The VM is the ultimate judge of what is correct. If it runs correctly, then the problem is with SWFDump.
-Darrell On Thu, Sep 19, 2013 at 4:14 PM, Alex Harui <aha...@adobe.com> wrote: > I'm using the new compiler's SWFDump. I thought about using MXMLC's > SWFDump, but it doesn't handle LZMA so I'd have to switch to uncompressed. > > When I run this SWF, I get a Verify Error about ABC corruption. When I > first tried SWFDump, it blew up on Vector.<Vector.<String>> in trying to > resolve the "Forward reference" when parsing the MultiName table in > ABCParser > > I'd never seen forward references in the MultiName table before. A look > at MXMLC's swfdump implies that it shouldn't work. So I'm wondering if > the player is also blowing up on this or whether it is allowed. > Otherwise, ABCEmitter should take the time to stick things in the pool in > a different order so there are no forward references. > > Thoughts? > -Alex > > On 9/19/13 12:11 PM, "Darrell Loverin" <darrell.love...@gmail.com> wrote: > > >Checkout ABCEmitter.java in falcon if you haven't already. > > > > > >If you are seeing an issue with the old compiler's SWFDump, then I believe > >that was a bug in SWFDump not in falcon's generated code. > > > > > >-Darrell > > > > > >On Thu, Sep 19, 2013 at 2:21 PM, Darrell Loverin > ><darrell.love...@gmail.com>wrote: > > > >> I don't know of a spec besides the AVM2 spec. I'd also look at the > >> swfdumps of code generated by falcon and mxmlc. > >> > >> > >> On Thu, Sep 19, 2013 at 1:38 PM, Alex Harui <aha...@adobe.com> wrote: > >> > >>> Darrell, > >>> > >>> Have you (or anybody) seen a spec for how a Vector ends up in the > >>> Multiname pool? > >>> > >>> Thanks, > >>> -Alex > >>> > >>> > >> > >
