What I need for OMB

[Evan Prodromou]

Seems like there's a lot of action going on on this project, which is
awesome.

With OpenMicroBlogging, we're hitting a lot of traction: a lot of
popularity, 3 (!) implementations right now, and more coming.

I need: a place to host or redirect the main domain name
(openmicroblogging.org); a mailing list; some tools for building
consensus; ways to publish documents (CMS, say); and ways to publicize
results (blog, say).

Is there any chance we'll be seeing that kind of infrastructure -- I'm
thinking SourceForge for standards -- in the next couple of months? Or
should I build it elsewhere and migrate to OWF if/when we get
accepted?

-Evan

[Brady Brim-DeForest]

I think that a sourceforge for standards would be an excellent idea.
Its first iteration could essentially be an application and services
stack.

--
Brady Brim-DeForest
www.brimdeforest.com

[Eran Hammer-Lahav]

There isn't any plan to offer any infrastructure other than an IPR policy and mentorship. While the board may eventually decide to offer such tools, this isn't the current focus so if you need such facilities I would suggest you create them or use existing options such as Yahoo! Groups, Google Groups, SourceForge, etc.

EHL

[M. David Peterson]

On Sat, Aug 16, 2008 at 3:15 PM, Eran Hammer-Lahav <er...@hueniverse.com> wrote:

There isn't any plan to offer any infrastructure other than an IPR policy and mentorship. While the board may eventually decide to offer such tools, this isn't the current focus so if you need such facilities I would suggest you create them or use existing options such as Yahoo! Groups, Google Groups, SourceForge, etc.

If not mistaken, Google Apps allow you to create and host content for a specific domain.  Coupled with Google Code and Google Groups, all of the pieces are there to accomplish what you need.

--
/M:D

M. David Peterson
Co-Founder & Chief Architect, 3rd&Urban, LLC
Email: m.d...@3rdandUrban.com | m.d...@amp.fm
Mobile: (206) 999-0588
http://3rdandUrban.com | http://amp.fm | http://www.oreillynet.com/pub/au/2354 | http://news.oreilly.com/m-david-peterson/

[Brady Brim-DeForest]

Evan - I think the idea is worth discussing further - perhaps outside of the context of the OWF.

[David Recordon]

I actually completly disagree with that. There has always been the
plan to provide hosting infrastructure for groups to work; just like
Apache does for source code projects. I'll catch up on this thread
when I'm home.

---
Sent from my iPhone classic.

On Aug 16, 2008, at 2:15 PM, Eran Hammer-Lahav <er...@hueniverse.com>
wrote:

>

[David Recordon]

What Evan asked for is something that OWF will be doing for projects that are accepted into the incubator. Just as Apache provides a license for projects not part of Apache, they also provide infrastructure for projects that are a part of it. OWF will do the same sort of thing.

---

Sent from my iPhone classic.

[Brady Brim-DeForest]

Good. It is painfully obvious to me that there is a subtantial need for these type of services in the open source standards community.

It makes more sense to consolidate these back office services with an IPR framework than to force standards organizations to go elsewhere for logistical support.

--
Brady Brim-DeForest
www.brimdeforest.com

[David Recordon]

There's obviously already been a bunch of discussion in this thread in only three hours, which is a great thing!  Obviously no one person is able to speak for the OWF community -- myself included -- but I do think as Brady said there is a definite need for what Even is asking for.

When announcing OWF at OSCON (http://www.slideshare.net/daveman692/supporting-the-open-web-oscon-2008/) slide 18 focused on four things that OWF will provide for communities developing open specifications for the web.  One of these was Incubation to help create new specification and another was Licensing to ensure that all OWF specs really are freely implementable.  Let's first take a look at licensing...

Traditionally communities like OpenID have used a non-assertion covenant once the specification is complete so that contributors to it can promise not to sue anyone for implementing the specification.  The non-assert can be a bit more complex than that, but that is the main goal.  The OpenID non-assert evolved from Sun's, Microsoft's, and work that had been done within formal standards bodies such as the W3C.  OAuth has since taken the OpenID non-assert and evolved it further.

One of my goals (which I think is a shared goal) for OWF is to create a single legal document (whether it be a non-assert or an actual license) which can be applied to specifications developed by anyone that means the specification is freely implementable.  This is analogous to an open source software license.  I can take the Apache 2.0 license and apply it to my code base whether I developed it within the Apache Software Foundation or somewhere else.  There are intricacies to this process such as making sure all of the contributions were made in such a way that the software can be released under this license, but the license itself doesn't deal with process.  Rather, best practices such as contributor license agreements before someone makes any contribution have emerged.  Having a small set of Open Source (note the capital O/S this time) licenses has helped companies and contributors clearly understand what is and is not allowed when using or contributing to code licensed in a certain way.  We don't yet have this clarity for open specifications and hope to help fix that.

