Well, the databuffer behavior is inconsistent across implementations
when it comes to WaitForDataXXX.
The original memory DB returns EOF when it waits for more data than it
has left.
The streaming DB returns OK regardless of how much data it gets when it
actually has to wait for data (if it reaches EOF while waiting). This is
a bug.
The original problem that the patch is meant to solve is the way the
JPEG image provider reads data. It always waits bor a (rather big) data
chunk, then it reads it all. Some JPEG files however end with an
incomplete buffer, even though they are perfectly valid - and they fail
to be decoded.
Sorin
On 05/23/2010 06:35 PM, Niels Roest wrote:
Hm... why would that be an error?
from docu:
This method blocks until at least the specified number of bytes is
available.
so seems that current implementation is correct behavior?
Greets
Niels
Vasi Floroiu wrote:
Hi,
I found a problem about improperly getting the end of file error
while loading a jpeg file into memory buffer.
The problem appeared when I tried to take less than buffer length
into memory buffer.
I attached the patch,
Thanks,
Vasile FLOROIU
------------------------------------------------------------------------
_______________________________________________
directfb-dev mailing list
directfb-dev@directfb.org
http://mail.directfb.org/cgi-bin/mailman/listinfo/directfb-dev
_______________________________________________
directfb-dev mailing list
directfb-dev@directfb.org
http://mail.directfb.org/cgi-bin/mailman/listinfo/directfb-dev
