[Ffmpeg-devel] [PATCH] Chinese AVS video decoder

The Wanderer inverseparadox
Tue Jul 4 08:19:49 CEST 2006


Roman Shaposhnik wrote:

> On Mon, 2006-07-03 at 16:51 -0400, The Wanderer wrote:

>> (I've been Googling for a copy of the C standard
> 
> http://dkuug.dk/JTC1/SC22/WG14/www/docs/n869/n869.pdf.gz

Thank you; this will presumably be quite useful for future reference.

>> If char is 16-bit, then either the statement which prompted my
>> original question was wrong (sizeof(char) would have to return 2,
>> not 1), or sizeof(foo) does not necessarily return the number of
>> bytes necessary to store foo - which I thought was the entire point
>> of sizeof in the first place.
> 
> I believe it is OK for char to be 16-bit. In that case, sizeof(char)
> == 1 (as per 6.5.3.4 para 3). I think the trouble you have with
> understanding the semantics of sizeof stems from the fact that you
> always assume bytes to be 8bits in size. Which is not true as far as
> C standard is concerned.

Ah. That would, indeed, be what was tripping me up; I had taken it as a
fundamental assumption underlying all modern computer systems that "one
byte is 8 bits", period. (Certainly attempting to cleanly handle
interactions between systems that define the width of a byte differently
looks, at a glance, like it would be a massive headache...)

> See a definition of byte on a page 7(17) para 3.4:

<snip>

Thank you. That resolves my confusion. I now return you to your
regularly scheduled on-topic list traffic.

-- 
       The Wanderer

Warning: Simply because I argue an issue does not mean I agree with any
side of it.

Secrecy is the beginning of tyranny.




More information about the ffmpeg-devel mailing list