way faster and a little tweak might make it even more so. EdB 1: http://jacksondunstan.com/articles/1305 On Tue, Sep 10, 2013 at 8:39 AM, Justin Mclean wrote: > HI, > >> If cellSelectionData[uid] is an Array, then yes, int lookup should be >> faster. If it is an ob

Re: ADG and cellIndex.toString()

HI, > If cellSelectionData[uid] is an Array, then yes, int lookup should be > faster. If it is an object, I would think it would be the same although > maybe the int gets converted to string in the VM so you save a function > call. It's an object but arrays extend object anyway. Do you know if t

Re: ADG and cellIndex.toString()

If cellSelectionData[uid] is an Array, then yes, int lookup should be faster. If it is an object, I would think it would be the same although maybe the int gets converted to string in the VM so you save a function call. On 9/9/13 10:13 PM, "Justin Mclean" wrote: >Hi, > >Was looking into the cel

Re: [DRAFT] Apache Flex September 2013 Report

> may be one of our sticking points. Compiling from source still requires > some voodoo magic if you are doing it all from scratch and you haven't done > it before.. Most committers (!) should be able to do this... And should actually have done this many, many times :-) EdB -- Ix Multimedia

ADG and cellIndex.toString()

Hi, Was looking into the cell selection code and noticed that the ADG has this throughout: cellSelectionData[uid][columnIndex.toString()] Is there any sane reason for not having the column index as just columnIndex (ie a plain int)? Int index lookup is faster than string index look up right? O

Re: [DRAFT] Apache Flex September 2013 Report

Let' not read too much into the fact that committers take vacations during their summer - and that summer is a different time of year in the southern hemisphere. On Sep 9, 2013, at 6:53 PM, Dasa Paddock wrote: > GitHub has some good tools for this too. For example, this shows that there > were

Re: git commit: [flex-sdk] [refs/heads/develop] - FLEX-33702: Added column sort type access to the datagrid columns.

Looks fine. One concern is that you changed an interface. We got burned by changing other interfaces in the past. Remember the IList snafu? But I don't know if there are any custom ISortField implementations in the wild so it may be ok. -Alex On 9/9/13 5:00 PM, "mkess...@apache.org" wrote:

Re: [DRAFT] Apache Flex September 2013 Report

GitHub has some good tools for this too. For example, this shows that there were six committers to the flex-sdk repo in July and August: https://github.com/apache/flex-sdk/graphs/contributors?from=2013-06-30&to=2013-08-31&type=c --Dasa On Sep 9, 2013, at 5:11 PM, Justin Mclean wrote: > Hi, >

Re: [DRAFT] Apache Flex September 2013 Report

Hi, It's looking at both the git and svn repos but is missing a few git repos eg falcon. I don't think the numbers are 100% correct but do give a fair indication, and I think it fair to say there are less active committers for the Flex SDK project than there were 6 months ago. Justin

Re: [DRAFT] Apache Flex September 2013 Report

On Sep 9, 2013, at 5:03 PM, Justin Mclean wrote: > Hi, > >> Do you think the number of committers is a threat to the project? > > I think it's a concern and may make further releases harder. It's something > that needs to be watched and users/people on dev list encouraged more to take > part

Re: [DRAFT] Apache Flex September 2013 Report

On Sep 9, 2013, at 4:56 PM, Justin Mclean wrote: > Hi, > >> Thanks for the input. What is the "release candidate vote issue"? > > An RC was voted in as released but not released. > > IMO we had more RC than we should due to lack of committer involvement (and > the release manager job harder)

Re: [DRAFT] Apache Flex September 2013 Report

On 9/9/13 4:56 PM, "Justin Mclean" wrote: >Hi, > >> Thanks for the input. What is the "release candidate vote issue"? > >An RC was voted in as released but not released. > >IMO we had more RC than we should due to lack of committer involvement >(and the release manager job harder). With the nu

Re: [DRAFT] Apache Flex September 2013 Report

I know I was in the middle of my testing suite when the vote expired. I was on vacation for two days and the full Mustella suite takes about 24 hours (if it runs correctly the first time) to run, so I missed it. We still don't make it easy for people to evaluate the non-binary setup which may be

Re: [DRAFT] Apache Flex September 2013 Report

Hi, Also it might be worth mentioning the (not insignificant) work on FlexUnit and the SWZ Framework donation (although not sure where that's at). Thanks, Justin

Re: [DRAFT] Apache Flex September 2013 Report

Hi, > Do you think the number of committers is a threat to the project? I think it's a concern and may make further releases harder. It's something that needs to be watched and users/people on dev list encouraged more to take part and become committers. Thanks, Justin

