[Tagging] Feature Proposal - RFC - temperature

[fly]

Am 04.02.2015 um 10:56 schrieb Kotya Karapetyan:
> Hi,
> 
> +1 for the proposal as such.
> 
> I have suggestions for some parts of the proposal though.
> 
> 1) I would discourage specification of the temperature without the
> scale indication. I have never lived in the US but I see from the Web
> that Americans like specifying temperature in degrees Fahrenheit
> without mentioning it (the same way as we in Europe use centigrade
> without underlying it). Taking into account the international nature
> of the OSM community, I foresee a significant risk that the map will
> get populated with invalid values. Warin is right about SI units, but
> SI is not even strictly followed in the technical and scientific
> community, not to mention the general public. Obviously, Americans in
> general ignore it by using inches, miles and degrees Fahrenheit :) I
> am afraid many people will not have heard about SI guidelines and will
> not have read the wiki page in significant detail.
> 
> Therefore, for the sake of clarity, I suggest always specifying "F" or
> "C" with the temperature value.

+1
Units for temperature are really wired and obviously Kelvin which I
would suspect to be the default is not really used in real live as
Celsius has the better scale for real life usage.

> 2) I suggest clarifying the verbal specification of the temperature.
> - Replace "chilled" with "cool" (by analogy with "warm") and also
> because "chilled" actually assumes that I know that the object was
> purportedly cooled down, which adds yet another uncertainty and is
> usually not very relevant;
> - remove the definition of "substantially colder" etc., because it
> doesn't add any clarity. I agree that it is important to distinguish
> between safe and unsafe situations, so let's just do that:
> 
> freezing
> cold — may be unsafe to handle
> cool
> warm
> hot — may be unsafe to handle
> boiling
> adjustable — the object temperature can be changed by consumer/user
> variable — the object temperature can vary on its own
> ambient — the object always remains at ambient temperature (note that
> this may include the object being "cold" and "warm", including being
> unsafe to handle, depending on the ambient temperature; think about
> water in Siberia rivers in January)

Only two values I could live with are cold and hot. Generally these
values are too ambiguous and an estimated value is much better.

> 3) For the numeric specification, I suggest adding:
> - "above"/"below" options
> - "approximate" value
> - range of temperatures (using above/below)
> 
> E.g.
> temperature:circa = 80 C
> temperature:above[:circa] = 300 C
> temperature:below[:circa] = 1000 C

I would add this in the value like:

temperature = < 10 C
temperature = > 300 C

We still can use source:temperature=estimated

4) How do we tag a shower with cold and hot water ?

temperature=4 C;80 C ?

Does this depend on the hose, e.g. one separate for each temperature or
a mix-batterie ?

cu fly