Re: roadmap status update/update request

[Alan Horkan <horkana maths tcd ie>]

[skip to the last two paragraphs for the improtant bit...]

On Tue, 8 Mar 2005, Alan Cox wrote:

> Date: Tue, 08 Mar 2005 10:26:36 +0000
> From: Alan Cox <alan lxorguk ukuu org uk>
> To: Eugenia Loli-Queru <eloli hotmail com>
> Cc: desktop-devel-list gnome org
> Subject: Re: roadmap status update/update request
>
> On Llu, 2005-03-07 at 18:21, Eugenia Loli-Queru wrote:
> > >A feature will be implemented if and only if there is a developer who wants
> > >to implement it, regardless of the number of votes it's received.
> >
> > Which is exactly why open source software will never replace commercial
> > software or Microsoft.

(pedantic: Open Source can be commericial software, better to say
proprietary software)

> I'd beg to disagree on this specific point. Commercial vendors get
> motivated by customers to fix actual problems and they then pay people
> to do stuff they might not otherwise have done. The community provides a
> base from which to do that and a place to give back the results but it
> will always do the bits it cares about.

Many volunteer developers do fix bugs that do not particularly interest
them.  If a project is healthy and well managed the lead developers can
cleverly delegate the minor less glamourous tasks to beginners who want to
learn.  Unfortunately the patience and management skill required to do
this is often more costly in the short term than a developer simply doing
the work themselves.  (I'm not saying anything new or insightful here, it
should be very familiar to anyone who has read about software management.)

If they were truely writing software purely for themselves would they
really release the source?

> The size of the developer pool (and thus of course how you treat
> newbies) is really important but most of them will do the stuff that
> matters to them.

Todays newbies are tommorrows developers.
Problem is you have to be just as polite to the misguided few who will
never become useful developers or you will scare off the talented few.

It doesn't take an experienced developer to recognise how successful
projects are managed.  With a bit of care and an optimistic "Yes, but ..."
approach enthusiastic users you can directed to do something helpful.

Clearly Eugenia is going to do some sort of poll no matter what anyone
says so it makes sense to try and tell her what you would like her to do
instead of what you dont want her to do.

> > You make Gnome sound ellitistic, and that's bad PR business to start with.

(did you mean elitist?  my pointing out a spelling mistake probably only
serves to make me seem elitist :( but I had to point it out, as I writer I
expect you dont mind.  I expect I've misspelled more than a few things in
this message.)

> Thats what technical people I interview keep saying btw. "Gnome is
> arrogant".


> Its dissapointing but I don't see the whole voting thing working very
> much - if it stays a way to help developers see the priorities then
> great. That might give a few people a shock too as to where the users
> idea of urgent begins and what it is.

I dont seem the harm in enabling voting in bugzilla so long as we can
_manage expectations_ and make it extremely clear that developers are are
almost all volunteers and will of course decide to work on whatever
interests them most.  I think if we have a reasonable caveat that is clear
about how unreliable web voting systems are and how any voting results
will only be treated as a polite suggestion it can only be an extra source
of information.

I didn't see Havoc's posts as a rejection of the idea in priciple but more
that he has very real concerns about potential side effects if such
feedback is not managed correctly or if users get unrealistic
expectations.

Sincerely

Alan Horkan
http://advogato.org/person/AlanHorkan/

Offlist replies are preferable, I'd just like to see voting enabled in a
way that wont have the cataclysmic effect that Havoc predicts.