Re: [DRAFT] Apache Flex September 2013 Report

- decreasing number of active committers [2] > > 2. http://www.ohloh.net/p/ApacheFlex/contributors/summary I dont think the Ohloh numbers are correct. They are not looking at all the projects and for most of them they are still looking at the old SVN repos [3] Someone with an Ohloh account can

Re: [DRAFT] Apache Flex September 2013 Report

Hi, > Thanks for the input. What is the "release candidate vote issue"? An RC was voted in as released but not released. IMO we had more RC than we should due to lack of committer involvement (and the release manager job harder). With the number of committers we have in the project a release

Re: [DRAFT] Apache Flex September 2013 Report

On Sep 9, 2013, at 4:40 PM, Justin Mclean wrote: > Hi, > >> Mailing list traffic is so high that folks have stopped reading emails. > Traffic has actually decreasing for the last 7 months month on month so not > sure that's correct [1]. I think it's just that some people prefer forums. > > Cur

Re: [DRAFT] Apache Flex September 2013 Report

Thanks for the input. What is the "release candidate vote issue"? Thanks, -Alex On 9/9/13 4:40 PM, "Justin Mclean" wrote: >Hi, > >> Mailing list traffic is so high that folks have stopped reading emails. >Traffic has actually decreasing for the last 7 months month on month so >not sure that's

Re: [DRAFT] Apache Flex September 2013 Report

Hi, > Mailing list traffic is so high that folks have stopped reading emails. Traffic has actually decreasing for the last 7 months month on month so not sure that's correct [1]. I think it's just that some people prefer forums. Currently we've had about 10 people sign up to the dev list and 20

Re: [Falcon]

No, I would fix the sdk. -Darrell On 9/9/13 4:17 AM, "Cyrill Zadra" wrote: >In flex sdk project there are a few "Warning: variable definition >duplicates function parameter" warnings if it s compiled by falcon. So I >check those lines and it seems for me that falcon compiler is correct and >sd

Re: Changes to LayoutBase?

Yeah. I saw that. The patch is only for 4.7 though. (I'm still using 4.5…) Not a big deal either way. Removing the layout tag is not a problem. Is layout supposed to be valid or not? On Sep 9, 2013, at 8:22 PM, Alex Harui wrote: > Because of a bug in FB. There is a patch available. See commen

RE: Changes to LayoutBase?

>Okay. So the tag is wrong. FB is creating a spark application as it would >create an mx one I guess? It's not that it's wrong, you can actually use layout as a property, even in the tag context to specify a layout class. It's just that a string isn't valid. Mike

Re: Changes to LayoutBase?

Okay. So the tag is wrong. FB is creating a spark application as it would create an mx one I guess? Sorry for the noise… ;-) Harbs On Sep 9, 2013, at 8:38 PM, OmPrakash Muppirala wrote: > I dont think this has anything to do with the FB bug. > > With a spark application, you cannot specify lay

Re: Changes to LayoutBase?

I dont think this has anything to do with the FB bug. With a spark application, you cannot specify layout like that. you need to specify layout like this: . . . Thanks, Om On Mon, Sep 9, 2013 at 10:33 AM, Harbs wrote: > Yeah. I saw that. The patch is only for 4.7 though

Re: [Falcon]

Fix SDK, certainly. EdB On Mon, Sep 9, 2013 at 6:47 PM, Darrell Loverin wrote: > No, I would fix the sdk. > > > -Darrell > > On 9/9/13 4:17 AM, "Cyrill Zadra" wrote: > >>In flex sdk project there are a few "Warning: variable definition >>duplicates function parameter" warnings if it s compile

Re: [DRAFT] Apache Flex September 2013 Report

Typo: "Apache Flex SDK 4.19.0" must be "Apache Flex SDK 4.10.0 " Download numbers are a bit off. I would put it like this: - Since beginning of this year, the Installer has been used to install over 16,000 copies of the SDK. There were more than 5000 installs of 4.10.0 in the month since its rel

Re: Changes to LayoutBase?

Because of a bug in FB. There is a patch available. See comments in the JIRA issue. On 9/9/13 10:12 AM, "Harbs" wrote: >Interesting. > >But that kind of begs the question: why is it happening? > >On Sep 9, 2013, at 6:09 PM, Erik de Bruin wrote: > >> Is this maybe related to: >> >> https://iss

[DRAFT] Apache Flex September 2013 Report

Here's the draft of our September report. -Alex - Apache Flex is an application framework for easily building Flash-based applications for mobile devices, the browser and desktop. RELEASES Apache Flex SDK 4.19.0 was released on 8/6/13. Apache Flex Installer 2.6.0 was released on 8/6/

Re: Changes to LayoutBase?

