Last 30 days

• Dec 12, 2009

  FAQ (FAQ or Frequently Asked Questions) Wiki page edited by debert   -   Revision r313 Edited wiki page through web user interface.

  Revision r313 Edited wiki page through web user interface.

Earlier this year

• Nov 12, 2009

  MemoryManagement (Understanding memory management) Wiki page edited by debert   -   Revision r312 Edited wiki page through web user interface.

  Revision r312 Edited wiki page through web user interface.
• Nov 11, 2009

  issue 110 (Flex SWC) commented on by debert   -   @Joe Nash Please do! The code is there, nothing is stopping you. As I said, how much time would that save people? $mxmlc br/com/stimuli/loading/BulkLoader.as -source-path . -o bulkloader.swc That's it, it's just one command away and you've got a swc, Cheers

  @Joe Nash Please do! The code is there, nothing is stopping you. As I said, how much time would that save people? $mxmlc br/com/stimuli/loading/BulkLoader.as -source-path . -o bulkloader.swc That's it, it's just one command away and you've got a swc, Cheers
• Nov 11, 2009

  issue 110 (Flex SWC) commented on by joe.n...@atticmedia.com   -   Shame... A SWC would have made Flex integration really quick! I might try and create one for myself. :)

  Shame... A SWC would have made Flex integration really quick! I might try and create one for myself. :)
• Nov 04, 2009

  issue 113 ([BulkLoader] Cannot use an instance that has been cleared fr...) commented on by debert   -   Hi Bach. The missing information here is if you are calling loader.clear(). If you aren't, it's a bug and please let me know so I can fix. If you are calling clear, why should be able to access the same loader? It's not the same loader. Once you've called clear, it's lost all of it's events, all of its assets, all of it's connections. Whatever you were willing to do with that loader instance will fail, so we nullify the reference to make that easier to grasp. What I mean is that it should have named the 'clear' method to 'nukeEverythingAndMakeThisInstanceTotallyUselessForever' If you wish to remove items from a instance and to keep using that instance, then the removeAll is the method to call. Cheers Arthur

  Hi Bach. The missing information here is if you are calling loader.clear(). If you aren't, it's a bug and please let me know so I can fix. If you are calling clear, why should be able to access the same loader? It's not the same loader. Once you've called clear, it's lost all of it's events, all of its assets, all of it's connections. Whatever you were willing to do with that loader instance will fail, so we nullify the reference to make that easier to grasp. What I mean is that it should have named the 'clear' method to 'nukeEverythingAndMakeThisInstanceTotallyUselessForever' If you wish to remove items from a instance and to keep using that instance, then the removeAll is the method to call. Cheers Arthur
• Nov 04, 2009

  issue 113 ([BulkLoader] Cannot use an instance that has been cleared fr...) commented on by bach44   -   Yes I'm familiar with the documentation, yet can think of no reason why I shouldn't always be able to retrieve a loader instance. In a component class within an added to stage handler, I am calling //CODE START protected var _loader:BulkLoader //then within the handler _loader = BulkLoader.getLoader( "imgsLoader" ) //CODE END This works the first time I reference it in a sub class. However, later on, when I need to reference the exact same loader, '_loader' is null WTF?

  Yes I'm familiar with the documentation, yet can think of no reason why I shouldn't always be able to retrieve a loader instance. In a component class within an added to stage handler, I am calling //CODE START protected var _loader:BulkLoader //then within the handler _loader = BulkLoader.getLoader( "imgsLoader" ) //CODE END This works the first time I reference it in a sub class. However, later on, when I need to reference the exact same loader, '_loader' is null WTF?
• Oct 30, 2009

  issue 109 (Flex Support) commented on by bl...@bentbit.com   -   Thats a great way to handle it! Thanks

  Thats a great way to handle it! Thanks
• Oct 30, 2009

  issue 106 (getProgressForItems return NaN for weightPercent until every...) Status changed by debert   -   This is fixed in revision 311. Regards

