On Fri, Apr 12, 2024 at 08:41:18AM +0100, Martin Baker wrote:
> I don't mind, I have given up with this software now.
>
> When I wrote the code I was hoping that this approach would replace old
> boot and C code with something much more maintainable and flexible. Also
> the scenegraph architecture makes it much easier to export to 2D and 3D
> graphics files like SVG and X3D and being written in SPAD it would have
> been easy to add other graphics formats.
Hmm, I am not sure what _exactly_ you expected. I have a book about
"object oriented program analysis". It stress that one should take
into account "critical factors for success". For me personally, and
I suspect that for most FriCAS users "critical factors for success"
is ability to use higher level interactive commands and pop up
windows.
I considerd hooking 'draw' to your framework. However, architecturaly
nice way would direct all graphics trough your framework. But at the
end _I_ want image in a window as one of options. Which means appropriate
"output format". Writing convertor from scengraph data structures to
say X11 drawing commands is a bit more than I am willing to do.
To say the truth I do know how much work is needed, I would first
go trough your code and analize what it needs.
Anyway, I hope to use your framework to produce SVG files, but as
a user most of my personal needs are covered by old framework. And
without implementing new featurs your framework can not do all
things that I need.
> I'm not critisising because I don't see the big picture but it seems to
> me that going from boot code to SPAD in lots of small steps rather than
> one big step is impossibly slow and and risks replacing messy boot code
> with messy SPAD code.
'src/interp' subdirectory contains about 62000 lines of code. I expect
replacement to have similar size. If you want to include in rewrite
C code, then considering that we have about 50000 lines of C code
we need 20000-30000 lines of additional code. Considering that FriCAS
is a volunteer part time project writing and debugging that amount of
code will take several years (say 5 years). If you go via big shot
rewrite FriCAS would probably remain broken during rewrite period.
And that is relatively optimistic estimate. Debugging tricky problems
can sink a lot of time. How many developers are going to stay in
a "broken" project? How many users are going to stay with project
that makes no useful release?
Krystian Baclawski attempted rewrite of Spad compiler as new code.
It took him few years and still his compiler is not usable on
algebra code, to finish it would require substantial work.
--
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/ZhnnCMv0NO3h7hpC%40fricas.org.
