Re: [DISCUSS] How do we want to handle Whiteboard?
On Apr 8, 2013, at 10:17 PM, Om wrote: On Mon, Apr 8, 2013 at 9:31 PM, Dave Fisher w...@apache.org wrote: This discussion seems familiar. Justin has the ASF viewpoint for the most part. Talk to infra and David Nalley about the issues in expanding out to github. Regards, Dave Before I bother Infra about this, do you think you can point me to some past discussion about this? The discussion was many months ago on the infra lists. I don't recall if it was public or private. The gist is that this PMC is responsible for keeping the IP in shape and doing the work in the open. The support of GIT has come a long way by Infra. You should ask because there may or may not yet be a document that discusses how far to go. If github is broken then please go to Infra to ask how the project can help. Regards, Dave Thanks, Om Sent from my iPhone On Apr 8, 2013, at 4:35 PM, Om bigosma...@gmail.com wrote: On Mon, Apr 8, 2013 at 4:26 PM, Justin Mclean jus...@classsoftware.com wrote: Hi, Think we may be missing the point here there's nothing to stop a committer from using github for their own experiments now, but if that code is eventually donated into the Flex project it would need to: 1) Any contributors who have a made a large contribution have signed a ICLA and agree to this code being donated. Just having an ICLA signed may not be enough as it may not been clear that the intent was to move the code to Apache Flex. So in git hub are all contributors easily identified and contactable? Notifying the contributer when they send a pull request to a committer's whiteboard project should serve exactly this purpose. 2) Development should be in the open and all current committers notified of progress. At the very least discussion about the code need to take place on the development list (and not elsewhere ie in github) and all changes/diffs to the code need to mailed to commit mailing list. We will experiment with the organization settings to see if all this is doable. Thanks, Om Having the whiteboards in git hub would be no different. Thanks, Justin
[FalconJX][FlexJS] Too self-centered?
Erik, I noticed that the FlexJS output has a lot of methods with “var self=this” in it, and source has a comment implying that you wish it didn’t. FalconJS currently adds “var self=this” only if there are nested functions inside. What is the history behind the “self” in FlexJS output? -- Alex Harui Flex SDK Team Adobe Systems, Inc. http://blogs.adobe.com/aharui
Re: [DISCUSS] How do we want to handle Whiteboard?
1) Any contributors who have a made a large contribution have signed a ICLA and agree to this code being donated. Just having an ICLA signed may not be enough as it may not been clear that the intent was to move the code to Apache Flex. So in git hub are all contributors easily identified and contactable? Before to think about the legal point of re-integrating the contributor work to ApacheFlex, I would like to remind you the technical problem to do it, until we work on Github only, that will be fine but if we want to make fit a one project per repo structure (github) into a multi-project per repo structure (our current per committer repo whiteboard), I can't figure out how to do that. Someone knows ? -Fred -Message d'origine- From: Om Sent: Tuesday, April 09, 2013 1:35 AM To: dev@flex.apache.org Subject: Re: [DISCUSS] How do we want to handle Whiteboard? On Mon, Apr 8, 2013 at 4:26 PM, Justin Mclean jus...@classsoftware.comwrote: Hi, Think we may be missing the point here there's nothing to stop a committer from using github for their own experiments now, but if that code is eventually donated into the Flex project it would need to: 1) Any contributors who have a made a large contribution have signed a ICLA and agree to this code being donated. Just having an ICLA signed may not be enough as it may not been clear that the intent was to move the code to Apache Flex. So in git hub are all contributors easily identified and contactable? Notifying the contributer when they send a pull request to a committer's whiteboard project should serve exactly this purpose. 2) Development should be in the open and all current committers notified of progress. At the very least discussion about the code need to take place on the development list (and not elsewhere ie in github) and all changes/diffs to the code need to mailed to commit mailing list. We will experiment with the organization settings to see if all this is doable. Thanks, Om Having the whiteboards in git hub would be no different. Thanks, Justin
Re: [DISCUSS] How do we want to handle Whiteboard?
Hi, I would like to remind you the technical problem to do it If the whiteboard is in SVN, Git or Github we have the same issue - the white board area doesn't have to mirror the current SDK structure. Answer is basically patch files in JIRA or branch existing SDK copy files over and see if you can merge and commit/push. The main difference is it easier to check out part of a repo in SVN so with Git or Github we may want a repo/committer rather than a single repo. Thanks, Justin
Re: [DISCUSS] How do we want to handle Whiteboard?
copy files over and see if you can merge and commit/push What would mean deleting the current tree and copy the new one over and not only copy over the existing tree as it would be a problem if parts of the tree have been renamed or moved, right ? The main difference is it easier to check out part of a repo in SVN so with Git or Github we may want a repo/committer rather than a single repo. It's not really the github way and would mean that a github contributor would be able to send pull requests touching files in a project is not initially contributing and has not the agreement of the committer, so to make it work, it would imply more checks for the committer reviewing the pull request before to integrate it but it should be possible I think. -Fred -Message d'origine- From: Justin Mclean Sent: Tuesday, April 09, 2013 8:45 AM To: dev@flex.apache.org Subject: Re: [DISCUSS] How do we want to handle Whiteboard? Hi, I would like to remind you the technical problem to do it If the whiteboard is in SVN, Git or Github we have the same issue - the white board area doesn't have to mirror the current SDK structure. Answer is basically patch files in JIRA or branch existing SDK copy files over and see if you can merge and commit/push. The main difference is it easier to check out part of a repo in SVN so with Git or Github we may want a repo/committer rather than a single repo. Thanks, Justin
Re: [FalconJX][FlexJS] Too self-centered?
The self = this needs to be more clever, you are correct. At one point in my race to understand all things ASJS it was probably convenient to add this to each function body and use self instead of this. JS does require a lot more this than AS, so adding one reference or another didn't seem like a big deal, especially since self is more certain. Then again, since using 'goog.bind' for this references in nested functions is a good idea, it is certainly worth another look. Added to my list :-) EdB On Tue, Apr 9, 2013 at 1:25 AM, Alex Harui aha...@adobe.com wrote: Erik, I noticed that the FlexJS output has a lot of methods with “var self=this” in it, and source has a comment implying that you wish it didn’t. FalconJS currently adds “var self=this” only if there are nested functions inside. What is the history behind the “self” in FlexJS output? -- Alex Harui Flex SDK Team Adobe Systems, Inc. http://blogs.adobe.com/aharui -- Ix Multimedia Software Jan Luykenstraat 27 3521 VB Utrecht T. 06-51952295 I. www.ixsoftware.nl
Re: [FalconJX][FlexJS] Dependency calculations
Alex, The current 'publisher' part of FalconJx uses the vanilla 'DepsGenerator' functionality of the Closure Compiler. Do you have a minimal test case of the kind of circular dependency you are talking about? It seems like that should be something that Google should have encountered and be able to handle. I'm willing to look into it, but I haven't come across it myself, so a test case might come in handy. EdB On Tue, Apr 9, 2013 at 1:36 AM, Alex Harui aha...@adobe.com wrote: Erik, Have you noticed that the google tools seem to be too sensitive to circular dependencies in the goog.requires? Back when I was using FalconJS, I was trying to use the google tools to compute the deps.js files and it kept failing with circular dependencies. I ended up writing my own deps calculator that kept track of which classes had already been seen and things seemed to work. In my “customer’s” test case, FalconJX is now returning a circular dependency error. I haven’t looked into the FalconJX deps logic, but are you also using some google code for that? Is there some other mechanism other than goog.requires() for dependencies that aren’t involved in the prototype definition? Thanks, -- Alex Harui Flex SDK Team Adobe Systems, Inc. http://blogs.adobe.com/aharui -- Ix Multimedia Software Jan Luykenstraat 27 3521 VB Utrecht T. 06-51952295 I. www.ixsoftware.nl
Re: [FalconJX][FlexJS] imports != require?
None of my test cases (and none of my real world code, I think they're lazy) uses wildcard imports, so I wasn't aware of this. But since they are a feature of AS, we need to be able to handle them. The compiler seems to be the proper handler for this, so I'll have a look at how FalconJS does this and see how we might re-use your solution. Added to the list. EdB On Mon, Apr 8, 2013 at 3:31 PM, Alex Harui aha...@adobe.com wrote: Erik, I haven’t found the code yet, but the FlexJS output seems to be using import statements as a way of setting up the goog.requires. The closure compiler subsequently doesn’t like wild card imports import foo.bar.*; And I think it will miss same-package class dependencies, unless that is being picked up somewhere else (and of course, will report unnecessary requires if you haven’t cleaned up your imports. FalconJS built up a list of dependencies as they were discovered by the compiler. This seems to be sufficient for the two test cases I’m working with. Any objections to switching away from imports to something more like how FalconJS does it? -- Alex Harui Flex SDK Team Adobe Systems, Inc. http://blogs.adobe.com/aharui -- Ix Multimedia Software Jan Luykenstraat 27 3521 VB Utrecht T. 06-51952295 I. www.ixsoftware.nl
Re: [DISCUSS] How do we want to handle Whiteboard?
On Tue, Apr 9, 2013 at 8:23 AM, Dave Fisher dave2w...@comcast.net wrote: ...The gist is that this PMC is responsible for keeping the IP in shape and doing the work in the open I'd also add that whoever commits (or pushes, in the Git model) to the Apache repository (where all releases must be cut from) takes responsibility for what they commit, under the iCLA that they signed. So it's ultimately up to that committer to make sure they have the right to commit what they're committing, and that the provenance of every line of code can be clearly established. Backing a contribution with an https://issues.apache.org issue where the contributor clearly expresses their intention to donate the code to Apache Flex is helpful, and having an iCLA for that contributor helps put the burden on them for checking that it's their own code that they are contributing. -Bertrand
[jira] [Commented] (FLEX-33480) Build error when locale is set to pt_PT
[ https://issues.apache.org/jira/browse/FLEX-33480?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13626369#comment-13626369 ] Justin Mclean commented on FLEX-33480: -- Confirmed and it also happens for some other locales as well but only if a s:TextArea is used, if a s:Label or other component is used it doesn't cause the issue. pt_PT resource bundles for textLayput do exist in the 4.9.1 SDK so might be a Flash Builder issues. Build error when locale is set to pt_PT --- Key: FLEX-33480 URL: https://issues.apache.org/jira/browse/FLEX-33480 Project: Apache Flex Issue Type: Bug Components: Spark: TextArea Affects Versions: Apache Flex 4.9.0 Environment: Tested using 4.9.1 Reporter: Dasa Paddock Locale is set to: pt_PT Test code is: ?xml version=1.0 encoding=utf-8? s:Application xmlns:fx=http://ns.adobe.com/mxml/2009; xmlns:s=library://ns.adobe.com/flex/spark s:TextArea/ /s:Application Build error is: Unable to resolve resource bundle textLayout for locale pt_PT. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Comment Edited] (FLEX-33480) Build error when locale is set to pt_PT
[ https://issues.apache.org/jira/browse/FLEX-33480?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13626369#comment-13626369 ] Justin Mclean edited comment on FLEX-33480 at 4/9/13 8:07 AM: -- Confirmed and it also happens for some other locales as well but only if a s:TextArea is used, if a s:Label or other component is used it doesn't cause the issue. pt_PT resource bundles for textLayput do exist in the 4.9.1 SDK so might be a Flash Builder issue. was (Author: jmclean): Confirmed and it also happens for some other locales as well but only if a s:TextArea is used, if a s:Label or other component is used it doesn't cause the issue. pt_PT resource bundles for textLayput do exist in the 4.9.1 SDK so might be a Flash Builder issues. Build error when locale is set to pt_PT --- Key: FLEX-33480 URL: https://issues.apache.org/jira/browse/FLEX-33480 Project: Apache Flex Issue Type: Bug Components: Spark: TextArea Affects Versions: Apache Flex 4.9.0 Environment: Tested using 4.9.1 Reporter: Dasa Paddock Locale is set to: pt_PT Test code is: ?xml version=1.0 encoding=utf-8? s:Application xmlns:fx=http://ns.adobe.com/mxml/2009; xmlns:s=library://ns.adobe.com/flex/spark s:TextArea/ /s:Application Build error is: Unable to resolve resource bundle textLayout for locale pt_PT. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Comment Edited] (FLEX-33480) Build error when locale is set to pt_PT
