04 Dec, 2009, Brinson wrote in the 1st comment:
Votes: 0
Has anyone here experimented with how many kb/s sustaining a constant connection to each user uses? I assume its kind of like streaming a song, where in if you stream at 8 kb/s, and can upload 80, you can sustain 10 users…anyone know what kind of upload's necessary to run a decent sized mud (50ish users at a time)?
04 Dec, 2009, Asylumius wrote in the 2nd comment:
Votes: 0
I've never looked into it, but I would imagine it varies a lot from MUD to MUD. What the players are doing and what game options they have set (brief, ascii map, etc) probably make a difference, as well as if you're using all those fancy protocols to do extra stuff with the client.

You could use some basic network analyzing tools to monitor the traffic though and figure out what kind of bandwidth you're using up on your MUDs port. I can't imagine its much.
04 Dec, 2009, David Haley wrote in the 3rd comment:
Votes: 0
It's really not like songs at all, because song streams are more or less constant in terms of data to be sent whereas MUDs are "bursty". If your maximal capacity is 100 (ignore units), and each user consumes 10 at peak, you can only sustain 10 if they all have constant peak activity. In practice, players aren't all at peak activity, so you could probably sustain 15-20 (made up numbers) or more depending on how the "burstiness" is distributed. In other words, a player's average bandwidth usage is not the only interesting number because usage is not constant; when that usage is consumed also matters.
04 Dec, 2009, KaVir wrote in the 4th comment:
Votes: 0
Brinson said:
Has anyone here experimented with how many kb/s sustaining a constant connection to each user uses?

I know someone did a comparison both with and without MCCP, but I can't recall where they posted it, or even which mud it was (was it Aardwolf?). Maybe someone else can remember the link.

EDIT: Found it, here you go (bottom of the page).
04 Dec, 2009, David Haley wrote in the 5th comment:
Votes: 0
Kayle has posted comparisons a few times as well, and the savings are considerable even for a relatively small MUD.
05 Dec, 2009, Kayle wrote in the 6th comment:
Votes: 0
Transfer data since Sun Sep 6, 2009 3:35:15 PM EDT
Received : 3MB 314KB 254B
Uncompressed sent : 270B
Compressed sent : 23MB 50KB 693B
Saved by MCCP : 211MB 127KB 499B
IMC data received : 36MB 5KB 569B
IMC data sent : 3MB 207KB 406B
Total Received : 39MB 319KB 823B
Total Sent : 26MB 258KB 345B
05 Dec, 2009, quixadhal wrote in the 7th comment:
Votes: 0
Hmmmm, perhaps IMC3 should require MCCP2. The mud in question could still support user connections without it, of course, but enforcing it as part of the IMC protocol should be simple enough, no?
05 Dec, 2009, Runter wrote in the 8th comment:
Votes: 0
quixadhal said:
Hmmmm, perhaps IMC3 should require MCCP2. The mud in question could still support user connections without it, of course, but enforcing it as part of the IMC protocol should be simple enough, no?


Enforcing some type of compression would be simple enough.
05 Dec, 2009, Mudder wrote in the 9th comment:
Votes: 0
That's a good suggestion. I like the enforced compression.
0.0/9