Skip to main content
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [List Home]
Re: [udig-devel] initial contribution - git branch
Cool, I will confirm that Frank is interested.

I will be in position to move on this from Friday onward - does that qualify as "quickly"?

We can keep discussion no this email list, and we will have to take your guidance on how to communicate with the IP team. Should we advise them on this idea of issuing a second "cut down" archive with their initial issues resolved? 

Frank if we are doing this then I want to focus on getting IP issues out of the way, not necessarily a functioning system. I think we could upgrade to GeoTools 9.0 release (I have a pull request doing the required API migration).

(Aside - Wayne I think you can change your signature now that EclipseCon is over)

-- 
Jody Garnett

On Tuesday, 9 April 2013 at 2:02 AM, Wayne Beaton wrote:

If this can be quickly addressed, then I think you should move on it.

If the IP team doesn't have to filter through the stuff you don't need, they'll be able to move faster.

Wayne
 
On 04/07/2013 01:51 AM, Jody Garnett wrote:
That is a *great* idea Wayne:

Wayne do you need to check with the IP team to see if they are fine with this idea? My understanding was we had to submit *something* in order to start this conversation.

If the IP team is going to stop dead, or be forced to restart when provided with a new link, then I would prefer to wait and do this once in a eclipse repository.

Assuming the idea can go ahead, and Frank is keen, we can start the branch and "quickly" address the initial range of concerns that was raised.

1. Andrea's GPL file - we can remove the file and that plugin will just be broken for a bit
2. The Service tutorial - transition to whatever BIRT uses for parsing CSV files
3. JAI+ImageIO native fragments
4. Remove unused folders (Jesse's original release scripts etc…)

OK, created a pull request for the origin repository to fix this issue: https://github.com/uDig/udig-platform/pull/167
Updated sphinx docs because otherwise the build would fail if the jar is not available.

In General : What is the working copy we can collaborate, is the initial contribution already a repository (possible a temporary) at locationtech? If so, how can we access it 
The initial contribution is just a CQ in our system. Jody included a pointer to the zipped repository in the CQ since it wouldn't let him attach such a large file. There is no repository from an IP team POV, just attachments on that CQ. We can obsolete and replace those attachments during the review process. Since it's very early in the review, now is an excellent time to provide a more concise set of bits for the IP team to review.

Perhaps it makes sense for you to create a branch in the existing repository to work out what really needs to be in the initial contribution. When you've made your final decision, you can zip it up and send it along to the IP team.

I do like the idea of using a Git repository for the IP team, but that's not how they work today.
> ./plugins/net.refractions.udig.jai.macosx/src/net/refractions/udig/jai_core.jar
> ./plugins/net.refractions.udig.jai.macosx/src/net/refractions/udig/jai_codec.jar
> ./plugins/net.refractions.udig.jai.macosx/src/net/refractions/udig/clibwrapper_jiio.jar
> ./plugins/net.refractions.udig.jai.macosx/src/net/refractions/udig/jai_imageio.jar
> ./plugins/net.refractions.udig.jai.macosx/src/net/refractions/udig/mlibwrapper_jai.jar

Third-party CQs needed (and it will be a pain).
OK, maybe we can start with work I've already done to make JAI/Image IO available as a bundle with native fragments. Normally JAI/JAI Image IO is provided in the libs/ext folder of the Jave Runtime installation. Because its definitely required to run uDIG we set up a guide how to build a JRE and provide it e.g. with the installer.
This is going to take more brain power than I have available right now. I'll get back to you.
Frank/Wayne - if the native code is too stressful we *can* run just with jars for a bit - I don't even think moovida would care since he is running 64 bit these days (Where no native code is available).

Wayne we have always expected that JAI/ImageIO will be the trouble part of this process, we did do some sanity checks prior to applying to LocationTech.

--
Jody Garnett


_______________________________________________
User-friendly Desktop Internet GIS (uDig)
http://udig.refractions.net
http://lists.refractions.net/mailman/listinfo/udig-devel

--
Wayne Beaton
Director of Open Source Projects, The Eclipse Foundation
Learn about Eclipse Projects
EclipseCon 2013
_______________________________________________
User-friendly Desktop Internet GIS (uDig)

Back to the top