oups, ACK'ed to fast, this is not 8, but 16 here

ups, I misread the code, just saw the SIGFOX_PAC_LENGTH >> 1 but didn't notice that it was the parameter for fmt_bytes_hex() and not for transfer_bytes... Will fix.

but wait, wouldn't it meant that the resulting string size is 17 then? The _read_pac() function copies 16 characters into the result string and then puts the \0 byte in the 17th position (array offset 16) of the result string, right?

You are right, I spoke too fast. That's what the driver test application does: see here

Perfect, then I will adapt this to the really correct numbers :-)

see here: https://github.com/RIOT-OS/RIOT/blob/master/drivers/include/ata8520e.h#L42

Please revert the doxygen 1 character shift of this enum, it seems unrelated.

no, it is not, see the first commit of this PR: acac429

Alignment in the master was off for one space for this block...

Alignment in the master was off for one space for this block...

I don't understand. Are you still talking of the "multiple of 4" indentations requirements? I don't consider this as a rule of thumb for alignment, thus the doxygen comment alignment is not 'off' in master. Maybe you should update your editor configuration ?

I'm refering this comment: "some doxygen blocks were not aligned to 4 spaces (this upset my text editor :-) )"

Yes, I am still talking about the 4-space alignment. In master, the this doxygen block is indented by 45 characters and 45 % 4 != 0. So I put moved it to 44 characters -> now 4-space aligned.

I don't consider this as a rule of thumb for alignment

But the coding convention does per default, as long as there is no good reason not to do so. I always interpreted this also applicable to doxygen blocks -> so yes, I say the 4-space alignment does apply here.

Furthermore, even if ignoring the 4-space alignment rule, it does not make sense to indent the block in questions here in a different style than the two blocks below...

@haukepetersen what about stopping discussing what I'm requested and revert this change ? Those doxygen comments are aligned on their own and the code is nice enough (and not only because I wrote it ;) ).

The only meaningful change of this PR is related to the documentation of payload sizes that are incorrect. Only this should be changed I think.

I think it is completely valid to piggy-back these kind of formatting fixes (and in this case it is a fix) with a related PR if the change in question is kept in its own commit. This is the case here, so let's not block this PR on a matter of opinion.