The second aspect then is an incubation process.  If you're proposing a project within the Apache Software Foundation they focus on more than just making sure your code is licensed in a certain way.  They focus a lot on helping a project build a healthy and diverse community and provide infrastructure (hosting, bug tracking, SVN, etc) which open source software projects need.  This is then an example of using the Apache 2.0 license within a project accepted into and hosted by the Apache Software Foundation.  They use a process which ensures that contributors sign a specific Contributor License Agreement so that the resulting software can be released under the Apache 2.0 license.

I thus see providing an analogous incubation process as something valuable that OWF can do.  Allowing a specification project like Open MicroBlogging or Portable Contacts API to have everything they need to be successful and develop a final specification within OWF should be a goal beyond just providing the legal documents they'll need.  I've started to work on describing the OWF Incubator (http://groups.google.com/group/open-web-discuss/browse_thread/thread/96a85e1647ecaffe) and really see it as one of the things I'll be spending a lot of my time on as OWF gets up and running.

So Evan, if you need web hosting, a mailing list, a wiki, a SVN repo to hold the raw spec, or other things then I'll be working on helping to provide them.  I have an idea in my mind what infrastructure and processes are needed to create a successful open specification for the web and a healthy community around it, but I'm sure we'll be learning more each time we do this. I hope that others agree with me that it is a useful piece of the puzzle that OWF can provide so that I'm doing more than just scratching my own itch. :)

--David

[M. David Peterson]

On Sat, Aug 16, 2008 at 6:09 PM, David Recordon <reco...@gmail.com> wrote:

I thus see providing an analogous incubation process as something valuable that OWF can do.

+1

 

So Evan, if you need web hosting, a mailing list, a wiki, a SVN repo to hold the raw spec, or other things then I'll be working on helping to provide them.  I have an idea in my mind what infrastructure and processes are needed to create a successful open specification for the web and a healthy community around it, but I'm sure we'll be learning more each time we do this. I hope that others agree with me that it is a useful piece of the puzzle that OWF can provide so that I'm doing more than just scratching my own itch. :)

I most certainly agree.  One thing I would recommend, at least for the time being, would be to push people towards setting up a Google Code project to get things rolling -- The reason being that Google makes it as easy as possible to either mirror the SVN repository at another location, or to move the repository to a new location if/when the need arises.  And all of the wiki updates are stored as a new SVN revision, so all of that content comes along for the ride, making the overall move that much easier.  Of course someone would need to write a parser for the wiki markup, but that wouldn't be all that difficult, and it wouldn't surprise me if Google was willing to open source their parser (if they haven't already?)

[Ben Laurie]

What is your objection to using existing infrastructure for this? What
is the advantage of running our own? I can see some pretty clear
disadvantages, for example:

1. Expense.

2. People's time.

3. Reliability (see 2).

[Eran Hammer-Lahav]

Replicating exiting (free) services is silly, just for the sake of having a custom domain for the mailing list. We seem to be doing pretty well using Google and Yahoo groups, Google Code, SourceForge, Get Satisfaction, various wiki solutions, etc. There are a few areas where these services lack, and that is facilities specific to the pains of managing an open specification, such as a way to manage who signed a CLA, a final license, etc. Manage contribution registration, employer approvals, etc.

When the Apache foundation was created, they didn’t have all these great options to choose from so it made sense for them to create it all from scratch. But this is not where we are today. My vote will always be to reuse existing tools if they can work equally as good as custom tools hosted by the OWF.

In addition, I think the same rules we apply to people – in which to become a member you need to first contribute and prove yourself – should be applied to tools. It would be great if David and others could setup a bunch of infrastructure at their own cost and time and let a few project play with. If these tools prove to be better than others and need help in managing their cost, that is when I think the OWF should get involved.

EHL

[M. David Peterson]

On Sun, Aug 17, 2008 at 9:37 AM, Eran Hammer-Lahav <er...@hueniverse.com> wrote:

When the Apache foundation was created, they didn't have all these great options to choose from so it made sense for them to create it all from scratch. But this is not where we are today. My vote will always be to reuse existing tools if they can work equally as good as custom tools hosted by the OWF.

+1.  Reuse where and when possible.  Focus on the core strength of the group and let those who have mastered the process of managing and hosting content and code handle that side of the process.

 

In addition, I think the same rules we apply to people – in which to become a member you need to first contribute and prove yourself – should be applied to tools. It would be great if David and others could setup a bunch of infrastructure at their own cost and time and let a few project play with. If these tools prove to be better than others and need help in managing their cost, that is when I think the OWF should get involved.

I assume that if a project gets accepted into the incubator by the foundation there will be a page on the OWF site that specifies as such.  In the end, there are three major pieces of the external facing puzzle that need to be in place: A web-based entry point, source and project control, and a list server.  If the OWF acts as the web-based entry point for all projects, and each project simply lists the URI to their choice of source and project control and list server, is there really anything else that is needed?  Between Google Code and Google Groups you get all of the above and then some.  In fact, if you can live with a generic domain for your home page, you can use both Google Code and Google Groups as your web-based entry point which is what we did in the Ruby.NET project (e.g. http://code.google.com/p/rubydotnetcompiler/ which links to http://rubydotnet.googlegroups.com/web/Home.htm ) and it worked perfectly.