  Status: Fixed

  This is fixed in revision 311. Regards

  Status: Fixed
• Oct 30, 2009

  issue 104 (MP4 Issues) Status changed by debert   -   Hi Krislocascio. I've tried the exact same code and mp4 here, with serveral plugin versions, and they all work. If you are still interested in this, maybe you can try running it on the browser, with another player version (either you have some other code interfering, or it's the cs4 player that's messing this up - which is unfortunately all too common). Regards Arthur

  Status: Invalid

  Hi Krislocascio. I've tried the exact same code and mp4 here, with serveral plugin versions, and they all work. If you are still interested in this, maybe you can try running it on the browser, with another player version (either you have some other code interfering, or it's the cs4 player that's messing this up - which is unfortunately all too common). Regards Arthur

  Status: Invalid
• Oct 30, 2009

  r311 (updated test suite for the name constructor parameter option...) committed by debert   -   updated test suite for the name constructor parameter optional

  updated test suite for the name constructor parameter optional
• Oct 30, 2009

  issue 109 (Flex Support) Status changed by debert   -   Hi Blake, Just checked in revision 310, which should do it. Now the 'name' parameter is optional, if passed null, bulkloader will generate a unique one for you. Cheers

  Status: Fixed

  Hi Blake, Just checked in revision 310, which should do it. Now the 'name' parameter is optional, if passed null, bulkloader will generate a unique one for you. Cheers

  Status: Fixed
• Oct 30, 2009

  r310 (Made 'name' constructor parameter optional, should fix issue...) committed by debert   -   Made 'name' constructor parameter optional, should fix issue #109 , thanks Blake.

  Made 'name' constructor parameter optional, should fix issue #109 , thanks Blake.
• Oct 30, 2009

  issue 111 (Problems with asdoc) Status changed by debert   -   Hi Ruy. Just fixed this in revision 309, thanks for the report. Oddly, my asdoc (sdk 3.3.0) was building the documentation without issues. Sigh.

  Status: Fixed

  Hi Ruy. Just fixed this in revision 309, thanks for the report. Oddly, my asdoc (sdk 3.3.0) was building the documentation without issues. Sigh.

  Status: Fixed
• Oct 30, 2009

  r309 (Fixed mal formed tags on asdoc, issue #111, thanks Ruy Adorn...) committed by debert   -   Fixed mal formed tags on asdoc, issue #111 , thanks Ruy Adorno

  Fixed mal formed tags on asdoc, issue #111 , thanks Ruy Adorno
• Oct 30, 2009

  r308 (Fixed events tags, issue #108 , thanks Blake) committed by debert   -   Fixed events tags, issue #108 , thanks Blake

  Fixed events tags, issue #108 , thanks Blake
• Oct 30, 2009

  issue 112 (Create svn tags with the stable releases) Status changed by debert   -   Hi Ruy. The idea is good, but in practice, it probably won't happen. The thing is, I am not sure what could constitute a stable version of BulkLoader. The core code doesn't change much (two exceptions[1] during the last few years, which were good candidates for stable releases ;). Also, the fact that I almost never get feedback about bugs introduced between versions make it harder to reason on those terms. I my self, use BulkLoader from trunk, and I advise folks to do the same. Out of the +300 revisions, 3 or 4 had regression bugs on them, and they're fixed pretty fast. Also BulkLoader has a pretty decent test suite at this point (over 230 tests) so large scale breakage is very unlikely. If I ever do any larger scale refactoring or api changes I'll try to keep that in mind though. What I've done so far is that I commit larger changes to trunk, let the folks on the mailing list know and ask them to test it. A few weeks without complaints and I upload a 'stable' download. Regards Arthur [1] one such change was the LI refactor branch, and another was the domain sharding one.

  Status: WontFix

  Hi Ruy. The idea is good, but in practice, it probably won't happen. The thing is, I am not sure what could constitute a stable version of BulkLoader. The core code doesn't change much (two exceptions[1] during the last few years, which were good candidates for stable releases ;). Also, the fact that I almost never get feedback about bugs introduced between versions make it harder to reason on those terms. I my self, use BulkLoader from trunk, and I advise folks to do the same. Out of the +300 revisions, 3 or 4 had regression bugs on them, and they're fixed pretty fast. Also BulkLoader has a pretty decent test suite at this point (over 230 tests) so large scale breakage is very unlikely. If I ever do any larger scale refactoring or api changes I'll try to keep that in mind though. What I've done so far is that I commit larger changes to trunk, let the folks on the mailing list know and ask them to test it. A few weeks without complaints and I upload a 'stable' download. Regards Arthur [1] one such change was the LI refactor branch, and another was the domain sharding one.

