After a lot of work, it seems like we can finally update the polygon data for Norway. They have merged a lot of counties, also the regions will change.
The reason it has taken time is that the site used to use (https://wambachers-osm.website/boundaries/) doesn't exist anymore. An old version of his existed at GitHub as open source and our initial plan was to try to use that. But everything needed wasn't there and I realized that it was easier to just implement it from scratch than trying to provide his source with data in a correct format.
That's how https://osm-boundaries.com/ was born! It's still in development, but it has gotten far enough for us to be able to download data and import it into Project-GC.
The changes for Norway are huge though, so that's the next challenge (pun not intended). The difference was so huge that the old compare-new-vs-old-polygons tool wasn't even close to good enough, so I have patched that up a bit as well. Besides better formatting of the output it now matches more properly. I hope I didn't miss any case.
A real case is that county "Re" will be removed. That matches a lot of scripts saying "Region" for example. Now it only matches things that looks like strings, on other words, when the name is surrounded by quotes or double quotes.
print "Region" - won't match Re
regions = [ "Re", "Skodje" ] - will match Re
In the next post I will add data for everything that will change and everything that will be affected. The changes are not live yet.
My plan is this:
* Go through all potentially affected scripts. If they seem to be affected, I will disable them.
** Related Challenges will then be without a challenge checker.
** If they don't get a new one or the script is fixed, the Challenges risks getting archived.
* I will disable all potentially affected tags, without going through them.
** The idea is that we will go through them one by one. Fix them if needed, and then enable them one by one.
* After fixing the scripts and disabling tags we will add the new polygon data to Project-GC. This will then be done before fixing the tags themselves.
This is something I plan to start on at Thursday this week. If there are objections/suggestions I surely will listen to them. This leaves quite a lot of work regarding the Challenge checkers, but I can't figure a better way. If there are feedback saying we should postpone, we will postpone.
Regions only in Norway:
Sogn og Fjordane
Regions only in _tmp_Norway:
Troms og Finnmark
Vestfold og Telemark
Potentially affected scripts (source code matches a region/county that has been added/removed):
scriptId:873 creator:kro8 scriptName:Alle cacher i en kommune i Hordaland matches:
scriptId:578 creator:Draupne scriptName:Generic number of finds in region matches:
scriptId:1004 creator:bjoff scriptName:SognOgFjordane, using boundaries from 31.12.2019 matches:
scriptId:494 creator:burncache scriptName:Roan65:815035 Chemistry Caching Challenge matches:
scriptId:535 creator:lillfiluren scriptName:combinationschecker matches:
scriptId:826 creator:vogelbird scriptName:GC code checker For the States ( I and L => 1) matches:
scriptId:1463 creator:famv17 scriptName:GC 1-100 matches:
scriptId:1495 creator:lingbeek scriptName:A periodic table matches:
scriptId:2486 creator:arisoft scriptName:GC-code substrings matches:
scriptId:2489 creator:vogelbird scriptName:GC code checker For the States ( I and L => 1) with Country and matches:
scriptId:731 creator:Draupne scriptName:Generic number of finds in county matches:
I have been going through the list, the following two Challenges needs new Checkers (when the new data is available):
Edited 2 time(s). Last edit at 06/09/2020 12:59PM by magma1447.
Needs a new checker that requires the Geocacher to have found all geocaches in a county within a list of counties. That list should be the counties that used to be in Hordaland.
* Granvin herad
* Kvam herad
* Os (Hordaland
* Ullensvang herad
* Ulvik herad
I have asked HQ regarding Challenges that uses the old definitions. As I see it we have two options:
1) Force-archive Challenges using counties that doesn't exist anymore.
2) Create new Challenge checker using user-defined polygons.
User-defined polygons aren't allowed for new Challenges. But as I see it, this is a different case. The Challenges already exists.
We'll see how fast HQ answers. Regardless I think we should work on at least disabling Checkers that won't work. Meanwhile those affected Challenges will be without challenge checkers, and we will then see what their destiny will become.
Script 1004 looks like the same situation - it was placed in 2017 and uses the geographical data from that time. And could currently be done using county/region data, but once the data is updated it would break. I think the CO organised the new checker on the assumption that PGC would update region/county data quicker than this.
That's actually quite good in my opinion. That's how I expect most of these Challenges will be solved. I haven't heard from HQ yet though. I expect that to take at least a week. Therefore I think that script developer had a smart approach on things. The question is if it should have been allowed by the reviewer though, but that's a bit off topic here, not really our concern.
Project-GC would have updated much sooner, at least already in February, if the site we used didn't stop updating its data back in January.
Challenges affected by this are considered "old" and were allowed when they got published. Therefore they have decided that they are allowed to keep using the old polygon data, just as their cache descriptions says.
So, for those where it's possible, we can convert the checkers into using a polygon script, if we have the polygon data. I still have Project-GC's source files for the polygons, I don't know how compatible they are yet though. I will look into that.
I know there was a challenge/script that already used its own custom polygons. If I recall correctly that was for regions, and that data should be usable for more then.