this is valid only for different releases of kicad (i.e. 5 vs 6) … there is no change in kicad internal format in a stable release cycle.
only the user can decide the ‘risk’ compared to the ‘advantage’ … moreover backups or git are there for a reason
When KiCAD started to gain some attention I remember that many people were playing exactly the same ‘risk’ attitude versus an open source sw compared to a commercial ‘trusted’ ECAD.
In general I prefer open source vs commercial closed source… in open source, what I miss I can improve
I just wonted to show what kind of problems I am thinking about, but the problem can arise also if there is no change in format. Are you sure (and can I be sure) that your loading/saving don’t looses something no one uses except me. For example I have added to only few pads in only few my footprints the Solder paste clearance parameter. I think that parameter is not used in any KiCad library footprint but for those my footprints is crucial. DRC will not notice if that parameters will be changed back to 0 and I am sure ‘my footprints are proved to be good’ so I don’t check it myself - I even don’t remember in what footprints I had to add this parameter. That way it can go to production not being noticed.
Will not help in situation I described above. I am afraid of ‘not noticed’ problems and not of file corruption I can use backup at once. Even if I will notice the problem in a process of verification of everything (close to finishing pcb) then I will have choices:
get 5 days old backup (when the bug was introduced) and do all work once more,
fix the bugs - I will have to check all footprints I’m not sure if that were they I have added some parameters no one except me uses.
But I rather expect that if such kind of bug would happen I will not notice it and I will get the information from contract manufacturer when all PCBs and paste mask will be already done and he will notice a problem with elements assembly.
I think I was enough good in programming but long time ago. I learned C++ from Borland Turbo C++ 1.0 we bought in 1988 and then from Bjarne Stroustrup book 1991. Those time I dreamed of writing PCB design program. But the train left and I stayed on the platform. Now I am not able to “what I miss I can improve ” as I think I will even have a very big problem to compile the KiCad myself as I just don’t understand how all that is currently organised. I also never tried what git really is
KSU will touch ONLY layers that are mechanical related (i.e. drawing objects, edges) … there is no risk in other layer/settings for the pcb file…
anyway I see you are on the conservative side and I respect that.
Anyway, in my vision, I need to rely on plugins to speed up my work an get features that I miss in KiCAD.
For example I had to use the fantastic @MitjaN replicate layout plugin which saved me many hours of work and many errors in multiple redesign a pcb section… the ‘risk’ (which was very little knowing @MitjaN coding ability and effort) was so little compared to the fast and accurate result I got for my complex pcb, which I produced without any issue.
And I could cite many other examples in which plugins are extremely useful in a professional pcb job.
So, as I already pointed out, anyone has to decide for his/her self… but IMO plugins are just useful and needed for KiCAD.
@maui, The tutorial is great! Thank you. I had not realized the extent to which KiCAD StepUp has progressed over time. It is truly an impressive and useful plugin.
Currently I just don’t know how to run any plugin in KiCad. KSU is the other thing - FC does everything for me to allow me use it
You need to use the the “Add-on Manager” which can be accessed via the tools menu to install plugins. There are a few useful workbench plugin’s available and a few macro’s too.