  Status: WontFix
• Oct 30, 2009

  issue 113 ([BulkLoader] Cannot use an instance that has been cleared fr...) Status changed by debert   -   Hi Bach44. If an instance has been cleared, nothing will work. It's not a bug really, but a feature. The clear method is a shortcut to remove everything from it, and therefore it must necessarily nullify it. This is documented on the clear doc : http://media.stimuli.com.br/projects/bulk- loader/docs/br/com/stimuli/loading/BulkLoader.html#clear() Cheers

  Status: Invalid

  Hi Bach44. If an instance has been cleared, nothing will work. It's not a bug really, but a feature. The clear method is a shortcut to remove everything from it, and therefore it must necessarily nullify it. This is documented on the clear doc : http://media.stimuli.com.br/projects/bulk- loader/docs/br/com/stimuli/loading/BulkLoader.html#clear() Cheers

  Status: Invalid
• Oct 30, 2009

  issue 113 ([BulkLoader] Cannot use an instance that has been cleared fr...) reported by bach44   -   What steps will reproduce the problem? 1. create a class passing a Bulkloader instance in the constructor 2. trace out the loader to ensure it exists, define it a a private variable (works) 3. later in the same class try tracing out the Bulkloader instance previously defined What is the expected output? What do you see instead? What version of the product are you using? On what operating system? OSX, FP10 Please provide any additional information below. What would cause a Bulkloader instance to be cleared from memory?

  What steps will reproduce the problem? 1. create a class passing a Bulkloader instance in the constructor 2. trace out the loader to ensure it exists, define it a a private variable (works) 3. later in the same class try tracing out the Bulkloader instance previously defined What is the expected output? What do you see instead? What version of the product are you using? On what operating system? OSX, FP10 Please provide any additional information below. What would cause a Bulkloader instance to be cleared from memory?
• Oct 28, 2009

  issue 112 (Create svn tags with the stable releases) reported by ruyadorno   -   I would like to have svn tags for each stable release of the project, it is a common concept as pointed here: http://svnbook.red-bean.com/en/1.1/ch04s06.html It is very useful for those of us using svn:externals properties to link the repository. Thanks,

  I would like to have svn tags for each stable release of the project, it is a common concept as pointed here: http://svnbook.red-bean.com/en/1.1/ch04s06.html It is very useful for those of us using svn:externals properties to link the repository. Thanks,
• Oct 28, 2009

  issue 111 (Problems with asdoc) commented on by ruyadorno   -   Arthur, Os problemas são só umas tags malformatadas, na classe BulkLoader começa na linha 484: * <td>priority</td. e segue nas outras linhas: * <td>maxTries</td. * <td>weight</td. * <td>headers</td. * <td>context</td. * <td>pausedAtStart</td. E na classe BulkProgressEvent a linha 104 ta assim: * @return A code>String</code> will loading information. Também um pequeno probleminha na tag <code>, só isso, mudei aqui e funcionou beleza! valeu! =) ---------- Anyway, I'm using flex4 sdk asdoc and the errors looked like this: Buildfile: build.xml properties: docs: [exec] Chargement du fichier de configuration /Users/ruy/Applications/Adobe Flash Builder Plug-in Beta2/sdks/4.0.0/frameworks/flex-config.xml [exec] Chargement du fichier de configuration /Users/ruy/Documents/workspace/lesite-flash-framework/build/config-doc.xml [exec] [Fatal Error] :20:36: The end-tag for element type "td" must end with a '>' delimiter. [exec] [Fatal Error] :1:22: The element type "test" must be terminated by the matching end-tag "</test>". [exec] Texte incorrectement mis en forme rencontr#. Veuillez consulter /Users/ruy/Documents/workspace/lesite-flash-framework/docs/validation_errors.log pour plus de d#tails. [exec] Source file /Users/ruy/Documents/workspace/lesite-flash-framework/docs/tempdita/packages.dita does not exist [exec] Result: 2 BUILD SUCCESSFUL Total time: 8 seconds

