Re: [OSM-talk] import of dataset for new zealand

2008-03-20 Thread Richard Fairhurst
Lester Caine wrote:

 When you jump into some of the links provided HERE you may have no  
 idea in the
 world where you are, so a banner line with 'Auckland, New Zealand'  
 would be
 very useful anyway?

In some circumstances. But over and above a basic attribution  
requirement, you should not and cannot restrict how people render/ 
visualise the OSM data. To do so would remove a fundamental reason  
for the project.

cheers
Richard

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk


Re: [OSM-talk] import of dataset for new zealand

2008-03-20 Thread Robin Paulson
On 20/03/2008, Richard Fairhurst [EMAIL PROTECTED] wrote:
  bear with me here, i'm not sure what you mean:
   what is the serious cartographic issue, and why is it a problem?


 It looks pug ugly!

fair point, i'll agree with that

  Not a lot more that I can say than that, but I'll try. It performs an
  interesting social function on the Osmarender layer. But a large part
  of the art of cartography is decluttering - conveying the maximum
  information with the minimum crowding. Having repeated (c) LINZ (or
  whatever) tags on every road effectively reduces the amount of other
  information you can convey, and the attractiveness of the map.

what about only including info for major data sources that demand
attribution, i.e. not the osm username info? this would at least keep
everything in a (very big ) area labelled the same, plus it would be
easier to explain to a non-copyright-savvy user? of course, only at
zoom 18 where there is generally plenty of space between objects.

or how about as a separate, floating layer that can be turned on and
off? something on by default (again, at zoom 18 only)? when the user
zooms into the area, the layer is turned on, but they are asked
(subtly) if they want it off from that point onwards? would using
cookies to determine if they have been to that zoom level in that area
be workable?

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk


Re: [OSM-talk] import of dataset for new zealand

2008-03-20 Thread Martijn van Oosterhout
On Thu, Mar 20, 2008 at 9:56 AM, Robin Paulson [EMAIL PROTECTED] wrote:
  in the first case, there is room for two attributions - one is the
  proposed read-only tag, called attribution, which is set to 'linz' or
  whatever. the second is from the edit history, retrieved by the API,
  set to the username of an osm user, updated whenever the item is
  tweaked (i.e. write and read)

  in the second case,well i see this as unlikely. how any orgs are
  willing to go to the cost of creating a duplicate data set for an area
  (by which i mean same data, same area), when they can get one that
  exists immediately?

I know you think this is hypothetical but it's a very real problem. In
NL we got the AND data and in the next year or two we will get access
to the NWB, which is the map compiled by the govt. It's not clear at
all that one is going to be better than the other and merging them is
going to be a real problem.

It's fortunate that neither ask for attribution or we'd have a serious
problem on our hands.

Have a nice day,
-- 
Martijn van Oosterhout [EMAIL PROTECTED] http://svana.org/kleptog/

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk


Re: [OSM-talk] import of dataset for new zealand

2008-03-19 Thread Lester Caine
[EMAIL PROTECTED] wrote:
 On Tuesday 18 March 2008 23:43:39 [EMAIL PROTECTED] wrote:
 The problem I can see with this is that the source attribute can be
 altered (or stripped).
 Which is as it should be. If an object originating from that import is
 significantly altered later, then it should be possible to reflect this in
 the source tag.

 
 That's a good question for their lawyers, do we have to maintain
 attribution for data which is found to be incorrect and hence
 corrected
 
 Using another (write once) tag would still enable the original upload to
 be recognized, even if the data is corrected at a later stage.

THIS one is a good point. The wiki/framework I'm working with personally has 
two user flags per item. Created and Last Edit. The created flag is only 
populated when an item is created and while it can be changed, the normal user 
access does not allow this. I would have thought THIS should be a norm for any 
of the bluk loaded data. How do we currently tag TIGER and other bulk data?

Of cause the second level on any good content framework is 'history' and I can 
see all the changes to an item and who entered them - but that is something 
that has been shot down already because of the quantity of traffic :(

-- 
Lester Caine - G8HFL
-
Contact - http://home.lsces.co.uk/lsces/wiki/?page=contact
L.S.Caine Electronic Services - http://home.lsces.co.uk
EnquirySolve - http://enquirysolve.com/
Model Engineers Digital Workshop - http://medw.co.uk//
Firebird - http://www.firebirdsql.org/index.php

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk


Re: [OSM-talk] import of dataset for new zealand

2008-03-19 Thread Cartinus
Anybody wanting (semi-)read-only tags in OSM should read the discussions about 
the edit war in Cyprus to see how futile they are for this purpose.

Can't edit a tag? Just delete the whole object and create it anew.

For a more practical example related to mass imports:
The waterbodies from the AND import in the Netherlands are often less accurate 
than the Yahoo aerial imaging that has recently become available here. Now I 
can tug at the polygon from the AND import till it conforms better to the 
Yahoo image and put something in the source tag that reflects both sources. 
If the source tag was read-only, then I'd delete the polygon from the AND 
import and trace a new one from the Yahoo image. The net effect would be a 
break in the edit history, because nothing links the new polygon to the old 
one.

-- 
m.v.g.,
Cartinus

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk


Re: [OSM-talk] import of dataset for new zealand

2008-03-19 Thread Andy Robinson (blackadder)
Robin Paulson wrote:
Sent: 19 March 2008 7:56 AM
To: Cartinus
Cc: talk@openstreetmap.org
Subject: Re: [OSM-talk] import of dataset for new zealand

On 19/03/2008, Cartinus [EMAIL PROTECTED] wrote:
 Anybody wanting (semi-)read-only tags in OSM should read the discussions
about
  the edit war in Cyprus to see how futile they are for this purpose.

  Can't edit a tag? Just delete the whole object and create it anew.

true, but is that really an issue here? why (in this context, with no
disputes of the magnitude of those in cyprus) would someone do that?
what if someone does delete a road sourced from linz, and then draws
it in again? the source of the information for that road is no longer
linz, it's the person who re-drew it, isn't it?


Correct, if someone deletes data and starts anew then the new data has no
link with the old, provided it wasn't derived from the old. Derived data is
much greyer, since it's likely that with time most objects that get placed
in the database will be further edited, either to change their spatial
aspects or to add additional information. Once the data has been through a
number of changes its really impossible to say that it originated from one
or more parties unless you minutely interrogate the history of each object. 

This is one of the reasons why I firmly believe that nobody really owns the
data they put in the database, the reality is that individuals and
organisations contribute to the pool. We all have to accept that we are
making these contributions for the good of the project, not because it puts
our or their names in lights for being associated with the most data. It
should be the same for organisations. If linz is happy to see their data
contributed for the good of all then that's great, if they want to see their
name associated with their data then we can do that by adding source tags
when its imported and using an appropriate username for the import. We can
also provide generic attribution like we give to anyone else on the wiki,
via the news/blog and elsewhere as we did for instance with AND. Beyond that
linz needs to be happy that their data will become diluted with time. This
is probably the most compelling reason why visual attribution on mapping
created from the database just cannot work.

I hope that you can persuade Linz that releasing their data under a modified
attribution approach to the one they prescribe is not about OSM wanting to
have their data only in its own terms. OSM appreciates greatly major
contributions of data and we like to give prominent coverage to those that
do so. The issue is about keeping the approach generic, so that we have one
rule that fits all, as anything less will be impracticable with time.

best of luck with the discussion

Cheers

Andy


  For a more practical example related to mass imports:
  The waterbodies from the AND import in the Netherlands are often less
accurate
  than the Yahoo aerial imaging that has recently become available here.
Now I
  can tug at the polygon from the AND import till it conforms better to
the
  Yahoo image and put something in the source tag that reflects both
sources.
  If the source tag was read-only, then I'd delete the polygon from the
AND
  import and trace a new one from the Yahoo image. The net effect would be
a
  break in the edit history, because nothing links the new polygon to the
old
  one.

if you deleted the AND data and then re-drew it, the lake would not
then be sourced from AND, so there is no need to attribute them. it's
a derivative work of you and yahoo.

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk


___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk


Re: [OSM-talk] import of dataset for new zealand

2008-03-19 Thread Frederik Ramm
Hi,

On 19.03.2008, at 09:14, Lester Caine wrote:
 Which is why FOR LEGAL REASONS most decent wiki type frameworks  
 don't have
 delete as an option for basic users!

[...]

 HOPEFULLY more countries will follow AND and New Zealand in  
 providing accurate
 data but only if that data is properly managed in the framework.

We don't manage other people's data on their behalf. We assimilate it  
into the collective, and it thereby becomes everybody's data. We  
modify it, delete it, replace it, change all of its properties and  
tags as we see fit. If people are not happy with this then they  
shouldn't give us their data.

Bye
Frederik

-- 
Frederik Ramm  ##  eMail [EMAIL PROTECTED]  ##  N49°00'09 E008°23'33




___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk


Re: [OSM-talk] import of dataset for new zealand

2008-03-19 Thread Lester Caine
Frederik Ramm wrote:
 Hi,
 
 On 19.03.2008, at 09:14, Lester Caine wrote:
 Which is why FOR LEGAL REASONS most decent wiki type frameworks don't 
 have
 delete as an option for basic users!
 
 [...]
 
 HOPEFULLY more countries will follow AND and New Zealand in providing 
 accurate
 data but only if that data is properly managed in the framework.
 
 We don't manage other people's data on their behalf. We assimilate it 
 into the collective, and it thereby becomes everybody's data. We modify 
 it, delete it, replace it, change all of its properties and tags as we 
 see fit. If people are not happy with this then they shouldn't give us 
 their data.

And trash perfectly good information just because you can?

We have already seen examples of the sort of 'vandalism' that disputes can 
cause. *IF* everybody plays nicely together then we do not have a problem, but 
we live in a world were people take pleasure in doing as much damage as they 
can and at present there is not an easy way of repairing damage *IF* someone 
decided to have major fun.

We don't have tools for modifying contributed tracks, and contributed data 
COULD be handled in a similar way? A layer that can be copied much like any 
other overlay. The 'contributed by' would then be clearly identified and we 
can then go back to the source data if we find that something HAS been 
'modified' for what ever reason.

I suppose at the end of the day a link to the source data such as is available 
for TIGER data may have to do, but having everything in the one easily 
handleable format makes things a lot more flexible? Given the major holes in 
even just the UK - anything that would encourage access to other data sources 
should be encouraged?

-- 
Lester Caine - G8HFL
-
Contact - http://home.lsces.co.uk/lsces/wiki/?page=contact
L.S.Caine Electronic Services - http://home.lsces.co.uk
EnquirySolve - http://enquirysolve.com/
Model Engineers Digital Workshop - http://medw.co.uk//
Firebird - http://www.firebirdsql.org/index.php

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk


Re: [OSM-talk] import of dataset for new zealand

2008-03-19 Thread Tom Evans
Ultimately I'm not sure we have a lot of wiggle room here because of the 
share-alike license.

We could offer that the main website does nice attributions at the bottom of 
the window, but we can't put any such _demands_ on the use of the data.  The 
reason being that I (speaking as one of many contributors) contributed my data 
under share-alike.  You are not allowed to redistribute _my_ data under a more 
restrictive license, such as one demanding selective attribution for some parts 
of the db.

I could see a future for 'soft' attribution, where OSMF promise to attribute 
where possible, but not for a 'hard' demand which must follow the planet file 
wherever it goes.  If we smile sweetly and offer this 'soft' attribution, will 
LINZ and others be happy with that?

Tom




___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk


Re: [OSM-talk] import of dataset for new zealand

2008-03-19 Thread Robin Paulson
On 20/03/2008, Stephen Gower [EMAIL PROTECTED] wrote:
   I absolutely agree with this (aka +1).

   OSM's baseline is user-collected data, freely given to project.
   Imports of large datasets are nice, but if they can't be given
   under the same terms as my GPS tracks, OSM should have to use
   alternative sources. I believe the proof of large-scale useful
   coverage in the UK and other places, entirely contributed by
   user-effort, should give the project confidence to walk-away from
   incompatable data sources, rather than try to jump through hoops to
   get them.

i do agree with the points on principle - however activity in uk does
not necessarily prove there will be the same activity elsewhere. the
reality of the matter, is that nz has an area about the size of uk,
and less than 1/15 the population..we could be waiting a while for
enough individual users to map the whole country :-(

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk


Re: [OSM-talk] import of dataset for new zealand

2008-03-19 Thread Robin Paulson
On 19/03/2008, Cartinus [EMAIL PROTECTED] wrote:
 Twice no.

  Objects in the database are more than just shapes. To concentrate on the
  second one: The shape from the Yahoo image is better, but there are no names
  visible in the aerial photographs. If I don't know the name (and whatever
  other attributes) before looking at the AND data, then AND is still source of
  part of the information contained in the object.

well, yes but you're moving the goal posts here. what you're
suggesting is no different in concept to a user looking up street
names in google maps/microsoft live maps/etc. and adding them to the
database, having traced the shape from the yahoo imagery. so, we treat
it in the same way: unless and until we (or the owners of the data)
suspect copying is going on, we assume people are playing by the rules
and getting their data legally, and do nothing

and for an org to get worried about this (the linz data contains
millions of items), a user would have to remove attribution from a
significant amount of data. this takes a lot of time. why would anyone
set out to do this? there are far easier ways of causing problems for
osm, if it's malice that drives them, and i see no practical benefit
that anyone could gain from doing this. if you want to be identified
as having worked on it, your name will be there as last editor by
tweaking the nodes

i understand you're looking at the worst case, and that is useful, but
so is looking at the practicalities and the likely actions of people

to anyone that might know: is there any evidence of this happening
with TIGER and AND data, people removing attribution tags after the
imports?

  Alternatively maybe not the whole lake needed to be redrawn, but that one
  read-only tag gave me no other choice.

well, no it doesn't give you no other choice. if part of the lake is
out of line, drag it to the correct position. if it's represented too
crudely, add some more points to the loop and drag those to the
correct point. data can always be modified to what we want, and
attribution will be kept

hell, talking hypothetically, there's no reason i couldn't convert
that lake into a motorway, a line of power cables/pylons, a large
building, or anything else represented by a way. drag the points to
new position, maybe break the link so it's not a polygon, change the
attributes, hey presto you're there and the linz attribution is kept.
pointless changes i know, but the principle stands

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk


Re: [OSM-talk] import of dataset for new zealand

2008-03-19 Thread Richard Fairhurst
Robin Paulson wrote:

 [adding attribution to the Mapnik layer]
 bear with me here, i'm not sure what you mean:
 what is the serious cartographic issue, and why is it a problem?

It looks pug ugly!

Not a lot more that I can say than that, but I'll try. It performs an  
interesting social function on the Osmarender layer. But a large part  
of the art of cartography is decluttering - conveying the maximum  
information with the minimum crowding. Having repeated (c) LINZ (or  
whatever) tags on every road effectively reduces the amount of other  
information you can convey, and the attractiveness of the map.

cheers
Richard

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk


[OSM-talk] import of dataset for new zealand

2008-03-18 Thread Robin Paulson
land information new zealand is a government org that holds data on
roads and properties for the entirety of new zealand. they have
recently given permission to use their data sets in osm, with a caveat
that we include an attribution statement:

Contains data sourced from Land Information New Zealand. Crown
Copyright reserved. Land Information New Zealand gives no warranty in
relation to the data, including its accuracy, reliability and
suitability and accepts no liability whatsoever in relation to any
loss, damage or other costs relating to the use of any data, any
compilations, derivative works or modifications of the data.

this is a huge data set, with a lot of very useful information. it
will immediately bring road coverage (currently very poor) for the
entire country up to 100%, and may include property information for
every land title in the country, which then opens the door for other
open data sets, such as zenbu.co.nz (business listings)

the key is the attribution

at present there is no method for attributing data in osm, so this is
a call to all:

a technical solution is sought, to display the corresponding
information whenever linz-sourced data is being viewed.

suggestions and comments, please

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk


Re: [OSM-talk] import of dataset for new zealand

2008-03-18 Thread Andy Allan
On Tue, Mar 18, 2008 at 9:44 AM, Robin Paulson [EMAIL PROTECTED] wrote:
 land information new zealand is a government org that holds data on
  roads and properties for the entirety of new zealand. they have
  recently given permission to use their data sets in osm, with a caveat
  that we include an attribution statement:

  Contains data sourced from Land Information New Zealand. Crown
  Copyright reserved. Land Information New Zealand gives no warranty in
  relation to the data, including its accuracy, reliability and
  suitability and accepts no liability whatsoever in relation to any
  loss, damage or other costs relating to the use of any data, any
  compilations, derivative works or modifications of the data.

  this is a huge data set, with a lot of very useful information. it
  will immediately bring road coverage (currently very poor) for the
  entire country up to 100%, and may include property information for
  every land title in the country, which then opens the door for other
  open data sets, such as zenbu.co.nz (business listings)

  the key is the attribution

  at present there is no method for attributing data in osm, so this is
  a call to all:

  a technical solution is sought, to display the corresponding
  information whenever linz-sourced data is being viewed.

  suggestions and comments, please

Aargh. This is so close to being extremely useful. But if that's going
to be a requirement for my cycle map to display that, and all the
sites that use the main slippy map, or those embedding the cycle map
tiles, and the garmin.img files need to display that, and trekbuddy
needs to display that, then there'll be a strong demand for a weekly
planet-no-NZ.osm.bz2, which is not exactly what we want to happen!

