incubator-general mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Michael Fitzner <Michael.Fitz...@bmw-carit.de>
Subject AW: Etch status (Was: [Incubator Wiki] Update of "March2012" by michaelfitzner)
Date Mon, 12 Mar 2012 13:05:11 GMT
Hi Jukka,
Thanks for your constructive feedback and the hint to [1]! The establishment of an active
and diverse community is currently one of the challenges in the Etch project. The task is
not so easy because we can do it just beside our main work! As Martijn already mentioned,
we try to publish an article about the Apache Etch Framework for a local (Germany) computer
science magazine. Furthermore I think it would be also good to publish some blog-posts e.g.
at [2] if possible. Another possibility would be to push Etch also in the Android direction
and provide some information how to use it there.

[1] http://markmail.org/message/s2le6gg3wxsa55dz
[2] https://blogs.apache.org

-----Ursprüngliche Nachricht-----
Von: Martijn Dashorst [mailto:martijn.dashorst@gmail.com] 
Gesendet: Montag, 12. März 2012 12:02
An: general@incubator.apache.org; etch-dev@incubator.apache.org
Betreff: Re: Etch status (Was: [Incubator Wiki] Update of "March2012" by michaelfitzner)

One article worth mentioning:

"Practical experiences on a communication middleware for IP-based in-car networks" [1]

Current in-car communication networks are based on automotive specific technologies like CAN,
FlexRay and MOST and use proprietary communication protocols. While CAN and FlexRay come with
a signal-based communication paradigm, MOST provides a more sophisticated interface based
on "function blocks" to the application programmer. In the next years, we expect IP-based
protocols and standard technologies like Ethernet to be deployed for more and more in-car
communication tasks. As a result, we need an IP-based communication middleware that provides
both signaland function-based interaction paradigms and works for all distributed applications
in vehicles. Main challenges are the large variety of embedded devices and operating systems
used in a single car in terms of footprint and compute power. Furthermore, it must be possible
to migrate existing interface definitions from legacy technologies to the new IP-based solution.
In this paper, we propose an IP-based in-car middleware framework based on an open source
solution, Apache Etch. We sketch how different, yet interoperable versions of the middleware
can be used to construct a scalable system that fits to both small and large devices.
Finally, we identify extensions to Etch that are necessary to qualify the solution for the
use in the automotive domain.

COMSWARE '11 Proceedings of the 5th International Conference on Communication System Software
and Middleware

[1] http://dl.acm.org/citation.cfm?id=2016563&dl=ACM&coll=DL&CFID=89538345&CFTOKEN=83544989

On Mon, Mar 12, 2012 at 9:45 AM, Martijn Dashorst <martijn.dashorst@gmail.com> wrote:
> On Sat, Mar 10, 2012 at 12:59 PM, Jukka Zitting <jukka.zitting@gmail.com> wrote:
>> On Mon, Mar 5, 2012 at 8:56 PM, Jukka Zitting <jukka.zitting@gmail.com> wrote:
>>> More generally, looking over the last quarter in Etch I see ongoing 
>>> technical work mostly by Michael Fitzner and Martin Veith but only 
>>> occasional input from others and not much focus on community-building.
>>> Do you have plans on how to increase community activity and 
>>> diversity, or are you happy with the status quo?
>>
>> Any thoughts on this from Etch mentors or the PPMC?
>
> Etch is a slow moving community. AFAICT the code is mostly done, but 
> actively maintained. The space Etch operates in is quite crowded, so 
> community raising efforts are difficult. The engineers working on such 
> code are usually very busy doing other stuff rather than improving the 
> library.
>
> I've tried to contact the most recent voted-in committer, but he has 
> yet to respond, and do actual work on Etch.
>
> The other committers and PPMC members have either jobs that focus on 
> implementing things other than directly working with Etch, meaning 
> that Etch development only happens when they return to a part that 
> uses it. For example the suitable name search is currently undergoing, 
> but finding time to complete it is difficult.
>
> Trying to build a community around a networking stack is difficult, 
> especially this one with its history. The original company open 
> sourcing Etch disbanded the development team when no diversity was 
> present, blowing up the project. I don't blame the company: they have 
> to do what is best for them. Unfortunately it did severe damage to the 
> community.
>
> IMO they should have a great community opportunity in the automotive 
> industry (the active contributors are from a large 3-letter car 
> manufacturer in Germany). My expectation is that this industry (in
> Germany) could expand the community with 3 or 4 committers. Raising 
> awareness by publishing an article about Etch in automotive 
> engineering magazines should help. I don't expect miracles to happen, 
> but if one or two companies start looking into etch, the diversity and 
> activity might just increase a bit. I've asked the guys to try to 
> write one or two articles in German magazines to raise awareness.
> Unfortunately that has taken quite some time, but what I hear is that 
> they are in the final stages.
>
> I was hoping that the new committer would become active, and that 
> graduating Etch would be around the corner. It appears that we need 
> other new blood to invigorate the community.
>
> Martijn



--
Become a Wicket expert, learn from the best: http://wicketinaction.com

---------------------------------------------------------------------
To unsubscribe, e-mail: general-unsubscribe@incubator.apache.org
For additional commands, e-mail: general-help@incubator.apache.org


Mime
View raw message