[Netwyrm]

Hi Mierza,

What you have described sounds a lot like the project I jumped into (four? five?) years ago deciding to write my game world. I settled on Java as my language to develop in, because it was very C-ish but didn't require manual memory management, and the lure of "write once" was pretty compelling.

In the intervening time I've abandoned using Java on the client--it's no testament to my rather weak programming skills but I had a dreadful time getting consistent behavior on the Mac and under Windows on the client side--not so much because of the Java language but because it was impossible to predict which incompatible virtual machine a given user would have installed on their machine!

(In the interface, I actually once engaged in a lengthy chat with a user in Romania who said they had two identically configured PCs, and one ran the game flawlessly while the other wouldn't even load it!

Anyway, the odd memory aside, I still use Java on the server side because I like it and it runs very well in a controlled environment where you can set up the parameters, but I'd strongly recommend looking into another client approach! It is very hard to get people to download a JVM if they don't have one that works already, and doing so usually breaks Java apps that already run for them, which can leave you with unhappy users asking for support to "put it back"...I have settled on a pure-DHTML client, which although not "perfect" is adequate for my rather limited display ambitions.

The other advice I would have to share is limit your art base...I started off with far too many races, and even after removing several am sorely afeared I will never complete the icon sets!

If I weren't so heavily invested in the current codebase (rewritten from scratch twelve times to date as I've learned more and better ways of doing things), I think I would start off writing a text MUD instead of a graphical one...the overhead of a graphical approach is *considerably* higher than text due to the art requirements--suddenly you must model *everything* that could have been created for the user in a few carefully chosen words.

There's my two coppers!

N.

P.S. I've noted that the signal to noise ratio on this board has improved markedly of late, and just wanted to say thanks to everyone for this interesting and informative thread!

[Mierza]

Hey Netwyrm, thanks for your reply!

I'm just curious about how your project turned out, like what things worked for you, and what didn't, etc.

One of my assistant coders specializes in java, so atleast it's good to know I've still got something that can be worked on by him.

As for races, I have about 4, possibly 5. Most of their concept art is already finished. I currently have 3 artists, so the workload of getting art done really isn't a problem. Now the only issue is.. since I guess I'll have to start the code over.. is it realistic for a new coder to come in and do the same job as my former -is it hard to code for the allowance of graphics in a seperate window.

We have this thing where there's the option of a graphical health status bar. Like we have the usual health bars and points for during battle sequences, but we have these way cool outlines that highlight different wounded areas and indicate what kind of injury it is. That's just an extra for RP value and skills.. like doctors are able to asses injuries and help treat them realistically. I really don't want to give these thoughts and efforts up so easily, especially since most of the effort put in was hours of work from dedicated team members. I doubt anyone else in my position would, either.

Uh.. anyway, back on track. Actually, I've not much more to add. Other than I totally agree with you on the thanks part -When I started this topic I didn't think it'd get so big, but the amount of experience and information shared here is awesome to read, and it really helps me learn alot more too.

[Kylotan]

Mierza, indeed anything can be accomplished with time and a bit of hard work, but these particular ideas require a bit more than most. As I mentioned before, the best feasibility test for someone who isn't a programmer is to compare the requirements to existing systems. If you're asking for something new that isn't done anywhere, that's a potential problem because there's nearly always a reason it's not been done yet. Or, look at it from the other side; what you're asking for are many of the features in a massively-multiplayer online game... and they usually take large teams of professionals years to produce.

Originally Posted by Before critisizing the 'high goals' please remember that when this project was started, these were realistic

Realistic according to who? Your lead coder who is no longer with you, who spent a year working (as an 'employee' ) and yet didn't leave you code or documentation? If you were paying him, you own that code and those docs. If you weren't, then I still can't see why you didn't arrange to have an up-to-date copy of these things at all times as part of the deal. Definitely something to keep in mind next time around.

You probably also know that you're unlikely to get great coders for your project by posting here. For every 1 coder there are 5 designers such as yourself looking for coders, each of whom thinks they have a great idea. And in some cases, the coder is (or believes they are) that designer and will make their own game instead. So maybe you'll get lucky, or maybe you'll end up with someone who will bail out when it gets tough. Good luck.

Do you have any screenshots or examples of some of these features you mention in action? Like those injury outlines? Or are these still just concepts on paper?

