Re: [OSM-talk] #AttributionIsNotOptional experiment on OSM France tile servers

[julien_thevenon--- via talk]

 




Le jeudi 12 mars 2020 à 18:33:15 UTC+1, Simon Poole  a écrit :

> -that- is not what is analogous to what we are discussing, it would be
more like displaying the message that you are late with the payment on
every TV in the country.

It would be only for every TV relying on my stream because I share it...
We do not broadcast the tiles. Here this affect only people consuming our tiles 
through this website.
is your point the distinction between the fact that the website only refer to 
us and is not in the middle (ie consuming and reserving our tiles )?

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

Re: [OSM-talk] #AttributionIsNotOptional experiment on OSM France tile servers

[Simon Poole]

Am 12.03.2020 um 16:06 schrieb THEVENON Julien:
> Le jeudi 12 mars 2020 à 15:43:17 UTC+1, Simon Poole  a écrit 
> : 
>
>> To use a completely different example: assume that you purchase a TV set 
>> paid by monthly instalments and you default on them. In civilised countries 
>> that doesn't give the seller the right to break in to your apartment and 
>> repossess the TV, they don't get to cut off electricity to the flat and they 
>> don't get the right to stick big notices on your doors. The seller needs to 
>> utilize the  whatever tools are provided by the legal system, totally 
>> regardless off how upset they are and how righteous they might feel about 
>> their actions.
> Hi Simon,
>
> My internet access provider provide IP TV. If I`m late to pay my TV display a 
> message saying I cannot look at stream until my debt is payed and this is 
> perfectly legal even it targets me precisely.
> My Internet access provider don't break into my appartment neither modify my 
> TV, this is just the stream it sent to me.
> In the case that interest us this is the stream of tile that is modified in 
> the same way except this is due to a non-respect of license instead of a debt.

-that- is not what is analogous to what we are discussing, it would be
more like displaying the message that you are late with the payment on
every TV in the country.

Simon

> Best regards
> Julien
>



signature.asc
Description: OpenPGP digital signature
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] #AttributionIsNotOptional experiment on OSM France tile servers

[Simon Poole]

Am 12.03.2020 um 15:56 schrieb Pierre Béland:
> Mar. 12 2020 10 h 43 min  UTC−4, Simon Poole wrote :
>
> > To use a completely different example: assume that you purchase
> a TV set paid by monthly instalments and you default on them. In
> civilised countries that doesn't give the seller the right to break in
> to your apartment and repossess the TV, they don't get to cut off
> electricity to the flat and they don't get the right to stick big
> notices on your doors. The seller needs to utilize the  whatever
> tools are provided by the legal system, totally regardless off how
> upset they are and how righteous they might feel about their actions.
> Simon
>
> An other example
> Let say we produce bricks standing outside of the shop.  Since too
> many are stolen, we use a trick to make the bricks flashing with a
> message when they get in inappropriate hands. Can somebody sue us
> because their house is flashing with message about where the bricks
> come from ?

Breach of a contract is not the same as stealing goods*, but depending
on legislation you could very well get sued for disclosing the alleged
criminals name.

Simon

* in IP legislation things are not quite so clear but that is really
going too far now.



signature.asc
Description: OpenPGP digital signature
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] #AttributionIsNotOptional experiment on OSM France tile servers

[THEVENON Julien via talk]

Le jeudi 12 mars 2020 à 15:43:17 UTC+1, Simon Poole  a écrit : 

> To use a completely different example: assume that you purchase a TV set paid 
> by monthly instalments and you default on them. In civilised countries that 
> doesn't give the seller the right to break in to your apartment and repossess 
> the TV, they don't get to cut off electricity to the flat and they don't get 
> the right to stick big notices on your doors. The seller needs to utilize the 
>  whatever tools are provided by the legal system, totally regardless off how 
> upset they are and how righteous they might feel about their actions.

Hi Simon,

My internet access provider provide IP TV. If I`m late to pay my TV display a 
message saying I cannot look at stream until my debt is payed and this is 
perfectly legal even it targets me precisely.
My Internet access provider don't break into my appartment neither modify my 
TV, this is just the stream it sent to me.
In the case that interest us this is the stream of tile that is modified in the 
same way except this is due to a non-respect of license instead of a debt.

