Re: [OSM-legal-talk] CT-compatible data recycling imported nodes

2012-04-07 Thread rhn
Hi again,

The inevitable data cleaning comes closer and closer, therefore I want to rise 
the issue again before it's too late (with the new knowledge I got from Simon's 
answers).

In short, I used old database objects (which were imports from sources 
incompatible with CT & ODBL) to enter entirely new data into OSM (IDs and some 
key-value pairs stayed the same). The new data are in commits marked as 
CT-compatible.
It appears that the automated cleanup process is going to remove that kind of 
data.

My question is: how to preserve the new data?
I believe that the fact that I surveyed the objects myself and ignored all 
existing positions/tagging breaks "time-continuity" of the objects, and 
therefore their old license.
Is it acceptable for me to wait until cleanup is finished, extract that kind of 
contributions and re-commit them? It's very unlikely that the original author 
of the imported data is going to relicense it. 

As much as I love contributing to OSM, I don't think I'm going to have the same 
willpower again to turn a blank 40x40km area into something useable. I don't 
want to sound negative, but combined with the fact that imports (which I 
consider a defining characteristic of open-source projects) are much harder now 
than a few years earlier, I might abandon OSM if this gets thrown away. I'm 
writing all this precisely because I don't want this to happen.

Cheers,
rhn

> 
> The v0 rule essentially states that allocating an object in the DB 
> doesn't create IP, so if you have an object that has lost all of the 
> attributes it originally had it is essentially a new object.
> 
> However in your case that really doesn't apply (IMHO), because what I've 
> seen from your examples is that you actually imported the data yourself 
> and at least some of the original tags have survived. Note that the data 
> would actually survive the redaction process at this point in time, but 
> naturally you shouldn't have agreed to the CTs in the first place.
> 
> The preferred way to proceed would be for you to get permission to 
> release the data you imported under the ODBL from the original creator 
> in the UMP project, as you probably know there is an effort under way to 
> organize exactly that in Poland.
> 
> Simon
> 
> Am 28.03.2012 22:43, schrieb rhn:
> > Three different examples; all of them were remapped&  verified in respect 
> > to location and tags (except of name=* in most cases). That doesn't mean 
> > the tags have changed though, sometimes they were imported just right.
> >
> > http://www.openstreetmap.org/browse/way/28099536/history
> > http://www.openstreetmap.org/browse/way/28099539/history
> > http://www.openstreetmap.org/browse/way/28099452/history
> >
> > Could you point me to the v0 rule you're referring to?
> >
> > Cheers,
> > rhn
> >
> >> If you essentially remapped the objects it may be that some or most of
> >> your data would be safe due to the v0 rule (regardless of any other
> >> developments wrt UMP). It is difficult to answer this more definitely
> >> we would need to see some examples.
> >>
> >> Simon
> >>
> >> Am 28.03.2012 22:12, schrieb rhn:
> >>> Hello,
> >>>
> >>> Please excuse me if my question has been asked before, I don't follow 
> >>> this list.
> >>>
> >>> Today I found information about the way data is going to be marked as 
> >>> incompatible - the way I understood it, all ways and nodes are going to 
> >>> be reverted to the latest compatible version (i.e. the one before first 
> >>> CC-only changeset).
> >>>
> >>> This worries me, as it seems the bulk of my changesets will be deleted.
> >>> I focused on an area with data coming nearly exclusively from an 
> >>> incompatible source (UMP). Before a license change was even in plans, I 
> >>> managed to replace the road network almost completely with GPS traces and 
> >>> some landuse data with WMS and traces.
> >>> The problem is, I never bothered too much with replacing the actual 
> >>> database objects (takes too much time), thinking removal of source=* 
> >>> would be enough. Let me mention that I removed source only from  nodes 
> >>> and ways that I had precise data about (and would have deleted if it 
> >>> wasn't a hassle).
> >>>
> >>> My questions are: Is it acceptable to copy the snapshot of my current 
> >>> data that would otherwise get deleted and restore it as CT-compatib

Re: [OSM-legal-talk] CT-compatible data recycling imported nodes

2012-03-30 Thread rhn
> 
> Am 29.03.2012 19:16, schrieb rhn:
> >  On a side note, relying on such a decision would be ironic - a lot of
> data I imported were only a copy of a PD map :) Cheers, rhn
> 
> Unluckily that the original source was PD doesn't make a difference
> (legally), what counts is the licence you received the data under. 
> 
I'm well aware of that, it's just something that I found interesting.

