Hello strk, Tuesday, August 28, 2007, 11:46:48 AM, you wrote: s> I still see it as a facility. The stream will just setup the buffer s> for the BitsReader, knowing size of each tag..
Ok, yes, whatever, but anyway this is a minor detail at the moment. :) s> Note that the users which are NOT willing to cache are tipically users s> that will cache themselves. You talk about JPEG tag loader for example? It could still use the already present get_underlying_stream() function, ie. don't care if they cache or not. Either you use that function or the bit reading functions for a tag. "cache" is not really the appropriate term. We just simplify the bit parsing process by taking advantage of native processor operations instead of loops and checks. s> I'm currently preparing work to expand their interfaces to accept a s> pre-made cache for reading. In that case, the stream (or s> BitsReader) should provide a pointer to the underlying buffer. Uhm, not sure I understand what you mean. Udo _______________________________________________ Gnash-dev mailing list [email protected] http://lists.gnu.org/mailman/listinfo/gnash-dev
