Re: [OSM-talk] OSM is a right mess (was: Craigslist OpenStreetMap Rendering Issue)

[Floris Looijesteijn]

oneway=no is very clear to me and different to not having a oneway value at
all.

this way it means it was surveyed and it is not a oneway.

maybe with lit=yes or lit=no is more clear that is actually adds something
of value?

greets,
floris

On Wed, Jun 3, 2015 at 2:04 AM, pmailkeey . 
wrote:

> iD shows oneway=unknown if it's not set. If it's unknown, iD should not
> show oneway at all.
>
> In OSM if oneway=no then it's not oneway and the oneway tag should not
> appear at all.
>
> The only time oneway should appear is in the case of oneway=yes - and the
> '=yes' is superfluous.
>
> OSM's k=v design is completely a serious and unnecessary flaw. Similarly
> are 'categories' like man_made', and 'amenity'.
>
> Why can we not simply stick to hard facts rather guessing what
> categor(ies) an object fits in
>
>
> A fountain is a fountain. It does not matter if it is
>
>- an amenity
>- man made
>- natural water (???!!)
>
> etc. Such categorization is semi-ambiguous; people think differently and
> are happy to categorize differently - and more - argue over categories.
>
> OSM is 90% argument, 5% dead-end discussions and 5% progress. The whole is
> not a marketable product; it's not fit to be rated as 'beta'. Is this a
> significant cause of ex-mappers ? It's a flipping brilliant project but
> sadly lacking a great leader.
>
> --
> Mike.
> @millomweb  -
> For all your info on Millom and South Copeland
> via *the area's premier website - *
>
> *currently unavailable due to ongoing harassment of me, my family,
> property & pets*
>
> T&Cs 
>
> ___
> talk mailing list
> talk@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/talk
>
>
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] OSM is a right mess (was: Craigslist OpenStreetMap Rendering Issue)

[pmailkeey .]

On 4 June 2015 at 19:39, Lester Caine  wrote:

> On 04/06/15 16:04, Paweł Paprota wrote:
> > Could you please move this discussion to the tagging list?
> While some elements being picked up on are simple 'tagging' questions,
> it is the general structure we are discussing which in my book is the
> whole point of OSM. I think there is still room to discuss the overall
> framework of how the layers of tagging evolve?
>
> --
> Lester Caine - G8HFL
> -
> Contact - http://lsces.co.uk/wiki/?page=contact
> L.S.Caine Electronic Services - http://lsces.co.uk
> EnquirySolve - http://enquirysolve.com/
> Model Engineers Digital Workshop - http://medw.co.uk
> Rainbow Digital Media - http://rainbowdigitalmedia.co.uk
>
>
Surely all OSM sub-discussion have a 'right' to be heard in the general OSM
list and tagging is quite key to the whole project. It is perhaps a shame
that these issues haven't been bashed out years ago - well they might have
but if newcomers continue to question the logic there's clearly an issue
that needs addressing.

-- 
Mike.
@millomweb  -
For all your info on Millom and South Copeland
via *the area's premier website - *

*currently unavailable due to ongoing harassment of me, my family, property
& pets*

T&Cs 
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] OSM is a right mess (was: Craigslist OpenStreetMap Rendering Issue)

[Lester Caine]

On 04/06/15 16:04, Paweł Paprota wrote:
> Could you please move this discussion to the tagging list?
While some elements being picked up on are simple 'tagging' questions,
it is the general structure we are discussing which in my book is the
whole point of OSM. I think there is still room to discuss the overall
framework of how the layers of tagging evolve?

-- 
Lester Caine - G8HFL
-
Contact - http://lsces.co.uk/wiki/?page=contact
L.S.Caine Electronic Services - http://lsces.co.uk
EnquirySolve - http://enquirysolve.com/
Model Engineers Digital Workshop - http://medw.co.uk
Rainbow Digital Media - http://rainbowdigitalmedia.co.uk

___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] OSM is a right mess (was: Craigslist OpenStreetMap Rendering Issue)

[Paweł Paprota]

Could you please move this discussion to the tagging list?

On Thu, Jun 4, 2015, at 16:57, Martin Koppenhoefer wrote:
> 
> 
> 
> 
> > Am 04.06.2015 um 01:48 schrieb pmailkeey . :
> > 
> > 
> > A value of residential here  seems to need a key to identify whether it 
> > relates to a building or landuse.
> 
> 
> there is also highway=residential 
> 
> 
> 
> > However, you suggest building=residential as possibly being redundant. In 
> > fact, I'd turn this on its head and make landuse=residential (with the 
> > exception of moles) redundant.
> 
> 
> 
> building=residential is bad tagging IMHO as it adds only rough
> information not going beyond what landuse already tells, typically people
> will use more specific values like apartments, house / detached, villa
> etc (if they specify more than yes)
> I can't deny there are quite some of them nonetheless:
> 
> http://taginfo.openstreetmap.org/keys/building#values
> 
> 
> > The only residential landuse is directly under a building but by using 
> > landuse=residential, such areas cover gardens and highways - which are 
> > clearly not residences.
> 
> 
> I agree for highways but gardens, terraces, garages etc are normally part
> of the residence, even if you don't sleep there...
> 
> 
> Cheers 
> Martin 
> ___
> talk mailing list
> talk@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/talk

___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] OSM is a right mess

[pmailkeey .]