In answer to the questions you posed in your reply to Netwyrm:
"since I guess I'll have to start the code over.. is it realistic for a new coder to come in and do the same job as my former" - well yes... starting from scratch is no harder the second time around, really. You may find some things get finished quicker while others take longer, however.
"is it hard to code for the allowance of graphics in a separate window." - not hard, but time-consuming. You need to develop a protocol for exchanging these graphics and write the interface for displaying them. Java is one platform you could use for this. I personally might use wxWindows. Although I'd also be tempted to see if much of what you want to do can be approximated with MXP so that you can use an existing client such as ZMud for testing and prototyping at least.

[Mierza]

Our project had been in development for about three months when the coder left, and we haven't been in much contact lately, but I am sure that I'll be able to (eventually) get the documentation and that from him. I just don't want to bet on it. Also, because he chose to use RUBY or whatever, I'd be open to whatever language a new coder would decide to use.. that way I have more chance of getting a coder, rather than just saying "You have to know this language". What language would then be best to use, I'll discuss with the new coder, and I'll have him/her read the informative posts here about your experiences with certain languages.

One reason I can think why this project hasn't been done before is.. well most people who have muds just borrow code that someone else wrote, to save time, and the fact that it is already tried and tested. Those that write their own, usually stay with a simple mud, changing the theme, and perhaps a different system of skills, etc, that they would prefer to use. I'm interested in taking MUDing a level higher, offering something a bit more unique that appeals to me.

I am interested in hearing what Netwrym has to say about the progress of his former project that was similar to mine.. just so I can find out, other than time consumption and the excess workload when compared to a normal MUD, what else contributes to the failure of a project.

Alot of the time when working out a feature, the ideas have come from other experiences. For example, there are some MUDs, like Archaea for example, that have their own little client. Is that an example of a java client or..? I'm not too sure.

And definately, a new coder comes and they'll have to report with weekly updates, and actually submit the work to my project manager, especially the documentation. My former coder just has a friendship trust as well, and he co-owned the project so I guess I wasn't too strict on telling him to submit his work. I've learned my lesson though. And hey, the main reason for this project is to get experience, for me as a game designer/producer, for the coders, and the artists too.

Whilst I want a game with some interesting features, they're not all set down in concrete, especially trivial things like video sequences and audio, but if there are people willing (and I've had them for other projects before) to do it, then why not add it in. These things could even just be run in a web browser if needed rather than actually accessable in client if we were to use some other client like ZMUD or MUSHclient. (I'll note that we were planning to use a regular client first for the development of the game, and build the other clients later on in the project.

Another area we were going to look into, is AI for mobs. But that's probably most definately going to be scratched unless I get an AI programmer to work on that himself.


Hmm.. This isn't the only place I'm looking for coders either, I am in the process of writing up a more specific job advert to put up on online gaming magazine's. There are people out there, you just have to be lucky enough to find them, or people that know them, and be persuasive enough to entice them to join you But thanks for the good luck, I'd appreciate some coming my way.

Ah sweet, I just got a message from a r/t contact of my mothers, he's a techy, gonna help me outline the job so I know what a coder needs to do, and to tell him what I want done. Excellente! I might post it here if/when I get it, if anyone is interested.

I remember a time when programmers used to be their own designers, like for games such as tetris and pacman and the like. But now that computer games are calling more for well devised stories and better features, it's my opinion that they usually begin to work with game designers for more involving projects, after all coders have code to worry about and as the demands are higher, so is their workload.

And yep, I have screen shots of the health outlines, and other features as well, although they're still at concept level and may be further edited as we go along. The reasons I hired artists so early in the project were things like: the art motivated my coder as he saw concepts and actual graphics coming up, he was more motivated.. it helped the game progress. And also for reasons like working out how we wanted the game system to go (The health outlines for eg). It's one thing to have it in your head, another to actually see it with your eyes.

I also appreciate your answer to my questions,Kylotan, thanks kindly.

P.S. If you're interested in seeing my concept for the health outline, just lemme know.

[Netwyrm]

Dear Mierza,

Thank you for your questions concerning my project. Let me begin to respond by saying my project is not "former" or "failed" at all...at least , *I* don't think so.

I've worked on it almost every night since it started and continue to pour all my free time into it! I've learned a LOT about many many subjects in the process, including DNS, servers, connectivity, languages, middleware, security, and UNIX, although I am by no means a master of any of these subjects. What I have learned in this process, some of it incredibly painful, serves me very well in my day job and keeps my skills increasing.

I began this project because I had to express my ideas, and a "MUD" like environment just seemed appropriate for what I had in mind. I'd already written a (bad) novel over several years in the setting, and computers with their ability for expression just fascinate me. However, I am a total Bartle "explorer" type...achievers  and killers *hate* my world because, well, there's not a lot to "achieve" or "kill"--yet, anyway!--and there aren't enough regulars for the "socialiser."

