All Posts by frumbert

*yawn*.

My level 45 cleric has been asleep for a year, and has woken up on Order (with slight amnesia) to find the world has changed. I can't remember where to begin again. Where should I be?

A class makes it easier for the programmer to weight things. If you have a system that, when you perform one action like swinging a sword, gain higher ability in that one skill, then it can be potentially exploited. How do you balance someone who macroed sword swinging at an impervious training post and then went on holiday for a week, and came back and found the strongest swordsman in the land? As a class, as your sword skills go up, other things can also go up (or down). Becoming a skilled melee fighter might make another stat go down, such as ranged fighting or magery. Classes make that easier in some cases for the people running the world. In a classless system, any change is a nerf.

I'm open to discussion on why a classless system is a good thing.

A lot of the MMORPG's that I've played have put a lot of effort into the newbie areas, with plenty of information about the game world and how to get started... and then you go off into a big empty world. Often I've had to go offline and onto the web to find out how I'm supposed to continue or to get a handle on how a particular thing works. I'd like a game where there didn't need to be a forum dedicated to discussing how elements of the game worked, I'd like it all to be in-game, obvious and overly abundant.If you can describe something in 20 words, describe it in 120 words and summarise it in 20 words, and give people the option for both. I really like games that can give me more than I ask for.

Are there any softwares out there that generate all the data for a world, given an input image or criteria?

E.g. "create a whole island, with heightmap data with rivers, oceans, lakes, mountains, etc. Make textures based on regional zone settings like tropical, desert, mountainous. generate a list of trees and foliage and terrain clutter points that I can use in my engine. Calculate optimal positions for towns based on a set of rules (e.g. must be near water, must not be on land greater than 1:3 slope, etc) and store those coordinates in a file"

I'm not talking about a "all-in-one" game creator - I want the data itself, not the actual finished product.

I use whatever the computer comes with. If I buy a Mac, it comes with MacOS. If I buy a PC, it comes with Windows.

I *could* hack MacOS to peices or put a different *nix distro on it, but I don't because MacOS does its job.

I *could* wipe off the pc and get Linux or BSD or something installed, but I don't because Windows does its job.

I currently can't buy a PC from a shopping mall computer store that runs Linux of any distro. I guess that says something about what the consumer is after...

I have nothing against Linux.

Genre?

Gameplay?

Is it a text adventure?

There are plenty of specific topic game creation sites - but what kind of game do you want to make?

Easier to work everything as if it were flat (like the server does). You can then calculate your position on the world which would then determine things like the time of day for you at that position of the world. Remember, all this stuff takes place on your client machine, not on the server, so your machine could make it night at your spot but a different person elsewhere on the planet is still seeing day because that's the time at their position in the game world. You could also simulate things coming up over the horizon by scaling down the height of objects taht are further away, so they seem to "grow" slightly as you get closer, the way a mountain does when you're driving towards it. That, combined with clever fogging, would probably convince most people that the world was round.

1) Write a text based MUD. It's a game, it has game mechanics that you have to work on. It has to have a server side and a client side (write both, don't just use telnet). If you write yourself a port server and a protocol it will give you a good start in programming experience for MMO's. There's no graphics engie worries to get sidetracked in.

2) Don't let people say "don't use this language or that tool" because you have to try these things out and see what works for you and your idea. They don't know your idea or how you work, so ignore them. They might have more experience than you at writing games, or they might have no experience at all and just want to hand out advice.

3) Don't take an idea you have no intention of following through and put it on a public open source repository like SourceForge. They are stuffed to the gills already with dead-ended projects with no releases and often no partial sources. The sites exist to benefit the community. I hardly use SourceForge anymore because it's becoming difficult to wade through the mles of crap that people put there one day and then never come back to. Uploading unfinished files to such repositories on the belief that people will look over your work and provide support and critique is a thing of the past.

4) Treat programming like playing a musical instrument. You start out sounding really bad but with practice and determination might get better. Be prepared to make games that don't work, or just make little applications that demonstrate a principal and show it to your buddies.

5) Don't listen to people who quote large sums of money or state that you need huge teams of people. Yes, writing a whole game (even a single player game) by yourself is bloody difficult, but there are in fact a LOT of small game companies out there who do just that - and release, and sell, and continue writing.

---

Originally posted by Jorvall
i've been programing for about a year now and still myself a beginner. i'm working with Visual Basic. i've just been learning the commands and putting together basic windows programs. is it possible to make a game using this language.
If not. how much differnet is DarkBasic. I understand that DarkBasic is geared towards games, is this correct?

---

If you're a vb programmer it might be the time to move to vb.net. It's a fairly big step to take, but a good one, as you get access to all the objects.