Best regards
Julien


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

Re: [OSM-talk] #AttributionIsNotOptional experiment on OSM France tile servers

[Pierre Béland via talk]

Mar. 12 2020 10 h 43 min  UTC−4, Simon Poole wrote :
> To use a completely different example: assume that you purchase a TV set 
> paid by monthly instalments and you default on them. In civilised 
> countries that doesn't give the seller the right to break in to your 
> apartment and repossess the TV, they don't get to cut off electricity to 
> the flat and they don't get the right to stick big notices on your doors. 
> The seller needs to utilize the  whatever tools are provided by the legal 
> system, totally regardless off how upset they are and how righteous they 
> might feel about their actions. Simon
An other example 
Let say we produce bricks standing outside of the shop.  Since too many are 
stolen, we use a trick to make the bricks flashing with a message when they get 
in inappropriate hands. Can somebody sue us because their house is flashing 
with message about where the bricks come from ?

 
Pierre 
 

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

Re: [OSM-talk] #AttributionIsNotOptional experiment on OSM France tile servers

[Simon Poole]

The process of

- creating a list of sites that you want to target

- creating a message and designing a method to display it on the 3rd
parties website

is very much deliberately scribbling on the third parties website.

To use a completely different example: assume that you purchase a TV set
paid by monthly instalments and you default on them. In civilised
countries that doesn't give the seller the right to break in to your
apartment and repossess the TV, they don't get to cut off electricity to
the flat and they don't get the right to stick big notices on your
doors. The seller needs to utilize the  whatever tools are provided by
the legal system, totally regardless off how upset they are and how
righteous they might feel about their actions.

Simon

Am 12.03.2020 um 15:09 schrieb Martin Koppenhoefer:
> Am Do., 12. März 2020 um 11:50 Uhr schrieb Simon Poole  >:
>
> So say you scribble on a German companies website,
>
>
>
> I am not talking about "scribbling" on someone else's website. The
> case at hand is about a specific website infringing attribution
> requirements and as a result they themselves integrating tiles (from
> the French tileserver) which makes people aware that they are
> infringing. "Scribbling" on someone else's homepage is very different
> to changing the addresses of images / renaming files on your own
> webserver.
>
> Cheers
> Martin
>


signature.asc
Description: OpenPGP digital signature
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] #AttributionIsNotOptional experiment on OSM France tile servers

[Martin Koppenhoefer]

Am Do., 12. März 2020 um 11:50 Uhr schrieb Simon Poole :

> So say you scribble on a German companies website,
>


I am not talking about "scribbling" on someone else's website. The case at
hand is about a specific website infringing attribution requirements and as
a result they themselves integrating tiles (from the French tileserver)
which makes people aware that they are infringing. "Scribbling" on someone
else's homepage is very different to changing the addresses of images /
renaming files on your own webserver.

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

Re: [OSM-talk] #AttributionIsNotOptional experiment on OSM France tile servers

[Simon Poole]

Am 12.03.2020 um 10:58 schrieb Martin Koppenhoefer:
> Am Mi., 11. März 2020 um 17:21 Uhr schrieb Simon Poole  >:
>
> I would be very very wary of doing anything that deliberately
> defaces a web site without consulting with a local (to the country
> the web site is in) lawyer, particularly if the message implies
> wrong doing."
>
> As I am not a lawyer in any country that a website could be
> displayed in, I'm really the wrong person to ask.
>
>
> So with "local lawyer" you were aiming at the country where the
> website could be displayed in? From my understanding, these local laws
> of the enduser only might matter for the service provider who uses the
> map tiles, while for the service that provides the map tiles (assuming
> reasonable ToS) their own local law would be the only relevant,
> especially when we are talking about B2B?

The "web site is in" should have been "the country the
business/organisation has its place of business in", as clarification,
but not excluding countries where they might not have a domicile but do
business in.

So say you scribble on a German companies website, if they are feeling
in the mood, you could be sued at least based on "Recht am
eingerichteten und ausgeübten Gewerbebetrieb" for damages (and naturally
to cease doing it)*. I suspect that you would not be able to reflect
this with terms in your ToUs, but as said you need to consult with
counsel in the know to be ale to asses the ricks of that happening.

