Re: Examples of pre-HTTP (i.e. session-based) protocol
Re: Examples of pre-HTTP (i.e. session-based) protocol
- Subject: Re: Examples of pre-HTTP (i.e. session-based) protocol
- From: Jens Alfke <email@hidden>
- Date: Mon, 30 Jun 2014 09:16:17 -0700
On Jun 30, 2014, at 1:45 AM, Daryle Walker < email@hidden> wrote: HTTP classically has been a one-and-done protocol; each resource requires a separate contact-and-response loop. I think that the 1.1 version supports persistent connections, to download a bunch of resources from a single server in a bunch.
Persistent connections (keep-alive) have been the norm since the mid-‘90s, really. No one uses HTTP 1.0 anymore; it’s much too expensive to have to open a new TCP connection for every request.
It helps to distinguish between the application-level semantics and the transport. HTTP is still stateless, but the underlying transport uses a persistent connection. (And there are now multiple types of connections that HTTP messages can be sent across; SPDY is another.) I want to know how to make a subclass that isn’t based on HTTP. The protocol is session-based. So part of what I want to know if Apple’s FTP implementation does brand new connections for every resource, or does it retain connections (for a while) in case a new resource request needs to go to an already-contacted server?
Both the FTP and HTTP implementations keep a pool of open connections and will send a request across one of them if possible. The HTTP implementation will open several connections to the same host (up to 4 on iOS) for better parallelism.
—Jens
|
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Macnetworkprog mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden