[9front] commit 1a09421f7db897e1d2e2a9210f79ca4389362c3a

[9front] commit 1a09421f7db897e1d2e2a9210f79ca4389362c3a

[Anthony Martin]

> kencc: clarify warning for unused assignments
>
> [...]
>
> - warn(Z, "set and not used: %B", bit);
> + warn(Z, "assignment not used: %B", bit);

What is unclear about "set and not used"?

This looks to me like change for change's sake.
Now it's even more tedious to compare the various
compiler forks, the symmetry with "used and not
set" is gone, and the reference to this specific
warning in /sys/doc/comp.ms is stale.

For what gain?

  Anthony

Re: [9front] commit 1a09421f7db897e1d2e2a9210f79ca4389362c3a

[Noam Preil]

"set and not used: foo" heavily implies that the variable is not used. 
This is not necessarily true.

int a = 0;
a = bar();
...

This would produce "set and not used: a" with the line number of the 
declaration, even if the subsequent assignment is used. "assignment not 
used" clarifies the actual warning.

In other words, the warning was changed because the prior form was simply 
wrong.

- Noam Preil

Re: [9front] commit 1a09421f7db897e1d2e2a9210f79ca4389362c3a

[Noam Preil]

Ah whoops, missed the point about the docs - will fix that.

Re: [9front] commit 1a09421f7db897e1d2e2a9210f79ca4389362c3a

[theinicke]

Quoth Anthony Martin <ality@pbrane.org>:
> > kencc: clarify warning for unused assignments
> >
> > [...]
> >
> > - warn(Z, "set and not used: %B", bit);
> > + warn(Z, "assignment not used: %B", bit);
> 
> What is unclear about "set and not used"?
> 
> This looks to me like change for change's sake.
> Now it's even more tedious to compare the various
> compiler forks, the symmetry with "used and not
> set" is gone, and the reference to this specific
> warning in /sys/doc/comp.ms is stale.
> 
> For what gain?
> 
>   Anthony

I totally agree with what Anthony wrote;
also in addition to all of this the change is half-assed:

Just looking at the diff one can see that there is
still the line with the typo "set an not used".

s/set an not used/set and not used/ would have been a
better substitution!

--
Tobias Heinicke

Re: [9front] commit 1a09421f7db897e1d2e2a9210f79ca4389362c3a

[hiro]

since you still wrote "used" it's just as unclear as before IMO.

maybe you want something like

assigned value is never read