[ https://issues.apache.org/jira/browse/FLEX-33480?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13626378#comment-13626378 ] Justin Mclean edited comment on FLEX-33480 at 4/9/13 8:19 AM: -- Looks like issue is that pt_PT, en_GB, en_CA, en_AU and el_GR bundles are commented out in textLayout's build.xml. Build.xml has same issue in Apache Flex develop branch. was (Author: jmclean): Looks like issue is that pt_PT, en_GB, en_CA, en_AU and el_GR bundles are commented out in textLayout's build.xml. Build.xml has same issue in Apache Flex 4.10. Build error when locale is set to pt_PT --- Key: FLEX-33480 URL: https://issues.apache.org/jira/browse/FLEX-33480 Project: Apache Flex Issue Type: Bug Components: Spark: TextArea Affects Versions: Apache Flex 4.9.0 Environment: Tested using 4.9.1 Reporter: Dasa Paddock Locale is set to: pt_PT Test code is: ?xml version=1.0 encoding=utf-8? s:Application xmlns:fx=http://ns.adobe.com/mxml/2009; xmlns:s=library://ns.adobe.com/flex/spark s:TextArea/ /s:Application Build error is: Unable to resolve resource bundle textLayout for locale pt_PT. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Assigned] (FLEX-33480) Build error when locale is set to pt_PT
[ https://issues.apache.org/jira/browse/FLEX-33480?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Justin Mclean reassigned FLEX-33480: Assignee: Justin Mclean Build error when locale is set to pt_PT --- Key: FLEX-33480 URL: https://issues.apache.org/jira/browse/FLEX-33480 Project: Apache Flex Issue Type: Bug Components: Spark: TextArea Affects Versions: Apache Flex 4.9.0 Environment: Tested using 4.9.1 Reporter: Dasa Paddock Assignee: Justin Mclean Locale is set to: pt_PT Test code is: ?xml version=1.0 encoding=utf-8? s:Application xmlns:fx=http://ns.adobe.com/mxml/2009; xmlns:s=library://ns.adobe.com/flex/spark s:TextArea/ /s:Application Build error is: Unable to resolve resource bundle textLayout for locale pt_PT. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (FLEX-33480) Build error when locale is set to pt_PT
[ https://issues.apache.org/jira/browse/FLEX-33480?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13626378#comment-13626378 ] Justin Mclean commented on FLEX-33480: -- Looks like issue is that pt_PT, en_GB, en_CA, en_AU and el_GR bundles are commented out in textLayout's build.xml. Build.xml has same issue in Apache Flex 4.10. Build error when locale is set to pt_PT --- Key: FLEX-33480 URL: https://issues.apache.org/jira/browse/FLEX-33480 Project: Apache Flex Issue Type: Bug Components: Spark: TextArea Affects Versions: Apache Flex 4.9.0 Environment: Tested using 4.9.1 Reporter: Dasa Paddock Locale is set to: pt_PT Test code is: ?xml version=1.0 encoding=utf-8? s:Application xmlns:fx=http://ns.adobe.com/mxml/2009; xmlns:s=library://ns.adobe.com/flex/spark s:TextArea/ /s:Application Build error is: Unable to resolve resource bundle textLayout for locale pt_PT. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Updated] (FLEX-33480) Build error when locale is set to pt_PT
[ https://issues.apache.org/jira/browse/FLEX-33480?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Justin Mclean updated FLEX-33480: - Attachment: textLayout_rb.swc pt_PT textLayout compiled resources Build error when locale is set to pt_PT --- Key: FLEX-33480 URL: https://issues.apache.org/jira/browse/FLEX-33480 Project: Apache Flex Issue Type: Bug Components: Spark: TextArea Affects Versions: Apache Flex 4.9.0 Environment: Tested using 4.9.1 Reporter: Dasa Paddock Assignee: Justin Mclean Attachments: textLayout_rb.swc Locale is set to: pt_PT Test code is: ?xml version=1.0 encoding=utf-8? s:Application xmlns:fx=http://ns.adobe.com/mxml/2009; xmlns:s=library://ns.adobe.com/flex/spark s:TextArea/ /s:Application Build error is: Unable to resolve resource bundle textLayout for locale pt_PT. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (FLEX-33480) Build error when locale is set to pt_PT
[ https://issues.apache.org/jira/browse/FLEX-33480?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13626407#comment-13626407 ] Justin Mclean commented on FLEX-33480: -- Fixed in Apache next. Attached is pt_PT textLayout_rb.swc copy to frameworks/locale/pt_PT/ in your 4.9.1 SDK to fix issue. Build error when locale is set to pt_PT --- Key: FLEX-33480 URL: https://issues.apache.org/jira/browse/FLEX-33480 Project: Apache Flex Issue Type: Bug Components: Spark: TextArea Affects Versions: Apache Flex 4.9.0 Environment: Tested using 4.9.1 Reporter: Dasa Paddock Assignee: Justin Mclean Attachments: textLayout_rb.swc Locale is set to: pt_PT Test code is: ?xml version=1.0 encoding=utf-8? s:Application xmlns:fx=http://ns.adobe.com/mxml/2009; xmlns:s=library://ns.adobe.com/flex/spark s:TextArea/ /s:Application Build error is: Unable to resolve resource bundle textLayout for locale pt_PT. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Resolved] (FLEX-33480) Build error when locale is set to pt_PT
[ https://issues.apache.org/jira/browse/FLEX-33480?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Justin Mclean resolved FLEX-33480. -- Resolution: Fixed Fix Version/s: Apache Flex 4.10.0 Build error when locale is set to pt_PT --- Key: FLEX-33480 URL: https://issues.apache.org/jira/browse/FLEX-33480 Project: Apache Flex Issue Type: Bug Components: Spark: TextArea Affects Versions: Apache Flex 4.9.0 Environment: Tested using 4.9.1 Reporter: Dasa Paddock Assignee: Justin Mclean Fix For: Apache Flex 4.10.0 Attachments: textLayout_rb.swc Locale is set to: pt_PT Test code is: ?xml version=1.0 encoding=utf-8? s:Application xmlns:fx=http://ns.adobe.com/mxml/2009; xmlns:s=library://ns.adobe.com/flex/spark s:TextArea/ /s:Application Build error is: Unable to resolve resource bundle textLayout for locale pt_PT. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
RE: FalconJX: bugs in external dependencies (Google Closure Compiler), what to do?
Unfortunately I could not find a workaround to this. Played with the paths trying to fool DepsGenerator but it didn't work. The other approach is to copy the DepsGenerator and PathUtil classes to FalconJX code base and fix it there. Technically this approach works, I tried it. The fix to PathUtil is trivial to do, only a few lines must be modified to make it handle Windows paths correctly. I did not need to touch DepsGenerator except to change the import to use 'our' fixed PathUtil. After the fix is applied compiling FlexJSTest_again works and produces correct deps.js on Windows and the example then runs fine in the browser. However I do not know what are full implications and if you practice such approach anyway. Looks like OK to me from legal point of view since the Closure Compiler is Apache licensed. Please let me know if what I am doing makes any sense or I should scrap what I did so far and wait for someone smarter than me to work on this. Is there any ToDo list for FalconJX so that I can go ahead and try to work on some outstanding issues? I now have Falcon projects in my Eclipse, git is configured and am now ready for a bit more serious work. Tigran. -Original Message- From: Erik de Bruin [mailto:e...@ixsoftware.nl] Sent: Monday, April 08, 2013 4:37 PM To: dev@flex.apache.org Subject: Re: FalconJX: bugs in external dependencies (Google Closure Compiler), what to do? Tigran, Both Alex and I are on a Mac, so that is probably why it escaped notice until now :-) Also, the publisher side of the code is very much a work in progress, which deserves all the attention you can give it. I'll answer your other email some time tonight, jetlag permitting. On this one: might manipulation the path(s) before they are fed into the 'depswriter' - working from memory, might be missing the point - help avoid this low level problem? EdB On Mon, Apr 8, 2013 at 6:27 AM, Tigran Najaryan tig...@gmail.com wrote: I debugged the problem that I reported earlier (generating incorrect paths in deps.js) and looks like it is due to a bug in com.google.javascript.jscomp.deps.PathUtil.isAbsolute. It is implemented by simply checking that the path starts by a forward slash which of course is incorrect under Windows. So what is the procedure when the bug is in an external dependency? It is not as if it is a single call that can be avoided. PathUtil.isAbsolute is called a few levels below in the call stack from the com.google.javascript.jscomp.deps.DepsGenerator.computeDependencyCalls , which is supposed to compute and return the correct text for dependencies file. I will debug this further but just wanted to understand if I should I try to work around the external bug when I meet one or I need to try something else? Tigran. P.S. Did anyone actually try running FalconJX on Windows or everybody is developing on a Mac? -- Ix Multimedia Software Jan Luykenstraat 27 3521 VB Utrecht T. 06-51952295 I. www.ixsoftware.nl
Re: FalconJX: bugs in external dependencies (Google Closure Compiler), what to do?
The other approach is to copy the DepsGenerator and PathUtil classes to FalconJX code base and fix it there. Technically this approach works, I tried it. The fix to PathUtil is trivial to do, only a few lines must be modified to make it handle Windows paths correctly. I did not need to touch DepsGenerator except to change the import to use 'our' fixed PathUtil. After the fix is applied compiling FlexJSTest_again works and produces correct deps.js on Windows and the example then runs fine in the browser. Forking the Google code to our project doesn't seem like the best option. We may consider donating your solution to the Closure project, which is Open Source... Please let me know if what I am doing makes any sense or I should scrap what I did so far and wait for someone smarter than me to work on this. Keep what you have, please, but don't push it to the 'develop' branch. You're probably the smartest person that will work on it, but we need to take care that we don't end up with a dependency on a fork of the closure tools where we actually want to be able to use their releases. Is there any ToDo list for FalconJX so that I can go ahead and try to work on some outstanding issues? I now have Falcon projects in my Eclipse, git is configured and am now ready for a bit more serious work. For FalconJX the main goal is to write tests for both MXML and AS parsing. At the moment there are some areas that are written specifically to allow for the correct parsing of the FlexJSTest_again example. This obviously is not ideal ;-) We need to make sure that all code is generic (or as much as possible) and that all output is properly tested so we don't kill any existing output when refactoring or adding code. Right now, there is only 2 tests specific for FlexJS MXML output, and those are @Ignored. But first, just look around, review what's there and see if what I/we did makes sense. EdB -- Ix Multimedia Software Jan Luykenstraat 27 3521 VB Utrecht T. 06-51952295 I. www.ixsoftware.nl
Re: FalconJX: bugs in external dependencies (Google Closure Compiler), what to do?
Ah, and a good way to start would be to create a JIRA ticket about the need for a better README and to attach a patch for that issue with the content of what you wrote in your original email. That should get you (and us) going on the road of coding and patching and will go some way towards getting you to committer-ship ;-) EdB On Tue, Apr 9, 2013 at 6:46 AM, Erik de Bruin e...@ixsoftware.nl wrote: The other approach is to copy the DepsGenerator and PathUtil classes to FalconJX code base and fix it there. Technically this approach works, I tried it. The fix to PathUtil is trivial to do, only a few lines must be modified to make it handle Windows paths correctly. I did not need to touch DepsGenerator except to change the import to use 'our' fixed PathUtil. After the fix is applied compiling FlexJSTest_again works and produces correct deps.js on Windows and the example then runs fine in the browser. Forking the Google code to our project doesn't seem like the best option. We may consider donating your solution to the Closure project, which is Open Source... Please let me know if what I am doing makes any sense or I should scrap what I did so far and wait for someone smarter than me to work on this. Keep what you have, please, but don't push it to the 'develop' branch. You're probably the smartest person that will work on it, but we need to take care that we don't end up with a dependency on a fork of the closure tools where we actually want to be able to use their releases. Is there any ToDo list for FalconJX so that I can go ahead and try to work on some outstanding issues? I now have Falcon projects in my Eclipse, git is configured and am now ready for a bit more serious work. For FalconJX the main goal is to write tests for both MXML and AS parsing. At the moment there are some areas that are written specifically to allow for the correct parsing of the FlexJSTest_again example. This obviously is not ideal ;-) We need to make sure that all code is generic (or as much as possible) and that all output is properly tested so we don't kill any existing output when refactoring or adding code. Right now, there is only 2 tests specific for FlexJS MXML output, and those are @Ignored. But first, just look around, review what's there and see if what I/we did makes sense. EdB -- Ix Multimedia Software Jan Luykenstraat 27 3521 VB Utrecht T. 06-51952295 I. www.ixsoftware.nl -- Ix Multimedia Software Jan Luykenstraat 27 3521 VB Utrecht T. 06-51952295 I. www.ixsoftware.nl
Re: Falcon tests fail: SecurityError: Error #2017: Only trusted local files may cause the Flash Player to exit.
Ah, you are correct... This is part of the README for the SDK, but isn't part of the README for Falcon(JS/JX). Since most developers will have built the SDK before getting to the compiler(s), we got away with this ;-) Another addition to the README, via a JIRA ticket? EdB On Tue, Apr 9, 2013 at 6:49 AM, Tigran Najaryan tig...@gmail.com wrote: Not sure where do I report this. If you get a Flash Player error when trying to run the compiler tests make sure to add full path to directory falcon/compiler.tests to your Flash Player trust file (for location see http://livedocs.adobe.com/flex/3/html/help.html?content=security2_25.html) Without trusting the directory the mxml.tags.MXMLArrayTagTests in feature.tests opens Flash Player and then fails with popup error message since it is not trusted to execute the exit() call. Adding directory falcon/compiler.tests to Flash Player Trust files fixes the problem. -- Ix Multimedia Software Jan Luykenstraat 27 3521 VB Utrecht T. 06-51952295 I. www.ixsoftware.nl
RE: FalconJX: bugs in external dependencies (Google Closure Compiler), what to do?
