Since the code is sensitive to changes in Graph, Model, and OpVisitor it
probably makes sense to release on cycle with Jena core.  I'll get the
source code repackaged, I think most files have the header already.

Is there a tool that will look for missing Apache headers?

Claude



On Thu, Aug 1, 2013 at 4:09 PM, Andy Seaborne <[email protected]> wrote:

> Claude,
>
>
> > How do I go about contributing the code?
>
> You already have an iCLA on file so you can contribute code at any time
> (provided it is yours to contribute! e.g. if your employer has any related
> policy).
>
> Are there any new dependencies?
>
> If no, this is mostly about you want to get it out to users.
>
> Is this an additional module a user might wish to use? (not something to
> hook deep into the standard distribution).
>
> I'm guessing it is, so the user adds "org.apache.jena:jena-**security"
> (or whatever it gets called) to their POM to add it.
>
> The next decision is how you want to release it - coupled to the next
> release or put out for discuss and feedback as a separate unit. (It can
> migrate into the main build later.)
>
> Is this likely (in the short term) to evolve faster than the main Jena
> release cycle?  Evolve means formal releases, not snapshots.  Given we seem
> to be on ~3 months rather than the loosely stated ~6, faster is tricky!
>
> If aligning to the main release is the right approach, then a module under
> trunk/ seems right.  It will be built and deployed with a release; it will
> not be in the apache-jena distribution or the related apache-jena-libs
> (jena-text and jena-fuseki aren't either).
>
> If you want, for the moment, a separate release cycle then put in under
> .../repos/asf/jena/**Experimental/ and announce it.
>
> Or, even if related to the main release cycle, you want a "first release"
> that is more about getting feedback, then released separately might be
> better.
>
> The quickest route is to put under Experimental/ and email users@. Build
> to the snapshot repo and people can try it out immediately. Actually,
> snapshot builds are an optional extra.  People can build themselves if
> necessary.  As first steps, early feedback is better than polishing the
> build process.
>
> However, there is one thing:
> You must add the Apache header to all files.
>
> fastest route:
> 1/ Align source code,
>      repackage to org.apache.jena.security
>      Add Apache header
> 2/ Import to /Experimental/????
> 3/ Email users@
> 4/ Then consider formal release and aligning to jena main releases.
>
>         Andy
>
>
> On 01/08/13 07:37, Claude Warren wrote:
>
>> I would like to contribute the Xenei Jena Security project to the Apache
>> Jena project (I assume as an experimental project).  The code is currently
>> at 
>> https://github.com/Claudenw/**JenaSecurity<https://github.com/Claudenw/JenaSecurity>
>> .
>>
>> The project applies security access security (Create, Read, Update and
>> Delete restrictions) on Graphs and Models.
>>
>> It does this by creating dynamic proxies to the Graph or Model
>> implementations and intercepting calls that require modification.  The
>> architecture uses a interface that must be implemented by the developer
>> using the package.  This interface provides the security layer with the
>> Principal of the current user, and determines if that user has specific
>> access to specific graphs and/or triples.
>>
>> There are several thousand test cases.
>> The Graph and Model  implementations pass the Jena Graph and Model test
>> suites.
>>
>> There is good javadoc coverage.
>>
>> There is some documentation.
>>
>> How do I go about contributing the code?
>>
>> Claude
>>
>>
>


-- 
I like: Like Like - The likeliest place on the web<http://like-like.xenei.com>
Identity: https://www.identify.nu/[email protected]
LinkedIn: http://www.linkedin.com/in/claudewarren

Reply via email to