Cumulus Support

In that case what happens if you use the rapidfire url / action ?

OR

If you delay your test by say 30m / 60m / 90m ?

beteljuice wrote:In that case what happens if you use the rapidfire url / action ?

I don't think that will help; I don't really understand how rapid fire works, but it extracts updates periodically for display in the 'history'. I don't know how it decides which ones to use, but I suspect that some of the uploads will get 'lost' if I use the rapidfire URL.

If you delay your test by say 30m / 60m / 90m ?

I've done some experimenting, feeding in updates manually and adjusting the timestamp. Any timestamp in the last 10 minutes gets ignored and the current time is used instead. I wonder if they are assuming that the PC clock is off slightly and are trying to 'correct' it? I'll send them an email and see if they have an explanation.

I wish I could help but I have no idea how coding works.

I will offer my thanks for giving it a go.

Randy

wd40 wrote:I will offer my thanks for giving it a go.

I think it's probably near enough as it is; it just means that the last "archive" update after start-up may get timestamped by WU a few minutes later than it should be. I've emailed them but have yet to receive a reply.

I will be glad to give it a try if that would help.
My data base is young enough that any loss or corruption of data would not be a problem.

Randy

wd40 wrote:I will be glad to give it a try if that would help.

You don't really need Cumulus, to see the effect. If you want to experiment (or anyone else, for that matter), and feel comfortable with doing this, you can copy one of the URLs quoted above (from 'http:' to 'updateraw' inclusive) , change it so it has your ID and password in it, and then copy it to your browser and experiment with the date and time values to see if you get the same result that I did. You'll have to work out what UTC is relative to your timezone. Wunderground will let you delete individual entries, so your WU database should be OK when you've finished.

You should find that a timestamp more than 10 minutes ago is reported correctly, but anything less than 10 minutes ago gets timestamped with the time you sent the update.

Wunderground will let you delete individual entries, so your WU database should be OK when you've finshed.

Steve, How you do this ?

I did a couple of experiments some days ago, but I stop the experiments avoiding messing up my database.

yv1hx wrote:Steve, How you do this ?

If you're logged in, you get a 'delete specific entries' button at the end of your tabular data: If you click on it, the data is redrawn with a 'delete entry' link on each entry:

I started by using Firefox 3.6 with the URL and my id and password. I wanted to see what would happen with the data up to date with the console plugged in to the USB and Cumulus running.

The return message from WU was the single word was "success"

That part worked so onward.

I changed the time several times and it inserted the data at the time I specified so I think it works just as you said it would.

How does WU know the time span of the backlog with only one time?

Randy

wd40 wrote:I changed the time several times and it inserted the data at the time I specified so I think it works just as you said it would.

But if you specify a time less than 10 minutes ago, it uses the current time instead?

How does WU know the time span of the backlog with only one time?

Each update is just a snapshot at the specified time; it doesn't cover a particular time span.

Steve, it want take the data until after the 10 min.

I tried the update set for 18:30 starting at 18:31, then at 18:33 then again at 18:40. The 18:40, ten minutes after the update I was inserting and that worked.

As you see, at 18:33, it did update but at Cumulus current values I think.

I am set for 5min updates, rapidfire. Cumulus was running at the time.


18:27 79.0 °F 57.8 °F 29.71in North 3.1mph 4.5mph 50% 0.00in Delete entry
18:30 52.5 °F 51.4 °F 29.82in SE 36.0mph 40.0mph 96% 0.05in Delete entry
18:32 78.8 °F 57.6 °F 29.72in North 3.8mph 4.5mph 50% 0.00in Delete entry
18:33 78.8 °F 57.0 °F 29.71in NE 3.8mph 4.5mph 49% 0.00in Delete entry
18:37 78.6 °F 58.1 °F 29.71in NE 3.1mph 5.4mph 51% 0.00in

If you're logged in, you get a 'delete specific entries' button at the end of your tabular data:

Good point!

Usually I'm not logged in WU while doing "experiments", because I have a PC running the Weather Station and I use my laptop looking the results as any other Internet user.

Thanks Steve!

wd40 wrote:Steve, it want take the data until after the 10 min

Yes, I think it must be some sort of correction for PC clocks which are slightly out. Other than waiting 10 minutes before uploading the archive entries (which seems daft to me), there's nothing I can do about it. We'll just have to accept that the last upload from the archive may be timestamped a few minutes late.

Surely if you are uploading 'historic' to WU, it must be because Cumulus was off - So won't the last record be 'current' anyway when Cumulus is restarted and caught up with station logs ?

The last 10m window is irrelevent assuming that is their minimum 'resolution' that they work to, or am I talking babblecrap ?