http://blog.gmane.org/gmane.comp.java.openlaszlo.devel hourly 1 1901-01-01T00:00+00:00 http://gmane.org/img/gmane-25t.png http://gmane.org http://comments.gmane.org/gmane.comp.java.openlaszlo.devel/28808 <pre> (Note: I am checking this changeset in.) Change philip-20130506-t0j by philip&lt; at &gt;philip-i7-2 on 2013-05-06 20:42:11 EDT     in /cygdrive/f/clients/laszlo/svn/openlaszlo/trunk     for http://svn.openlaszlo.org/openlaszlo/trunk Summary: Improve performance of html component New Features: Bugs Fixed: LPP-10276 Technical Reviewer: (pending) QA Reviewer: antun Doc Reviewer: (pending) Documentation: Release Notes: Overview: Details: I think some recent changes I made to reduce closures resolved the memory leak (I did not observe them during testing). I did make a change to cache the location of the iframe to improve performance. The update runs via the idle timer and monitors the iframe position and zoom value. Most of the method is needed only when the window size changes. Tests: Test case in jira. I do not see a memory leak in IE9, but I suspect this is from an earlier change where I removed a bunch of closures. I did enable some code to cache the position of the iframes. Performance improvement is only about 1%, but it should help reduce any flickering when the iframe is moved. Files: M       lps/includes/source/iframemanager.js Changeset: http://svn.openlaszlo.org/openlaszlo/patches/philip-20130506-t0j.tar </pre> promanik< at >laszlosystems.com 2013-05-07T00:52:08 http://comments.gmane.org/gmane.comp.java.openlaszlo.devel/28802 <pre>How are custom type objects instantiated when the value of the attribute is set using CSS? I have a . custom type with a constructor defined. When setting the attribute at runtime, the type's &lt;accept&gt; tag is used to trigger the instantiation of the new type instance. When I set the value using CSS, the constructor of the type is called, passing an instance of the type instead of the CSS property value. Is there any documentation online for the type system? - Raju </pre> Raju Bitter 2013-04-16T21:14:07 http://comments.gmane.org/gmane.comp.java.openlaszlo.devel/28801 <pre>I found another bug with the boxmodel mixin and the basefocusview. The basefocusview sets the dimensions of a view ignoring possible borderwidth values &gt; 0, meaning the focus rectangle is smaller than the target view with borders. http://jira.openlaszlo.org/jira/browse/LPP-10352 - Raju </pre> Raju Bitter 2013-04-11T01:09:15 http://comments.gmane.org/gmane.comp.java.openlaszlo.devel/28799 <pre>When I subscribed the list I got this automated message from Laszlo Systems (not CP!): "The mailing lists are hosted by Laszlo Systems, but Laszlo Systems employees participate in the lists as members of the community, essentially on their own time and in the spirit of the collaborative nature of any open source project. In particular, the lists are not an official support mechanism and you should not expect an answer to any particular question, from Laszlo Systems employees or anyone else." Note this sentence: "In particular, the lists are not an official support mechanism and you should not expect an answer to any particular question, from Laszlo Systems employees or anyone else." The welcome message needs to be updated, unless CP is trying to hide the fact that OpenLaszlo is a Critical Path product right now. It looks like the Critical Path Webtop system is a dead product. They try to create a nice marketing message: http://www.cp.net/index.php/news-2/280-news-radicati-top-player "The Radicati Group names Critical Path ‘Top Player’ ... Messaging platforms are robust, highly scalable, brandable, and localizable offerings." How robust and stable is your messaging platform if you don't maintain one of the key platforms you are using for building highy-priced products. Critical Path has a bad reputation anyway, haven't heard anything good about the company from people in the Bay Area and people working for one of the other companies they acquired. I wouldn't be surprised if all former Laszlo employees have left the company by now. Stevie </pre> Steven Cho 2013-04-10T02:03:21 http://comments.gmane.org/gmane.comp.java.openlaszlo.devel/28798 <pre>The number of serious bugs with the view system, boxmodel, corner radius. drop shadow, clipping of views, click masking of views is concerning. I've never seen the OpenLaszlo platform in such a state in the past 8 1/2 years. Is there anyone within the management at CP aware of those bugs, and are those bugs going to be addressed? If not, the platform will become be unusable very