Hi,

Sorry to reply to my own post but regarding advantages of a type 2 driver -

Mark O'Dunohue wrote a while back

>Unfortunately Im more linux inclined, I hope someone else can help - but
>classic with type2 jdbc driver is I think a *VERY* good idea.

I guess there is a good reason why a type2 driver would be good with CS -
but I dont know what it is - so there are probarbly further advantages and
disadvatages that i'm not aware of.

Thanks
Ryan

-----Original Message-----
From: Ryan Baldwin [mailto:ryan.baldwin@...]
Sent: 04 March 2003 18:30
To: Firebird-Java@yahoogroups.com
Subject: RE: [Firebird-Java] Another set of Class 2 Driver modification
files. These ones are pretty much complete and stable(-:


Hi Rick,

>Can you give us a rundown on what you had to change for this. I thought
>your mods were for the Java JayBird side only.

They are only on the Java JayBird in the existing code. Sorry if my post was
potentially
confusing. The c++ code refered to is a piece of code seperate from all the
existing code. This code builds a dll/shared library that exposes a number
of entry points which match up to native methods defined in the java code.
The jvm loads this dll and resolves these entry points when the
corrisponding java native method is called.

This dll in turn loads and resolves entry points on GDS32.dll or equivilant.
Calls to the java native method entry points are mapped on to GDS32 entry
points.

So in short: Nothing that exists outside Java JayBird code.

>What are the advantages/disadvantages of using the type 2 as opposed to

type

>4 in this case?

The reason I wrote this initially was to make it possible to load
fbembed.dll in process with a JVM. Although it turns out that FB may make
use of faster IPC methods then sockets on localhost connections(XNET - FB1.5
??? - FB1.0 ) - in this situatuion it seems there may be a performance
advantage.

The disadvantages are that:

1) The firebird client side must be installed on the client
machine(fbclient.dll - gds32.dll).

2) The native dll with the type 2 modifications must be installed on the
client machine.

3) Foreign native coded is loaded into the jvm which might get boored and
decide its time to bring the process crashing down.(The native code might
crash the vm process).

Hope this helps to clarify things.

Thanks
Ryan

-----Original Message-----
From: Rick Fincher [mailto:rnf@...]
Sent: 04 March 2003 17:40
To: Firebird-Java@yahoogroups.com
Subject: Re: [Firebird-Java] Another set of Class 2 Driver modification
files. These ones are pretty much complete and stable(-:


Hi Ryan,

Can you give us a rundown on what you had to change for this. I thought
your mods were for the Java JayBird side only.

What are the advantages/disadvantages of using the type 2 as opposed to type
4 in this case?


Rick

----- Original Message -----

> Hi,
>
> I've uploaded another set of files containing binaries and source for the
> type 2 driver modifications. These are:
>
> class2drv_bin_0_3.zip
> class2drv_src_0_3.zip
>
> The c++ code has been tideyed up and improved. Many bugs and resource
leaks
> from the previous release have been fixed and it now appears to be usable
as
> a complete and stable class 2 driver. It should 'drop in' anywhere where
> jaybird is currently used and I am not aware of any bugs at present.
>
> Only windows binaries are included - if someones interested in another
> platform I(or somone) could look at building the shared library for other
> platforms but otherwise I will probarbly leave this for a while and look
at
> tideying the java side of the modifications.
>
> I have had some interesting performance results with these modifications
> which I will post about these seperatly.
>
> Thanks
> Ryan



To unsubscribe from this group, send an email to:
Firebird-Java-unsubscribe@yahoogroups.com



Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/





To unsubscribe from this group, send an email to:
Firebird-Java-unsubscribe@yahoogroups.com



Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/