On Thu, Feb 18, 2021 at 08:24:08AM +0100, Ralf Hemmecke wrote:
> > There were plans in early period of open source Axiom. AFAIK there
> > are no visible results. There would be some benefit if one could
> > compile algebra using Aldor. Relatively recently I did a little
> > experiment and it seems that there is a lot of use of few constructs
> > that are valid and perfectly good Spad but which are rejected by
> > Aldor. IMO if one seriously thinks about using Aldor to compile
> > FriCAS algebra one first needs to modify Aldor to accept those
> > constructs.
>
> Waldek, can you give a list of those constructs so that we can show them
> publicly in "Differences between Aldor and SPAD"?
See attached file and Aldor compilation report (I sent it before
but I include it in case it got lost).
>
> > Also, on files that I tried Aldor compiler was about 10 times slower
> > than Spad compiler. For me Spad compiler is slow and 10 times
> > slowdown would be step back.
>
> That is one issue. But honestly, it is more important to me if the
> compiled code runs faster and is safer.
Yes, safer code is among expected benefits. Concerning speed:
AFAIK there is no inlining between Aldor and Spad code. So
if you use Spad types from Aldor (via interface) all operations
would go via function calls and consequently almost all Aldor
optimizations will be ineffective. So you will get better
speed only when all time-critical parts will be pure Aldor.
IIUC in your interface few basic Spad types are entirely replaced
by low Aldor types, but my impression is that Spad compiler
can profitably inline more than that.
--
Waldek Hebisch
--
You received this message because you are subscribed to the Google Groups
"FriCAS - computer algebra system" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to fricas-devel+unsubscr...@googlegroups.com.
To view this discussion on the web visit
https://groups.google.com/d/msgid/fricas-devel/20210218173657.GA12410%40math.uni.wroc.pl.
