felix-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Richard S. Hall" <he...@ungoverned.org>
Subject Re: OSGi Initial Provisioning implementation wanted?
Date Tue, 27 Sep 2011 15:41:44 GMT
On 9/27/11 11:35 AM, Jeremias Maerki wrote:
> That was fast. ;-)
>
> I wrote it out of a desire to simplify app deployment in a distributed
> environment by using standard mechanisms defined by the OSGi spec. It
> looks clean and simple. I'm currently at the experimentation stage. At
> some point, the security features will certainly become important to me
> but I cannot say when that will be. I can currently only invest time as
> far as my requirements go, so I cannot make any promises. It could
> effectively result in a code dump, but then, someone else might pick it
> up. If that's not good enough, no hard feelings and I'll just publish
> the code via my website.

Well, let's see if others are interested in picking it up...

-> richard

>
> You can take a look at the source code here (without the test suite):
> http://jeremias-maerki.ch/download/osgi/ch.jm.osgi.provisioning-0.9.0.dev-sources.jar
>
> On 27.09.2011 17:20:14 Richard S. Hall wrote:
>> I think it would be interesting, assuming that there would be ongoing
>> development of it. Do you plan to continue to work on it to fill the gaps?
>>
>> ->  richard
>>
>> On 9/27/11 11:15 AM, Jeremias Maerki wrote:
>>> Is there interest in an implementation of the OSGi Initial Provisioning
>>> 1.2 service? AFAICT, there is none in Apache Felix. Over the past few
>>> days, I've written a clean-room implementation that I could donate to
>>> Felix if there's interest. The implementation runs on J2SE-1.4, the test
>>> suite on J2SE-1.6 (due to Pax Exam).
>>>
>>> There are a few gaps still but it already seems to do the job in my case:
>>> - rsh and https protocols are not supported, yet. Only file and http.
>>> - InitialProvisioning-Entries manifest header is not yet interpreted.
>>> Only ZIP extra fields and file extensions are used to determine the
>>> entry type.
>>> - There's no guard against endless recursions that might happen with
>>> poorly designed ZIP chains.
>>> - I'm currently using my own Ant-based build system. A Maven POM would
>>> have to be written by someone who knows what he's doing. ;-)
>>>
>>> I'll gladly post the source code for review if there's interest. Let me
>>> know.
>>>
>>> Cheers,
>>> Jeremias Maerki
>>>
>
>
>
> Jeremias Maerki
>

Mime
View raw message