[FFmpeg-soc] [soc]: r5877 - in mms: mms.c mms.h mmsh.c mmst.c

[Ronald S. Bultje]

Hi,

On Mon, Aug 2, 2010 at 5:12 AM, Benoit Fouet <benoit.fouet at free.fr> wrote:
> On Mon, 2 Aug 2010 16:17:14 +0800 zhentan feng wrote:
>> Hi
>>
>> On Mon, Aug 2, 2010 at 4:02 PM, Benoit Fouet <benoit.fouet at free.fr> wrote:
>>
>> > On Sun, 1 Aug 2010 14:22:22 -0400 Ronald S. Bultje wrote:
>> > > Hi,
>> > >
>> > > On Sun, Aug 1, 2010 at 3:51 AM, spyfeng <subversion at mplayerhq.hu> wrote:
>> > > > Modified: mms/mms.c
>> > > >
>> > ==============================================================================
>> > > [..]
>> > > > +int ff_read_data(MMSContext *mms, uint8_t *buf, const int buf_size)
>> > >
>> > > this will be a public symbol, please make it a little more specific,
>> > > ff_mms_read_some_kind_of_data(). Same for the other function (make it
>> > > ff_mms_asf_header_parser()).
>> > >
>> >
>> > BTW, what is the purpose of making an 'int' parameter 'const'?
>> >
>>
>> hmm, usually, const will be used for a pointer.
>> here, It means buf_size is read only.
>
> IMO, this is useless. I don't see any situation where it will be of
> use ; modifying buf_size would only modify it for the scope of this
> function anyway.

We discussed it a while ago, and although we all agree it doesn't do
anything, i was agreed that if the developer implementing it wants it,
we can keep it. Therefore, I will keep it, since Zhentan is the
developer and he apparently prefers it.

Ronald