I assumed m_current_timestamp was in milliseconds, but you're right documentation is missing. If milliseconds is confirmed I'd use an integer value for that as well.
--strk; On Wed, May 30, 2007 at 12:41:47PM +0100, Martin Guy wrote: > 2007/5/30, Martin Guy <[EMAIL PROTECTED]>: > >* server/asobj/NetStreamFfmpeg.cpp: Two casts to quiet compiler warnings > > There is another warning which may be a bug: > > server/asobj/NetStreamFfmpeg.cpp:1218: warning: passing 'double' for > argument 1 to 'bool FLVParser::isTimeLoaded(uint32_t)' > > here: > if (m_go && m_pause && m_start_onbuffer && m_parser.get() && > m_parser->isTimeLoaded(m_current_timestamp+m_bufferTime)) > > since isTimeLoaded expects a uint32_t but m_current_timestamp is a double > (m_buffer_time is a uint32_t too). > > Unfortunately I can't easily see whether this is correct or not since > none of the "time" variable declarations say, as they should, what > units the times are measured in (seconds? milliseconds? SWF frames? > ticks of some kind?) and when they are measured from (from start of > flash movie in realtime seconds? from start of video clip in video > time?) > > m_current_timestamp seems to be in seconds to a 1/1000 second accuracy; > the others...? > > Could someone who knows this code better check this statement is right > (and maybe add the missing comments here and there?) > > Cheers > > M > > > _______________________________________________ > Gnash-commit mailing list > [email protected] > http://lists.gnu.org/mailman/listinfo/gnash-commit -- () ASCII Ribbon Campaign /\ Keep it simple! _______________________________________________ Gnash-commit mailing list [email protected] http://lists.gnu.org/mailman/listinfo/gnash-commit
