Bump clipboard-win from v3.0.2 to v4.5.0

[rhysd]

I confirmed examples worked on my Windows 10 machine.

Small trick in this patch is that clipboard_win::SystemError does not implement std::error::Error so we need to wrap the error instance by ourselves. This should be okay because the wrapped error instance is hidden by Box<dyn Error> trait object.

Fixes #65

[chrisduerr]

WrappedError is a pretty bad name, would probably be better to make it WindowsError, WindowsClipboardError, WindowsSystemError, ClipboardSystemError, or something similar.

[rhysd]

Sure. I will update it.

[rhysd]

@chrisduerr I chose WindowsSystemError and force-pushed the change to this branch: 795726f

[kchibisov]

The only reason I haven't bumped dep is because there was no point in doing so given that I don't remember that it provided any value. However I was waiting for their 5.0 version, which will use windows-rs so we should be able to drop winapi.

[rhysd]

However I was waiting for their 5.0 version, which will use windows-rs so we should be able to drop winapi.

I noticed it, but I'm not sure when it is actually released. And a new major version tends to be buggy. Once 5.0 is released, I'll try to make another PR. Meanwhile, can we use v4.5.0?

[kchibisov]

And a new major version tends to be buggy

Tends or is? It's not like you can't try to update and test yourself.

but I'm not sure when it is actually released.

You can ask and explain why you want a bump?

[chrisduerr]

Considering version 4 is the most widely used version out there, it's probably just useful for deduplication purposes. Though not sure why you'd have two versions of this in your tree.

That said, the maintenance isn't amazing so not updating is probably the safe choice.

[rhysd]

Sorry I may not get the point of discussion. My intention was upgrading the dependency to the most stable latest version v4.5.0 and then upgrading the version to v5 once it is released. What is a problem with it?

[kchibisov]

I mean, you could try to work on upgrade straight to v5 and work with upstream on getting stable version. They already have pre-release, I don't mind bump, but it seems like we'll bump dep anyway to v5, which is preferred.

[chrisduerr]

My intention was upgrading the dependency to the most stable latest version v4.5.0 and then upgrading the version to v5 once it is released. What is a problem with it?

Well technically if 4.Y.Z is receiving constant updates, but doesn't provide anything over 3.Y.Z which is working without drawbacks without updates, then 3.Y.Z is the more stable version.

[rhysd]

Hmm, OK, I'll give up this upgrade and try other crate.

[Tastaturtaste]

That said, the maintenance isn't amazing so not updating is probably the safe choice.

How is not getting any patches safer than getting some patches?

Not updating this dependency means this bug still causes crashes today even though it was fixed in 2022.

Obviously you are free to do whatever you want with your crate in your own time, but, like @rhysd, I will consider arboard over this crate for my needs because of this.

[kchibisov]

You can update to v5.0.0 as was said. abroad serves different purpose compared to this crate, but again, if you don't do windowing you don't need this crate.