On 3 June 2015 at 07:00, Maarten Deen  wrote:

> On 2015-06-03 02:04, pmailkeey . wrote:
>
>> iD shows oneway=unknown if it's not set. If it's unknown, iD should
>> not show oneway at all.
>>
>
> I agree.
>
>  In OSM if oneway=no then it's not oneway and the oneway tag should not
>> appear at all.
>>
>
> Here I don't agree.
>
>  The only time oneway should appear is in the case of oneway=yes - and
>> the '=yes' is superfluous.
>>
>
> Some roads are implied oneway. E.g. junction=roundabout and
> highway=motorway both imply that the road is one-way only. If for some
> reason the object in case is not oneway, a oneway=no tag is very much
> needed.
>
> I agree that in every case where oneway=yes is not implied, oneway=no is
> superfluous (in a network design way), but that does not make oneway=no
> superfluous.
>
> There is also the occurence of oneway=-1 in case someone reverses the
> direction of a way. What should be done when the only possibility for
> oneway is either set or unset and the direction gets reversed? Should
> reversing be disallowed? Should you get a warning "oneway street can not be
> reversed"?
>
> Maarten
>
>
Are the world of random renderers going to look for junction=roundabout and
make the same oneway assumption ? Would it not be better for
'junction=roundabout' to cause a mechanical edit by adding the oneway tag -
so that rather than saying =no, the tag could simply be removed ?

What reason is there for reversing the way - as presumably all
direction-dependent tags have + / - options ? Leads to the question as to
why make oneway an exception to this rule - it seems most logical to have
oneway as the direction as indicated rather than against.

Them's my thoughts !

-- 
Mike.
@millomweb  -
For all your info on Millom and South Copeland
via *the area's premier website - *

*currently unavailable due to ongoing harassment of me, my family, property
& pets*

T&Cs 
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] OSM is a right mess

[Martin Koppenhoefer]

> Am 04.06.2015 um 02:11 schrieb Warin <61sundow...@gmail.com>:
> 
> No.
> Landuse - a facility (what it gets used for)
> The governmental authority here says an area is for residential use ..
> They do that before there are any houses thus it is not just the area under a 
> building.


No
;-)
You are referring to zoning, the planning of landuse, after approval becomes 
the prescription of landuse.
In osm, the tag "landuse" describes the actual landuse, regardless of what 
should or could be at that place according to the law or planning.

cheers 
Martin 
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] OSM is a right mess (was: Craigslist OpenStreetMap Rendering Issue)

[Martin Koppenhoefer]

> Am 04.06.2015 um 01:48 schrieb pmailkeey . :
> 
> 
> A value of residential here  seems to need a key to identify whether it 
> relates to a building or landuse.


there is also highway=residential 



> However, you suggest building=residential as possibly being redundant. In 
> fact, I'd turn this on its head and make landuse=residential (with the 
> exception of moles) redundant.



building=residential is bad tagging IMHO as it adds only rough information not 
going beyond what landuse already tells, typically people will use more 
specific values like apartments, house / detached, villa etc (if they specify 
more than yes)
I can't deny there are quite some of them nonetheless:

http://taginfo.openstreetmap.org/keys/building#values


> The only residential landuse is directly under a building but by using 
> landuse=residential, such areas cover gardens and highways - which are 
> clearly not residences.


I agree for highways but gardens, terraces, garages etc are normally part of 
the residence, even if you don't sleep there...


Cheers 
Martin 
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] OSM is a right mess

[John Eldredge]

In urban areas, it is common to have mixed-use buildings as well, with 
retail or services on the ground floor and residential units on upper floors.


--
John F. Eldredge -- j...@jfeldredge.com
"Darkness cannot drive out darkness; only light can do that. Hate cannot 
drive out hate; only love can do that." -- Martin Luther King, Jr.




On June 3, 2015 7:13:39 PM Warin <61sundow...@gmail.com> wrote:


On 4/06/2015 9:48 AM, pmailkeey . wrote:
>
>
>
> A value of residential here  seems to need a key to identify whether
> it relates to a building or landuse. However, you suggest
> building=residential as possibly being redundant. In fact, I'd turn
> this on its head and make landuse=residential (with the exception of
> moles) redundant. The only residential landuse is directly under a
> building but by using landuse=residential, such areas cover gardens
> and highways - which are clearly not residences.
>

No.
Landuse - a facility (what it gets used for)
The governmental authority here says an area is for residential use ..
They do that before there are any houses thus it is not just the area
under a building.

building - a physical item (what is there on the ground)
A building that is used for residences has a particular set of features
that distinguish it for other building types e.g. a mall.

--
I use my garden! I go out there and read, email .. even do OSM ... in
the garden.
The street in front of my home gets used by the kids to play soccer,
tennis ...

The land use is 'residential' .. even in the garden. And the neighbours
swimming pool.
The house is also residential.

I and my neighbours reside here - we use all of it.

___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk




___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] OSM is a right mess

[Maarten Deen]

On 2015-06-04 01:48, pmailkeey . wrote:

On 3 June 2015 at 09:45, Lester Caine  wrote:


On 03/06/15 01:04, pmailkeey . wrote:

OSM's k=v design is completely a serious and unnecessary flaw.

Similarly

are 'categories' like man_made', and 'amenity'.
Why can we not simply stick to hard facts rather guessing what
categor(ies) an object fits in


This is a bit like saying XML is the wrong base format. and actually
I
would agree with that, but the majority of material only works with
a
k=v structure. While for a few VALUES there are potentially only one
k=v, they are very few and far between.


