Re: [MiNT] Cool new AES features

[Evan Langlois <Evan@CoolRunningConcepts.com>]

On Wed, 2005-07-06 at 20:50 +0200, Arnaud BERCEGEAY wrote:

> Ok, so now you know what we are talking about when we wrote about  
> "gradients" of color like XGEM does.

Gradients was just one of the many issues.  Does XGEM handle radial
gradients?   How is it doing this with the VDI.  

> I hope you agree that we *can* implement that in AES. No need for an  
> optimised powerfull graphic card driver.

The AES can draw whatever the VDI can draw.   Period.  I'm talking about
adding the capability to do that.  You are talking about some library
that hacks it in there so it displays on screen.  I'm talking device
independant general purpose graphics functions.

Oh no, we don't need optimized graphics drivers.  Thats a stupid
statement.

> It's not the problem. I don't want XGEM "as it is" in AES (it's a bad  
> idea).

Exactly what are you proposing?  Some magical library can do it, and so
we'll make the AES do it too.   What about some programmer that wants to
make use of the basic drawing primitives that this "unidentified" body
of code (since you said you aren't using XGEM) makes use of?  How do I
put this gradient in my program.  If the AES can do it, why can't I?  If
the AES draws it, shouldn't there be some VDI commands that do it?
There is no way that using the VDI (as it stands) for random gradient
creation is in any way an optimal solution!

Now - if you have to write some routine to make this gradient, since you
don't have an open-source solution, why not make the gradient routine as
flexible as possible and allow the application-programmer to make use it
it.  What API would drawing gradient go into?  What else can we add that
programmers might like to use?  Imagine!   Be creative!  Dream!

> I don't know about the licence for XGEM library, and it's not the  
> question. The question was about the feasability of gradient with the  
> current situation (VDI layer)... and it can be done. This item can then be  
> add to the whish list, it's not a stupid request.

Ah, so we can do 1 thing on the list.  Forget about planning for the
future.  We'll just hack in that one thing and be done with it.
Nevermind thinking ahead.   We'll just keep hacking in features 5 years
after everyone else does it, and keep things as kludgey as possible.