I realize there's a natural desire to do everything on your own if and when you can.  But if there's any one thing I have learned during all my years of working in open source communities it's to push as much of the load as you can to third-party service providers and focus on your core competencies.  If you do you gain the luxury of actually being able to focus on the primary purpose you got involved with any given project instead of dealing with all of the micro-management details that you'll soon come to despise. Macro-management is *so much more productive* than micro-management when you can get away with it.

[Gabe Wachob]

+1 to reusing existing services, and having the OWF site be the
"starting point"/"authoritative site" for OWF projects.

--
Gabe Wachob / gwa...@wachob.com \ http://blog.wachob.com

[Evan Prodromou]

On Aug 16, 5:15 pm, Eran Hammer-Lahav <e...@hueniverse.com> wrote:
> There isn't any plan to offer any infrastructure other than an IPR policy and mentorship. While the board may eventually decide to offer such tools, this isn't the current focus so if you need such facilities I would suggest you create them or use existing options such as Yahoo! Groups, Google Groups, SourceForge, etc.
>

OK, thanks for the info. I've registered and set up OpenWebForge as a
GForge instance:

http://openwebforge.org/

Still need a lot of tuning, a logo, a real SSL cert, testing all the
components, better copy, and all that jazz, but at least it's up.

Obviously, I'd be happy to hand over the domain (and .com/.net/.info)
and working instance to the OWF when and if this becomes a priority.
And I think my company would gladly fund the server for whatever time
period is needed. Until then, please feel free to point other spec
projects towards openwebforge.org for infrastructure.

Thanks to all for your feedback,

-Evan

[Eran Hammer-Lahav]

This is awesome! Thanks for doing this. It would be helpful to have a sandbox we can play with and try different ideas.

For the most part, I have happy to do my spec work using existing tools which proved to be successful. However, I do recognize the need for some additional tools to help facilitate the parts of the process we are not very good at (controlling contributions, IPR, etc.). So I think that is where the opportunity is.

EHL

[Brady Brim-DeForest]

The solution could be as simple as publishing a recommended stack of applications/services.

[Evan Prodromou]

On Aug 18, 11:26 am, Eran Hammer-Lahav <e...@hueniverse.com> wrote:

> For the most part, I have happy to do my spec work using existing tools which proved to be successful. However, I do recognize the need for some additional tools to help facilitate the parts of the process we are not very good at (controlling contributions, IPR, etc.). So I think that is where the opportunity is.
>

My experience with developing specs (OASIS for SAML, JSF) has been
that having a single participants' Web site or other infrastructure
used for organization is always going to be a sticking point.

Google is not (yet) a participant in OpenMicroBlogging, but I hope
they become one (through Jaiku). It would just be inappropriate to
host such a project on Google Code or Google Groups.

-Evan

[Ben Laurie]

Why?

>
> -Evan
> >
>

[Evan Prodromou]

On Aug 18, 1:34 pm, "Ben Laurie" <b...@google.com> wrote:
> > It would just be inappropriate to
> > host such a project on Google Code or Google Groups.
>
> Why?

Google Code and Groups are fine free-of-cost tools for the Web
community. But: the very appearance of ownership is inappropriate for
an independent body.
The reason we have independent standards organizations is so that no
one company, group or individual is in control of the specification.

Every page on Google Code and Google Groups has the Google logo at the
top of it. The URL says google.com. There are links to other Google
services. I use my Google account to post. Maybe for standards where
Google has only an indirect interest in the outcome, this would be
acceptable, but for others, it's just not.

Anyways, that's my opinion on the matter. I just wouldn't home OMB on
a Google site, any more than I'd home it at Microsoft or Twitter or
Yahoo. Hosting is cheap and easy to manage, and there's plenty of good
Free Software out there to do the job. Volunteers love running these
kind of things. The value of free hosting just isn't worth the trade-
off.

-Evan

[Eran Hammer-Lahav]

I disagree. Anyone who thinks the appearance of a Google, Yahoo, or other logo on a hosted groups page means any hint of influence is being silly. If people and companies want to offer alternatives to the current tool set, that is great. But I would not like the OWF to mandate any such tools. Free market always work better.

EHL

[Alvaro Lopez Ortega]

Evan Prodromou wrote:
> On Aug 18, 1:34 pm, "Ben Laurie" <b...@google.com> wrote:
>>> It would just be inappropriate to
>>> host such a project on Google Code or Google Groups.
>> Why?
>
> Google Code and Groups are fine free-of-cost tools for the Web
> community. But: the very appearance of ownership is inappropriate for
> an independent body.

Fully agree with this for a really simple reason. That is what I would
think if I saw any of the other independent organization hosted at Google.

Of course it would not mean it were not independent, but that would
certainly be the first impression for many people.

Besides, hosting a wiki and a mailman is not that hard nor expensive,
and the foundation would had a neutral and independent looking.

--
Greetings, alo.
http://www.alobbs.com/