@ch3root deviating from spec is okay IFF documented, especially important in compilers
-
-
-
Only if there's a unified option for "adhere to the standard" rather than N different esoteric ones you have to know
-
Yeah, let's introduce an option to choose a particular standard, like -std=c11. Oh, wait... :-)
-
Exactly. My point was that the standards-conforming options should all be implied by -std=c11, etc.
-
if a vuln was found in a standard should we require extra flags to disable feature, should it be default?
-
I don't know what "a vuln in the standard" would look like.
-
AFAICT it's a meaningless term. The closest thing might be funcs like "gets" that are impossible to use right.
-
But then the only fix is teaching programmers not to use them, or amending the standard to remove them.
- 1 more reply
New conversation -
-
-
@ch3root this is a perfectly reasonable question to ask if you're trying to allocate scarce dev time to the most important bugs. -
I.e. rather than "go away, standards weenie," read as "how should we prioritize this vs other bugs?"
-
Changing the default value for -fmerge-all-constants is that difficult?
-
Changing the value is easy. Dealing with the fallout of recursive functions in existing code that depend on the old \
-
behavior to not blow out the stack might require a fair bit of support and hand-holding.
-
they could take GCC's "suck-it" approach.
-
The "suck-it" approach is tempting, but doesn't exactly endear you to compiler users =).
End of conversation
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.