soon. - Raju </pre> Raju Bitter 2013-04-09T14:34:11 http://comments.gmane.org/gmane.comp.java.openlaszlo.devel/28797 <pre>In DHTML runtime, getMouse() adds the borderwith values to the x/y position of the mouse pointer. containsPt() uses those incorrect values and is broken as well. http://jira.openlaszlo.org/jira/browse/LPP-10351 I'll assign the issue to Amy, since Max does not seem to be active any more. - Raju </pre> Raju Bitter 2013-04-09T14:28:18 http://comments.gmane.org/gmane.comp.java.openlaszlo.devel/28795 <pre>There's a but in the LzCanvasGradient class for linear gradients in SWF runtime. The details and a link to a test app to veryify the bug have been added to the issue. - Raju </pre> Raju Bitter 2013-04-07T19:34:35 http://comments.gmane.org/gmane.comp.java.openlaszlo.devel/28793 <pre>Is Critical Path / Laszlo still working on a new component set for OpenLaszlo? Looking at the number of bugs I'm running into when using the drawview component with features like rounded corners and drop shadows it seems that there's no much QA being done at the moment. Are any other developers using these features for their apps? Without some serious bugfixing and QA work drawview and some of the LFC APIs for drop shadows and rounded corners are pretty broken. If we provide patches and bug reports, is there any interest in actually fixing those bugs? And is a 5.0 release realistic for 2013? - Raju </pre> Raju Bitter 2013-04-07T20:52:10 http://comments.gmane.org/gmane.comp.java.openlaszlo.devel/28785 <pre>When unloading resources in a view using the LaszoView#unload() method, shouldn't the &lt; at &gt;resource value be null? It's set to null on the sprite object (view.sprite.resource is null after unload), but the sprite's owner property remains set to the old value. Looks like a bug to me. - Raju </pre> Raju Bitter 2013-03-26T18:09:26 http://comments.gmane.org/gmane.comp.java.openlaszlo.devel/28780 <pre>It seems that the only people responding to questions in this list are Amy and Phil. Is anyone else still reading messages sent to this list? Thanks, Raju </pre> Raju Bitter 2013-03-26T13:37:34 http://comments.gmane.org/gmane.comp.java.openlaszlo.devel/28778 <pre>While clipping of views with rounded corners works in SWF10, it's not working in DHTML runtime. Has anyone found a workaround for that? - Raju </pre> Raju Bitter 2013-03-26T12:17:39 http://comments.gmane.org/gmane.comp.java.openlaszlo.devel/28777 <pre>Views with a borderwidth &gt; 0 are not clipped correctly. The clipping is applied based on the width of the view WITHOUT borders. Is the OL team aware of that bug? - Raju </pre> Raju Bitter 2013-03-26T11:48:11 http://comments.gmane.org/gmane.comp.java.openlaszlo.devel/28776 <pre>http://jira.openlaszlo.org/jira/browse/LPP-9457 Does anyone still know what algorithm Andre was using for the corner radius? The only optimization which should be done at the moment is to use a lower radius value for the inner corner, when a borderwidth is a set on a view with a cornerradius &gt; 0. - Raju </pre> Raju Bitter 2013-03-26T11:37:40 http://comments.gmane.org/gmane.comp.java.openlaszlo.devel/28775 <pre>I guess that http://jira.openlaszlo.org/jira/browse/LPP-1840 can be closed, since there's no need for the roundrect class any more with cornerradius support added to the LFC. - Raju </pre> Raju Bitter 2013-03-26T11:20:49 http://comments.gmane.org/gmane.comp.java.openlaszlo.devel/28774 <pre> (Note: I am checking this changeset in.) Change philip-20130325-nv2 by philip&lt; at &gt;philip-i7-2 on 2013-03-25 12:12:41 EDT     in /cygdrive/f/clients/laszlo/svn/openlaszlo/trunk     for http://svn.openlaszlo.org/openlaszlo/trunk Summary: Alternate fix for shift-click contextmenu on the Mac New Features: Bugs Fixed: LPP-10331 Technical Reviewer: (pending) QA Reviewer: dju Doc Reviewer: (pending) Documentation: Release Notes: Overview: Details: My previous fix turned off the onmouseup event to prevent the context menu from closing. This fix moved the check inside the context menu to prevent the menu from closing when the onmouseup event is received. Tests: See jira Files: M       WEB-INF/lps/lfc/kernel/dhtml/LzMouseKernel.js M       WEB-INF/lps/lfc/contextmenu.js M       WEB-INF/lps/lfc/contextmenu.lzx Changeset: http://svn.openlaszlo.org/openlaszlo/patches/philip-20130325-nv2.tar </pre> promanik< at >laszlosystems.com 2013-03-25T16:17:56 http://comments.gmane.org/gmane.comp.java.openlaszlo.devel/28772 <pre>http://jira.openlaszlo.org/jira/browse/LPP-9969 is a duplicate of http://jira.openlaszlo.org/jira/browse/LPP-10195, which has been fixed by Phil