Re: [Antbear-devel] Architecture questions

[Adam Ferla]

Apart from the advantages listed in my comment below I can see no other 
advantages in using a pre-made database, except for the saving of time. 
Also, the major issues in using a pre-made database that thorsten listed are 
valid. Making our own database and single threaded client processing seem 
the way to go. The integrated database would not have to be extrodinarily 
complex and Java's concurrency measures would be adequate to support our 
needs (I think).

Cheers
Adam






> From: wolpers <address@hidden>
> To: address@hidden
> Subject: Re: [Antbear-devel] Architecture questions
> Date: 12 Nov 2002 11:57:42 +0100
>
> On Tue, 2002-11-12 at 09:47, Adam Ferla wrote:
> > Hey peoples
> >
> > Just a quick question regarding Thorstens comment about using a single
> > thread to process client request. Are we using a premade database or
> > building a simple one ourselves? If we use a premaid database the 
> database
> > will probably have functionality to allow multiple reads, etc to a 
> database,
> > in which case multithreaded client processing would be an advantage., 
> i.e. a
> > listener thread creates a new thread for each client request. However, 
> if we
> > make our own database then I agree that single threaded client 
> processing
> > would be cool.
>
> I have been thinking about that and so far I like the idea of an
> integrated,selfmade database better.  If you think different please
> teach me.  When thinking about using a premade database a few points
> come to my mind:
>
> - I do not like the idea of creating an antbear that requires a certain
> (type of) database to be installed before it can be used.  So we would
> have to find a database we can integrate into the antbear.  One point
> would be the license, the other one would be the idea that I do not
> really want antbear to require  access to installed files.  I would like
> antbear to be usable 'out of the box' with no requirements, but Java, an
> optional settings file and optional parameters.  So we would require a
> pure Java 'all in memory/swap' database.  The amount of data we would
> store is not too large. (about 30000 server addresses whith details on
> each server: runningmod, name, map, maxplayers, currentplayers,
> playernames).
> -  Would a premade database give us any advantage in performance as
> opposed to a selfmade one?  Might the interface between antbear and the
> database slow things down? I have no experience with that.
> - There was one more, but I forgot 8( .
>
> >
> > P.S. sorry my involvement has been rather minimal lately. I just 
> recently
> > found out that hotmail was sending most of the antbear emails to my junk
> > mail folder (even though it hasnt in the past). This problem has been
> > rectified.
>
> 8)
>
> - Thorsten
>
>
>
>
> _______________________________________________
> Antbear-devel mailing list
> address@hidden
> http://mail.nongnu.org/mailman/listinfo/antbear-devel


_________________________________________________________________
Tired of spam? Get advanced junk mail protection with MSN 8. 
http://join.msn.com/?page=features/junkmail