Simon

*just in case you believe I'm just making things up: I was sued two
decades ago as a CEO of company on that base just for stating in a press
release that we intended to start providing services in Germany in the
foreseeable future and having a website that was accessible there. The
underlying problem was a  name and trademark conflict with a German
company of the same name (different area of business though). The net
result after multiple 10'000 of Euros of court, damages and lawyer
costs, not to mention renaming the company, all products and so on, is
that I still have a multiple 100'000 Euro per infringement injunction
against me personally in that matter. And that was a case in which we
were not clearly doing something that was wrong,  very different than
what we are discussing here.



signature.asc
Description: OpenPGP digital signature
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] #AttributionIsNotOptional experiment on OSM France tile servers

[Martin Koppenhoefer]

Am Mi., 11. März 2020 um 17:21 Uhr schrieb Simon Poole :

> I would be very very wary of doing anything that deliberately defaces a
> web site without consulting with a local (to the country the web site is
> in) lawyer, particularly if the message implies wrong doing."
>
> As I am not a lawyer in any country that a website could be displayed in,
> I'm really the wrong person to ask.
>

So with "local lawyer" you were aiming at the country where the website
could be displayed in? From my understanding, these local laws of the
enduser only might matter for the service provider who uses the map tiles,
while for the service that provides the map tiles (assuming reasonable ToS)
their own local law would be the only relevant, especially when we are
talking about B2B?

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

Re: [OSM-talk] #AttributionIsNotOptional experiment on OSM France tile servers

[Marc M.]

Hello,

Simon Poole wrote
> it is clear that using a neutral message and clear ToUs, 
> is definitely less risky than an aggressive > message without ToUs.

I fully agree with that. so I suggest that we include in the ToU :
- the "caller" may include a means of contact in its call to the tiles (
parameter like ?email=xxx in the url or http header to define).
if this one is defined, the supplier will be able to (and MAY) report
the anomaly and in an automatic way (because for the moment, it is the
volunteer who must consult the site, find the means of contact,
sometimes register on the site in order to report the violation).
Of course we may also use whois contact info.
- add in the ToS that the supplier reserves the right to add overlay
text and/or to replace and/or block tiles access, in the event of
non-compliance with the ToU.
 - add also a disclaimer stating that the provider does its best to
detect the lack of attribution but that the service is provided free of
charge, an error at this level cannot lead to a compensation, the user
being free to produce the tiles himself (with switch2som link)

this in no way prevents the experiment in progress,
on the contrary, it would increase its legitimacy.

Regards,
Marc

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

Re: [OSM-talk] #AttributionIsNotOptional experiment on OSM France tile servers

[Simon Poole]

Well the other bit that I wrote was:

"Enforcing attribution for services that you are providing directly (aka
tiles in some form) only has a small overlap with the goals of the
attribution guideline, and the avenues open to you depend on your ToUs /
contracts with your users and the legal situation in the countries you
are providing the service in.

I would be very very wary of doing anything that deliberately defaces a
web site without consulting with a local (to the country the web site is
in) lawyer, particularly if the message implies wrong doing."

As I am not a lawyer in any country that a website could be displayed
in, I'm really the wrong person to ask. Things vary widely by country,
that is in particular: inclination and costs to sue and protection
afforded to buisiness undertakings. But it is clear that using a neutral
message and clear ToUs, is definitely less risky than an aggressive
message without ToUs.

Things that need to be considered:

- any message you display will be shown to customers / users of the site
in question and needs to toned down enough so that you don't cause
unwarranted damage to the reputation of the sites operators. It is
likely that if you get sued that the range of options for action
available to you will be considered, you will need to show that what you
did was appropriate.

- everybody makes mistakes, so you -will- miss existing attribution.
Anything you do needs to be able to be undone with a simple "sorry it
was a mistake", except if you have deep pockets.

I'm very aware that getting peoples attention is sometimes difficult,
from experience things that work: sending a fax, registered mail, phone
calls,worst case publicly messaging on social media.

