I meant: 
I wouldn't like to commit that in this state.
Frédéric THOMAS

> From: webdoubl...@hotmail.com
> To: dev@flex.apache.org
> Subject: RE: [FDB] Integration
> Date: Tue, 29 Apr 2014 17:35:50 +0100
> 
> The integration of the worker code itself which requires, in order to be 
> total, to modify the Ant script that build FDB in order to link the Falcon 
> sources (AST classes used to evaluate the expression), at the moment I didn't 
> do this part and then still use the old code which required some integration 
> code to work, I would like to commit that but once the build script modified 
> and everything is working, I will commit. Is it fine for you like that ?
> 
> Bugs: I finally been able to reproduce the ones mentioned by Alexander but 
> I'm not sure the back and forth discussion have some big interess, anyway, 
> that's something that will go soon in JIRA
> 
> Frédéric THOMAS
> 
> > From: aha...@adobe.com
> > To: dev@flex.apache.org
> > Subject: Re: [FDB] Integration
> > Date: Tue, 29 Apr 2014 16:23:38 +0000
> > 
> > IMO, unless the discussion is going to cover proprietary aspects of
> > IntelliJ, it might be more like the "Apache Way" to share the code in a
> > git branch and continue to discuss on the list.
> > 
> > On 4/29/14 8:03 AM, "Alexander Doroshko"
> > <alexander.doros...@jetbrains.com> wrote:
> > 
> > >Yes, I'd gladly try it. We can switch to personal emails in order not to
> > >spam the list.
> > >
> > >On 29.04.2014 18:34, Frédéric THOMAS wrote:
> > >> Hi Alexander,
> > >>
> > >> I downloaded the files you indicated me + some missing files (UI,
> > >>ShineMP3, metallica.wav), created an IntelliJ project, build and run
> > >>successfully the application with Apache Flex SDK 4.12 / AIR 4.0 + my
> > >>version of FDB and was unable to reproduce both of the issues, it worked
> > >>as expected, so, that's a good news.
> > >>
> > >> Also, not on this test app but on another one [1], I've been able to
> > >>set a BP on the worker using IntelliJ once the worker has been loaded
> > >>but the variables was not readable and futur stops to the BP didn't
> > >>work, IJ shows up the SystemManager class instead.
> > >>
> > >> If you are interested to test my version, just tell me and I will send
> > >>it to you.
> > >>
> > >> Unfortunately, I haven't been able to work on it yesterday but will
> > >>continue tonight.
> > >>
> > >> Frédéric THOMAS
> > >>
> > >> [1] https://github.com/doublefx/downloadFileWorker
> > >>
> > >>
> > >>> Date: Tue, 29 Apr 2014 15:25:23 +0400
> > >>> From: alexander.doros...@jetbrains.com
> > >>> To: dev@flex.apache.org
> > >>> Subject: Re: [FDB] Integration
> > >>>
> > >>> Hi Frédéric,
> > >>> that's great that you are ready to put the effort into fdb with workers
> > >>> support!
> > >>>
> > >>> I have to say that IntelliJ IDEA still doesn't support debugging apps
> > >>> with workers. I spent some time on it but faced some bugs in fdb that
> > >>> were not that easy to workaround. That's not an excuse for me and I
> > >>>hope
> > >>> to finally support workers debug in the next version. Luckily fdb is
> > >>>not
> > >>> a black box for me any more.
> > >>>
> > >>> Below are some of the bugs that I faced. I played with this project:
> > >>> http://gotoandlearn.com/play.php?id=169. 'Download files' link contains
> > >>> already compiled MP3Worker.swf which is the main app.
> > >>>
> > >>> 1. When the worker is loaded it is impossible to switch to main thread
> > >>> using 'worker 0' command. Fdb says that active worker has changed but
> > >>> asking 'info workers' again says that actually it hasn't. See listing
> > >>>[1].
> > >>>
> > >>> 2. More important bug for the IDE integration is inability to set
> > >>> breakpoint without knowing file id. Files that belong to different
> > >>> workers may have the same id. When I ask to set a breakpoint in a file
> > >>> with given name but current active worker doesn't have file with this
> > >>> name then fdb sets the breakpoint inside completely different file that
> > >>> does belong to the current worker and has the same id as the file where
> > >>> I asked to set breakpoint. See listing [2].
> > >>>
> > >>> Alexander
> > >>>
> > >>> [1]
> > >>>
> > >>> C:\flex\air_sdk_13.0_asc_2.0\bin>fdb
> > >>> C:\flex\air_sdk_13.0_asc_2.0\bin>java -Dsun.io.useCanonCaches=false
> > >>>-Xms32m -Xmx512m
> > >>>-Dapplication.home="C:\flex\air_sdk_13.0_asc_2.0\bin\.." -jar
> > >>>"C:\flex\air_sdk_13.0_asc_2.0\bin\../lib/legacy/fdb.jar"
> > >>> Adobe fdb (Flash Player Debugger) [build development]
> > >>> Copyright (c) 2004-2007 Adobe, Inc. All rights reserved.