I'm not convinced. A value of yes as a stand-alone item is meaningless
but a value of hedge is sufficient to indicate we're talking about a
barrier. (please read below before responding to this item)


But equally important is a consistent way of offering your data. And 
offering not necesserily to end-users but most importantly to 
application builders.
Nothing is more annoying than seeing multiple types of data being 
offered because for each type you have to program something differently.


Sure, "residential" on an object may be sufficient to see that it is a 
road and it is classified as a residential road. But now you want to get 
all roads. So you have to select residential, tertiary, secondary... 
etc. How much more easy is it then to select all ways with key 
"highway".


I agree that some values may be self explanatory, but I'm sure it does 
not weigh up to the added burden of having to work with them.


Maarten


___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] OSM is a right mess

[Warin]

On 4/06/2015 9:48 AM, pmailkeey . wrote:




A value of residential here  seems to need a key to identify whether 
it relates to a building or landuse. However, you suggest 
building=residential as possibly being redundant. In fact, I'd turn 
this on its head and make landuse=residential (with the exception of 
moles) redundant. The only residential landuse is directly under a 
building but by using landuse=residential, such areas cover gardens 
and highways - which are clearly not residences.




No.
Landuse - a facility (what it gets used for)
The governmental authority here says an area is for residential use ..
They do that before there are any houses thus it is not just the area 
under a building.


building - a physical item (what is there on the ground)
A building that is used for residences has a particular set of features 
that distinguish it for other building types e.g. a mall.


--
I use my garden! I go out there and read, email .. even do OSM ... in  
the garden.
The street in front of my home gets used by the kids to play soccer, 
tennis ...


The land use is 'residential' .. even in the garden. And the neighbours 
swimming pool.

The house is also residential.

I and my neighbours reside here - we use all of it.

___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] OSM is a right mess (was: Craigslist OpenStreetMap Rendering Issue)

[Lester Caine]

On 04/06/15 00:48, pmailkeey . wrote:
> A value of residential here  seems to need a key to identify whether it
> relates to a building or landuse. However, you suggest
> building=residential as possibly being redundant. In fact, I'd turn this
> on its head and make landuse=residential (with the exception of moles)
> redundant. The only residential landuse is directly under a building but
> by using landuse=residential, such areas cover gardens and highways -
> which are clearly not residences.

building=residential is only redundant if the area it is in is
'residential'. The whole point here is that 'zoning/planning' identifies
housing estates, the related retail and amenity areas, and recreation
areas within that residential estate. Micro mapping would provide
individual buildings either detached, in pairs, in terraces, and rather
than every building having 'residential', 'detached/semi/terraced', the
fact they are on a residential estate provides that data? When zooming
out, once the buildings become too small to display, the residential
area still identifies the estate, and any associated retail areas. If
park areas are big enough they will appear in the residential area.

-- 
Lester Caine - G8HFL
-
Contact - http://lsces.co.uk/wiki/?page=contact
L.S.Caine Electronic Services - http://lsces.co.uk
EnquirySolve - http://enquirysolve.com/
Model Engineers Digital Workshop - http://medw.co.uk
Rainbow Digital Media - http://rainbowdigitalmedia.co.uk

___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] OSM is a right mess (was: Craigslist OpenStreetMap Rendering Issue)

[pmailkeey .]

On 3 June 2015 at 09:45, Lester Caine  wrote:

> On 03/06/15 01:04, pmailkeey . wrote:
> > OSM's k=v design is completely a serious and unnecessary flaw. Similarly
> > are 'categories' like man_made', and 'amenity'.
> > Why can we not simply stick to hard facts rather guessing what
> > categor(ies) an object fits in
>
> This is a bit like saying XML is the wrong base format. and actually I
> would agree with that, but the majority of material only works with a
> k=v structure. While for a few VALUES there are potentially only one
> k=v, they are very few and far between.
>

I'm not convinced. A value of yes as a stand-alone item is meaningless but
a value of hedge is sufficient to indicate we're talking about a barrier.
(please read below before responding to this item)


>
> That there are a few keys that are confusing even when one is used to
> how they work is the real problem.
> amenity===shop===leisure===sport===tourism===building===landuse
> Just what combination is right for a single retail building with a
> leisure facility on the ground floor and accommodation above?
>
> building=detached is another peculiarity where the closed way identifies
> that fact and what is actually needed is simply building=residential
> unless it's contained in an area of landuse=residential ... is the
> building=residential now redundant?
>

A value of residential here  seems to need a key to identify whether it
relates to a building or landuse. However, you suggest building=residential
as possibly being redundant. In fact, I'd turn this on its head and make
landuse=residential (with the exception of moles) redundant. The only
residential landuse is directly under a building but by using
landuse=residential, such areas cover gardens and highways - which are
clearly not residences.


