[Python-Dev] Enhanced tracker privileges for dangerjim to do triage.

[Steven D'Aprano]

On Tue, 27 Apr 2010 12:59:26 pm Stephen J. Turnbull wrote:
> Steve Holden writes:
>  > Yes, in the last year in particular there has been some excellent
>  > effort of maintaining the issue tracker content. But the question
>  > still remains - who are we worried about offending?
>
> In this thread, we did worry about offending Sean and dangerjim.  Now
> that Sean has commented, I don't think anybody is worrying about
> offending anybody; there is an understanding that there's a process
> issue to be resolved.  The question is how best to build the
> community.
>
> There are two camps, the quantity camp ("low cost of contribution
> means more contributors, and that's good"), and the quality camp
> ("more interaction within the community, especially of experienced
> developers with newcomers, means more effective contributors and
> that's good").

I'm not sure that is a relevant division between the two camps. I think 
both sides recognise the need to increase the number of contributors 
without compromising on their quality. I haven't heard anyone say that 
we have enough people to do the work that needs doing, or that we 
should take any warm body who can spell PC.

As I see it, the two camps are divided purely on the question of how to 
get increased privileges. Both sides agree that merit is a requirement, 
but the disagreement is on how to prove you have such merit.

One side insists that the only way to prove merit is to go through a 
period of untrusted, unprivileged contributions, with no exceptions. 
One argument for this is that unless we treat everyone identically, 
some would-be contributors will see the process as nepotism, or 
otherwise be offended that they didn't get the "special treatment".

I believe this misses the point that we *don't* treat people 
identically, nor can we in a meritocracy. People are treated 
differently according to not just the quality of their patches, but 
also the speed at which they submit them, and even more importantly, 
their ability to gain recognition for their merit. It's not enough to 
be good at what you do, people have to know it. Ten high-quality 
patches for high-profile bugs in a week may get you enhanced 
privileges, while thirty high-quality patches for low-profile bugs in 
six years might not, simply because nobody notices you.

The other side says that a second way of proving merit is through 
reputation and having a trusted contributor vouch for you, and offer to 
mentor you. The major difference here is that it's not mandatory to 
prove your merit to the entire community, but sufficient to prove it to 
a single trusted member of the community who is willing to stake his 
own reputation on your ability to perform.

Suppose the PSF were to hire somebody specifically to work on patches in 
the tracker, chances are they would be somebody well-known to the 
community. But suppose they weren't -- suppose the hirer read dozens of 
CVs, performed interviews, and determined that the best person for the 
job was somebody utterly unknown to the community, somebody who had 
been working quietly away doing brilliant things in Python but with no 
public profile, and offered her the job of committing patches. Would 
she get elevated privileges immediately?



[...]
> *By definition*, a community is not diverse in the most fundamental
> sense.

I think you're using a definition of community that doesn't appear in 
any dictionary I'm aware of, nor do I understand what you mean by "most 
fundamental sense" of diverse. Talking about diversity within a single 
community is not an oxymoron.

> As long as Pythonicity is important to Python, there is 
> danger as well as opportunity in more rapid influx of newcomers.

This at least is true. I can't dispute that.



-- 
Steven D'Aprano