> Sender: [EMAIL PROTECTED]
> From: Juliusz Chroboczek <[EMAIL PROTECTED]>
> Date: 18 Apr 2002 18:22:13 +0100
> To: [EMAIL PROTECTED]
> Subject: Re: [Xpert]limitations of TinyX?
> -----
> SP> Is there any paper anywhere that talks about TinyX and the limitations
> SP> behind it?
>
> Not to my knowledge.
>
> Contrary to what some believe, KDrive (TinyX) is a complete
> implementation of X11R6.5.
>
> There are three limitations of KDrive that you're likely to notice.
> First, it's currently Linux-only (but should not be difficult to
> port).
Yes, the port should be simple.
> Second, it doesn't have the XAA layer, which makes writing
> drivers for it much more tedious than for XFree86; whence the small
> number of drivers for it (although for many of the expected
> applications of KDrive the generic fbdev or vesa drivers should be
> enough); in addition, the KDrive drivers tend to accelerate fewer
> operations than their XFree86 counterparts.
Not particularly: experience shows that for most applications, only
a few operations need acceleration these days. The amount of code
to take good advantage of a chip is typically quite small. XAA is almost
certainly overkill (though certainly much better than using the original
interfaces), particularly on the kinds of intended embedded platforms.
And yes, for most intended uses in embedded systems, frame buffer is just fine.
Though having Xv support is nice for DVD's :-).
> Third, it doesn't support
> the XFree86 module loader.
Yup. Both a feature and a bug...
- Jim
--
Jim Gettys
Cambridge Research Laboratory
Compaq Computer Corporation
[EMAIL PROTECTED]
_______________________________________________
Xpert mailing list
[EMAIL PROTECTED]
http://XFree86.Org/mailman/listinfo/xpert
