[FrameworkBundle] HttpCache is not longer abstract #26356
Conversation
|
this was never necessary AFAICT. It looks like a mistake by trying make it parallel to AppKernel extending the Symfony class, while there was actually nothing to implement in AppCache from day one. |
|
However, this class has a typehinting issue in the constructor. It typehints HttpKernelInterface, but the implementation actually expects a KernelInterface (which has much more methods) |
|
So is it a bug fix ? |
|
lets' make it a bug fix and fix the type hint? |
|
OK, I will do that |
87baaf1 to
ded042a
Compare
|
done |
| { | ||
| protected $cacheDir; | ||
| protected $kernel; | ||
|
|
||
| /** | ||
| * @param HttpKernelInterface $kernel An HttpKernelInterface instance | ||
| * @param string $cacheDir The cache directory (default used if null) | ||
| * @param KernelInterface $kernel An KernelInterface instance |
There was a problem hiding this comment.
- An KernelInterface instance
+ A KernelInterface instanceded042a to
4d075da
Compare
|
I'm confused: is this a bug fix or a new feature? The branch is |
|
@javiereguiluz I updated the PR to be a "bug fix" as requested. Indeed the milestone should be changes. And this change does not require any change in end-users code. |
|
Thank you @lyrixx. |
This PR was merged into the 2.7 branch. Discussion ---------- [FrameworkBundle] HttpCache is not longer abstract | Q | A | ------------- | --- | Branch? | master | Bug fix? | no | New feature? | no | BC breaks? | no | Deprecations? | no | Tests pass? | yes | Fixed tickets | | License | MIT | Doc PR | --- I don't really know why this class was abstract in the first place. But it's not needed. A fresh symfony 4 application can use this class directly without extending it. Commits ------- 4d075da [FrameworkBundle] HttpCache is not longer abstract
|
How is this a bug? This is definitely a "new feature". It does not fix anything. |
|
@fabpot initially the PR targeted master. But Nikos asked me to put it as a bug fix.
WDYT? Should I revert the PR and re-open a new one against master ? |
|
"But I thinks it does not harm" is exactly what I don't like. If we start doing that, most PRs do not harm (at least, that's the initial reasoning), but something it does. So, we should follow our policy strictly: only bug fixes on non-master branches. In this case, I would revert the abstract change in all branches but master (you can do that directly in the code without PRs as it would mean 2 extra PRs). |
|
Ok I got it. Sorry about that. I have added 6073066 to the 2.7 branch. I wanted to merge 2.7 in 2.8 but there was conflicts about other changes. I'm not really able to solve theme :/ Thanks |
I don't really know why this class was abstract in the first place.
But it's not needed. A fresh symfony 4 application can use this class
directly without extending it.