--
You received this message because you are subscribed to the Google Groups "KiCad Developers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to devlist+u...@kicad.org.
To view this discussion visit https://groups.google.com/a/kicad.org/d/msgid/devlist/0b60bf33-cc03-4bac-89f4-95f8aa19946bn%40kicad.org.
To view this discussion visit https://groups.google.com/a/kicad.org/d/msgid/devlist/6a69106c-9717-45e8-bda8-59c1f57d0ce0n%40kicad.org.

--
You received this message because you are subscribed to the Google Groups "KiCad Developers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to devlist+u...@kicad.org.
To view this discussion visit https://groups.google.com/a/kicad.org/d/msgid/devlist/f343dde9-a9c9-491b-a3af-f61b9d48d3f1%40inti.gob.ar.
It was fine for KiCad 6.
For 7, 8, 9 and 10 was always outdated. I had to adapt my code on every single release and never found documentation for the changes.
It doesn't help that a lot of the changes are just cosmetic. Others are just a lack of abstraction in the file format, which seems to reflect some internal details.
Is also a pitty that the change from the ad-hoc format used in 4/5 was done to s-expressions + JSON, instead of just JSON. A lot of inconsistencies, that were then changed, are just a consequence of the format.
I understand KiCad team lacks resources, but I don't agree with the huge priority of additions, and the low priority on consistency and reproducibility. I can't believe the huge amount of bugs introduced in minor releases. Stable releases should remain stable, let the fancy stuff for the unstable branch. I see too much cherry-picking instead of backporting fixes. Yes, it takes more time, which can't be dedicated to the funny new stuff, but this is what makes a project reliable.
And this policy gets reflected in the file format. The lack of documentation makes things even worst.
Perhaps the team should have a policy about changes in the file format without its corresponding documentation, just reject them.
BTW: is too hard to contribute to the code base, MR takes for ever to be incorporated, even some trivial ones. The code is organized in a way that a small change takes hours to be compiled, this makes verifications frustrating. I also think the poor coverage of the tests is one really weak point, the team doesn't have a mechanism to be confident about a change, and this leads to a state where the code is never stable.
Is not my intention to sound harsh, but after years developing plug-ins and contributing to the project I'm really frustrated.