Sorry for long rant!
1. How do I get/set the list of base units (try "foot" and then ask
yourself "which foot is this?" [4])
2. How do I get/set derived units (Siemens is the inverse of Ohm, that is
S=Ω⁻¹ [3])
3. How do I add prefixed units (1kΩ, and 1mΩ, and note there is a bunch of
non-standard prefixes - not to forget localized ones! [5] I hate the mess
on Wikipedia... And note the mess with kilogram[7])
4. How to normalize a unit (it is (nearly) always µF, even when you write
4700µF [6] - this text is so messy and it does not really address the
problem)
5. Is there any plan to handle deprecated units (the weight prototype
gaining weight[1], and the new proposed standard [2] is one known problem
6. How to disambiguate units (the feet-problem in another version)
7. Is there any plan to add warnings about units that needs disambiguation
(the feet-problem is well-known, but how about kilogram? And note that is
the kilogram that is the standard unit, not the gram.)
8. How to handle incompatibilities between unit systems (you can't convert
some old units to newer ones.)
On 1, perhaps we could make Wd-entries for the different feets, but then
the lookup-list will be very long. Old classical units for length, area,
volume, and weight are the biggest problems. Some of them also coincide
with 8, as accurate conversions isn't possible.
On 2, some derived units can be transformed from one form into another.
Siemens is one of them. Others can be expressed in different ways, but all
variants is really just one and the same. We could use aliases for this, as
for example Farad (F) is s⁴A²⋅m⁻²kg⁻¹ and a bunch of other. An other
solution is to cluster descriptions, that makes a better solution for 1.
On 3, the simple solution is to add the SI-prefixes to everything. It
almost works, except we have units like kilogram (kg) which should retain
the "k". It will also create problems with kph and a bunch of other such
localized units.
On 4, don't confuse normalization of the unit with normalization of the
value. Normalization of the unit is highly domain specific.
On 5, note that there is a subtle difference here between an unit that goes
out of common use and a unit that is deprecated through law. Not sure if we
need to differentiate those, hope not!
On 6, I think foot is a good example on how long this list can be. Note
that in some countries different trade unions used different length of a
foot, and even some cities defined their own foot. I would like to define
my foot as the new standard unit.
On 7, note that the accuracy (error bounds) on the number should trigger a
need for disambiguation. Also note that precision imply a set level of
accuracy. Accuracy and precision is not the same, but precision can be used
as a proxy for accuracy.
On 8, there are several posts about this problem. Some claim you can avoid
the problem by setting the accuracy in the conversion sufficiently high. I
don't think that would be a valid solution. Perhaps we should have a
property for valid conversions, with constants for each one of them and
with proper error bounds. If a conversion isn't listed, then it isn't valid.
[1]
http://www.livescience.com/26017-kilogram-gained-weight.html
[2]
http://www.dailymail.co.uk/sciencetech/article-3161130/Reinventing-kilogram…
[3]
https://en.wikipedia.org/wiki/Siemens_%28unit%29
[4]
https://en.wikipedia.org/wiki/Foot_%28unit%29
[5]
https://en.wikipedia.org/wiki/Decametre
[6]
https://www.westfloridacomponents.com/blog/is-mf-mfd-the-same-as-uf/
[7]
http://www.bipm.org/en/bipm/mass/ipk/
On Sun, Dec 20, 2015 at 11:57 AM, Lydia Pintscher <
lydia.pintscher(a)wikimedia.de> wrote:
On Sun, Dec 20, 2015 at 10:08 AM, John Erling Blad
<jeblad(a)gmail.com>
wrote:
Can someone give an explanation why development
of units are so
difficult,
or what seems to be the problem? Is there
anything other people can do?
It seems to me like this has a serious feature creep...
https://phabricator.wikimedia.org/T77977
We have done the minimum version and deployed it. You're able to enter
and retrieve information with quantities and units. Now that the
minimum is in place other things got higher priority. That was/is
mainly data quality, properly linking to other sources in out export
formats and a UI cleanup including separating out identifiers. Those
are still in progress. Once we've brought those further along we'll
pick up the remaining work for units as well.
The main thing that is left now is unit conversion for the query service.
Cheers
Lydia
--
Lydia Pintscher -
http://about.me/lydia.pintscher
Product Manager for Wikidata
Wikimedia Deutschland e.V.
Tempelhofer Ufer 23-24
10963 Berlin
www.wikimedia.de
Wikimedia Deutschland - Gesellschaft zur Förderung Freien Wissens e. V.
Eingetragen im Vereinsregister des Amtsgerichts Berlin-Charlottenburg
unter der Nummer 23855 Nz. Als gemeinnützig anerkannt durch das
Finanzamt für Körperschaften I Berlin, Steuernummer 27/681/51985.
_______________________________________________
Wikidata mailing list
Wikidata(a)lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/wikidata