Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Activity
    Jody Garnett
    @jodygarnett
    (admitted some of that motivation evaporated with Java 8 collecting different options for “support”)
    Torben Barsballe
    @tbarsballe
    I think a milestone would probably be the better option, last time we changed the actual release schedule it took quite a while to go back to normal
    Jody Garnett
    @jodygarnett
    @tbarsballe updating the spreadsheet to reflect moved classes
    do we still hope to make a SED script or publish this thing?
    for the work done post-sprit I do not see the moves being recorded?
    Torben Barsballe
    @tbarsballe
    Good question. Did we ever determine if a sed script would actually be feasible? (i.e. has anyone tested a minimal case yet). A sed script would be nice, but given the size of the changes is likely going to be a lot of work to do.
    As for the moves, yeah, you’re right. I’ll update everything I’ve moved
    Jody Garnett
    @jodygarnett
    @jnh5y indicated he had experience
    as long as we are matching the full classnames a sed script should be both possible, and get client code 95% of the way there (reflection for example may fail).
    Jody Garnett
    @jodygarnett
    https://stackoverflow.com/questions/51608196/replacing-multiple-strings-in-multiple-files
    the good news is the most recent changes to gt-complex did not have any affect on the geoserver codebase
    Torben Barsballe
    @tbarsballe
    A couple of the changes I made had minor effects
    See geoserver/geoserver#3256
    Torben Barsballe
    @tbarsballe
    Just had a thought - at current, the nightlies won’t work for Java 11, because the only ones we publish are built on Java 8.
    Jody Garnett
    @jodygarnett
    I thought if we build on Java 8 … it can work on Java 11
    Torben Barsballe
    @tbarsballe
    Maybe, thought there was a reason it couldn’t; I’m verifying now
    Jody Garnett
    @jodygarnett
    It was only the reverse that was not true (ie build on Java 11 the result cannot run on Java 8 because of type narrowing API changes)
    Torben Barsballe
    @tbarsballe
    Yeah, you’re right - just confirmed that the current 2.15.x nightly works fine on Java 11
    I’d got my directions the wrong way around
    Jody Garnett
    @jodygarnett
    so for the spreadsheet - sqlite?
    Torben Barsballe
    @tbarsballe
    Do you know what gt uses spatiallite for
    I don’t, and don’t really know what might break if we drop it
    (I’m currently working on the GS Running Java 11 docs)
    Jody Garnett
    @jodygarnett
    I added a line for the geotools docs and am working on that
    Jody Garnett
    @jodygarnett
    I am having trouble testing locally since I hava GDAL / OGR 1.x on the path which ends up failing geoserver tests
    Torben Barsballe
    @tbarsballe
    Hmm. That shouldn’t fail the tests
    I’ve got GDAL 1.11.1 locally and my GeoServer tests all pass
    Jody Garnett
    @jodygarnett
    I will need to look at it later
    okay geotools docs are added to our branch, for your review?
    back in a couple hours
    Torben Barsballe
    @tbarsballe
    I’ll take a look
    Torben Barsballe
    @tbarsballe
    Reviewed.
    I’ve pushed up some GeoServer Java 11 docs to the GeoServer branch, if you want to take a look
    They feel a little sparse, but there’s not a whole lot to say for GeoServer...
    Jody Garnett
    @jodygarnett
    thanks, answering your feedback now, please advise.
    James Hughes
    @jnh5y
    @tbarsballe @jodygarnett most of my knowledge is captured by things like what's in the JTS sed commands: https://github.com/locationtech/jts/blob/master/MIGRATION.md
    that said, I do believe it is possible to build up a sed script with a list of transformations (e.g. one sed script with 500 changes rather than 500 scripts each with one change)
    Jody Garnett
    @jodygarnett
    Want to give it a go? We have our sheet with somewhat less then 500 changes :)
    James Hughes
    @jnh5y
    Ah... this week is rather bad for me to jump into something like this...
    what's the timeline for needing something?
    Torben Barsballe
    @tbarsballe
    We want to try and push out a release early-mid December, so before then?
    Jody Garnett
    @jodygarnett
    or longer term before Jan RC :)
    Mostly want to get an milestone out for geoserver users
    geotools developers can always use master if they are keen
    Jody Garnett
    @jodygarnett
    Some sprint results to share:
    Jody Garnett
    @jodygarnett
    Attending FOSS4G-NA code sprint? https://wiki.osgeo.org/wiki/FOSS4GNA_Code_Sprint_2019
    María Arias de Reyna
    @Delawen
    I wish :)
    Andrea Aime
    @aaime
    Unlikely
    Jody Garnett
    @jodygarnett
    General idea for these sprints that involve an extra day of accommodation for attendees; it may be good to offer a solid count on the conference ticket as a good way to encourage community participation (or acknowledge community participation).
    Jody Garnett
    @jodygarnett
    Pictures are up https://flic.kr/s/aHsmDv6FwQ