On 22-4-2022 8:47, Rowan Tommins wrote:
On 22 April 2022 02:02:58 BST, php-internals_nos...@adviesenzo.nl wrote:
I agree it would be a good idea to run a package analysis, but to be fair, in
all honesty that should have been done for the original RFC, which was
completely missing an impact analysis.
That's a fair point, but of course "somebody else should have done it already"
isn't a good excuse not to do it now.
3. As for the pattern being common or not - the fact that I found it so easily
in multiple random projects which I elected to test the sniff against, makes me
believe the pattern is not _uncommon_.
Since there has been some misunderstanding, the usage I think we need to look for is where a deprecation notice would
*not* be useful. In other words, are there people using is_callable in such a way that even if a value like
"parent::foo" changes from returning "true" to "false", there won't be anything that
needs changing, because the code is equally "happy" with both return values?
I freely admit that I can't think of any such usage off the top of my head, but
I wouldn't have thought of some of the examples already raised either.
If an actual search fails to find such usages, or provides evidence that it is
very rare, then I am absolutely in favour of adding a deprecation notice.
Ah! So you're basically asking for the impossible. Search for a code
pattern where a deprecation would not be useful, while noone has been
able to come up with one.
In that case: search done. Nothing found. Let's get this fixed.
