activemq-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Weiqi Gao <>
Subject Re: [DISCUSS} HornetQ & ActiveMQ's next generation
Date Fri, 27 Mar 2015 20:31:06 GMT
 From one "user"'s perspective:

I joined the list as soon as I learned of the code donation.  I saw the 
ActiveMQ6 guys hard at work and making a lot of progress.  I saw the 
ActiveMQ5 guys hard at work making fixes.  I saw the "consensus" of 
about merging the two code bases and communities in the thread from last 
year.  And I saw the ActiveMQ 6 name being proposed not by the former 
HornetQ guys, but by a member of the then existing ActiveMQ community.

All good and harmony.  Until the release vote, when six months of hard 
work was done on the ActiveMQ6 repo.

I agree with Daniel Kulp and David Jencks on the general direction.

Weiqi Gao

On 3/27/2015 2:11 PM, Daniel Kulp wrote:
>> On Mar 27, 2015, at 2:46 PM, Hadrian Zbarcea <> wrote:
>> After this thread, it's painfully clear that there won't be any convergence and it's
best for both communities to evolve independently.
> Huh?   I don’t agree with that at all.   It appears to me like it’s just you that
feels there won’t be any convergence.  The Hornet folks are doing a ton of work to get the
“code” in a good state to support it, most of the “AMQ5” devs are supportive of the
efforts although a bit short on time to help.  From this thread, it looks like even the “users”
are supportive of getting the Hornet codebase into a state to be AMQ6.   Lets get a “milestone”
release out (give it a code name if you really object to 6.0-m1) to help foster some excitement
around it, start getting contributions and committers and eventually PMC members, and hopefully
we can even back port some of the ideas and such to 5.x.   This is exactly the kind of thing
this community needs to help foster diversity and growth and all that.   Sitting around doing
the "status quo” has obviously done very little to change anything.
> Dan
>> Sure, it is expected that some of the developers would move to the new kid on the
block project, that's fine, actually great for the new project. The new project could reuse
whatever they want from ActiveMQ, grow a community. If at a later time there is a desire for
convergence it can still happen.
>> Continuing like this, I fear, will be a big distraction for both projects, not good
for any of the two communities. I hope some sort of resolution will happen soon.
>> Cheers,
>> Hadrian
>> On 03/27/2015 01:28 PM, Hiram Chirino wrote:
>>> I've been trying to keep quite to get an idea of different folks view
>>> points.  At this point I think it's fair to say that the ActiveMQ
>>> project has not reached consensus that the HornetQ code contribution
>>> is ready to become the successor to ActiveMQ 5.
>>> So calling the git repo for the code donation activemq-6, was probably
>>> a bad idea.  A this point I think the code donation should follow the
>>> path the apollo took and switch to a code name.  It should continue to
>>> do milestone release and solicit the help of ActiveMQ 5.x
>>> users/developers to help mature into the successor that it wants to
>>> become.
>>> We can then revisit renaming to an ActiveMQ N, once it has matured to
>>> the point there is little objection to it becoming the successor.
>>> On Fri, Mar 27, 2015 at 11:22 AM, Rich Bowen <> wrote:
>>>> [I see that some of what I put in this email has already been said by
>>>> others, but I'm going to go ahead and send it, because it needs to be
>>>> heard.)
>>>> On 03/26/2015 12:02 PM, Hiram Chirino wrote:
>>>>> Hi Chris,
>>>>> If you take a peek at the source code for the code grant I think
>>>>> you'll notice that all the original HornetQ references have been
>>>>> removed/replaced by ActiveMQ.  So I think we are ok from a TM
>>>>> perspective.
>>>> A much larger concern (at least to me) is not merely the naming, but the
>>>> perception that a completely new codebase has been brought to the project,
>>>> replaced the existing work wholesale, and been called the next version. This
>>>> is how it's been described to me by several different members of the project
>>>> community, and their perception is that this has been done without the
>>>> consent of the community. This is, of course, a fairly serious accusation.
>>>> Related to this is the assertion that the PMC has been somewhat biased on
>>>> who has been invited to join their numbers, based on corporate affiliation
>>>> an even more serious accusation.
>>>> The analogy that was offered to me was that of the IIS code being imported
>>>> into the Apache httpd code tree, and released as httpd 3.0, by virtue of
>>>> majority Microsoft presence on the PMC.
>>>> I recognize that this is a very harsh accusation. The folks that have
>>>> brought this concern to me have done so privately because they feel that
>>>> their voice is ignored on the PMC list.
>>>> In terms of how this situation might be resolved, two things have been
>>>> suggested.
>>>> 1) Invite some of your 30+ non-PMC committers onto the PMC.
>>>> 2) Go ahead and release something based on HornetQ, just don't call it the
>>>> next version of ActiveMQ over the objections of the minority. (I see that
>>>> this solution has been addressed by others, recommending that the code be
>>>> taken to the incubator.)
>>>> --
>>>> Rich Bowen - - @rbowen
>>>> - @apachecon

View raw message