>
> The discussion recently on how a 'university' should be tagged probably
> encompasses all of the circularity with the current tagging practice,
> and where the hierarchy of tagging could best be agreed.
>
> Campus area === landuse=university
> Some universities are on several campuses, but there there are other
> landuse areas between, AND there should be no other landuse areas
> contained within the campus area, other tagging should identify water,
> woodland, grass, building, and the like WITHIN the area. Same for a
> shopping quarter, residential area and so on. (leisure=park rather than
> landuse=park)
>
> A retail facility is a building=retail and it's 'amenity'=convenience
> although a number of uses can be listed. So rather than
> amenity,shop,leisure,sport,tourism we have use= attached to building= or
> area= and we use landuse= to wrap the major function of the whole ground
> area ... hospital, theme_park, school, and so on ... just need an
> equivalent to building= for area= where there is no actual building so
> area=park between the building=apartments of a residential area.
>
> --
> Lester Caine - G8HFL
> -
> Contact - http://lsces.co.uk/wiki/?page=contact
> L.S.Caine Electronic Services - http://lsces.co.uk
> EnquirySolve - http://enquirysolve.com/
> Model Engineers Digital Workshop - http://medw.co.uk
> Rainbow Digital Media - http://rainbowdigitalmedia.co.uk
>
> ___
> talk mailing list
> talk@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/talk
>



-- 
Mike.
@millomweb  -
For all your info on Millom and South Copeland
via *the area's premier website - *

*currently unavailable due to ongoing harassment of me, my family, property
& pets*

T&Cs 
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] OSM is a right mess (was: Craigslist OpenStreetMap Rendering Issue)

[Lester Caine]

On 03/06/15 12:57, Janko Mihelić wrote:
> Yes, we need semantic meaning and structure, but I still don't
> understand how do two strings help you with that. Key and value can only
> help someone who is digging through a xml file to quickly help
> themselves. But the true meaning of a k=v combination is in our wiki.
> The structure is also in our wiki, and hopefully in a future wikibase.
> If we are depending on a combination of two strings to give our data
> structure and meaning, we are in a lot of trouble.

I don't see how you can do anything else. Some keys have many values and
those values have a different meaning when attached to a different key.
The key is an element for grouping sub-sets of values or for identifying
just what a number or date applies to, and elements like 'name', 'ref',
'id' identify free format data and how it fits into the data model. *I*
don't see how you can do away with a pair of strings as the base
structure. Yes a few keys can be used without a value, but the major
number of items require a value of some sort to go with the key.

PERSONALLY I would prefer that keys that have a well defined list of
values were stored as a simple number ... then the table used to display
them as text can simply be a version in the correct language for the
user! But far too many of these keys also have values which are not in
the 'approved' list. Enforcing the approved set may have some
advantages, but it is still a list of k/v tags. And on can add ANY
key/value data to that list.

-- 
Lester Caine - G8HFL
-
Contact - http://lsces.co.uk/wiki/?page=contact
L.S.Caine Electronic Services - http://lsces.co.uk
EnquirySolve - http://enquirysolve.com/
Model Engineers Digital Workshop - http://medw.co.uk
Rainbow Digital Media - http://rainbowdigitalmedia.co.uk

___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] OSM is a right mess (was: Craigslist OpenStreetMap Rendering Issue)

[Lester Caine]

On 03/06/15 12:02, Janko Mihelić wrote:
> I don't get the problem with k=v instead of just v. There is a function
> called concatenate which solves that.
> 
> We should move away from mapping directly with tags, and to a system
> like iDs, with descriptions that hide tags. For a database it's 
> irrelevant if it's "oneway=yes" or "4658". All this talk about correct
> k=v combinations is bikeshedding to me. Only DB admins and tag proposers
> should ever see what a tag looks like.

Sorry, but that does not make any sense at all. The starting point here
is that OSM is ONLY a database of information, and the schema of how
that data is made up is what we are discussing. The key values are
fundamental to the schema, and while different database models can make
things easier from one view or another, the current structure is the
only way to make things flexible across all users. That some users can
hide elements of the schema is only a matter of the user application,
the raw data has to be in a format that can be translated as required.
In the database we need to know if an object has a property 'oneway' and
what values are allowed for that property. YES "4658" can be a valid
property, or "4658" can be pulled out of some other database as
"oneway=yes", but we still need to know what the key oneway means and
how it fits into the rest of the data structure. That a road is defined
as 'highway' and has keys for oneway, speed, and other free format keys
is fundamental to building the data model.

-- 
Lester Caine - G8HFL
-
Contact - http://lsces.co.uk/wiki/?page=contact
L.S.Caine Electronic Services - http://lsces.co.uk
EnquirySolve - http://enquirysolve.com/
Model Engineers Digital Workshop - http://medw.co.uk
Rainbow Digital Media - http://rainbowdigitalmedia.co.uk

___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] OSM is a right mess (was: Craigslist OpenStreetMap Rendering Issue)

[Janko Mihelić]

I don't get the problem with k=v instead of just v. There is a function
called concatenate which solves that.

We should move away from mapping directly with tags, and to a system like
iDs, with descriptions that hide tags. For a database it's  irrelevant if
it's "oneway=yes" or "4658". All this talk about correct k=v combinations
is bikeshedding to me. Only DB admins and tag proposers should ever see
what a tag looks like.