Forking the Google code to our project doesn't seem like the best option. Yes, that what concerns me too. We may consider donating your solution to the Closure project, which is Open Source... That may be a solution in the long term but I am not sure that helps to fix the bug in Falcon compiler right now. Anyway submitting a patch to Closure will require more work on my part. Although the bug fix I propose is generic for PathUtil I still did not examine all the dependencies from that class in the Closure itself and am not sure my patch is generic enough and correct for the Closure itself. Honestly I would rather not digress and stay concentrated on contributing to Flex. Keep what you have, please, but don't push it to the 'develop' branch. You're probably the smartest person that will work on it, but we need to take care that we don't end up with a dependency on a fork of the closure tools where we actually want to be able to use their releases. I guess that means I will keep my patch for myself only until either somebody else needs it or a better way to fix this bug is found. For FalconJX the main goal is to write tests for both MXML and AS parsing. At the moment there are some areas that are written specifically to allow for the correct parsing of the FlexJSTest_again example. This obviously is not ideal ;-) We need to make sure that all code is generic (or as much as possible) and that all output is properly tested so we don't kill any existing output when refactoring or adding code. Right now, there is only 2 tests specific for FlexJS MXML output, and those are @Ignored. But first, just look around, review what's there and see if what I/we did makes sense. OK, I will spend a few more days just getting familiar with the code base and see if I can find and fix some other bugs before trying to make functional changes to the compiler. Tigran.
Re: FalconJX: bugs in external dependencies (Google Closure Compiler), what to do?
Also, the current FlexJS implementation for both MXML and AS contains several giant methods that simply beg to be broken up into more manageable sets of support functions which maybe even share some common code (like the emission of JSDoc type headers)... That might also be a good way to get your feet wet without needing to change any of the APIs. EdB On Tue, Apr 9, 2013 at 7:11 AM, Tigran Najaryan tig...@gmail.com wrote: Forking the Google code to our project doesn't seem like the best option. Yes, that what concerns me too. We may consider donating your solution to the Closure project, which is Open Source... That may be a solution in the long term but I am not sure that helps to fix the bug in Falcon compiler right now. Anyway submitting a patch to Closure will require more work on my part. Although the bug fix I propose is generic for PathUtil I still did not examine all the dependencies from that class in the Closure itself and am not sure my patch is generic enough and correct for the Closure itself. Honestly I would rather not digress and stay concentrated on contributing to Flex. Keep what you have, please, but don't push it to the 'develop' branch. You're probably the smartest person that will work on it, but we need to take care that we don't end up with a dependency on a fork of the closure tools where we actually want to be able to use their releases. I guess that means I will keep my patch for myself only until either somebody else needs it or a better way to fix this bug is found. For FalconJX the main goal is to write tests for both MXML and AS parsing. At the moment there are some areas that are written specifically to allow for the correct parsing of the FlexJSTest_again example. This obviously is not ideal ;-) We need to make sure that all code is generic (or as much as possible) and that all output is properly tested so we don't kill any existing output when refactoring or adding code. Right now, there is only 2 tests specific for FlexJS MXML output, and those are @Ignored. But first, just look around, review what's there and see if what I/we did makes sense. OK, I will spend a few more days just getting familiar with the code base and see if I can find and fix some other bugs before trying to make functional changes to the compiler. Tigran. -- Ix Multimedia Software Jan Luykenstraat 27 3521 VB Utrecht T. 06-51952295 I. www.ixsoftware.nl
[jira] [Created] (FLEX-33481) Improve README for FalconJX compiler
Tigran Najaryan created FLEX-33481: -- Summary: Improve README for FalconJX compiler Key: FLEX-33481 URL: https://issues.apache.org/jira/browse/FLEX-33481 Project: Apache Flex Issue Type: Documentation Components: .Unspecified - Compiler Reporter: Tigran Najaryan Priority: Minor The current README file is sparse and needs more details. Here is proposed README file content: Apache FalconJx Prototype Cross Compiler Note: steps 1-4 and 7 below describe how to get and build Apache Flex SDK, develop branch. If you already have the the develop branch of SDK skip these steps. To get and build FalconJX compiler follow these steps: 1. Create a working directory for Apache Flex (repo below). 2. cd repo 3. Get bleading-edge of SDK: git clone https://git-wip-us.apache.org/repos/asf/flex-sdk.git sdk cd sdk git checkout develop cd .. 4. Get TLF git clone https://git-wip-us.apache.org/repos/asf/flex-tlf.git tlf 5. Get Falcon and switch to develop branch: git clone https://git-wip-us.apache.org/repos/asf/flex-falcon.git falcon cd falcon git checkout develop cd .. 6. Get ASJS git clone https://git-wip-us.apache.org/repos/asf/flex-asjs.git asjs cd asjs git checkout develop cd .. 7. Build SDK. Follow instructions in sdk/README, then: cd sdk ant main (Select all default answers when prompted, i.e. just hit ENTER until the build finishes). cd .. 8. Build Falcon compiler. Follow instructions in falcon/compiler/README, then: cd falcon/compiler ant -Dsdk.branch=path-to-repo/sdk cd ../.. (Note: the compiler/build.xml assumes Flex SDK is in ../flex-sdk dir, which is not what we have created in step 3 above, so you need to supply correct dir to Ant via sdk.branch parameter). 9. Build FalconJX. cd falcon/compiler.jx ant cd ../.. 10. Build ASJS and FlexJSUI.swc: cd asjs ant cd frameworks/as ant -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
RE: Falcon tests fail: SecurityError: Error #2017: Only trusted local files may cause the Flash Player to exit.
Ah, you are correct... This is part of the README for the SDK, but isn't part of the README for Falcon(JS/JX). Since most developers will have built the SDK before getting to the compiler(s), we got away with this ;-) My mistake. Since it is in the README for the SDK and it is a pre-requisite for Falcon I think that is sufficient.
Re: Git commit emails
I have the feeling I read an update on this somewhere in one of the INFRA messages, but I can't find it... Anyway, the git commit emails are still not formatted the way we discussed. EdB On Wed, Mar 27, 2013 at 12:44 PM, Om bigosma...@gmail.com wrote: I made this request to INFRA a while ago. I will ping them again now. (I dont think we can do this ourselves, but if someone knows how, please feel free to) Thanks, Om On Wed, Mar 27, 2013 at 10:43 AM, Erik de Bruin e...@ixsoftware.nl wrote: Is this something we can implement ourselves, or must this go through INFRA? EdB On Mon, Mar 18, 2013 at 6:45 PM, Om bigosma...@gmail.com wrote: On Mon, Mar 18, 2013 at 10:20 AM, Gordon Smith gosm...@adobe.com wrote: The email subject is currently something like git commit: The files/directories to exclude while checkin. Since we have 4 different repos, can we customize this to show which repo the commit was for? (Why? Because I want to pay more attention to commits to flex-falcon than to flex-sdk.) I suggest something like git commit [flex-sdk] The files/directories to exclude while checkin which makes it easy for the eye to filter on the repo. - Gordon There is documentation on how this works: https://git-wip-us.apache.org/docs/switching-to-git.html#commit-email-subject-format I guess the format you want for Falcon would be something like: git commit [%(repo_name)s]: %(subject)s If folks want to come up with a commit email subject format for each repo, I can give that to Infra and they set it up for us. If everyone likes this format, we can keep this format for all the repos. Thanks, Om -- Ix Multimedia Software Jan Luykenstraat 27 3521 VB Utrecht T. 06-51952295 I. www.ixsoftware.nl -- Ix Multimedia Software Jan Luykenstraat 27 3521 VB Utrecht T. 06-51952295 I. www.ixsoftware.nl
[jira] [Updated] (FLEX-33481) Improve README for FalconJX compiler
[ https://issues.apache.org/jira/browse/FLEX-33481?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Erik de Bruin updated FLEX-33481: - Assignee: Erik de Bruin Improve README for FalconJX compiler Key: FLEX-33481 URL: https://issues.apache.org/jira/browse/FLEX-33481 Project: Apache Flex Issue Type: Documentation Components: .Unspecified - Compiler Reporter: Tigran Najaryan Assignee: Erik de Bruin Priority: Minor The current README file is sparse and needs more details. Here is proposed README file content: Apache FalconJx Prototype Cross Compiler Note: steps 1-4 and 7 below describe how to get and build Apache Flex SDK, develop branch. If you already have the the develop branch of SDK skip these steps. To get and build FalconJX compiler follow these steps: 1. Create a working directory for Apache Flex (repo below). 2. cd repo 3. Get bleading-edge of SDK: git clone https://git-wip-us.apache.org/repos/asf/flex-sdk.git sdk cd sdk git checkout develop cd .. 4. Get TLF git clone https://git-wip-us.apache.org/repos/asf/flex-tlf.git tlf 5. Get Falcon and switch to develop branch: git clone https://git-wip-us.apache.org/repos/asf/flex-falcon.git falcon cd falcon git checkout develop cd .. 6. Get ASJS git clone https://git-wip-us.apache.org/repos/asf/flex-asjs.git asjs cd asjs git checkout develop cd .. 7. Build SDK. Follow instructions in sdk/README, then: cd sdk ant main (Select all default answers when prompted, i.e. just hit ENTER until the build finishes). cd .. 8. Build Falcon compiler. Follow instructions in falcon/compiler/README, then: cd falcon/compiler ant -Dsdk.branch=path-to-repo/sdk cd ../.. (Note: the compiler/build.xml assumes Flex SDK is in ../flex-sdk dir, which is not what we have created in step 3 above, so you need to supply correct dir to Ant via sdk.branch parameter). 9. Build FalconJX. cd falcon/compiler.jx ant cd ../.. 10. Build ASJS and FlexJSUI.swc: cd asjs ant cd frameworks/as ant -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
Re: [FalconJX][FlexJS] imports != require?
I think I have copied the Falconjs implementation to falconjx. Sent via the PANTECH Discover, an ATT 4G LTE smartphone. Erik de Bruin e...@ixsoftware.nl wrote: None of my test cases (and none of my real world code, I think they're lazy) uses wildcard imports, so I wasn't aware of this. But since they are a feature of AS, we need to be able to handle them. The compiler seems to be the proper handler for this, so I'll have a look at how FalconJS does this and see how we might re-use your solution. Added to the list. EdB On Mon, Apr 8, 2013 at 3:31 PM, Alex Harui aha...@adobe.com wrote: Erik, I haven’t found the code yet, but the FlexJS output seems to be using import statements as a way of setting up the goog.requires. The closure compiler subsequently doesn’t like wild card imports import foo.bar.*; And I think it will miss same-package class dependencies, unless that is being picked up somewhere else (and of course, will report unnecessary requires if you haven’t cleaned up your imports. FalconJS built up a list of dependencies as they were discovered by the compiler. This seems to be sufficient for the two test cases I’m working with. Any objections to switching away from imports to something more like how FalconJS does it? -- Alex Harui Flex SDK Team Adobe Systems, Inc. http://blogs.adobe.com/aharui -- Ix Multimedia Software Jan Luykenstraat 27 3521 VB Utrecht T. 06-51952295 I. www.ixsoftware.nl
Re: [FalconJX][FlexJS] imports != require?
Locally? Or did I miss a commit? EdB On Tue, Apr 9, 2013 at 8:10 AM, Alex Harui aha...@adobe.com wrote: I think I have copied the Falconjs implementation to falconjx. Sent via the PANTECH Discover, an ATT 4G LTE smartphone. Erik de Bruin e...@ixsoftware.nl wrote: None of my test cases (and none of my real world code, I think they're lazy) uses wildcard imports, so I wasn't aware of this. But since they are a feature of AS, we need to be able to handle them. The compiler seems to be the proper handler for this, so I'll have a look at how FalconJS does this and see how we might re-use your solution. Added to the list. EdB On Mon, Apr 8, 2013 at 3:31 PM, Alex Harui aha...@adobe.com wrote: Erik, I haven’t found the code yet, but the FlexJS output seems to be using import statements as a way of setting up the goog.requires. The closure compiler subsequently doesn’t like wild card imports import foo.bar.*; And I think it will miss same-package class dependencies, unless that is being picked up somewhere else (and of course, will report unnecessary requires if you haven’t cleaned up your imports. FalconJS built up a list of dependencies as they were discovered by the compiler. This seems to be sufficient for the two test cases I’m working with. Any objections to switching away from imports to something more like how FalconJS does it? -- Alex Harui Flex SDK Team Adobe Systems, Inc. http://blogs.adobe.com/aharui -- Ix Multimedia Software Jan Luykenstraat 27 3521 VB Utrecht T. 06-51952295 I. www.ixsoftware.nl -- Ix Multimedia Software Jan Luykenstraat 27 3521 VB Utrecht T. 06-51952295 I. www.ixsoftware.nl
[jira] [Resolved] (FLEX-33481) Improve README for FalconJX compiler
[ https://issues.apache.org/jira/browse/FLEX-33481?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Erik de Bruin resolved FLEX-33481. -- Resolution: Fixed Improve README for FalconJX compiler Key: FLEX-33481 URL: https://issues.apache.org/jira/browse/FLEX-33481 Project: Apache Flex Issue Type: Documentation Components: .Unspecified - Compiler Reporter: Tigran Najaryan Assignee: Erik de Bruin Priority: Minor The current README file is sparse and needs more details. Here is proposed README file content: Apache FalconJx Prototype Cross Compiler Note: steps 1-4 and 7 below describe how to get and build Apache Flex SDK, develop branch. If you already have the the develop branch of SDK skip these steps. To get and build FalconJX compiler follow these steps: 1. Create a working directory for Apache Flex (repo below). 2. cd repo 3. Get bleading-edge of SDK: git clone https://git-wip-us.apache.org/repos/asf/flex-sdk.git sdk cd sdk git checkout develop cd .. 4. Get TLF git clone https://git-wip-us.apache.org/repos/asf/flex-tlf.git tlf 5. Get Falcon and switch to develop branch: git clone https://git-wip-us.apache.org/repos/asf/flex-falcon.git falcon cd falcon git checkout develop cd .. 6. Get ASJS git clone https://git-wip-us.apache.org/repos/asf/flex-asjs.git asjs cd asjs git checkout develop cd .. 7. Build SDK. Follow instructions in sdk/README, then: cd sdk ant main (Select all default answers when prompted, i.e. just hit ENTER until the build finishes). cd .. 8. Build Falcon compiler. Follow instructions in falcon/compiler/README, then: cd falcon/compiler ant -Dsdk.branch=path-to-repo/sdk cd ../.. (Note: the compiler/build.xml assumes Flex SDK is in ../flex-sdk dir, which is not what we have created in step 3 above, so you need to supply correct dir to Ant via sdk.branch parameter). 9. Build FalconJX. cd falcon/compiler.jx ant cd ../.. 10. Build ASJS and FlexJSUI.swc: cd asjs ant cd frameworks/as ant -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (FLEX-33479) Conflict between allowBrowserInvocation and captiveruntime packaging of AIR application
[ https://issues.apache.org/jira/browse/FLEX-33479?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13626630#comment-13626630 ] Abhishek Dwevedi commented on FLEX-33479: - @ Frederic, Do you believe its a bug or limitation of captive run time ? Is this very much expected thing ? Conflict between allowBrowserInvocation and captiveruntime packaging of AIR application - Key: FLEX-33479 URL: https://issues.apache.org/jira/browse/FLEX-33479 Project: Apache Flex Issue Type: Bug Affects Versions: Adobe Flex SDK 4.6 (Release) Environment: ALL Reporter: Abhishek Dwevedi Labels: adt, air, allowbrowser, captiveruntime, packaging Hi Team, I am trying to create an air app to reboot itself, its running fine with native installer or signed package but when i package it using captive runtime its throws issues Case 1: Packagin by Captive Runtime allowBrowserInvocationtrueallowBrowserInvocation Issue Browsers invocation is not supported for this target Case 2: Packagin by Captive Runtime allowBrowserInvocationfalseallowBrowserInvocation Application Doesn't works. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
Re: Git commit emails
On Apr 9, 2013 5:42 AM, Erik de Bruin e...@ixsoftware.nl wrote: I have the feeling I read an update on this somewhere in one of the INFRA messages, but I can't find it... Anyway, the git commit emails are still not formatted the way we discussed. EdB I was told by Infra that this issue was fixed. Is the format wrong or did it not change at all? Thanks, Om On Wed, Mar 27, 2013 at 12:44 PM, Om bigosma...@gmail.com wrote: I made this request to INFRA a while ago. I will ping them again now. (I dont think we can do this ourselves, but if someone knows how, please feel free to) Thanks, Om On Wed, Mar 27, 2013 at 10:43 AM, Erik de Bruin e...@ixsoftware.nl wrote: Is this something we can implement ourselves, or must this go through INFRA? EdB On Mon, Mar 18, 2013 at 6:45 PM, Om bigosma...@gmail.com wrote: On Mon, Mar 18, 2013 at 10:20 AM, Gordon Smith gosm...@adobe.com wrote: The email subject is currently something like git commit: The files/directories to exclude while checkin. Since we have 4 different repos, can we customize this to show which repo the commit was for? (Why? Because I want to pay more attention to commits to flex-falcon than to flex-sdk.) I suggest something like git commit [flex-sdk] The files/directories to exclude while checkin which makes it easy for the eye to filter on the repo. - Gordon There is documentation on how this works: https://git-wip-us.apache.org/docs/switching-to-git.html#commit-email-subject-format I guess the format you want for Falcon would be something like: git commit [%(repo_name)s]: %(subject)s If folks want to come up with a commit email subject format for each repo, I can give that to Infra and they set it up for us. If everyone likes this format, we can keep this format for all the repos. Thanks, Om -- Ix Multimedia Software Jan Luykenstraat 27 3521 VB Utrecht T. 06-51952295 I. www.ixsoftware.nl -- Ix Multimedia Software Jan Luykenstraat 27 3521 VB Utrecht T. 06-51952295 I. www.ixsoftware.nl
Re: Git commit emails
Well, maybe my email client is messing things up, but my commits this afternoon show up like: git commit: %(subject)s while I expected: git commit [%(repo_name)s]: %(subject)s EdB On Tue, Apr 9, 2013 at 9:23 AM, OmPrakash Muppirala bigosma...@gmail.com wrote: On Apr 9, 2013 5:42 AM, Erik de Bruin e...@ixsoftware.nl wrote: I have the feeling I read an update on this somewhere in one of the INFRA messages, but I can't find it... Anyway, the git commit emails are still not formatted the way we discussed. EdB I was told by Infra that this issue was fixed. Is the format wrong or did it not change at all? Thanks, Om On Wed, Mar 27, 2013 at 12:44 PM, Om bigosma...@gmail.com wrote: I made this request to INFRA a while ago. I will ping them again now. (I dont think we can do this ourselves, but if someone knows how, please feel free to) Thanks, Om On Wed, Mar 27, 2013 at 10:43 AM, Erik de Bruin e...@ixsoftware.nl wrote: Is this something we can implement ourselves, or must this go through INFRA? EdB On Mon, Mar 18, 2013 at 6:45 PM, Om bigosma...@gmail.com wrote: On Mon, Mar 18, 2013 at 10:20 AM, Gordon Smith gosm...@adobe.com wrote: The email subject is currently something like git commit: The files/directories to exclude while checkin. Since we have 4 different repos, can we customize this to show which repo the commit was for? (Why? Because I want to pay more attention to commits to flex-falcon than to flex-sdk.) I suggest something like git commit [flex-sdk] The files/directories to exclude while checkin which makes it easy for the eye to filter on the repo. - Gordon There is documentation on how this works: https://git-wip-us.apache.org/docs/switching-to-git.html#commit-email-subject-format I guess the format you want for Falcon would be something like: git commit [%(repo_name)s]: %(subject)s If folks want to come up with a commit email subject format for each repo, I can give that to Infra and they set it up for us. If everyone likes this format, we can keep this format for all the repos. Thanks, Om -- Ix Multimedia Software Jan Luykenstraat 27 3521 VB Utrecht T. 06-51952295 I. www.ixsoftware.nl -- Ix Multimedia Software Jan Luykenstraat 27 3521 VB Utrecht T. 06-51952295 I. www.ixsoftware.nl -- Ix Multimedia Software Jan Luykenstraat 27 3521 VB Utrecht T. 06-51952295 I. www.ixsoftware.nl
RE: FalconJX: bugs in external dependencies (Google Closure Compiler), what to do?
Also, the current FlexJS implementation for both MXML and AS contains several giant methods that simply beg to be broken up into more manageable sets of support functions which maybe even share some common code (like the emission of JSDoc type headers)... That might also be a good way to get your feet wet without needing to change any of the APIs. Can you list a few file names with those giant methods so that I can have a look?
Re: [FalconJX][FlexJS] imports != require?
Locally. Sent via the PANTECH Discover, an ATT 4G LTE smartphone. Erik de Bruin e...@ixsoftware.nl wrote: Locally? Or did I miss a commit? EdB On Tue, Apr 9, 2013 at 8:10 AM, Alex Harui aha...@adobe.com wrote: I think I have copied the Falconjs implementation to falconjx. Sent via the PANTECH Discover, an ATT 4G LTE smartphone. Erik de Bruin e...@ixsoftware.nl wrote: None of my test cases (and none of my real world code, I think they're lazy) uses wildcard imports, so I wasn't aware of this. But since they are a feature of AS, we need to be able to handle them. The compiler seems to be the proper handler for this, so I'll have a look at how FalconJS does this and see how we might re-use your solution. Added to the list. EdB On Mon, Apr 8, 2013 at 3:31 PM, Alex Harui aha...@adobe.com wrote: Erik, I haven’t found the code yet, but the FlexJS output seems to be using import statements as a way of setting up the goog.requires. The closure compiler subsequently doesn’t like wild card imports import foo.bar.*; And I think it will miss same-package class dependencies, unless that is being picked up somewhere else (and of course, will report unnecessary requires if you haven’t cleaned up your imports. FalconJS built up a list of dependencies as they were discovered by the compiler. This seems to be sufficient for the two test cases I’m working with. Any objections to switching away from imports to something more like how FalconJS does it? -- Alex Harui Flex SDK Team Adobe Systems, Inc. http://blogs.adobe.com/aharui -- Ix Multimedia Software Jan Luykenstraat 27 3521 VB Utrecht T. 06-51952295 I. www.ixsoftware.nl -- Ix Multimedia Software Jan Luykenstraat 27 3521 VB Utrecht T. 06-51952295 I. www.ixsoftware.nl
[jira] [Commented] (FLEX-33482) FalconJX generates incorrect paths in deps.js
[ https://issues.apache.org/jira/browse/FLEX-33482?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13626754#comment-13626754 ] Alex Harui commented on FLEX-33482: --- Is all of the code on the same drive? Google expects URIs relative to a deployed app on a server, not file paths scattered over multiple drives. FalconJX generates incorrect paths in deps.js - Key: FLEX-33482 URL: https://issues.apache.org/jira/browse/FLEX-33482 Project: Apache Flex Issue Type: Bug Components: .Unspecified - Compiler Environment: Windows Reporter: Tigran Najaryan Steps to reproduce: 1. Compile FlexJSTest_again example using FalconJX on Windows. 2. Check deps.js in FlexJSTest_again/bin/js-debug/library/closure/goog/ deps.js contains invalid paths like '../../../../D:\WORK\Projects\ApacheFlex\repo\falcon\compiler.jx\bin/D:\WORK\Projects\ApacheFlex\repo\asjs\examples\FlexJSTest_again\bin\js-debug\org\apache\flex\net\HTTPHeader.js' Due to incorrect paths in deps.js when trying to open js-debug /index.html in Firefox I get a bunch of JavaScript errors, the first being SyntaxError: malformed Unicode character escape sequence at deps.js line 4. More details: The bug comes from com.google.javascript.jscomp.deps.PathUtil.isAbsolute() function. It is implemented by simply checking that the path starts by a forward slash which of course is incorrect under Windows. com.google.javascript.jscomp.deps.PathUtil.isAbsolute() is used by com.google.javascript.jscomp.deps.DepsGenerator.computeDependencyCalls() which is called by MXMLFlexJSPublisher to generate deps.js. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
Re: [FalconJX][FlexJS] Dependency calculations
I don't have a test case I can make public. It appears to be ClassA requires ClassB requires ClassC requires ClassA. Which is all valid because the circularity doesn't involve prototype definitions. When I was working with FalconJS, I tried the calcdeps.py and it kept complaining about circular deps. I may just end up copying my deps calculator over to FalconJX. I think the key is that the google tools don't really know the kind of dependency and must report circularities, but we can trust that circularities are benign in our case because the AS compiler would have caught it. I've got a lot of changes locally right now trying to get my customer test case to work. Hopefully I'll get through it soon. On 4/9/13 12:27 AM, Erik de Bruin e...@ixsoftware.nl wrote: Alex, The current 'publisher' part of FalconJx uses the vanilla 'DepsGenerator' functionality of the Closure Compiler. Do you have a minimal test case of the kind of circular dependency you are talking about? It seems like that should be something that Google should have encountered and be able to handle. I'm willing to look into it, but I haven't come across it myself, so a test case might come in handy. EdB On Tue, Apr 9, 2013 at 1:36 AM, Alex Harui aha...@adobe.com wrote: Erik, Have you noticed that the google tools seem to be too sensitive to circular dependencies in the goog.requires? Back when I was using FalconJS, I was trying to use the google tools to compute the deps.js files and it kept failing with circular dependencies. I ended up writing my own deps calculator that kept track of which classes had already been seen and things seemed to work. In my ³customer¹s² test case, FalconJX is now returning a circular dependency error. I haven¹t looked into the FalconJX deps logic, but are you also using some google code for that? Is there some other mechanism other than goog.requires() for dependencies that aren¹t involved in the prototype definition? Thanks, -- Alex Harui Flex SDK Team Adobe Systems, Inc. http://blogs.adobe.com/aharui -- Alex Harui Flex SDK Team Adobe Systems, Inc. http://blogs.adobe.com/aharui
[jira] [Commented] (FLEX-33482) FalconJX generates incorrect paths in deps.js
[ https://issues.apache.org/jira/browse/FLEX-33482?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13626803#comment-13626803 ] Tigran Najaryan commented on FLEX-33482: Yes, everything is on drive D: FalconJX generates incorrect paths in deps.js - Key: FLEX-33482 URL: https://issues.apache.org/jira/browse/FLEX-33482 Project: Apache Flex Issue Type: Bug Components: .Unspecified - Compiler Environment: Windows Reporter: Tigran Najaryan Steps to reproduce: 1. Compile FlexJSTest_again example using FalconJX on Windows. 2. Check deps.js in FlexJSTest_again/bin/js-debug/library/closure/goog/ deps.js contains invalid paths like '../../../../D:\WORK\Projects\ApacheFlex\repo\falcon\compiler.jx\bin/D:\WORK\Projects\ApacheFlex\repo\asjs\examples\FlexJSTest_again\bin\js-debug\org\apache\flex\net\HTTPHeader.js' Due to incorrect paths in deps.js when trying to open js-debug /index.html in Firefox I get a bunch of JavaScript errors, the first being SyntaxError: malformed Unicode character escape sequence at deps.js line 4. More details: The bug comes from com.google.javascript.jscomp.deps.PathUtil.isAbsolute() function. It is implemented by simply checking that the path starts by a forward slash which of course is incorrect under Windows. com.google.javascript.jscomp.deps.PathUtil.isAbsolute() is used by com.google.javascript.jscomp.deps.DepsGenerator.computeDependencyCalls() which is called by MXMLFlexJSPublisher to generate deps.js. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
Re: FalconJX: bugs in external dependencies (Google Closure Compiler), what to do?
org.apache.flex.compiler.internal.codegen.mxml.flexjs.MXMLFlexJSEmitter.java is a good place to start :-) EdB On Tue, Apr 9, 2013 at 10:13 AM, Tigran Najaryan tig...@gmail.com wrote: Also, the current FlexJS implementation for both MXML and AS contains several giant methods that simply beg to be broken up into more manageable sets of support functions which maybe even share some common code (like the emission of JSDoc type headers)... That might also be a good way to get your feet wet without needing to change any of the APIs. Can you list a few file names with those giant methods so that I can have a look? -- Ix Multimedia Software Jan Luykenstraat 27 3521 VB Utrecht T. 06-51952295 I. www.ixsoftware.nl
Re: [FalconJX][FlexJS] imports != require?