DarkBasic is a kind of Basic - not based on VB but with similar language syntax. Take alook at http://darkbasic.thegamecreators.com/ . There are books on darkbasic too - http://darkbasic.thegamecreators.com/?f=beginners_guide is decent for a beginner.

It's pretty common for people to shun VB because a lot of programmers approach vb from another language, expecting to find all the features of C++ or Delphi (which is pascal) in it. Before VB.Net came along these features just were there. VB has had a lot of time to suck as a language which it's done in spades, but that's changed now. VB.Net and C# are very capable new languages, which compile down to near-identical code (C# lets you handle unmanaged code a little easier than VB, so I presume it optimises slightly differently). I strongly urge you to either grab DaskBasic and live within its framework (enough for a lot of people), or look in either one of the .Net languages (VB.Net or C#) for your project before looking at C or C++

[quote]Originally posted by Athon
You obviously failed to read the entire sig.
Athon[/b][/quote]

"Designing Virtual Worlds" - Actually I saw that.

Yes, but fret just a little about them. Horizons is the worst offender I can think of - there seems to be one type of "golem", which has the same mesh scaled up and down and recoloured about 50 times. Ditto for wolves. There's so little physical difference between the different levels of critters that you get tired of them fast. That game seems to fave forgotten the golden rule of modern games:

Changing skins does not a new monster make.

As axala said, a little physical variation in a base model mest goes a long way - horns, boots, a tail, teeth, a third eyeball - these things can all extend a base mesh very quickly and give you a fairly different looking critter.

---

Originally posted by Fluffmcduff
I think it should be more on bugs, not just focused on ants. You should be able to choose what type of ants and then fight other types of bugs with different levels, i have thought of something like this myself.

---

why? why can't the mmorpg be about just ants? They have to dig for water and shelter, they have to hunt for food, they have to go beserk and fight each other to fertalise the queen or get sent out to create a new colony somewhere. Colonies could battle each other for resources or just battle each other for domination.

sounds a bit "final fantasy" to me. Sounds like fun!

HMM.

Why does "science fiction" conjure up images of "space", "time travel" and "ray guns". The story "Dr Jekel and Mr Hyde" was science fiction because it used chemistry to alter biology. It can't happen like in that story, but that's all it was: a story. That's all science fiction is: look at Dr Who or Star Trek: none of the things explained in there are plausible except in context to itself. Things happen in particular ways in these made up worlds which are acceptable within the stories.

So why isn't there a science fiction MMO set in the 1500's, where electricity was invented, or set in the year 2000 where the wheel wasn't invented, or in the year 1-bazillion where the universe has vanished and everybody floats around in a giant jelly mold filled with treacle?

I'm starting to think "SF" stands for "Space Fiction".

---

Originally posted by Athon
If you want to do some graphical games, try cloning one of the retro games. My first graphical C++ program was Pong, and I'm currently planning to do something like Space Invaders for my second.

Athon Solo

---

If that's all you want to do then a good game creator package would be DarkBasic. It's not C++ but it's a package created for writing games in. Always a good idea to flesh out a game in some easy-to-learn language such as this (flash and director and other capable packages which are easy to learn) before getting involved in a hard core language like C++ -- you want to know if the game design will work before you go implementing everything. Do it easy to start, and work up.

---

Originally posted by Axala
This brings up a good point. Some of us should get together and write a guide for what is needed to make a MMORPG. Not a how to exactly, but a guide to show what is needed, in depth, so everytime someone asks we can link to the guide and watch them promptly move to a flash site or something.

---

There are of course books on the subject, each with scarey AI algorithms and network library code in them. Perhaps just link to suitable example pages from a few of these and anyone not scared off by that is probably a little more serious!

Yeah that's what I thought. I'd probably markey such a game to a completely different area - role players. Because its such a role play style game (you start up, you play your role, the game ceases) it would be differently marketed than games whos gameplay style is "you start up, you play, and play forever". I think this style game would have only limited success in todays climate in western lands (US, europe, middle east) , but might find better success in lands where "role playing" is more important sometimes that the game itself (japan, korea, india).

I like these ideas, but they need to be fleshed out a hell of a lot more.

---

Originally posted by Axala
1. Give the NPCs a routine. Don't just have them stand in the same spot for 300 years. The come to work at a certain time, go home at a certain time, have family and kids they can talk about. Maybe visit a pub 3 times a week. This is just a better thought out static NPC but adds to the game.

---

What happens to players who are seeking the NPC to get or finish a quest and find him gone? They ask around and people say he's gone home. The player would have to either be able to go there and seek him out, or wait until the next day. This tends to piss players off - in some MMO's when an NPC vanishes during an update (happened a lot in early Horizons days) players would be left stumped waiting to find NPC's to complete quests which could in turn time out, invalidating the quest. I agree that a routine would "add more to the game" but I don't think it should be an AI routine.

