The handling of the ANETZ precipitation has been fully redone in order to be based on hourly correlations (instead of the original 6 hours). The proper coefficients will be provided soon...

Small inconsistency (although it is not sure if we will not just stop trying to support VC++ when its symbol export rules are an absolute non-sense)

Adding some potentially missing includes (<algorithm>) and addressing Cmake's policy CMP0054 (changes in the strings comparisons)

The snowpack_psum versus imis_psum behavior was not in agreement with the documentation... This has been fixed and should behave in a more appropriate way

Following an issue with an wrongly chosen coordinate system, a few improvements have been developed: the documentation has been improved, UTM zones can be provided as {num}{letter} or the opposite, and a bug has been fixed in NetCDFIO (in some cases, the opened file would not be closed).

Code cleanup in NetCDFIO (as well as mapping for the Albedo grids), extra code comment in data_converter.cc

When compiling with gcc>5 AND linking with OCCI, there is a special option to add otherwise the linking fails...

For better consistency, now using the grid's size() method instead of getNx()*getNy(). The debug messages in libinterpol2D have been removed after performing the checks (ie that our implementation of Winstral does conserve mass).

the ListonRH algorithm has been tweaked: when only 1 station is available, no detrending is applied (instead of having the detrending returning an error if no default lapse rate has been provided).

The proper gracerc file (that has to be copied in $HOME/.grace) has been committed. In order for the direct import of smet files, the smet2agr.sh script must be in the PATH.

Minor code cleanup in the filters.

Unfortunatelly, the '#' char that was used to mark hexadecimal colors is also a comment marker in smet files... So it has been removed. The colors are now specified *without* the '#' char.

Non-applicable units specs have been replaced by "-" in smet files. A toString() method has been implemented in SmetWriter for easier debugging and an error message improved for clarity.

It now works when no colors are provided in the smet file. Instead of forcing bach, now dash or bach can be invoked (it relies on /bin/sh).

Code cleanup in libsmet, better min/max for ILWR in PNGIO, support for color, description, units, min and max in smet files in SMETIO and smet2agr. The SMET specifications will still have to be updated to reflect these recent additions, though...

the key BUFF_CHUNK_SIZE has been renamed as BUFFER_SIZE for more clarity and its default value is now 370 (which makes more sense for most applications). Some documentation about it has been written.

More includes cleanup and code cleanup in the plugins (getting rid of the "stationindex" thing that was pretty much deprecated in all plugins)

The method "waterSaturationPressure" has been renamed as "vaporSaturationPressure" and a reference added.

Trying to bring clarity to the data creator / data generator concept... The API has been slightly change (renaming both "generate" into a "generate" and a "create") and the documentation restructured and expanded. The raw data editing step is now visible in the meteoio workflow figure.

Now the dataCreators are defined in the [Input] section with a {parameter}::create = {algorithm} statement. They behave the same as the dataGenerators but are called *before* the filtering takes place and always work with the whole vector of data (ie whole dataset). In order to make life easier for the devs, a new structure will be devised...

The PotentialSW filter can now handle RSWR, most of the (useful) constants defined in the Shade and UnventillatedT filters are now in ProcessingBlock (for easy reuse), the Shade filter now has an easier structure (the Sun object can now do internal caching making the external caching such as performed in the filter not needed anymore).

A new filter has been implemented that compares the measured ISWR with the potential radiation. This would filter out stations that are shaded, stations that fill missing data with zeroes, etc

A warning has been fixed for Clang on FreeBSD by setting a compilation flag only for one specific file. A clear sky, potential short wave generator has been implemented as this is very useful when working with SW radiation. A bug in the AllSkySW generator has been fixed as it was not using TAU_CLD if available in one of the two possible calls. Its documentation was also not proper and has been improved.

When moving to gcc5, we discovered that its new ABI is incompatible with the current Oracle OCCI. A "legacy" API compatibility compilation flag solves the problem.

Since the IIR filter implementation can do both High and Low pass, it has been renamed to IIR and options have been added to use it as either low or high pass filter. The documentation has been improved and stability constraints now lead to a warning message if the filter would start to be unstable.

The Butterworth filter has been replaced by a critically damped filter implemented as IIRLowPass that properly corrects for the phase (ie behaves as zero-phase filter).

Well, the Butterworth filter is changing a lot... It is not Butterworth anymore but a critically damped filter. Although the options are currently hard-coded, it will offer the choice of bidirectional filtering with a technique to suppress the overshoot that are created in this case (at the cost of higher computing time).

a zero-phase version of the two poles Butterworth filter has been implemented. This is based on backward/forward filtering. The problem is that this doubles the order of the filter and this seems to be too much for our signals (tested on HS).

The virtual stations numbering was not kept in the virtual stations' names and IDs. This has been fixed. Thanks Sebastian W. for finding this bug!