Why's it so hard to create a character generator that rocks?

[WingOver]

I partly answered my own question. Making anything that does everything is obviously hard, especially if its software and the rules are for D20. But besides that, it seems that the D20 rules are mechanically precise overall and lends itself to be transferred into software better than previous editions. So why hasn't this been done?

Neverwinter Nights and The Temple of Elemental Evil had some serious D20 rules under the hood, even if they didn't quite implement everything ("real-time rules" for NWN, templates, monster races, etc). But it seems that it would only take a few more steps for those companies to create an all-encompassing, slick character generator.

My post isn't intended to slam the exisiting generators out there (e-tools, PCGen, Redblade, Campaign Suite, DM's Familiar and Genie, etc.) I have the utmost repect for their work. However, I downloaded and used these apps but wasn't satisified because:

1. Didn't trust that the stats were properly calculated 100% of the time. If I have to double-check, then what's the use?
2. GUI not intuitive, or too busy. If I know D20, but can't figure out how the software is implementing it, then again, what's the use?
3. Unable/difficult to create new data (races, classes, magical items, etc).

D20 has been out for over 3 years. There's a lot of talented software designers out there. So why isn't there a single, all-encompassing, rocking character generator available yet? Is it too hard to design? Is it a huge project that needs a large team and too much money? Is D&D too "small-time" to bother with? What do you think?

 

[trancejeremy]

Well, as someone who has written a d20 character generator of sorts (I say that because it mostly works, but is far from complete), is that it's a whole lot of work. A whole lot of work. And it's really really boring work. And besides the programmings, there's tons of data entry.

NWN, Temple of Elemental Evil, etc, had large staffs of professional programmers that were well paid.

I mean, the character rules are generally pretty easy. But then you get all sorts of extra stuff, like equipment, spells, familiars, lots and lots of feats. And there are some tricky bits, like Rogue's special abilities - you have to support choosing those.

Plus, the OGL is kinda murky when it comes to software, I think. A big company would no doubt want to license D&D from Hasbro (which is also pretty murky, as I think another company has electronic rights to D&D). The net result is that it's probably not worth the investment.

 

[WingOver]

And it's really really boring work. And besides the programmings, there's tons of data entry.

Try programming software for the electricty industry or accounting. Now that's boring work! Developing a D&D character generator synchs with our hobby at least.

The OGL/D20/licensing issue is a good point. Maybe those big guys don't want to mess with all the legal wrangling. Making a D20 compliant generator is difficult, and using the OGL would be easier programmatically, but lousy for marketing.

Maybe the D&D market's not big enough for them?

 

[frugal]

I partly answered my own question. Making anything that does everything is obviously hard, especially if its software and the rules are for D20. But besides that, it seems that the D20 rules are mechanically precise overall and lends itself to be transferred into software better than previous editions. So why hasn't this been done?

D&D rules like most things in life follow the 80/20 rule: 80% of the rules are straightforward and easy, 20% of the rules are a complete nightmare.

Have you ever looked at the monster advancement rules for 3.0?

A decent PC generator has to cope with 3.0 and 3.5 the differences in rules are subtle and complicated. We have been trying for 12 months to come up with and XML representation of the rules and keep hitting "exception cases".

Consider something simple like "Greater Weapon Focus", you can only take GWF for a weapon that you have already taken "Weapon Focus" for. So you need to do the following:

1. Mark a character as having a particular feat
2. Mark the particular feat (weapon focus) with a particular weapon
3. Be able to examine the user to see if they have the weapon Focus feat
4. Be able to examine the feat assigned to the character to get the weapons that it specifices
5. assign a new feat to the character based on one of those weapons.

And you need to be able to dop all of that without hard coding the feats as a special case. PCGen is the only character generator I have seen that does this and even then it hardcodes some parts of that operation.

You also need to be able to query the character in quite complex ways: "You are only allowed to take this class if you can cast 3 different arcane spells at level 4 or above". That is quite difficult unless you hardcode the concepts of spells and spell levels into the program. It is possible to do it with generic "leveled objects" but you still need a querying language that can extract the information.

So not only do you need to encode the concepts in a generic way, but you need to write a query language that is both generic and high level enough that users can work with it.

1. Didn't trust that the stats were properly calculated 100% of the time. If I have to double-check, then what's the use?
2. GUI not intuitive, or too busy. If I know D20, but can't figure out how the software is implementing it, then again, what's the use?
3. Unable/difficult to create new data (races, classes, magical items, etc).

D20 has been out for over 3 years. There's a lot of talented software designers out there. So why isn't there a single, all-encompassing, rocking character generator available yet? Is it too hard to design? Is it a huge project that needs a large team and too much money? Is D&D too "small-time" to bother with? What do you think?

D&D utilities are written by gamers to satisfy a particular need. 90% of the people who write open suorce programs are not professional programmers, and of the 10% who are they do not want to spend their time writing high level design documents to make it easy for newcomers to the project (they get enough of that at work I can assure you).

In general assume that getting any 2 people to work on a project (they are not being paid for) in the same way is like trying to get an octopus in a string bag. the more people you add to the project the larger the holes in the bag get ;O)

There is no one on the PCGen team (that I know of) who is skilled in Interface design and ergonomics. Professional teams can hire a designer or two to make the product polished, free product have to hope that they can fnid a gamer who works as a designer and who care enough to put up with the developers who just want to hack something together and do not want to spend 3 times as long writing the interface as they currently do.

With regard to editing races, classes etc. As far as PCGen goes the developers do not generally enter data so they do not use the editors, nor do they know what a user wants from the editors. The data entry guys use a text editor because they want to work at a lower level and are entering large lumps of data. So you end up with a few users who try to enter their own classes with no one to support them.

Frugal
PCGen developer
Professional Software Engineer.

 

[frugal]

My post isn't intended to slam the exisiting generators out there (e-tools, PCGen, Redblade, Campaign Suite, DM's Familiar and Genie, etc.) I have the utmost repect for their work. However, I downloaded and used these apps but wasn't satisified because:

Just as a quick metric PCGen has over 900 classes and 300,000 lines of code. To make it slick and user friendly would probably double that to 1,800 classes and 600,000 lines of code.

 

[WingOver]

Thanks for the reply, Frugal. So basically the reasons there isn't an uber-generator:

1. D20 rules are complex and extremely difficult to put into code.
2. It's a lot of work.
3. There is little motivation to overcome 1, 2 without getting paid.
4. Most of the programmers are gamers, without necessarily all the exact skills required.

I'm willing to bet that reason #1 is the primary reason, and that the 20% of nightmare rules are the cause.

 

[Joshua Randall]

I like PCGen, and I use it, but its main problem is this: it's trying to do too much, all at once. I'd like to see PCGen get stripped down to just the RSRD (3.5 rules), and then be tested to make sure everything works. Right now it's iffy even with basic issues like accurate melee attack bonuses or armor classes. And that is unacceptable... or as unacceptable as a free, open-source program for D&D can be.

 

[WingOver]

... or as unacceptable as a free, open-source program for D&D can be.

I appreciate the efforts of the PCGen team. Every hour spent coding/maintaining PCGen is one less hour of free time to play D&D, write adventures, or just relax.

 

[Davin]

I agree with the basics of what others have said. D&D/d20 rules seem neat and clean (and compared to previous versions, they certainly are!), but trying to code them brings out the hidden issues. Basic functionality is trivial (let's guess about 5% of the total workload for about 50% functionaltiy) -- and that's about what you get from most spreadsheet-style generators. Handling the majority of cases isn't too bad (about +10% of the workload for another +45% of capability). But that last 5% of exceptional conditions could easily amount to the remaining 85% worth of effort. And that's just for coding -- data entry and other issues add considerably to the difficulties.

Another reason that you see incomplete systems are that people tend to write their software for just what they need, and that tends to also be among the easiest parts to do. But for a good, general, commercial-like package you have to build in all the little quirks to make a really complete system, useful to anyone and everyone. For the reasons outlined above, few (if any) people have the motivation to take it that far.

 

[Joshua Randall]

WingOver, I appreciate what the PCGen team has accomplished, too. And I groove on the software. I just get frustrated that there are a bazillion non-core d20 sources being added -- most of which break the functionality in some wierd way -- while the core RSRD portion of the program still doesn't work properly. But as Davin said, people work on what they want to for their game. Not what I need. I can understand that, but it's still (mildly) frustrating.

Does anyone have the new, Code Monkey Publishing version of E-Tools? I would hope that it implements the RSRD rules properly?