1 |
I think on some platforms having another set of X11 libraries and headers |
2 |
inside prefix is redundant. |
3 |
|
4 |
X11 is very conservative technology. Recent changes are in the field of direct |
5 |
hardware interaction, but I believe it is not possible to use DRI inside prefix |
6 |
if X11 implementation in prefix is not compatible with host one, so it could not |
7 |
be reason for need of keeping updated X.org version inside prefix. |
8 |
|
9 |
On Linux host we already have X.org, maybe different version than in portage |
10 |
But API and ABI is keeped stable between releases. Mac OS X >= 10.5 also |
11 |
provides X.org. |
12 |
|
13 |
Some platforms provide alternative implementations of X11, and I guess they must |
14 |
be API compatible with X.org. These include Windows SFU, older Mac OS X (<= 10.4) |
15 |
and older Linux using Xfree86, maybe other Unixes too. |
16 |
|
17 |
I propose to add virtual ebuilds for using host X11 libraries instead of building own. |
18 |
Any objections? |
19 |
|
20 |
-- |
21 |
Regards, |
22 |
Konstantin |