Monday, September 22, 2008

More Cards, More Errors

I saw the statement somewhere that said, “Maybe you should fix the current card errors rather than trying to add new cards.” I figured other people might also have this question, so I might as well answer it.

My overall goal is to add as many Magic cards as I can and if sometimes they don’t work correctly, oh well. I know some people might say that I’m taking the easy route and I certainly am. Testing a card for me means that I might play it once or twice to see if it generally works. Sometimes I don’t even check to see if the computer can play it correctly, I just hope it does.

The problem with more extensive testing is that it means fewer cards. If I test each card in five different situations, I’m going to be bored and program fewer cards. I also get a thrill by adding new cards and debugging old ones are pretty low on my agenda. For version 2, I plan to be more strict in my card implementations.

Version 1 has a good number of cards and also a fair amount of errors. While the errors are occasionally annoying, I think my time would be better spent on version 2. So I personally think that version 1 is “good enough.” If someone wants to improve version 1, go ahead, you can join my Google Project or start your own project and use my source code as a starting point.

Version 1 has many fundamental errors that can’t be fixed like the computer can’t play instants during combat, no X spells, no “removed from play” triggers, and a million other things. Version 2 is a clean slate that incorporates all of the things that I learned while programming version 1.

5 comments:

Gando the Wandering Fool said...

Yeah I dont think anyone expects perfect code the first run out with new cards. Thats why the client is perpetually in beta test mode. When we find the errors we tell ya (usually).

Anonymous said...

I agree with Gando. In a way, we are all unofficial betatesters on this project. ;-)

Anonymous said...

actually the whole project would become easier if you share the annoying work, make your already awesome work even better by making it moddable.

Meaning. If you would allow .xml input as a card database then the forge would have a complete database of all cards in no time. People would spend their time in order to create such a database for you.
(in fact it is quite easy, but a litte tedious, someone needs to write a few macros in excel in order to let excel extract the information from spoiler.txt files and then export the same into .xml)
I already started such an attempt and I could finetune it.

BUT it would be your job to make the engine understand the keywords and this is the more interesting part any how.

Hardcoding each card IS tedious.

Forge said...

Database, lol, funny. I don't have no database, maybe later. With version 1 I was just trying to get something working. Some many Magic projects have been started, by smarter people, but they never get working at all

Gando the Wandering Fool said...

"Meaning. If you would allow .xml input as a card database then the forge would have a complete database of all cards in no time. People would spend their time in order to create such a database for you."

Welcome to this blog (check out the forums for more info input). The Open Magic Data project that was started years ago did this very thing. Hogan Long and David Hibbs were the captains of that project (http://omd.sourceforge.net). Ive mentioned this several times already. :)

As Forge has said he is not improving Version 1 and is busy working on Version 2, however there are others working on Version 1 still. (see the forum.)