I was terribly, terribly naive when I began...I honestly did not know that MUDs existed! My background is pencil and paper RPG. I had just discovered JavaScript in Netscape Navigator 4.7 for Mac, and went completely nuts building something very similar to what has evolved today...only to buy an ISP account as it was working very well and discover performance was *completely backward* over a connection than it was from the local hard drive. Frames didn't load in the right order, items didn't fire because the frames didn't load, etc.!

(I was also naive enough to assume that Netscape for Mac and Netscape for PC would behave in even remotely similar fashion...HA! At the time I thought Netscape, which had a 90% market, would be a stable, safe platform. I didn't know any better and BOY! was I wrong )

From these experiences I learned a hard lesson that I have stuck with to this day:

"Always put the code into a test environment exactly like the real world it will run in. Don't develop on the desktop for a distributed environment."

I maintain a development machine, and a server running my website and the prototype game, and make sure to keep it up to date every single time some portion advances. Over the years, I have kept the engine, in whatever state it is in, through all of its incarnations, available in a build on my site for anyone who was curious to come in and, well, break so I could see what broke! (The link is in my profile, the rules here state don't put URLs in your post although it would seem germane to the discussion here.)

Other mistakes: my setting calls for nine of everything. I really ought to have limited that to two or three...nine times anything is a lot more work than three times anything. Nine noble houses, each with a province to be detailed and pictures drawn was just a colossal error from a producibility standpoint. If this was being developed on a commercial budget, I'd fire myself, really.

The rule here: "Limit yourself. It hurts, but do it anyway. You can put it in later, but once it's in and someone is using it, you can't remove it."

Coders will also have a much easier time meeting a detailed down to the "this happens and then this happens" level than being left to figure anything out while coding. It's also very, very hard to produce such a specification. That's another lesson I've learned:

"Always write down what you are thinking. It's much easier to throw away an unworkable idea in a note than to remember anything accurately a week later, no matter how clear and shining it may seem."

If you can afford to hire artists you will do a LOT better. I've had a number of friends promise to help at various points, but even after putting immense amounts of time into writing up hows and specs to make their jobs easy, none have ever actually come through and produced anything--at this point I find it much easier just to learn how to do it. Now, I am decidedly anti-social, and also a terrible agoraphobe, so perhaps that is just my approach to things, but you do sound like you have a good grasp on the management side with some real clue as to the tasks that need to be performed.

You will definitely gain the "game designer" experience of which you speak, and it sounds as though your vocabulary is already better on the subject, and that you are more well-informed about MUDs than I am! (Really, it wouldn't be hard--but I'm serious when I say these things.)

I would suggest nailing down the language to be used yourself, as the project manager, rather than allowing it to float. As you've already noticed, people tend to drift in their allegiance and interest level, and there are many fewer coders out there who understand Ruby or Python than Java or C, from a practical standpoint of having to replace someone. It takes me about two years to become passably fluent in a language, and maybe I'm just slow, but I've learned enough to be able to make that statement with some confidence. Then there is finding a server where you will be allowed to run the language of your choice, if you do not intend to host yourself. Many commercial hosting services will limit what you may run, your level of access, or insist you provide your own machine for colocation if you want to do "that."

Also, you're going to need expertise in more than one subject to pull it off...there is the server side, client side, and any support systems, each of which takes quite a bit of effort, creativity and time. Do the hard stuff first, because you'll need it to support the softer, more fun parts of development, and you can't stick the foundation underneath the building after it is built. No matter how unfun it is, your game will be only the tip of the iceberg of your systems design.

Certainly it sounds as though you have already learned the hard lesson that your game IS your code...keep it managed and under lock and key. Make backups--complete backups--regularly and don't skip them, ever. And make sure the same goes for your staff.

Now good luck, and good night. I'm off to get to work on--heh--my game.

N.

[Mierza]

Wow! Thank's so much Netwrym for your post, I've found it very informative, and encouraging to know that I'm not the only one persuing a dream for a MUD like mine.

As you may guess, from reading the first post of this thread, I too originally wanted to learn to code my own MUD.

However, I was only wanting to learn enough to take an already made codebase and go from there, to learn just enough to get me started.

Ofcourse then others came along, and since my strengths lay mostly in leadership, design, and production, that's where I naturally positioned myself, leaving coders to code, artists to create art, and learning a little about each in the process of putting it all together.

You sure sound like you had some learning experience, alright, hah! But good on you. I get most of my design type experience from the real world, from novels, from movies, etc. Alot also comes from research, and I seriously suggest that people who are interested, do some research and find places like GameDev. The opinions of experienced game designers and programmers alike, can seriously alter the way you view the creation of a MUD.

