On Tue, Apr 06, 2010 at 11:01:21AM -0400, Christopher Faylor wrote: >On Tue, Apr 06, 2010 at 04:10:28PM +0100, Dave Korn wrote: >>On 06/04/2010 15:15, Christopher Faylor wrote: >>> On Tue, Apr 06, 2010 at 07:55:19AM -0400, George Barrick wrote: >>>> What I am suggesting is that the new cygwin.dll has difficulties >>>> implementing the kind of pipe that octave uses to send its graphing >>>> commands to gnuplot. The specific details of the pipe that serves as >>>> an example of this can be seen in the __gnuplot_get_var__.m script >>>> inside the GNU Octave package. >>> >>> Sorry but unless someone can narrow this down to a test case which >>> does not requier installing gnuplot I don't intend on investigating >>> this. >> >> Here ya go. I took a look at the octave module, and narrowed it down to the >>equivalent C testcase attached, which appears to reproduce the same failure >>mode for me: > >Thanks for tracking it down but the test case is probably simpler than this. > > $ mkfifo /tmp/fifo > $ cat /tmp/fifo > >That demonstrates the problem. I'll fix this ASAP.
This should be fixed in the latest snapshot: http://cygwin.com/snapshots/ cgf -- Problem reports: http://cygwin.com/problems.html FAQ: http://cygwin.com/faq/ Documentation: http://cygwin.com/docs.html Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple
