Closed fdouglis closed 3 years ago
Thanks, this is certainly a bug. The the true intention should be exactly as you propose (i.e., comparison rather than assignment).
Aha, thanks. And you are most welcome.
I fixed this. @fdougliy, could you please pull and check if the compilation error is gone?
@geisserf: I accidentally violated our workflow in the development repo, and decided after that first violation not to go through the hassle of undoing everything but ignore the workflow altogether. For this reason, this issue is already merged.
Yup, builds without complaint now. And I see the real test was for the bounds, and not that the parameters were already set to the bounds. So the change I'd made wasn't really right either, though it didn't complain at runtime.
Fixed in commit de09e63f34e06e38108979d3e02468cfa8ba230e
When I moved to a new system and rebuilt prost, I ran into a set of complaints in one file. One example:
At first I thought this was a typo and it was supposed to be == (and perhaps it is). I changed it to ==, and it built, but I hit that other error I mentioned with the new code for identifying those errors (#128). I tried changing it to what I think the original code did:
assert((minRes = numeric_limits<double>::max()) != 0);
but I hit the same bug.I suspect this has nothing to do with #128 and simply an artifact of building on a very recent system (FC33), and it's a syntax issue you might want to address in the master repo at your convenience.