:: [DNG] On talk.do and Web forums
Page principale
Supprimer ce message
Répondre à ce message
Auteur: hellekin
Date:  
À: dng
Sujet: [DNG] On talk.do and Web forums
On IRC #devuan: <golinux> hellekin: Can we PLEASE move on from
Discourse. Very few people are using it and those who are loathe it for
the most part. It's holding back Devuan's community growth!

Let me state it one last time, by email, so that it can be read by
everyone and we can have a discussion about it, instead of a persistent
and frankly annoying counter-productive attitude regarding talk.devuan.org.

No, we're not moving on from Discourse.

Yes, very few people are using it.

Yes, *some* people insist on loathing it, and although I have tried to
address every single of their arguments, they keep coming at it without
even trying (e.g., to *not* use the Web after their account is
registered.) Maybe there's some missing documentation about how to use
Discourse as a mailing list only.

I don't think anybody here as much to say against mailing list apart
from the well-known issues:

- threads casually explode as people's MUA don't keep the thread
reference (e.g., when replying from a digest)
- multiple threads talk about the same thing, adding "where?" to the
archaeology of remembering what was said.
- titles are not consistent within a thread (drifts happen before the
title is changed)
- archives are very much an accumulation of dead letters that require
reading sometimes entire threads to figure out pertinent content.
- mailing lists can get invaded by trolls
- etc.

For each of these issues, Discourse provides an elegant solution:

- 'categories' (I prefer calling them 'conferences') and topics are
malleable and can be transformed at will without losing the URLs.
- posts and topics can be merged into a single *long lived* topic to
prevent many threads repeating themselves
- 'drifts' can be selected-out of a conversation into a new topic, to
keep the discussion consistent and encourage on-topic contributions
- some contributions (preferably the opening post in a topic) can be
made a wiki page so that information accumulated during the time of a
conversation can be turned into accessible knowledge.
- actual participation *over time* provide an effective defense against
trolls, spammers, and other undesirable 'participants' that infest
mailing lists.

*Most of these wanted features* are *not available* in other forum
software, that:

- use CAPTCHAs or external anti-spam services susceptible to harm
people's nerve and privacy
- accumulate an infinity of repeating topics that often end up in
dead-ends polluting search engines with useless contents
- promote a reactionary culture of 'self-expression' and novelty
(because people prefer "opening a new topic" rather than searching for
existing contents--also because searching existing topics is a pain)
rather than a long-term conversation towards collective intelligence and
collaborative knowledge management

So, I understand that Discourse is not for everyone, its web interface
requires some heavy Javascript that can be slow, and it doesn't look
like you would like, and certainly not like any forum out there, where
you just sit and shout. But you need to understand that:

1) we're thinking of a *tool* towards a *goal*: knowledge management
2) it can be *used by email* (as a mailing list, requiring very few
interaction with the Web at all)
3) nothing prevents anyone from *setting up another forum software*

Which brings me to the last argument: "It's holding back Devuan's
community growth!"

Really, golinux, do you think it does? FriendsOfDevuan has a wiki that
is mentioned in the official documentation page on
devuan.org/os/documentation while it's not operated by the VUA, so be my
guest and make a popular forum that will help grow the Devuan community.
I think the two objectives are orthogonal, and certainly not incompatible.

I know you've been arguing the talk.do was a threat to the mailing
lists. Yet, officially, DNG has been replaced by devuan-discuss and
devuan-announce mailing lists, which see seldom traffic so far. That
means we're not in an univocal world where "the VUA decide" and "the
community follows". I urge you to look around and watch the hilarious
Ethereum fork story that unfolded this summer. In fact, we can just
*propose ways*, and some people will understand and participate, and
some will prefer keeping using what's there *and there is no problem at
all with that*. We live in a complex, multipolar world, where *one
vision does not exclude other visions*.

With Devuan we're trying to give another take on what an universal OS
means, and for that we want to have a compact set of tools that enables
more diversity in the expression of what is, how to make, and who makes
"a distro". *In my opinion*, talk.do has an important role to play in
this strategy, as does the devuan-sdk, and Amprolla and build
automation, etc. But I certainly do not support the idea that my
opinion is the only valid one. I'd rather not have *another* forum
software under devuan.org to avoid dissipating energies. But I
certainly cannot prevent the community from deciding that my vision is
moot and go on setting up something that may eventually replace it.
Devuan is not Python: there's no one-true-way here.

<3

==
hk

P.S.: I'm tempted to post this to devuan-discuss and talk.do, but hey,
let's not cross-post :)

-- 
 _ _     We are free to share code and we code to share freedom
(_X_)yne Foundation, Free Culture Foundry * https://www.dyne.org/donate/