Bug fixes I threw patches in the past

[9rnsr]

No description provided.

[donc]

Bug 5080 cannot be applied until opEquals is fixed. I think the others are ready.

[9rnsr]

Patch of issue 5080 changes only implicit conversion behavior. See Steven's comment #9.
http://d.puremagic.com/issues/show_bug.cgi?id=5080#c9

[donc]

Please ignore Steven's comment. It's misleading. opEquals is WRONG. This patch prevents the only known workaround.
The problem is, that we have legal code which compiles right now because of this bug. Putting this patch in right will break existing code. And the problem is NOT with the code, or with this patch; it's with opEquals. BTW almost all compiler regressions are like this: they break something which used to work, but for incorrect reasons. The fact they worked for incorrect reasons does not stop it being a regression. opEquals needs to be fixed first.

[9rnsr]

Thanks for the detailed explanation. I'm looking forward to your fix.

[WalterBright]

In the future, please have a separate pull request for each issue.

[WalterBright]

5178 is incorrect, see the bugzilla entry for it.

[9rnsr]

I understood your reason. I also think that converting from rvalue to lvalue is correct behavior.
Apart from that, it seems to me that StructLiteral should be rvalue.
This behavior is need to write resource management wrapper type like std.typecons.Unique.
I posted issue 5769 with same reason.
What do you think?