[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [MiNT] gradients,FreeMiNT 1.18



Paul Wratt wrote:

> ok, I still have not had a look at how much has bee done, but I can
> guess it is similar to what I did in 1.16.3, except that the gradient
> is produced as a separate binary that XaAES loads at run time.

Correct.

> I suggest that the easiest way to handle these changes is to forget
> that they are gradient/texture specific and think of them as ANY
> config changes. So XaAES needs to be able accept a signal/protocol to
> load a config, not nesescarily the main one, in which case (using
> gradients as an example) only a limited set of changes will be applied

I'm thinking of using the alert-pipe for this. It could contain one or more
cnf-commands (like lang=nz, or an app_option or whatever) that would instantly be
parsed and executed.


> this has the advantage of changing other XaAES + system variable on
> the fly (without actually restarting XaAES)
>
> I believe that a preview may not be useful on some systems (resource
> related), but if it is implimented, then there is no reason why
> certain apps could not use specific settings (which they already do)
> that include gradients/textures. that is a coding issue tho.. and a
> nice feature..

That's a different story - I personally don't think it would look very
good to have different styles for different windows/apps, but it would
be an extension to the current state anyway.

> before we get that far tho, both XA_XTOBJ.RSC and IMG also need to be
> configurable in .CNF, just as WIDGETS is

Where are they needed?

> Thanks for the screen shots of the gradients Helmut (it may have

You mean widgets?

> If a preview is possible from XaAES, the a preview app would be simple.

Maybe it's possible to pass parts of the grd-file - only change
window-titles, leave the rest and so on.

> app was NOT written in C.. same goes for textures, although specific
> object rendering is not as clear as with gradients..

Should textures precede gradients - i.e. no gradient if a texture is
given for an object?

> My original idea was, at runtime, to be able to load and parse the
> actual c files as used by XaAES at compile time (to build the default
> gradient), but I believe it is probably outside the scope of XaAES,
> unless someone has a simple solution.

Compilers are fast ...

> That idea also required all gradients to be defined in 1 file, even
> though 2 separate parts must be included by to separate c files at
> compile time.. (for those that did not know)

??

> Also remember that XaAES has the ability to know about screen capture
> app, why not preview app as well..

Once the on-the-fly-config is there this would be rather trivial.

> Hope that answer most of your questions

But this does not answer joksas question: How do the gradients actually
work, and what impact do the values in the structs have?

I guess he needs to know this when he writes his style-editor ..


-Helmut