Windows default plugin has changed so we need OS/2 changes.
Was this ever checked in? On Seamonkey 1.0a, it appears to me that the stream isn't being killed for the plugin data: NPN_DestroyStream(pInstance, stream, NPRES_DONE); Can't we point these to http://plugindoc.mozdev.org/OS2.html instead? #define DEFAULT_PLUGINFINDER_URL "http://plugins.netscape.com/plug-in_finder.adp" #define JVM_SMARTUPDATE_URL "http://plugins.netscape.com/plug-in_finder.adp"
Nope, I never checked this in because I thought they were going to port the new plugin code to SeaMonkey so this wouldn't be relevant. I'll take a look at the code again.
Created attachment 207847 [details] [diff] [review] sync anew Update of Mike's patch, also fixes some warnings and has two whitespace cleanups. It also changes the two URLs to point to Steve's mozdev page. Compiles on trunk.
Hmm, it just doesn't work. If I go to http://www.dtmb.de/Rundgang/Neubau/body_schifffahrt_Aufbau01.html and double click the jigsaw piece I am asked by the dialog to download the QT plugin from http://www.apple.com/quicktime/download/ and when I click on "Get the Plug-in" it goes to that website. With the new version the same dialog appears but it does not go to that page any more. Hmm, it seems I have to do some debugging (tomorrow or when I get the time).
OK, that's because "application.use_ns_plugin_finder" is false by default, so the patch is OK.
Comment on attachment 207847 [details] [diff] [review] sync anew r=mkaply
Fix checked in.
Checked in on just the trunk, or branches too?
trunk should I put it on the branches?
Yes, please. :)
In principle I agree that this should go into the branches, but I think we should test the trunk version first.
Steve, does the new version (from trunk or in my newest unofficial releases) work the way you expect it to?
(In reply to comment #13) > Steve, does the new version (from trunk or in my newest unofficial releases) > work the way you expect it to? Yes, the streams do stop, thanks!
Comment on attachment 207847 [details] [diff] [review] sync anew OK, so we should get this on the branche(s). This is OS/2 only and was shown on the trunk to work.
Comment on attachment 207847 [details] [diff] [review] sync anew approved for 1.8.0 branch, a=dveditz for drivers
OK, checked into both branches.