On Tuesday, November 27, 2012 11:47:01 AM UTC+1, rew wrote: > > On the other hand, programs like "enfuse" are deterministic. They > should yield exactly the same output given the same input. This > means that "intermittent" problems are a hint that your hardware > is broken. >
This is not necessarily true for multi-threaded programs, like the OpenMP enabled version of enblend. Example: worker thread A may finish before thread B in one run of the program, or later in another run (depending on e.g. the load of the CPU the thread is run on). Now, if the results of both worker threads are merged in a wrong way, this may result in unexpected behavior, but not necessarily always on the same input. Or both threads may for instance use the same variable for both read and write, without making a local copy first. Such an (implementation) error is easy to overlook. The chances of such an error are way higher than the chance of hardware failure. -- Bart -- You received this message because you are subscribed to the Google Groups "Hugin and other free panoramic software" group. A list of frequently asked questions is available at: http://wiki.panotools.org/Hugin_FAQ To post to this group, send email to hugin-ptx@googlegroups.com To unsubscribe from this group, send email to hugin-ptx+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/hugin-ptx