Aye, these are valid points, and there are very real issues with what we have (and don't have) currently.

But why do we need a separate organisation to address them? Why can they not be addressed at home, on mw.org? Why hasn't any of this consolidation happened already, before adding another level of complexity?

Some comments inline, too, but they're less relevant to my main point/question.

On 04/08/14 16:27, Derric Atzrott wrote:

This reply is as I understand the situation to be. If anyone else can provide a bit more insight into things, or correct me where I am wrong, I would be grateful.

...

Hi,

What would be the purpose of this organisation and separate community, exactly? Has there been any demonstrated need or even want for such an organisation amongst the community it would proportedly serve?

This is something that has been in the works on Wikitech-l for some time now. The folks on Wikitech-l and the WMF have come to the decision that they can't really dedicate the resources required to handle third parties properly (someone correct me if I am wrong on this).

I thought that was what the release managers were supposed to be doing. I could be wrong.

They are going to be working with the folks over at Debian, among other places, to get the vary Linux distributions packages up-to-date and keep them up-to-date.

They are also going to be giving the installer a little bit more love than it currently has, and working on additional database support. The WMF is paying them for their work. Its a contracted deal to offload some of the development and release tasks that primarily benefit third-party users to someone who can actually dedicate the time and effort to listen to third- party users.

But is there any reason these would need to be off mw.org etc and out of the existing communities? Why they would need an entirely new community? How would that help?

Given the luck that enteprise has had in the past at getting some features added to Mediawiki (without coding them ourselves that is), I believe that a need has definitely been demonstrated. As an enteprise user myself, I personally want this and like the idea, so while I can't speak for everyone, at least one person in the community wants it.

In open source, that's how it works - unless you get lucky or it's fairly generic, you need to code specific pieces you need yourself, or explicitly pay someone else to do it. Developer time isn't free, even if they are donating it.

A new organisation would not change that.

...

I ask in particular because as a third-party sysadmin myself, it's hard enough following all the relevant discussion and information that concerns releases as it is already.

I think the idea is that they will be consolodating these as much as possible. Though, I will say that I think they should keep using this list as the primary list for enteprise instead of having their own mailing list as well. Unless the enterprise list is to be deprecated.

It would be nice to just be able to subscribe to two lists as a third-party user, mediawiki-enteprise-l and announcements-l.

If you have custom extensions, large scale, and/or like to keep up to date, wikitech-l is pretty much a requirement too, since architectural changes and whatnot are discussed there, and they will affect how you do things even if you don't have a particular stake in the outcome.

If you do, however, the only way to ensure your voice is heard is to speak up for yourself, or hire someone directly to do so. This is unfortunate because some discussions (such as the Architecture Summit) can be particularly time-consuming and potentially costly, but my experience there and on various lists (with the WMF, with the Marks, and others) has been that nobody else will speak for you unless they have an explicit reason to do so. Someone else paying them to do so is not such a reason, because people answer to the people who pay them, not random people out on the sidelines.

...

Adding another organisation on top of that, with its own lists and websites to check and follow, and another layer of community to go through to get things upstreamed, seems highly premature when we can't even consolidate the basics (release notes, date announcements, even testing) at home.

I suspect that they will be more able to help third-party users get things upstreamed. Or at least that is my hope.

They should also be handling release notes and date announcements entirely now for third-party users (assuming I've understood correctly). I think that this will lead to more consistent and easier to understand release notes and announcements.

I can't really say anything about testing. I'm not terribly familiar with any of our testing infrastructure to be honest.

These things can easily be done without a new organisation and community. They just take time and resources regardless of where they wind up.

The new organisation and community would also take time and resources. On top of that.

...

Considering we also have no guarantee that any new organisation would be more receptive to the needs and concerns of the third-party end users than the WMF is currently, and there would still be things we would need to go to the WMF directly about anyway (thus making it even harder to figure out where to go for something), I find this all very worrying.

They are being paid to be more receptive, so I hope that they will be.

I should hope that they will also be able to act as a liason between third-party users and the Mediawiki development community.

Personally, I don't find it worrying, I actually find it quite refreshing. Its about time third-party users were treated as first-party citizens! :D

Thank you, Derric Atzrott

Aye, we need better. But it's been a year already with a much more contained scope, and even that has proven far more difficult than those involved expected, with subsequently little progress. In light of this, extending the scope seems highly premature.

Hence worrying. I want the better to actually happen.

-I