Hi David,
I have put trace and stop control on my TODO list (aka.
README-4-compliance).
May take a while, though.
/// Jürgen
On 04/06/2014 09:56 AM, David B. Lamkins wrote:
Oops. Never mind. That's an IBM-ism; not part of ISO APL.
On Sat, 2014-04-05 at 23:20 -0700, David B. Lamkins wrote:
Hi Elias,
yes, please send the patch.
/// Jürgen
On 04/06/2014 06:12 PM, Elias Mårtenson wrote:
The current mechanism to load native libraries first attempts to find
a library with the suffix .so. If that fails, it will try again with
the suffix .dylib (the OSX library extension).
This pos
The current mechanism to load native libraries first attempts to find a
library with the suffix .so. If that fails, it will try again with the
suffix .dylib (the OSX library extension).
This poses a problem, since if the .so library failed for any reason other
than the file not existing, the error
Hi,
the current solution seems to be (master == thread-0):
*for (int c = 1, c < core_count; ++c) thread-0 waits for thread-c*
One could instead do something this:
*for (int dc = 1; dc < core_count); dc += dx)
{
parallel(
thread-n waits for thread-n+d
What part of the join should be parallel? The join itself is essentially
the main thread waiting for all other threads to finish. What is it that
can be parallelised?
Regards,
Elias
On 6 April 2014 22:32, Juergen Sauermann wrote:
> Hi,
>
> one more plot that might explain a lot. I have plotted
Hi,
one more plot that might explain a lot. I have plotted the startup times
and the total times
vs. the number of cores (1024÷1024 array).
For small core counts (i.e. < 6...10), the startup time is moderate and
the total time decreases rapidly.
For more cores, the total time increases agai
Hello,
as another preparation for parallel execution, I have rewritten
SkalarFunction::eval_AB() so that it is not recursive anymore (SVN 188).
That should allow more equal distribution of work for nested values.
Next step is to make all Cell functions nothrow.
/// Jürgen
Oops. Never mind. That's an IBM-ism; not part of ISO APL.
On Sat, 2014-04-05 at 23:20 -0700, David B. Lamkins wrote:
> A suggestion:
>
> For compatibility with APL2, reserve names beginning with t∆ and s∆. GNU
> APL should indicate that trace and stop control are unimplemented when
> attempting t
