You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
There is a need for a community edition of WorldWind. This will essentially be a community-managed fork that incorporates pull requests not belonging in the stable upstream version, but still being useful enough to the community that it should be maintained and kept up-to-date. See the discussion at pull-request #77 for some background on this.
I think two pull requests that we should already consider a priority is the JOGL upgrade (see pull-request #1) and the GDAL upgrade (see pull-request #146). I mean the JOGL pull request has been waiting for a merge now for over 3 years.
These pull-requests: #18, #19, #30, #138, #139 that are bug-fixes have also been unmerged for roughly 2 to 3 years now. Not to mention the more recent pull-requests such as #150 and #157 that should probably be inspected and merged.
A community edition with an updated design would be great, count me in for
what is worth.
I think a javafx por would be great. I don't know if a javafx version would
use jogl.
Also maven would be great!
If someone builds it, we will follow :)
Cheers,
Tomas Lund Petersen
El vie., 5 abr. 2019 09:14, Wiehann Matthysen <notifications@github.com>
escribió:
A community edition with an updated design would be great, count me in for what is worth. I think a javafx por would be great. I don't know if a javafx version would use jogl. Also maven would be great! If someone builds it, we will follow :) Cheers, Tomas Lund Petersen El vie., 5 abr. 2019 09:14, Wiehann Matthysen notifications@github.com escribió: …
I think two pull requests that I would already consider a priority is the JOGL upgrade (see issue #1 <#1>) and the GDAL upgrade (see issue #146 <#146>). — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#81 (comment)>, or mute the thread https://github.com/notifications/unsubscribe-auth/AB_pYAATm6AVFSIWtjxkep1s_xuWHjdeks5vdz4qgaJpZM4Ldu1G .
Not sure about the JavaFX port either. I know that JavaFX has its own 3D API. I haven't personally worked with it. I am not sure whether work has already been done to port WorldWind to run in JavaFX. But any contributions with respect to that would be a welcome addition to the community edition. I know it can at least be embedded in a SwingNode and displayed in a JavaFX application (but that is different than porting it to JavaFX's 3D API).
I think we should just find a proper home for this project. It would probably be best if it lives alongside the other community editions, hence the reason I mentioned Bruce Schubert's (@emxsys) community fork of WebWorldWind.
I also think we should concentrate on providing replacement for the basemap
layers that point to the nasa servers that are going down may 4rth.
Bing aereal, sentinel2, places and roads seems to me te most importants. I
guess elevations would be important to some people also.
Cheers,
Tomas
El vie., 5 abr. 2019 09:33, Tomas Lund Petersen <kotogadekiru@gmail.com>
escribió:
A community edition with an updated design would be great, count me in for
what is worth.
I think a javafx por would be great. I don't know if a javafx version
would use jogl.
Also maven would be great!
If someone builds it, we will follow :)
Cheers,
Tomas Lund Petersen
El vie., 5 abr. 2019 09:14, Wiehann Matthysen ***@***.***>
escribió:
> I think two pull requests that I would already consider a priority is the
> JOGL upgrade (see issue #1
> <#1>) and the GDAL
> upgrade (see issue #146
> <#146>).
>
> —
> You are receiving this because you are subscribed to this thread.
> Reply to this email directly, view it on GitHub
> <#81 (comment)>,
> or mute the thread
> <https://github.com/notifications/unsubscribe-auth/AB_pYAATm6AVFSIWtjxkep1s_xuWHjdeks5vdz4qgaJpZM4Ldu1G>
> .
>
I also think we should concentrate on providing replacement for the basemap layers that point to the nasa servers that are going down may 4rth. Bing aereal, sentinel2, places and roads seems to me te most importants. I guess elevations would be important to some people also. Cheers, Tomas El vie., 5 abr. 2019 09:33, Tomas Lund Petersen kotogadekiru@gmail.com escribió: …
A community edition with an updated design would be great, count me in for what is worth. I think a javafx por would be great. I don't know if a javafx version would use jogl. Also maven would be great! If someone builds it, we will follow :) Cheers, Tomas Lund Petersen El vie., 5 abr. 2019 09:14, Wiehann Matthysen @.***> escribió: > I think two pull requests that I would already consider a priority is the > JOGL upgrade (see issue #1 > <#1>) and the GDAL > upgrade (see issue #146 > <#146>). > > — > You are receiving this because you are subscribed to this thread. > Reply to this email directly, view it on GitHub > <#81 (comment)>, > or mute the thread > https://github.com/notifications/unsubscribe-auth/AB_pYAATm6AVFSIWtjxkep1s_xuWHjdeks5vdz4qgaJpZM4Ldu1G > . >
Yes, I think that is also important. I have a Sentinel2 layer configuration file that I can share to replace the LandSat layer. It connects directly to the EOX::Maps server to get the s2cloudless layer. I don't know about an OpenStreetMap layer that we can use. Maybe we can get a layer from EOX::Maps as well. The elevation data is very important. I don't know if the NASA guys have made any plans to host it somewhere else. It would be nice to at least get some info on how to setup and host the data ourselves.
