GeoTools/ GeoServer meeting 7/3/17
Attending: Jody, Ian, Torben, Kevin, Jukka, Andrea
Agenda:
1.
Release planning for release candidate
2.
CodeSprint planning and technical decisions
3.
GEOT-5632 SLD 1.0 / SE 1.1 Differences
4.
Library Upgrades / Wicket Upgrade Discussion
5.
Bug Stomp Discussion / Cleanup
Actions:
-
JG/?? Release Thursday/Friday
-
DB: Look at dependency upgrade for GEOS-7920
-
All: Look for code sprint sponsors
-
TB/JG: Find most complicated REST finder, and sort out an approach as
part of prep
-
Release planning for Release Candidate
Release delayed for bug stomp: GT 17-RC1 / GS 2.11-RC1 / GWC 1.11-RC1
Release volunteer(s):
-
Jody Volunteered to work on this Thursday/Friday (after bug stomp)
-
Need to make new builds and branches
Did we get any response to the beta on Jira?
-
nothing specific
-
Review for anything serious / regressions
GeoServer Search Here
<https://osgeo-org.atlassian.net/issues/?jql=project%20%3D%20GEOS%20AND%20affectedVersion%20%3D%202.11-beta>
(14 issues - 5 open)
-
GEOS-7997 Vector Tiles are not displaying correctly using Mapbox GL
(improvement)
-
GEOS-7920 Vector Tile Extension on Geoserver look "ugly" on MapBox GL JS
-
GEOS-8010 Regression: dbtype and filetype store arguments should not be
visible
GeoTools Search Here
<https://osgeo-org.atlassian.net/issues/?jql=project%20%3D%20GEOT%20AND%20resolution%20%3D%20Unresolved%20AND%20affectedVersion%20%3D%2017-beta%20ORDER%20BY%20priority%20DESC%2C%20updated%20DESC>
(8 issues)
-
GEOT-5590 Build Failure on Windows
-
GEOT-5588 SVGGraphicsFactoryTest testLocalURLXEE windows failure
Nothing above should hold up the RC1, we could take a run on the windows
build failures during RC timeline.
Consider: Upgrading dependencies now for RC1 (see agenda topic below)
CodeSprint planning and technical decisions
Planning:
-
Hotel? Talk to Simone for details/recommendations
-
Transport? Consider rental car (should be able to share or carpool)
-
Email treasu...@osgeo.org with travel receipt for reimbursement
-
We are WAY OVER BUDGET ON TRAVEL (due to everyone booking late)
-
Please reach out to sponsors; or we will be making the next teams
sprint planning harder (Looking to make up 1-2k here)
-
Astun - can we ask them?
-
Vivid - may be interested due to work on REST API
-
Please consider who you can ask
A lot of prep is needed before we hit the ground March 27th.
-
We dug into REST API docs (that is not our biggest technical risk)
-
Want to have a working example to start from
-
Toreben recommend the Style endpoint as being straightforward
-
Approach
-
1) make sure we have decent tests to start from before migration
-
XML good
-
JSON weaker (check and possibly add)
-
Ideally we want an actual response for a regressions test
-
2) Migrate to Spring MVC
-
Resource → Controller class annotated with end points
-
Use example for guide to use of annotations
-
Intend to heavily use annotations
-
Pull variable names out of paths
-
3) Migrate JSON and XML bindings (should be able to re-use)
-
JSON 1.2 upgrade planned
-
4) Carefully migrate controller path management
-
One controller per resource, from then on in configured by
annotations
-
risk: Finders can have some logic
-
Action: Find most complicated finder, and sort out an approach
as part of prep
-
5) Endpoint discussion
-
6) Risk Importer
-
This is not normal rest config
-
This is a very flexible
-
7) Risk Resource Store
-
Simple design, but supports arbitrary paths
8) Other extensions / community modules to look out for
-
Extensions
-
Monitor, importer, xslt
-
Community
-
sld service, scripting, rest-upload, jms-cluster, geofence,
backup/restore
-
Geogig has a custom rest endpoint
-
9) Doc prep
-
Automation
-
RST Example
-
Output of prep:
-
Notes and examples
-
https://github.com/geoserver/geoserver/wiki/REST-API-Refresh
Any open technical questions / risks:
-
Doc automation? Jody does not care we just need to pick one :)
-
Solid example of migrating bindings showing use of xstream
Action:
-
JG/TB/KS have some prep time
-
Expect some discussion on the email list
GEOT-5632 and SLD 1.0 / SE 1.1 Differences
Link: https://osgeo-org.atlassian.net/browse/GEOT-5632
SE 1.1 has anchorPoint x=0.5 y=0.5 (for both labelPlacement and
pointSymbolizer)
SLD 1.0 has anchorPoint x=0.0 y=0.5 for labelPlacement, and mandates 0.5
and 0.5 for the point symbolizer, without any way to change it
So we sort this out in the parser, by the time we hit the SLDStyleFactory
it is too late :(
Formats:
-
SLD - this is our library default
-
SE - fill in the default values
-
CSS - maps direct to SLD 1.0 defaults (so it should be fine)
-
ysld - should fill all the values (it does not rely on geotools defaults)
-
mbstyle - should fill all the values (it does not rely on geotools
defaults)
Action:
-
Warning that fixing GEOT-5632 will have some visual changes for the
release
-
Take SLD 1.0/SE 1.1 into a new bug!
[image: achorPoint.png]
Library Upgrades / Wicket Upgrade Discussion
Library upgrades:
-
wicket (see email for 7.6.0 recommendation)
Consider:
-
Jetty? etc..
-
going to stick with wicket
Bug Stomp Discussion / Cleanup
-
See GEOT-5632 above
-
GEOT-5461 Ian asks for help encoding the transaction
-
dblasby found that gt-wfs-ng was sending a WFS 2.0 request the wrong
version of GML??
-
Justin may be able to help
--
Ian Turton
------------------------------------------------------------------------------
Announcing the Oxford Dictionaries API! The API offers world-renowned
dictionary content that is easy and intuitive to access. Sign up for an
account today to start using our lexical data to power your apps and
projects. Get started today and enter our developer competition.
http://sdm.link/oxford
_______________________________________________
Geoserver-devel mailing list
Geoserver-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/geoserver-devel
