OpenStreetMap
For tomorrow’s eclipse I built a small free tool that reconstructs the ground-level view anywhere in the visibility zone, with buildings and forests from OSM extruded over real terrain, and the eclipse computed minute by minute. I posted the technical details on the community forum this week. The tool is at 2026eclipse.eu. Feedback on how your area renders is very welcome, especially where height tagging is sparse.
Discussion
Comment from mccartney436 on 12 August 2026 at 06:07
Thanks. Nice tool!
I’ve just tried this out and for my location it seems to show the times off by one hour (?). Some timezone issue (?).
https://2026eclipse.eu/#52.13663,21.04434,18.02
Comment from spiregrain on 12 August 2026 at 07:33
Brilliant!
Comment from mccartney436 on 12 August 2026 at 07:52
Re my timezone issue. The page says: “ (local time UTC+1) P1 start 18:14:47 Maximum 19:06:08 · mag 0.862 P4 end 19:55:22 “ While my local time is UTC+2 (CEST). Which confirms it’s a timezone issue.
Comment from RenaudZi on 12 August 2026 at 09:26
Thanks @mccartney436 for the report, that was a useful one.
The timezone normally comes from the weather API, which returns the offset for the location itself. When that call fails, the site was falling back to a rough estimate based on longitude, which ignores daylight saving time. For Warsaw that gives UTC+1 instead of UTC+2, so everything shows one hour early. Exactly what you saw.
It’s now fixed: the fallback uses a real timezone table instead of the longitude guess. Reload the page and you should get the right times.
One thing I’m curious about, if you still have it open: did the panel show “UTC+1” instead of “GMT+2” at the bottom, and did you get a message about the weather being unavailable? That would confirm the diagnosis.
Comment from mccartney436 on 12 August 2026 at 09:47
@RenaudZi:
Also - thanks for the tool! It’s super useful as we are planning which place to pick for tonight.