I like your tips. Especially the one about limiting yourself. I've seen games with thousands of races, and that's great to offer variety for a more simple, traditional mud. However in cases like yours and mine, I agree that where a more complex system is involved, content should be simplified to begin with. I was thinking of like.. 8-9 races when I first started designing, but ended up deciding on 4, possibly 5 original ones.

Thanks for your suggestion about nailing a language, lol. I do intend to learn enough for me to work efficiently as a game designer. Unfortunately there's no way I could code a game to my standards, and have it meet my standards. You yourself said you've been learning for the last few years. Mmm.. I know there has to be a coder out there somewhere that finds it appealing to be involved and challenge themself with an innovative MUD design? Even if this game of mine isn't to go commercial, this is just one step in the rung of a ladder. Either my company will slowly grow, OR quite simply, my team members can take our finished product, and their development steps and show these to potential employers in the future. ~That's what I intend to do, to help me get game design experience to get into the industry (as well as school education, of course) -and I think that has been one major motivation for coders and artists that I know as well.

I have another point to add to your "Always write down what you're thinking."

My point is brainstorm -An initially 'stupid' idea that can be brainstormed, can turn out to be a great concept further along on the track. Getting together with the rest of your development team (or even just some like-minded friends if you're working solo) can be a great source of fun and motivation for progress.

I think your last section of the post, about backups of code, is extremely important. -You were right, I did learn my lesson the hard way, and I hope that other people can read this and learn from my mistake.

[Kylotan]

Coders are very opinionated with regards to the language they use So I doubt having them read this topic will change anything, but it can't hurt. All I would say is that in this day and age, speed of development is more important than speed of execution, especially when it comes to what is essentially just a server that sends out text to various clients. And especially when you have a team like yours which is ready to get to work, if only the code was in place. With that in mind, rapid development languages such as Python and Ruby are surely a better choice than optimised systems programming languages like C or C++. But it'll all come down to who is available to work for you, I'm sure. If I didn't already have 2 MUDs in development, I'd write you a Python base myself...

Originally Posted by One reason I can think why this project hasn't been done before is.. well most people who have muds just borrow code that someone else wrote, to save time, and the fact that it is already tried and tested. ... I'm interested in taking MUDing a level higher, offering something a bit more unique that appeals to me.

I think you should dig a little deeper and see what is out there. You might be surprised at just how advanced some of the less well-known custom games are. Hit locations, graphics sent to the client, custom clients, embedded scripting languages for builders... all been attempted before. The difference is that few people are attempting all of them in one game, because it's a lofty goal and a lot of work for volunteers. I'm not saying you can't succeed, just that it's very tough. My one piece of advice in that regard is to take it a feature at a time... incremental development.

Originally Posted by For example, there are some MUDs, like Achaea for example, that have their own little client. Is that an example of a java client or..? I'm not too sure.

Achaea is a commercial enterprise and I expect they probably are able to bring more resources to bear upon a problem such as wanting a custom client. I don't remember you mentioning it so I'm guessing you're not paying your team at this stage.

Originally Posted by Alot of the time when working out a feature, the ideas have come from other experiences.

It's very easy to cherry-pick the best ideas from a variety of games, and it's quite obvious that combining them gives you a better than average chance of creating a better game in turn. However the chances of failure increase at a faster rate due to complexities of scale, increased pressure on your developers, and so on. It's a tough balancing act, between The Vision and The Reality.

Originally Posted by Another area we were going to look into, is AI for mobs. But that's probably most definately going to be scratched unless I get an AI programmer to work on that himself.

Don't be fooled into thinking AI is a very complicated subject. There are a lot of simple approximations that can make for very convincing behaviour. Whatever you do, don't let someone come in throwing the terms "neural networks" and "genetic algorithms" at you, claiming that they'll solve all your problems. Good AI is simple; it's just that there seems to be little demand for it when most games use creatures purely for target practice. The main thing is sorting out what you mean by "AI". Break it down into what exactly you're looking for, and then it may be quite easy to tackle.

Originally Posted by I remember a time when programmers used to be their own designers, like for games such as tetris and pacman and the like. But now that computer games are calling more for well devised stories and better features, it's my opinion that they usually begin to work with game designers for more involving projects, after all coders have code to worry about and as the demands are higher, so is their workload.

