Tell HN: 17 years on the same game

Are psionists still useless at night?

I never thought about it before, but I now realise how important it is for immersion to let the player imagine their own voice in the game (as in GTA, Skyrim, and Red Dead Redemption).

Allowing players to use their imagination is so important.

[0] https://www.reddit.com/r/MUD/comments/5uzx3z/list_of_current...

Even tiled ASCII characters would be a huge boon; I think it's the fact that I can't easily imagine the entire game area through text. If I was able to at least "see" the squares around my character, I could better spatially orient myself. Almost every time I've quit a MUD has been because I got bored trying to pour through the docs/guides to figure out how I ended up somewhere.

I imagine an ASCII multiplayer game is more complex when it comes to netcode though. Perhaps the reason so many MUDs have survived is because they don't require nearly as much server infrastructure to support?

That's not to say that there is still room for something like this! There are heaps of great indie games on Steam that are lightweight and deep.

If you don't mind graphics, I'd recommend checking out Streets of Rogue and Death Road to Canada, though they are not RPGs per-se and more Roguelites.

If you want something a bit more hardcore I'd check out Caves of Qud, still haven't managed to figure that one out.

Also, see Rimworld, an accessible Dwarf Fortress like game.

Upgrading from ASCII to graphics would mean kicking out all your friends with older modems, those with poor phone lines, those without Win 3.1, etc.

I also think one of the great parts of developing a MUD as a developer is that you don't need to be good at graphics. Just code and writing.

Modern day MUDs contain a large portion of blind/sight-impaired people so moving towards a graphic based game might turn away most of your players. As it's pretty much the only kind of multiplayer game you can play as a blind person.

It was a fun experiment though, I learned a lot about websockets and networked game programming!

I remember trying to get it working for Discworld, but the richness of the world descriptions (which varied depending on weather, light level, and all sorts of other factors), along with spontaneous info messages, made it very hard to extract reliably.

My complete lack of knowledge about parsing & text processing probably didn't help matters, and I learned a lot building triggers and other (mostly tolerated) automation back in the day.

One particular feature/challenge I remember was the UU library, which in order to navigate around, switched from cardinal directions (go east; go west; ...) to relative (turn left; forward; forward; turn right), so you had to keep track of your rotation. Combined with deviously constructed routes, it didn't take much to get hopelessly lost, which as the entire point I think :)

A lot of that subtlety would be lost in a simple tiled map, as would all the (often extremely verbose and entertaining) world descriptions.

Add that to the fact that almost all were player/volunteer-built worlds, and graphical art is a much harder field to dabble in than text, probably explains why graphics-first waited until commercial developments like EQ.

[1] http://www.oocities.org/timessquare/dungeon/6091/automap1.gi...

I'd go so far even as to say that IRE's MUDs made me who I am today. A commitment to integrity of role-playing my characters helped shape my identity while I was growing up, particularly as a social exercise. And my first exposure to scripting in the form of simple aliases and triggers led to the pursuit of programming skills. Without MUDs, I'd be a different person entirely, I have no doubt.

Can you just telnet in and avoid the web based interface completely ?

I would assume so ... then you could start your game in a screen session and just re-attach anytime you like.

Can you, indeed, telnet (or ssh) to discworld ?

The only thing you need is good internet. If your internet connection doesn't tend to completely fail (a small bit of latency and the rare crash shouldn't be the end of the world) then you're honestly good to go.

If you want to run this on a server, you're looking at around $10/mo for something decent, and if you scout around hard you may be able to get something for less. Google Compute Engine quietly let you have 600MB of RAM and 1GB of bandwidth per month for free, for example. :P

--

Your profile mentions PHP and Node.JS. Of the two, Node.JS is probably the better choice to build something like this with - PHP's network/socket handling has a few sharp edges that make it occasionally fall over, and for a social project like a MUD, you'll get death-by-a-thousand-papercuts in the form of "...PHP?!" all the time, so there's that as well.

Obviously Java would work fine too; Node's advantages over Java in a context like this are lower grammatical verbosity and zero compile time.

--

The first thing I'd recommend is poking around to understand how to do both telnet and websocket based networking.

Websockets will handle modern browser clients and cover 99% of a fledgling MUD's connections. Telnet will handle the die-hards who, if they come back, might stay for a while :P

Start with telnet first: you need to decide how you want the text UI to work. You can leave telnet out until you have everything else doing interesting things, but I strongly advise having at least a basic understanding of how VT100 emulation works from the start (while you're still in the design stage), or whatever you build may be extremely difficult to get working inside telnet.

Some web MUD clients just wrap a VT100 session hosted on the server with some websocket trickery and do things that way. That's a good way to do things, because it means people who want to stick with telnet don't really lose anything.

Websockets is just a networking system and is mostly fun to work with; to get started just find a library you like and tinker with it.

--

As for gamedev, I recommend studying existing MUD systems, not with the intention of drowning yourself in how they all work (which will be disasterous), but to get an idea of how they structure their world info - what's where, how navigation works, etc etc.

Ultimately there are (AFAIK) few gotchas and MUD development is quite simple - rooms are connected to other rooms, and things happen - but it's a good idea to think through how you want to architect how rooms connect together so you can eg easily get at the bigger picture, and think about how you want interactions and cause/effect state to be maintained and associated with players.

--

Disclaimer: I'm the author.

It exists on GitHub as well: https://github.com/nickgammon/mushclient

I used to use zMUD (http://www.zuggsoft.com/page.php?file=zmud/zmudinfo.htm), but looks like it maybe never made it past Windows XP.