modular -> monolithic

[Keith Packard]

On Tue, 2008-01-22 at 21:28 -0600, Pat Kane wrote:

> In the good old days RWS  did that.

Where by 'good old days', you mean before 1988. After that point, the X
consortium staff (including RWS and me) were paid by the corporations
doing the development, and it wasn't (yet) clear that the corporations
shouldn't have a direct say in what went into the distribution.

Thus we suffered through XKB, XInput, XIE, PEX, VEX, Xt and a wealth of
other really bad ideas, along with the constant requirement that
'nothing old stop working'. Shockingly enough, the only things which
survive unscathed from that era are things which were *not* developed
using this model (Shape and Xv being the only two largely intact and
non-trivial extensions that we use today).

It's really only in the last couple of years that we've taken a strong
stand against some of those mistakes, along with an acceptance that
we've still got a lot of catching up to do. Having spent the last 8
years adding a bunch of missing functionality, we're only now starting
to work on areas of the system which cannot easily be fixed in small
incremental steps.

One thing the kernel never faced was fifteen years of fundamental
stagnation with a wealth of kludge-arounds piled on top.

-- 
keith.packard at intel.com
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 189 bytes
Desc: This is a digitally signed message part
URL: <http://lists.x.org/archives/xorg/attachments/20080122/a4a554b3/attachment.pgp>