[OSM-talk] Big mess- fight or flee?

[Andrew Errington]

Hi everyone,

I have been mapping in Korea a lot.  In July I discovered a problem
because I took a trip to an area I had mapped before.  When I went to use
my new data to check against the map I thought I was going mad.  I was
sure I had mapped certain roads, and they were there on the map, but my
name was no longer in the history.  In fact, there seemed to be only one
version of the road in existence.

Here is one way I know I mapped:
http://www.openstreetmap.org/browse/way/44818037

It was deleted as part of this changeset, which is rather large:
http://www.openstreetmap.org/browse/changeset/7300872

Here's another:
http://www.openstreetmap.org/browse/way/83432223/history

>From this large changeset:
http://www.openstreetmap.org/browse/changeset/7575051

I contacted the mapper responsible, who apologised to me and basically
said that they messed something up, or something went wrong during the
editing session, couldn't fix it, so did the best they could to put things
back.

So, I am happy that it was not malicious, but not happy that a lot of my
work (and others' work) has been lost/needlessly altered, and there is no
continuity in the history.  Anyway, I'd like to know what is the best
course of action.  I still have my GPS traces and photos of streetsigns
and other detail, so I could reload it and re-map the area, but it's a big
area...

Should I...
a) Ask someone to investigate what happened and show me what tools to use
to recover/restore deleted data, bearing in mind that the current data
seems to be identical and would overlap restored data?
b) Check/re-map the areas using whatever data I have?
c) Ignore the issue?
d) None of the above

I know that in some ways the database is 'self-healing' since mappers who
spot discrepancies (due to whatever cause) will fix them.  But these
errors don't need to be discovered- I know they are there (and now you do
too!).

I also wonder if there should be some mechanism to stop (or at least draw
attention to) massive edits/deletions before too much time goes by.

Thanks,

Andrew


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

Re: [OSM-talk] How to start to remove non-CT compliant data..

[Павел Фомин]

What about this case:
v1 is CT-compliant.
v2 adds a new tag and is not CT-compliant.
Then, v3 changes this tag and adds a bunch of other tags.
Will these other tags be considered compliant?

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

Re: [OSM-talk] How to start to remove non-CT compliant data..

[Anthony]

On Thu, Sep 1, 2011 at 2:57 AM, Ian Sergeant  wrote:
> When I have a v1 object that is non-CT compliant, then we have to assume the
> further revisions may be derivatives.

Why do we have to assume this?

> If CT-agreed mappers have added tags
> from a survey in later revisions, then we can possibly grab those, but apart
> from that it is a remapping effort that needs to be undertaken.

How can that remapping effort avoid making a derivative?

> However, when we have a v2 non-CT compliant object based on a v1
> CT-compliant one, it is a different story.

Sure, but how do we recognize a v1 CT-compliant object?  The average
mapper does not have the legal expertise to determine CT compliance.

> The objective is a CT-clean database, with the absolute minimum data loss.
>
> The discussion is about the best way to accomplish that, especially where we
> have CT-agreed versions of objects that we want to leverage.

I would suggest that having amateurs determine what is and is not
compliant is most certainly not the best way to accomplish this.

Furthermore, the goal is not to have a CT-clean database.  You already
have a CT-clean database.  The goal, apparently, is to have an
ODbL-clean database.

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

Re: [OSM-talk] How to start to remove non-CT compliant data..

[80n]

On Thu, Sep 1, 2011 at 12:44 PM, Anthony  wrote:

> Furthermore, the goal is not to have a CT-clean database.  You already
> have a CT-clean database.  The goal, apparently, is to have an
> ODbL-clean database.
>
> I think you mean a CT-clean contributor-base.  Much of the database content
is un-infected by the CTs.
___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] How to start to remove non-CT compliant data..

[Anthony]

On Thu, Sep 1, 2011 at 8:14 AM, 80n <80n...@gmail.com> wrote:
> On Thu, Sep 1, 2011 at 12:44 PM, Anthony  wrote:
>>
>> Furthermore, the goal is not to have a CT-clean database.  You already
>> have a CT-clean database.  The goal, apparently, is to have an
>> ODbL-clean database.
>>
> I think you mean a CT-clean contributor-base.  Much of the database content
> is un-infected by the CTs.

What I mean is that the database is compatible with the CTs, as the
CTs allow the database to be released under CC-BY-SA.  Alternatively
put, the CTs do not require any content to be removed.  Licensing the
database under the ODbL *would* require content to be removed.  But
the CTs do not require this.

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

Re: [OSM-talk] Big mess- fight or flee?

[Toby Murray]

On Thu, Sep 1, 2011 at 2:20 AM, Andrew Errington
 wrote:
> Should I...
> a) Ask someone to investigate what happened and show me what tools to use
> to recover/restore deleted data, bearing in mind that the current data
> seems to be identical and would overlap restored data?
> b) Check/re-map the areas using whatever data I have?
> c) Ignore the issue?
> d) None of the above
>
> I know that in some ways the database is 'self-healing' since mappers who
> spot discrepancies (due to whatever cause) will fix them.  But these
> errors don't need to be discovered- I know they are there (and now you do
> too!).
>
> I also wonder if there should be some mechanism to stop (or at least draw
> attention to) massive edits/deletions before too much time goes by.

There are tools to revert changesets but I think too much time has
passed since this change for them to be used easily since there will
be lots of conflicts with things that have been touched since then. If
the user had asked for help immediately when they noticed the mistake,
things could have been reverted easily.

