Mailing List Archive

Brian Harring wrote:

> Guess I'll be the killjoy, and throw in the -1 on it.
>
> Reasons are pretty straightforward (at least to me):

I originally agreed with you, but after giving it some thought I think
it might help.

> 1) Creating such channels is just attempting to shift the problem out
> of sight.

This is true, if you consider the problem to be that
a) we are required to be subscribed to -dev
b) we don't want to spend our time sorting the signal from the noise
(where noise is defined as politics or non-technical debates or rhetoric)
c) such noise kills developer interest and motivation and generally
makes us frowny-faced.

Shifting it out of sight is kinda the point. We've already tried
(extensively) to make people get along together and it's obviously not
working. We need to acknowledge that and try another approach.

> 2) Shifting said problem into a concentrated arena means the incidence
> of idiot conflicts/trolling/needling/whatever is likely to increase

I don't think so. Every rule of conduct that currently applies to -dev
should also apply to -project. It's not OTW, just the non-technical
half (5/6ths? ;)) of -dev.

> 3) said increase means proctors/devrel have more work (meaning more
> random outbursts at the proctors/devrel when folks realize that they
> *are* going to enforce the behaviour rules, and that the outburstes
> can be punished too).

It should probably be made clear beforehand then that these rules are
still in effect.

> 4) look through -dev history; the issue isn't OT discussion, it's
> people needling/harassing/trolling/(chose your verb) kicking off yet
> another "mine is bigger" last word battle on the ml.

By making -dev 'technical discussions only', the vast majority of that
needling/harassing/trolling becomes OT. Now, of course, you can still
have a firefight in a technical debate, but history shows it to be far
less common than in a political discussion.

> Basically, what does this solve? If the intention is to create an OTW
> equivalent for the forums, sure, go nuts, but I strongly doubt it'll
> improve things on -dev.

This is nothing like OTW. Posts still need to be on-topic and we still
need to pretend we actually like each other. ;)

Of course, neither of us has a crystal ball (at least I know I don't),
so either one of us could be wrong.

PS. this thread is a good example of something that would belong on
gentoo-project. ;)


--
dirtyepic salesman said this vacuum's guaranteed
gentoo org it could suck an ancient virus from the sea
9B81 6C9F E791 83BB 3AB3 5B2D E625 A073 8379 37E8 (0x837937E8)

--
gentoo-dev@gentoo.org mailing list

Kent Fredric wrote:
> On 6/12/07, Alexander Gabert <pappy@gentoo.org> wrote:
>> There are others like him and there will be others after him. There
>> were even people doing that before him.
> As with trolls, theres more where they came for, but that doesn't make
> gentoo-ML 'different' to as to how we slay a troll.
>
> I agree with matthias. If somebody's a troll, then you ignore them,
> you make no attempt to reply to their post, and thus escape a
> massacre. Its just that some times people think that by _not_ replying
> to a trollish argument, that somehow the troll 'wins' . I figure, that
> this being a gentoo-dev room, most the people in here have their head
> screwed on and can know what a troll looks like when they see one, and
> act accordingly to the 'leave them alone and they'll go away' policy.
> Otherwise all your reply does is _guarantee_ they've won, because
> nothing you say or do about a troll, bar ignoring them, will make any
> good come to pass. Then either the troll will go away, or stop
> trolling.
>
I agree with you in the general case but not the specific. In this case, I
feel the constant drip-drip effect is depressing. Furthermore, outsiders
read the list and see these melodramatic claims about the uselessness of
the current development process which are never answered. From what you're
saying this is embarrassed silence, as with a senile relative. Until I was
informed of the error, I assumed it was because the troll was actually
right. He certainly has vocal allies on every Gentoo medium. I actually
used to believe their claims on the forums, I am forced to admit.

Normally a troll dies from being ignored as the attention is their reward.
In this case however, I feel the intention is political, to gain acceptance
for Paludis as the one true package manager, since after all Paludis is
useless without the portage tree. (Hence the troll's insistence that
the "ebuild tree" defines Gentoo, when in application terms ebuild is no
more or less than the portage file format. Using the same format to do the
same task in OpenOffice doesn't stop doc being the Word file format, for
example.) Why he can't just fork is beyond me; it's not like access to the
portage tree is restricted, as sabayon can attest.

The irony of the troll's complaints about Gentoo QA when his closest ally is
QA lead, is vintage though.


--
gentoo-dev@gentoo.org mailing list

Stephen P. Becker wrote:
>> So (without a Portage tree) it replaces the oldgrown single-liner
>> wget foo; tar -xzf foo; cd foo; ./configure; make; make install
>
> Are you implying that there would be much more involved with anything
> currently in the gentoo tree in the absence of portage?
>
> /me cracks the bell
>
Er the discussion was about paludis without Gentoo ebuilds, not upstream
software, or Gentoo without its package-manager(?!) If it's so great
and "The Portage tree is not the only package repository out there..." why
not prove it with a whole maintainable OS install using Paludis and zero
Gentoo ebuilds?

Personally, I'd do Paludis for sourcemage, although I don't know whether
anyone would want to switch from the approved package manager on that
distro either. Still, since it's so amazing, I am sure you would be able to
prove it was better, and it would win on technical merit.

(BTW posting links to an external website's code when specifically asked
about algorithms on a developer list is bad form imo. It presumes on the
time of your audience, some of whom actually work, and might have
intellectual property constraints on whose code they can read. In future
please just outline the algorithm for the issue at hand, if you have one.)


--
gentoo-dev@gentoo.org mailing list