Muds?

I used to play Zynna back when it was called Zhing!. Coincidentally, I spent a tiny portion of this morning thinking about playing again to see what's new.

I wrote a MUD engine several years ago, and deployed it for a few weeks before becoming bored with it and taking it offline.

I probably still have the source kicking around.

I have played MUDs in the past and I wrote my own MUD server a while back. Presently one of my many programming projects is a MUD along with a team of other people.

Admin and time waster on
http://cryosphere.net. We're shiny.

I'm currently working on a C# MUD. Still in the second week of programming though, so I don't have much to show for it besides the network engine.


I'm using an asynchronous socket model to support many connections, but I ended up creating a demultiplexer for the work thread in order to avoid race conditions and deadlocks.

For the future, I think I may end up creating some sort of world segmentation system so I can create N worker threads, where each one will work on a single segment of the game. However, the current network engine can handle 500 spamming connections on a dual core machine before the worker thread starts falling behind real-time processing, so that may be something far in the future.


It's kind of funny that even with 500 connections, my CPU usage was peaking at 55-60% (50% for the core that the worker thread was running on, and 5-10% for the asynchronous sockets in the threadpool).

I used to run a mud...long time ago...called Battle of the Immortals...aka BOTI.


yeah, that was a while ago.

Interesting. I'm honestly surprised to see any responses on this at all. For those of you who are working on your own, are you planning on setting it up for the public and actually running it, or is it just an amusing exercise to build an engine and leave it at that?

Quote:

Original post by xanin Interesting. I'm honestly surprised to see any responses on this at all. For those of you who are working on your own, are you planning on setting it up for the public and actually running it, or is it just an amusing exercise to build an engine and leave it at that?

I honestly have no idea. It's just something that I've always wanted to do, and for some reason I feel motivated to do it at this time, even though I know the concept is hideously out of date.

However, I have been designing the engine with possible future graphical additions in mind.

Quote:

Original post by Mithrandir Quote: Original post by xanin Interesting. I'm honestly surprised to see any responses on this at all. For those of you who are working on your own, are you planning on setting it up for the public and actually running it, or is it just an amusing exercise to build an engine and leave it at that? I honestly have no idea. It's just something that I've always wanted to do, and for some reason I feel motivated to do it at this time, even though I know the concept is hideously out of date. However, I have been designing the engine with possible future graphical additions in mind.

Speaking of which, how tough do you think it would be to do an avatar-chat front-end, ala the Unity avatar chat at http://www.moock.org/unity/uchatavatar/

Seems like it'd just be a matter of sending "hidden" messages like "user Zippy moves to x=100 y=50", then also sending everyone's position information to clients who join the room.

Just musing.

Quote:

Original post by Mithrandir I'm currently working on a C# MUD. Still in the second week of programming though, so I don't have much to show for it besides the network engine.

Will you be writing anything about it? Lessons learned since Mud Game Programming or something like that?

Quote:

Original post by BradDaBug Quote: Original post by Mithrandir I'm currently working on a C# MUD. Still in the second week of programming though, so I don't have much to show for it besides the network engine. Will you be writing anything about it? Lessons learned since Mud Game Programming or something like that?

Possibly.


Here are some of the biggies so far, from BetterMUD 1.0 to 2.0 (the one I'm working on):

* Integrating Python into C++ was a disaster. I should have done the whole thing in either Python or C++, or used a framework that is designed for integrating different languages together (ie: .NET)
* Using polled sockets is hideously inefficient, and it only takes a little more work to implement a much more scalable asynchronous socket server
* Trying to support Linux and Windows at the same time was pure folly; I spent more time working out quirks on both systems than I did actually making the game


I'm sure there's more that I'm not thinking of right now; it's been a few years since I touched BM1 and I've forgotten a lot about it.

Quote:

Original post by johnhattan Quote: Original post by Mithrandir Quote: Original post by xanin Interesting. I'm honestly surprised to see any responses on this at all. For those of you who are working on your own, are you planning on setting it up for the public and actually running it, or is it just an amusing exercise to build an engine and leave it at that? I honestly have no idea. It's just something that I've always wanted to do, and for some reason I feel motivated to do it at this time, even though I know the concept is hideously out of date. However, I have been designing the engine with possible future graphical additions in mind. Speaking of which, how tough do you think it would be to do an avatar-chat front-end, ala the Unity avatar chat at http://www.moock.org/unity/uchatavatar/ Seems like it'd just be a matter of sending "hidden" messages like "user Zippy moves to x=100 y=50", then also sending everyone's position information to clients who join the room. Just musing.

Probably not tough at all. BM2 works right now by using customizable "Command" objects; they all have a chain-of-notification so they know who to notify when they are executed, so all you'd need to do is program a new command to handle sub-room movement. Since all entities in the game are dynamic objects, you can easily tack on new variables to the objects that are relevant only to the new command.

I could program something like this for Phase 2 or 3 of the project.

Quote:

Original post by Mithrandir It's kind of funny that even with 500 connections, my CPU usage was peaking at 55-60% (50% for the core that the worker thread was running on, and 5-10% for the asynchronous sockets in the threadpool).

Interesting. Can you get cache miss rates for the two cores? It seems likely that a worker thread would have much better locality than a pool of threads.
Now optimize your worker thread so it only gets 1% cache misses! [grin]

Quote:

Original post by Mithrandir[...]It's kind of funny that even with 500 connections, my CPU usage was peaking at 55-60% (50% for the core that the worker thread was running on, and 5-10% for the asynchronous sockets in the threadpool).

You might find the table in Sample Chapter from Network Programming for Microsoft® Windows®, Second Edition by Anthony Jones and Jim Ohlund interesting.

I'm currently playing dragon swords. It's a fun medieval RPG. Newbie-friendly too!

Quote:

Original post by Zeraan I'm currently playing dragon swords. It's a fun medieval RPG. Newbie-friendly too!

400 mortal levels? How does that not begin to grind? Or is level advancement fairly rapid, in which case one must ask if there are things to do besides gain levels - involved crafting / political systems / etc?

A lot of MUDs, unlike modern MMORPGs, have actual interesting things to do that give you experience points and thus new skills.

Retromud, for example, gives you experience points for exploration -- you can clear a few dozen levels just by travelling between rooms and enjoying the sights (it has over half a million rooms). Discworld has some experience points for speechcraft, and Materia Magica is primarily combat-oriented but has a bunch of experience points for misdirection and chicanery.

I started a MUD in Lisp. It never reached much past the chatroom stage, though :)

Quote:

Original post by Hedos I started a MUD in Lisp. It never reached much past the chatroom stage, though :)

Most don't, heh.

:P