Remove Obsolete Attributes

[rprouse]

We obsoleted several attributes like FixtureSetup with the 3.0 release. It is probably about time we removed them.

Other candidates for removal are some of the fluent syntax for constraints where we came up with better naming.

[stevenaw]

I can take this on.
Can I please be assigned to it?

[CharliePoole]

Hi @stevenaw Sure thing. I sent you an invitation as a contributor. After you accept I'll be able to make assignments to you.

[stevenaw]

Thanks @CharliePoole , I've accepted the invite 🙂

[Lette]

Just a thought:

If we're following SemVer, we have to consider that removing these attributes might be a breaking change, which will require a bump of the major version. So, this is probably something that shouldn't be done before the 4.0 release.

[CharliePoole]

That's a point. Not sure how much into semver everyone is at this point though. @rprouse ?

[NN---]

Please do it in 4.0.
It will break my code and many other projects too.

[CharliePoole]

@rprouse How do you want to handle this? Should we have a 4.0 Milestone?

[rprouse]

If we follow SemVer to the letter, then we probably should hold off on this to a 4.0 release. I think you could argue though that since most of the attributes were obsoleted in 3.0 and there have been many releases to fix the warnings it isn't a breaking change.

AFAIK, the only obsolete feature that isn't a simple search and replace update is the AssertHelper. AssertHelper has only been obsolete for one release, so removing it may be pre-mature. Updating everything else should only take a few minutes, even for large solutions. @nn or @Lette is there anything other than AssertHelper that you see as a major or difficult change?

Personally, I am surprised that people haven't already updated their code, but maybe I just hate warnings more than other people 😄 I am also surprised that people continue to use AssertHelper since it is so limited compared to the other asserts.

[stevenaw]

I have some of these changes sitting in a branch. Should I go back to it, wrap it up, and open a PR? Or should I wait until we know when we'll be releasing it?

…

On Tue, Jun 27, 2017, 09:04 Rob Prouse ***@***.***> wrote: If we follow SemVer to the letter, then we probably should hold off on this to a 4.0 release. I think you could argue though that since most of the attributes were obsoleted in 3.0 and there have been many releases to fix the warnings it isn't a breaking change. AFAIK, the only obsolete feature that isn't a simple search and replace update is the AssertHelper. AssertHelper has only been obsolete for one release, so removing it may be pre-mature. Updating everything else should only take a few minutes, even for large solutions. @nn <https://github.com/nn> or @Lette <https://github.com/lette> is there anything other than AssertHelper that you see as a major or difficult change? Personally, I am surprised that people haven't already updated their code, but maybe I just hate warnings more than other people 😄 I am also surprised that people continue to use AssertHelper since it is so limited compared to the other asserts. — You are receiving this because you were assigned. Reply to this email directly, view it on GitHub <#2219 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AAa2JNn3PuJAEzRw_L5jW5jpwR-2oOUvks5sIP35gaJpZM4Nvhbo> .

[NN---]

I have some obsolete attributes. Not really much work to find and replace.

The main difficulty was ExpectedException attribute removal in 3.0 which I can say was a very good idea.
Too bad it existed from the beginning..

[rprouse]

@stevenaw if you want to do a PR, it might help us visualize what is changing and what the impact will be without searching through the code ourselves. If you do the PR, please put a do not merge in the description and you should probably hold off on doing more work until we make a final decision on this issue. @Lette and @NN--- have raised good points and I think we should evaluate the impact of this change.

[stevenaw]

@rprouse @CharliePoole I've put up a sample PR here: #2338
It doesn't look like I'm able to update labels, so I've put a big "DO NOT MERGE" heading in the description

[stevenaw]

Sorry, I forgot to structure my GH comment so that this would close automatically.
I'm unable to manually close it. @rprouse @CharliePoole @ChrisMaddock can you guys help out?

I see we're handling obsoleted non-attributes (constraints) in a different issue (#2342)