Alex Tweedly  wrote:

> Richard wrote:
>> This excercise raises a question:  rather than invent another protocol,
>> why not use HTTP?
>
>
> Usually, because HTTP is quite decidedly a client-server protocol.
> If you have a peer-peer protocol need, then you have to bend HTTP
> out of shape :-)

Sockets work by having a listening and a caller.  Having both on both sides is 
P2P.

What "bending" of HTTP is needed over any similar protocol, like the one being 
designed here?


> Or, because in the event of a breakdown in the firewall, a 
> mal-intentioned outsider could more easily circumvent your protections.

That would also be a concern with P2P, where ports must be opened.

And that's also one of the reasons much of the P2P buzz of the early part of 
this century faded as cloud rose: well maintained firewalls can make devices 
unreachable.

Skype is a good example.  Originally designed as P2P with a client-server 
fallback when blocked by a firewall, it's mostly used via client-server these 
days.


> Or, just because it's more interesting :-)

Learning is always a valid reason.

I've built and thrown away a good many sytems over the decades, for custom 
protocols to network DBs and a couple of half-baked CMSes. I don't regret the 
time invested.  Tho these days I tend to use standard tooling where available 
for the case at hand, the time spent with home-grown experiments gave me a 
deeper understanding of the problem space, allowing me to use standards a bit 
better than if I'd come in cold.

--
Richard Gaskin
FourthWorld.com

_______________________________________________
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription 
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode

Reply via email to