  Arthur, Os problemas são só umas tags malformatadas, na classe BulkLoader começa na linha 484: * <td>priority</td. e segue nas outras linhas: * <td>maxTries</td. * <td>weight</td. * <td>headers</td. * <td>context</td. * <td>pausedAtStart</td. E na classe BulkProgressEvent a linha 104 ta assim: * @return A code>String</code> will loading information. Também um pequeno probleminha na tag <code>, só isso, mudei aqui e funcionou beleza! valeu! =) ---------- Anyway, I'm using flex4 sdk asdoc and the errors looked like this: Buildfile: build.xml properties: docs: [exec] Chargement du fichier de configuration /Users/ruy/Applications/Adobe Flash Builder Plug-in Beta2/sdks/4.0.0/frameworks/flex-config.xml [exec] Chargement du fichier de configuration /Users/ruy/Documents/workspace/lesite-flash-framework/build/config-doc.xml [exec] [Fatal Error] :20:36: The end-tag for element type "td" must end with a '>' delimiter. [exec] [Fatal Error] :1:22: The element type "test" must be terminated by the matching end-tag "</test>". [exec] Texte incorrectement mis en forme rencontr#. Veuillez consulter /Users/ruy/Documents/workspace/lesite-flash-framework/docs/validation_errors.log pour plus de d#tails. [exec] Source file /Users/ruy/Documents/workspace/lesite-flash-framework/docs/tempdita/packages.dita does not exist [exec] Result: 2 BUILD SUCCESSFUL Total time: 8 seconds
• Oct 28, 2009

  issue 111 (Problems with asdoc) commented on by debert   -   Hi Ruy. Thanks for the heads up. I can (and routinely do ) run asdoc on BulkLoader without issues. Rather than just apply the fixes, I'd rather know __what__ went wrong. If you could post full version information for the asdoc tool and the sdk, and also the arguments to the command line, it would help a lot. Thanks Arthur

  Hi Ruy. Thanks for the heads up. I can (and routinely do ) run asdoc on BulkLoader without issues. Rather than just apply the fixes, I'd rather know __what__ went wrong. If you could post full version information for the asdoc tool and the sdk, and also the arguments to the command line, it would help a lot. Thanks Arthur
• Oct 28, 2009

  issue 111 (Problems with asdoc) reported by ruyadorno   -   What steps will reproduce the problem? 1. Try to compile the documentation using asdoc utility 2. The compiler will give fatal erros and generate a log error file 3. I used the head repository version ------------- Here is the log error: Text for param in br.com.stimuli.loading:BulkLoader/add is not valid. org.xml.sax.SAXParseException: The end-tag for element type "td" must end with a '>' delimiter. Text for return in br.com.stimuli.loading:BulkProgressEvent/loadingStatus is not valid. No matching start tag. ------------- The first one seems to be a problem with the columns and arrows in the asdoc of add() method of BulkLoader class, the problem started because we have multiple non-closed <td> tags starting at line 484. The second issue is the <code> tag malformated at line 104 of BulkProgressEvent class ------------- I attached the files with the necessary corrections. Thanks, Ruy

  What steps will reproduce the problem? 1. Try to compile the documentation using asdoc utility 2. The compiler will give fatal erros and generate a log error file 3. I used the head repository version ------------- Here is the log error: Text for param in br.com.stimuli.loading:BulkLoader/add is not valid. org.xml.sax.SAXParseException: The end-tag for element type "td" must end with a '>' delimiter. Text for return in br.com.stimuli.loading:BulkProgressEvent/loadingStatus is not valid. No matching start tag. ------------- The first one seems to be a problem with the columns and arrows in the asdoc of add() method of BulkLoader class, the problem started because we have multiple non-closed <td> tags starting at line 484. The second issue is the <code> tag malformated at line 104 of BulkProgressEvent class ------------- I attached the files with the necessary corrections. Thanks, Ruy
• Oct 28, 2009

  issue 110 (Flex SWC) commented on by bl...@bentbit.com   -   Sounds good. The project is really easy to work with as is.

  Sounds good. The project is really easy to work with as is.
• Oct 28, 2009

  issue 109 (Flex Support) commented on by debert   -   Yes, but we cannot require users to use only one instance of BulkLoader. Either it works with n instances, or it doesn't.

  Yes, but we cannot require users to use only one instance of BulkLoader. Either it works with n instances, or it doesn't.
• Oct 28, 2009

  issue 110 (Flex SWC) Status changed by debert   -   Hi Blake. While I do mostly pure as3 projects, I also compile with the mxmlc, and I don't feel that having a swc will ease integration. I also tend to encourage people to look at the source code and tweak things to their liking. The incremental compilation with fsch means that the overhead of compiling BulkLoader (which is a small lib) is also negligible. Thanks for volunteering, though. Regards, Arthur Debert

  Status: WontFix

  Hi Blake. While I do mostly pure as3 projects, I also compile with the mxmlc, and I don't feel that having a swc will ease integration. I also tend to encourage people to look at the source code and tweak things to their liking. The incremental compilation with fsch means that the overhead of compiling BulkLoader (which is a small lib) is also negligible. Thanks for volunteering, though. Regards, Arthur Debert

  Status: WontFix
• Oct 28, 2009

  issue 109 (Flex Support) commented on by bl...@bentbit.com   -   The solution mentioned above might not be ideal but it does allow a single instance of the control to be created in the flex environment. It additionally allows the user to create component instances with custom names using actionscript. I have been playing with the code in flex and it works great with this method as long as you use a single instance of the component.

  The solution mentioned above might not be ideal but it does allow a single instance of the control to be created in the flex environment. It additionally allows the user to create component instances with custom names using actionscript. I have been playing with the code in flex and it works great with this method as long as you use a single instance of the component.
• Oct 28, 2009

  issue 110 (Flex SWC) reported by bl...@bentbit.com   -   Reccomend adding a compiled SWC download to project for easy integration. I am able to assist in flex environment.

  Reccomend adding a compiled SWC download to project for easy integration. I am able to assist in flex environment.
• Oct 28, 2009

  issue 109 (Flex Support) changed by debert   -   I understand the use case, but this solution won't work. BulkLoader has a registry pattern of sorts where all intances are accessible through a static look up, using the name as a key (BulkLoader.getLoader("photo-gallery"). This makes it a requirement that each instance's name be unique. Therefore, hardcoding a default name will break once you intantiate a second instance. I am in favor of null constructors too (helps a few other pain points as well), so I rather find a solution that answers both needs. Regards

  Status: Accepted

  Labels: Type-Enhancement Priority-Low −Type-Defect −Priority-Medium

  I understand the use case, but this solution won't work. BulkLoader has a registry pattern of sorts where all intances are accessible through a static look up, using the name as a key (BulkLoader.getLoader("photo-gallery"). This makes it a requirement that each instance's name be unique. Therefore, hardcoding a default name will break once you intantiate a second instance. I am in favor of null constructors too (helps a few other pain points as well), so I rather find a solution that answers both needs. Regards

  Status: Accepted

  Labels: Type-Enhancement Priority-Low −Type-Defect −Priority-Medium
• Oct 28, 2009

  issue 109 (Flex Support) reported by bl...@bentbit.com   -   Reccomend making the name attribute on the BulkLoader function in the BulkLoader class an optional attribute. This will allow the class to be created via MXML in Flex. name : String = "oMainBulkLoader" public function BulkLoader(name : String = "oMainBulkLoader", numConnections : int = BulkLoader.DEFAULT_NUM_CONNECTIONS, logLevel : int = BulkLoader.DEFAULT_LOG_LEVEL){

