http://gwt-code-reviews.appspot.com/1277801/diff/1/3 File plugins/npapi/NPVariantWrapper.h (right):
http://gwt-code-reviews.appspot.com/1277801/diff/1/3#newcode105 plugins/npapi/NPVariantWrapper.h:105: // numerical variants out of V8. On 2011/01/11 21:17:27, jat wrote:
Have we talked to them about this change? That seems like a pretty
big change
that could impact performance and break a lot of plugins, so I would
wonder if
it is worth making this change if they are going to revert it.
I had a quick chat with Tony today. He says this makes them more safari-like and that plugins should be more lenient about types. If you read through the WebKit changelog, they did this in order to fix some other plugins so I think it's probably 6 one way half dozen the other. http://gwt-code-reviews.appspot.com/1277801/diff/1/3#newcode106 plugins/npapi/NPVariantWrapper.h:106: return static_cast<int>(NPVARIANT_TO_DOUBLE(variant)); On 2011/01/11 21:17:27, jat wrote:
Should there be a check that the value actually is an integer? I
don't know if
it is a good idea to silently truncate a value here.
It might be better to have isInt accept doubles that have integral
values
instead, and would require no changes to callers that expect integers.
yeah, patch updated. http://gwt-code-reviews.appspot.com/1277801/diff/1/5 File plugins/npapi/main.cpp (right): http://gwt-code-reviews.appspot.com/1277801/diff/1/5#newcode241 plugins/npapi/main.cpp:241: //Debug::log(Debug::Spam) << "NPP_HandleEvent(instance=" << instance << ")" << Debug::flush; On 2011/01/11 21:17:27, jat wrote:
Why commented out? Spam seems low enough level to leave it on.
this function gets called a ton on the mac build and just creates noise when trying to debug. http://gwt-code-reviews.appspot.com/1277801/show -- http://groups.google.com/group/Google-Web-Toolkit-Contributors