Janko

sri, 3. lip 2015. 12:22 Lester Caine  je napisao:

> On 03/06/15 10:30, Martin Koppenhoefer wrote:
> >> Just what combination is right for a single retail building with a
> >> > leisure facility on the ground floor and accommodation above?
> >
> > I don't think a combination of tags is the best answer to this question,
> it's rather a combination of objects (3, one for the building, one for the
> ground floor facility, one for the accommodation facility)
>
> I think my point here was that there is a single building, but a number
> of uses within that outline. We don't have a convenient method of
> handling different objects on different floors within the one set of
> 'ways'. The areas I'm trying to tidy up are small shopping malls which
> have shops within shops on multiple levels ... with accommodation above.
> A combination of objects is just what I'm looking for, but with
> different outlines for each object ...
>
> --
> Lester Caine - G8HFL
> -
> Contact - http://lsces.co.uk/wiki/?page=contact
> L.S.Caine Electronic Services - http://lsces.co.uk
> EnquirySolve - http://enquirysolve.com/
> Model Engineers Digital Workshop - http://medw.co.uk
> Rainbow Digital Media - http://rainbowdigitalmedia.co.uk
>
> ___
> talk mailing list
> talk@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/talk
>
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] OSM is a right mess (was: Craigslist OpenStreetMap Rendering Issue)

[Martin Koppenhoefer]

> Am 03.06.2015 um 12:21 schrieb Lester Caine :
> 
> The areas I'm trying to tidy up are small shopping malls which
> have shops within shops on multiple levels ... with accommodation above.
> A combination of objects is just what I'm looking for, but with
> different outlines for each object ...


this is much less a problem of our data model than it is to edit with current 
2D osm editors, and to survey with common methods. Most people will only put 
nodes for the shops inside the mall polygon and often they will not be at their 
actual position but just inside- what is sufficient for many usecases but not 
satisfactory if you'd want to render floor plans.


Cheers 
Martin 
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] OSM is a right mess

[Martin Koppenhoefer]

> Am 03.06.2015 um 11:36 schrieb Warin <61sundow...@gmail.com>:
> 
> landcover rather than landuse? (for woodland, grass, forest etc)


there is a landcover tag proposal, grass is a proposed value, but forest and 
grassland are deliberately not, please use "trees" (regardless of any 
forest/wood/other groups of trees-distinction). "woodland" could be a 
key:natural object (name etc), for landcover it would be split into trees and 
grass, bushes etc. (if you went to that level of detail).

shouldn't this be discussed on tagging?


cheers 
Martin 
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] OSM is a right mess

[Maarten Deen]

On 2015-06-03 12:08, Shaun McDonald wrote:

On 3 Jun 2015, at 07:00, Maarten Deen  wrote:


I agree that in every case where oneway=yes is not implied, oneway=no 
is superfluous (in a network design way), but that does not make 
oneway=no superfluous.




There are some cases where oneway=no is useful. For example an area
where there is lots of one way streets and only a few that are two
way, adding the oneway=no confirms that the data is correct rather
than the oneway=yes being missing. Similarly where a street was oneway
previously and has recently been made two way, this makes it explicit
that it is now two way in addition to whatever changeset note there
may be.


Yes, that's why I said "in a network design way". Looking at the data, 
oneway=no is not necessary on any object where it is not implied. 
However, adding it does make clear to people editing the map that it is 
not an omission if surrounding objects are all oneway=yes.


Regards,
Maarten

___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] OSM is a right mess

[Lester Caine]

On 03/06/15 11:08, Shaun McDonald wrote:
>> I agree that in every case where oneway=yes is not implied, oneway=no is 
>> superfluous (in a network design way), but that does not make oneway=no 
>> superfluous.
>> > 
> There are some cases where oneway=no is useful. For example an area where 
> there is lots of one way streets and only a few that are two way, adding the 
> oneway=no confirms that the data is correct rather than the oneway=yes being 
> missing. Similarly where a street was oneway previously and has recently been 
> made two way, this makes it explicit that it is now two way in addition to 
> whatever changeset note there may be.

The one element of this which is not easy to tag currently is where a
section of road is one direction in the morning and another in the
afternoon? That the routing software can use yes and -1 depending on the
time of day is practical and another example of where a simple single
oneway value is limiting. In this case 'reverse' is more accurate, but
as long as values are well documented the exact value is somewhat
irrelevant. The couple of examples I know off are a little more complex
in that the signage is under manual control rather than via a clock :)

-- 
Lester Caine - G8HFL
-
Contact - http://lsces.co.uk/wiki/?page=contact
L.S.Caine Electronic Services - http://lsces.co.uk
EnquirySolve - http://enquirysolve.com/
Model Engineers Digital Workshop - http://medw.co.uk
Rainbow Digital Media - http://rainbowdigitalmedia.co.uk

___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] OSM is a right mess (was: Craigslist OpenStreetMap Rendering Issue)

[Lester Caine]

On 03/06/15 10:30, Martin Koppenhoefer wrote:
>> Just what combination is right for a single retail building with a
>> > leisure facility on the ground floor and accommodation above?
> 
> I don't think a combination of tags is the best answer to this question, it's 
> rather a combination of objects (3, one for the building, one for the ground 
> floor facility, one for the accommodation facility)

I think my point here was that there is a single building, but a number
of uses within that outline. We don't have a convenient method of
handling different objects on different floors within the one set of
'ways'. The areas I'm trying to tidy up are small shopping malls which
have shops within shops on multiple levels ... with accommodation above.
A combination of objects is just what I'm looking for, but with
different outlines for each object ...

