ignoring a process' last words: SIGPIPE ...

[Marcin ‘Qrczak’ Kowalczyk]

Dnia 28-03-2008, Pt o godzinie 01:58 +0000, Michael Meeks pisze:

> Clearly if we used a single Watch (& hence poll record) for both read &
> write the code would be cleaner, simpler & more efficient wrt. handling
> this case ( and less invasive: down to ~1 new line instead of ~30).

But then it would be impossible to enable and disable both ends
independently.

Or is it valid to change watch flags during its life?

-- 
   __("<         Marcin Kowalczyk
   \__/       qrczak at knm.org.pl
    ^^     http://qrnik.knm.org.pl/~qrczak/