Simon


Am 11.03.2020 um 16:17 schrieb joost schouppe:
> Hi Simon,
>
> In a volunteer community, fun things are more likely to happen at all.
> So I do think the idea is worth exploring, even if the current
> implementation might be risky to OSMfr or to OSMF if implemented
> without much further thought.
>
> I would personally be interested in a more in depth analysis from you.
> I personally don't see how a more neutral message ("This map is based
> on OpenStreetMap data. Please add the required attribution to your
> website. Contact us at X if you need help.") would be more defacing or
> likely to lead to a liability claim than just a blacked out map, but I
> would not mind at all to be enlightened.
>
> Joost
>
> Op wo 11 mrt. 2020 om 15:39 schreef Simon Poole  >:
>
> As I wrote (conveniently ignored in the noise of the vigilante
> rampage): "The safe, I admit also the less fun, option, is to
> simply block access after giving any required notice."
>
> Simon
>
> Am 11.03.2020 um 14:49 schrieb joost schouppe:
>> Simon,
>>
>> I guess with small overlap you mean it's only about people who
>> use osm.org  tiles, not people who use other
>> services?
>> While that is true, the double whammy of both heavily using
>> resources and also not attributing does seem like a good subgroup
>> to start with some measures.
>>
>> In the case of the OSM.org tiles, I suppose this is regulated by
>> https://wiki.osmfoundation.org/wiki/Terms_of_Use . At first
>> glance I didn't see anything providing people who do not respect
>> those terms. Am I missing something, or is this a naive approach
>> to the problem?
>> Even if the ToU's could be lacking in detail, couldn't we simply
>> change them? The final section talks about changes, which we seem
>> to be able to just do when we want to.
>>
>> I would think the biggest challenge on OSMF side would be the
>> workload for OWG/sysadmins.
>>
>> Best,
>> Joost
>>
>> Op zo 8 mrt. 2020 om 12:18 schreef Simon Poole > >:
>>
>> Just for the record:
>>
>> Enforcing attribution for services that you are providing
>> directly (aka tiles in some form) only has a small overlap
>> with the goals of the attribution guideline, and the avenues
>> open to you depend on your ToUs / contracts with your users
>> and the legal situation in the countries you are providing
>> the service in.
>>
>> I would be very very wary of doing anything that deliberately
>> defaces a web site without consulting with a local (to the
>> country the web site is in) lawyer, particularly if the
>> message implies wrong doing. The safe, I admit also the less
>> fun, option, is to simply block access after giving any
>> required notice.
>>
>> Simon 
>>
>> Am 08.03.2020 um 11:04 schrieb Yves:
>>> This looks at first as a nuisance that could be perceived as
>>> a bad move, but the feedback you're receiving rather prove
>>> the contrary.
>>> Well done!
>>> Ps: would you share your nginx partial redirect, I may
>>>

Re: [OSM-talk] #AttributionIsNotOptional experiment on OSM France tile servers

[Yves]

If the terms of use is sufficiently clear, and a couple of warnings have been 
issued beforehand, then what can possibly happen? 
This is a rhetorical question. 
Yves 

Le 11 mars 2020 16:10:55 GMT+01:00, Simon Poole  a écrit :
>
>Am 11.03.2020 um 15:48 schrieb Mateusz Konieczny via talk:
>> Mar 11, 2020, 15:37 by si...@poole.ch:
>>
>> As I wrote (conveniently ignored in the noise of the vigilante
>> rampage): "
>>
>> I guess that people were irritated by describing gentle reminder
>about
>> license requirements
>> using pejorative terms ("deface") where their applicability was
>dubious.
>Sorry, but that is exactly the appropriate term, and anything that will
>inevitably get you on the wrong end of being sued if you do it often
>enough, is not a "gentle reminder".
>>
>> The safe, I admit also the less fun, option, is to simply block
>> access after giving any required notice."
>>
>> And note that in case of OSMF-served tiles no notice is required.
>>
>I'm not sure why you feel it necessary to tell me about terms that I'm
>completely aware off (and in some cases that I actually co-wrote), the
>subject matter in this thread is not -just- about the OSMF operated
>servers, but of those of OSM-FR and others.
>
>Simon
>
>> See https://operations.osmfoundation.org/policies/tiles/
>>
>> "Clearly display license attribution." is in explicit requirements,
>> and given overloaded servers
>>
>> "should any users or patterns of usage nevertheless cause problems to
>> the service, access may still be blocked without prior notice."
>>
>> applies anyway.
>>
>> "access may be withdrawn at any point" is later repeated.
>>
>>
>> ___
>> 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] #AttributionIsNotOptional experiment on OSM France tile servers

[Mario Frasca]

On 11/03/2020 09:37, Simon Poole wrote:


As I wrote (conveniently ignored in the noise of the vigilante 
rampage): "The safe, I admit also the less fun, option, is to simply 
block access after giving any required notice."



Hi Simon.  let me provide you my individual reason for ignoring your option.

