Basemaps in geographic coordinates in the future?

[FrancescaRiolo]

Hello,
Anybody knows if ESRI is planning to provide the new ArcGIS Online basemaps also in geographic coordinates (wkid:4326) in the near future? This was the spatial reference used in the old basemaps (currently under "Extended support" status) while the new map services are all in WebMercator (wkid: 102100). This means always having to project data stored in geographic coordinates.
Thank you!

[JoeHershman]

ESRI changed these to match with Google and Bing which use WebMercator.  This way the tiling scheme is the same for all the primary base maps sources.  I cannot see that they would ever change back to WGS as it would mean having to support two different sets of cached maps.

I agree that WGS84 makes a lot more sense to use as a projection

[derekswingley1]

If you want to convert coordinates from web mercator to wgs84, use esri.geometry.webMercatorToGeographic.

The coordinate conversion is done client side and does not require anything from a server.

[JeffPace]

i actually feel its a "miracle" everyone (GOOGLE, BING, ESRI, OPEN STREET MAPS) found a way to all agree on a coordinate system.  I think lat/long would be a monster step backwards

I think we are actually at a state where the coordinate system doesnt really matter, and as developers should be nearly or completely transparent to the end user.

[FrancescaRiolo]

I don't know if we can say they agreed or maybe that everybody felt forced to follow Google...but I guess they picked the best system.
Still, after years of "avoiding on the fly projection" to improve performance I am reluctant to go for it now... I don't know if the fact that is run client side makes me feel reassured! Maybe is better to start storing data in web Mercator. Time and testing will tell.
Thank you everybody for the input!

[JeffPace]

ok maybe agree wasnt the right word, how about "begrudgingly work together"

For us, not a single agency we work with (including ourselves) would ever stored data in lat/long.  We have been using on the fly projection in the data frame to web mercator and performance has been more than satisfactory.

Just tile more, and the coversion is precomputed, at least for display 🙂