Release 4.3.0

[jonorossi]

I think its just about time to cut a new release.

Thinking 4.3.0 would be a good version number since we've got a few improvements along with a handful of defect fixes.

Thoughts?

/cc @fir3pho3nixx @stakx

[stakx]

Sounds great to me! Increasing the minor version seems a sensible thing to do because, among other reasons, there has been a change to the targeted frameworks.

_{My only thought was whether to include a (partial) fix for #341 in this release. At first glance, it should be relatively easy to duplicate custom attributes on the generated proxy type ctors. I'd be willing to take a stab at it today. Only the reproduction of default values would pose some problems. (I've done some preliminary work on that in #291, in order to solve #341 completely, that work would likely have to go through first, which would take me some more time to do properly... too long for this release.)}

_{That being said, I'd also be totally fine with #341 being resolved in the next release instead of this one.}

On second thought, I wouldn't be surprised if including a last minute fix for #341 led to problems later on. Perhaps better to give this enough time and test it thoroughly.

[jonorossi]

#341 was actually the only remaining issue I was wondering if we could get fixed before release, it would be great if you had a chance today to look at it. I'd ignore default values for now (that repro doesn't have them anyway), I wouldn't want to rush that in with the number of problems we've had with default values in the past.

[stakx]

@jonorossi - OK. I'll take a look and will let you know later on today how things stand. If I'm confident that I can submit a fix in the next two days or so, I'll get started on it.

[stakx]

@jonorossi - I've just submitted #355 as a proposed fix for #341.

[jonorossi]

We've progressed a whole bunch of things forward in the past 29 days.

@castleproject/committers is there anything else we want in 4.3.0 before release?

[stakx]

I think we're good to go! 🚀

(There's one general topic -- async interception, discussed in #145 and #238 -- as well as two reported defects, #106 and #295, all of which would profit from a closer look instead of a quick fix. Regarding the defects, there's an easy workaround for the former, and the latter doesn't directly break user code or anything serious, so I would say let's deal with them later instead of rushing things.)

[ghost]

Think we are good to go.

I am also very interested in the async stuff but lost touch for a bit. We should definitely not rush anything.

[ghost]

@stakx please feel free to call me into anything you think I could help on. I am busy with Windsor. If you need me to drop that I can to help out.

[stakx]

@fir3pho3nixx - Will do, cheers! 👍 If you're interested in the async issues, I'll happily defer them to you... perhaps that makes good sense too, since I haven't actually taken a closer look at those yet anyway.

Now that most easily fixable bugs are gone, I am planning to focus a bit longer on the various reflection bugs in the .NET runtimes (e.g. their handling of parameter default values and custom modifiers), I think these will increasingly cause trouble for DynamicProxy, given the new C# language features. I estimate that this will keep me busy for several weeks (I've delved into CoreCLR's reflection code and it's, uhm, somewhat convoluted, to put it nicely 😄).

[jonorossi]

Just tagged v4.3.0, confirmed successfully pushed to nuget.org, and tweeted.

[ghost]

Can I upgrade Windsor please?

[jonorossi]

Can I upgrade Windsor please?

Go for it.

[ghost]

Done. Noticed the commit that fixed the CHANGELOG.md. Will follow that pattern. Thanks.