How would you then avoid "multivalues" ?

For example for brand=*

If a shop sells several brands (Honda;Yamaha;Suzuki;BMW;KTM),
how would you tag this ?

In the clothes example, we could certainly split the characteristics up in


But there might still be multiple values for each,
which clearly mean a listing to me if separated by semicolon.

Am 03.09.2017 um 14:32 schrieb Martin Koppenhoefer:
> sent from a phone
>> On 2. Sep 2017, at 10:58, Rafael Avila Coya <> wrote:
>> So if you have a shop that is "fashion" and women only, you could tag it as
>> shop=clothes
>> clothes=women;fashion
> this isn't semantically clear, it could just as well mean clothes for women 
> and fashion. What about clothes=bridal;fur, does this mean bridal clothes in 
> fur? ;-)
> There are many different kinds of things and properties in the clothes tag, 
> which would require multivalues and lead to ambiguities if we'd continue this 
> "system", because you don't know if it's "AND" or "OR".
>  it's unclear what property "clothes" is describing, e.g. "women", "men", 
> "babies", "kids" are orthogonal to "fur", "leather", "denim", are orthogonal 
> to "underwear", "hats" are orthogonal to "sports", "wedding", "workwear", are 
> orthogonal to "traditional", "fashion" - while others are a mix ("bridal" or 
> "lingerie" imply women and exclude men, kids, babies).
> It's also inconsistent to map shoe shops with their own top level tag, but 
> hats shops or wedding shops not. I surely wouldn't lump specialized shops 
> like a bridal shop together with generic clothing shops and distinguish them 
> only in a subtag, that's absurd, because nobody will want to find the bridal 
> shop in his search results when s/he wants to buy clothes - unless it's for 
> your own wedding, which typically is very infrequently, and for which you'd 
> likely will search specifically.
> cheers,
> Martin 
> _______________________________________________
> Tagging mailing list


Thilo Haug
72764 Reutlingen

Mobil: +49 177 3185856
Festnetz : +49 7121 3826414

Tagging mailing list

Reply via email to