I fear that blocking access will push people to other services. whatever 
you say about costs and resources, favouring Mr.G. can never be an 
acceptable outcome.


I also think that "giving any required notice" needs an acknowledgement, 
like others have remarked: you write an email to the web master, and if 
anybody ever receives it, it's not the relevant people.


this French experiment is a very nice one, it has proved effective, and 
I invite the author to share more of their positive experience.  may I 
suggest identifying metrics, estimate them from the French experiment, 
and use them to adapt and repeat the experience from a different 
server.  that is, if anybody is noticing any similar abuse.  maybe the 
French were in some sort of special situation.


in the altered experiment, you may even include the formal approach: 
"write to web master, wait one week without reply, write again, no 
reply, block the service, check how long it takes them to show Google Maps".


Mario


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

Re: [OSM-talk] #AttributionIsNotOptional experiment on OSM France tile servers

[Rory McCann]

Hi Christian,

That's a interesting approach, and interesting to see that it gets results.

If you think it's too rude, or aggressive, you should remember that 
polite words sometimes don't work. If you don't like people shouting at 
you, then it's your fault for ignoring them when they speak calmly. 
Don't complain when this is the only resort that works.


Most of the complaints about missing OSM attribution don't come from 
people using osm.org tile servers. So I'm not sure if doing it on 
OSM.org is very useful. Never the less, there might be benefit in OSM 
taking the next level of action for our attribution. Concerns about 
libelling/defaming someone are legitimate. We don't want to cause much 
more headaches for ourselves.


Rory

On 08/03/2020 10:14, Christian Quest wrote:

Here is a hort report on this experiment...

I started a week ago by searching OSM France tile server logs for 
referer and checked manually if the map on the refering page was 
correctly attributed.


This allowed me to create a short list of 20 entries of sites using the 
french styled tiles and the humanitarian tiles (yes, it is made by OSM 
France).



I then modified our nginx based proxy_cache configuration, to redirect 
some tiles to an "attribution tile" only for the domain in the list.


For two of them, I tweeted about it... the most visible one is the 
moroco yellow page service, generating a little less than a million 
daily tile requests on our servers.


https://twitter.com/cq94/status/1234516075695525888

In less than 24 hours, the attribution appeared and I removed them from 
the list.


https://twitter.com/cq94/status/1234779931537739776


Then I included an email address in the attribution reminder tile... and 
got emails back within a few hours.


Some were asking how to do the attribution, others telling me the 
attribution was now ok and asking how to remove the reminder tiles.


In my answers, I also remind that our tile service made by volunteers on 
donated hardware is not unlimited and inviting them to have a look at 
switch2osm to setup their own tile server or use a commercial provider.


Up to now, nobody complained :)


Yesterday, I've started automating attribution checking using selenium. 
For each referer, a python script loads the page, searches for tiles, 
then looks for attribution text or link. The result is stored in a 
postgresql database which allows to group referers by url, hostname and ip.


The attribution percentage I currently see is around 70-80% which is not 
that bad.


My next major step is to use the same technique to remind about tile 
usage policy...



To do something similar on osm.org, a first step is to extract referers 
from the cache logs, then use the automated attribution check to 
evaluate the situation.



Le 08/03/2020 à 01:52, Nuno Caldeira a écrit :
That would be a good option for those that use third party providers 
of OSM. But to be honest, from my experience I highly doubt that even 
corporate members of OSMF, like Mapbox would do it, when their client 
Facebook (also corporate member of OSMF) after one year and half, 
still has maps with lack of attribution or attributed to HERE, when 
it's clearly OSM.


On Sun, 8 Mar 2020, 00:46 Phil Wyatt, > wrote:


I am sure others may have seen this 'blacklist' implementation for
showing a reminder about attribution.

https://twitter.com/cq94/status/1234528717604577282

Worthy of consideration for openstreetmap.org
?

Cheers - Phil


--
Christian Quest - OpenStreetMap France


___
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] #AttributionIsNotOptional experiment on OSM France tile servers

[joost schouppe]

Hi Simon,

In a volunteer community, fun things are more likely to happen at all. So I
do think the idea is worth exploring, even if the current implementation
might be risky to OSMfr or to OSMF if implemented without much further
thought.

I would personally be interested in a more in depth analysis from you. I
personally don't see how a more neutral message ("This map is based on
OpenStreetMap data. Please add the required attribution to your website.
Contact us at X if you need help.") would be more defacing or likely to
lead to a liability claim than just a blacked out map, but I would not mind
at all to be enlightened.

