[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: What will improve Debian most?

On Mon, Mar 23, 2009 at 04:49:21PM -0700, Russ Allbery wrote:
> Anthony Towns <aj@erisian.com.au> writes:
> >     Over the next twelve months, what single development/activity/project
> >     is going to improve Debian's value the most? By how much? How will
> >     you be involved?
> Maybe you meant to ask "what are you going to work on over the next twelve
> months that will improve Debian's value?"  Because the right answer to
> "what one thing is going to improve $FOO the most in the next twelve
> months," for pretty much any value of $FOO, is "ask us in twelve months
> and we'll tell you."

Well, there was that bonus question, which asked about the previous
twelve months, that I haven't seen any answers to. I'm not sure it's
actually any easier to answer, tbh.

Ultimately, I don't see any problem asking what people see as the biggest
forthcoming improvement for Debian though; sure it's impossible to say for
certain, and most people will pick something that's somewhere between the
second most important thing and the hundredth. But that's still useful
and interesting; and finding out what other people are focussing on in
preference to what you think is most important is interesting no matter
who ends up being right.

FWIW, I chose the focus ("what's important, how will you help" rather
than "what are you doing, why's that important") deliberately. Maybe the
best thing for Debian is something that you can't work on directly --
but it can still help to identify that thing, both so the people who
are working on it feel appreciated, and so people who could help if only
they were aware of the challenge become aware of it and do help.

> Personally, here's some things that I plan on working on over the next
> twelve months that I think will improve Debian's value:
> * [...]
> None of these are exponential and I don't care about percentages with any
> of them.  I'm okay with that.  I don't think those are useful measures,
> and I don't think exponential growth is a useful or interesting goal.

I thought about this some more, and wrote up some of those thoughts on my
blog at [0]. YMMV obviously.

I'd still be interested in seeing how much of an effect you think those
changes will have (and for whom), FWIW. 

It'd be nice to know what people's actual priorities are these days, in
some way that's independent from people pointing blindly at the social
contract at just repeating "our users and free software". Are Debian
developers mostly interested in doing things that help themselves? Other
developers? People who file bugs? People in their LUGs? How about Ubuntu
or Eee or Mi users? Desktop or server users? Embedded stuff? Upstream
developers? Something else entirely?

> One thousand people all improving the parts of Debian that they care about
> produces incredibly impressive results.

Yes, that's completely true. To double Debian's usefulness, 1000 people
only need to individually contribute a change that improves Debian's
usefulness by a little less than 0.07%. (1.000693387..^1000 = 2) If you
somehow had a thousand people that each increased Debian's value by an
appreciable amount, say just 1.5%, you'd get an overall 292,443,586%
improvement -- that's higher than Zimbabwe's annualised inflation rate

Which is to say, definitely, and I don't think that sense of things --
that a lot of people each contributing in small amounts adds up amazingly
-- is incompatible with making numerical estimates.

For comparison, my estimate of the value of a small speedup on m68k
was a 0.2% improvement, which is already quite a bit better than the
0.07% above.

If it turns out Debian's aiming for a thousand 0.07% improvements,
there's no point identifying which is the 0.071% improvement and which
is the 0.069% improvement; but getting an idea of the scale is still
useful. Personally, as was probably implicit in my examples, I think
there's at least a handful of 10% or better improvements that could
happen over the next year or two.

On Mon, Mar 23, 2009 at 06:04:47PM -0700, Russ Allbery wrote:
> I should probably note here that it looked like Anthony had carefully
> phrased his question to apply to the entire project, not just the DPL
> candidates, and I replied in that context.  If it was intended as a DPL
> candidate question, er, never mind.  :)

It was something that I think DPL candidates ought to be able to provide
some interesting answers for; but I'm more interested in the question
(or its answer) than in how particular people happen to respond to it. In
other words, thanks for the answer. :)


[0] http://www.erisian.com.au/wordpress/2009/03/28/exponential-growth

[1] http://www.guardian.co.uk/world/2008/oct/09/zimbabwe

Attachment: signature.asc
Description: Digital signature

Reply to: