Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Repo info
Activity
  • Aug 01 2018 17:48
    @jnh5y banned @matrixbot
Jody Garnett
@jodygarnett
good night
Martin Davis
@dr-jts
good night! (not in PST right now?). Thanks for doing the release!
James Hughes
@jnh5y
@jodygarnett @dr-jts congrats on the release! Apologies for not helping push the final buttons. (I'll try and line up time better in the future.)
Jody Garnett
@jodygarnett
I am at the bolsena code sprint, so this was a "good" sprint activity.
@jnh5y no worries about pushing the final buttons, the important part was having good release steps to follow. I did clarify a few things (so you could review?).
Does the https://projects.eclipse.org/projects/locationtech.jts/releases/1.19.0 page need to be updated somehow now that the release is out?
James Hughes
@jnh5y
Yes! That page should be updated with a few more commits that made it into the release!
Jody Garnett
@jodygarnett
I thought TWKB was addressed? Is that not the case locationtech/jts#246
James Hughes
@jnh5y
Yes, that issue was addressed. It got merged after the release notes were created. That's at least one of the things which needs to be updated in the release notes (there may be others)
Jody Garnett
@jodygarnett
James Hughes
@jnh5y
Are you able to take a spin through the release notes and update them? (If not, lemme know and maybe I can get to it later?)
Jody Garnett
@jodygarnett
I really cannot, sprint is wrapping up today and I have many things at 70%
indeed Ian is working to make a pull request for geotools to use the new JTS
a lot of the database tests fail due too JTS doing a better job of dimensinon and measure
geotools has its own litecoordinatesequence that is causing fun
James Hughes
@jnh5y
Sounds good. Say hi to the rest of the gang! And sorry that I didn't catch the changes in GeoTools... I must have hit some issues running the unit tests in my branch previously
Jody Garnett
@jodygarnett
these tests needed a database; it is okay.
LiteCooordinateSequence should have its own test coverage; and perhaps be retired now that JTS alternatives
Björn Harrtell
@bjornharrtell
Any chance of reviving interest in locationtech/jts#714 ? I'm still generallty concerned about it and blocked by it.
Jody Garnett
@jodygarnett
Do we have a lead on the cause / fix …
Do we have a lead on cause / fix?
Björn Harrtell
@bjornharrtell
I'm actually waiting for your answer there 🙂
The cause it's described in the issue and is clear enough I hope.
Martin Davis
@dr-jts
I'll look at this again. Note that #715 has a conflict now
would be great if there's some interest in trying this out
Björn Harrtell
@bjornharrtell
@dr-jts: conflict resolved :)
Matt Gilene
@mdgilene
Hey, trying to figure out how to store 3D polys in PostGIS using a combination of spring-data, hibernate and JTS objects. Is this possible? From the DB side it looks like it should be supported via the POLYHEDRASURFACE type but I don't can't seem to find any support in Hibernate or JTS for that type.
Martin Davis
@dr-jts
JTS doesn't support 3D objects.
Martin Davis
@dr-jts
Sneak preview of upcoming development in JTS - Polygonal Coverage Simplfication
image.png
Chris Hodgson
@cmhodgson
Martin, is the intent to maintain the coverage relationship(s) between the polygons, eg. the resulting polygons would share edges and not overlap, at least if the input meets those criteria? It's not entirely apparent from your example but then the resolution of the input is so fine that you can't really be sure
Martin Davis
@dr-jts
Yes - and that image is not the right one! It shows the effect of the current VWSimplifier, which doesn't preserve coverage topology (or polygonal topology, either). Stand by...
image.png
This image shows the effect of the upcoming CoverageSimplifier - coverage topology is preserved.
(If you look closely you can see Luxembourg and Andorra still in existence :)
Björn Harrtell
@bjornharrtell
Phew I thought your definition of topology had become screwed up @dr-jts. 😂
Chris Hodgson
@cmhodgson
Ya that's exactly what I was getting at too, Bjorn. Though my first guess was that the input topology was not perfect and that was causing the unexpected output... I'm guessing that things like slivers (unintended, small overlaps or holes in the coverage) must cause this algorithm some trouble? It would be amazing if you could clean up slivers smaller than the simplification tolerance, while simplifying :D
Martin Davis
@dr-jts
Don't worry, I haven't lost my topological marbles! That was just the wrong image (thanks to @cmhodgson for pointing that out)
And yes, if the coverage has gaps or (invalid) overlaps, the algorithm will still produce a result, but is likely to make the problems worse. GIGO, in other words.
Björn Harrtell
@bjornharrtell
@dr-jts : any time to consider locationtech/jts#715 ?
Björn Harrtell
@bjornharrtell

I'm unable to compile master, getting the following error:

[ERROR] Failed to execute goal org.apache.maven.plugins:maven-checkstyle-plugin:3.1.2:check (validate) on project jts: Failed during checkstyle execution: Failed during checkstyle configuration: unable to parse configuration stream: com.puppycrawl.tools.checkstyle.api.CheckstyleException: Property ${checkstyle.header.file} has not been set -> [Help 1]

Anyone know what this is and what can be done about it?

Martin Davis
@dr-jts
Are you building using mvn clean install ?
Martin Davis
@dr-jts
That works for me
Jody Garnett
@jodygarnett
I set it up with a single shared checkstyle configuration file.
Björn Harrtell
@bjornharrtell
Hmm looks like mvn clean install helped. I did mvn install before to no effect...
Björn Harrtell
@bjornharrtell
Sorry but I will continue to nag about locationtech/jts#715 until you tell me to stop. Please consider accepting it.
Martin Davis
@dr-jts
Still not crazy about adding that complexity for (AFAIK) a single user.... wish there was another simpler way to solve your issue
Björn Harrtell
@bjornharrtell
Not sure why you think it's so complex? As I see it it's just a simple switch to make JTS behave like it did in 1.17 and before for those concerned with Z values (which I admit is likely a minority) and it's entirely optional.