[Imports] YahooJapan/ALPSMAP imports

Satoshi IIDA nyampire at gmail.com
Sat Oct 6 07:24:35 BST 2012


Thank you for comments!

> it is best to stop the imports and fix the data first, if possible.
I confirmed to dataset creator, he could do it. (although needs a bit times..)
So I will talk it out with Japanese mappers.

> the ~ character
I think it could be converted into ~ or - , and removing "m" letter in
the value, when we reconstruct the dataset.
But as the opposite opinion, "yh: WIDTH" is a mere reference value.
Wiki documentation is enough, I think.
(Of cource, the explanation should be added on wiki docs. I will add it later)

> The first step should be to have the ongoing imports changed so that they don't include those tags.
> The second step is to add the tags to the list of tags that are automatically removed by JOSM. Potlatch can't automatically remove tags like this yet.
> A potential third step is to remove it from existing ways with a bot, but this shouldn't be considered until the first two steps are done.
Very clear roadmap!
If we could change the .osm dataset, it can be substitute for step 2.




2012/10/5 Paul Norman <penorman at mac.com>:
>> From: Satoshi IIDA [mailto:nyampire at gmail.com]
>> Sent: Tuesday, October 02, 2012 7:04 AM
>> To: Paul Norman
>> Cc: imports at openstreetmap.org
>> Subject: Re: [Imports] YahooJapan/ALPSMAP imports
>>
>> Japanese mappers got consent to delete almost all "yh:* " tags.
>> Only "yh:WIDTH" tag needs stay.
>>
>> http://wiki.openstreetmap.org/wiki/JA:YahooJapanALPS_Data/Highway/DataSt
>> ructure
>>
>> "yh:WIDTH" tag is not duplicate information.
>> Although a bit vague information, but it may help our survey.
>> (it will be deleted manually after survey)
>
>
> Looking more closely at the data, I see this now - although it does appear to duplicate yh:WIDTH_RANK.
>
> Can it be changed so that it doesn't use the ~ character as this is not a character in the standard ASCII character set? I would suggest either - or ~ as the character. I know that anyone using OSM data (particularly in Japan) needs to support UTF but there's nothing to be gained by ~ using instead of something else.
>
> The tag should also be documented on the wiki so that people know what it means.
>
>> Then, 2 subjects rise.
>> Any comments are appliciated!
>>
>> * How to delete?
>> I think BOT is the best way, but we don't have "know how" to run bots.
>>
>> * When do we start to modify?
>> Currently imports are still ongoing.
>> Most of them are imported already, but some are remains.
>>
>> Which is the better choice to modify?
>>   * after all the files are imported.
>>   * modify them several times (if any bots are run)
>
> The first step should be to have the ongoing imports changed so that they don't include those tags.
>
> The second step is to add the tags to the list of tags that are automatically removed by JOSM. Potlatch can't automatically remove tags like this yet.
>
> A potential third step is to remove it from existing ways with a bot, but this shouldn't be considered until the first two steps are done.
>



-- 
Satoshi IIDA
mail: nyampire at gmail.com
twitter: @nyampire



More information about the Imports mailing list