IMPORTANT - New build of 1.8.4 available

[steve]

I'm repeating this announcement here as it's a serious bug affecting Fine Offset users. All versions before 1.8.4 build 604 area affected.

I have replaced the existing download of version 1.8.4 with a new build - 604.

This fixes a bug which is important for Fine Offset users, who should upgrade as soon as possible. The bug is in the processing of the history, when handling the wrap-around from the end of the memory back to the start, and causes gaps in the Cumulus data logs. I thought releasing a new build of 1.8.4 was simpler than going for a new version at this point.

http://sandaysoft.com/downloads

Steve

[steve]

Unbelievably, I only managed to half-fix the bug. I have now uploaded build 605. Please download and install this build.

My only excuse is that I am on two lots of medication trying to shake off the remains of this damn virus and a consequent chest infection.

Many, many apologies.

Steve

[keithatrochdale]

Any ideas why I have no min pressure for today and max/min temp and max/min temp pressures for yesterday showing?

On 605, but is the same with earlier versions.

[steve]

Problems like this are usually to do with incorrect rollovers from one day to the next. Possibly the pre-604 bug. If you downloaded the archive data using a version with the bug, you would still have the potential for incorrect data. Have a look in dayfile.txt and see if there is an entry for yesterday in there.

When you say you have no minimum pressure for today, what is it actually displaying?

Steve

[keithatrochdale]

0.0 mb

Will check the file later, off flying for now.

many thanks and hope you feel better soon.

[steve]

I think in that case it actually got a zero pressure value from somewhere. Do your all-time records show this zero too? A quick fix would be to stop Cumulus and edit the [Pressure] section in today.ini to put the current value in for the low (or some other better estimate). The displayed 'yesterday' values are transient anyway and will disappear tomorrow. You could look in yesterday.ini to see what values and date are in there. Please feel free to send me those two files and I'll see if I can work out what happened. Dec08log.txt might give a few clues too.

Steve

[harrym1byt]

Thanks Steve...

Might it be worth winding back the data using one of the backup directories?

[steve]

Might it be worth winding back the data using one of the backup directories?

Yes, it's worth trying.

Steve

[keithatrochdale]

Ok here are the files, guess we can see why there is no yesterday reading; now the question is why.

I do close the prog. by clicking on the top right red X and then answer yes I do want to close, could this be not saving the data?

Thanks

[steve]

It does seem to be a problem of the day rollover not happening correctly. It looks like you shut down last night at about 23:40 and started up again at about 0907? There does seem to be a problem with shutting down just before midnight, and I wonder if the fact that you station logger is set to 30 minutes is a factor? Perhaps if you shut down less than your logger interval before midnight... But I thought I had code in to cater for that case.

If you were to rewind to the backup files saved when you started up yesterday (at about 1000?) I think it would all correct itself - but of course you would lose the 5-minute logging from Cumulus, replaced by the 30-minute logging from your station.

Could you please send me the files from your Diags folder? And I'll look at the code and see if I can spot the bug.

Thanks,
Steve

[keithatrochdale]

Ahh, I just expected the close down process to do an automatic save.

Will sent Diag via e-mail, attaching them is causing problems, does not like the extensions.

[steve]

Yes, it does save everything when you shut down. The bug is in detecting the start of a new day when processing the data from the logger. I think I've found the bug, it is pretty much as I suggested. If the first entry in the archive to be processed causes a day rollover (e.g. yours was the entry at 00:07) , it performs the day rollover correctly, but it hasn't actually read any data yet from the logger so the data that it's saving for the end of day is invalid (probably all zero, hence your zero pressure reading).

I'll check the files you've sent me, but I'm pretty sure that I've found the bug. I need to work out the best way of fixing it. In the meantime, the circumvention is to not shutdown just before midnight within your station logger period - i.e. in your case with your 20-minute logger interval, don't shut down between 23:30 and midnight). I realise this is a bit of a pain, and I'll try to get a fix asap.

Steve

[keithatrochdale]

Ok, got the max pressure sorted ok, but I have now noticed that the min pressure is 0.0mb at 00:07 and min temp is 0.0 C at 00:07.

(Steve; closed software down at 11:47 without problem)

[steve]

Ok, got the max pressure sorted ok, but I have now noticed that the min pressure is 0.0mb at 00:07 and min temp is 0.0 C at 00:07.

This is with the 1.8.5 beta, having closed at 23:47 the previous night? Have you had the same problem today?

Steve

[keithatrochdale]

This is with the 1.8.5 beta, having closed at 23:47 the previous night? Have you had the same problem today?

Steve

Sorry, still no min temp or min pressure for yesterday I am afraid.