Interesting. But that kind of begs the question: why is it happening? On Sep 9, 2013, at 6:09 PM, Erik de Bruin wrote: > Is this maybe related to: > > https://issues.apache.org/jira/browse/FLEX-33636 > > EdB > > > > On Mon, Sep 9, 2013 at 4:01 PM, Harbs wrote: >> Trying to set layout in mx

RE: Mustella question

Thanks Alex for the details, so, from what you say, it should probably be more SWF timeout than step timeout which would explained the difference in the number of tests ran. My server point the only one mustella/tests folder I have on this machine and I can see the list on http://locahost Unfort

Re: Mustella question

On 9/9/13 6:01 AM, "Frédéric THOMAS" wrote: >> "./mini_run.sh -all" will run all the main Mustella tests (close to >4) tests, taking about 8 - 10 hours. > >I ran "-all" on purpose and it gave me about 20500 tests passes and 20 >failed included 17 from the Marshall plan > >> If any of these

Re: Mustella question

On 9/9/13 8:43 AM, "Alex Harui" wrote: > >> >>My problems are: >>1- Why the MarshallPlan tests don't pass ? >Marshall plan tests require a localhost or localhost:8080 server pointed >at the mustella tests folder so we can load child SWFs into a sandbox. I see that you already set up a server. H

Re: Changes to LayoutBase?

Is this maybe related to: https://issues.apache.org/jira/browse/FLEX-33636 EdB On Mon, Sep 9, 2013 at 4:01 PM, Harbs wrote: > Trying to set layout in mxml in a Spark Application (layout="absolute") > results in the following error in version 4.10.0: > > Initializer for 'layout': values of ty

Re: Mustella failures

Set the launch URL to be something like 'http://a.a' and while it is waiting to connect, launch another debug enabled SWF. Tom On 09/09/2013 04:37, Alex Harui wrote: way more efficient to use FDB (for me). Supposedly there is a way to create a generic project in FB, launch it in the debugger,

Changes to LayoutBase?

Trying to set layout in mxml in a Spark Application (layout="absolute") results in the following error in version 4.10.0: Initializer for 'layout': values of type spark.layouts.supportClasses.LayoutBase cannot be represented in text. Was this an intentional change? Harbs

Re: Mustella question

Running -all should give you exactly 35966 tests (this is from the last run of the suite on the Mustella VM)... So it might be worth your time first figuring out why you are missing about 15000 tests before you go into more details. EdB On Mon, Sep 9, 2013 at 3:01 PM, Frédéric THOMAS wrote: >>

RE: Mustella question

> "./mini_run.sh -all" will run all the main Mustella tests (close to 4) tests, taking about 8 - 10 hours. I ran "-all" on purpose and it gave me about 20500 tests passes and 20 failed included 17 from the Marshall plan > If any of these tests fail, those tests will be written to 'failures.tx

Re: Mustella question

Fred, "./mini_run.sh -all" will run all the main Mustella tests (close to 4) tests, taking about 8 - 10 hours. If any of these tests fail, those tests will be written to 'failures.txt'. If you want to re-run these failed tests, you simply run "./mini_run.sh -failures". No additional arguments

Mustella question

Hi, That’s several time I try to run the Mustella tests “ ./mini_run.sh –all “, in more than 3 failed tests, it gives me a bunch of failures regarding the Marshall plan, though I run my web server on the port 80 to serving mustella/tests. So, my first question is, what could be wrong ?

Re: [Falcon]

In flex sdk project there are a few "Warning: variable definition duplicates function parameter" warnings if it s compiled by falcon. So I check those lines and it seems for me that falcon compiler is correct and sdk code should be fixed. Any other suggestions? https://issues.apache.org/jira/brows

Re: Jenkins build is back to normal : flex-sdk_mustella-mobile #129

Until I break it again :-)

Re: Jenkins build is back to normal : flex-sdk_mustella-mobile #129

Awesome! On Sun, Sep 8, 2013 at 11:42 PM, Erik de Bruin wrote: > Yes! > > Thanks all for getting Mustella back to 'zarro boogs found' ;-) > > As a bonus: the SDK (main) tests pass for 11.9/3.9! > > EdB > > > > On Mon, Sep 9, 2013 at 4:13 AM, wrote: > > See