Closed GoogleCodeExporter closed 9 years ago
Is it possible your openvpn ends a line with \r\n or somehow includes \r\n
somewhere
in its output?
In this case the statement lnr.read() would probably read two characters and
report
them as \n. This would invalidate the mark.
I changed the readLimit in lnr.mark to 2.
Fix is in repo, please test.
Original comment by friedrich.schaeuffelhut
on 8 Mar 2010 at 1:51
Actually openvpn terminates each line on the mgmt interface with \r\n.
If for some reason the call to lnr.read hits a \r\n the mark would become
invalid.
This would most likely happen if an empty line was reported by openvpn on the
mgmt
interface.
Original comment by friedrich.schaeuffelhut
on 8 Mar 2010 at 2:07
Actually openvpn terminates each line on the mgmt interface with \r\n.
If for some reason the call to lnr.read hits a \r\n the mark would become
invalid.
This would most likely happen if an empty line was reported by openvpn on the
mgmt
interface.
Original comment by friedrich.schaeuffelhut
on 8 Mar 2010 at 2:07
Seems to work correctly again.
Thanks
Original comment by christophevandeplas@gmail.com
on 8 Mar 2010 at 5:56
Original issue reported on code.google.com by
christophevandeplas@gmail.com
on 7 Mar 2010 at 10:07