FAQ: Difference between revisions

All available via the [[[[Downloads|Downloads page]]]] in this Wiki

As of 2017, development of both Cumulus 1 and Cumulus MX was abandoned by Steve Loft. Although there have been a couple of patches to change the years listed in drop downs with Cumulus 1, its code is frozen as the relevant development environment is no longer available.  

* The highest of the station's 3-second readings over the previous 10 (configurable by adding ''AvgSpeedMinutes=N'' in [[Cumulus.ini#Section:_Station]] where 'N' is number of) minutes becomes the Cumulus 'Gust (available via <#wgust>, no matching direction web tag). Note that the WMO definition of a wind gust is the mean wind speed over a 2 to 3 second sample period (so Cumulus reports this within an extended 10, or as configured, minute period, not the 'latest' gust).

Select value for 'N' using ''AvgSpeedMinutes=N'' in [[Cumulus.ini#Section:_Station]].

* A 10 (configurable by adding ''AvgSpeedMinutes=N'' in [[Cumulus.ini#Section:_Station]] where 'N' is number of) minutes average of the station's 'Wind' value will then display as Cumulus 'Average'.

See ''LCMaxWind=X'' parameter in [[Cumulus.ini#La_Crosse_specific:]] where X is the maximum (uncalibrated) wind speed in m/s that you want to allow.

**As you will see if you look in [[diags]], Cumulus stores the 'total rain count' figure at rollover, so by maintaining the current figure it can subtract off the start of day figure to deduce the rainfall today and this is shown on the front screen in Cumulus 1, labelled ''Today''. In Cumulus 1 the '''Edit''' menu gives access to the ''Today's rain'' screen and that provides the ability to edit the start of day figure so that the figure shown for today is correct. Equivalent functionality is provided in Cumulus MX. If there is no rain in a day, the start of day count will be the same at the start of the next day.

**Important, for the <#ryear> web tag, and for the 'This year' display on the ''main'' Cumulus screen, the processing reads the ''dayfile.txt'' for the rainfall season starting with the month set on the station settings screen and stored as the value for attribute ''RainSeasonStart'' in [[Cumulus.ini#Section:_Station]] (its default is January).   

For some Oregon Scientific models, the temperature from an extra sensor can replace the temperature from the main sensor for processing by Cumulus 1, but all extra sensor values are stored in a series of [[Extra_Sensor_Files| monthly log]] files.  

No other processing is performed for Davis extra sensors, but past values are retained in a series of [[Extra_Sensor_Files| monthly log]] files.  The file [[strings.ini]] is used to customise the description associated with extra sensors.

Cumulus was originally something Steve Loft 'knocked together' quickly when I got my first weather station, simply because I couldn't find any other software anywhere which I was prepared to use. No 'design' went into it, it was really just a prototype. Some of the things I did were just quick 'hacks' that I intended to fix when I did it 'properly'. But I never got around to going back and doing it 'properly', and it was never my intention that anyone else would use it. But I was persuaded to make it available for others, and then to add more and more enhancements.  

So yes, I am aware that some of things that Cumulus does may seem strange, but it is simply not possible for me to start again from scratch, knowing what I know now. And it is not possible for me to effectively produce separate versions of Cumulus specifically for each type of weather station. If you don't like the way Cumulus does certain things, do what I did and write your own software!

*It is possible to back up the Cumulus files from one installation running off a weather station, and briefly have another reduced functionality copy just looking at historic data without ever any updating (the second Cumulus is set up as if reading from an EasyWeather.dat file, but not given a path/file name) - see [[Cumulus.ini#Introduced_for_problems_with_Fine_Offset_family:|Easyweather.dat section here]].

{{Version badge 1}}The supplied web page templates are part of Cumulus and will be over-written on upgrade. If you want to create your own web pages, don’t edit the supplied ones, place them in a different folder and use the ‘extra files’ facility. From version 1.8.9, the installer offers the option of not over-writing the web templates; it unconditionally also installs them to a sub-folder of the web folder called 'originals', so you can refer to them if required, even if you chose not to over-write those in the web folder.