> Just so that things are clear, which account did you use that hasn't
> accepted the CTs yet?
> 
I own 2 accounts:
rhn - I used it to import in addition to create, so I didn't mark it as 
CT-compliant by accepting CT.
rhn_CT - created after CT was made mandatory for new users. Unfortunately, I 
didn't contribute much with it yet.

Cheers,
rhn

___
legal-talk mailing list
legal-talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/legal-talk


Re: [OSM-legal-talk] CT-compatible data recycling imported nodes

2012-03-29 Thread rhn
Thanks for the explanations.
I missed a very important detail: I didn't accept CT with this account. However 
one of the Polish community members promised to mark all my changeset that were 
not imports as CT-compatible.
>From this point of view, a lot of my data is clean starting from some commit, 
>perhaps except of those which were imported already tagged properly 
>(paradoxically...), and those with untouched name.
In other areas, I edited data imported by other people in the same way.

My main hope is that, since I can precisely and automatically extract data 
pieces that I created from scratch, the data in question is not considered to 
be derived from CC data and therefore not bound by CC.
Regardless of the decision of UMP members, I don't want to rely solely on it in 
order to keep months of my work alive - that's why I'm asking for an 
alternative resolution here.

On a side note, relying on such a decision would be ironic - a lot of data I 
imported were only a copy of a PD map :)

Cheers,
rhn

> 
> The v0 rule essentially states that allocating an object in the DB 
> doesn't create IP, so if you have an object that has lost all of the 
> attributes it originally had it is essentially a new object.
> 
> However in your case that really doesn't apply (IMHO), because what I've 
> seen from your examples is that you actually imported the data yourself 
> and at least some of the original tags have survived. Note that the data 
> would actually survive the redaction process at this point in time, but 
> naturally you shouldn't have agreed to the CTs in the first place.
> 
> The preferred way to proceed would be for you to get permission to 
> release the data you imported under the ODBL from the original creator 
> in the UMP project, as you probably know there is an effort under way to 
> organize exactly that in Poland.
> 
> Simon
> 
> Am 28.03.2012 22:43, schrieb rhn:
> > Three different examples; all of them were remapped&  verified in respect 
> > to location and tags (except of name=* in most cases). That doesn't mean 
> > the tags have changed though, sometimes they were imported just right.
> >
> > http://www.openstreetmap.org/browse/way/28099536/history
> > http://www.openstreetmap.org/browse/way/28099539/history
> > http://www.openstreetmap.org/browse/way/28099452/history
> >
> > Could you point me to the v0 rule you're referring to?
> >
> > Cheers,
> > rhn
> >
> >> If you essentially remapped the objects it may be that some or most of
> >> your data would be safe due to the v0 rule (regardless of any other
> >> developments wrt UMP). It is difficult to answer this more definitely
> >> we would need to see some examples.
> >>
> >> Simon
> >>
> >> Am 28.03.2012 22:12, schrieb rhn:
> >>> Hello,
> >>>
> >>> Please excuse me if my question has been asked before, I don't follow 
> >>> this list.
> >>>
> >>> Today I found information about the way data is going to be marked as 
> >>> incompatible - the way I understood it, all ways and nodes are going to 
> >>> be reverted to the latest compatible version (i.e. the one before first 
> >>> CC-only changeset).
> >>>
> >>> This worries me, as it seems the bulk of my changesets will be deleted.
> >>> I focused on an area with data coming nearly exclusively from an 
> >>> incompatible source (UMP). Before a license change was even in plans, I 
> >>> managed to replace the road network almost completely with GPS traces and 
> >>> some landuse data with WMS and traces.
> >>> The problem is, I never bothered too much with replacing the actual 
> >>> database objects (takes too much time), thinking removal of source=* 
> >>> would be enough. Let me mention that I removed source only from  nodes 
> >>> and ways that I had precise data about (and would have deleted if it 
> >>> wasn't a hassle).
> >>>
> >>> My questions are: Is it acceptable to copy the snapshot of my current 
> >>> data that would otherwise get deleted and restore it as CT-compatible?
> >>> If yes, should the backup be performed now or is there going to be a way 
> >>> to access CC data after the license change?
> >>> If not, is there any other way to preserve the data?
> >>>
> >>> Cheers,
> >>> rhn
> >>>
> >>> ___
> >>> legal-talk mailing list
> >>> legal-talk@openstreetmap.or

Re: [OSM-legal-talk] CT-compatible data recycling imported nodes

2012-03-28 Thread rhn
Three different examples; all of them were remapped & verified in respect to 
location and tags (except of name=* in most cases). That doesn't mean the tags 
have changed though, sometimes they were imported just right.

http://www.openstreetmap.org/browse/way/28099536/history
http://www.openstreetmap.org/browse/way/28099539/history
http://www.openstreetmap.org/browse/way/28099452/history

Could you point me to the v0 rule you're referring to?

Cheers,
rhn

> If you essentially remapped the objects it may be that some or most of 
> your data would be safe due to the v0 rule (regardless of any other 
> developments wrt UMP). It is difficult to answer this more definitely  
> we would need to see some examples.
> 
> Simon
> 
> Am 28.03.2012 22:12, schrieb rhn:
> > Hello,
> >
> > Please excuse me if my question has been asked before, I don't follow this 
> > list.
> >
> > Today I found information about the way data is going to be marked as 
> > incompatible - the way I understood it, all ways and nodes are going to be 
> > reverted to the latest compatible version (i.e. the one before first 
> > CC-only changeset).
> >
> > This worries me, as it seems the bulk of my changesets will be deleted.
> > I focused on an area with data coming nearly exclusively from an 
> > incompatible source (UMP). Before a license change was even in plans, I 
> > managed to replace the road network almost completely with GPS traces and 
> > some landuse data with WMS and traces.
> > The problem is, I never bothered too much with replacing the actual 
> > database objects (takes too much time), thinking removal of source=* would 
> > be enough. Let me mention that I removed source only from  nodes and ways 
> > that I had precise data about (and would have deleted if it wasn't a 
> > hassle).
> >
> > My questions are: Is it acceptable to copy the snapshot of my current data 
> > that would otherwise get deleted and restore it as CT-compatible?
> > If yes, should the backup be performed now or is there going to be a way to 
> > access CC data after the license change?
> > If not, is there any other way to preserve the data?
> >
> > Cheers,
> > rhn
> >
> > ___
> > legal-talk mailing list
> > legal-talk@openstreetmap.org
> > http://lists.openstreetmap.org/listinfo/legal-talk
> 
> 
> ___
> legal-talk mailing list
> legal-talk@openstreetmap.org
> http://lists.openstreetmap.org/listinfo/legal-talk

___
legal-talk mailing list
legal-talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/legal-talk


[OSM-legal-talk] CT-compatible data recycling imported nodes

2012-03-28 Thread rhn
Hello,

Please excuse me if my question has been asked before, I don't follow this list.

Today I found information about the way data is going to be marked as 
incompatible - the way I understood it, all ways and nodes are going to be 
reverted to the latest compatible version (i.e. the one before first CC-only 
changeset).

This worries me, as it seems the bulk of my changesets will be deleted.
I focused on an area with data coming nearly exclusively from an incompatible 
source (UMP). Before a license change was even in plans, I managed to replace 
the road network almost completely with GPS traces and some landuse data with 
WMS and traces.
The problem is, I never bothered too much with replacing the actual database 
objects (takes too much time), thinking removal of source=* would be enough. 
Let me mention that I removed source only from  nodes and ways that I had 
precise data about (and would have deleted if it wasn't a hassle).

