Titus Winters

Related note: I'm a little hesitant about recommending move-into-place by default until we've got better tooling for detecting use-after-move. The (usually minor) gains in most places can be easily wiped...

If we're telling everyone to move into sink functions (those that store or modify their inputs), there'll be more calls to move, and more use-after-move. We're already seeing it for...

I'll claim that most containers (counted by instantiation or CPU usage or both) are built up and then used in a mostly read-only fashion. That means that although there are...

The policy I've been advocating is roughly this (for cases where the parameter is being sinked/consumed): If the *body* of the function is doing work that is comparable to the...

Nit: In "Copy" you're saying "his", but programming isn't gendered. I'm not an owner for core guidelines, but your summary mostly matches my understanding. I don't know that there is...

Yep, that all sounds accurate.

@MikeGitb - You're correct that it'd be difficult to catch this in the general sense. There isn't a simple mechanism yet to identify "Does this type have reference semantics? (or...

Tentative plans, yes. As always, there's a lot of work to be done checking the design, putting things into Abseil, resolving old issues before making them public. If we release...

I'm not aware of anything good quality in C++ that abstracts over both native files and GCS. You might do better with filesystem plugins to make GCS look like a...

Like I said internally - I could *swear* we did some of the sign conversion cleanup before release, so I'm surprised that this is busted. But at the same time,...