kicad-developers team mailing list archive

[Brian Sidebotham <brian.sidebotham@xxxxxxxxx>]

On 3 September 2014 17:03, Maciej Sumiński <maciej.suminski@xxxxxxx> wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
>> On Wed, Sep 03, 2014 at 10:22:03AM +0200, Maciej Sumiński wrote:
>>> Healing could be done another way. Instead of moving vertices,
>>> it could be better to add the missing segments. If they are no
>>> longer than a line thickness, then it is not noticeable to bare
>>> eye and should fix the manufacturing problem. User should be
>>> notified of such changes, so he could verify them.
>
> Jean-Pierre,
>
> What do you think about it? Maybe it is better to connect edges if
> there is a break smaller than the thickness of segments instead of
> simply accepting them as they are?
>
> Regards,
> Orson

I think we could do it in a similar fashion to EESCHEMA (it's a more
pleasant experience when things work in a similar way) Any vertexes on
the Edge.Cuts layer that do not match *exactly* should be highlighted
with a square centred on the disconnected vertex(es).

Secondly, we can apply similar logic to pads being magnetic to tracks
by having Edge.Cuts vertexes being able to snap to currently off-grid
Edge.Cuts vertexes as well as to the grid.

This way anyone can guarantee they have at least connected all of
their Edge.Cuts drawings. Secondly, part of the DRC check can also
list disconnected Edge.Cuts vertexes - because that should at least be
a warning, if not an error.

I don't think there's any reason to have disconnected vertexes on the
Edge.Cuts layer is there? Perhaps for drawing a rout line, but then I
draw a rectangle with the correct width of the router bit I want used
so it is less ambiguous.

Best Regards,

Brian.