emacs-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Final(?) patch for server sockets and datagram (UDP) support.


From: Al Petrofsky
Subject: Re: Final(?) patch for server sockets and datagram (UDP) support.
Date: Wed, 13 Mar 2002 16:50:12 -0800

It's always been a bit confusing that emacs uses the term process for
sockets that have no associated process.  That confusion will get a
little worse now that emacs processes will include server and datagram
sockets, which don't even share processes' stream-like nature.

I'm not saying we should rename everything now to fix this, but I
think it would help if the start of the "make-network-process" doc
string immediately disclaimed any relationship to a unix process.

> :host HOST -- HOST is name of the host to connect to, or its IP
> address.  If specified for a server process, only clients on that host
> may connect.  The symbol `local' specifies the local host.

Don't you mean something like "If specified for a server process, it
must be a valid name or address for the local host, and only clients
connecting to that address will get through"?

> :local ADDRESS -- ADDRESS is the local address used for the
> connection.  This parameter is ignored when opening a client process.
> When specified for a server process, the HOST and SERVICE are ignored.
> 
> :remote ADDRESS -- ADDRESS is the remote partner's address for the
> connection.  This parameter is ignored when opening a server process.
> When specified for a client process, the HOST and SERVICE are ignored.

These seemed pointless until I read the process-contact doc, and the
NEWS.  (The format of the address wasn't documented in either
function's doc string.)  I don't think it's a good idea to add two
more arguments to make-network-process just so that process-contact is
easier to document.  Is there some other point?

-al



reply via email to

[Prev in Thread] Current Thread [Next in Thread]