On 10/11/15 12:23, Michael Schnell wrote: > On 11/10/2015 01:14 PM, Engelbert Buxbaum wrote: >> >> as far as I can see the only advantage of running the math-routine >> in a thread would be that the program could be terminated at any >> time, rather than only in the main loop. > If a long winding calculation runs in a Thread, the complete main > thread (including the GUI) is not blocked and the program can do any > other stuff (e.g. as a reaction on TTimer events, and also as a > reaction on Events this or any other thread fires via > TThread.Queue). > I'd just add, but,
> Moreover on a multi-CPU system (which is standard nowadays), multiple > threads can do long winding calculations at the same time and finish > by far faster than a single threaded program. > - this, only if the calculation itself is possible to be multithreaded (e.g. formula possible to calculate in chunks/portions/arbitrary bounds etc). > >> >> Even if one did it like that, the original problem remains: who is >> the sender in the FormPaint call? > As any GUI stuff is only allowed in the main thread it can only be > the Main thread (if this is what you are asking) > > -Michael > el es -- _______________________________________________ Lazarus mailing list [email protected] http://lists.lazarus.freepascal.org/mailman/listinfo/lazarus
