What makes Debian a "distribution" rather than just a random
collection of miscellaneous software is integration. This is an
integration wishlist. It has to happen at the distribution level if
it is to happen anywhere. I don't understand why you want to close
this issue---the logic seems to be just that it's hard to address, but
that doesn't seem like a very compelling reason. And I don't see how
it can be filed against any (or many) particular packages---it is an
issue that requires policy and coordination between many packages, and
hence belongs on Package: general.

--Barak.



--
To UNSUBSCRIBE, email to debian-devel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org

Hi Barak,

On Saturday 20 September 2008 21:09, Barak A. Pearlmutter wrote:
> What makes Debian a "distribution" rather than just a random
> collection of miscellaneous software is integration. This is an
> integration wishlist. It has to happen at the distribution level if
> it is to happen anywhere. I don't understand why you want to close
> this issue---the logic seems to be just that it's hard to address, but
> that doesn't seem like a very compelling reason. And I don't see how
> it can be filed against any (or many) particular packages---it is an
> issue that requires policy and coordination between many packages, and
> hence belongs on Package: general.

As said, feel free to reopen. And, as usual.., are you willing to work on this
goal?

Also IMO it's not really about integration (yet). First, some general plan and
then an implementation needs to be found, maybe within freedesktop.org, maybe
not. Then, packages would need to be changed to implement this plan. And then
a mass bug filing could be done, if there is consensus on debian-devel@ that
this should be done. In the bug report I read no consensus _and_ no activity
since 5 years.

But, as said, feel free to reopen.


regards,
Holger

> As said, feel free to reopen.

Will do.

Debian has made analogous contributions in similar domains: see
package libpaper, or tempfile(1) in debianutils, or emacsen-common.
So it not just our mandate; we even have a history.

> And, as usual.., are you willing to work on this goal?

Willing? Yes. Able? No, not realistically: too many other plates
spinning in the air. But I'm happy to bounce things around with
people.

> ... First, some general plan and then an implementation needs to be
> found, maybe within freedesktop.org, maybe not. Then, packages
> would need to be changed to implement this plan. And then ...

Exactly. There are a number of high-level approaches that could be
taken, and it is unclear which would be most sensible. (Library, like
papersize? Executable, like tempfile? Location network server and
protocol?) But at root, we should recognize that this is not a very
hard problem; merely a tedious one. Like many design decisions, it
would be good for a group to talk through some options in order to
come up with a minimal but extensible (eg, accept and/or broadcast
location via avahi) design appropriate to this particular problem.
That's what debian-devel is for, at its best.

Then the real hassle starts: rejiggering all the plumbing to hook
things into a shiny new system.

--Barak.



--
To UNSUBSCRIBE, email to debian-devel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org

Hi Barak,

On Sunday 21 September 2008 19:13, Barak A. Pearlmutter wrote:
> > As said, feel free to reopen.
> Will do.

And you did :-)

> Willing? Yes. Able? No, not realistically: too many other plates
> spinning in the air. But I'm happy to bounce things around with
> people.

If there are people who are willing and able to do the work... that was
actually part of the reason I closed it. Appearantly for years there was no
one to tackle this.

(And if there is no progress again within the next two years or so, I'll
likely close this again.)

> Exactly. There are a number of high-level approaches that could be
> taken, and it is unclear which would be most sensible. (Library, like
> papersize? Executable, like tempfile? Location network server and
> protocol?) But at root, we should recognize that this is not a very
> hard problem; merely a tedious one. Like many design decisions, it
> would be good for a group to talk through some options in order to
> come up with a minimal but extensible (eg, accept and/or broadcast
> location via avahi) design appropriate to this particular problem.
> That's what debian-devel is for, at its best.

But no one did. Maybe it's worth having a BoF about this at next DebConf?


good luck & have fun,
Holger

> (And if there is no progress again within the next two years or so,
> I'll likely close this again.)

Guess we have different ideas about what "wishlist" bugs are for.
My attitude is they're for wishes, like the sea is for fishes.
Sometimes someone picks one up, perhaps even a big wily old fat one.
Maybe takes it on as a summer-of-code project, or whatever.
It might swim around for many years, that's okay. They form a pool of
ideas, and sometimes someone fishes around and finds one they want to
take to heart.

--Barak.



--
To UNSUBSCRIBE, email to debian-devel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org

On Tue, 23 Sep 2008 21:35:23 +0100
"Barak A. Pearlmutter" <barak@cs.nuim.ie> wrote:

> > (And if there is no progress again within the next two years or so,
> > I'll likely close this again.)
>
> Guess we have different ideas about what "wishlist" bugs are for.
> My attitude is they're for wishes, like the sea is for fishes.

Wishes should still have some possibility of attainment, otherwise it
is wishful-thinking not wishlist. (Subtle difference, at least to me
- and Holger by the sounds of it too).

> Sometimes someone picks one up, perhaps even a big wily old fat one.
> Maybe takes it on as a summer-of-code project, or whatever.
> It might swim around for many years, that's okay. They form a pool of
> ideas, and sometimes someone fishes around and finds one they want to
> take to heart.

Issues that get no response in years, despite all the changes that
happen between the releases that occur within that time, should just be
considered as 'dead'. They had their time, nobody thought they were
good enough ideas to be worth investing any significant amounts of
effort. If it was a good idea, the bug report is still there, it is
still archived. Someone can reopen it *IF* they can make time available
to turn the wish into a proposal.

If ideas get positive feedback and the bug report has lots of
discussion, maybe it is worth making a Wiki page for the idea (as long
as the discussion has moved beyond painting the bike-shed).

--


Neil Williams
=============
http://www.data-freedom.org/
http://www.nosoftwarepatents.com/
http://www.linux.codehelp.co.uk/

Hi,

On Tuesday 23 September 2008 23:10, Neil Williams wrote:
> > Guess we have different ideas about what "wishlist" bugs are for.
> > My attitude is they're for wishes, like the sea is for fishes.
> Wishes should still have some possibility of attainment, otherwise it
> is wishful-thinking not wishlist. (Subtle difference, at least to me
> - and Holger by the sounds of it too).

Yup.

> Issues that get no response in years, despite all the changes that
> happen between the releases that occur within that time, should just be
> considered as 'dead'. They had their time, nobody thought they were
> good enough ideas to be worth investing any significant amounts of
> effort. If it was a good idea, the bug report is still there, it is
> still archived. Someone can reopen it *IF* they can make time available
> to turn the wish into a proposal.
>
> If ideas get positive feedback and the bug report has lots of
> discussion, maybe it is worth making a Wiki page for the idea (as long
> as the discussion has moved beyond painting the bike-shed).

Exactly.

Sorry for the low signal here, but I thought I should state this as I have
been dealing with those general bugs a lot recently. I do consider the
general bugs everyDDs bugs though and if there is consensus to leave such
bugs open forever, I'm happy to let them be. I just think what Neil
summarized above... :-)


regards,
Holger

P.S.: please respect the reply-to: header and don't cc: this bug on replies.