On Sun, Feb 27, 2011 at 3:43 PM, Michael Wood <esiotrot@gmail.com> wrote:
On 27 February 2011 14:52, Rupert Swarbrick <rswarbrick@gmail.com> wrote:
> Michael Wood <esiotrot@gmail.com> writes:
>> 2011/2/27 Rupert Swarbrick <rswarbrick@gmail.com>:
>>> How should I make sure that the process finishes too? (I don't mind if
>>> I have to wait for it: this doesn't need to run in parallel)
>>
>> Just a guess, but try changing ":wait nil" to ":wait t"?
>
> Surely that won't work? Won't that block and then I don't get to send it
> any data? (I haven't tried this, but I'm not at a computer with ecl atm)

hmmm...  Yes, I suppose you're right.  OK, so it was a bad guess :)

Actually it should have worked, but for some reason it does not. I have tried various things and as a last resort I had to introduce a signal handler for SIGCHLD that updates a new list of external (and active) processes.

Juanjo

--
Instituto de Física Fundamental, CSIC
c/ Serrano, 113b, Madrid 28006 (Spain)
http://juanjose.garciaripoll.googlepages.com