-- 
Lester Caine - G8HFL
-
Contact - http://lsces.co.uk/wiki/?page=contact
L.S.Caine Electronic Services - http://lsces.co.uk
EnquirySolve - http://enquirysolve.com/
Model Engineers Digital Workshop - http://medw.co.uk
Rainbow Digital Media - http://rainbowdigitalmedia.co.uk

___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] OSM is a right mess

[Shaun McDonald]

> On 3 Jun 2015, at 07:00, Maarten Deen  wrote:
> 
> On 2015-06-03 02:04, pmailkeey . wrote:
>> iD shows oneway=unknown if it's not set. If it's unknown, iD should
>> not show oneway at all.
> 
> I agree.
> 
>> In OSM if oneway=no then it's not oneway and the oneway tag should not
>> appear at all.
> 
> Here I don't agree.
> 
>> The only time oneway should appear is in the case of oneway=yes - and
>> the '=yes' is superfluous.
> 
> Some roads are implied oneway. E.g. junction=roundabout and highway=motorway 
> both imply that the road is one-way only. If for some reason the object in 
> case is not oneway, a oneway=no tag is very much needed.
> 
> I agree that in every case where oneway=yes is not implied, oneway=no is 
> superfluous (in a network design way), but that does not make oneway=no 
> superfluous.
> 

There are some cases where oneway=no is useful. For example an area where there 
is lots of one way streets and only a few that are two way, adding the 
oneway=no confirms that the data is correct rather than the oneway=yes being 
missing. Similarly where a street was oneway previously and has recently been 
made two way, this makes it explicit that it is now two way in addition to 
whatever changeset note there may be.

Shaun


___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] OSM is a right mess

[Warin]

On 3/06/2015 6:45 PM, Lester Caine wrote:


The discussion recently on how a 'university' should be tagged probably
encompasses all of the circularity with the current tagging practice,
and where the hierarchy of tagging could best be agreed.

Campus area === landuse=university
Some universities are on several campuses, but there there are other
landuse areas between, AND there should be no other landuse areas
contained within the campus area, other tagging should identify water,
woodland, grass,


landcover rather than landuse? (for woodland, grass, forest etc)

  .. and around we go...



___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] OSM is a right mess (was: Craigslist OpenStreetMap Rendering Issue)

[Martin Koppenhoefer]

> Am 03.06.2015 um 10:45 schrieb Lester Caine :
> 
> Just what combination is right for a single retail building with a
> leisure facility on the ground floor and accommodation above?


I don't think a combination of tags is the best answer to this question, it's 
rather a combination of objects (3, one for the building, one for the ground 
floor facility, one for the accommodation facility)


cheers 
Martin 
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] OSM is a right mess (was: Craigslist OpenStreetMap Rendering Issue)

[Lester Caine]

On 03/06/15 01:04, pmailkeey . wrote:
> OSM's k=v design is completely a serious and unnecessary flaw. Similarly
> are 'categories' like man_made', and 'amenity'.
> Why can we not simply stick to hard facts rather guessing what
> categor(ies) an object fits in

This is a bit like saying XML is the wrong base format. and actually I
would agree with that, but the majority of material only works with a
k=v structure. While for a few VALUES there are potentially only one
k=v, they are very few and far between.

That there are a few keys that are confusing even when one is used to
how they work is the real problem.
amenity===shop===leisure===sport===tourism===building===landuse
Just what combination is right for a single retail building with a
leisure facility on the ground floor and accommodation above?

building=detached is another peculiarity where the closed way identifies
that fact and what is actually needed is simply building=residential
unless it's contained in an area of landuse=residential ... is the
building=residential now redundant?

The discussion recently on how a 'university' should be tagged probably
encompasses all of the circularity with the current tagging practice,
and where the hierarchy of tagging could best be agreed.

Campus area === landuse=university
Some universities are on several campuses, but there there are other
landuse areas between, AND there should be no other landuse areas
contained within the campus area, other tagging should identify water,
woodland, grass, building, and the like WITHIN the area. Same for a
shopping quarter, residential area and so on. (leisure=park rather than
landuse=park)

A retail facility is a building=retail and it's 'amenity'=convenience
although a number of uses can be listed. So rather than
amenity,shop,leisure,sport,tourism we have use= attached to building= or
area= and we use landuse= to wrap the major function of the whole ground
area ... hospital, theme_park, school, and so on ... just need an
equivalent to building= for area= where there is no actual building so
area=park between the building=apartments of a residential area.

-- 
Lester Caine - G8HFL
-
Contact - http://lsces.co.uk/wiki/?page=contact
L.S.Caine Electronic Services - http://lsces.co.uk
EnquirySolve - http://enquirysolve.com/
Model Engineers Digital Workshop - http://medw.co.uk
Rainbow Digital Media - http://rainbowdigitalmedia.co.uk

___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] OSM is a right mess

[Maarten Deen]

On 2015-06-03 02:04, pmailkeey . wrote:

iD shows oneway=unknown if it's not set. If it's unknown, iD should
not show oneway at all.


I agree.


In OSM if oneway=no then it's not oneway and the oneway tag should not
appear at all.


Here I don't agree.


The only time oneway should appear is in the case of oneway=yes - and
the '=yes' is superfluous.


Some roads are implied oneway. E.g. junction=roundabout and 
highway=motorway both imply that the road is one-way only. If for some 
reason the object in case is not oneway, a oneway=no tag is very much 
needed.


I agree that in every case where oneway=yes is not implied, oneway=no is 
superfluous (in a network design way), but that does not make oneway=no 
superfluous.


There is also the occurence of oneway=-1 in case someone reverses the 
direction of a way. What should be done when the only possibility for 
oneway is either set or unset and the direction gets reversed? Should 
reversing be disallowed? Should you get a warning "oneway street can not 
be reversed"?


Maarten


___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] OSM is a right mess (was: Craigslist OpenStreetMap Rendering Issue)

[pmailkeey .]

