Hiram, I believe you, but this has nothing to do with the ability of
building a community and hence a sustainable project.
Apollo 3 years ago (more?) was probably a technically superior option.
It has nothing to do with resistance to change. A business running
ActiveMQ in production has totally different concerns than a vendor who
focuses on features and the coolness factor of the new technology.
Thinking more about it (and I should have done that earlier, my bad),
having a sub-project, which is what hornet is, having the name of the
future version of the main project, effectively undermines the option of
the main project to migrate to a newer version.
That would be ok, if the community decides to build/refactor the next
generation as a subproject, but when a completely new codebase comes
from the outside, it's very bad.
I really want to see people here talking about the community not the
technical merits. That's the issue we need to sort out now.
Hadrian
On 03/24/2015 04:10 PM, Hiram Chirino wrote:
Look, I'm they guy who wrote most of ActiveMQ Apollo. I would not
have taken on such a task if there was not a good reason. I also
would not be welcoming the hornetq project if I did not feel it was
better path forward than Apollo. Please feel free to run the SPEC JMS
benchmark against HornetQ and ActiveMQ if you want some metrics.
Seems you don't believe in that there is a technical market out there
that this project is competing in. Trust me there is, and it's not
all roses for ActiveMQ right now. ActiveMQ is only as popular as it
is because of how long it's been around and because of the
relationships it it's built with other projects like OpenEJB,
MCollective etc.
On Tue, Mar 24, 2015 at 3:47 PM, artnaseef <[email protected]> wrote:
That's all marketing. Are there metrics to back it up?
Last I understood, there is a very strong community using ActiveMQ.
Old isn't much of an argument. For example, it was built before NIO, and
yet it now supports NIO.
Are there more specifics? Feedback on which no action can be taken is
purely criticism, and I don't appreciate ActiveMQ being called a dinasour.
Java, C, C++, Javascript, and others have been around even longer than
ActiveMQ - will you deprecate them all?
If there are critical issues behind ActiveMQ, let's get them out in the
open. This is reminding me of the HawtIO discussion from last year - a lot
of emotion behind it, but ultimately not much strength of reasoning.
Let me turn this around. If we can't clearly enumerate the benefit to the
ActiveMQ community (which is a deal-breaker for me), then why not start
HornetQ as its own Apache project? Where is the downside?
--
View this message in context:
http://activemq.2283324.n4.nabble.com/VOTE-Apache-ActiveMQ-6-0-0-tp4692911p4693757.html
Sent from the ActiveMQ - Dev mailing list archive at Nabble.com.
