Where does alltime.rec get its data from?

[bigharps]

I searched thru quite a few posts and found nothing directly relating to my problem, which really isn't earth shattering, but annoying nonetheless.
I've attached a jpeg of my alltime records screen. The alltime low of -6.9 is a furphy - it's unlikely we will get as low as +6.9 all winter.
Being a tinkerer, I thought I'd look at May10log.txt and fix the delinquent entry and see if that corrected the records. When I looked at May10log.txt, it actually shows a sensible figure of 16.3 - consistent with other readings about that time.
23/05/10,16:49,16.3,70,10.7,3.2,6.8,32,0.0,0.0,1021.0,1648.3,17.9,56,3.2,16.3,16.3
23/05/10,16:50,16.3,70,10.7,0.0,6.8,0,0.0,0.0,1021.0,1648.3,17.9,56,0.0,16.3,16.3
23/05/10,16:51,16.3,70,10.7,6.1,6.8,29,0.0,0.0,1021.0,1648.3,17.9,56,6.1,16.3,16.3
23/05/10,16:52,16.3,70,10.7,2.9,6.8,29,0.0,0.0,1021.1,1648.3,17.9,56,2.9,16.3,16.3
23/05/10,16:53,16.3,70,10.7,1.8,6.8,29,0.0,0.0,1021.1,1648.3,17.9,56,1.8,16.3,16.3
23/05/10,16:54,16.3,70,10.7,0.0,6.1,0,0.0,0.0,1021.2,1648.3,17.9,56,0.0,16.3,16.3
23/05/10,16:55,16.3,70,10.7,4.0,6.1,30,0.0,0.0,1021.2,1648.3,17.9,56,4.0,16.3,16.3
23/05/10,16:56,16.3,71,10.9,2.2,6.1,30,0.0,0.0,1021.2,1648.3,17.9,56,2.2,16.3,16.3
23/05/10,16:57,16.2,70,10.6,2.5,6.1,29,0.0,0.0,1021.1,1648.3,17.9,56,2.5,16.2,16.2
23/05/10,16:58,16.3,71,10.9,2.9,6.1,29,0.0,0.0,1021.1,1648.3,17.9,56,2.9,16.3,16.3
23/05/10,16:59,16.2,71,10.8,0.0,6.1,0,0.0,0.0,1021.1,1648.3,18.0,57,0.0,16.2,16.2
So.... where else did alltime.rec get its info?

And related, but not the same...
The high wind speed reading on January 31 is no way correct at 184km/h.
The logfile for January shows that Mr LaCrosse fed a whole lot of inncorrect data around that time. Who knows why? It is hard wired, so it's unlikely to be an RF fault.

31/01/10,11:57,24.9,81,21.4,5.4,5.8,27,0.0,5.2,1015.1,1414.2,25.2,76,5.4,24.9,24.9
31/01/10,11:58,24.9,81,21.4,5.4,5.4,22,0.0,5.2,1015.2,1414.2,25.3,76,5.4,24.9,24.9
31/01/10,11:59,24.9,81,21.4,5.4,5.4,22,0.0,5.2,1015.2,1414.2,25.3,76,5.4,24.9,24.9
31/01/10,12:00,24.9,81,21.4,5.4,5.4,22,0.0,5.2,1015.2,1414.2,25.3,76,5.4,24.9,24.9
31/01/10,12:01,24.9,81,21.4,184.0,184.0,22,0.0,5.2,1015.1,1414.2,25.3,76,184.0,24.9,24.9
31/01/10,12:02,24.9,81,21.4,184.0,184.0,23,0.0,5.2,1015.0,1414.2,25.4,75,184.0,24.9,24.9
31/01/10,12:03,24.9,81,21.4,184.0,184.0,22,0.0,5.2,1015.1,1414.2,25.4,76,184.0,24.9,24.9
31/01/10,12:04,24.9,81,21.4,184.0,184.0,22,1.5,5.2,1015.1,1414.2,25.6,76,184.0,24.9,24.9
31/01/10,12:05,0.0,0,0.0,184.0,184.0,22,0.0,5.2,1015.1,1414.2,25.5,76,184.0,0.0,0.0
31/01/10,12:06,0.0,0,0.0,184.0,184.0,22,0.0,5.2,1015.2,1414.2,25.6,76,184.0,0.0,0.0
31/01/10,12:07,0.0,0,0.0,184.0,184.0,22,0.0,5.2,1015.0,1414.2,25.7,75,184.0,0.0,0.0
31/01/10,12:07,0.0,0,0.0,184.0,184.0,22,1.0,5.2,1015.1,1414.2,25.8,75,184.0,0.0,0.0
31/01/10,12:08,0.0,0,0.0,184.0,184.0,22,1.0,5.2,1015.0,1414.2,25.8,74,184.0,0.0,0.0
31/01/10,12:09,0.0,0,0.0,184.0,184.0,22,1.0,5.2,1015.0,1414.2,25.8,74,184.0,0.0,0.0
31/01/10,12:10,25.5,78,21.3,0.0,184.0,0,1.0,5.2,1015.0,1414.2,25.8,74,0.0,25.5,25.5
31/01/10,12:11,25.6,78,21.4,6.5,184.0,27,1.0,5.2,1015.0,1414.2,25.9,74,6.5,25.6,25.6
31/01/10,12:12,25.6,78,21.4,3.6,184.0,27,1.0,5.2,1015.0,1414.2,25.9,73,3.6,25.6,25.6
31/01/10,12:13,25.7,78,21.5,4.3,184.0,27,1.0,5.2,1015.1,1414.2,25.9,73,4.3,25.7,25.7
31/01/10,12:14,25.8,78,21.6,0.0,184.0,0,1.0,5.2,1015.0,1414.2,25.9,73,0.0,25.8,25.8
31/01/10,12:15,25.9,77,21.5,3.6,184.0,29,0.5,5.2,1015.1,1414.2,25.9,73,3.6,25.9,25.9
31/01/10,12:16,25.9,77,21.5,4.3,184.0,30,0.5,5.2,1015.0,1414.2,25.9,73,4.3,25.9,25.9
31/01/10,12:17,25.9,77,21.5,3.2,184.0,31,0.5,5.2,1015.0,1414.2,25.9,73,3.2,25.9,25.9
31/01/10,12:18,26.1,77,21.7,9.7,13.3,37,0.5,5.2,1014.9,1414.2,26.0,74,9.7,26.0,26.1
31/01/10,12:19,26.1,76,21.4,2.5,13.3,37,0.0,5.2,1014.8,1414.2,26.0,73,2.5,26.1,26.1
31/01/10,12:20,26.1,76,21.4,9.7,13.3,35,0.0,5.2,1015.1,1414.2,26.0,73,9.7,26.1,26.1
My question is, if I insert sensible entries for wind speed and temperatute for the stupid stuff on Jan 31, will that update the alltime.rec file, especially seeing the correct temp on May 23 is corrupted for the alltime records?
I am loathe to just try changing the records in case the data to alltime.rec comes from somewhere else, seeing alltime does not agree with the May 23 data on the log file?
I hope this is not too long winded. Like the program and the professional looking web pages for displaying the weather.
My version is 1.8.8, build 881

[PaulMy]

Bigharps,
The highs and lows can be edited directly in the [Edit] [All time records] screen and stays that way until a current new record occurs. Not sure why the records got corrupted but when I had seen obvious errors in my high/lows (and like yours, the monthly data appeared to be valid) I searched my data (with Weatherlink's Strip Charts was quite easy to find the historical high/lows) and edited/inserted the correct value, date, and time in Cumulus. If I recall, the current Cumulus version logs the high/low changes and/or edits.

Paul

[bigharps]

Thanks PaulMy - I sort of knew the answer was going to be too simple. Well getting the chart correct anyway.
Still a bit of a puzzle for Steve(?) where that -6.9 came from.
Thanks.

[steve]

Each data reading is examined to see if it creates a new all time record. There are many more data readings than are logged in the log files, even at 1-minute intervals - I believe a wired La Crosse updates every 8 seconds? So it is quite possible for a bad reading to register an all-time record but not appear in the log (which is just a snapshot in Cumulus 1).

You could use the EWTempDiff setting in cumulus.ini to filter out any sudden erroneous changes in temperature: http://wiki.sandaysoft.com/a/Cumulus.ini

Cumulus 1.8.9 has a check for the 'error' value which the La Crosse generates when it can't supply a wind speed; I think your 184 corresponds to that value. You might also want to try the LCMaxWind setting in cumulus.ini, if you tend to get other 'too high' values which aren't the 'error' value.