ID: | 5995 |
From: | rn...@igor2.repo.hu |
Date: | Mon, 13 Mar 2023 12:23:05 +0100 (CET) |
Subject: | Re: [pcb-rnd] Bug: extended objects are considered components for |
in-reply-to: | 5994 from Gabriel Paubert <pa...@iram.es> |
replies: | 5997 from Bdale Garbee <bd...@gag.com> |
Hi Gabriel, On Mon, 13 Mar 2023, Gabriel Paubert wrote: > Hello Igor2, > >On Sat, Mar 11, 2023 at 10:04:30AM +0100, rnd2@igor2.repo.hu wrote: >> Hello Gabriel, >> >> On Sat, 11 Mar 2023, Gabriel Paubert wrote: >> >> > Hi Igor2, >> > >> >since you introduced extended objects in pcb-rnd, I've used them a bit >> >for differential lines and busses. >> > >> >But they have a funny side effect, these extended objects appear in the >> >bom file, in the assembly window (":asm"), and also in the xy file with >> >the following message: >> > >> >xy: can't get subc side for (unknown) >> > >> >I strongly suspect that the bug is that extended objects are included in >> >these lists by mistake. >> > >> >Example lht file attached, with bom and xy. >> >> >> Thank you, the example files were very useful. >> >> I've commit a series of small fixes ending at r37346 to make some of the >> plugins ignore extended objects - typically those that deal with physical >> parts. The 3 you've found are included. >> >> Please test! > >Thank you, this fixes all the issues I've met when exporting .xy and >.bom files, as well as hand assembly. Cool, thanks for reporting! >On the other hand I've got a problem with DRC but I have trouble >distilling a test case: a layout which gives 0 drc errors with an old >version finds hundreds of them with a more recent version. > >I'm still trying to find the cause, it might be a pilot error. >(I don't know how to make regression testing with svn, is there an >equivalent of "git bisect"?) We have a bisect helper script in trunk/util/bisecter but since we have librnd in a separate repo, bisecting became harder because you always need to make sure the right major version of librnd is available and we just had a major version change there recently. Plus in some cases the bug you are after is not in pcb-rnd but in librnd. At the end of the day a test case or a recipe for reproducing the problem is more valuable for me than a revision number where it broke. Being able to reproduce the problem is a must for fixing it and validating the fix. Given a reproduction path, bisecting to the revision where it broke is cheap. Hiven a revision where somehting broke, constructing an example that demonstrates how it breaks is often much harder. Best regards, Igor2
Reply subtree:
5995 Re: [pcb-rnd] Bug: extended objects are considered components for from rn...@igor2.repo.hu
5997 Re: [pcb-rnd] Bug: extended objects are considered components for from Bdale Garbee <bd...@gag.com>
5998 Re: [pcb-rnd] Bug: extended objects are considered components for from rn...@igor2.repo.hu
5999 Re: [pcb-rnd] Bug: extended objects are considered components for from Bdale Garbee <bd...@gag.com>