Skip to content

TriggerTime: fix wrong integer division

Jens Kroeger requested to merge jekroege/corryvreckan:fix_a_dot into master

trigger timestamp was wrong due to integer division, bringing back the dot to make it a double-divison

Discovered with @tvanat. Since stamp is always between 0 and 11 it was resulting in zero all the time.

In @fpipper's 43f50246 it is present and that's what he was using for his in-depth timing analysis.

43f50246

@simonspa, I remember we discussed that it would be better coding avoiding situations in which a dot makes a difference. If this is the case here, we could for instance do

static_cast<double>(timestamp + static_cast<long long int>(stamp)) / 12 / (8. * 0.04);

i.e. move the bracket of the cast, right? What do you think?

Merge request reports