> > >>> (fdb) run
> > >>> Waiting for Player to connect
> > >>> Player connected; session starting.
> > >>> Set breakpoints and then type 'continue' to resume the session.
> > >>> [SWF] C:\work\flex_projects\MP3Worker\bin-debug\MP3Worker.swf -
> > >>>175?169 bytes after decompression
> > >>> (fdb) c
> > >>> [WorkerCreate] 1
> > >>> Additional ActionScript code has been loaded from a SWF or a frame.
> > >>> To see all currently loaded files, type 'info files'.
> > >>> Active worker has changed to worker 1
> > >>> Set additional breakpoints as desired, and then type 'continue'.
> > >>> (fdb) info workers
> > >>> Main Thread 0 - Running
> > >>> Worker 1 - Suspended (Active)
> > >>> [SWF] C:\work\flex_projects\MP3Worker\bin-debug\MP3Worker.swf -
> > >>>493?264 bytes after decompression
> > >>> Set additional breakpoints as desired, and then type 'continue'.
> > >>> (fdb) worker 0
> > >>> *Active worker has changed to worker Main Thread*
> > >>> Set additional breakpoints as desired, and then type 'continue'.
> > >>> (fdb) info workers
> > >>> *Main Thread 0 - Running*
> > >>> *Worker 1 - Suspended (Active)**
> > >>> *Set additional breakpoints as desired, and then type 'continue'.
> > >>> (fdb)
> > >>>
> > >>> [2]
> > >>>
> > >>> ...
> > >>> (fdb) info workers
> > >>> Main Thread 0 - Suspended (Active)
> > >>> Worker 1 - Suspended
> > >>> (fdb) show files
> > >>> 1 /Users/leebrimelow/Documents/Adobe Flash Builder
> > >>>4.6/MP3Worker/src/MP3Worker.as, MP3Worker.as (Main Thread)
> > >>> 2 /Users/leebrimelow/Documents/Adobe Flash Builder
> > >>>4.6/MP3Worker/src/Workers.as, Workers.as (Main Thread)
> > >>> 1 /Users/leebrimelow/Documents/Adobe Flash Builder
> > >>>4.6/MP3Worker/src/BackWorker.as, BackWorker.as (Worker 1)
> > >>> 2 
> > >>>/Users/kikko/work/tests/shineMP3_alchemy/src/fr/kikko/lab/ShineMP3Encode
> > >>>r.as, ShineMP3Encoder.as (Worker 1)
> > >>> *(fdb) break BackWorker.as:26**
> > >>> **Breakpoint 2: file MP3Worker.as, line 26**
> > >>> *(fdb)
> > >>>
> > >>>
> > >>> On 27.04.2014 4:01, Frédéric THOMAS wrote:
> > >>>> Hi,
> > >>>>
> > >>>> To resume a bit what I did and discover today:
> > >>>>
> > >>>> 1- I merged the donated code to the the current FDB, keeping the
> > >>>>existing fixes and the current ExpressionEvaluator (I guess It is
> > >>>>possible to use the new Falcon based implementation modifying the ANT
> > >>>>script which builds it, in order to link the Falcon sources).
> > >>>>
> > >>>> 2- The new commands are "worker <workerID>" used to select a
> > >>>>parlicular worker by its ID and "info workers", this is an example of
> > >>>>what gives a break inside a worker and what displays the 2 commands :
> > >>>>
> > >>>> (fdb) info workers
> > >>>> !mainThread! 0 - !workerRunning!
> > >>>> !inWorker! {worker=1} - !workerSuspended! !workerSelected!
> > >>>>
> > >>>> (fdb) worker 0
> > >>>> !workerChanged! !mainThread!
> > >>>>
> > >>>> (fdb) info workers
> > >>>> !mainThread! 0 - !workerRunning! !workerSelected!
> > >>>> !inWorker! {worker=1} - !workerSuspended!
> > >>>>
> > >>>> (fdb) worker 1
> > >>>> !workerChanged! 1
> > >>>>
> > >>>> (fdb) info workers
> > >>>> !mainThread! 0 - !workerRunning!
> > >>>> !inWorker! {worker=1} - !workerSuspended! !workerSelected!
> > >>>>
> > >>>> Note for IDE integration:
> > >>>>
> > >>>> 1- At the moment, watch, displays and breakpoints are set for 1
> > >>>>particular worker, it means, they need to be defined again when a
> > >>>>worker starts.
> > >>>> 2- "info breakpoints" does well display the breakpoints for the same
> > >>>>line and on different worker of the same class but without the
> > >>>>relative worker ID (I guess it could be easily fixed)
> > >>>>
> > >>>> I'll be back on it on Monday.
> > >>>>
> > >>>> Frédéric THOMAS                                        
> > >>                                          
> > >
> > 
>                                         
                                          

Reply via email to