Skip to main content

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [List Home]
Re: [incubation] wrap binaries into a plugin

Hi Jens,
 
Thanks a lot for this clarifcation. EMO eproves it I would suggest to put it into the wiki into the new project handbook. It helped me a lot to understand the sitation.
 
So for 4diac FORTE we are currently mainly doing source only releases. We only have very basic binaries for Windows and Linux (and hopefully soon for MacOS) allwoing users to do some quick tests and evaluations. But these binaries have only OS dependencies.
 
Furthermore all all our "special" dependenices are opt in dependencies. That means the user needs to add them in the compilation step and have the dependencies available during compialtion (i.e. our option 2).
 
We have for some systems vendor specific (closed source) dependencies (i.e., I would like to compare it to a Windows dependency). In this case our user typically needs anyhow to buy the hardware and/or the according tools and libraries. So we can not deliver binaries for these.
 
So making the long story short. If I understood your explantion we sofar do not need to any CQs as long as we stick to source releases.
 
BR,
Alois
 
Gesendet: Dienstag, 28. Juni 2016 um 13:33 Uhr
Von: "Jens Reimann" <jens.reimann@xxxxxxxxxxxxxxx>
An: incubation@xxxxxxxxxxx
Betreff: Re: [incubation] wrap binaries into a plugin
Hi,

I do think that there are a few scenarios here.

1) You port your runtime to a new operating system. If that does not
bring any new dependencies (like libraries) then it should be not an
issue at all. I know that some operating systems just behave a bit
differently every new and then. For example porting from Linux i386 to
Linux x86_64 may require to change a few things here and there. But in
the end your code base supports both with one code set. I don't see a
reason to have any CQ here.

2) You port your runtime to a new target which requires additional
dependencies. If you want to distribute that final runtime (binary) you
need a CQ and the code has to go through the full IP check.

3) Like 2) but you don't want to distribute any binaries. Then you could
go with a "exempt pre-reqs", as Mike said. People just have to get those
dependencies first before they can compile your runtime themselves.

4) You want to add functionality to your runtime which requires 3rd
party dependencies which are not compatible with the EPL or are closed
source. Now I don't think that is possible in this case, since it would
require using "#include"s and function calls to proprietary/closed
source/incompatible code. So it is not only a matter of "when its there
is works", but it is a part of your runtime (4DIAC) then.

But I also do think that there is a simple solution to #4. If your
runtime supports dynamic loading of extensions (which also works in
C/C++) then you could define some sort of plugin API. The plugin system
can be distributed easily through the Eclipse Foundation. And all other
extensions, which require problematic dependencies, can by loaded when
the runtime initializes. Which is in fact what the Eclipse IDE does all
the time.

I think your scenario matches #2 or #3. Since, if I understood you
right, you compile your runtime, together with some O/S blob into a
firmware blob which is a combination of your code and their O/S. So you
can decide if you only distribute the source code (#3) or a full blob (#2).

Jens

PS: @EMO please correct me if I was wrong

On 06/24/2016 08:56 PM, Alois Z. wrote:
> Mike,
>
>> Operating systems are typically treated as "exempt pre-reqs" because they are already installed on the target machine. I/O access libraries have to be treated on a case-by-case basis, and it makes a great deal of difference if you are distributing the library with your code, or simply calling a library that is already installed on the target machine.
>
> Thanks for the clarification. How would we treat operating systems as often sued on smaller IoT devices where the operating system is a library linked to the application (e.g., eCos, freeRTOS). Would these be treated like any library?
>
> I'll read through the docs. Thanks for the pointers.
>
> Alois
>
>> See http://www.eclipse.org/org/documents/ for the following documents:http://www.eclipse.org/org/documents/Eclipse_Policy_and_Procedure_for_3rd_Party_Dependencies_Final.pdf[http://www.eclipse.org/org/documents/Eclipse_Policy_and_Procedure_for_3rd_Party_Dependencies_Final.pdf]
>> http://www.eclipse.org/org/documents/GPL_CE_Policy.php
>> http://www.eclipse.org/org/documents/LGPL_API_Policy.pdfHope that helps.
>
> --
> Mike Milinkovich
> mike.milinkovich@xxxxxxxxxxx[mike.milinkovich@xxxxxxxxxxx]
> +1.613.220.3223
> _______________________________________________ incubation mailing list incubation@xxxxxxxxxxx To change your delivery options, retrieve your password, or unsubscribe from this list, visit https://dev.eclipse.org/mailman/listinfo/incubation[https://dev.eclipse.org/mailman/listinfo/incubation]
> _______________________________________________
> incubation mailing list
> incubation@xxxxxxxxxxx
> To change your delivery options, retrieve your password, or unsubscribe from this list, visit
> https://dev.eclipse.org/mailman/listinfo/incubation


--
IBH SYSTEMS GmbH
D-85235 Pfaffenhofen an der Glonn
Läutenring 43
Geschäftsführer / CEO: Dr. Thomas Heitzig

Amtsgericht München
Handelsregister Nummer HRB 197959
USt ID: DE267945175

Office Munich
D 80992 München
Agnes-Pockels-Bogen 1
T +49 89 18 9 17 49 0

The information transmitted is intended only for the person or entity
to which it is addressed and may contain confidential and/or pivileged
material. Any review, retransmission, dissemination or other use of,
or taking of any action in reliance upon, this information by persons
or entities other than the intended recipient is prohibited. If you
received this in error, please contact the sender and delete the
material from any computer.


_______________________________________________
incubation mailing list
incubation@xxxxxxxxxxx
To change your delivery options, retrieve your password, or unsubscribe from this list, visit
https://dev.eclipse.org/mailman/listinfo/incubation

Back to the top