Yep, because everybody seems to come in with the mindset we don’t know what we already are doing.
There’s just a lot to do and we aren’t trying to be open-source Alitum, or open-source PADS or open-source Cadence which some people with strong opinions want to make KiCad. There are also tons of technical reasons we just do and dont do some things currently. It just takes time to evolve and refactor.
Even the computational geometry side of KiCad which is basically everything from the router to just filling polygons is extremely wild ride of algorithms and logic that has to be slowly evolved.
Also to remark on:
there is no critical error analysis system. The program crashed twice and there is not even something like coredump that allows you to recreate the cause of the program crash.
We have sentry data capture on Windows. We get small mini dumps to analyze. Any larger dumps are too dangerous and stray into PII handling territory as such significant dumps may contain too much detail. We don’t want to run afoul of EU or California law on PII. Other than that, full dumps are massive. Users can trigger that themselves in Windows if they really want, shitting up a normal non-programmers disk space with core dumps is otherwise inappropriate.
Linux has no dump functionality because we are currently forbidden by our build platforms and distros generally being averse to it. For example on Ubuntu, Launchpad does not allow us to connect to the internet at all during build. So we can’t upload the kicad debug symbols to sentry for later use. Fedora otoh as a distro actually captures crashes for us through a separae service that’s already operating.
If only there were some sort of forum environment where the developers could interact with the users in an almost real time continuous fashion. I guess we can dream.
Any new program will be a “shock” and have some learning curve, deficiencies/differences will be found quickly.
Shortcuts are configurable, maybe more choices are OK.
Intuitiveness in programs is a first choice paradigm. Reading a manual, that’s for women and accountants, unless defusing a bomb.
First start-up with prompts is a good idea, so is more YouTube videos.
3D viewing of PCB is not (for me) a top priority, knowing boundary dimensions of parts is. Moreover, all my footprints include a documentation layer for physical size, height noted in mm, connectors have an outline of the mated connector and cable. (Finding mated connector dimensions is hopeless, even with step files, to be sure buy relevant connectors and measure to be 100% sure). Not much need for 3D with good footprints.
Designing a board from copper, ouch. So wrong on many levels. Amateur hour. All design starts with schematic.
“Via-flanges”, “collar” you mean Annular Rings for Vias and Pads?
I have not seen any Rules or Setup files files in Ki, Design rules setting for various vendors, (CAM files) or Design Rules files that can be imported. No heavy math here.
I just started with Ki after using another program for 15 years, that may be disappearing.
I’m sure I’ll find out a lot of stuff.
Welcome to Ki.
I moved from Protel 3 to KiCad in 2017. I have never used 3D before. Now I use it only to add photo-like looking pictures to PCB documentation. But when I imagine someone designing cell-phone I think he can’t do without 3D .
I have not KiCad at hand here, but I remember there is a possibility to import Board Rules from other PCB so if you need you can have empty PCB designs for various vendors and when needed import rules from there.
Well for me the only thing i really makes me use kicad is the 3d model facility, for multiphysics simulation and its nice for verifying the mechanical assembly fitment. All other functions are a nice plus. So YMMV
.