Ah, hard for me to review ;-) Do you plan to commit it, or do you want me to look at the current implementation and try to fix that? EdB On Tue, Apr 9, 2013 at 10:49 AM, Alex Harui aha...@adobe.com wrote: Locally. Sent via the PANTECH Discover, an ATT 4G LTE smartphone. Erik de Bruin e...@ixsoftware.nl wrote: Locally? Or did I miss a commit? EdB On Tue, Apr 9, 2013 at 8:10 AM, Alex Harui aha...@adobe.com wrote: I think I have copied the Falconjs implementation to falconjx. Sent via the PANTECH Discover, an ATT 4G LTE smartphone. Erik de Bruin e...@ixsoftware.nl wrote: None of my test cases (and none of my real world code, I think they're lazy) uses wildcard imports, so I wasn't aware of this. But since they are a feature of AS, we need to be able to handle them. The compiler seems to be the proper handler for this, so I'll have a look at how FalconJS does this and see how we might re-use your solution. Added to the list. EdB On Mon, Apr 8, 2013 at 3:31 PM, Alex Harui aha...@adobe.com wrote: Erik, I haven’t found the code yet, but the FlexJS output seems to be using import statements as a way of setting up the goog.requires. The closure compiler subsequently doesn’t like wild card imports import foo.bar.*; And I think it will miss same-package class dependencies, unless that is being picked up somewhere else (and of course, will report unnecessary requires if you haven’t cleaned up your imports. FalconJS built up a list of dependencies as they were discovered by the compiler. This seems to be sufficient for the two test cases I’m working with. Any objections to switching away from imports to something more like how FalconJS does it? -- Alex Harui Flex SDK Team Adobe Systems, Inc. http://blogs.adobe.com/aharui -- Ix Multimedia Software Jan Luykenstraat 27 3521 VB Utrecht T. 06-51952295 I. www.ixsoftware.nl -- Ix Multimedia Software Jan Luykenstraat 27 3521 VB Utrecht T. 06-51952295 I. www.ixsoftware.nl -- Ix Multimedia Software Jan Luykenstraat 27 3521 VB Utrecht T. 06-51952295 I. www.ixsoftware.nl
Re: [DICUSS] Release InstallApacheFlex 2.5.4 - RC5 (was: Release InstallApacheFlex 2.5.3 - RC4)
Thanks Alex, I thought about taking another Vote as example already but where can I find the list of the PMC Members to make the difference between binding and and no binding votes ? -Fred -Message d'origine- From: Alex Harui Sent: Monday, April 08, 2013 8:51 PM To: dev@flex.apache.org ; Frédéric THOMAS Subject: Re: [DICUSS] Release InstallApacheFlex 2.5.4 - RC5 (was: Release InstallApacheFlex 2.5.3 - RC4) You can probably use the website or old vote result thread. Sent via the PANTECH Discover, an ATT 4G LTE smartphone. Frédéric THOMAS webdoubl...@hotmail.com wrote: Hi, I will count the votes tomorrow, is there anything specific I have to know ? where is the official list of the PMCs ? -Fred -Message d'origine- From: Justin Mclean Sent: Saturday, April 06, 2013 1:37 AM To: dev@flex.apache.org Subject: Re: [DICUSS] Release InstallApacheFlex 2.5.4 - RC5 (was: Release InstallApacheFlex 2.5.3 - RC4) Hi, Is it ok ? Can I process to a Vote ? Good to go you can call fro a vote. Justin
Re: [FalconJX][FlexJS] Dependency calculations
I'll try to write some tests that reproduce this. Let's not change any code until we have a way to pinpoint the issue? EdB On Tue, Apr 9, 2013 at 11:24 AM, Alex Harui aha...@adobe.com wrote: I don't have a test case I can make public. It appears to be ClassA requires ClassB requires ClassC requires ClassA. Which is all valid because the circularity doesn't involve prototype definitions. When I was working with FalconJS, I tried the calcdeps.py and it kept complaining about circular deps. I may just end up copying my deps calculator over to FalconJX. I think the key is that the google tools don't really know the kind of dependency and must report circularities, but we can trust that circularities are benign in our case because the AS compiler would have caught it. I've got a lot of changes locally right now trying to get my customer test case to work. Hopefully I'll get through it soon. On 4/9/13 12:27 AM, Erik de Bruin e...@ixsoftware.nl wrote: Alex, The current 'publisher' part of FalconJx uses the vanilla 'DepsGenerator' functionality of the Closure Compiler. Do you have a minimal test case of the kind of circular dependency you are talking about? It seems like that should be something that Google should have encountered and be able to handle. I'm willing to look into it, but I haven't come across it myself, so a test case might come in handy. EdB On Tue, Apr 9, 2013 at 1:36 AM, Alex Harui aha...@adobe.com wrote: Erik, Have you noticed that the google tools seem to be too sensitive to circular dependencies in the goog.requires? Back when I was using FalconJS, I was trying to use the google tools to compute the deps.js files and it kept failing with circular dependencies. I ended up writing my own deps calculator that kept track of which classes had already been seen and things seemed to work. In my ³customer¹s² test case, FalconJX is now returning a circular dependency error. I haven¹t looked into the FalconJX deps logic, but are you also using some google code for that? Is there some other mechanism other than goog.requires() for dependencies that aren¹t involved in the prototype definition? Thanks, -- Alex Harui Flex SDK Team Adobe Systems, Inc. http://blogs.adobe.com/aharui -- Alex Harui Flex SDK Team Adobe Systems, Inc. http://blogs.adobe.com/aharui -- Ix Multimedia Software Jan Luykenstraat 27 3521 VB Utrecht T. 06-51952295 I. www.ixsoftware.nl
Re: [VOTE] Release InstallApacheFlex 2.5.4 - RC5
+1 binding. All my usual tests pass. Thanks, Om On Apr 8, 2013 6:15 AM, Frédéric THOMAS webdoubl...@hotmail.com wrote: +1 Thanks guys, -Fred (Win7 64, fr_FR) On 4/5/13 6:35 PM, Frédéric THOMAS webdoubl...@hotmail.com wrote: *Issues addressed in this Release Candidate:* 1. Enable Flex SDK download stats tracking 2. https://issues.apache.org/**jira/browse/FLEX-33426https://issues.apache.org/jira/browse/FLEX-33426(UI fix for license screen (show regular checkboxes)) 3. https://issues.apache.org/**jira/browse/FLEX-33151https://issues.apache.org/jira/browse/FLEX-33151(Auto-update logic fix) 4. French and Dutch language locale fixes. 5. Added the SDK version number to be downloaded in the window title. 6. https://issues.apache.org/**jira/browse/FLEX-33202https://issues.apache.org/jira/browse/FLEX-33202(more issues). 7. https://issues.apache.org/**jira/browse/FLEX-33419https://issues.apache.org/jira/browse/FLEX-33419(added germain language). This is the first official release of the InstallApacheFlex AIR application. The official Apache distribution is the source kit which can contain only source. The source distributions for Windows and Mac are available here: https://dist.apache.org/repos/**dist/dev/flex/installer/2.5/** RC5/sources/https://dist.apache.org/repos/dist/dev/flex/installer/2.5/RC5/sources/ The binary distributions as a convenience for the respective platforms, available here: https://dist.apache.org/repos/**dist/dev/flex/installer/2.5/** RC5/binaries/https://dist.apache.org/repos/dist/dev/flex/installer/2.5/RC5/binaries/ Along with the installer, we are releasing the 'badge installer' that can be embedded in our site as well as third-party websites. Here is a preview: http://flex.apache.org/**installer.htmlhttp://flex.apache.org/installer.html *Before voting please review the section,* What are the ASF requirements on approving a release? at http://www.apache.org/dev/**release.html#approving-a-**releasehttp://www.apache.org/dev/release.html#approving-a-release *Please vote to approve this release:* [ ] +1 Approve the release [ ] -1 Veto the release (please provide specific comments) This vote will be open for at least 72 hours. Regards, Om Justin Fred Apache Flex PMC members + Release managers -- Alex Harui Flex SDK Team Adobe Systems, Inc. http://blogs.adobe.com/aharui
Re: [jira] [Commented] (FLEX-33482) FalconJX generates incorrect paths in deps.js
Can you please copy your exact command line arguments to here (or in a comment on the JIRA ticket)? I remember that the relative paths only resolve when the inputs exactly match and all resources are correctly positioned with regard to the files being published... EdB On Tue, Apr 9, 2013 at 11:48 AM, Tigran Najaryan (JIRA) j...@apache.org wrote: [ https://issues.apache.org/jira/browse/FLEX-33482?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13626803#comment-13626803 ] Tigran Najaryan commented on FLEX-33482: Yes, everything is on drive D: FalconJX generates incorrect paths in deps.js - Key: FLEX-33482 URL: https://issues.apache.org/jira/browse/FLEX-33482 Project: Apache Flex Issue Type: Bug Components: .Unspecified - Compiler Environment: Windows Reporter: Tigran Najaryan Steps to reproduce: 1. Compile FlexJSTest_again example using FalconJX on Windows. 2. Check deps.js in FlexJSTest_again/bin/js-debug/library/closure/goog/ deps.js contains invalid paths like '../../../../D:\WORK\Projects\ApacheFlex\repo\falcon\compiler.jx\bin/D:\WORK\Projects\ApacheFlex\repo\asjs\examples\FlexJSTest_again\bin\js-debug\org\apache\flex\net\HTTPHeader.js' Due to incorrect paths in deps.js when trying to open js-debug /index.html in Firefox I get a bunch of JavaScript errors, the first being SyntaxError: malformed Unicode character escape sequence at deps.js line 4. More details: The bug comes from com.google.javascript.jscomp.deps.PathUtil.isAbsolute() function. It is implemented by simply checking that the path starts by a forward slash which of course is incorrect under Windows. com.google.javascript.jscomp.deps.PathUtil.isAbsolute() is used by com.google.javascript.jscomp.deps.DepsGenerator.computeDependencyCalls() which is called by MXMLFlexJSPublisher to generate deps.js. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira -- Ix Multimedia Software Jan Luykenstraat 27 3521 VB Utrecht T. 06-51952295 I. www.ixsoftware.nl
RE: Git commit emails
It didn't change at all. - Gordon -Original Message- From: omup...@gmail.com [mailto:omup...@gmail.com] On Behalf Of OmPrakash Muppirala Sent: Tuesday, April 09, 2013 7:24 AM To: dev@flex.apache.org Subject: Re: Git commit emails On Apr 9, 2013 5:42 AM, Erik de Bruin e...@ixsoftware.nl wrote: I have the feeling I read an update on this somewhere in one of the INFRA messages, but I can't find it... Anyway, the git commit emails are still not formatted the way we discussed. EdB I was told by Infra that this issue was fixed. Is the format wrong or did it not change at all? Thanks, Om On Wed, Mar 27, 2013 at 12:44 PM, Om bigosma...@gmail.com wrote: I made this request to INFRA a while ago. I will ping them again now. (I dont think we can do this ourselves, but if someone knows how, please feel free to) Thanks, Om On Wed, Mar 27, 2013 at 10:43 AM, Erik de Bruin e...@ixsoftware.nl wrote: Is this something we can implement ourselves, or must this go through INFRA? EdB On Mon, Mar 18, 2013 at 6:45 PM, Om bigosma...@gmail.com wrote: On Mon, Mar 18, 2013 at 10:20 AM, Gordon Smith gosm...@adobe.com wrote: The email subject is currently something like git commit: The files/directories to exclude while checkin. Since we have 4 different repos, can we customize this to show which repo the commit was for? (Why? Because I want to pay more attention to commits to flex-falcon than to flex-sdk.) I suggest something like git commit [flex-sdk] The files/directories to exclude while checkin which makes it easy for the eye to filter on the repo. - Gordon There is documentation on how this works: https://git-wip-us.apache.org/docs/switching-to-git.html#commit-email-subject-format I guess the format you want for Falcon would be something like: git commit [%(repo_name)s]: %(subject)s If folks want to come up with a commit email subject format for each repo, I can give that to Infra and they set it up for us. If everyone likes this format, we can keep this format for all the repos. Thanks, Om -- Ix Multimedia Software Jan Luykenstraat 27 3521 VB Utrecht T. 06-51952295 I. www.ixsoftware.nl -- Ix Multimedia Software Jan Luykenstraat 27 3521 VB Utrecht T. 06-51952295 I. www.ixsoftware.nl
Re: [DICUSS] Release InstallApacheFlex 2.5.4 - RC5 (was: Release InstallApacheFlex 2.5.3 - RC4)
Hmm. The new website no longer shows who is committer-only. We should probably keep that around for vote counting purposes. I think you're the only one added since the last vote. Here are the PMC ids: aharui bigosmallm cframpton cherreman cmastrandrea erikdebruin espenskogen frishy fthomas gordonsmith greddin igorcosta iwo jeff jhouser jmclean jonbcampos jtellier junheider kkorngut labriola leifwells masuland mheidegger mijordan mschmalle myaiser pent peterelst quetwo rduartes s9tpepper sdelamater tink wave On 4/9/13 9:56 AM, Frédéric THOMAS webdoubl...@hotmail.com wrote: Thanks Alex, I thought about taking another Vote as example already but where can I find the list of the PMC Members to make the difference between binding and and no binding votes ? -Fred -Message d'origine- From: Alex Harui Sent: Monday, April 08, 2013 8:51 PM To: dev@flex.apache.org ; Frédéric THOMAS Subject: Re: [DICUSS] Release InstallApacheFlex 2.5.4 - RC5 (was: Release InstallApacheFlex 2.5.3 - RC4) You can probably use the website or old vote result thread. Sent via the PANTECH Discover, an ATT 4G LTE smartphone. Frédéric THOMAS webdoubl...@hotmail.com wrote: Hi, I will count the votes tomorrow, is there anything specific I have to know ? where is the official list of the PMCs ? -Fred -Message d'origine- From: Justin Mclean Sent: Saturday, April 06, 2013 1:37 AM To: dev@flex.apache.org Subject: Re: [DICUSS] Release InstallApacheFlex 2.5.4 - RC5 (was: Release InstallApacheFlex 2.5.3 - RC4) Hi, Is it ok ? Can I process to a Vote ? Good to go you can call fro a vote. Justin -- Alex Harui Flex SDK Team Adobe Systems, Inc. http://blogs.adobe.com/aharui
Re: [FalconJX][FlexJS] imports != require?
