Am 11.10.2022 um 21:20 schrieb Kevin Kofler via Development:
"locking you down to a vendor" is a funny argument when the Wikipedia article: https://en.wikipedia.org/wiki/Pragma_once cannot name a single compiler that does not support #pragma once, and 20 that do.
Yes, it is supported in a way, that each compiler recognizes this #pragma and does something compiler specific with it, but what exactly a compiler does with it, how each compiler determinates identity is purposefully not listed in that table. Well, and that's exactly the problem: Nobody really can tell, what any compiler does with the pragma and if behavior might change with the next micro patch version.
I am pretty sure, absolutely nobody on this list would object the use of #pragma once, if compiler vendors would have documented somewhere, what exactly #pragma once means, what identity means, how it is recognized, how it is to be implemented. If #pragma once would be a well documented and properly implemented industry standard really nobody could object its use, as Qt and its users could rely on the spec and ask tooling and compiler vendors to fix their software if it doesn't match this specification. Still: It isn't. We basically are discussion the use of a feature that's entirely undefined and totally in the void. Doing so is something for religion, not for engineering.
Ciao Mathias _______________________________________________ Development mailing list Development@qt-project.org https://lists.qt-project.org/listinfo/development
