[Talk-GB] apostrophe's WAS "The last 2%"

[David Groom]

 ----- Original Message ----- 
From: "Shane Reynolds" <shane28r at gmail.com>
To: <talk-gb at openstreetmap.org>
Sent: Tuesday, August 17, 2010 2:09 PM
Subject: Re: [Talk-GB] The last 2%


> Hi,
>
> I am the developer who works on a number of products including OSM 
> Analysis
> for ITO.
>
> I am slightly confused about making the OS Locator box a dashed box if the
> not:name tag is present. When we do the processing if any street is found
> with the not:name tag matching an OS Locator street then these should not 
> be
> flagged in the stats nor should they be drawn on the map. Are you saying
> that this is not the case and they are still appearing? If this is true 
> then
> possibly there is a bug that I may want to address (we are usually a day
> behind the updated planet file so possibly its something to do with that?)
>
> With regard to getting the figures down (in slight fear of being flamed as 
> I
> am a very novice mapper) - with apostrophes, if the road sign has an
> apostrophe then I would say the apostrophe should be in OSM and if it is 
> not
> then it is correct to flag it as a difference. However if the road sign 
> does
> not have an apostrophe and OS Locator does then probably the road should
> have a not:name tag added with the apostrophe version of the name. This
> would remove any apostrophe issues.

Shane,
great work on your OSM analysis tools.

My recollection of the discussion on this list in early June was that 
"apostrophe" differences would not be tagged with not:name as this might 
have a tendency to flood the not:name reports with minor issues.

It was also suggested (by me I'll admit) that on the ITO tiles that 
differences due solely to apostrophes could be marked just in one pale 
colour, so they could quickly and easily be distinguished from other errors.

David


>With regard to rural roads - if they
> have no road signs is it not better to use the OS Locator name rather than
> have no name at all as I think in general OS Locator has been proved to be
> pretty accurate? (apologies if in my ignorance I do not know that there is 
> a
> good reason not to do this).
>
> Shane