On 3 June 2015 at 01:37, Tom MacWright  wrote:

> Hi Mike,
>
> Please propose an alternative.
>

I see an awful lot of good in OSM and I think it's a great project. I've
had it agreed with another about it being such a mess - but the fact it's
such a worthwhile project it's worth battling on with it - even if, sadly,
others have succumbed to the nightmares around every corner. An alternative
would be wrong, it is this that needs fixing and it needs a lot of 'tricky'
effort to see the good from the not so good. It's like one team at both
ends of a thick 'tug of war' rope, not really sure which way to pull for
the best whilst each has in their hand only a mere strand to that rope. I
think everything is overwhelmed yet the whole thing is clearly in its
infancy - with diseases, viruses and god-knows what being thrown at it at
all times from all angles.

The freedom to make up any tags is brilliant while likely being one of the
biggest problems - that not simply throws simple problems but really
complex multi-dimensional ones with 'language' issues for one thing, the
fact OSM is secretly 2 maps (at least!!) in one where in the main the two
are compatible and other places where clearly they're not. Computers
frequently come up with an answer. OSM rarely does; likely many answers or
none at all. Digital, analogue, fuzzy logic all put into a blender and
whizzed for a few seconds - and each time this is tried a different result
is found. Consistently inconsistent.

I think in the end it will work - but the effort required to get there will
be far greater than the sum of its parts.

Take one aspect - quality - there's the whole gamut from true to false -
anyone can't assume anything about OSM data - and if different people
around the world wrote a report on it, none of the resulting reports would
agree !

I can't produce some magic answer without some agreement :)

-- 
Mike.
@millomweb  -
For all your info on Millom and South Copeland
via *the area's premier website - *

*currently unavailable due to ongoing harassment of me, my family, property
& pets*

T&Cs 
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] OSM is a right mess (was: Craigslist OpenStreetMap Rendering Issue)

[Warin]

On 3/06/2015 10:37 AM, Tom MacWright wrote:

Hi Mike,

Please propose an alternative.


He has been .. on other threads..


On Tue, Jun 2, 2015 at 8:22 PM, Eugene Alvin Villar > wrote:


On Jun 3, 2015 8:06 AM, "pmailkeey ." mailto:pmailk...@googlemail.com>> wrote:
> OSM's k=v design is completely a serious and unnecessary flaw.
[...] OSM is 90% argument, 5% dead-end discussions and 5%
progress. The whole is not a marketable product; it's not fit to
be rated as 'beta'. Is this a significant cause of ex-mappers ?
It's a flipping brilliant project but sadly lacking a great leader.

It seems you are deeply unsatisfied with how OSM works. And your
broad assertions such as that OSM is "not fit" or is "90%
argument" are completely unfounded. Sure, OSM is not perfect but I
seriously doubt that the k=v design or some other point you have
raised is the culprit.

Feel free to leave and create a separate project. You can even be
the "great leader" for that new project that you think OSM needs.
If your ideas are indeed better, then your project will succeed
and you can then prove OSM wrong.


___
talk mailing list
talk@openstreetmap.org 
https://lists.openstreetmap.org/listinfo/talk




___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk


___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] OSM is a right mess (was: Craigslist OpenStreetMap Rendering Issue)

[Steve Coast]

Fair point, I meant in the context of the list, as I thought others did too.

Steve




On Tue, Jun 2, 2015 at 6:07 PM -0700, "Eugene Alvin Villar"  
wrote:










On 6/3/15, Steve Coast  wrote:
>
>> On Jun 2, 2015, at 6:22 PM, Eugene Alvin Villar  wrote:
>> On Jun 3, 2015 8:06 AM, "pmailkeey ." > > wrote:
>> > OSM's k=v design is completely a serious and unnecessary flaw. [...] OSM
>> > is 90% argument, 5% dead-end discussions and 5% progress. The whole is
>> > not a marketable product; it's not fit to be rated as 'beta'. Is this a
>> > significant cause of ex-mappers ? It's a flipping brilliant project but
>> > sadly lacking a great leader.
>>
>> It seems you are deeply unsatisfied with how OSM works. And your broad
>> assertions such as that OSM is "not fit" or is "90% argument" are
>> completely unfounded.
>>
>
> I don’t know; there are a bunch of fairly key and active OSM people who
> unsubscribed from the lists precisely because they felt it was mostly
> circular argument.

Yes, people leave mailing lists because of the endless arguments and
constant bike-shedding. But that does not constitute 90% of OSM. I am
willing to bet that majority if not 90% of OSM activity is of mappers
actually mapping. Mailing list discussions is a really small slice of
the overall OSM activity.___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] OSM is a right mess (was: Craigslist OpenStreetMap Rendering Issue)

[Eugene Alvin Villar]

On 6/3/15, Steve Coast  wrote:
>
>> On Jun 2, 2015, at 6:22 PM, Eugene Alvin Villar  wrote:
>> On Jun 3, 2015 8:06 AM, "pmailkeey ." > > wrote:
>> > OSM's k=v design is completely a serious and unnecessary flaw. [...] OSM
>> > is 90% argument, 5% dead-end discussions and 5% progress. The whole is
>> > not a marketable product; it's not fit to be rated as 'beta'. Is this a
>> > significant cause of ex-mappers ? It's a flipping brilliant project but
>> > sadly lacking a great leader.
>>
>> It seems you are deeply unsatisfied with how OSM works. And your broad
>> assertions such as that OSM is "not fit" or is "90% argument" are
>> completely unfounded.
>>
>
> I don’t know; there are a bunch of fairly key and active OSM people who
> unsubscribed from the lists precisely because they felt it was mostly
> circular argument.

