Re: signal-to-noise on d-d-l

[Luis Villa <louie ximian com>]

On Mon, 2004-02-16 at 23:53, Jeff Waugh wrote:
> <quote who="Luis Villa">
> 
> > Jeff, I don't think your proposals really solve the problem, they just
> > diffuse the problem over more lists, as far as I can see- most people
> > will still need to read all of them if they want to understand what is
> > going on, and wade through the attendant crap. Is my gut sense on this
> > wrong?
> 
> Not over more lists -> we shift the 'global' discussion to gnome-hackers,
> and let d-d-l stay on-topic for desktop issues. It makes no sense that the
> general development list in "The GNOME Project" is d-d-l.

Given that GNOME is a desktop, yeah, it sort of does. I'm not sure how
you draw a line between the two for the kinds of things that have been
causing so much wasted bandwidth, like flags (required for a desktop
module), evo, or scripting. 

>  It also means that
> if you want to be involved in decision making in a certain area of the
> project, you have to be involved in the list, instead of getting an end run
> around everyone simply by being on the cool kids list.

Eh?

> > * d-d-l becomes subscribers-only for posting, with a readonly list. [I
> > know the admins hate this, but... I think it's important in this case.]
> 
> d-d-l already is subscribeers-only for posting (just about all GNOME lists
> are). But a big hearty NO NO NO to any list elitism via readonly lists and
> stupid crap like that. You're essentially proposing to dunk d-d-l into the
> same fetid mess that gnome-hackers/gnome-private have been, and ruling out
> contribution from sensible people, simply because there are some noisy
> useless twats around the place.
> 
> PLEASE let's not use a technical solution to a social problem.

I don't consider this a technical solution- it's a
trust/relationship-based solution, as opposed to 'anyone who can operate
mailman is qualified to comment on GNOME', which is the current state of
things.

>  If someone is
> posting regularly to a major project list, and only providing useless noise,
> TELL THEM privately, and if they persist, TELL THEM publically.

We've done that to several people lately, with no discernable effect.

> No more barriers! I've proposed removing barriers from gnome-hackers.

I'm fine with removing barriers from g-h, as it doesn't currently have a
horrendous noise problem. However, I think that if you intend to move
'substantive' discussion there without thinking about the noise issue on
d-d-l, you'll also get crossover noise on g-h.

>  If we
> add more to a project-specific list like d-d-l, then we're just going
> further down a terrible path. We started on that path when GNOME was having
> serious problems - let's keep solving problems, instead of sticking our
> heads in the sand.

I'm attempting to solve a problem, please don't tar me with that brush.

> > * we make sure the moderators are active- I volunteer myself for this,
> > since I read all the crap anyway, so it can't really make my life worse,
> > per se, and hopefully it'll make it better for others.
> 
> Moderators are good generally.

Was probably unclear here- didn't mean to imply they aren't doing good
work now. Just that if we ask there to be more work involved in
moderation and/or list membership, there would need to be more work and
energy on their part, or more moderators.

> > * all foundation members (someone else has already done the screening,
> > right?) get subscribed to start, and after that the bar is low-
> > moderators should accept subscriptions from anyone who has done even a
> > modicum of work for any module, inc. hacking, translation, bugs, docs,
> > marketing, a11y, UI, etc.. Enough to show that you've paid some
> > attention to GNOME and made effort to play nicely with others- nothing
> > huge, so that we make our best effort to encourage commentary and
> > contributions from new contributors.
> 
> Foundation members may not have anything at all to do with development on
> the desktop... 

They have to contribute to the project, which means contributing to the
desktop.

Look, I'm not really excited about this idea/proposal, and agree that
solving things through more direct social pressures is better. But that
hasn't worked lately, and 'we're all on one list' is not the reason why
it hasn't worked. [It's a problem, sure, but it's not the cause of this
problem.] So... why hasn't 'regular' social pressure worked?
Insufficient stick? Do we need to ask moderators to go Bowie on people
more often when they aren't useful? The current standard is that you
have to be a total asshole, not just totally clueless- is that standard
too high? Being applied the wrong way? Dunno, but if you want to go
about things a different way, I think that is the question that needs
answering.

Luis