jsmorley wrote: ↑June 7th, 2020, 7:29 pm
Yeah, maybe I don't fully understand, but I have never noticed a problem. Of course I NEVER use either the long or the short name for the current day, as when it is Saturday, that is "Today / Tonight" for me, and never "Saturday / Sat" as that just feels weird to me. I'm less anal about "Tomorrow", as I'm fine with either "Tomorrow" or "Sunday" for that.
No, I meant the fact that there is a delay / jump from / to the next day around 3 AM. I noticed that a long time ago.
Regarding the day names, if I correctly understand the issue raised by SilverAzide, that's just a consequence of the aforementioned day delay / jump in the middle of the night. Me, I'm using only day abbreviations, never tomorrows and such, but as I said, that wasn't necessarily what I was referring to, just the delay / jump part.
jsmorley wrote: ↑June 7th, 2020, 7:40 pm
Just as an aside, it just changed from "today" to "tonight" for me, about 3:15pm or 3:30pm or so my time. I'm pretty ok with that, as I would postulate that a "forecast" for "today" becomes less relevant about that time, and the forecast for "tonight" becomes more interesting. I guess any decisions you are going to make based on today's weather have already been made by then.
Yes! If you point your weather skin at someplace like Vietnam or where ever it is coming up on 3AM, you'll be able to watch the transition from "tonight" to "today". The About screen will show the mismatches in day names (the short day name will be ahead of the long day name).
Yincognito wrote: ↑June 7th, 2020, 7:42 pm
No, I meant the fact that there is a delay / jump from / to the next day around 3 AM. I noticed that a long time ago.
Regarding the day names, if I correctly understand the issue raised by SilverAzide, that's just a consequence of the aforementioned day delay / jump in the middle of the night. Me, I'm using only day abbreviations, never tomorrows and such, but as I said, that wasn't necessarily what I was referring to, just the delay / jump part.
Well, I must just be stupid today, but I don't see how subtracting 3 or 4 hours from the timestamp helps at all EXCEPT during that narrow window of time, as soon as the site DOES have the correct day, you are likely to see "yesterday" for at least some period of time. This feels very Rube Goldberg to me.
jsmorley wrote: ↑June 7th, 2020, 7:48 pm
Well, I must just be stupid today, but I don't see how subtracting 3 or 4 hours from the timestamp helps at all EXCEPT during that narrow window of time, as soon as the site DOES have the correct day, you are likely to see "yesterday" for at least some period of time. This feels very Rube Goldberg to me.
Yeah, I wasn't the one mentioning that subtraction, but SilverAzide - you'd have to ask him about it.
SilverAzide wrote: ↑June 7th, 2020, 7:48 pm
Yes! If you point your weather skin at someplace like Vietnam or where ever it is coming up on 3AM, you'll be able to watch the transition from "tonight" to "today". The About screen will show the mismatches in day names (the short day name will be ahead of the long day name).
Hm.. Yeah. If I change to A Lu, Vietnam, I get this, and it is still "tonight" there.
1.jpg
You do not have the required permissions to view the files attached to this post.
I'm not sure I'm in love with any solution for this. The "cleanest" seems to me to build a series of Substitute statements in the language variables file and use that to change "Sunday" into "Sun", but ick. That is going to take some doing for all the languages out there.
How about the date itself? Because for the defunct wxdata (same source, after all), that was offset as well, for like 2-3 hours in the late night / early morning...
P.S. I ask this because if the date is correct, then the day of week name can be adjusted based on that. Right?
Yincognito wrote: ↑June 7th, 2020, 7:57 pm
How about the date itself? Because for the defunct wxdata (same source, after all), that was offset as well, for like 2-3 hours in the late night / early morning...
In my A Lu, Vietnam example, it looks like this:
1.jpg
It is more or less 4am there, and the date and long name are correct.
You do not have the required permissions to view the files attached to this post.
Feels to me that the problem is not with the "short" day name based on the Observation timestamp, that is correct. It is the "long" day name, which is passed as-is in the data. But if we start replacing that with a calculation, then we presumably have to do that with a lot of things, including the "part" data, like "Today" and "Tonight".
I mean technically I guess 3am is "tonight", but not really in my view. It's "today" or at least "This morning"...
Obviously weather.com doesn't see it that way, as you are still getting "nighttime" weather at 3am.