I hope you can find some way around this restriction they are asking for!

Cheers,
Andy

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk


Re: [OSM-talk] import of dataset for new zealand

2008-03-18 Thread Tom Chance

Hello,

On Tue, 18 Mar 2008 11:03:48 +, Andy Allan [EMAIL PROTECTED]
wrote:
 On Tue, Mar 18, 2008 at 9:44 AM, Robin Paulson [EMAIL PROTECTED]
 wrote:
 land information new zealand is a government org that holds data on
  roads and properties for the entirety of new zealand. they have
  recently given permission to use their data sets in osm, with a caveat
  that we include an attribution statement:

  Contains data sourced from Land Information New Zealand. Crown
  Copyright reserved. Land Information New Zealand gives no warranty in
  relation to the data, including its accuracy, reliability and
  suitability and accepts no liability whatsoever in relation to any
  loss, damage or other costs relating to the use of any data, any
  compilations, derivative works or modifications of the data.
 
 Aargh. This is so close to being extremely useful. But if that's going
 to be a requirement for my cycle map to display that, and all the
 sites that use the main slippy map, or those embedding the cycle map
 tiles, and the garmin.img files need to display that, and trekbuddy
 needs to display that, then there'll be a strong demand for a weekly
 planet-no-NZ.osm.bz2, which is not exactly what we want to happen!

What if we upload all of that data under a particular username? In that
sense they are getting attribution as much as any other contributor to OSM.
I wonder if they would be happy with their notice going on the OSM wiki,
with a note pointing to that under their user account and any OSM wiki
pages about NZ?

Tata,
Tom


___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk


Re: [OSM-talk] import of dataset for new zealand

2008-03-18 Thread Robin Paulson
On 19/03/2008, Ian Darwin [EMAIL PROTECTED] wrote:
   the key is the attribution
  
   at present there is no method for attributing data in osm, so this is
   a call to all:


 I presume you've verified that putting it on the OSM web site would not
  be adequate? I think there needs to be a separate page there listing all
  the bulk contributors (Holland, USA, Spain, and now NZ, et al) with a
  link from the front page.

no, it's early days yet, approval to use the data only came through
yesterday. we've not verified anything attribution wise, this is the
first step - asking for possible solutions, which we will then present
to LINZ for their approval. your suggestion has been added to the LINZ
import page on the wiki, feel free to add more/edit as you see fit

http://wiki.openstreetmap.org/index.php/LINZ

  I would *think* this would cover it, but IANAL. Good luck!

  Hmm, if this works, we might be able to get some maps for Canada too.

  P.S. Amused by the cross-post: I too am both an OM-er and and OSM-er.
  I did most of the roads around my home:
  http://www.openstreetmap.org/?mlon=-79.93265mlat=44.012150zoom=13

well, that was actually a mistake on my part.but there's a lot of
people working on/aware of both - i'm particularly impressed to see a
number of osm compatible mapping/routing applications developed for
openmoko/the neo

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk


Re: [OSM-talk] import of dataset for new zealand

2008-03-18 Thread Cartinus
On Tuesday 18 March 2008 23:43:39 [EMAIL PROTECTED] wrote:
 The problem I can see with this is that the source attribute can be
 altered (or stripped).

Which is as it should be. If an object originating from that import is 
significantly altered later, then it should be possible to reflect this in 
the source tag.

-- 
m.v.g.,
Cartinus

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk


Re: [OSM-talk] import of dataset for new zealand

2008-03-18 Thread Robin Paulson
On 19/03/2008, [EMAIL PROTECTED] [EMAIL PROTECTED] wrote:
   Which is as it should be. If an object originating from that import is
   significantly altered later, then it should be possible to reflect this in
   the source tag.

of course, defining 'significantly altered' becomes an issue

maybe we can ignore it, and keep it as linz in perpetuity, no matter
how much it changes? i don't see too many issues with this, and
anyway, i would imagine only a small number would change enough to
warrant complete removal?

 That's a good question for their lawyers, do we have to maintain
  attribution for data which is found to be incorrect and hence
  corrected

i'll put this point forward

  Using another (write once) tag would still enable the original upload to
  be recognized, even if the data is corrected at a later stage.

yes, providing double-attribution - linz/some other data set
originator and the last editor of the item

  Thinking more on how osmarender places a copyright on one side of a way,
  could the other sided be used for an 'attribution' tag?

how much of an issue would it be to do this in mapnik as well?

___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/cgi-bin/mailman/listinfo/talk