Joost

Op wo 11 mrt. 2020 om 15:39 schreef Simon Poole :

> As I wrote (conveniently ignored in the noise of the vigilante rampage):
> "The safe, I admit also the less fun, option, is to simply block access
> after giving any required notice."
>
> Simon
> Am 11.03.2020 um 14:49 schrieb joost schouppe:
>
> Simon,
>
> I guess with small overlap you mean it's only about people who use osm.org
> tiles, not people who use other services?
> While that is true, the double whammy of both heavily using resources and
> also not attributing does seem like a good subgroup to start with some
> measures.
>
> In the case of the OSM.org tiles, I suppose this is regulated by
> https://wiki.osmfoundation.org/wiki/Terms_of_Use . At first glance I
> didn't see anything providing people who do not respect those terms. Am I
> missing something, or is this a naive approach to the problem?
> Even if the ToU's could be lacking in detail, couldn't we simply change
> them? The final section talks about changes, which we seem to be able to
> just do when we want to.
>
> I would think the biggest challenge on OSMF side would be the workload for
> OWG/sysadmins.
>
> Best,
> Joost
>
> Op zo 8 mrt. 2020 om 12:18 schreef Simon Poole :
>
>> Just for the record:
>>
>> Enforcing attribution for services that you are providing directly (aka
>> tiles in some form) only has a small overlap with the goals of the
>> attribution guideline, and the avenues open to you depend on your ToUs /
>> contracts with your users and the legal situation in the countries you are
>> providing the service in.
>>
>> I would be very very wary of doing anything that deliberately defaces a
>> web site without consulting with a local (to the country the web site is
>> in) lawyer, particularly if the message implies wrong doing. The safe, I
>> admit also the less fun, option, is to simply block access after giving any
>> required notice.
>>
>> Simon
>> Am 08.03.2020 um 11:04 schrieb Yves:
>>
>> This looks at first as a nuisance that could be perceived as a bad move,
>> but the feedback you're receiving rather prove the contrary.
>> Well done!
>> Ps: would you share your nginx partial redirect, I may consider it for
>> Opensnowmap tiles policy?
>>
>> Le 8 mars 2020 10:14:58 GMT+01:00, Christian Quest
>>   a écrit :
>>>
>>> Here is a hort report on this experiment...
>>>
>>> I started a week ago by searching OSM France tile server logs for
>>> referer and checked manually if the map on the refering page was correctly
>>> attributed.
>>>
>>> This allowed me to create a short list of 20 entries of sites using the
>>> french styled tiles and the humanitarian tiles (yes, it is made by OSM
>>> France).
>>>
>>>
>>> I then modified our nginx based proxy_cache configuration, to redirect
>>> some tiles to an "attribution tile" only for the domain in the list.
>>>
>>> For two of them, I tweeted about it... the most visible one is the
>>> moroco yellow page service, generating a little less than a million daily
>>> tile requests on our servers.
>>>
>>> https://twitter.com/cq94/status/1234516075695525888
>>>
>>> In less than 24 hours, the attribution appeared and I removed them from
>>> the list.
>>>
>>> https://twitter.com/cq94/status/1234779931537739776
>>>
>>>
>>> Then I included an email address in the attribution reminder tile... and
>>> got emails back within a few hours.
>>>
>>> Some were asking how to do the attribution, others telling me the
>>> attribution was now ok and asking how to remove the reminder tiles.
>>>
>>> In my answers, I also remind that our tile service made by volunteers on
>>> donated hardware is not unlimited and inviting them to have a look at
>>> switch2osm to setup their own tile server or use a commercial provider.
>>>
>>> Up to now, nobody complained :)
>>>
>>>
>>> Yesterday, I've started automating attribution checking using selenium.
>>> For each referer, a python script loads the page, searches for tiles, then
>>> looks for attribution text or link. The result is stored in a postgresql
>>> database which allows to group referers by url, hostname and ip.
>>>
>>> The attribution percentage I currently see is around 70-80% which is not
>>> that bad.
>>>
>>> My next major step is to use the same technique to remind about tile
>>> usage policy...
>>>
>>>
>>> To do something similar on osm.org, a first step is to extract referers

Re: [OSM-talk] #AttributionIsNotOptional experiment on OSM France tile servers

[Simon Poole]

