Would be great to have a new release of this module.
From @Simon Georges post in #1563648-10: Taking a break from geocoder development
Miscellaneous RTBC issues:
- #1908428: Remove trailing spaces and CVS keywords
- #1911770: Typo: Get the geomtry of a GPX string or file
- #1515372: Caching Results
Regarding bulk-update:
- #1451942: Provide drush-command to re-geocode all empty values
- #1735046: Generate geocode info for existing nodes with address field (->Batch update)
- #1928280: Geocode existing nodes
Regarding preserving existing results in various cases:
- #1509988: Add flag to prevent geocoding
- #1686620: Skip Geocoding if values already exist
- #1493496: When programmatically creating many nodes, 10k+ is there a way to skip geocoding if it has already been geocoded?
Regarding geocoding accuracy (and especially country support):
- #1872086: "Casilla" doesn't geocode properly
- #1380570: Reject geocoding if accuracy requirements are not met
- #1748412: Geofield Mapquest Geocoder
- #1871962: Use Countries module to improve geocoding results
Additional services / settings support:
- #1542842: Bing Geocoding Handler
- #1580512: Google geocoding with a Google business enterprise account
- #1549254: Google routing support
- #1914132: Entity property
& @rudiedirkx's in #13
Comments
Comment #1
rudiedirkx commentedI think #1914132: Entity property and #2159925: Geocode from virtual fields/entity property instead of just real fields try to solve (almost) the same thing: encode from a non-Field. Different approach. The boss should pick one.
Comment #2
basvredeling@maintainers If we get this list cleaned up, do you feel confident to release another stable minor like 1.3?
Comment #3
simon georges commentedI will publish a new release as soon as this list (or current Needs Review / RTBC issues) are cleaned up.
Comment #4
basvredelingRemoved duplicate entry
Comment #5
polWe're slowly getting there :-)
Comment #6
rudiedirkx commented2 new important (related) bugs need fixing first though:
Comment #7
rudiedirkx commentedPol, could you choose which one to keep and mark the other one as duplicate?, because they have a combined fix for how errors and caching work (together). And maybe add it in this issue's body?
Comment #8
polHi,
I think the most important bug before the release is this one: #2649096: Fix exception throwing and empty result handling
After that, we're good for a release, I think.
Comment #9
rudiedirkx commentedI agree that's the issue, but the notice isn't the most important. And the other one has a
!wrong, so you have to fix that. I don't think we agree on the what-makes-an-error, so that's up yo you, but empty results should be cacheable, so fix the combination, not just the notice.I will make a patch tomorrow for how I would see the error & caching as perfect. Up to you then what you use.
Comment #10
rudiedirkx commentedMaybe #1493496: When programmatically creating many nodes, 10k+ is there a way to skip geocoding if it has already been geocoded? is still important, even though it's 3y old. With the new 'caching instead of diffing' method, we (or I) kinda broke it for sites with many geocoded entities that update many at the same time. Since their cache is still empty, it'll lookup all of them at once, even though nothing's changed, to fill the cache. I didn't think about that. Relevant for many sites with programmatic updates.
Comment #11
polWe have to give some love to #1580512: Google geocoding with a Google business enterprise account and we're ready I think.
Comment #12
polIt's done, thank you all !!!
Comment #13
basvredelingAwesome, Thanks!