It looks like you're new here. If you want to get involved, click one of these buttons!
Hi,
I am running a DRC code checking for a separation between two shapes on two different layers:
lay1.separation(lay2, 4.45.um, projection).output("lay1 separation to lay2", "lay1 separation to lay2 - 4.45um")
I get an error marker with and the following edge pairs:
lay1 separation to lay2 - 4.45um
edge-pair: (700,-0.7;699.998,-0.7)/(700,-5.15;700.004,-5.15)
However, (-0.7) - (-5.15) = 4.45 and so I am not sure why these edges are flagging.
Any help would be appreciated.
Thanks,
Vikas
Comments
Hi Vikas,
they don't event project ... I tried to reproduce the issue with two rectangles (0,-0.7,700,0) and (700,-6,1400,-5.15), but I could not find a way to get this error.
Can you give the two polygons involved in this check?
Thanks,
Matthias
Hi Matthias,
I can place the GDS and code on Dropbox. Is there any way to send you the link privately or do I need to post it in the forum?
Thanks,
Vikas
Hi Vikas,
you can send GDS and/or code to the mail address listed on the "Contact" page. Another channel is to [submit a GitHub issue here: https://github.com/klayoutmatthias/klayout/issues.
Please don't send confidential data. Although I won't disclose a test case if you ask me to, I can't give warranties beyond a promise.
In doubt, you can try to reduce the test case to the polygons which allow reproducing the issue.
Thanks,
Matthias
Thanks Matthias. Let put together a package for you.
Hi Vikas,
thanks for the test data. I have an explanation. I have tried to provide an image for a better description:
The image shows an exaggerated view of the situation. The arrows indicate that there are actually positions where the distance is less than 4.45. Even if the edges are 4.45 apart vertically, their projected distance is less. I little, but sufficient to create this marker.
There are actually three interactions to consider:
So the situation is a kind of border case. In a strict mathematical sense, the whole right side should be flagged as a violation. Practically, the results in such a case are difficult to predict. But my conclusion is that the violation is not entirely off-topic.
Kind regards,
Matthias