No, Heraldry failed because the two companies responsible for the
majority of OpenID implementations at the time didn't want to work
within the ASF's process. This is one of the reasons why community
based open source development is important beyond just corporate
backed development.
I think Chris' proposal is sound, he has buy in from various library
contributors, we have a way to let people like Mart continue
developing on GitHub, and I'm not seeing a concrete alternative
proposal with someone willing to lead it and make it happen like Chris
is. So I'm sorry, but can we please move forward?
If we believe that the best path forward is for Chris to first make http://openid.net/code
then lets do that, but I agree with him that an OpenID Google Code
project is a demonstrable piece of forward momentum. The wider
developer community has expressed many times over that OpenID's
libraries are not of the quality that they need to be and it is the
Foundation's job to help fix that.
--David
On Jun 1, 2009, at 8:38 AM, Johannes Ernst wrote:
On Sun, May 31, 2009 at 10:00 PM, Johannes Ernst
<[email protected]> wrote:
We had this discussion before and it lead to the Apache incubator
named Heraldry. Admittedly that one failed, but I don't think it
was because of the name ;-)
If it wasn't the name, can you describe why it failed. I've heard
of Heraldry, but am not familiar with its structure or fate.
The idea was to incubate within the Apache Software Foundation an
open-source project developing OpenID-related functionality.
Libraries were donated into it, and an entire OpenID provider was
donated into it. There was broad support from all parts of the
OpenID community. We figured being associated with the ASF would not
be a bad idea, and the Apache license sounded good, too.
The incubation process failed because basically nobody "did
anything" in terms of writing code.
I am curious how you think that the foundation should best go about
creating or facilitating the creation of the circumstances that
would lead to world-class open source OpenID libraries being
developed.
I haven't heard alternative proposals, but I have received some
negative feedback towards my proposals, and yet the libraries are
still not writing themselves.
Well, from what I can see the openid4java project has some traction.
It is my understanding that code from that project has been
incorporated into some large-scale commercial offerings. It's a
small community but it is active and has been for a while. So they
are doing something right. Perhaps one could attempt to broaden that
project beyond Java?
I think a similar question needs to be asked about commercial/
proprietary implementations. There aren't a whole lot of those
either. I would stipulate that it is for the same reason.
Now stop me because I'm about the speculate why that is. ;-) But
that wasn't your question.
Cheers,
Johannes.
_______________________________________________
board mailing list
[email protected]
http://openid.net/mailman/listinfo/board
_______________________________________________
board mailing list
[email protected]
http://openid.net/mailman/listinfo/board