{{Version badge 1}}You’re probably running Windows Vista/Windows 7/Windows 8 (or one of the corresponding server versions of Windows). Look in C:\Users\[username]\AppData\Local\VirtualStore\Program Files\Cumulus - note that this is a 'hidden' folder. Ideally, to avoid this situation, you would install to somewhere outside the Program Files hierarchy, as recommended in the readme.txt.

{{Version badge 1}}You've probably corrupted your Cumulus.ini file in the default Cumulus folder. See [[Cumulus.ini#Section:_NOAA]]. It can only be edited either (if Cumulus is stopped) with an external text file editor or (with Cumulus running) using the ''NOAA Setup'' in the configuation menu.  Pay attention to where the quotes are required for the file names, and check that you have an allowed date format (see [[Webtags#Time.2FDate_.27format.27_Parameter]]).

You've almost certainly entered your co-ordinates incorrectly, or not entered them at all - on the station settings screen. You must enter DEGREES, MINUTES, and SECONDS, no other format will work. Make sure to use '''unsigned whole numbers''' for degrees, minutes, and seconds, with a 'W' for your longitude if you're West of the Greenwich meridian (e.g. if you are in the United States), and the appropriate letter for your latitude (N for northern hemisphere, S for southern hemisphere).  

If you are viewing or editing log files at the time when rollover takes place, it is likely that some parts of the daily rollover will fail.  Although Cumulus 1 provides the functionality to view and (except for the data file) update the log files, make sure you have exited out of the built-in viewer or editor before rollover is due. If you inadvertently forget this, the cure of overwriting files mentioned above is applicable if your weather station has an internal log and it is around 2 logging intervals (see [[#Cumulus_takes_a_long_time_to_download_the_data_from_my_Davis_station_at_start_up]]) after rollover. When using an external editor, close Cumulus 1 or Cumulus MX before opening any Cumulus file.

Choosing the wrong station type means individual items of information are not correctly assigned to weather parameters.  Cumulus knows what is at what location for each station type.  

In general, if Cumulus 1.x.x identifies an odd value it will reuse the last good value instead, assuming that the problem will either right itself, or be spotted and dealt with quickly.

'''Fine Offset stations''' are prone to reporting the odd non-sensical value, and Cumulus has the ability to compare successive readings:

*From the '''Configuration''' menu select [[Cumulus_Screenshots#Configuration_Menu_Screens|''Calibration'']] and in the '''Spike removal''' frame you can specify "the maximum '''differences''' between successive values" that Cumulus is to tolerate. See the '''Help''' for that screen, and note that spike removal applies while Cumulus is running, not to catch-up readings from the station's logger.  

The [[Diags]] log often contains useful information for sorting out what has been read, which values have been ignored, and which highs or lows may have been updated incorrectly. Note that automatic backups are made just after roll-over time because restarting Cumulus when there are no logger records to read before roll-over has been known to cause problems. Therefore, the backup to choose might be one with the date before when the problem occured, but remember that any calculations Cumulus makes are based on readings at least every minute when Cumulus is running, but for the catch up period such calculations can only use the readings available at your station's logging interval. Note if you choose a later backup and it does not work, you can always then repeat the 'rewind' with a earlier backup.

Easy - correct today's total using the [[Today.ini#Editing_rainfall_in_today.ini_within_Cumulus | 'today's rain']] editor on the edit menu (select from main screen in Cumulus 1, or find in the MX browser interface).

# The counts are for completed days, not part days. The threshold for whether a day counts as a wet day is set in the optional parameter [[Cumulus.ini#Section:_Station|RainDayThreshold= -1]], and the parameter [[Cumulus.ini#Section:_Station|NoFlashWetDryDayRecords=1]] is also relevant.

Note that this method does not work for the Oregon WMR200 station. This station deletes logger entries when they are downloaded, and there is no mechanism for retrieving logger entries for a particular period - it just sends all the entries that it has. You may be able to circumvent this by creating a today.ini file in the data folder before you run Cumulus 1, with just this in it:

Note that Cumulus MX stores some time-stamps '''in ISO format''' (yyyy/MM/dd HH:mm:ss) within '[[today.ini]]', the format example above only works for Cumulus 1, see the article on the log file itself for specifics.

