93f58b4538
* Release 2.0: Heavy rewrite with too many new features to enumerate here in the ChangeLog file. * NEWS: List of important changes since 1.x releases. * weather, weather.py: Implemented support for Python 3000 as requested by ptchinster on behalf of Arch Linux, conditions/forecast searches by latitude/longitude requested by Brandt Daniels, support for newer NOAA forecasts pointed out by Darryl Mouck and Richard Dooling, custom URIs requested by Michel Pelzer, international weather stations requested by Milton Hubsher, and fixed a metric conversion issue with negative values reported by Jochen Keil, Michiel Appelman and Stefan Metzlaff. Thanks to everyone for your input and assistance!
52 lines
2.4 KiB
Plaintext
52 lines
2.4 KiB
Plaintext
=================================================
|
|
New Version Information for the Weather Utility
|
|
=================================================
|
|
|
|
:Copyright: (c) 2006-2012 Jeremy Stanley <fungi@yuggoth.org>. Permission
|
|
to use, copy, modify, and distribute this software is
|
|
granted under terms provided in the LICENSE file distributed
|
|
with this software.
|
|
|
|
.. contents::
|
|
|
|
2.0 Release
|
|
-----------
|
|
The 2.0 release involves a major rewrite of the underlying code and
|
|
addition of large volumes of previously-unneeded location correlation
|
|
data. In the Spring of 2011 the USA NOAA/NWS made significant changes to
|
|
the way they organized and published forecast data such that it could
|
|
no longer be supported by design assumptions inherent within this
|
|
utility. Attempts were made to preserve backward compatability with 1.x
|
|
command-line usage and configuration file formats where possible, but
|
|
some regressions are unfortunately unavoidable. The aurl, city, flines,
|
|
furl, id, murl, st and zones options have been removed. Minimal logic
|
|
was retained to recognize and continue supporting limited use of city,
|
|
id and st in configuration files, though these are deprecated for
|
|
eventual removal in a future release.
|
|
|
|
On a positive note, this provided an opportunity to design out
|
|
some reported bugs and add in numerous requested features. Highlights
|
|
include:
|
|
|
|
* Because NOAA/NWS now treats forecast data in the same way as alerts,
|
|
the alert reporting features are now much better integrated and no
|
|
longer considered a beta test.
|
|
|
|
* The lack of memorable alert/forecast zone coding in the new scheme
|
|
drove development of intelligent search functionality, allowing users
|
|
to find stations and zones through a variety of methods like place
|
|
names, IATA/FAA/ICAO/FIPS/ZIP codes and even raw coordinates.
|
|
|
|
* To reduce unnecessary load on NOAA/NWS servers, the utility now
|
|
caches retrieved data for a configurable period of time.
|
|
|
|
* Airport code lists are no longer maintained in configuration (though
|
|
they are still easily overridden through configuration), and are
|
|
instead now separately managed in a manner similar to the other
|
|
pregenerated correlation data.
|
|
|
|
Worth noting, however, is that the new forecast data publication format
|
|
(now essentially identical to the alert format) is all-caps freeform
|
|
prose, and not easily parsed as a result. Due to the lengthy nature of
|
|
this output, piping it through a pager is highly recommended.
|