MYSQL and Worldgroup

Malakai · Post by **Malakai** » Tue Dec 02, 2008 2:07 am

First of all, is there any way to get the btrieve database files converted to a format mysql can read? There is a program called Conversion Tool which cost like $299 that may convert btrieve to xml (and possibly convert it back to sql?) but that's a bit pricey.

Has any one ever tried to install MYSQL as the worldgroup database? If any one has had any success, I'm sure a few of us would like to know how.

Questman · Post by **Questman** » Sat Dec 06, 2008 3:24 pm

You could use an ODBC-based driver like Powerbase or whatever it was called, but ODBC is slow and not optimal.

You can't just convert the Btrieve data files - they are not stored like a SQL database would be; there's no table definition to the files - the record structure of the underlying program defines how to parse the data, which is stored in a binary format. Unfortunately, that means that a conversion program would have to be specifically written for each file.

One of the things I was doing was trying to SQLify the databases - that's the idea behind the SQLite integration. The advantage of SQLite is that it is embedded, so it's fast, like BTrieve, while using a standard database format. The disadvantage is that you can't setup a database architecture separate from the software.

Making a RDBMS available wouldn't be difficult, but it would require re-coding of the programs that would wish to be able to use it.

Malakai · Post by **Malakai** » Sun Dec 07, 2008 2:11 pm

OK.. So, odbc stuff seems out for now..

Back to authentication.... Since worldgroup can create a textfile with all user signup information on it, with no binary or encrypted stuff, how hard would it be to get programs that use sql to read that file and use it for all of the authentication?

I know, at minimum, you'd have to take out any signup information from the external sql programs (forums, logd, bnt, phaos, whatever) and all signups would have to go through worldgroup.

On the top line of userlist.txt, it basically tells you what the fields are between each quotes.

Questman · Post by **Questman** » Mon Dec 08, 2008 3:11 am

It could be done... it'd be better to create an addon that supported standard SSO techniques.

Stoneslinger76 · Post by **Stoneslinger76** » Sun Apr 05, 2009 10:17 am

Questman wrote:It could be done... it'd be better to create an addon that supported standard SSO techniques.

Any progress on GALSQL.DLL or some form of sql compatibility?

Questman · Post by **Questman** » Mon Apr 06, 2009 2:06 am

The DLL works, and it's primitive. I'd like to make the routines more user friendly, I'd like to create a C++ object version of it for smarter coding, and I'd like to performance/battle test it.

If you think this is something that would be of value I could focus on it as a sooner term thing...

Gardner Denver · Post by **Gardner Denver** » Mon Apr 06, 2009 2:37 am

A fresh MajorMUD install can be done using Pervasive SQL as the database manager versus btrieve but that is only the actual mud data files not the WG files.

I can provide information if interested.

Questman · Post by **Questman** » Mon Apr 06, 2009 3:20 am

What would be better received -

(a) embedded SQLite databases (plus: fast like BTrieve, but in industry standard SQL format; minus: file-based, ultimately, and no networked db)

(b) mySQL support (plus: could use embedded version potentially OR networked version; minus: slower, requires separate Db install)

(c) either-or (plus: flexibility, easier for add-on developers to choose which they prefer; minus: take longer to do)

Stoneslinger76 · Post by **Stoneslinger76** » Tue Apr 07, 2009 10:18 am

Gardner Denver wrote:A fresh MajorMUD install can be done using Pervasive SQL as the database manager versus btrieve but that is only the actual mud data files not the WG files.

I can provide information if interested.

I have tested the Pervasive 9.5 with pervasives btreive .dat's updated to a new version so pervasive can extract info from the .dats. I had wg running nicely on the updated dats (only for test purposes) and created many of the tables for mmud items list's etc but couldnt get pervasive to communicate with mysql. Unfortunatly one of the harddrives on my test server died and havent got back to it yet in hope of Questman's completion of galsql.dll

I will see what you have over their.

Stoneslinger76 · Post by **Stoneslinger76** » Tue Apr 07, 2009 11:15 am

Questman wrote:What would be better received -

(a) embedded SQLite databases (plus: fast like BTrieve, but in industry standard SQL format; minus: file-based, ultimately, and no networked db)

(b) mySQL support (plus: could use embedded version potentially OR networked version; minus: slower, requires separate Db install)

(c) either-or (plus: flexibility, easier for add-on developers to choose which they prefer; minus: take longer to do)

A: Probably the most compatible to worldgroup but doesnt make for an easy interface for the self educated wanna be programmer. The non networked db would be a drawback.

B: I had this running thru xampp, but didnt seem to be all that reliable. Even with just a message base running in it would eat up system resources and crash for no apparant reason while idle. Maybee the dual p3 600 cpu's and gig of ram were not enough (win 2k server).

C: Either- OR I like that concept

this would be the best for overall for addons a sysop wants to add. You could have the ability to login with WG's user records. Be able to run many of the opensource projects and other sql oriented programs. Also setup intergrated scores/help docs and other usefull information for the WG games and WG modules. Getting the various sql's and db's to work in one interface would be a challenge to say the least.

I am curious as to what other sysop's thoughts are.

Stoneslinger76 · Post by **Stoneslinger76** » Tue Apr 07, 2009 11:24 am

Questman wrote:The DLL works, and it's primitive. I'd like to make the routines more user friendly, I'd like to create a C++ object version of it for smarter coding, and I'd like to performance/battle test it.

If you think this is something that would be of value I could focus on it as a sooner term thing...

I would be interested in testing it as it would give us a better idea of what direction is best.

We could start with option A and then move on to more complex configurations as needed.