I have an idea how to implement async subprocesses with Tramp, and also how to implement the splitting of stdout and stderr together with incremental output production.
We launch an async subprocess by saying "$CMD 1>/tmp/out.$$ 2>/tmp/err.$$". We then use "echo $!" to figure out the process id of that subprocess. We remember all subprocesses currently running. Every couple of seconds, we look whether output has arrived from one of the subprocesses. We do this by getting the size of the file and comparing it with the remembered size of the file. We get the new output in the following manner: using dd, we get the bytes from where we left off last time to the current file size we just obtained. We find out if the process is still alive by checking for the process with ps. This procedure is complex and error-prone, I'm afraid. What do people think? -- ~/.signature _______________________________________________ Tramp-devel mailing list [EMAIL PROTECTED] http://mail.nongnu.org/mailman/listinfo/tramp-devel
