Programming question: Geek speak

[Bonegavel]

A question for Marc V., Jim G., Marcus H., or any others that program for a living:

How do you handle the database aspects of your games? I refuse to use an mdb, and have researched c++ collection classes to handle the massive lists and whatnot, but would like to know how the pros do it. If you can share, I would appreciate any help in this regards.

The sweetest thing would be to have a scaled down version of a SQL engine to power the data, but that would most likely require royalty fees etc.

gracias

[primelord]

Originally posted by BoneGavel
The sweetest thing would be to have a scaled down version of a SQL engine to power the data, but that would most likely require royalty fees etc.

Not if you use a free SQL implementation such as MySQL or Postgresql.

[Bonegavel]

Originally posted by primelord
Not if you use a free SQL implementation such as MySQL or Postgresql.

I should have been a little more precise. If there is a product, it will be a Windows EXE and I want everything internal to the exe, no external database (more in line with the text-sims we all know and love).

MySQL has a mini-engine, but it is a for-sale item.

[Solecismic]

I just condense the numbers, store them in binary files, and if they're frequently used, keep them in integer arrays in RAM while the game is open to maximize speed.

I would only use a database if I needed to conduct complex sorts on a large quantity of data before using it. I'm sure there are places in FOF and TCY where a database would be useful from a purely geek perspective, but I'm still at a point where the longest delays are due to internal processing that can't be done with SQL, and with the actual display of the data.

[Ksyrup]

Originally posted by Solecismic
I just condense the numbers, store them in binary files, and if they're frequently used, keep them in integer arrays in RAM while the game is open to maximize speed.

It's reading things like this that gives me comfort that I've chosen the proper career path.

In other words...wuh?

[Castlerock]

a) MySql is free and runs on Windows. It is quite good.
b) Microsoft SQL Server is a full-blown DB package and is expensive.
c) MSDE is a scaled down version of SQL Server and is free. It has a throttle but it is good enough for even fairly large apps.

[Fritz]

Originally posted by BoneGavel
A question for Marc V., Jim G., Marcus H., or any others that program for a living:

How do you handle the database aspects of your games? I refuse to use an mdb, and have researched c++ collection classes to handle the massive lists and whatnot, but would like to know how the pros do it. If you can share, I would appreciate any help in this regards.

The sweetest thing would be to have a scaled down version of a SQL engine to power the data, but that would most likely require royalty fees etc.

gracias

not that this is the way to go, but it a free stripped down version of sqlserver (7)

http://msdn.microsoft.com/vstudio/do...e/default.aspx

[Castlerock]

Oh... I missed your requirement that it be entirely self-contained.

I think you are stuck using language data types. Arrays, collections, etc. Another alternative to storing data in a binary data files might be XML.

[mrsimperless]

XML is the way I am going with my game, but I am developing in .Net, and everything (at this point) is completely hosted. I don't know about storing raw XML files on a client machine... I guess if people try hard enough they will hack anything.

[Bonegavel]

Originally posted by Solecismic
I just condense the numbers, store them in binary files, and if they're frequently used, keep them in integer arrays in RAM while the game is open to maximize speed.

I would only use a database if I needed to conduct complex sorts on a large quantity of data before using it. I'm sure there are places in FOF and TCY where a database would be useful from a purely geek perspective, but I'm still at a point where the longest delays are due to internal processing that can't be done with SQL, and with the actual display of the data.

thanks Jim (and the others).

So you read from disk most of the time? Seems like that is a lot of disk I/O, but most of the files in the universe folder appear small. As far as the larger files, are those read in all at once, or is the data flagged where you only load portions?