Mail archive

Re: [alpine-devel] RFC: Project goals and guiding principles

From: Chloe Kudryavtsev <>
Date: Sat, 30 Mar 2019 18:42:40 -0400

On 3/30/2019 5:53 PM, Oliver Smith wrote:
> Regarding "Do the right thing over doing the written thing". What is the
> written thing? If it is the rules outlined in the initial post, then I
> don't see a valid reason to go against these rules.

We're in the process of trying to re-organize internal structure.
This is done for a variety of reasons (for instance, due to the internal
process, PRs can take exorbitant amounts of time to get through) - and
the goal of the re-organization is to improve that.
The path towards that has ended up being through a change in governance
model (decentralizing efforts and avoiding bottlenecks, among other things).

While this process has been public (in that everything that happened is
available to the public (including comments), and a few callouts have
been made), no full/finalized proposal has been sent in for ratification

In order to expedite the conversation, we decided to add project goals,
as well as guiding principles.
This way parts (or, ideally, all) of the proposal could be justified
through them, thus defeating any arguments ahead of time.
The example given was someone that would push for harsher conditions
under the impression that that would create higher efficiency.

In short, the "written thing" would be the (longer) contents of the new
governance and organization proposal.

> Chloe also wrote:
>> On 3/29/2019 2:35 PM, Natanael Copa wrote:
>> The axioms involved are as such:
>> - There is always a correct course of action, based on the project goals.
>> - It is possible to identify this course of action.
>> - This course of action could, in theory, require breaking an existing rule or policy.
>> Under the latter conditions, the idea here is to allow people to do said correct thing without (necessarily) waiting on consensus.
> IMHO that makes it more complicated than necessary. I don't think that
> the Alpine team would deadlock itself because of a rule they made
> earlier that does not hold up anymore. So I would rather remove the "Do
> the right thing..." line.

One of the side effects we want (and foresee) of the re-organization is
a higher influx of contributors *with* git access (but not necessarily
git access to most/everything).
While the bar for inclusion would be relatively high (decent volumes of
prior good work), once you have access you would be expected to handle
PRs, which inherently would require a sort of decentralized decision making.
This *can* cause deadlock (mostly because it's not a "single" alpine
team anymore), and the idea is to allow for an acceleration of resolving
potentially problematic situations, in which it's the obviously correct
thing to do (even if, say, it breaks a specific policy in an unexpected

This has, however, been the most discussed and controversial principle.
For more details as to where and why it came from, see Drew Devault's
latest email (he was the one that suggested its inclusion).

Received on Sat Mar 30 2019 - 18:42:40 UTC