Perhaps better placement of said NPC in a location that you would expect to find him, and then perform simple client-side AI to move him around, make him get bored and sit down, mumble to himself, walk around in circles, and so on. If the AI was clientside, the NPC could be positioned in the same physical location on each client machine, but in effect instanced across each client machine. Interactions with the NPC would still go back to the server. The NPC would still always be in his "area" but he would look more animated, bored, whatever. It would give more believability to the NPC without having to change the server side code at all. A less intrusive update.

---

Originally posted by Axala
2. Now give the NPCs desires and wants. They want a better house, or to have kids, or to take a vacation, and when the opportunity presents itself they can fulfill those wants. (say a merchant wants to give his daughter a good wedding and so needs cash, doesn't have it; then some theif comes by to fence some gear. Normally the merchant might not do this, but here is an opportunity to get that $$ he needs so in this case he does).

---

Yikes! I've been working a database that can handle needs and desires type AI and it sure gets complicated fast. The more complicated the more easy it is to introduce bugs, and sometimes complex systems such as desire systems can raise bugs that only exist as emergent behabiours. The issue with want/need based AI is in it's balancing. Take a monster mob, for instance. You have to create a pattern so that the mob wants to attack players, but also another desire to stay in a given area. These two allow the mob to run around in its desired area, but then also attack players if they get too close. You may also want the mob to run in a pack (using a swarming algorithm usually), so you introduce another desire to the mob so that it wants to stay close to other mobs (but not too close). But when the player attacks a single mob, the swarming code makes the other mobs get close to the attacked one, and they all start to attack the player. This is an obvious example, and the solution might be to space the pack out more - but it illustrates the point of one desire or need affecting another. Sometimes you don't see any problems until several or dozens of behaviours are taken into account for a single action, when suddenly the NPC goes crazy - this is an emergent behaviour and is very hard to control. Also because you are taking into account so many factors it is very CPU intensive. You can't really share behaviour process blocks between units either because everything is taken into account as a whole for each NPC.

Secondly, in your "merchant needs to marry off his daughter" scenario, what happens if a) the merchant no longer gets any business because something elese takes the business? or b) the merchant makes enough to marry his daughter off? If you simply state that that was his overall desire then players will pick up on it quickly and find it no more believable than the existing "NPC standing in the one spot for all eternity" problem.

---

Originally posted by Axala
3. Give the NPCs a friends and enemies list. They can then remember who they like and hate, in fact if they hate you enough they will come after you. Talk about interactive.

---

Better still, if they hate you enough give another player the quest to kill you. Likewise, if they love you enough hand ot quests to reward a given player. Unfortunately, what happens if the NPC hands out quests to find someone who suddenly quits the game or doesn't log in as frequently as the person who received the quest?

---

Originally posted by Axala
4. Allow the above and environmental concerns to rewrite their routine. Now you don't have a static NPC anymore but one whose very nature changes over time. And there we have a much smarter, more interesting NPC. Nothing hard in there at all.

---

Environmental concerns change over time, so does the NPC change with it? Its not "hard" to conceptualise these systems. The only "hard" but is the coding (ask any AI programmer!), and that's unfortunately expensive.

sounds good - except i'd have to watch 3 and I don't be doing that until it's on its second repeat on free-to-air television.

It's a pity with a site like mmogchart that SOE have chosen to obsfucate all figures related to games by lumping them all into a single system. The figures of the games become inaccurate. It can only have been done to satisfy a manager - they like figures across projects to be the same, and SOE have guarenteed that at least from a reporting point of view this will be the case. SWG certainly doesn't have the numbers reported by this report. In the excel chart I'm surprised by Horizons, which doesn't seem to have reported the number of subscribers for the last several months - maybe they don't want to scare the last remaining number out of the game.

AI is great when it works, and in a well-planned game it can be fantastic - but it's also hideously CPU intensive. You might think that while most MMO's might already have a pretty expensive rendering pipeline as it is, having to cache and control the world, the players you see around you, etc, it would still have plenty of time to do AI (because you have a really fast computer). But AI has to take place on the server: mobs exist not only for you but also for everyone else at that location in the world. Server side AI is incredibly complex - having to work out mobs in relation with how players are interacting with them (both singularly and in a group). Putting AI down on the client might mean mobs appear in different spots for different players, which could give people unfair advantage in certain situations. The complexity slows the server, which can cause other problems such as lag. It's a balancing thing; as more affordable fasters server become available, the code they can run will also push their limits - AI will get better.