> < ^ Date: Mon, 01 Mar 1993 21:57:07 +0100
> < ^ From: Martin Schoenert <martin.schoenert@math.rwth-aachen.de >
< ^ Subject: Re: porting GAP to OS/2 2.0 with emx

In his e-mail message of 1993/03/01, Harald Boegeholz writes:
Sinc I just joined the GAP-forum, I don't know about "the
kbhit()-problem". Could you tell me what that was?

In the DOS port the function 'kbhit()' was used by GAP to poll whether
the user had entered <ctr>-'C'. Since this meant switching from
protected mode to real mode and back, it took rather long, and since it
was called quite frequently, it reduced GAP's performance significantly.

How does the 'emx' version test whether the user entered <ctr>-'C'?
Since you write that 'emx' simulates a UNIX environment, does the
extender catch the keyboard interrupt and deliver it to GAP as a signal,
i.e., is 'SyAnsIntr' called if the user enters <ctr>-'C'?

I suggest that you upload your executable to

'samson.math.rwth-aachen.de:pub/tmp'

once you feel it is ready. Then we can all experiment whether it runs
not at all, slower, as fast, or faster as the current DOS version. Write
me and I will set the permissions of 'pub/tmp' accordingly.

Martin.

-- .- .-. - .. -.  .-.. --- ...- . ...  .- -. -. .. -.- .-
Martin Sch"onert,   Martin.Schoenert@Math.RWTH-Aachen.DE,  +49 241 804551
Lehrstuhl D f"ur Mathematik, Templergraben 64, RWTH, D 51 Aachen, Germany

> < [top]