  Reccomend making the name attribute on the BulkLoader function in the BulkLoader class an optional attribute. This will allow the class to be created via MXML in Flex. name : String = "oMainBulkLoader" public function BulkLoader(name : String = "oMainBulkLoader", numConnections : int = BulkLoader.DEFAULT_NUM_CONNECTIONS, logLevel : int = BulkLoader.DEFAULT_LOG_LEVEL){
• Oct 28, 2009

  issue 108 (Missing error event metadata) reported by bl...@bentbit.com   -   The BulkLoader class is missing the error event metadata. [Event(name="error", type="flash.events.ErrorEvent")]

  The BulkLoader class is missing the error event metadata. [Event(name="error", type="flash.events.ErrorEvent")]
• Oct 28, 2009

  issue 106 (getProgressForItems return NaN for weightPercent until every...) commented on by fumanchu12   -   sorry wrong scenario: it happens only on 0-byte file loading and i didn't see that one of my PHP scripts returned nothing it's still a bad idea to divide by something that can be 0 :)

  sorry wrong scenario: it happens only on 0-byte file loading and i didn't see that one of my PHP scripts returned nothing it's still a bad idea to divide by something that can be 0 :)
• Oct 22, 2009

  issue 107 (An error is caught by the unloadAndStop check in ImageItem::...) Status changed by debert   -   Hi John. Thanks a lot for the detailed report. I've committed the fix on revision 307, that should do it. Regards Arthur

  Status: Fixed

  Hi John. Thanks a lot for the detailed report. I've committed the fix on revision 307, that should do it. Regards Arthur

  Status: Fixed
• Oct 22, 2009

  r307 (Fixed bug where swf compiled for player 10 and executed on p...) committed by debert   -   Fixed bug where swf compiled for player 10 and executed on player 9 would crash, issue #107 , thanks John Reevess.

  Fixed bug where swf compiled for player 10 and executed on player 9 would crash, issue #107 , thanks John Reevess.
• Oct 22, 2009

  issue 106 (getProgressForItems return NaN for weightPercent until every...) Status changed by debert   -   Hi Fumanchu. There is definitely something fishy with this, because I cannot reproduce the problem at all here. Can you post the minimal test case (if possible with the online urls for the issue) that will cause it? I can then take a look and see what's wrong. Thanks

  Status: MoreInfoneeded

  Hi Fumanchu. There is definitely something fishy with this, because I cannot reproduce the problem at all here. Can you post the minimal test case (if possible with the online urls for the issue) that will cause it? I can then take a look and see what's wrong. Thanks

  Status: MoreInfoneeded
• Oct 22, 2009

  issue 107 (An error is caught by the unloadAndStop check in ImageItem::...) reported by johnreeves1   -   What steps will reproduce the problem? 1. The application must be compiled with the Flex SDK Pure SDK (the NON FP10 version) and then run in the FP10 runtime. What is the expected output? What do you see instead? An exception is caught internally with the following stacktrace: [BulkLoader] Error while removing item from key:LoadingItem url: ../../../../diablo_static/versioned/flash/assets/world/items/decor/doors/brown_door.swf, type:movieclip, status: finished ReferenceError: Error #1069: Property unloadAndStop not found on flash.display.Loader and there is no default value. at br.com.stimuli.loading.loadingtypes::ImageItem/destroy()[/projects/diablo_as3/workspace/utils_as3/src/br/com/stimuli/loading/loadingtypes/ImageItem.as:134] What version of the product are you using? On what operating system? Bulkloader r.306 Flex SDK 0 Pure (NOT the FP10 version) Flash Player 10 runtime. Please provide any additional information below. The issue can be fixed by changing the destroy() method to include: if (loader.hasOwnProperty("unloadAndStop") && loader["unloadAndStop"] is Function) { loader["unloadAndStop"](true); }