Stop Cumulus and edit [[Cumulus.ini#Section:_Station | Cumulus.ini (station section)]] within the main Cumulus folder. Change the "StartDate=" line. Be sure to use the same format for the date.

[[File:Badge vMx.png]] However in calculating and updating extremes, Cumulus MX will ignore any monthly log file data prior to the start date set, so it is crucial to make sure the start date is right before you (re-)start MX.

{{Version badge 1}}It really represents "Cumulus began on", so there is an alternative of leaving cumulus.ini unchanged, but editing default Cumulus ''recordsT.htm'' template twice:

# that you do not have a 'EWpressureoffset=x.y' line in the [Station] section of [[Cumulus.ini#Introduced_for_problems_with_Fine_Offset_family:| cumulus.ini]],

Normally when Cumulus starts, it reads in [http://www.jim-easterbrook.me.uk/weather/mm/ the fixed block] in the console a pair of absolute/station pressure and relative/sea level pressure readings and uses the difference between these as the ‘offset’ in millibar for the station;  Cumulus uses this offset subsequently to calculate the sea level pressure (each time when reading the absolute pressure in the weather memory locations of the station), but it is possible to define the offset in '''Cumulus.ini''' instead.  

'''IMPORTANT'''  The above applies to Cumulus 1.  bvs79669 did some investigation in January 2019 that reveals the mapping of memory bytes to pressure reported by Jim Easterbrook and used by Steve Loft is wrong. The two most significant bits included in the mapping create impossibly high readings for pressure. From build 3047, MX excludes these two bits and the reliability of reading correct pressure from a Fine Offset station has been significantly improved.

{{Version badge 1}}You probably have an error in one or more data files. Look near the beginning of the latest file in the Diags folder for an error message which will tell you which file is in error and which line is causing the problem. It may simply be a blank line. Edit the file with Cumulus stopped. Note that you may need to repeat this process; only the first error is shown.

If you've been running Cumulus for more than a few days, but the graph is still empty, the most likely explanation is that it was unable to read your dayfile.txt file. Look at the latest file in the 'diags' folder (the files are plain text) for error messages mentioning dayfile.txt.

You can over-ride the start up check by adding a line to the [Station] section of [[Cumulus.ini#Section:_Station|cumulus.ini]], thus: ''NoSensorCheck=1''

If you are using the "Synchronise" setting but are still getting frequent console interface lockups, then the clocks on the console and the platform running Cumulus may be drifting too far between synchronisations. Look in the diags files for the word "drift" and see how many seconds apart the clocks are drifting. If it is more than 3 seconds, you will need to adjust the period that Cumulus allows where it doesn't read the console data. Stop Cumulus and edit Cumulus.ini. In the [Station] section, add a line:

*With MX stopped, edit Cumulus.ini. In the [Station] section set (or add if missing) the two entries...

*If you want to stop Cumulus uploading the indoor temperature graph, you can un-tick 'Include standard images' on the 'files' tab of the internet settings, and list all of the images that you still want to be uploaded as extra files. Cumulus 1 configuration only allows you to add 10 files so you would need to do this by editing Cumulus.ini, ironically these notes do not apply to Cumulus MX but in that all 100 slots can be edited within the settings pages.

*Or you have edited/customised the template and added an incorrectly specified Cumulus [[webtag]] that has caused [[Customised_templates#What_is_meant_by_.27Cumulus_processes_templates.27| Cumulus processing]] to abort for that template.  If you stop cumulus you can use a text editor to open the latest [[Diags|diagnostic file]]. Look at entries time-stamped just before the hour and repeated at your auto-update ''interval'' (just below red arrow in screen image above). You might find a message like "Error processing C:\web\indexT" (i.e. mentioning whichever template is at fault) followed by a brief hint as to the nature of the error.

There's a limit of 10 entries on the files tab in Cumulus 1.9.x (in Cumulus MX all 100 entries are included in the settings interface).  Entries for the full 100 files can also be amended by editing [[Cumulus.ini#Section:_FTP_site|cumulus.ini]], with Cumulus stopped. Follow the format of any existing entries for specifying additional files.