-
Notifications
You must be signed in to change notification settings - Fork 4.8k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[EXPERIMENTATION ONLY] - Prevent Reflection's MakeGenericType()
from Allowing the Instantiation of Abstract Classes
#105844
[EXPERIMENTATION ONLY] - Prevent Reflection's MakeGenericType()
from Allowing the Instantiation of Abstract Classes
#105844
Conversation
Needs a test |
Is there indication the original PR is abandoned? The author looks to be waiting for next steps from area owners - might want to ping the author on the PR. It's not a good look to take someone's work and commit under own name. I see milestone 9.0 - do we still have runway to do breaking changes? |
It looks abandoned since it hasn't had any activity for the last 2-3 months. But I will ping the author. Also I marked it as draft for a reason. I wanted to do some experimentation with the CI first. Was planning to initiate the conversation once I had checked what I wanted to check.. |
MakeGenericType()
from Allowing the Instantiation of Abstract ClassesMakeGenericType()
from Allowing the Instantiation of Abstract Classes
We have push access to the branches with PRs. So for example if that PR is missing a test and the author doesn't react anymore, it's possible to just push a commit with the test to the branch, even if the author is non-responsive (not everyone can work on our schedules). We tend to do that so that we don't rob people of their contributions. Contributors GitHub handle shows up e.g. here: https://dotnet.microsoft.com/en-us/thanks. It wasn't clear to me this is for experimentation, I guess I'll see the intent here later. |
I've rephrased to make it extra clear that it is for experimentation purposes now. And I did not know we could just push to anyone's PR so that's good to know.. |
Original PR is #101963.
This PR aims to experiment with fixing the loophole where
MakeGenericType()
does not fail when attempting to create an object from an abstract class, and thus resulting in incorrect behavior.