Yes, people leave mailing lists because of the endless arguments and
constant bike-shedding. But that does not constitute 90% of OSM. I am
willing to bet that majority if not 90% of OSM activity is of mappers
actually mapping. Mailing list discussions is a really small slice of
the overall OSM activity.

___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] OSM is a right mess (was: Craigslist OpenStreetMap Rendering Issue)

[pmailkeey .]

On 3 June 2015 at 01:36, Bryce Nesbitt  wrote:

> Perhaps http://wikimapia.org/ will better match your needs, and offer
> more peace for your family, property and pets.
>

Well, on Monday the 'prison' idea went out the window. Having me and my
brother accused of ill-treating our mother - two , nay, 3 visits to court
in preparation for a trial - but now the prosecution say they've no
evidence.



It's a ton of worry off my shoulders



Just another 19 similar tons to go.

God, it's just so bloody awful what they do to 'vulnerable adults' (Mum has
dementia) in this country.


It was the National Health Service that caused severe distress to my cat -
needing veterinary treatment.


-- 
Mike.
@millomweb  -
For all your info on Millom and South Copeland
via *the area's premier website - *

*currently unavailable due to ongoing harassment of me, my family, property
& pets*

T&Cs 
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] OSM is a right mess (was: Craigslist OpenStreetMap Rendering Issue)

[Steve Coast]

> On Jun 2, 2015, at 6:22 PM, Eugene Alvin Villar  wrote:
> On Jun 3, 2015 8:06 AM, "pmailkeey ."  > wrote:
> > OSM's k=v design is completely a serious and unnecessary flaw. [...] OSM is 
> > 90% argument, 5% dead-end discussions and 5% progress. The whole is not a 
> > marketable product; it's not fit to be rated as 'beta'. Is this a 
> > significant cause of ex-mappers ? It's a flipping brilliant project but 
> > sadly lacking a great leader.
> 
> It seems you are deeply unsatisfied with how OSM works. And your broad 
> assertions such as that OSM is "not fit" or is "90% argument" are completely 
> unfounded. 
> 

I don’t know; there are a bunch of fairly key and active OSM people who 
unsubscribed from the lists precisely because they felt it was mostly circular 
argument.

Maybe it’s like San Francisco - everything was built decades ago and now it’s 
illegal to build things, so we just argue over whether the golden gate should 
have a suicide-proof railing or if rich people should be allowed to live in the 
mission or not.

Best

Steve___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] OSM is a right mess (was: Craigslist OpenStreetMap Rendering Issue)

[Tom MacWright]

Hi Mike,

Please propose an alternative.

On Tue, Jun 2, 2015 at 8:22 PM, Eugene Alvin Villar 
wrote:

> On Jun 3, 2015 8:06 AM, "pmailkeey ."  wrote:
> > OSM's k=v design is completely a serious and unnecessary flaw. [...] OSM
> is 90% argument, 5% dead-end discussions and 5% progress. The whole is not
> a marketable product; it's not fit to be rated as 'beta'. Is this a
> significant cause of ex-mappers ? It's a flipping brilliant project but
> sadly lacking a great leader.
>
> It seems you are deeply unsatisfied with how OSM works. And your broad
> assertions such as that OSM is "not fit" or is "90% argument" are
> completely unfounded. Sure, OSM is not perfect but I seriously doubt that
> the k=v design or some other point you have raised is the culprit.
>
> Feel free to leave and create a separate project. You can even be the
> "great leader" for that new project that you think OSM needs. If your ideas
> are indeed better, then your project will succeed and you can then prove
> OSM wrong.
>
> ___
> talk mailing list
> talk@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/talk
>
>
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] OSM is a right mess (was: Craigslist OpenStreetMap Rendering Issue)

[Bryce Nesbitt]

On Tue, Jun 2, 2015 at 5:04 PM, pmailkeey . 
wrote:

> iD shows oneway=unknown if it's not set. If it's unknown, iD should not
> show oneway at all.
> OSM's k=v design is completely a serious and unnecessary flaw. Similarly
> are 'categories' like man_made', and 'amenity'.
> Why can we not simply stick to hard facts rather guessing what
> categor(ies) an object fits in
>
> --
> Mike.
> @millomweb  -
> For all your info on Millom and South Copeland
> via *the area's premier website - *
> *currently unavailable due to ongoing harassment of me, my family,
> property & pets*
>

Perhaps http://wikimapia.org/ will better match your needs, and offer more
peace for your family, property and pets.
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] OSM is a right mess (was: Craigslist OpenStreetMap Rendering Issue)

[Eugene Alvin Villar]

On Jun 3, 2015 8:06 AM, "pmailkeey ."  wrote:
> OSM's k=v design is completely a serious and unnecessary flaw. [...] OSM
is 90% argument, 5% dead-end discussions and 5% progress. The whole is not
a marketable product; it's not fit to be rated as 'beta'. Is this a
significant cause of ex-mappers ? It's a flipping brilliant project but
sadly lacking a great leader.

It seems you are deeply unsatisfied with how OSM works. And your broad
assertions such as that OSM is "not fit" or is "90% argument" are
completely unfounded. Sure, OSM is not perfect but I seriously doubt that
the k=v design or some other point you have raised is the culprit.

Feel free to leave and create a separate project. You can even be the
"great leader" for that new project that you think OSM needs. If your ideas
are indeed better, then your project will succeed and you can then prove
OSM wrong.
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk