On 17 August 2024 22:33:03 BST, Rob Landers <rob@bottled.codes> wrote:
>I wouldn't consider it a BC break, no. But (ironically?), Symfony crashes with
>this change. It really shouldn't but ...
I don't think it makes sense to say "it breaks existing code, but it's not a
compatibility break".
Perhaps what you're saying is "it's only a BC break for code that's not
following best practices"?
But more relevant than whether you think the current code is "correct" is the
fact that a) it will need to be changed to work with your proposal; and b) the
change is simple and can be done in advance.
So the RFC should acknowledge this BC break, but could argue that it's small
enough to include in a minor version. This is actually really common - RFCs
that introduce a new global function often acknowledge that it would break
existing userland functions with that name. Between that and obviously serious
BC breaks like *removing* a function, there's a big grey area where we have to
make a judgement call.
Regards,
Rowan Tommins
[IMSoP]