On 4/9/13 9:52 AM, Erik de Bruin e...@ixsoftware.nl wrote: Ah, hard for me to review ;-) Do you plan to commit it, or do you want me to look at the current implementation and try to fix that? If it builds the test app and my customer's app then I will check it in. It sounds like you agree that using imports is insufficient. -- Alex Harui Flex SDK Team Adobe Systems, Inc. http://blogs.adobe.com/aharui
Re: [FalconJX][FlexJS] imports != require?
The FalconJX::MXMLFlexJSEmitter implementation keeps track of which types (instances) are actually used in the MXML and writes the goog.requires from that list. The FalconJX::JSGoogEmitter (the parent class of JSFlexJSEmitter) uses the getAllImports method on the ScopedNode representing the AS class. These two seem to be sufficiently generic to work. I think you are referring to the handling of the imports in a fx:script tag in MXML, correct? I'm sure we can use one of the methods above to handle those as well. Or do you have a rewrite that handles all three types of imports? EdB On Tue, Apr 9, 2013 at 12:25 PM, Alex Harui aha...@adobe.com wrote: On 4/9/13 9:52 AM, Erik de Bruin e...@ixsoftware.nl wrote: Ah, hard for me to review ;-) Do you plan to commit it, or do you want me to look at the current implementation and try to fix that? If it builds the test app and my customer's app then I will check it in. It sounds like you agree that using imports is insufficient. -- Alex Harui Flex SDK Team Adobe Systems, Inc. http://blogs.adobe.com/aharui -- Ix Multimedia Software Jan Luykenstraat 27 3521 VB Utrecht T. 06-51952295 I. www.ixsoftware.nl
Re: [FalconJX][FlexJS] imports != require?
I still think using imports is insufficient.
Here's a simple example:
ClassA.as
package comps
{
public class A
{
public function A()
{
var foo = new B();
}
}
}
ClassB.as
package comps
{
public class B
{
public function B()
{
}
}
}
TestApp.as
package
{
import comps.A;
public class TestApp
{
public function TestApp()
{
var bar = new A();
}
}
}
In this example, when TestApp is compiled, it will compile A and B, but
there are no imports for B so it won't add a goog.requires(comps.B);
And then there is handling wild-card syntax (import comps.*) and unnecessary
goog.requires if I forget to remove an import statement. For example, what
do you want the list of requires to look like for this sloppy code:
TestApp2.as
package
{
import comps.HugeClassA;
import comps.HugeClassB;
public class TestApp2
{
public function TestApp2()
{
}
}
}
IMO, there shouldn't be any requires. Someone simply forgot to clean up
their imports.
It turns out that Falcon builds out a set of dependencies between
compilation units that don't rely on import statements. It relies on the
attempt to resolve indentifiers when compiling code. This is much more
reliable and robust. That's how FalconJS works and it was pretty easy to
copy it to FalconJX.
In summary, relying on imports for dependencies is not the right approach.
For MXML, relying on tags you found helps, but doesn't take care of
dependencies in the script block or event handlers.
On 4/9/13 10:44 AM, Erik de Bruin e...@ixsoftware.nl wrote:
The FalconJX::MXMLFlexJSEmitter implementation keeps track of which
types (instances) are actually used in the MXML and writes the
goog.requires from that list.
The FalconJX::JSGoogEmitter (the parent class of JSFlexJSEmitter) uses
the getAllImports method on the ScopedNode representing the AS
class.
These two seem to be sufficiently generic to work.
I think you are referring to the handling of the imports in a
fx:script tag in MXML, correct? I'm sure we can use one of the methods
above to handle those as well.
Or do you have a rewrite that handles all three types of imports?
EdB
On Tue, Apr 9, 2013 at 12:25 PM, Alex Harui aha...@adobe.com wrote:
On 4/9/13 9:52 AM, Erik de Bruin e...@ixsoftware.nl wrote:
Ah, hard for me to review ;-)
Do you plan to commit it, or do you want me to look at the current
implementation and try to fix that?
If it builds the test app and my customer's app then I will check it in. It
sounds like you agree that using imports is insufficient.
--
Alex Harui
Flex SDK Team
Adobe Systems, Inc.
http://blogs.adobe.com/aharui
--
Alex Harui
Flex SDK Team
Adobe Systems, Inc.
http://blogs.adobe.com/aharui
Re: Git commit emails
Sorry, I should have checked the commit emails. I have alerted Infra about this. Hopefully they will fix it soon. Thanks, Om On Tue, Apr 9, 2013 at 10:14 AM, Gordon Smith gosm...@adobe.com wrote: It didn't change at all. - Gordon -Original Message- From: omup...@gmail.com [mailto:omup...@gmail.com] On Behalf Of OmPrakash Muppirala Sent: Tuesday, April 09, 2013 7:24 AM To: dev@flex.apache.org Subject: Re: Git commit emails On Apr 9, 2013 5:42 AM, Erik de Bruin e...@ixsoftware.nl wrote: I have the feeling I read an update on this somewhere in one of the INFRA messages, but I can't find it... Anyway, the git commit emails are still not formatted the way we discussed. EdB I was told by Infra that this issue was fixed. Is the format wrong or did it not change at all? Thanks, Om On Wed, Mar 27, 2013 at 12:44 PM, Om bigosma...@gmail.com wrote: I made this request to INFRA a while ago. I will ping them again now. (I dont think we can do this ourselves, but if someone knows how, please feel free to) Thanks, Om On Wed, Mar 27, 2013 at 10:43 AM, Erik de Bruin e...@ixsoftware.nl wrote: Is this something we can implement ourselves, or must this go through INFRA? EdB On Mon, Mar 18, 2013 at 6:45 PM, Om bigosma...@gmail.com wrote: On Mon, Mar 18, 2013 at 10:20 AM, Gordon Smith gosm...@adobe.com wrote: The email subject is currently something like git commit: The files/directories to exclude while checkin. Since we have 4 different repos, can we customize this to show which repo the commit was for? (Why? Because I want to pay more attention to commits to flex-falcon than to flex-sdk.) I suggest something like git commit [flex-sdk] The files/directories to exclude while checkin which makes it easy for the eye to filter on the repo. - Gordon There is documentation on how this works: https://git-wip-us.apache.org/docs/switching-to-git.html#commit-email-subject-format I guess the format you want for Falcon would be something like: git commit [%(repo_name)s]: %(subject)s If folks want to come up with a commit email subject format for each repo, I can give that to Infra and they set it up for us. If everyone likes this format, we can keep this format for all the repos. Thanks, Om -- Ix Multimedia Software Jan Luykenstraat 27 3521 VB Utrecht T. 06-51952295 I. www.ixsoftware.nl -- Ix Multimedia Software Jan Luykenstraat 27 3521 VB Utrecht T. 06-51952295 I. www.ixsoftware.nl
Re: [FalconJX][FlexJS] imports != require?
I'm writing test cases for all of the issues you mentioned to see how
the current code behaves (and so we can test the new code if it's
needed).
EdB
On Tue, Apr 9, 2013 at 2:05 PM, Alex Harui aha...@adobe.com wrote:
I still think using imports is insufficient.
Here's a simple example:
ClassA.as
package comps
{
public class A
{
public function A()
{
var foo = new B();
}
}
}
ClassB.as
package comps
{
public class B
{
public function B()
{
}
}
}
TestApp.as
package
{
import comps.A;
public class TestApp
{
public function TestApp()
{
var bar = new A();
}
}
}
In this example, when TestApp is compiled, it will compile A and B, but
there are no imports for B so it won't add a goog.requires(comps.B);
And then there is handling wild-card syntax (import comps.*) and unnecessary
goog.requires if I forget to remove an import statement. For example, what
do you want the list of requires to look like for this sloppy code:
TestApp2.as
package
{
import comps.HugeClassA;
import comps.HugeClassB;
public class TestApp2
{
public function TestApp2()
{
}
}
}
IMO, there shouldn't be any requires. Someone simply forgot to clean up
their imports.
It turns out that Falcon builds out a set of dependencies between
compilation units that don't rely on import statements. It relies on the
attempt to resolve indentifiers when compiling code. This is much more
reliable and robust. That's how FalconJS works and it was pretty easy to
copy it to FalconJX.
In summary, relying on imports for dependencies is not the right approach.
For MXML, relying on tags you found helps, but doesn't take care of
dependencies in the script block or event handlers.
On 4/9/13 10:44 AM, Erik de Bruin e...@ixsoftware.nl wrote:
The FalconJX::MXMLFlexJSEmitter implementation keeps track of which
types (instances) are actually used in the MXML and writes the
goog.requires from that list.
The FalconJX::JSGoogEmitter (the parent class of JSFlexJSEmitter) uses
the getAllImports method on the ScopedNode representing the AS
class.
These two seem to be sufficiently generic to work.
I think you are referring to the handling of the imports in a
fx:script tag in MXML, correct? I'm sure we can use one of the methods
above to handle those as well.
Or do you have a rewrite that handles all three types of imports?
EdB
On Tue, Apr 9, 2013 at 12:25 PM, Alex Harui aha...@adobe.com wrote:
On 4/9/13 9:52 AM, Erik de Bruin e...@ixsoftware.nl wrote:
Ah, hard for me to review ;-)
Do you plan to commit it, or do you want me to look at the current
implementation and try to fix that?
If it builds the test app and my customer's app then I will check it in. It
sounds like you agree that using imports is insufficient.
--
Alex Harui
Flex SDK Team
Adobe Systems, Inc.
http://blogs.adobe.com/aharui
--
Alex Harui
Flex SDK Team
Adobe Systems, Inc.
http://blogs.adobe.com/aharui
--
Ix Multimedia Software
Jan Luykenstraat 27
3521 VB Utrecht
T. 06-51952295
I. www.ixsoftware.nl
[jira] [Commented] (FLEX-33169) Spark List selection on mobile doesn't work properly
[ https://issues.apache.org/jira/browse/FLEX-33169?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13627069#comment-13627069 ] Fréderic Cox commented on FLEX-33169: - This is one of those issues that causes people to say things like Well it doesn't work as smooth/good as real native apps. Would love to get this committed in a next apache flex release Spark List selection on mobile doesn't work properly Key: FLEX-33169 URL: https://issues.apache.org/jira/browse/FLEX-33169 Project: Apache Flex Issue Type: Bug Components: Spark: List Affects Versions: Adobe Flex SDK 4.5.1 (Release), Adobe Flex SDK 4.6 (Release) Environment: iOS 5.x, Android 3.2/4.0, Air 3.0 - 3.3 Reporter: Al Birdy Attachments: FLEX-33169_Mobile_List_Selection.patch, ListExt.as The touch selection of list items on the actual device does not work properly. Sometimes, even though the touch has definitely been recognized by the device, the list selection change isn't committed, instead the previously selected item remains selected. To reproduce the issue, please set up a new Flex mobile project and add the most basic list with inline renderer to a view: s:List id=myList width=100% height=100% s:dataProvider s:ArrayCollection fx:Object label=Item1 / . fx:Object label=Item30 / /s:ArrayCollection /s:dataProvider /s:List Deploy the app to an actual device (this bug does NOT appear in the emulator!) and randomly try to select items, scroll the list, select again and so on. You'll notice that the down color is sometimes set, but the selection isn't changed. For a more detailed description of this bug please see http://stackoverflow.com/questions/11740074/selection-on-spark-mobile-list-only-works-3-out-of-4-times -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
Re: [FalconJX][FlexJS] imports != require?
