Checking in media::DataBuffer, a simple implementation of WritableBufferInterface.

Checking in media::DataBuffer, a simple implementation of WritableBufferInterface.

Removed media/base/media.cc, since it's no longer needed to generate media.lib.

Also added media/using_media.scons and updated scons files for Linux build.

[cpu_(ooo_6.6-7.5), 2008-12-10 21:42:45]

http://codereview.chromium.org/13682/diff/1/2
File media/base/data_buffer.cc (right):

http://codereview.chromium.org/13682/diff/1/2#newcode24
Line 24: }
worth mentioning that DataBuffer assumes memory was allocated with 'new char[x]'

http://codereview.chromium.org/13682/diff/1/3
File media/base/data_buffer.h (right):

http://codereview.chromium.org/13682/diff/1/3#newcode21
Line 21: virtual int64 GetTimestamp();
hmmm, consider making this function const later on.

http://codereview.chromium.org/13682/diff/1/4
File media/base/data_buffer_unittest.cc (right):

http://codereview.chromium.org/13682/diff/1/4#newcode12
Line 12: };
afaik, you can remove this no-op class.

http://codereview.chromium.org/13682/diff/1/4#newcode31
Line 31: 
There are two flavors of testing: with EXPECT_xx and with ASSERT_xx, use the
former in most cases unless the failure is truly fatal in the sense of making
something down the road crash:

ASSERT_TRUE(data)  ok
ASSERT_TRUE(buffer.get())  use EXPECT_EQ

[scherkus (not reviewing), 2008-12-10 23:46:03]

http://codereview.chromium.org/13682/diff/1/2
File media/base/data_buffer.cc (right):

http://codereview.chromium.org/13682/diff/1/2#newcode24
Line 24: }
On 2008/12/10 21:42:45, cpu wrote:
> worth mentioning that DataBuffer assumes memory was allocated with 'new
char[x]'

Done.

http://codereview.chromium.org/13682/diff/1/3
File media/base/data_buffer.h (right):

http://codereview.chromium.org/13682/diff/1/3#newcode21
Line 21: virtual int64 GetTimestamp();
On 2008/12/10 21:42:45, cpu wrote:
> hmmm, consider making this function const later on.

Done.

I added const to all getters in the interface definitions.

http://codereview.chromium.org/13682/diff/1/4
File media/base/data_buffer_unittest.cc (right):

http://codereview.chromium.org/13682/diff/1/4#newcode12
Line 12: };
On 2008/12/10 21:42:45, cpu wrote:
> afaik, you can remove this no-op class.

Done.

http://codereview.chromium.org/13682/diff/1/4#newcode31
Line 31: 
On 2008/12/10 21:42:45, cpu wrote:
> There are two flavors of testing: with EXPECT_xx and with ASSERT_xx, use the
> former in most cases unless the failure is truly fatal in the sense of making
> something down the road crash:
> 
> ASSERT_TRUE(data)  ok
> ASSERT_TRUE(buffer.get())  use EXPECT_EQ
> 
> 
> 
> 

Done.