Mail Archives: geda-user/2016/07/28/03:27:06
On Thu, 28 Jul 2016, Brane2 (brane2 AT s5tehnika DOT net) [via geda-user AT delorie DOT com] wrote:
> On 25. 07. 2016 04:18, gedau AT igor2 DOT repo DOT hu wrote:
>>
>>
>> On Mon, 25 Jul 2016, Brane2 (brane2 AT s5tehnika DOT net) [via
>> geda-user AT delorie DOT com] wrote:
>>
>>> On 25. 07. 2016 02:54, Brane2 (brane2 AT s5tehnika DOT net) [via
>>> geda-user AT delorie DOT com] wrote:
>>>
>>> <SNIP>
>>>> Also, If i start drawing hole over several polygons that partially cover,
>>>> like top and bottom layer, for some reason hole will stick to polygon on
>>>> top layer, even if I happen to be on the bottom layer at that time.
>>> Errm, scratch that last part. Now this works.
>>>
>>> But I was definitely able to f**k this up once or twice and triggered
>>> polygon copy to some obscure layer. It seems that it is possible to screw
>>> up operation internal state for all future holes by lucky selection of
>>> first hole. I did it once, but can't repeat it now.
>>
>> I have high hopes for r2309 fixing this too: if the first click's search
>> didn't result in any polygon (user clicked far away from polygons) the code
>> went into invalid state: hole drawing activated with no polygon chosen.
> Close, but no cigar.
> Now I can click first point outside polygon,
This shouldn't happen with r2309 in place. It should not allow you to
start a hole not clicking in the body of on any of the visible polygons.
If you are sure you are using r2309 or later, we need to check what the
poly search finds on the first click.
> but as long
> as any subsequent hole vertex lands in it, with final vertex hole gets created
> on the polygon, as it should.
> Problem is, thing goes berserk if none of the hole vertices actually hit the
> meat of the polygone. Eg. when you draw one hole and after that another,
> within first one.
> Or e.g put 4 vias inside poly. Then try to make a hole in poly, that has
> vertices inside via clearances or vias themselves. This situation usually
> arises when you have e.g poly over dip connector and you want to get rid of
> many small, unconnected islands. If you draw a hole over such place, chances
> are all its vertices will hit pins and not enclosing poly and thing can do
> anything from deleting whole poly, to drawing another one on some obscure
> layer ( white silk) or killing itself.
>
> This happens only when placing the hole and program searches for the poly to
> apply it to. Once you place the hole, you can place whatever over its
> vertices, it will be fine.
Tried this, it worked for me with pcb-rnd HEAD. Which rev are you using?
In HEAD: if I click to start the hole between the pins, where there's "no
polygon pixels" left because of the clearance, it doesn't even start
drawing the hole. Tried it also with very narrow poly bridges between pins
and if I managed to click them with a fine-enough grid, it worked.
Regards,
Igor2
- Raw text -