On Fri, Jan 14, 2011 at 6:15 AM, Glenn Knickerbocker <n...@bestweb.net> wrote: > On 01/13/2011 05:08 PM, Rob van der Heij wrote: >> I must be more dumb than usual today, but what's the role of a gate >> that is never triggered? > > It's not *never* triggered; it's triggered when PIPMOD STOP kills > whatever stages are waiting for asynchronous input. It's there to sever > the pipeline streams when the user shuts off the external sources of input.
Right, I missed that. Not sure it was more dumb than usual... Talked to Finn and he explained (and gave credits to the Piper for the solution). He could not get eof propagation to fold up his pipes properly when the asynchronous stages terminate. Rather than computing a large delay that will not overflow the clock, you could probably wait until midnight next day over and over again (and accept a tiny cost) literal 24:00:00 | dup * | delay | literal | take last | g: gate | Rob
