Re: [Gnash-dev] make check results

[Benjamin Wolsey]

> FAIL: trace-test-runner: arguments
> FAIL: trace-test-runner: super
> FAIL: trace-test-runner: this
>  --=[ testsuite/misc-swfmill.all/trace-as3 ]=--
> FAIL: trace-test-runner: stack1
> 

The swfmill version is too old. There should be a version check in
configure, and it should warn at least.

> XMLSocket has problems on some older machines:
> 
> FAIL: Test case XMLSocketTester still running after 600 seconds, killing
> it (deadlock?)

It relies on Perl closing the socket before it finishes to test the
XMLSocket.onClose method. I never found a very reliable way of doing
this, so if anyone can improve the test (it fails to complete
occasionally for me too) it would be helpful.

> None of these other issues is addressed by my checkin of 12197, as this
> was focused on just ExternalInterface. Note that this checkin also
> disabled Gnash from checking for XML messages from the browser. My plan
> was to do something similar as FSCommand to have movie_root execute the
> ExternalInterface::update(), which is what looks for incoming messages
> in a thread safe manner. How ::update() gets into movie_root I'm open to
> suggestions, or I'll follow my current idea of a plan.

I suspect it would make the code much cleaner if you separated the
ExternalInterface API from the AS implementation much more strictly.
This could involve moving the actual ExternalInterface API
implementation into a separate file and attaching the whole thing to
movie_root.

Because ExternalInterface isn't a proper class, using a Relay is
overkill. It will work up to a certain point, but it's more difficult
than it needs to be and there are (admittedly quite weird) test
scenarios that it would fail.

Converting the Relay class into a different type and using it as a
movie_root member would probably be relatively trivial to do (I haven't
looked at the code in much depth).

Then if the class is owned by movie_root anyway, you can simply call an
update function on every advance.

You know the requirements of ExternalInterface better than I do - are
there any problems with this approach?

bwy

--
The current release of Gnash is 0.8.7
http://www.gnu.org/software/gnash/

Benjamin Wolsey, Software Developer - http://benjaminwolsey.de
C++ and Open-Source Flash blog - http://www.benjaminwolsey.de/bwysblog

xmpp:address@hidden
http://identi.ca/bwy

signature.asc
Description: Dies ist ein digital signierter Nachrichtenteil