Re: Asynchronous DNS

[Eli Zaretskii]

> From: Lars Ingebrigtsen <address@hidden>
> Cc: address@hidden,  address@hidden
> Date: Tue, 09 Feb 2016 11:47:01 +1100
> 
> Do you think that
> 
> (process-send-string (make-network-process ... :nowait t) "foo")
> 
> shouldn't signal an error?

Yes, I do.

> It currently does.  I'm not sure that's the right decision, but I
> think I'm leaning towards thinking that that's more helpful than
> blocking and pretending nothing's wrong...

I think it's just an accident.  When these APIs were originally
written, computers were sufficiently slow as to have the connection
completed by the time any other API is called.  Computers got much
faster since then, so we will see more and more of these errors.  (I'm
quite sure I see them in smtpmail from time to time, because I use a
very distant SMTP server.)