  What steps will reproduce the problem? 1. The application must be compiled with the Flex SDK Pure SDK (the NON FP10 version) and then run in the FP10 runtime. What is the expected output? What do you see instead? An exception is caught internally with the following stacktrace: [BulkLoader] Error while removing item from key:LoadingItem url: ../../../../diablo_static/versioned/flash/assets/world/items/decor/doors/brown_door.swf, type:movieclip, status: finished ReferenceError: Error #1069: Property unloadAndStop not found on flash.display.Loader and there is no default value. at br.com.stimuli.loading.loadingtypes::ImageItem/destroy()[/projects/diablo_as3/workspace/utils_as3/src/br/com/stimuli/loading/loadingtypes/ImageItem.as:134] What version of the product are you using? On what operating system? Bulkloader r.306 Flex SDK 0 Pure (NOT the FP10 version) Flash Player 10 runtime. Please provide any additional information below. The issue can be fixed by changing the destroy() method to include: if (loader.hasOwnProperty("unloadAndStop") && loader["unloadAndStop"] is Function) { loader["unloadAndStop"](true); }
• Oct 21, 2009

  issue 106 (getProgressForItems return NaN for weightPercent until every...) reported by fumanchu12   -   easy fix: localWeightLoaded += (item._bytesLoaded / item._bytesTotal) * item.weight; >>> if (item._bytesTotal) localWeightLoaded += (item._bytesLoaded / item._bytesTotal) * item.weight;

  easy fix: localWeightLoaded += (item._bytesLoaded / item._bytesTotal) * item.weight; >>> if (item._bytesTotal) localWeightLoaded += (item._bytesLoaded / item._bytesTotal) * item.weight;
• Sep 30, 2009

  issue 104 (MP4 Issues) commented on by krislocascio   -   Here is the requested mp4.

  Here is the requested mp4.
• Sep 30, 2009

  issue 105 (MP4 Issues Cont.) Status changed by debert   -   This is a dupe of #104

  Status: Duplicate

  This is a dupe of #104

  Status: Duplicate
• Sep 30, 2009

  issue 104 (MP4 Issues) commented on by krislocascio   -   Sorry, I didnt see that you could add a file in the same issue, i reposted, let me know if you need me to reupload to this string, so you can delete the unnecessary issue.

  Sorry, I didnt see that you could add a file in the same issue, i reposted, let me know if you need me to reupload to this string, so you can delete the unnecessary issue.
• Sep 30, 2009

  issue 105 (MP4 Issues Cont.) reported by krislocascio   -   Thanks for the quick response. I have been able to load the mp4 normally. The file is attached.

  Thanks for the quick response. I have been able to load the mp4 normally. The file is attached.
• Sep 30, 2009

  issue 104 (MP4 Issues) commented on by debert   -   Hi Krislocascio. Can you make sure there is nothing wrong with your mp4 file? e.g can you load it through a naked NetStream? We do have mp4 netstreams no the test suite, and they are running fine. The player is a bit picky on the mp4 enconding, and sometimes it just won't play them. If you're postive it's not your NetStream, can you please attach it to this issue? Regards

  Hi Krislocascio. Can you make sure there is nothing wrong with your mp4 file? e.g can you load it through a naked NetStream? We do have mp4 netstreams no the test suite, and they are running fine. The player is a bit picky on the mp4 enconding, and sometimes it just won't play them. If you're postive it's not your NetStream, can you please attach it to this issue? Regards
• Sep 30, 2009

  issue 104 (MP4 Issues) reported by krislocascio   -   What steps will reproduce the problem? 1. ... loader.add("main.mp4", {"id":"video-main", type:"video"}); 2. ... function onAllItemsLoaded(evt : Event) : void{ var video : Video = new Video(); var theNetStream : NetStream = loader.getNetStream("main.mp4"); addChild(video); video.attachNetStream(theNetStream); theNetStream.resume(); } What is the expected output? What do you see instead? I should see video, the video comes up blank. If I replace the mp4 with an flv it works fine. What version of the product are you using? On what operating system? The newest build, OSX Leopard, CS4 Please provide any additional information below.

  What steps will reproduce the problem? 1. ... loader.add("main.mp4", {"id":"video-main", type:"video"}); 2. ... function onAllItemsLoaded(evt : Event) : void{ var video : Video = new Video(); var theNetStream : NetStream = loader.getNetStream("main.mp4"); addChild(video); video.attachNetStream(theNetStream); theNetStream.resume(); } What is the expected output? What do you see instead? I should see video, the video comes up blank. If I replace the mp4 with an flv it works fine. What version of the product are you using? On what operating system? The newest build, OSX Leopard, CS4 Please provide any additional information below.
• Sep 25, 2009

  r306 (Major code cleanup (indentation, blank lines).) committed by debert   -   Major code cleanup (indentation, blank lines).