Ok, too much effort (time) to create project like test cases right
now (so we can test dependencies between files), I'll add that to my
ToDo list.
I did see that the getAllImports() method the compiler provides
apparently doesn't resolve dependencies (like I expected it to do,
since in the compiler and all) but only lists the import statements
that are in the file. So, my bad for wasting your time Alex, it looks
like your solution is the one we should have been looking for all
along. Just goes to show that the strength and weakness of unit
testing; weakness in the sense that if you don't write a test, it
won't fail ;-)
EdB
On Tue, Apr 9, 2013 at 3:16 PM, Erik de Bruin e...@ixsoftware.nl wrote:
I'm writing test cases for all of the issues you mentioned to see how
the current code behaves (and so we can test the new code if it's
needed).
EdB
On Tue, Apr 9, 2013 at 2:05 PM, Alex Harui aha...@adobe.com wrote:
I still think using imports is insufficient.
Here's a simple example:
ClassA.as
package comps
{
public class A
{
public function A()
{
var foo = new B();
}
}
}
ClassB.as
package comps
{
public class B
{
public function B()
{
}
}
}
TestApp.as
package
{
import comps.A;
public class TestApp
{
public function TestApp()
{
var bar = new A();
}
}
}
In this example, when TestApp is compiled, it will compile A and B, but
there are no imports for B so it won't add a goog.requires(comps.B);
And then there is handling wild-card syntax (import comps.*) and unnecessary
goog.requires if I forget to remove an import statement. For example, what
do you want the list of requires to look like for this sloppy code:
TestApp2.as
package
{
import comps.HugeClassA;
import comps.HugeClassB;
public class TestApp2
{
public function TestApp2()
{
}
}
}
IMO, there shouldn't be any requires. Someone simply forgot to clean up
their imports.
It turns out that Falcon builds out a set of dependencies between
compilation units that don't rely on import statements. It relies on the
attempt to resolve indentifiers when compiling code. This is much more
reliable and robust. That's how FalconJS works and it was pretty easy to
copy it to FalconJX.
In summary, relying on imports for dependencies is not the right approach.
For MXML, relying on tags you found helps, but doesn't take care of
dependencies in the script block or event handlers.
On 4/9/13 10:44 AM, Erik de Bruin e...@ixsoftware.nl wrote:
The FalconJX::MXMLFlexJSEmitter implementation keeps track of which
types (instances) are actually used in the MXML and writes the
goog.requires from that list.
The FalconJX::JSGoogEmitter (the parent class of JSFlexJSEmitter) uses
the getAllImports method on the ScopedNode representing the AS
class.
These two seem to be sufficiently generic to work.
I think you are referring to the handling of the imports in a
fx:script tag in MXML, correct? I'm sure we can use one of the methods
above to handle those as well.
Or do you have a rewrite that handles all three types of imports?
EdB
On Tue, Apr 9, 2013 at 12:25 PM, Alex Harui aha...@adobe.com wrote:
On 4/9/13 9:52 AM, Erik de Bruin e...@ixsoftware.nl wrote:
Ah, hard for me to review ;-)
Do you plan to commit it, or do you want me to look at the current
implementation and try to fix that?
If it builds the test app and my customer's app then I will check it in.
It
sounds like you agree that using imports is insufficient.
--
Alex Harui
Flex SDK Team
Adobe Systems, Inc.
http://blogs.adobe.com/aharui
--
Alex Harui
Flex SDK Team
Adobe Systems, Inc.
http://blogs.adobe.com/aharui
--
Ix Multimedia Software
Jan Luykenstraat 27
3521 VB Utrecht
T. 06-51952295
I. www.ixsoftware.nl
--
Ix Multimedia Software
Jan Luykenstraat 27
3521 VB Utrecht
T. 06-51952295
I. www.ixsoftware.nl
Re: [FalconJX][FlexJS] imports != require?
And then I discovered that, even with my own dependency generator, when you
pass this set to the closure compiler for optimization it barfs on the
circularity and there won't be any working around that.
I'm testing out a change to the dependency list logic to try to detect
circularities up front. The logic is basically: if someone has already
require'd that class then you shouldn't need to require it.
But I'm still wondering if goog.require is supposed to be for all
requirements or just for prototype requirements and if we're supposed to
resolve other dependencies some other way. For example:
public class A extends Base
{
public function A()
{
var foo = new B();
bar = new C(this);
}
private var bar:C;
}
public class C
{
public function C(owner:Object)
{
this.owner = owner;
}
private var owner:A;
}
The only hard dependency is on Base.
It is in the adding of requires for B and C that you get circularities. But
technically, it doesn't matter when B and C get loaded as long as they get
loaded before you instantiate A.
For a more real example, look at FlexJSTest_again.mxml and
MyController.as. FlexJSTest_again instantiates MyController, and
MyController wants a back reference to FlexJSTest_again. MyController
doesn't import FlexJSTest_again because FlexJSTest_again is in the default
namespace, but the new dependency code will add a requires to MyController
for FlexJSTest_again and a circularity is formed. It seems like only
dependencies like the base class and instance and static initializers should
be in the requires list for a class and maybe all other dependencies should
be in another list of requires for the main app.
On 4/9/13 2:22 PM, Erik de Bruin e...@ixsoftware.nl wrote:
Ok, too much effort (time) to create project like test cases right
now (so we can test dependencies between files), I'll add that to my
ToDo list.
I did see that the getAllImports() method the compiler provides
apparently doesn't resolve dependencies (like I expected it to do,
since in the compiler and all) but only lists the import statements
that are in the file. So, my bad for wasting your time Alex, it looks
like your solution is the one we should have been looking for all
along. Just goes to show that the strength and weakness of unit
testing; weakness in the sense that if you don't write a test, it
won't fail ;-)
EdB
On Tue, Apr 9, 2013 at 3:16 PM, Erik de Bruin e...@ixsoftware.nl wrote:
I'm writing test cases for all of the issues you mentioned to see how
the current code behaves (and so we can test the new code if it's
needed).
EdB
On Tue, Apr 9, 2013 at 2:05 PM, Alex Harui aha...@adobe.com wrote:
I still think using imports is insufficient.
Here's a simple example:
ClassA.as
package comps
{
public class A
{
public function A()
{
var foo = new B();
}
}
}
ClassB.as
package comps
{
public class B
{
public function B()
{
}
}
}
TestApp.as
package
{
import comps.A;
public class TestApp
{
public function TestApp()
{
var bar = new A();
}
}
}
In this example, when TestApp is compiled, it will compile A and B, but
there are no imports for B so it won't add a goog.requires(comps.B);
And then there is handling wild-card syntax (import comps.*) and unnecessary
goog.requires if I forget to remove an import statement. For example, what
do you want the list of requires to look like for this sloppy code:
TestApp2.as
package
{
import comps.HugeClassA;
import comps.HugeClassB;
public class TestApp2
{
public function TestApp2()
{
}
}
}
IMO, there shouldn't be any requires. Someone simply forgot to clean up
their imports.
It turns out that Falcon builds out a set of dependencies between
compilation units that don't rely on import statements. It relies on the
attempt to resolve indentifiers when compiling code. This is much more
reliable and robust. That's how FalconJS works and it was pretty easy to
copy it to FalconJX.
In summary, relying on imports for dependencies is not the right approach.
For MXML, relying on tags you found helps, but doesn't take care of
dependencies in the script block or event handlers.
On 4/9/13 10:44 AM, Erik de Bruin e...@ixsoftware.nl wrote:
The FalconJX::MXMLFlexJSEmitter implementation keeps track of which
types (instances) are actually used in the MXML and writes the
goog.requires from that list.
The FalconJX::JSGoogEmitter (the parent class of JSFlexJSEmitter) uses
the getAllImports method on the ScopedNode representing the AS
class.
These two seem to be sufficiently generic to work.
I think you are referring to the handling of the imports in a
fx:script tag in MXML, correct? I'm sure we can use one of the methods
[jira] [Commented] (FLEX-33169) Spark List selection on mobile doesn't work properly
[ https://issues.apache.org/jira/browse/FLEX-33169?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13627343#comment-13627343 ] Matt Krokosz commented on FLEX-33169: - After pulling my hair out on this one and trying the suggested patch without much luck on my custom item renderer, I found another solution / work-around that works very well for me...I hope it works for everyone else too. The solution was very simple, just set mouseEnabled to false on all of the components within your List item. That prevents the mouse up/down events from targetting those components and creating the problem. Spark List selection on mobile doesn't work properly Key: FLEX-33169 URL: https://issues.apache.org/jira/browse/FLEX-33169 Project: Apache Flex Issue Type: Bug Components: Spark: List Affects Versions: Adobe Flex SDK 4.5.1 (Release), Adobe Flex SDK 4.6 (Release) Environment: iOS 5.x, Android 3.2/4.0, Air 3.0 - 3.3 Reporter: Al Birdy Attachments: FLEX-33169_Mobile_List_Selection.patch, ListExt.as The touch selection of list items on the actual device does not work properly. Sometimes, even though the touch has definitely been recognized by the device, the list selection change isn't committed, instead the previously selected item remains selected. To reproduce the issue, please set up a new Flex mobile project and add the most basic list with inline renderer to a view: s:List id=myList width=100% height=100% s:dataProvider s:ArrayCollection fx:Object label=Item1 / . fx:Object label=Item30 / /s:ArrayCollection /s:dataProvider /s:List Deploy the app to an actual device (this bug does NOT appear in the emulator!) and randomly try to select items, scroll the list, select again and so on. You'll notice that the down color is sometimes set, but the selection isn't changed. For a more detailed description of this bug please see http://stackoverflow.com/questions/11740074/selection-on-spark-mobile-list-only-works-3-out-of-4-times -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira
AIR 3.7 and Apache Flex
Hi, AIR 3.7 is out but no Apache Flex compatible version yet. http://labs.adobe.com/technologies/flashruntimes/air/ We are continuing to release beta updates of AIR 3.7.x, including the addition of the AIR SDK for Flex developers. A little odd that they have a release version and are continuing to release 3.7 beta's IMO but at least they have stated it's coming. Thanks, Justin
Re: AIR 3.7 and Apache Flex
Yeah... seemed pretty odd, but I'm assuming it's because the new AIR has only been tested with the new Falcon compiler, and not old compiler.. -Nick On Tue, Apr 9, 2013 at 9:51 PM, Justin Mclean jus...@classsoftware.comwrote: Hi, AIR 3.7 is out but no Apache Flex compatible version yet. http://labs.adobe.com/technologies/flashruntimes/air/ We are continuing to release beta updates of AIR 3.7.x, including the addition of the AIR SDK for Flex developers. A little odd that they have a release version and are continuing to release 3.7 beta's IMO but at least they have stated it's coming. Thanks, Justin
SVG button with skin states and JS
Hi, Here is some more experimental work based on FXG to SVG transformation. http://people.apache.org/~bigosmallm/fxg2svg/buttonskinjs/embedButtons.html This time, all the skin states are in one .svg file along with the JS: http://people.apache.org/~bigosmallm/fxg2svg/buttonskinjs/buttonSkin.svg Notes: * My earlier approach of using each .svg file as a background of each button state has a major problem - it is not runtime scriptable because of security restrictions. * In this version, I am using the embed tag with the src attribute pointing to the .svg skin. * Inside the SVG, the states are being managed using JavaScript * The button's label and click callback function is injected into the SVG file via the embed tag's attributes. * Inside the SVG, the onInit() function reads the window.frameElement.attributes properties to read and set the required variables * Chrome and Firefox restrict accessing window.frameElement when working on the filesystem (cross domain scripting security precaution) But it works fine when running from a server. I will figure out a way around this for development ease soon. * Communication from SVG to the parent HTML is done via calling top[injectedCallbackFunctionName](); * In this function call, we can send any parameters we choose. For now, I am sending the the original MouseEvent and the this (SVG window) object. * In the HTML page, the embed objects are styled via CSS as usual. * Tested to work fine with Firefox, Chrome and Android Browser. I will test with IE9 and iOS in a bit. If someone can test it in the meantime, that would be great. * I need to play with the pointer-events property of the SVG objects to improve the hit area of the button. In Chrome, clicking on the text seems to work, but not on the button background. Please let me know your feedback on this approach. Thanks, Om
RE: How to get a clean working tree for flex-sdk