As for detecting these changes... Do you know about OWL[1]? Although
this may not be much help since it sounds like this didn't happen in
your "home" area. You can watch change feeds from anywhere but at some
point a large enough area will start to overwhelm someone trying to
watch it.

I personally usually have a copy of LiveMapViewer[2] running at home.
Every once in a while I check on activity I see. I have found a couple
instances of new users being clueless and have been able to contact
them and resolve things quickly. I think that's about all we have at
the moment.

[1] http://wiki.openstreetmap.org/wiki/OWL_(OpenStreetMap_Watch_List)
[2] http://wiki.openstreetmap.org/wiki/LiveMapViewer

Toby

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

[OSM-talk] Removing non-CT data method?

[Mike N]

It seems to me to be premature to start removing map features.   Edit, 
remap from aerial and acceptable sources would be OK, IMO.   Just not 
something I'll be spending time on.


   I was checking out a local change and was surprised that the church 
was missing since it is a prominent local feature that I drive by 
frequently.   So I added it back.  Then I looked at a local area extract 
to find it in the history:


http://www.openstreetmap.org/browse/node/469532416/history

  The last 2 edit authors have accepted the CTs, but the feature is 
still deleted?



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

Re: [OSM-talk] Removing non-CT data method?

[Nathan Edgars II]

On 9/1/2011 5:39 PM, Mike N wrote:

http://www.openstreetmap.org/browse/node/469532416/history

The last 2 edit authors have accepted the CTs, but the feature is still
deleted?


Looks like vandalism by rw__.

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

Re: [OSM-talk] Removing non-CT data method?

[Apollinaris Schoell]

Why? it's replaced by a building polygon. so it's improvement of data and
license status.


On Thu, Sep 1, 2011 at 3:06 PM, Nathan Edgars II  wrote:

> On 9/1/2011 5:39 PM, Mike N wrote:
>
>> http://www.openstreetmap.org/**browse/node/469532416/history
>>
>> The last 2 edit authors have accepted the CTs, but the feature is still
>> deleted?
>>
>
> Looks like vandalism by rw__.
>
> __**_
> talk mailing list
> talk@openstreetmap.org
> http://lists.openstreetmap.**org/listinfo/talk
>
___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Removing non-CT data method?

[Thomas Davie]

Look at the whole change set, notably, it includes adding this way:

http://www.openstreetmap.org/browse/way/128541629

Bob
if (*ra4 != 0xffc78948) { return false; }

On 1 Sep 2011, at 23:06, Nathan Edgars II wrote:

> On 9/1/2011 5:39 PM, Mike N wrote:
>> http://www.openstreetmap.org/browse/node/469532416/history
>> 
>> The last 2 edit authors have accepted the CTs, but the feature is still
>> deleted?
> 
> Looks like vandalism by rw__.
> 
> ___
> talk mailing list
> talk@openstreetmap.org
> http://lists.openstreetmap.org/listinfo/talk

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

Re: [OSM-talk] Removing non-CT data method?

[Nathan Edgars II]

On 9/1/2011 6:20 PM, Thomas Davie wrote:

Look at the whole change set, notably, it includes adding this way:

http://www.openstreetmap.org/browse/way/128541629


Which was (and is) not labeled as the church. Mike had to add the church 
back: http://www.openstreetmap.org/browse/node/1420378996


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

Re: [OSM-talk] Removing non-CT data method?

[Mike N]

On 9/1/2011 6:14 PM, Apollinaris Schoell wrote:

Why? it's replaced by a building polygon. so it's improvement of data
and license status.


The building polygon has no tags except building=yes.  An anonymous 
building has less value to OSM than a searchable POI in my opinion.


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

Re: [OSM-talk] How to start to remove non-CT compliant data..

[Ian Sergeant]

Павел Фомин  wrote on 01/09/2011 09:24:30 PM:

> What about this case:
> v1 is CT-compliant.
> v2 adds a new tag and is not CT-compliant.
> Then, v3 changes this tag and adds a bunch of other tags.
> Will these other tags be considered compliant?

This highlights one of the issues.  The v3 may or not be derived, and 
telling whether it is will depend on a curious blend of logical 
heuristics, subsequent human evaluation combined with an assessment 
against developing multi-national legal precedents.  At the end of the day 
we may choose to radically include a v3 object where all non-CT-compliant 
tags have been overwritten or removed, or we may conservatively choose to 
remove anything that has a possibility of being tainted by an earlier 
revision.

If the v3 editor can optionally just modify the v1 CT-compliant version of 
the object, then the problem is minimised, and the task simplified.

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

Re: [OSM-talk] Removing non-CT data method?

[Apollinaris Schoell]

You have been too fast in adding them and fix the data.
Certainly agree that a building alone has not much value. drawing nice
looking maps is not much value compared to a verified POI

On Thu, Sep 1, 2011 at 5:34 PM, Mike N  wrote:

> On 9/1/2011 6:14 PM, Apollinaris Schoell wrote:
>
>> Why? it's replaced by a building polygon. so it's improvement of data
>> and license status.
>>
>
> The building polygon has no tags except building=yes.  An anonymous
> building has less value to OSM than a searchable POI in my opinion.
>
> __**_
> talk mailing list
> talk@openstreetmap.org
> http://lists.openstreetmap.**org/listinfo/talk
>
___
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] Removing non-CT data method?

[Nathan Edgars II]

On 9/1/2011 9:08 PM, Apollinaris Schoell wrote:

You have been too fast in adding them and fix the data.
Certainly agree that a building alone has not much value. drawing nice
looking maps is not much value compared to a verified POI


I can't figure out what you're trying to say or who you're directing it at.

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