last week. - Raju </pre> Raju Bitter 2013-03-25T11:45:45 http://comments.gmane.org/gmane.comp.java.openlaszlo.devel/28771 <pre>I'm doing some heavy testing of the drawview/boxmodel classes with shadows, borders, view clipping, etc. I've run into a number of bugs when I'm using the setAttribute() API to set those attributes at runtime. I've created a JIRA issue as a container for all these bugs: http://jira.openlaszlo.org/jira/browse/LPP-10336 The individual entries I've created so far are: http://jira.openlaszlo.org/jira/browse/LPP-10333 http://jira.openlaszlo.org/jira/browse/LPP-10337 http://jira.openlaszlo.org/jira/browse/LPP-10338 I'm going to create another 3-4 issues tomorrow. It would be good to fix these issues before the 5.0 release. - Raju </pre> Raju Bitter 2013-03-24T23:47:13 http://comments.gmane.org/gmane.comp.java.openlaszlo.devel/28770 <pre> (Note: I am checking this changeset in.) Change philip-20130322-aGu by philip&lt; at &gt;philip-i7-2 on 2013-03-22 16:07:26 EDT     in /cygdrive/f/clients/laszlo/svn/openlaszlo/trunk     for http://svn.openlaszlo.org/openlaszlo/trunk Summary: Rebuild dev console without old debugging statements New Features: Bugs Fixed: LPP-10332 Technical Reviewer: (pending) QA Reviewer: dju Doc Reviewer: (pending) Documentation: Release Notes: Overview: Details: I removed a call to the new setTimeout method inside of embednew.js because it can load before the file that defines it. I rebuilt the dev console again. A diff shows I had some debugging statements running inside the dev console and this is probably why it would not always work. Tests: Run any application in dhtml with and without debugging. Files: M       lps/includes/source/embednew.js M       lps/admin/dev-console.lzx.js M       lps/admin/dev-console.lzx.swf M       lps/admin/lps/includes/lfc/LFCdhtml.js Changeset: http://svn.openlaszlo.org/openlaszlo/patches/philip-20130322-aGu.tar </pre> promanik< at >laszlosystems.com 2013-03-22T20:13:26 http://comments.gmane.org/gmane.comp.java.openlaszlo.devel/28769 <pre>I found another bug related to the shadow/border functionality in the SWF10/AS3 runtime: http://jira.openlaszlo.org/jira/browse/LPP-10333 Looks like no one ever checked if destroying a view with shadows and borders works in the SWF10 runtime. If http://jira.openlaszlo.org/jira/browse/LPP-10330 is fixed, this should be fixed as well. - Raju </pre> Raju Bitter 2013-03-22T12:11:23 http://comments.gmane.org/gmane.comp.java.openlaszlo.devel/28768 <pre> (Note: I am checking this changeset in.) Change philip-20130321-i1e by philip&lt; at &gt;philip-i7-2 on 2013-03-21 09:31:04 EDT     in /cygdrive/f/clients/laszlo/svn/openlaszlo/trunk     for http://svn.openlaszlo.org/openlaszlo/trunk Summary: Fix context menu with ctrl+click New Features: Bugs Fixed: LPP-10331 Technical Reviewer: (pending) QA Reviewer: dju Doc Reviewer: (pending) Documentation: Release Notes: Overview: Details: OL detects ctrl+click and opens the context menu. However, the onmouseup event that follows will close it. I added a flag to ignore the onmouseup event when the context menu is opened on a Mac. Normal context menu opening with a right mouse click is not affected. Tests: Load any application on OSX and try ctrl+click. The context menu should open and stay open. Files: M       WEB-INF/lps/lfc/kernel/dhtml/LzMouseKernel.js Changeset: http://svn.openlaszlo.org/openlaszlo/patches/philip-20130321-i1e.tar </pre> promanik< at >laszlosystems.com 2013-03-21T13:37:34 http://comments.gmane.org/gmane.comp.java.openlaszlo.devel/28767 <pre> (Note: I am checking this changeset in.) Change philip-20130321-8wW by philip&lt; at &gt;philip-i7-2 on 2013-03-21 08:57:03 EDT     in /cygdrive/f/clients/laszlo/svn/openlaszlo/trunk     for http://svn.openlaszlo.org/openlaszlo/trunk Summary:  Rebuild dev console New Features: Bugs Fixed: LPP-10332 Technical Reviewer: (pending) QA Reviewer: dju Doc Reviewer: (pending) Documentation: Release Notes: Overview: Details: I did find one file that I did not update, but I'm not sure if it would resolve the issue. I rebuilt the dev console. Tests: Make sure your cache is clean and load any application with the debugger enabled. Files: M       lps/includes/source/embednew.js M       lps/admin/dev-console.lzx.swf M       lps/admin/lps/includes/lfc/LFCdhtml.js M       lps/admin/dev-console.lzx.js Changeset: http://svn.openlaszlo.org/openlaszlo/patches/philip-20130321-8wW.tar </pre> promanik< at >laszlosystems.com 2013-03-21T13:18:12 Search the mailing list at Gmane query http://search.gmane.org/?group=$group=gmane.comp.java.openlaszlo.devel