span delta does not start from NOW?

[mrzimgjugl asdf]

Thanks to Andrew Milner, I can now fix wrong min/max lines using span with delta. Now min/max correspond mostly with the graphs above them. Good thing WeeWX is almost escellent and almost anything can be achieved with it. So, I am expecting this should be fixable.

Using span delta, I am getting some wierd results not seen on the graphs directly above to which I tried to sync these min and max lines.

Problem may be with the reference from which span delta starts and that reference seems not to be NOW. Possibly, the midnight before NOW is taken as a reference and starting point for the span? I can see the problem in many cases. How to fix that?

Current report was generated at monday 9:15. Graph correctly starts at 9:00 at Sunday and it just added new segment after 9:00 monday, so it shows 27 hours. I have modified summary min/max text below the graph into:

max: $span(hour_delta = 27).outTemp.max at $span(hour_delta = 27).outTemp.maxtime.format("%A %k:%M")<br/>
min: $span(hour_delta = 27).outTemp.min at $span(hour_delta = 27).outTemp.mintime.format("%A %k:%M")

But the text says:
max: 25,6°C at Sunday 15:25
min: 15,0°C at Sunday 6:55

Sunday 6:55 should be out of scope since I generated report at monday 9:15. I do not see in manual a precise definition what is the starting point for span delta. So, I estimate it is not using NOW as the end starting point but something else.

My question: what is the reference point in time for commands like:
$span(hour_delta=27)
$span(day_delta=7)
$span(day_delta=30)

I want spans to summarise the same period as the graphs above them. They all summarise form the moment when the report is generated back for given time: 27 hours or 7 or 30 days. Not take entire calendar days from the last midngiht or something like this. I even tried to give all spans in hours, but I do not think that helps. This is not the simplest issue to test.

[Pila]

I have just realized I probably misspoke and invented a problem!

I made a mistake of looking the start time of the graph and assuming it to be 27 hours ago! It is not since graph steps in 3 hours intervals.

Since I was working on this some time between other things I do, timeline (my reports) changed in the meantime and I failed to take that into account.

Span actually works as advertised taking NOW as a reference point. And does it literary. Graph is not. And both work fine.

Am I right to say I was wrong?