My questions are: Is it acceptable to copy the snapshot of my current data that 
would otherwise get deleted and restore it as CT-compatible?
If yes, should the backup be performed now or is there going to be a way to 
access CC data after the license change?
If not, is there any other way to preserve the data?

Cheers,
rhn

___
legal-talk mailing list
legal-talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/legal-talk


Re: [OSM-legal-talk] What should be considered legal?

2009-10-28 Thread rhn
Martin Koppenhoefer wrote:
>> Thanks. The reason I asked that was that I frequently forget where the GPS 
>> trace was taken - was it a road or a track, which village or whatever else. 
>> This usually happens in areas where OSM map is pitch white :) Yahoo maps 
>> aren't very helpful there either.
> 
> well, you can still upload the traces as they are always usefull (also
> more than one on the same place), especially in white areas, but
> without further information (road name, road class, physical state,
> reference number, restrictions etc.) you should tag them as
> highway=road if you decide to do it (and if it wasn't cross country).
> Btw: I guess you ment yahoo aerial imagery, as we have no right at all
> to trace yahoo maps or take information from it.
> 
> cheers,
> Martin
> 

I do tag them as highway=road or I try to strain my memory (and the map 
accuracy), because I don't like leaving roads like that.
My bad, I meant Yahoo imagery + Landsat.

___
legal-talk mailing list
legal-talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/legal-talk


Re: [OSM-legal-talk] What should be considered legal?

2009-10-25 Thread rhn
Matt Amos wrote:
> On 10/24/09, Martin Koppenhoefer  wrote:
>> 2009/10/21 rhn :
>>> I'm a mapper for more than a year, and I know a little bit about
>>> intellectual property, but some questions have been puzzling me for quite
>>> some time.
>>>
>>> First of them - how much is allowed when referring to proprietary maps? Is
>>> it right to look at the street names to see whether I got them right? Or
>>> can I compare topology of the streets with the external map? See if I got
>>> the village placement right and adjust it?
>> IMHO (IANAL) you can always compare your map to others, but if the
>> don't match, you will not know, who's right, unless you recontroll.
> 
> i'd agree - it's OK to compare OSM to proprietary maps and use that to
> figure out where needs surveying. but it's not OK to take information
> from that proprietary map - if there is a difference then you'll have
> to go out and survey the difference.
> 
> so (imho) it wouldn't be OK to adjust village placement based on
> proprietary maps; if there's a difference you'd have to look at other
> allowable sources like Y! aerial imagery or out-of-copyright maps, or
> go out and survey it with a GPS.
> 
> cheers,
> 
> matt

Thanks. The reason I asked that was that I frequently forget where the GPS 
trace was taken - was it a road or a track, which village or whatever else. 
This usually happens in areas where OSM map is pitch white :) Yahoo maps aren't 
very helpful there either.

Cheers
rhn

___
legal-talk mailing list
legal-talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/legal-talk


[OSM-legal-talk] What should be considered legal?

2009-10-21 Thread rhn
I'm a mapper for more than a year, and I know a little bit about intellectual 
property, but some questions have been puzzling me for quite some time.

First of them - how much is allowed when referring to proprietary maps? Is it 
right to look at the street names to see whether I got them right? Or can I 
compare topology of the streets with the external map? See if I got the village 
placement right and adjust it?

The second I've been wondering about is copying the other way around. Suppose 
I'm working for a mapping company who has to trace satellite data to make a 
vector map. Not all roads are visible through trees and sometimes, the streets 
look similar to footways. Can I look at (printed, not overlayed) OSM maps to 
see where streets are to avoid mapping footways and paths as streets? Is my 
work a derived work of OSM if I do that? There is no tracing over OSM involved.

I hope someone could explain it to me :)

Cheers
rhn

___
legal-talk mailing list
legal-talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/legal-talk