feature request solar time widget

[nutpantz]

I think it wound be cool to have a widget that just shows the solar time for the location. Am/pm,24h, metric times as options.. Mostly for the cool factor.

[forrestguice]

This is in the clockwidget branch, but it will be a few weeks before its merged/released. Its still missing some options (the "cool factor") but basically done I think.

[akostadinov]

Nice! Do you have a build one can try?

[forrestguice]

Give this a try.

[akostadinov]

Due to certificate difference, had to uninstall completely to install this build. There seems to be a small bug though. Solar time shows13:40 while TZ time shows 13:54. While "-5m 5s eot" (whatever eot means).

I installed app "Sun Time" 13:38 at the same time. So it appears to me that eot if it means difference is correct but widget shows around 1 minute more.

[forrestguice]

hrm. I'll take a closer look. The eot is the equation of time value, the offset between local mean time and apparent solar time. Adding eot to apparent solar time should yield local mean time.

[forrestguice]

Due to certificate difference, had to uninstall completely to install this build.

Sorry about that. I'm not really accustomed to building betas. The apks on F-Droid are the official release (and they sign using their own key). The linked apk is signed with my personal key. I should have put a ".beta" on the end of the package name to allow it to coexist. I'll do that next time.

Solar time shows 13:40 while TZ time shows 13:54. While "-5m 5s eot" (whatever eot means). I installed app "Sun Time" 13:38 at the same time.

I installed "Sun Time" and can see the discrepancy. Initially it was at least a minute or two, but now its showing a four minute difference after enabling gps. I think that would make sense if my location was 1 degree west. Since the "Sun Time" app doesn't tell me my current location (and it seems to be automatic), I actually suspect a bug here.

Next I verified the values returned by Suntimes by comparing against http://www.jgiesen.de/astro/suncalc/index.htm. It's close to expected; on 2019/01/06 at 12:59 local time (Mountain Standard; UTC-7) at location 34.58, -112.52 the solar time offset is -35.065 for a solar time of 12.399. The widget reports 12:23 while the "Sun Time" app reports 12:22. Hrm, very close, so what's going on here..

On a hunch I reconfigured the location in Suntimes. I changed 34.58 to 35 and -112.52 to -113. Now I'm getting perfect agreement between "Sun Time" and Suntimes. Maybe a peek into the "Sun Time" source would reveal whats happening but I suspect some kind of decimal to integer rounding bug.

Maybe somebody else can verify all this. Does anybody really know what time it is?

[akostadinov]

I'm more concerned about the fact that local time + eot != apparent sun time. "Sun Time" might be doing some rounding, it doesn't look like a very sophisticated app. I just gave it as an additional example that something might be a little off.

[forrestguice]

local time + eot != apparent sun time

Can you show me this? I'm seeing agreement. The sign - might be confusing. In this case a value of -5m 5s indicates apparent solar time is 5m slow. If the sundial shows 1:07pm that indicates local mean time is 1:12.

[forrestguice]

Solar time shows 13:40 while TZ time shows 13:54. While "-5m 5s eot" (whatever eot means). I installed app "Sun Time" 13:38 at the same time.

Its not clear what TZ is in this case, but I suspect its the local time in your current timezone. That isn't the same as "local mean time" (which is the UTC time offset by longitude). For example in my case UTC-7 (the timezone boundary) is actually 30 minutes east of my location. When the clock says "2:00 PM" mountain time, its actually "1:30" local mean time (or 1:25 apparent solar time).

[forrestguice]

I just gave it as an additional example that something might be a little off.

Sorry if I'm sounding a little defensive. I do think something might be a little off.
I think its the "Sun Time" app calculation up to a full degree longitude.

After double checking everything I think my own calculations are more or less correct.

...edit: I actually have no idea why "Sun Time" appears to always be 2m slow. I went searching for a bug there I expected to see and failed to find it. However I'm very confident in my own local mean time calculation, as well as the equation of time result (I see some error there but its reasonable, ~6s).