I don't want to get into a big debate about this (as I used to cover this topic a lot on the Gamedev.net forums), but a lot of designers share your opinion... for obvious reasons, and a lot of programmers do not... also for obvious reasons. It's important to note that many of the best games were developed by someone who was both a designer and a programmer (often with other designers and programmers below them on the team). There are many stories of designers with amazing ideas that are just way too impractical to be coded. At the other end of the scale, there are programmers who write games with boring designs; but then such a game is better than no game at all. By having someone with both skills, you save time on communication (and translation!, you cut out the impractical ideas, and foster the extra creativity that a fusion of subject areas can bring.

[Mierza]

Hah! I like you, Kylotan. I actually agree with most everything you said.

If you ever are free of some workload, feel free to give me a yell, I'm sure I can conjur up plenty lol

You're right -I do need to research more into what has been done in the boundries of MUD's. *adds that to her ever growing list of things to research* -If anyone knows some games of prior example off the top of their head, that'd be great too.

And I'll take your advice on the one feature at a time. At the moment I'm just working on testing the game theory with some guinea pigs in a forum RP, then we'll compare the skill/class system concepts and ensure it's what we need, try it out in an IRC RPG and hopefully if those two stages are successful, I'll have had the code well on it's way for the MUD... and alot more experience.

No, you're right, at the moment my team members are volunteers. Gimme a year or two though.

I understand that Archaea is commercial, I was just curious as to if their client was java or not -I'm trying to get a better idea of what is what.

As for the AI, I was interested in it because of the greater interaction it could offer a player. For example, rather than just having static mobs to be killed, or just as space fillers, NPC's would be more for quests. I've seen some mobs that are programmed, it has it's limits. Wether AI can help here, I'm not sure. But I was enticed mostly by what my former coder said about mobs and their recognition of players. I'll just post a part of what he said.

- AI Module
* Designed to give its full interface to the scripting module to allow for very believable npc's and to make the player feel like they are indeed interacting in a believeable world.


lol -okay, I can see how game designers think that they have their jobs, whilst coders feel that they can design and program. Everyone loves to have their own say, I guess. None the less, one day I intend to run a successful entertainment company, and this is just one rung in my ladder of experience.

I defiantely agree that knowledge of programming is alot more efficient and practical too, but right now I'm just trying to get my project done, and I hope to learn a sufficient amount of the programming side of things to get me through.

Again, thanks for your help.

~Jade. (aka Mierza)

[KaVir]

Originally Posted by I remember a time when programmers used to be their own designers, like for games such as tetris and pacman and the like. But now that computer games are calling more for well devised stories and better features, it's my opinion that they usually begin to work with game designers for more involving projects, after all coders have code to worry about and as the demands are higher, so is their workload.

In every company I've ever worked at or applied for, software engineers are expected to be able to both design and code (as well as the other parts of the software development lifecycle).

Someone who can code, but not design, is typically referred to as a "code monkey".

[Azeroth]

I am not even a programmer and I thought that was funny (nice link to the jargon site as well) *chuckle*

[Kylotan]

Originally Posted by (KaVir @ Jan. 08 2004,06:08) In every company I've ever worked at or applied for, software engineers are expected to be able to both design and code (as well as the other parts of the software development lifecycle).

That's true, but game design means something very distinct from software design in the game industry. The best way of putting it is that they design the game, rather than the program.

[Kylotan]

Originally Posted by (Mierza @ Jan. 07 2004,23:50) I understand that Archaea is commercial, I was just curious as to if their client was java or not -I'm trying to get a better idea of what is what.

Yeah, they have a Java client on their website. No idea what it offers though because I can't run it on my browser for some reason.

Originally Posted by As for the AI, I was interested in it because of the greater interaction it could offer a player. For example, rather than just having static mobs to be killed, or just as space fillers, NPC's would be more for quests.

Ok. You're unlikely to get anything that can fool a player into thinking a mob is another player. But you can certainly make them seem more intelligent. For example, simple pathfinding is one important step (no pun intended) along the way; mobs that actually go places and guards that patrol a given route seem more realistic than those who wander erratically or stand perfectly still forever. Couple that with time-triggered progs and you can have mobs that go to their shop in the morning, go to the tavern in the evening, and go home at night. Or mobs that take a bribe and agree to meet you at midnight in the Square to give you further details on a mission. For a good coder, the pathfinding shouldn't take more than a day or two to implement in C or C++, maybe a few hours in Python or Ruby. But the gameplay rewards can be immense.

Another small touch that might work wonders is a basic reputation system. Kill an elf, and elves will stop trading with you, and kill several, and maybe they'll start being aggressive on sight. And maybe losing favour with elves involves gaining it to a lesser degree with goblins. Again, a very simple mechanic, but so many possibilities.

[GenmaC]

Their java client is simply a telnet interface, as most java clients for MUDs seem to be.