Thomas Down wrote > I'm not a Windows user. But one thing which occurs to > me: you're handling the output of the process on the same > thread as the Process itself. I always use separate threads > (I don't know if this is the `official' way of doing things, > but it's a pragmatic approach which always seems to be reliable):
I've written code using Runtime.getRuntime().exec which used the same thread for handling the standard output and standard error. This code worked fine on a linux platform. When I tried it under windows, it failed. I ended up having to create a separate thread for handing the standard output and standard errors. It seemed to be a problem only when the output of either was more than a small number of characters (80 or 100 from memory). Once I rewrote the code to use separate threads, it ran fine on both operating systems. I can dig into my code archive to send a snippet of what it took to get the multiple threads working if required. It will take me a day or two to dig it up, just let me know if you want it. cheers tony _______________________________________________ Biojava-l mailing list - [EMAIL PROTECTED] http://biojava.org/mailman/listinfo/biojava-l
