kicad-developers team mailing list archive

["Tim Hanson" <sideskate@...>]

Hi all,

Yes, I guess it is not best for the users to have one tool written in
a completely different language. (but I wanted to learn a new
language!) To ameliorate this, I've added a binary 'release' for
linux to the sourceforge page (
http://sourceforge.net/projects/kicadocaml/ ) in case anyone would
like to try kicadocaml without getting the dependencies. It now
features the ability to move/rotate modules, crossprobe to eeschema &
has a proper realtime-updating minimum-spanning-tree ratnest algorithm
(this was fun to write :) I have a windows computer and may try to
build on that next to get a windows binary... though I'm not so
familiar with this ...

If we do eventually get python support (back) into pcbnew, then I'll
be happy to port the arraying functionality over. On the other hand,
users have been asking for undo, and I think this would be relatively
easy in ocaml as compared to C++.

Tim

>
>
> Hi,
>
> nice tool. Although I am kind of sceptical about the whole Ocaml story :-)
> Although your motivations (from a developer's point of view) may be
> understandable, this is pretty sad from a user's point of view:
>
> Basically, it adds even more dependencies to KiCad - given that most
> average
> users are interested in kicadocamls features, which I think they should be,
> because they do look really interesting. So what I am saying is that this
> increases the entry burden for the average user.
>
> The pros of ocaml you name ("quicker to add features") seem to vanish once
> we
> have python support in KiCad. So maybe that would be a better long-term
> solution.
>
> Jonas
>