Again, all sounds good. Merging in pull request #1, a Gradle build, and alternative base layers would all be great. Not so sure about JavaFX, both because of OpenGL issues, and its exclusion from future JDKs, but I'm not too clued up on that stuff. Might be best to make anything that relies on JavaFX a separate optional module?
Though, once a home for this project has been found then issues/features can be decided upon & tracked there, and that location can be advertised. Probably worth dropping @emxsys an e-mail to get his thoughts, if he hasn't already gotten notifications from being tagged here, and then moving forward from there.
so our JavaFX port works in parallel to the Swing components. so no issue. it also uses openGL.
we also have our own jogl and GDAL patches.... as probably most of you.
and we are also on maven...
we even have tried to file fundamentaly bugs like a math bug in matrix determinant computation... and they havent picked it...
So is this version available? Id like to give it a try.
Where can i get a hold of it?
Tks,
Tomas
El vie., 5 abr. 2019 10:21, caller <notifications@github.com> escribió:
we havent polished our javaFX version jet for release... that might take some weeks in our processes. I might also need mgmt approval. Im working for a quite big company ;-)
Again, all sounds good. Merging in pull request #1, a Gradle build, and alternative base layers would all be great. Not so sure about JavaFX, both because of OpenGL issues, and its exclusion from future JDKs, but I'm not too clued up on that stuff. Might be best to make anything that relies on JavaFX a separate optional module?
Though, once a home for this project has been found then issues/features can be decided upon & tracked there, and that location can be advertised. Probably worth dropping @emxsys an e-mail to get his thoughts, if he hasn't already gotten notifications from being tagged here, and then moving forward from there.
I bumped some threads in the forum. I think Bruce Schubert would probably be notified via GitHub as a result of all the mentions. I'll email him too just in case.
Reacted by Shitesh Arihunta, Alessio Iannone and Marco Möller
I found this new server for the image layers to wich we could all switch i
think.
*https://worldwind.earth/explorer <https://worldwind.earth/explorer>*
Also id like to bring to notice that the date for the black out has been
moved from april 9th to may 3rd.
El vie., 5 abr. 2019 10:33, Alessio Iannone <notifications@github.com>
escribió:
Yes I think that we need to create separated module of the Nasa World Wind SDK, isolating math-projection library, common utility to the core features
I think that is a great idea. I mentioned it in issue #29 back in the day (though it got shot down). The idea of a WorldWindCore project would be very practical. It could contain all of the common classes such as math and geo-related classes. Then, the Android version could also depend on it and not have duplicated code in its code-base.
Speaking of the Android version. Anyone interested in getting a community for for that going as well? Or would it form part of this effort?
we where thinking on getting the android version running on iOS as well via gluon framework and porting things to openGL ES.... together with WorldWind Core this would make a true cross platform java globe
.. anyone interested? we havent started jet... only played a bit
we where thinking on getting the android version running on iOS as well
via gluon framework and porting things to openGL ES.... together with
WorldWind Core this would make a true cross platform java globe
.. anyone interested? we havent started jet... only played a bit
—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
<#81 (comment)>,
or mute the thread
<https://github.com/notifications/unsubscribe-auth/AB_pYMCp55Jpv5za2SI0gD2LlO3xKehTks5vd1MrgaJpZM4Ldu1G>
.
Bruce Schubert has established the 'official' fork of Web WorldWind that does incorporate the code improvements that NASA has not yet joined into the main branch.
https://www.linkedin.com/in/bdschubert/https://github.com/WorldWindEarth
Plus he has written up the 'recipe' for setting up a server with the elevation data. It would be good to have a few redundant sources for that.
Bruce is also working on a JSON file that would not require a recompile for WorldWind to recognize the available servers whenever that might change, allowing for it to be continually and easily updated.
Patrick Hogan, Former WorldWind PM
https://www.linkedin.com/in/phogan/
Just getting up to speed on this -- does this fork already exist? I love what you guys are thinking. Having JavaFX capability but also keeping the Swing parts working would be perfect.
Some other good (free) WMS layers are available here:
Bruce got back to me and mentioned that we can host the Java version of the community-fork on WorldWindEarth (that is: https://github.com/WorldWindEarth).
NASA Ames Research Center, where the WorldWind project has been continuously hosted since 2002 had proposed a suspension of the project, but the kick-back from other Federal agencies 'encouraged' a reconsideration. So NASA is back-in-the-saddle with regard to the WorldWind project. The suspension has now been 'suspended' for a month (until May 3), but it will likely stay alive and well ad infinitum. That is not to suggest the community should not take this opportunity to accelerate this technology 'to infinity and beyond!' Controlling our own destiny so to speak, with a community effort at a global scale. This will be beautiful, and show what's possible when working in concert 'for the benefit of all.'
@PJHogan, I agree. I think we should go forward with a community-fork irrespective of the suspension (or the suspension of the suspension). That is, if we can get a community fork going. A community fork could be very fruitful in my opinion in the long run, seeing as there are a lot of developers using WorldWind and benefiting from its continued development (myself included). I also pick up that there are developers that are very willing to contribute code-changes to WorldWind, but the way in which pull-requests and bug-fixes on this project has been managed the last couple of years has been very frustrating to say the least. That is not to say that myself and everyone using WorldWind are very grateful towards NASA and its developers for releasing WorldWind as an open-source project. I just think that going forward, the fork itself should be lead by the community.
Wiehann, of course!
And I think forks can also feed each other. :~}
Bruce has generously offered to host ^it^ at https://github.com/WorldWindEarth/
I think that would do a lot to focus the community effort in making this thing really go!
@wcmatthysen I'm ready to move forward. I can create the WorldWindJava fork with issue tracker, and a "WorldWindJava" team within the WorldWindEarth GitHub organization.
@emxsys Great, let's do it. I would love to help out where I can. I don't know exactly where I'll fit in, but I'd like to contribute some of my code and help with the ongoing maintenance of the code-base. Thanks for the link on the opensource leadership and governance.
Marco, I can only read the tea leaves, but my sense is that NASA got a wake-up call and so will continue to develop WorldWind, particularly the web version, especially given ESA has been coordinating with NASA on that development for a few years now. And the ESA contribution has been HUGE!
Unfortunately, the NASA 'official' numbered releases have been quite slow-on-the-draw (branches of the repository have many capabilities waiting for an 'official' release). FORTUNATELY https://worldwind.earth/ incorporates these, along with other 'select' contributions and bug fixes from the community.
So, my reading is that the ESA+NASA joint effort will continue.
But if you want the best of that, now and to come, it's here, https://github.com/WorldWindEarth/
Virtual globe technology is essential to so much operational and infrastructure management, as well as Earth status, scientific and other. This open source platform has much to offer the world, not only in terms of support, but also facilitating both the opportunity for innovation and the sharing of solutions. WorldWind is the epitome of an essential software technology 'for the benefit of all.'
Thanks for your contribution of this wonderful gift to the world!
Can someone clarify the branch we have to use for daily development, I
suppose it should be develop branch, can you confirm this?
Il giorno dom 7 apr 2019 alle ore 19:14 Wiehann Matthysen <
notifications@github.com> ha scritto:
Yes, as far as I understand the develop branch is the latest unstable branch where development happens. You create a new branch that branches off from the develop branch when developing a new feature (ideally with a short name that describes the feature). Then, when the feature is ready the pull request is merged back into the develop branch. From time to time the develop branch gets merged into master when a stable release is made.
@Beak-man, are the WorldWind developers at least going to try and work with the open-source community this time? Because last time I checked, and if I can quote @pdavidc on this, the policy was:
It's also worth noting that NASA World Wind is an open source project, but is closed development. Our NASA administrator can provide more insight into this if you'd like. We accept pull requests sparingly.
That didn't really inspire a community to become involved in this project to be frank. But hey, that is just my 2-cents on this.
Also, you might have a look at what we did over at the community-edition. We managed to fix a lot of outstanding issues that were never addressed here. Will you actually merge in those changes, or are you going to re-commit the changes as your own like @zglueck did in this pull-request?
@wcmatthysen , officially speaking, the point of contact to address such inquiries can be found here.
I'll tell you what I would personally like, though: Setting up the project in a similar way to this one (particularly the part about NASA's CLA) in order to improve our engagement with the open source community, which of course is part of our developer community. I'll bring it up with management and we'll see how it goes from there.
Let's allow the FUTURE tohappen!
Get this planet onto one page for spatial data!
Come on NASA and community!
Please find a way to get it together 'for the benefit of all.'
@Beak-man, @PJHogan, I'm going to be honest and say that I'm a bit pessimistic about this "renewed commitment" from NASA given the history of how the WorldWind devs managed this repository. I can already see a pull-request with work being done that has already being done by the community to some extent. The JOGL upgrade was done way back in 2016 by @xranby in pull-request #1. If you are sincere with regards to getting a community involved in this project you need to interact with them. You could start by merging some of the outstanding pull-requests that should have been pulled in a long time ago. Those stale pull-requests are just sitting there cluttering up the repository and it gives the impression that contributions are not accepted by the developers of this repository. But, from what I've seen how this repository was managed, it looks like you are going to run this again as an exclusive club where only the WorldWind devs interact with one another. Hopefully I'm wrong and you guys can rebuild a community. I just get the feeling that most of the community have already left and moved on.
@wcmatthysen You are quite accurate in your concern!
And thanks dearly for your incredible support over the years. This repository certainly deserves the kind of respect you so accurately describe as lacking.
It will be interesting to see if NASA can finally rise to the occasion.
@wcmatthysen Thanks for the feedback. A couple of points. 1. You should raise your concerns to the project management folks, they can be reached via the contact page on the web site. The people who respond to you here don't make the decisions about project direction. 2. We will be reviewing the old PRs as we work on the WWJ code. PR #1 is likely invalid as we're working with the new unreleased JOGL code (last modified on Sept 4 or so.) so as to be Java 11+ compatible.
@wcmatthysen A third point. The NASA version of WorldWind has to be curated in a way that makes certain government agencies comfortable with using the software. That being said, I'm hearing positive noises from project management regarding community engagement. Unfortunately, only time will tell if your pessimism is justified or not.
NASA WorldWind just got new releases for Java and Web.
Web WorldWind 0.10.0 and WorldWind Java 2.2.0 are now available on GitHub. The new version of Web WorldWind addresses potential vulnerabilities in the code and underlying packages. The new version of WorldWind Java focuses on upgrading to Java 11 and JOGL 2.4 as well as a switch to Apache 2.0 license. WorldWind’s API remains largely unchanged in this release and we are committed to maintaining a consistent API in future releases. More information on the release can be found at these links: Web WorldWind 0.10.0 and WorldWind Java 2.2.0.
Activity
wcmatthysen commentedon Apr 5, 2019
I think two pull requests that we should already consider a priority is the JOGL upgrade (see pull-request #1) and the GDAL upgrade (see pull-request #146). I mean the JOGL pull request has been waiting for a merge now for over 3 years.
wcmatthysen commentedon Apr 5, 2019
These pull-requests: #18, #19, #30, #138, #139 that are bug-fixes have also been unmerged for roughly 2 to 3 years now. Not to mention the more recent pull-requests such as #150 and #157 that should probably be inspected and merged.
kotogadekiru commentedon Apr 5, 2019
wcmatthysen commentedon Apr 5, 2019
Not sure about the JavaFX port either. I know that JavaFX has its own 3D API. I haven't personally worked with it. I am not sure whether work has already been done to port WorldWind to run in JavaFX. But any contributions with respect to that would be a welcome addition to the community edition. I know it can at least be embedded in a SwingNode and displayed in a JavaFX application (but that is different than porting it to JavaFX's 3D API).
I think we should just find a proper home for this project. It would probably be best if it lives alongside the other community editions, hence the reason I mentioned Bruce Schubert's (@emxsys) community fork of WebWorldWind.
kotogadekiru commentedon Apr 5, 2019
wcmatthysen commentedon Apr 5, 2019
Yes, I think that is also important. I have a Sentinel2 layer configuration file that I can share to replace the LandSat layer. It connects directly to the EOX::Maps server to get the
s2cloudlesslayer. I don't know about an OpenStreetMap layer that we can use. Maybe we can get a layer from EOX::Maps as well. The elevation data is very important. I don't know if the NASA guys have made any plans to host it somewhere else. It would be nice to at least get some info on how to setup and host the data ourselves.arihunta commentedon Apr 5, 2019
Again, all sounds good. Merging in pull request #1, a Gradle build, and alternative base layers would all be great. Not so sure about JavaFX, both because of OpenGL issues, and its exclusion from future JDKs, but I'm not too clued up on that stuff. Might be best to make anything that relies on JavaFX a separate optional module?
As for visibility, it might be worth bumping your original thread in the forum, and posting on this one as well: https://forum.worldwindcentral.com/forum/world-wind-community/worldwind-general/158839-worldwind-suspension.
Though, once a home for this project has been found then issues/features can be decided upon & tracked there, and that location can be advertised. Probably worth dropping @emxsys an e-mail to get his thoughts, if he hasn't already gotten notifications from being tagged here, and then moving forward from there.
caller commentedon Apr 5, 2019
so our JavaFX port works in parallel to the Swing components. so no issue. it also uses openGL.
we also have our own jogl and GDAL patches.... as probably most of you.
and we are also on maven...
we even have tried to file fundamentaly bugs like a math bug in matrix determinant computation... and they havent picked it...
caller commentedon Apr 5, 2019
#155
kotogadekiru commentedon Apr 5, 2019
quonn77 commentedon Apr 5, 2019
Yes I think that we need to create separated module of the Nasa World Wind SDK, isolating math-projection library, common utility to the core features
quonn77 commentedon Apr 5, 2019
So which are the next steps? Some one has get in touch with Bruce Schubert's?
caller commentedon Apr 5, 2019
we havent polished our javaFX version jet for release... that might take some weeks in our processes. I might also need mgmt approval. Im working for a quite big company ;-)
wcmatthysen commentedon Apr 5, 2019
I bumped some threads in the forum. I think Bruce Schubert would probably be notified via GitHub as a result of all the mentions. I'll email him too just in case.
caller commentedon Apr 5, 2019
more things I might have for sharing:
quonn77 commentedon Apr 5, 2019
I have added support for DXF (Autocad) and MapBox
kotogadekiru commentedon Apr 5, 2019
wcmatthysen commentedon Apr 5, 2019
I think that is a great idea. I mentioned it in issue #29 back in the day (though it got shot down). The idea of a WorldWindCore project would be very practical. It could contain all of the common classes such as math and geo-related classes. Then, the Android version could also depend on it and not have duplicated code in its code-base.
Speaking of the Android version. Anyone interested in getting a community for for that going as well? Or would it form part of this effort?
caller commentedon Apr 5, 2019
maybe I am interested in android as well... not sure jet
caller commentedon Apr 5, 2019
we where thinking on getting the android version running on iOS as well via gluon framework and porting things to openGL ES.... together with WorldWind Core this would make a true cross platform java globe
.. anyone interested? we havent started jet... only played a bit
kotogadekiru commentedon Apr 5, 2019
PJHogan commentedon Apr 5, 2019
wcmatthysen commentedon Apr 5, 2019
Bruce mentioned these links regarding alternative servers for imagery and elevation:
https://worldwind.earth/#nasa-worldwind-project-suspension
https://worldwind.earth/elevation-server.html
https://worldwind.earth/elevation-data.html
Also check:
https://gist.github.com/emxsys/f8c7a8dd5cf0060387ce50aa3f186bac
https://gist.github.com/emxsys/e3eac29221eec86d1d42525517acb3ad
mainstringargs commentedon Apr 5, 2019
Just getting up to speed on this -- does this fork already exist? I love what you guys are thinking. Having JavaFX capability but also keeping the Swing parts working would be perfect.
Some other good (free) WMS layers are available here:
https://developer.earthdata.nasa.gov/gibs/gibs-api-for-developers
Although it requires updating the WMSTiledImageLayer to capitalize the url query parameters:
https://github.com/nasa/World-Wind-Java/blob/7c9886ab67ac03d53bdb04f161b9605d3f3dd810/WorldWind/src/gov/nasa/worldwind/wms/WMSTiledImageLayer.java#L189
mainstringargs commentedon Apr 5, 2019
And I vote for gradle
PJHogan commentedon Apr 5, 2019
wcmatthysen commentedon Apr 6, 2019
Bruce got back to me and mentioned that we can host the Java version of the community-fork on WorldWindEarth (that is: https://github.com/WorldWindEarth).
PJHogan commentedon Apr 6, 2019
PJHogan commentedon Apr 6, 2019
NASA Ames Research Center, where the WorldWind project has been continuously hosted since 2002 had proposed a suspension of the project, but the kick-back from other Federal agencies 'encouraged' a reconsideration. So NASA is back-in-the-saddle with regard to the WorldWind project. The suspension has now been 'suspended' for a month (until May 3), but it will likely stay alive and well ad infinitum. That is not to suggest the community should not take this opportunity to accelerate this technology 'to infinity and beyond!' Controlling our own destiny so to speak, with a community effort at a global scale. This will be beautiful, and show what's possible when working in concert 'for the benefit of all.'
wcmatthysen commentedon Apr 6, 2019
@PJHogan, I agree. I think we should go forward with a community-fork irrespective of the suspension (or the suspension of the suspension). That is, if we can get a community fork going. A community fork could be very fruitful in my opinion in the long run, seeing as there are a lot of developers using WorldWind and benefiting from its continued development (myself included). I also pick up that there are developers that are very willing to contribute code-changes to WorldWind, but the way in which pull-requests and bug-fixes on this project has been managed the last couple of years has been very frustrating to say the least. That is not to say that myself and everyone using WorldWind are very grateful towards NASA and its developers for releasing WorldWind as an open-source project. I just think that going forward, the fork itself should be lead by the community.
PJHogan commentedon Apr 6, 2019
Wiehann, of course!
And I think forks can also feed each other. :~}
Bruce has generously offered to host ^it^ at https://github.com/WorldWindEarth/
I think that would do a lot to focus the community effort in making this thing really go!
emxsys commentedon Apr 6, 2019
@wcmatthysen I'm ready to move forward. I can create the WorldWindJava fork with issue tracker, and a "WorldWindJava" team within the WorldWindEarth GitHub organization.
emxsys commentedon Apr 6, 2019
For consideration: https://opensource.guide/leadership-and-governance/
wcmatthysen commentedon Apr 6, 2019
@emxsys Great, let's do it. I would love to help out where I can. I don't know exactly where I'll fit in, but I'd like to contribute some of my code and help with the ongoing maintenance of the code-base. Thanks for the link on the opensource leadership and governance.
PJHogan commentedon Apr 6, 2019
Wahoo!
WorldWind Java, WWJ, going from world-class to global-class!
Android can't be far behind!
caller commentedon Apr 6, 2019
Hi Patrick,
Just to clarify: nasa will host tile servers but they will probably stop writing new code. Is that a correct summary?
I have seen some NASA ESA collaborations on WW over the years. Do you know the status of the ESA engagement?
Thx Marco
PJHogan commentedon Apr 6, 2019
Marco, I can only read the tea leaves, but my sense is that NASA got a wake-up call and so will continue to develop WorldWind, particularly the web version, especially given ESA has been coordinating with NASA on that development for a few years now. And the ESA contribution has been HUGE!
Unfortunately, the NASA 'official' numbered releases have been quite slow-on-the-draw (branches of the repository have many capabilities waiting for an 'official' release). FORTUNATELY https://worldwind.earth/ incorporates these, along with other 'select' contributions and bug fixes from the community.
And I believe the ESA Copernicus Sentinel apps are also pulling from there:
https://itunes.apple.com/us/app/copernicus-sentinel/id1036738151
https://play.google.com/store/apps/details?id=it.acsys.android.esaersapp
So, my reading is that the ESA+NASA joint effort will continue.
But if you want the best of that, now and to come, it's here,
https://github.com/WorldWindEarth/
caller commentedon Apr 7, 2019
Thx Patrick,
So we should still go full speed ahead on the java and android side with our community approach. 😊
PJHogan commentedon Apr 7, 2019
Virtual globe technology is essential to so much operational and infrastructure management, as well as Earth status, scientific and other. This open source platform has much to offer the world, not only in terms of support, but also facilitating both the opportunity for innovation and the sharing of solutions. WorldWind is the epitome of an essential software technology 'for the benefit of all.'
Thanks for your contribution of this wonderful gift to the world!
wcmatthysen commentedon Apr 7, 2019
Just a heads-up to anyone that comes across this:
The community-fork of WorldWindJava has been created at: https://github.com/WorldWindEarth/WorldWindJava
thanks to Bruce (@emxsys).
Please feel free join in and contribute. We need all the help we can get to make this a success.
quonn77 commentedon Apr 8, 2019
wcmatthysen commentedon Apr 8, 2019
Yes, as far as I understand the develop branch is the latest unstable branch where development happens. You create a new branch that branches off from the develop branch when developing a new feature (ideally with a short name that describes the feature). Then, when the feature is ready the pull request is merged back into the develop branch. From time to time the develop branch gets merged into master when a stable release is made.
Beak-man commentedon Sep 3, 2019
Hello all,
Fortunately, WorldWind has received renewed commitment from NASA for continued development and support.
More information in the project's website:
https://worldwind.arc.nasa.gov/news/2019-08-21-project-update/
wcmatthysen commentedon Sep 4, 2019
@Beak-man, are the WorldWind developers at least going to try and work with the open-source community this time? Because last time I checked, and if I can quote @pdavidc on this, the policy was:
That didn't really inspire a community to become involved in this project to be frank. But hey, that is just my 2-cents on this.
Also, you might have a look at what we did over at the community-edition. We managed to fix a lot of outstanding issues that were never addressed here. Will you actually merge in those changes, or are you going to re-commit the changes as your own like @zglueck did in this pull-request?
Beak-man commentedon Sep 4, 2019
@wcmatthysen , officially speaking, the point of contact to address such inquiries can be found here.
I'll tell you what I would personally like, though: Setting up the project in a similar way to this one (particularly the part about NASA's CLA) in order to improve our engagement with the open source community, which of course is part of our developer community. I'll bring it up with management and we'll see how it goes from there.
PJHogan commentedon Sep 7, 2019
Let's allow the FUTURE tohappen!
Get this planet onto one page for spatial data!
Come on NASA and community!
Please find a way to get it together 'for the benefit of all.'
wcmatthysen commentedon Sep 7, 2019
@Beak-man, @PJHogan, I'm going to be honest and say that I'm a bit pessimistic about this "renewed commitment" from NASA given the history of how the WorldWind devs managed this repository. I can already see a pull-request with work being done that has already being done by the community to some extent. The JOGL upgrade was done way back in 2016 by @xranby in pull-request #1. If you are sincere with regards to getting a community involved in this project you need to interact with them. You could start by merging some of the outstanding pull-requests that should have been pulled in a long time ago. Those stale pull-requests are just sitting there cluttering up the repository and it gives the impression that contributions are not accepted by the developers of this repository. But, from what I've seen how this repository was managed, it looks like you are going to run this again as an exclusive club where only the WorldWind devs interact with one another. Hopefully I'm wrong and you guys can rebuild a community. I just get the feeling that most of the community have already left and moved on.
PJHogan commentedon Sep 7, 2019
@wcmatthysen You are quite accurate in your concern!
And thanks dearly for your incredible support over the years. This repository certainly deserves the kind of respect you so accurately describe as lacking.
It will be interesting to see if NASA can finally rise to the occasion.
markpet49 commentedon Sep 8, 2019
@wcmatthysen Thanks for the feedback. A couple of points. 1. You should raise your concerns to the project management folks, they can be reached via the contact page on the web site. The people who respond to you here don't make the decisions about project direction. 2. We will be reviewing the old PRs as we work on the WWJ code. PR #1 is likely invalid as we're working with the new unreleased JOGL code (last modified on Sept 4 or so.) so as to be Java 11+ compatible.
markpet49 commentedon Sep 8, 2019
@wcmatthysen A third point. The NASA version of WorldWind has to be curated in a way that makes certain government agencies comfortable with using the software. That being said, I'm hearing positive noises from project management regarding community engagement. Unfortunately, only time will tell if your pessimism is justified or not.
DJViking commentedon Aug 13, 2020
NASA WorldWind just got new releases for Java and Web.
The news article on the shutdown seems to have been removed
https://worldwind.arc.nasa.gov/news/...ension-notice/
Instead there is now an news article that says NASA has recommitted to WorldWind
https://worldwind.arc.nasa.gov/news/...roject-update/
Does not look like the community fork is needed.