Background: I'm trying to get 'git status' to report nothing after simply doing a build of the SDK (which is what you expect because you haven't changed any source files). The problem is that a copy of flex-config.xml and a replace of Version.java in our Ant scripts seem to misbehave on Cygwin and have the side effect of doing chmod +x on these two files. This mode change then makes them show up as changed in 'git status'. Using Ant's chmod task to set -x doesn't seem to work. I don't think Alex's save-a-copy approach will work, because it use copy, which is causing the problem in the first place. I looked into his suggestion of doing a git checkout of the problematic files to put them back into their original states. This solves the problem. But is it OK for the Ant scripts to assume that a command-line git is installed? Some people might only have a Git GUI. And if so, is it OK for them to assume that git is on the path, or should GIT be another environment variable that you have to set? If we can't use Git, I think I would need to replace the copy and replace tasks with an exec that executes a small Java program (which we would have to compile from source). Is anybody else doing Apache Flex development using Cygwin? - Gordon -Original Message- From: Alex Harui [mailto:aha...@adobe.com] Sent: Thursday, April 04, 2013 10:44 PM To: dev@flex.apache.org Subject: Re: How to get a clean working tree for flex-sdk Can we get the build to save a copy of these files before modification and then replace them at the end of a build even if the build fails? Or maybe just add git checkout to the build script? On 4/4/13 9:28 PM, Gordon Smith gosm...@adobe.com wrote: After adding 3 entries to .gitignore to suppress some build output files from showing up in 'git status', I'm now at the point where 'git status' after a build shows only this: gosmith@GOSMITH-WIN7 /cygdrive/d/asf/flex-sdk $ git status # On branch develop # Changes not staged for commit: # (use git add file... to update what will be committed) # (use git checkout -- file... to discard changes in working directory) # # modified: frameworks/flex-config.xml # modified: modules/asc/src/java/macromedia/asc/util/Version.java # no changes added to commit (use git add and/or git commit -a) Git is detecting that the SDK's Ant build scripts change the file mode of flex-config.xml (which gets diddled with an Ant copy) and Version.java (diddled with an Ant replace). These tasks seem to have the side effect of doing chmod +x for some reason. I've tried using chmod file=... perm=-x to fix this, but it doesn't seem to work. Both of these issues are probably some unfortunate interaction between Ant and Cygwin, but googling does't confirm this.(It does confirm that the people who wrote Ant don't think much of Cygwin. If anybody has an idea of how to fix this, so that we get clean 'git status' output after a build -- other than configurating Git to ignore file modes -- let me know. I suppose one solution is to just use Command Prompt instead of Cygwin, but then I have to give up other stuff. And another is to use a Mac, which is what most people seem to be doing these days. - Gordon -Original Message- From: Alex Harui [mailto:aha...@adobe.com] Sent: Thursday, April 04, 2013 8:46 PM To: dev@flex.apache.org Subject: Re: How to get a clean working tree for flex-sdk Awesome! Thanks. On 4/4/13 5:24 PM, Gordon Smith gosm...@adobe.com wrote: OK, I've cleaned up the bad file modes in flex_tlf. There don't seem to be any in flex-falcon or flex-asjs. - Gordon -Original Message- From: Gordon Smith [mailto:gosm...@adobe.com] Sent: Thursday, April 04, 2013 1:31 PM To: dev@flex.apache.org Subject: RE: How to get a clean working tree for flex-sdk I've pushed the file mode fixes to flex-sdk. 'ant clean main checkintests' works for me in Cygwin on Windows. Please let me know if I screwed anything up on Macs. I'll clean up the bad file modes in flex-tlf tonight. - Gordon -Original Message- From: Gordon Smith [mailto:gosm...@adobe.com] Sent: Thursday, April 04, 2013 1:14 PM To: dev@flex.apache.org Subject: RE: How to get a clean working tree for flex-sdk After you're done, will a simple pull correct the local repo's? I'm not a Git expert, but I'm pretty sure that it will. However, the fact that many files were marked as executable (+x) when they shouldn't be is not causing any problems that I know of. Fixing this is just housecleaning. - Gordon -Original Message- From: Kessler CTR Mark J [mailto:mark.kessler@usmc.mil] Sent: Thursday, April 04, 2013 10:25 AM To: dev@flex.apache.org Subject: RE: How to get a clean working tree for flex-sdk After you're done, will a simple pull correct the local repo's? -Mark -Original Message- From: Gordon Smith
Re: SVG button with skin states and JS
On Apr 9, 2013 7:40 PM, OmPrakash Muppirala bigosma...@gmail.com wrote: Hi, Here is some more experimental work based on FXG to SVG transformation. http://people.apache.org/~bigosmallm/fxg2svg/buttonskinjs/embedButtons.html This time, all the skin states are in one .svg file along with the JS: http://people.apache.org/~bigosmallm/fxg2svg/buttonskinjs/buttonSkin.svg Notes: * My earlier approach of using each .svg file as a background of each button state has a major problem - it is not runtime scriptable because of security restrictions. * In this version, I am using the embed tag with the src attribute pointing to the .svg skin. * Inside the SVG, the states are being managed using JavaScript * The button's label and click callback function is injected into the SVG file via the embed tag's attributes. * Inside the SVG, the onInit() function reads the window.frameElement.attributes properties to read and set the required variables * Chrome and Firefox restrict accessing window.frameElement when working on the filesystem (cross domain scripting security precaution) But it works fine when running from a server. I will figure out a way around this for development ease soon. * Communication from SVG to the parent HTML is done via calling top[injectedCallbackFunctionName](); * In this function call, we can send any parameters we choose. For now, I am sending the the original MouseEvent and the this (SVG window) object. * In the HTML page, the embed objects are styled via CSS as usual. * Tested to work fine with Firefox, Chrome and Android Browser. I will test with IE9 and iOS in a bit. If someone can test it in the meantime, that would be great. * I need to play with the pointer-events property of the SVG objects to improve the hit area of the button. In Chrome, clicking on the text seems to work, but not on the button background. A simple rect with no fill and pointer-events set to 'all' did the trick. And testing on iOS worked fine as well. Thanks, Om Please let me know your feedback on this approach. Thanks, Om
Re: How to get a clean working tree for flex-sdk
Hi, JFYI I'm running into a similar issue in windows in OSX. FlexComponentKit.mxi seems to have differing permissions but not sure if it should be added to .gitignore or not. You may also need to modify build.properties depending on your setup (eg TLF path) - how do we deal with that? Thanks, Justin
Seeing what 'ant clean' isn't cleaning
If you do 'git status --ignored' after 'ant clean', it will reveal any (normally-ignored) build output that didn't get cleaned. - Gordon
Re: Seeing what 'ant clean' isn't cleaning
Hi, If you do 'git status --ignored' after 'ant clean', it will reveal any (normally-ignored) build output that didn't get cleaned. Any reason why these are ignored? # modules/compiler/src/java/flex2/compiler/mxml/ParseException.java # modules/compiler/src/java/flex2/compiler/mxml/Parser.java # modules/compiler/src/java/flex2/compiler/mxml/ParserConstants.java # modules/compiler/src/java/flex2/compiler/mxml/Token.java # modules/compiler/src/java/flex2/compiler/mxml/TokenManager.java # modules/compiler/src/java/flex2/compiler/mxml/TokenMgrError.java Justin
Re: How to get a clean working tree for flex-sdk
The .mxi is a source file so it shouldn't be in gitignore. The Extension Manager changes its permissions to executable for unknown reason. I thought I checked in an .mxi file with executable permissions to try to cure this problem. Maybe I put it in the wrong branch or maybe it didn't work on Windows. On 4/9/13 8:34 PM, Justin Mclean jus...@classsoftware.com wrote: Hi, JFYI I'm running into a similar issue in windows in OSX. FlexComponentKit.mxi seems to have differing permissions but not sure if it should be added to .gitignore or not. You may also need to modify build.properties depending on your setup (eg TLF path) - how do we deal with that? Thanks, Justin -- Alex Harui Flex SDK Team Adobe Systems, Inc. http://blogs.adobe.com/aharui
Re: How to get a clean working tree for flex-sdk
I probably do not understand enough of the problem you are describing, and probably don't have enough smarts to provide anything of value, but, for what it is worth: A. The cygwin version of the 'cp' utility has a '-p' option which, in my experience, keeps its promise; the target file will have precisely the same mode, ownership and timestamp as the source file did. B. Assuming that your ant tasks are just a convenient means for scripting *nix commands, it would seem that you ought to be able to get ant to use the 'cp -p'' syntax to accomplish anything that is merely a copy operation. C. If, on the other hand, what is being accomplished is not a simple copy, rather it is an update -- to roll a revision number, perform macro replacement or the like -- the output is different from the input, so if the process reports that as a change, it does so correctly. D. If it is necessary and desired that a modified target be able to masquerade as if it were the source, follow the copy operation with chmod using the -reference option where the referent will point to the source. On Tue, Apr 9, 2013 at 8:18 PM, Gordon Smith gosm...@adobe.com wrote: Background: I'm trying to get 'git status' to report nothing after simply doing a build of the SDK (which is what you expect because you haven't changed any source files). The problem is that a copy of flex-config.xml and a replace of Version.java in our Ant scripts seem to misbehave on Cygwin and have the side effect of doing chmod +x on these two files. This mode change then makes them show up as changed in 'git status'. Using Ant's chmod task to set -x doesn't seem to work. I don't think Alex's save-a-copy approach will work, because it use copy, which is causing the problem in the first place. I looked into his suggestion of doing a git checkout of the problematic files to put them back into their original states. This solves the problem. But is it OK for the Ant scripts to assume that a command-line git is installed? Some people might only have a Git GUI. And if so, is it OK for them to assume that git is on the path, or should GIT be another environment variable that you have to set? If we can't use Git, I think I would need to replace the copy and replace tasks with an exec that executes a small Java program (which we would have to compile from source). Is anybody else doing Apache Flex development using Cygwin? - Gordon -Original Message- From: Alex Harui [mailto:aha...@adobe.com] Sent: Thursday, April 04, 2013 10:44 PM To: dev@flex.apache.org Subject: Re: How to get a clean working tree for flex-sdk Can we get the build to save a copy of these files before modification and then replace them at the end of a build even if the build fails? Or maybe just add git checkout to the build script? On 4/4/13 9:28 PM, Gordon Smith gosm...@adobe.com wrote: After adding 3 entries to .gitignore to suppress some build output files from showing up in 'git status', I'm now at the point where 'git status' after a build shows only this: gosmith@GOSMITH-WIN7 /cygdrive/d/asf/flex-sdk $ git status # On branch develop # Changes not staged for commit: # (use git add file... to update what will be committed) # (use git checkout -- file... to discard changes in working directory) # # modified: frameworks/flex-config.xml # modified: modules/asc/src/java/macromedia/asc/util/Version.java # no changes added to commit (use git add and/or git commit -a) Git is detecting that the SDK's Ant build scripts change the file mode of flex-config.xml (which gets diddled with an Ant copy) and Version.java (diddled with an Ant replace). These tasks seem to have the side effect of doing chmod +x for some reason. I've tried using chmod file=... perm=-x to fix this, but it doesn't seem to work. Both of these issues are probably some unfortunate interaction between Ant and Cygwin, but googling does't confirm this.(It does confirm that the people who wrote Ant don't think much of Cygwin. If anybody has an idea of how to fix this, so that we get clean 'git status' output after a build -- other than configurating Git to ignore file modes -- let me know. I suppose one solution is to just use Command Prompt instead of Cygwin, but then I have to give up other stuff. And another is to use a Mac, which is what most people seem to be doing these days. - Gordon -Original Message- From: Alex Harui [mailto:aha...@adobe.com] Sent: Thursday, April 04, 2013 8:46 PM To: dev@flex.apache.org Subject: Re: How to get a clean working tree for flex-sdk Awesome! Thanks. On 4/4/13 5:24 PM, Gordon Smith gosm...@adobe.com wrote: OK, I've cleaned up the bad file modes in flex_tlf. There don't seem to be any in flex-falcon or flex-asjs. - Gordon -Original Message- From: Gordon Smith [mailto:gosm...@adobe.com] Sent: Thursday, April 04,
Re: SVG button with skin states and JS
Good progress. I didn't look at your implementation, but some comments inline. On 4/9/13 7:40 PM, OmPrakash Muppirala bigosma...@gmail.com wrote: Hi, Here is some more experimental work based on FXG to SVG transformation. http://people.apache.org/~bigosmallm/fxg2svg/buttonskinjs/embedButtons.html This time, all the skin states are in one .svg file along with the JS: http://people.apache.org/~bigosmallm/fxg2svg/buttonskinjs/buttonSkin.svg Notes: * My earlier approach of using each .svg file as a background of each button state has a major problem - it is not runtime scriptable because of security restrictions. If you don't need runtime scriptability, is separate svg files a better solution? Maybe we need to support both? Especially because a quick search of how-to seemed to find the separate svg file approach. * In this version, I am using the embed tag with the src attribute pointing to the .svg skin. * Inside the SVG, the states are being managed using JavaScript Do the states get tied to CSS pseudo-states or something else. If we do have to support two mechanisms (embedded js states and separate svg files with css, is there a higher level abstraction that can encapsulate both? * The button's label and click callback function is injected into the SVG file via the embed tag's attributes. Why is the callback for click part of the view? * Inside the SVG, the onInit() function reads the window.frameElement.attributes properties to read and set the required variables * Chrome and Firefox restrict accessing window.frameElement when working on the filesystem (cross domain scripting security precaution) But it works fine when running from a server. I will figure out a way around this for development ease soon. Try launching chrome with these additional command-line parameters: --allow-file-access-from-files --disable-web-security * Communication from SVG to the parent HTML is done via calling top[injectedCallbackFunctionName](); What kind of communication is required? I would hope that most skins don't have events or other output. -- Alex Harui Flex SDK Team Adobe Systems, Inc. http://blogs.adobe.com/aharui
