Re: lock down features

[Luis Villa <louie ximian com>]

On Thu, 2002-11-14 at 07:36, Glynn Foster wrote:
> Hey,
> 
> While I'd love to see some of this stuff going into 2.2, I think I would
> prefer to concentrate getting the various GNOME preferences a little
> saner in terms of -
> 
> 	o GConf notification
> 	o sensitivity of widgets with non-writable configuration
> 	o handling --break-key, --recursive-unset etc..
> 	o making it easier to setup default preferences *

Absolutely. The panel gconf usage is... painful to look at. And I don't
even know much about gconf. :)

> because I'm not sure if lock down really makes sense until we've got the
> other pieces of the puzzle solved, although I'd love to see this all
> happen.
> 
> I'm definitely in agreement with the 2.2 schedule worries

I'm still not terribly worried about the 2.2 schedule overall; I'm very,
very worried about specific components. This is /especially/ true of
panel, which is unusable and untestable for me right now.

>  and given the state of 2.1.x recently

Agreed. See (for starters):
http://makeashorterlink.com/?C1AA22272

Note that many of those are in panel[1]. Basically, were my panel
functional at all right now, I'd have very little problem with
shoehorning this in past feature freeze. But it's not. It crashes
whenever a launcher is used; it crashes whenever I try to use the window
list applet.  

>  I think we probably have other things to sort out that, perhaps, are
>  more important.

Agreed. Give me a stable panel and I'll look the other way on panel
features. But I don't have one.

Or let me put it another way. Someone said something along the lines of
'Luis will throw a fit at this idea.' I am willing to throw a fit. But
it's not because of some generic notion of a feature freeze, which is
what the original email implied my motivation would be. I'm flexible on
that and have even requested at least one exception from the release
team for an app that I considered stable and sane and able to benefit
from a minor feature addition. 

The fit would be because the proposed new features are being added to a
code base that's not (as far as I can see) stable or ready or /even
moving in the right direction/, which is _exactly_ what the feature
freeze was supposed to prevent.

Luis

[1]Nautilus's new icon theming has also broken more things than I care
to think about, and the new theme stuff also scares me. FWIW.