Hi all…in the latest 0.9.0 release I create a brand new map, and then create a Java applet from that map.
1) There is an erroneous <title/> tag in the HTML header that prevents Chrome from opening up the file (bug submitted, no need to worry about that).
2) The main problem: Chrome (once the header is fixed) is able to open the Java applet once just fine…see here (redactions are mine….don't mind the black boxes):
…but then hitting reload, or navigating away and then back (even in separate tabs) results in squished / compressed nodes…see here:
Anyone know what is going on?
Also, let me add that the only way to fix this is to close down Chrome and start it up again…at which time it works the first time, then has the same behavior as above.
Firefox 4 beta works fine on initial and all reloads…..
Figured out a bit more….the only boxes that are squished are the ones that have had CR / line returns entered into them…..maybe that helps someone figure something out?
About the "main" problem, I can confirm that I have the same problem… The applets generated by FreeMind work fine for me on Fedora 14 Firefox 3.6.17, Windows 7 Firefox 4.0.1, Windows 7 IE8 but not on Fedora 14 Chrome…
Now some new input :)
I am trying to integrate this Freemind-applet in a Content Management System (CMS) named Plone (see http://plone.org), so that all freemind-files in the CMS can be viewed in browser. Therefor, I take the applet code, and change the archive-attribute and the browsemode_initial_map-parameter. I change the parameters so that the refer to a real internet-address e.g. http://some-domain.org/freemindbrowser.jar and http://some-domain.org/data/freemind.mm. When I use this local file template.html, It is the same story as above. But when this applet code is integrated in Plone, and I open multiple freemind-files, the issue of the squished boxes happens when I load my second (different) mindmap in ALL my test-configurations (Fedora 14 Firefox 3.6.17, Windows 7 Firefox 4.0.1, Windows 7 IE8, Fedora 14 Chrome)… I love platform-independent stuff ;)
Has anyone an idea of what is going wrong?
@pendragn: Do you have new input/suggestions?
I forgot to mention that the squished boxes also appear in nodes which don't have carriage returns but which are too long to be displayed on one line. You can see that in the freemind.mm example (child of "Table of key mappings") which is located in freemind-bin-max-0.9.0/browser/freemind.mm
I found a bypass on http://java.sun.com/developer/technicalArticles/javase/newapplets/
It is a additional parameter in the applet-code: separate_jvm
archive="freemindbrowser.jar" width="100%" height="100%">
<param name="type" value="application/x-java-applet;version=1.4">
<param name="scriptable" value="false">
<param name="modes" value="freemind.modes.browsemode.BrowseMode">
<param name="browsemode_initial_map" value="./freemind.mm">
<param name="initial_mode" value="Browse">
<param name="selection_method" value="selection_method_direct">
[b] <param name="separate_jvm" value="true">
This solves the squished nodes, because every (re)load of the applet, a complete new Java Virtual Machine (JVM) is started. Normally, all applets run in the same JVM. This does mean it puts a heavy load (CPU/RAM) on your system. Certainly, if you open multiple applets, multiple separate JVM's are started…
Just re-visited and found this…thanks so much for the assistance…yes that does work well! Maybe something about something not being cleaned up when re-using the original JVM….