Another fun quirk.

[adam5]

Hi Steve,

Had another interesting CWOP issue this morning. Temperatures here dipped a few degrees below 0F this morning. After the temperature reached the -1F mark, Cumulus starting sending some interesting temperature data to CWOP. I've pasted in the CWOP data strings over a couple hour stretch, starting with the date/time group for brevity. As soon as the temp "warmed" back to 0F, the data was fine again. Data sent to WU as well as the 10-minute logged data during this time was unaffected. Logged data in WeatherLink is fine as well, so this does not appear to be a sensor problem. Could it be a negative number issue?

170858z4459.66N/08439.48W_273/000g000t000r000P000h86b10287eCumulusDsVP
170903z4459.66N/08439.48W_272/000g000t-00r000P000h86b10287eCumulusDsVP
170908z4459.66N/08439.48W_273/000g000t-18446744073709551615r000P000h86b10285eCumulusDsVP
170913z4459.66N/08439.48W_272/000g000t-18446744073709551615r000P000h86b10285eCumulusDsVP
170918z4459.66N/08439.48W_272/000g000t-18446744073709551615r000P000h87b10285eCumulusDsVP
170923z4459.66N/08439.48W_272/000g000t-18446744073709551615r000P000h86b10285eCumulusDsVP
170928z4459.66N/08439.48W_272/000g000t-18446744073709551614r000P000h86b10284eCumulusDsVP
170933z4459.66N/08439.48W_272/000g000t-18446744073709551614r000P000h86b10283eCumulusDsVP
170938z4459.66N/08439.48W_273/000g000t-18446744073709551614r000P000h86b10282eCumulusDsVP
170943z4459.66N/08439.48W_272/000g000t-18446744073709551614r000P000h86b10282eCumulusDsVP
170948z4459.66N/08439.48W_273/000g000t-18446744073709551614r000P000h86b10282eCumulusDsVP
170953z4459.66N/08439.48W_272/000g000t-18446744073709551614r000P000h86b10282eCumulusDsVP
170958z4459.66N/08439.48W_272/000g000t-18446744073709551614r000P000h86b10281eCumulusDsVP
171003z4459.66N/08439.48W_272/000g000t-18446744073709551614r000P000h87b10279eCumulusDsVP
171008z4459.66N/08439.48W_272/000g000t-18446744073709551615r000P000h87b10279eCumulusDsVP
171013z4459.66N/08439.48W_273/000g000t-18446744073709551615r000P000h86b10278eCumulusDsVP
171018z4459.66N/08439.48W_272/000g000t-18446744073709551615r000P000h87b10277eCumulusDsVP
171023z4459.66N/08439.48W_272/000g000t-18446744073709551615r000P000h86b10277eCumulusDsVP
171028z4459.66N/08439.48W_272/000g000t-18446744073709551615r000P000h87b10278eCumulusDsVP
171033z4459.66N/08439.48W_273/000g000t-18446744073709551615r000P000h87b10277eCumulusDsVP
171038z4459.66N/08439.48W_272/000g000t-18446744073709551615r000P000h87b10277eCumulusDsVP
171043z4459.66N/08439.48W_272/000g000t-00r000P000h87b10277eCumulusDsVP

Certainly made for an interesting data plot in CWOP!

http://www.findu.com/cgi-bin/wxpage.cgi?call=DW3024

-Brian

[adam5]

Not sure how my post showed up twice. I might have hit the browser "back" button and hit refresh.

My apologies.

-Brian

[adam5]

OK, discovered the delete feature and deleted to duplicate post.

-Brian

[steve]

The Delphi routine that I use to format the temperature into CWOP format must be doing something strange with negative numbers - I'll investigate.

[steve]

I've found the problem; I'll release a fix tomorrow.

[adam5]

Awesome,

What was the issue, if you don't mind me asking.

-Brian

[steve]

if you don't mind me asking.

I do, actually.

The required format for the temperature is a 3-digit integer, or if negative, a minus sign and a 2-digit integer. For negative values, I intended to negate the value, convert to a two digit string (using a system function) and manually stick a minus on the front. I forgot to negate, and the system function I use didn't like the negative value very much.

But it was good that we actually found and fixed a bug in Cumulus!

[adam5]

Hey Steve,

Thought I'd give you follow up report since your software update to correct the below 0F issue. Have since had at least one morning with temps below 0F and the software worked just fine. Thought you might want to know.
-Brian

[steve]

Thanks, Brian.