Fabs has side effects

[ben_lubar]

Yes, Microsoft. Taking the absolute value of a number changes the original number.

Or was it the subtraction? Yeah, those subtraction-with-side-effects things are pretty hard to work around.

[chubertdev]

Getting the z property has a 0 chance of changing vecStart or vecEnd.

[Captain]

Wow, that is major WTFery.

Is your fabs over-ridden? Or maybe the (-)?

I'm so glad my tools reject improperly typed side effects.

[too_many_usernames]

You guys know that you can't do function calls in a hardware-based conditional breakpoint, right?

(edit: clarified breakpoint type)

[IMil]

Shame on you, Microsoft engineers, for not being able to verify that a function has no side effects! I mean, this cannot be an unsolvable or hard problem, can it?

[dhromed]

NP stands for "No Problemo!" right?

[Onyx]

I'm more concerned that they consider being fabulous a problem.

---

Filed under: That's what it means, right?

[anotherusername]

So, I'm assuming that

vecEnd.z - vecStart.z > 4 || vecStart.z - vecEnd.z > 4

would have probably worked just fine?

[fatbull]

fabs might call _matherr.

[Captain]

I can't read Algol-based languages anymore. It's like a foreign language to me.

[dtech]

Just rewrite everything in Haskell. Problem solved.

[jpa]

Let me introduce you to the FPU status word:

Archiveddocs

_status87, _statusfp