malformed overview! batches

Russ Allbery rra at stanford.edu
Tue Jan 23 23:27:48 UTC 2001


Todd Olson <tco2 at cornell.edu> writes:

> Hi Russ

> I have no data my self, ... and perhaps I'm misinterpretting what was
> written in the message that started this thread, but in that message
> Jaye Mathisen <mrcpu at internetcds.com> gave an example that had a line of
> data with what looked like a missing front end .....

My understanding from the previous discussion (and I've not checked the
code, so I could have this wrong) is that INN tries to write to the pipe,
gets a last partial write in, decides the pipe is full, and then starts
buffering to a file by writing out the remainder of that last line that it
tried to write.  If we made it back up and write out the whole line, that
should solve that part of the problem without needing a checksum.

-- 
Russ Allbery (rra at stanford.edu)             <http://www.eyrie.org/~eagle/>



More information about the inn-workers mailing list