the tutorial gives me:
“Each symbol needs a unique reference designator assigned to it. This process is also known as annotation.”
yes good, and…
where is the annotation used, for what purpose… (I have some reminder that it connects the symbol and the fotprints, and that mixing creates kaos…)
guess this is trivial…
thanks anyways
Georg
The Symbol is a symbolic representation of the component, the Footprint is the physical representation of the component. If both have the same reference designator you now have a way of representing the schematic diagram on the PCB layout - that’s the ratsnest or netlist.
Now “simply” place tracks to achieve all the connections shown in the ratsnest . . . et voila ! you should now have a PCB that does what is described in your schematic . . . . is your schematic correct ? 
Annotation is for your BOM so you know where to place the components.
Give all your ICs U1 symbol and all your resistors R1 symbol and then try to explain someone (using only words) how your circuit is expected to work.
It is what for annotation is used.
Footprint assigning to symbols is not annotation.
Ideally the RefDes is only for communicating with Humans. From explaining the working of a circuit to PCB repair work where the RefDes is printed on the silkscreen and can be easily referenced to the schematic.
Internally KiCad uses UUID’s to match the symbols with the footprints, but this can be overridden for “unusual workflows”, for example when a Schematic and a PCB is imported form another program and KiCad does not know yet how to match the symbols with the footprints. In that case the RefDes mapping can be used (once!) to create the UUID’s.
In practice it’s a bit more complicated. Because of historical reasons KiCad uses the RefDes in more situations then it should, and as a result the annotation is also required before for example the footprint assignment tool can be used, or before the footprints can be put on the PCB.
Hi ! Thank you all for your respective views.
As I get it the RefDes (Read out Reference Designator ?) is a field stored and used in different places, and also is the target of “annotation”. ? It is an “UUID” (well how universal ? within your projecct?)
The first and prime use seem to be to connect the sym and PCB footprint entities, and also to give the values in the BOM, silkscreen, where to mount what…
It kind of looks likes there would be a need for a refman, where in shortest/clearest possible language, the techinical use and implication of things in the KiCad would be described.
georg
You could just read the KiCad Documentation . . .
Reference designators are unique identifiers for components in a design. They are often printed on a PCB and in assembly diagrams, and allow you to match symbols in a schematic to the corresponding components on a board.
If you don’t want to read the documentation, asking ChatGPT is also a great way to avoid bothering actual people with lazy questions.
Just google for uuid and you can see universal means universal. Covering the whole universe
Ahh sorry. Had not seen that. Just got stuck on tutorial. Guess shoiuld have looked better.
They are randomly generated. However, if you copy-paste a project, PCB file, or schematic sheet via the file system, you will inherit the old UUIDs from the source file and KiCad won’t care. So it is both not only not guaranteed that UUIDs are unique between different projects, it is quite likely in some cases.
But they should be unique within a project.
In general, KiCad doesn’t expect users to need to know about UUIDs, which is why they’re not shown in the UI. Normally, the only place you may be aware that such a thing exists is that is how the symbols and footprints stay linked even if you, say, change “R42” to “R43” in the schematic.