the small footprint MQTT server in the
model of RSMB/Mosquitto is what I had in mind. Very much designed
to run at the "edge of network" on edge server boxes i.e. the
type of box that typically bridges from a sensor area network to the internet
/ wide area network. The edge server then communicates upstream to
a general purpose servers like Apollo. The small footprint Java
MQTT server would also enable simple integration into an Eclipse Java developers
Until it exists providing an OSGi wrapper
for RSMB/Mosquitto seems like a good approach. The main externals
to RSMB/Mosquitto are via an ini file which should simplify the task or
"wrapping" it for OSGi.
All the best
Paul Fremantle <paul.fremantle@xxxxxxxxxx>
discussions for paho project <paho-dev@xxxxxxxxxxx>
ActiveMQ and its successor Apollo are both general purpose
message brokers. I think they are excellent choices in many circumstances.
I would also like to see MQTT support added to other general purpose brokers,
for example Apache QPid.
However, I would suggest that RSMB and Mosquitto implement
a different pattern that aligns very well with MQTT which is the idea of
a very small footprint broker that specializes in MQTT. I think if we build
any new work in Java we should aim for this model. Otherwise it will just
be YAMB (Yet Another Message Broker).
On 7 October 2013 10:05, Julien Vermillard <jvermillar@xxxxxxxxxxxxxxxxxx>
Active MQ embedded could be a solution no? ActiveMQ is
and play wells in OSGi environements.
On Mon, Oct 07, 2013 at 01:35:26AM -0700, Dave Locke wrote:
> Hi Scott,
> as one of the project leads I would love to see a small footprint
Java MQTT server hosted in Paho or a project under the M2M umbrella.
With Mosquito and RSMB being contributed, a small footprint MQTT server
that runs well in an OSGi container is one of the main missing pieces of
a solid Eclipse MQTT based ecosystem.
> All the best
> From: Scott Lewis <slewis@xxxxxxxxxxxxx>
> To: General development discussions for
paho project <paho-dev@xxxxxxxxxxx>
> Date: 05/10/2013 05:57
> Subject: Re: [paho-dev] Paho plans
> Sent by: paho-dev-bounces@xxxxxxxxxxx
> Hi Mike,
> I happen to disagree with the wisdom of the Board's definition, but
that doesn't have to do with anything...as I'm not questioning based upon
impl language...I don't really care what language it's written in.
> What I do care about is being able to use a framework provided by
another EF project...seems to me that everyone should want other EF projects
as consumers...but I guess community means different things to different
> As I've already made clear, I would be one such contributor for a
java based Mqtt broker...but sadly, without the commitment of the project
leads that seems unlikely to happen for a new project in incubation
> I never said anything about a requirement. I simply thought that a
new project might want some consumers/community members/supporters.
> In the mean time ECF has little choice other than to use the work
of other, non-EF projects in this area. Not what we would prefer to do.
> Mike Milinkovich <mike.milinkovich@xxxxxxxxxxx>
> The "powers that be" would be referring to the Eclipse Board.
> Your viewpoint on what defines an Eclipse project is out of date.
The Eclipse Board decided quite some time ago that we are willing to accept
projects based on languages and platforms other than Java and OSGi. (Sorry,
I'm on my phone or I'd provide a link. I'll do that in the morning. )
> If sufficient contributors show up willing to create a Java MQTT broker
as part of Mosquitto, that would be great. But there is no requirement
for the existing project to port to Java.
> Mike Milinkovich
> From: Scott Lewis
> Sent: Friday, October 4, 2013 6:50 PM
> To: paho-dev@xxxxxxxxxxx
> Reply To: General development discussions for paho project
> Subject: Re: [paho-dev] Paho plans
> Hi Roger,
> On 10/4/2013 3:20 PM, Roger Light wrote:
> > Hi Scott,
> >> I suppose that this is more a question for Mosquitto than
> >> since Mosquitto is a project proposal I suppose I should
ask it here: If
> >> Mosquitto doesn't integrate with OSGi runtimes...or Eclipse...in
> >> way...would it make sense to host the project at Eclipse?
> > A good question to raise. The Eclipse Foundation "About"
page makes no
> > mention of Java or OSGi :) It does talk in terms of building
> > frameworks for building and maintaining software, but my impression
> > that with the M2M side of thing the powers that be have a desire
> > broader approach on the technology front.
> I'm not sure who you are referring to with the 'powers that be'. Could
> you explain?
> I have no objection to using C as an implementation
> language...particularly as I imagine it could provide size and/or
> performance benefits. But it would be nice if it also ran on the EF
> runtime platform (OSGi framework) so that other EF projects could
> actually use it as a framework. Currently that's not possible...i.e.
> the vast majority of EF projects can't even use the framework. This
> severely limits the usefulness to other EF projects...which hopefully
> are at least a desired consumer of this work.
> I'm not asking that the C implementation be at all
> abandoned...rather...why not do what SWT did...and provide a java
> that uses the C implementation? Like I said, I would be willing to
> contribute to such an effort...so that other EF projects could actually
> consume/use this framework (ECF in my case, but I suspect that other
> projects could be interested also).
> Note I'm not asking this for the IDE/tooling per se...rather ECF is
> runtime project, and is primarily focused on supporting OSGi runtimes
> and standards (e.g. OSGi remote services). So this request isn't about
> the IDE...or even tooling more generally. A lot of work in the RT
> project at EF is focused on OSGi server development, and in the abstract
> M2M would be an ideal combination IMHO...lightweight brokers for
> modular, lightweight OSGi servers.
> All I'm asking is that you look to other EF projects as a prospective
> community, and make it possible for them/us to consume this work.
> that the existing codebase be abandoned or rewritten.
> I'm of the opinion that supporting other EF projects should be some
> of metric for all EF projects...so that we don't end up a loose
> collection of unrelated projects...with no/little relationship other
> than that we are hosted at eclipse.org...but that's just my opinion.
> > It's without a doubt true
> > that a big part of how EF works is down to the IDE, and that
> > bit odd when the C and Python that I'm interested do have nothing
> > do with the IDE, but I think that's just one of those things.
> > genuinely was the opinion that it wouldn't make sense to host
> > project at Eclipse, then fair enough, no hard feelings, but I
> > think that is the case. It just fits less under the IDE umbrella
> > the M2M umbrella, which is by its nature a bit of a different
> > Ponte and Paho (Java client notwithstanding; there are C, Lua,
> > related.
> > Cheers,
> > Roger
> > _______________________________________________
> > paho-dev mailing list
> > paho-dev@xxxxxxxxxxx
> > https://dev.eclipse.org/mailman/listinfo/paho-dev
> paho-dev mailing list
> paho-dev mailing list
> Unless stated otherwise above:
> IBM United Kingdom Limited - Registered in England and Wales with
> Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire