Naming issue of the FooThread and alike

[izgzhen]

This is a follow-up of discussion started from #11189.

The naming issues include:

• Thread is not IpcSender: we type-alias a lot like FooThread = IpcSender<FooMsg>. The "thread" here might not be indicating what we want
• net_traits::IpcSend is coupled with Ipc which might not be a good name (as suggested by @nox)
• net_traits::CoreResourceMsg and others renamed from original ControlMsg and ResourceThread. But the CoreResourceMsg manages a lot of duties and used by a lot of other code, which might add difficulty to name it cleanly.

[KiChjang]

Off the top of my head, we can rename it as FooChannel or FooPipe, since it is some sort of message passing system.

[izgzhen]

+1 for the FooChannel one

[izgzhen]

BTW we might use FooChan if the longer one is too verbose.

Also the irc paste for reference here:

10:46 PM <ajeffrey> in fact I prefer not introducing type synonyms for IpcSender<FooMsg> in general,
10:46 PM <nox> ajeffrey: That doesn't sound great.
10:46 PM <ajeffrey> as it makes it trickier to track down the FooMsg definition.
10:46 PM <nox> We don't care whether it's IPC or not. We care about the interface IMO.

[Ms2ger]

-100 on "channel" or "chan". The thing you send on is a "sender"; a "channel" is something with two sides.

[KiChjang]

How about FooMsgSender?

[nox]

FooSender seems ok to me, or just not introducing type synonyms at all as @asajeffrey seemed to prefer.

[nox]

Do we still care about this?

[asajeffrey]

Yes, but it's low priority. While we're at it, the naming of the *Msg types is a bit erratic.

[nox]

The naming is starting to get on my nerves. ><