Opened 11 years ago
Closed 5 years ago
#651 closed defect (fixed)
std::isnan detection should be via autoconf test, not OS ifdef
| Reported by: | gdt | Owned by: | strk |
|---|---|---|---|
| Priority: | minor | Milestone: | 3.8.0 |
| Component: | Build/Install | Version: | main |
| Severity: | Annoyance | Keywords: | |
| Cc: |
Description
in include/geos/platform.h.in, there are explicit ifdefs on OS to decide if ISNAN should be ::isnan or std::isnan.
- If standards call for std::isnan, perhaps that should just be used.
- If it really does need to be different, then there should be a feature test in configure.ac, rather than a list of operating systems.
Change History (6)
comment:1 by , 11 years ago
| Milestone: | GEOS Future → 3.5.0 |
|---|
comment:2 by , 9 years ago
| Milestone: | 3.5.0 → 3.6.0 |
|---|
comment:3 by , 8 years ago
| Milestone: | 3.6.0 → 3.7.0 |
|---|
comment:4 by , 7 years ago
| Milestone: | 3.7.0 → 3.8.0 |
|---|
comment:5 by , 5 years ago
I think this can be closed since ISNAN is no longer used as of a582d479fed26f3c17e0e0e49112cb74e039e62c/git, correct?
comment:6 by , 5 years ago
| Resolution: | → fixed |
|---|---|
| Status: | new → closed |
Yeah, in theory now that we're using c++11 and the standard way of getting that functionality, the platform tests are irrelevant and have been removed.
Note:
See TracTickets
for help on using tickets.
Ticket retargeted after milestone closed