  Major code cleanup (indentation, blank lines).
• Sep 24, 2009

  r305 (Indentation fixes to LazyBulkLoader.) committed by debert   -   Indentation fixes to LazyBulkLoader.

  Indentation fixes to LazyBulkLoader.
• Sep 24, 2009

  r304 (Fixed bug where gzip content would create nuts progress repo...) committed by debert   -   Fixed bug where gzip content would create nuts progress reports. This fix is an ugly work around. The player will not give access to the response headers, so we have to guees. When gzip compression is turned on you will end up with changing, unreliable, smaller than bytesLoaded values for bytesTotal. So what we do (not sure if it is **the** best answer here) is to check for moot values and make sure that: - bytesRemaining will always be 0 when loading is done. - bytesTotal will always be equal to bytesLoaded when loading is done. - Give no bizarre percentage midway

  Fixed bug where gzip content would create nuts progress reports. This fix is an ugly work around. The player will not give access to the response headers, so we have to guees. When gzip compression is turned on you will end up with changing, unreliable, smaller than bytesLoaded values for bytesTotal. So what we do (not sure if it is **the** best answer here) is to check for moot values and make sure that: - bytesRemaining will always be 0 when loading is done. - bytesTotal will always be equal to bytesLoaded when loading is done. - Give no bizarre percentage midway
• Sep 18, 2009

  r303 (Exposed the source xml from LazyXMLLoader so that end users ...) committed by debert   -   Exposed the source xml from LazyXMLLoader so that end users can retrieve it.

  Exposed the source xml from LazyXMLLoader so that end users can retrieve it.
• Sep 18, 2009

  r302 (Fixed bug where hasItem would return true even for items not...) committed by debert   -   Fixed bug where hasItem would return true even for items not yet loaded. This is a regression bug, also added tests. Thanks benny (http://groups.google.com/group/bulkloader-users/browse_thread/thread/29f4bdbe1fe03759?hl=en)

  Fixed bug where hasItem would return true even for items not yet loaded. This is a regression bug, also added tests. Thanks benny (http://groups.google.com/group/bulkloader-users/browse_thread/thread/29f4bdbe1fe03759?hl=en)
• Sep 14, 2009

  issue 103 (maxTries = 0 -> retires the asset still 3 times) Status changed by debert   -   Hi Jens. This is updates in revision 301, can you verify this works as expected. As a side note, the issue is that trying to load an asset once should be done using maxTries:1, not 0 (there is no way to load something without trying it at least once), right? Cheers Arthur Debert

  Status: Fixed

  Hi Jens. This is updates in revision 301, can you verify this works as expected. As a side note, the issue is that trying to load an asset once should be done using maxTries:1, not 0 (there is no way to load something without trying it at least once), right? Cheers Arthur Debert

  Status: Fixed
• Sep 14, 2009

  r301 (Maxtries will work even when set to 0, this fixes issues #10...) committed by debert   -   Maxtries will work even when set to 0, this fixes issues #103, thanks Jense Franke

  Maxtries will work even when set to 0, this fixes issues #103, thanks Jense Franke
• Sep 14, 2009

  issue 103 (maxTries = 0 -> retires the asset still 3 times) reported by jensenfranke   -   I only want to try loading an assets one time. If you use bulkInstance.add("/something.php", {maxTries: 0}); the assets will retry 3 times. Maybe the problem is in line 780: if(i.status == LoadingItem.STATUS_ERROR && i.numTries == i.maxTries) Thanks in advance Jens

  I only want to try loading an assets one time. If you use bulkInstance.add("/something.php", {maxTries: 0}); the assets will retry 3 times. Maybe the problem is in line 780: if(i.status == LoadingItem.STATUS_ERROR && i.numTries == i.maxTries) Thanks in advance Jens