nsPluginStreamListener doesn't necessarily pass data to the plugin. The implementation requires that OnProgress be called before OnDataAvailable since OnProgress sets up the peer. However, OnProgress gets called before OnStartBinding, and it's during OnStartBinding that the nsPluginStreamListener gets created. So the nsPluginStreamListener never does get the OnProgress call. My solution was to always set up the peer in nsPluginStreamListener::OnStartBinding: <! if ((PR_TRUE == mGotProgress) && (nsnull == mPeer) && !> // RC: possible that OnProgress happened before listener is set up !> // in that case, if we don't set up now, we'll get OnDataAvailable with no peer set to receive it !> if (/*(PR_TRUE == mGotProgress) &&*/ (nsnull == mPeer) &&

We're actually in the midst of making a big change to how streams are done in the new Plugin API - it involves communicating with the plugin entirely through a StreamListener type of interface. It's also a bug in netlib that OnProgress() gets called before OnStartBinding(). I'll report it to the netlib folks. But meanwhile, I'll test and hopefully check in your fix soon. Thanks!

Setting all current Open/Normal to M4.

per leger, assigning QA contacts to all open bugs without QA contacts according to list at http://bugzilla.mozilla.org/describecomponents.cgi?product=Browser

reassigning Greg Lynn as QA contact

QA assigned set to whitebox, gerardok.

This should be fixed in the latest build.

cohn, I don't know how to reproduce the failure caused by this bug. Please can you mark it as verified or provide a testcase to verify it? Thanks!

I'm not in a position right now where I can update to the latest build, and I probably won't be for about two weeks. And I don't have a simple test case, just my plugin that I can't release yet. I will verify the fix when I'm able to update to the current build.

Alex, please could you verify this bug or provide a testcase to verify it. The bug description does not specify how to reproduce the application failure. Thanks!

Sorry for the delay, but I was finally able to verify that this problem is still there. The same fix works. OnProgress is still getting called before OnStartBinding. We now have a pre-alpha version of our plug-in that we can hand out under NDA if that's of interest to you.

this is a M4 bug which was re-opened and we are now on M7. Moving to M7.

The Necko landed since the last comment. Could you please verify it still there and provide with a test case?

This code is no longer relevant. Marking invalid.