Couple of things (I read their report earlier, made a post about it actually:
- # of events =/= # of votes; an event is any step in the process that one ballot goes through (ballot scanned, ballot stored, ballot analysed, data presented to screen, written to disc; n.b.: these are random examples, the report does not specify);
- errors are always present in an error log, the amount, or rather, severity of the error logged varies on the settings one inputs for the recording of the error log, so an error could be any minor thing in the process, which does not affect the process itself, and is not presented to the user; the report states that the vast majority of these errors are actually errors related to the hardware used, the most common ones; they specify only a few, but the most serious one was "not enough memory available to process", which means the system had to use more memory than was originally allocated to the software, and it just grabbed some from elsewhere; does not affect the process;
- the only thing the report states it found, was that there were possibilities to misuse the software, i.e., commit fraud, but nowhere does it state they found that actually happened; they, for instance, state that the config of the software was so broad, that an administrator
could have flipped votes, however, they found no evidence of that happening in the logs;
- the 0.7/ 1.3 nonsense was debunked the second it came out, and is found nowhere in the released report.
How's that for leg work
@glenwo2? Of to bed now, leg day tomorrow, coincidentally.