incubator-general mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Brian McCallister <>
Subject Re: Blaze and Openness of Standards (was Re: [Proposal] Blaze)
Date Mon, 31 Jul 2006 17:12:04 GMT
On Jul 29, 2006, at 4:45 PM, Cliff Schmidt wrote:

> Does anyone have any further concerns about this proposal?

<snip />

> - There was also the question about how the AMQP specification will be
> handled and licensed.  I started this thread with my feelings about
> that aspect (short version: it looks better than some other currently
> incubating projects, but I'd like us to come up with guidelines about
> what is acceptable at Apache, and then make sure this project adheres
> to those guidelines before graduating from the incubator).

I am still uncomfortable with the AMQP spec ownership and process for  
two reasons.

1) The pessimistic and defensive one: Entering incubation at Apache  
implies Apache's endorsement. This is not what we mean, but it is how  
the world will react. This endorsement is partly the point of the  
proposal -- getting the ASF behind AMQP will give it a boost, and  
incubation is still not well understood, even inside the ASF :-(

2) The "go conquer the world" one: If the goal is to create a  
standard protocol for messaging stuff, this requires a lot of buy in  
from a wide range of parties. Keeping the protocol behind closed  
doors and with a mysterious future sabotages this. Transparency is, I  
believe, a major requirement for accomplishing this goal, and the  
process is anything but transparent at the moment.

Both of these points would be lightened if the folks presently  
involved with the specification process seemed to recognize them as  
issues. To my reading, they are not recognized as issues, and there  
has been no public discussion by the folks actually involved with the  
protocol spec about this. The extent of it has been to say, more or  
less, that they doesn't think there is a problem. Lots of uninvolved  
people have chimed in with thoughts, but we (as I am one) are the  
peanut gallery and have no say in the current specification system.

Finally, is the specification forkable if it becomes an  
insurmountable problem?

If I weren't already committed to other incubating projects I would  
offer to help mentor as I really want this to succeed, meaning AMQP  
to become a de facto standard and the Apache implementations to be  
the best. Hope folk don't mind if I do stick in thoughts at least :-)


ps: Tuscany being even more closed is not a justification for this  
getting it wrong. Tuscany's spec relationship is, in my opinion, a  
mistake, and one we have, hopefully, learned from.

To unsubscribe, e-mail:
For additional commands, e-mail:

View raw message