[Members] XSF @ 10

jehan at zemarmot.net jehan at zemarmot.net
Sat Jul 16 08:05:52 UTC 2011


On Fri, 15 Jul 2011 21:20:23 -0600, Peter Saint-Andre
<stpeter at stpeter.im> wrote:
> First, I agree with Jehan and others that it's not clear what, if
> anything, XSF members can or should be doing. I'd like that to be
> clearer, and I'd like it if members didn't feel the need to join teams
> with leaders and such in order to get stuff done. I don't think that
> structures like the Communications Team have been successful. The
> Infrastructure Team is more successful, perhaps because it predates the
> whole idea of official teams.

About this specific issue, I think too we could get rid of this old
"team" stuff in its current form of "sub-organization, permanent, which
does all and everything (in some very generic area) with its own
hierarchy". As I introduced before, we could replace this by a more
temporary logics of working groups, easier to create and dissolve fast
when needed, dedicated to a specific task.
The team logics is just too generic in my opinion, and in the same time
too limited and constraining.

Note that they could remain in some form, probably with some relation
to the more specific working group. But not in the current "do all and
everything" form.

> Second, I can tell that our community is losing energy. To some extent,
> XMPP is going into maintenance mode. We're not working on so many
> extensions any longer because we've defined a lot of extensions over the
> years and we don't have that much more to define. In addition, IM is no
> longer the hot thing, most developers and end users are more interested
> in things like social networking and microblogging, and anything based
> on XML is seen as a stodgy old technology that's just not cool anymore
> (don't worry, the same fate awaits JSON, just give it time).

It is not seen as the "hot" thing as an end itself, BUT it is used
everywhere. All these new "social networking" are using IM through a
form or another. And for several of them, this IM part is done through

As for me, I would say that's good. XSF should not want XMPP to be an
end. This is a tool, like others, and what we do is making it useful,
making it a goal by itself. In my own opinion, people should not have
care or even know what is "behind", nor how things work. That's also
I think all the debates about naming so puerile. Who knows (I mean,
amongst non technician people) that behind the web is
http/html/css/etc.? That should be the same for IM. The future is not
that people speak of XMPP; the future is simply that people use IM as a
daily tool (whether for enjoyment or work) and when they meet, they
don't ask to acquaintances their XMPP or Jabber address. No they ask an
IM address, like one ask an email address. Our job, in my opinion,
should hence be to make all the IM addresses interoperable and the
network open.

> So I think we need to plan for a smaller community of developers. What
> happens 3 or 5 or 10 years from now when the XSF has only 20 members?
> How do we elect a Board and a Council at that point?
> Paradoxically, I continue to hear a lot about developers and companies
> and other organizations *using* XMPP, but that doesn't mean that there's
> a lot of work for the XSF to do, because they're using it as a stable,
> mature technology. If we've done our job well, people will be able to
> use XMPP for many years to come. However, it's not clear to me how we
> need to structure and position the XSF to keep working in that kind of
> environment.

I think, but that's maybe just me, that we can still do a lot of
things. It is true, many companies use them as a stable technology. But
that's in fact because they actually use only the stable part, which is
far not the most exciting one. They use for IM and related parts, but
all know XMPP can be much more: collaborative work (white board or
collaborative document writing, etc.) as well as gaming, conferences,
file sharing, integration in another environment to improve the whole
user experience, and so on. And when they use it for something
that's often limited to the company and the advanced features never
really get out in other implementations and in the federation.
In this environment, I think we could set a pace, and give directions.

For instance something which was said is that jabber.org could be much
more a state of the art. Ok so I know that jabber.org is not a XSF
service (I know this comment is going to come). But we can't deny it is
closely related to the XSF (changes to the services must even be
approved by the XSF board!) and even from the inside this is still
blurry. I heard that jabber.org should soon host a jingle node. That's a
very good example. But I am sure we can do more.

Also things like interop events could have importance. We can't set a
pace for client side, and what is a problem is that so many clients have
various incompatible implementations. Interop events could really
improve this a lot by really helping developers knowing where their
implementations lose compatibility with the next one. And that needs to
be real a interop event with test cases well thought which will return
actual useful debug information (I read the result of the 2010 event, I
wondered if it actually benefited to any implementation with data they
did not already have).
Statistics on the the user land could be very interesting, to
understand what people use the more (service, software, features, etc.)
and what we fail to provide reliably.

Also some things could be improved with money (servers, events, etc.).
Fundraising could definitely be improved as I understand that the XSF is
not that rich. Not only fundraising from corporate sponsor (I still
don't understand the state of the XSF with sponsors, and I am inside the
XSF. So how do you think it is perceived from outside?), but also from
individuals. If we were to promote better the work we do, I know some
individuals would like to help. And in any case, we don't risk anything
by trying. That's also a topic I proposed some time ago (1 year maybe?)
and the topic once again just disappeared in the members@ list black

In any case, I don't think we should content ourselves with current
XMPP state. It is spread and used in many companies, yes. But that's
still a pretty basic and badly interoperable XMPP we have out there.
Really I can see that XMPP is not that well perceived (even in Free
software's community) and receive a lot of critics.
So yes XSF definitely still has a lot of work to do.

> Again, I'm mostly thinking out loud here. This conversation has been
> helpful, at least to me. :)
> Peter

I think it is helpful for everybody, and hopefully for the XSF as well!


More information about the Members mailing list