On Tue, Apr 5, 2016 at 1:45 PM, Linus Torvalds torvalds@linux-foundation.org wrote:
Mind sending me an xml file with your dive just as a test-case, I'll take a look.
Ok, thanks, that shows that the failure mode wasn't quite the one I expected - but yes, the "Info" window shows you with a SAC rate of 0.25 (very impressive!) while the "stats" page shows you with a SAC rate of 0.50 (rather more believable).
Looking at it all, I'm not *quite* sure how to handle it. One the one hand, the 0.25 SAC rate is clearly completely bogus, because it is calculated as if you had only used the first cylinder, while the actual "stats" page just looks at your overall SAC-rate using your total gases used. So the "info" side is clearly wrong.
At the same time, the info side is *meant* to show the per-cylinder gas use, and without gas switches, it basically has an impossible job.
I suspect the only option for us is to simply have a special case, and notice that
(a) you had a gas that clearly was used
(b) you didn't have a gas switch event *to* that gas
and as a result just say that "we cannot give you per-cylinder SAC rates".
Let me see how trivial that is to do. It doesn't sound hard, it just sounds _tedious_.
Linus