Re: [Xenomai] Xenomai community meeting 2018, meeting minutes

2018-02-09 Thread Karim Yaghmour


Hi Wolfgang,

On 02/09/2018 11:45 AM, Wolfgang Denk wrote:

Dear Karim,

In message <71b1c406-6eb5-8c62-36e6-c48341baf...@opersys.com> you wrote:



- usage stats could be extracted from our website, downloads,
  mailinglists


We have to be very careful here.  German law is pretty strict about
what you can (or rather must not) do with such data...


I'm obviously not in a position to provide any sort of legal advice nor
would I recommend violating any laws. However, it would probably be
worth getting some specifics on this instead of just shying away because
of an unspecified potential issue. The point is, it's my understanding
that there would be no selling nor advertising of this nor use for any
commercial purpose. The only purpose here is to get a better
understanding of who is using Xenomai and where they are physically
located to plan activities accordingly and provide a more accurate
picture to the larger open source community as to the composition of the
Xenomai user-base.


This is not intended to "shy away" anybody.  But the laws are really
strict. Yes,we can easily generate information like numners of
downloads or such, without problems.  But just storing accessing IP
addresses (even whenyou never ever disclose these to anybody) means
you are storing pesonal information, and you might require
permission from the accessing user.  So for exmaple trying to
evaluate accessing domain names to generate statistics about the
number of companies who are interested in Xenomai would be something
which I would not do without prior legal clearance.


Understood. FWIW, I don't have access to any of this myself. Just trying 
to help.



I doubt this will actually work.  In my experience, there is a
pretty large base of users who intentionally do NOT advertise their
use of Xenomai.


That's a good point. Still, Xenomai has historically suffered from
having a primarily "anonymous" user base. Yet, that user base is
hampered by the lack of seeming traction given that few are willing to
"endorse" the project.


This is only natural, I think. We are not one of the classic open
source communities like Linux of U-Boot have.  You don't run Xenomai
on your toy project to scratch some private itches.  Xenomai is
intended for and primarily applied in industrial environments.
The overwhelming majority of Xenomai users known to me don't do this
for fun, but for business. And many of them are not permitted to
disclose any details of their work, often not even the name of the
company.  This has not much to do with being willing or not, but
with being permitted.


That makes sense. I wouldn't say most people using U-Boot are doing it 
for toy projects, but I get what you mean. Being a Linux or U-Boot user 
may be more anonymizing that declaring that you use Xenomai.



That's the core issue. If no one is going to show public and strong
support for this project, it's going to be close to impossible to show
the value of the project to the wider open source community, especially
the Linux kernel developers. The project's value can only be argued for
so long simply based on logical arguments when compared to other approaches.


This problem is as old as Xenomai.  We tried before to overcome
this, for example with the first (and so far only, sic!) Xenomai
User's Meeting of 2009 where we were actually able to find a few
companies willing to publicly confirm their use of Xenomai ([1]).
We did a number of projects based on Xenomai since, and please
don't assume that we did not try to get more such publicity.

I really don't want to discourage you, but please don't hold your
breath!


True, this is an old problem. But it's going to have to be solved at 
some point. Because it's those very users who depend on Xenomai who are 
ultimately going to pay the price if it ceases to be maintainable.


The idea was that by having regular meetings at a proper location, 
possibly co-located with another event such as ELC or ELCE, then people 
could show up "uncommitted" to an adjunct event. Yet, still, this would 
foster at least some form of community around the project.


Cheers,

--
Karim Yaghmour
CEO - Opersys inc. / www.opersys.com
http://twitter.com/karimyaghmour

___
Xenomai mailing list
Xenomai@xenomai.org
https://xenomai.org/mailman/listinfo/xenomai


Re: [Xenomai] Xenomai community meeting 2018, meeting minutes

2018-02-09 Thread Karim Yaghmour


Hi Wolfgang,

On 02/09/2018 06:00 AM, Wolfgang Denk wrote:

In message <20180208145121.7551c...@mmd1pvb1c.ad001.siemens.net> you wrote:

   - usage stats could be extracted from our website, downloads,
 mailinglists


We have to be very careful here.  German law is pretty strict about
what you can (or rather must not) do with such data...


I'm obviously not in a position to provide any sort of legal advice nor 
would I recommend violating any laws. However, it would probably be 
worth getting some specifics on this instead of just shying away because 
of an unspecified potential issue. The point is, it's my understanding 
that there would be no selling nor advertising of this nor use for any 
commercial purpose. The only purpose here is to get a better 
understanding of who is using Xenomai and where they are physically 
located to plan activities accordingly and provide a more accurate 
picture to the larger open source community as to the composition of the 
Xenomai user-base.



- for now we do not look at that and try to not be nosy


I can't parse that?


- usage stats could be very useful for users as well
 - get an idea how popular Xenomai is
 - where in the world
 - which versions especially


I doubt this will actually work.  In my experience, there is a
pretty large base of users who intentionally do NOT advertise their
use of Xenomai.


That's a good point. Still, Xenomai has historically suffered from 
having a primarily "anonymous" user base. Yet, that user base is 
hampered by the lack of seeming traction given that few are willing to 
"endorse" the project.


That's the core issue. If no one is going to show public and strong 
support for this project, it's going to be close to impossible to show 
the value of the project to the wider open source community, especially 
the Linux kernel developers. The project's value can only be argued for 
so long simply based on logical arguments when compared to other approaches.


Cheers,

--
Karim Yaghmour
CEO - Opersys inc. / www.opersys.com
http://twitter.com/karimyaghmour

___
Xenomai mailing list
Xenomai@xenomai.org
https://xenomai.org/mailman/listinfo/xenomai