It seems that some of these issues are arguments against Apache process,
so I am not sure what to say about that.
Further, I am not really certain about what is being said here. This
thread seems to imply that if we did "rm -rf framework" in our trunk
directory, then it would be possible for our bundles to be seen as
framework independent and good OSGi citizens. However, since one of our
subprojects happens to be a framework implementation, then all of our
subprojects are tainted and seen as "Felix-only"? Is that right?
If so, then why would anyone want to work on OSGi-related subprojects at
Felix? By this perception, they would be ideologically limiting their
audience. I would hope that this isn't the case or if it was it could be
corrected.
I would think that if a Felix PMC member felt that we were not doing a
good enough job promoting the use of our subprojects on other
frameworks, that he would try to propose ways to remedy that situation,
not encourage.
For example, perhaps it would be worthwhile for us to create some sort
of standard template/table/graphic for each subproject documentation
page to clearly inform users whether it is applicable to other
frameworks and/or whether it has been tested on other frameworks.
Further, I think we could go as far as setting up or documenting
mechanisms that make it possible to automatically test our subprojects
on other frameworks.
Clearly, not everything need be or can be developed at Felix, but the
implication that we might as well accept that Felix bundles will only be
used by Felix users seems awfully counter to all of the concepts for
which OSGi stands.
-> richard
Alin Dreghiciu wrote:
Beside Niclas points there are some additional facts related to OPS4J:
* it's open participation meaning that virtually anybody can
contribute. No need for patches.
* release process. we tend to release often.
Alin
On Jan 13, 2008 1:05 PM, Niclas Hedhman <[EMAIL PROTECTED]> wrote:
On Thursday 10 January 2008 23:09, Richard S. Hall wrote:
There is no attempt by any of our sub-projects to specifically tie
themselves to the Felix framework as far as I am aware. I think we want
all of our work to be interoperable where possible, so I think this is a
non-issue.
Yes, that is a good goal and I salute that. In general, we are more often than
not met with positive comments when trying to resolve cross-platform issues.
But, I was thinking more "mentality-wise". People who use KF, first go and
check the KF's set of bundles, the Equinox-based folks will search the
Eclipse site first... and so on.
Why is that? Because the KF developed bundles are tested on KF only, the
Eclipse stuff is practically only tested on Equinox and so forth. This sends
the signal that it is a "safer bet" to choose from within the same community.
It's all in our heads!
At OPS4J we *try* to ensure that everything gets tested on all the frameworks
we claim to support. And with Pax Runner, we try to force all the frameworks
into a configuration that is as close to each other as possible...
End of the day, I think we want to remain "un-associated" with a framework,
and a strong cross-framework interoperability focus. I think it will benefit
us all.
Cheers
--
Niclas Hedhman, Software Developer
I live here; http://tinyurl.com/2qq9er
I work here; http://tinyurl.com/2ymelc
I relax here; http://tinyurl.com/2cgsug
