/**
@page page-capabilities Capabilities

@tableofcontents

@section overview Overview

This section describes the capabilities and fundamentals of the
library.

@subsection data-records Data Records

@note miniSEED data are constructed as records, typically anywhere
from 128 to 4096 bytes in length.  Each record is independent,
identifying the data source, encoding type, etc. in header structures.
Reconstructing a continuous time series from an arbitrary number of
data records is an excersize for a format reader, e.g. this library.

A miniSEED record is represented in the library using the MS3Record
structure.  This structure is used for both unpacking and packing of
miniSEED records.  When unpacking with msr3_parse() this structure is
populated.  When packing with msr3_pack() this structure is used as a
template for the resulting data records and as a source of samples to
be packed.

@subsubsection Data Sample Encodings

The library supports all commonly used @ref encoding-values for
decoding and a subset of the most common values for encoding.  The
following data encoding formats are supported for both unpacking and
packing: ASCII, INT16, INT32, FLOAT32, FLOAT64, STEIM1 and STEIM2.
The INT and FLOAT encodings each have two versions for quantities with
a different number of bits of representation.  The STEIM compression
encodings work with 32-bit integers.  The following data encoding
formats are supported for unpacking only: GEOSCOPE (24-bit, 16/3 and
16/4 gain ranged), CDSN, SRO and DWWSSN.

@subsection traces Traces

In order to reconstruct, manage and process continuous trace segments
the library provides @ref trace-list functionality.

@ref trace-list structures are nested to contain an arbitrary number
of identified time series, each containing an arbitrary number
continuous time series segments, aka traces.  While a trace list is
normally used to hold trace information and associated data samples
they can also be used without data samples as a means of tracking data
coverage without actual sample values.

@subsection time-representation Time representation

Internally, time stamps are represented as nanoseconds since the
Unix/POSIX epoch via the nstime_t data type.  A number of functions
exist to parse or generate date-time strings to or from this
representation.  See @ref time-related.

*Note:* nanosecond time resolution does not mean the data included
 this resolution.

@subsection time-series-data Time series data formats

Time series data samples are supported in a few different formats
depending on how they are unpacked or will be packed.  Samples can be
either ASCII, 32-bit integer, 32-bit floats or 64-bit double precision
floats.  These types are represented internally using the
single-character @ref sample-types.

@subsection log-messages Log messages

All of the log and diagnostic messages emitted by the library
functions use a central logging facility.  It is common for programs
that use the library to also use these logging functions.
Furthermore, the output from these logging interfaces can be
controlled by directing output to callback functions.  This is useful
when the library will be embedded in a larger system with it's own
logging facility.  See the @ref logging page for more details.

*/