Am 11.03.2020 um 15:48 schrieb Mateusz Konieczny via talk:
> Mar 11, 2020, 15:37 by si...@poole.ch:
>
> As I wrote (conveniently ignored in the noise of the vigilante
> rampage): "
>
> I guess that people were irritated by describing gentle reminder about
> license requirements
> using pejorative terms ("deface") where their applicability was dubious.
Sorry, but that is exactly the appropriate term, and anything that will
inevitably get you on the wrong end of being sued if you do it often
enough, is not a "gentle reminder".
>
> The safe, I admit also the less fun, option, is to simply block
> access after giving any required notice."
>
> And note that in case of OSMF-served tiles no notice is required.
>
I'm not sure why you feel it necessary to tell me about terms that I'm
completely aware off (and in some cases that I actually co-wrote), the
subject matter in this thread is not -just- about the OSMF operated
servers, but of those of OSM-FR and others.

Simon

> See https://operations.osmfoundation.org/policies/tiles/
>
> "Clearly display license attribution." is in explicit requirements,
> and given overloaded servers
>
> "should any users or patterns of usage nevertheless cause problems to
> the service, access may still be blocked without prior notice."
>
> applies anyway.
>
> "access may be withdrawn at any point" is later repeated.
>
>
> ___
> talk mailing list
> talk@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/talk


signature.asc
Description: OpenPGP digital signature
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] #AttributionIsNotOptional experiment on OSM France tile servers

[Mateusz Konieczny via talk]

Mar 11, 2020, 15:37 by si...@poole.ch:

>
> As I wrote (conveniently ignored in the noise of the vigilante  rampage): 
> "
>
>
I guess that people were irritated by describing gentle reminder about license 
requirements
using pejorative terms ("deface") where their applicability was dubious.

> The safe, I admit also the less fun, option, is to  simply block access 
> after giving any required notice."
>
And note that in case of OSMF-served tiles no notice is required.

See https://operations.osmfoundation.org/policies/tiles/

"Clearly display license attribution." is in explicit requirements,
and given overloaded servers

"should any users or patterns of usage nevertheless cause problems to
the service, access may still be blocked without prior notice."

applies anyway.

"access may be withdrawn at any point" is later repeated.

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

Re: [OSM-talk] #AttributionIsNotOptional experiment on OSM France tile servers

[Simon Poole]

As I wrote (conveniently ignored in the noise of the vigilante rampage):
"The safe, I admit also the less fun, option, is to simply block access
after giving any required notice."

Simon

Am 11.03.2020 um 14:49 schrieb joost schouppe:
> Simon,
>
> I guess with small overlap you mean it's only about people who use
> osm.org  tiles, not people who use other services?
> While that is true, the double whammy of both heavily using resources
> and also not attributing does seem like a good subgroup to start with
> some measures.
>
> In the case of the OSM.org tiles, I suppose this is regulated by
> https://wiki.osmfoundation.org/wiki/Terms_of_Use . At first glance I
> didn't see anything providing people who do not respect those terms.
> Am I missing something, or is this a naive approach to the problem?
> Even if the ToU's could be lacking in detail, couldn't we simply
> change them? The final section talks about changes, which we seem to
> be able to just do when we want to.
>
> I would think the biggest challenge on OSMF side would be the workload
> for OWG/sysadmins.
>
> Best,
> Joost
>
> Op zo 8 mrt. 2020 om 12:18 schreef Simon Poole  >:
>
> Just for the record:
>
> Enforcing attribution for services that you are providing directly
> (aka tiles in some form) only has a small overlap with the goals
> of the attribution guideline, and the avenues open to you depend
> on your ToUs / contracts with your users and the legal situation
> in the countries you are providing the service in.
>
> I would be very very wary of doing anything that deliberately
> defaces a web site without consulting with a local (to the country
> the web site is in) lawyer, particularly if the message implies
> wrong doing. The safe, I admit also the less fun, option, is to
> simply block access after giving any required notice.
>
> Simon 
>
> Am 08.03.2020 um 11:04 schrieb Yves:
>> This looks at first as a nuisance that could be perceived as a
>> bad move, but the feedback you're receiving rather prove the
>> contrary.
>> Well done!
>> Ps: would you share your nginx partial redirect, I may consider
>> it for Opensnowmap tiles policy?
>>
>> Le 8 mars 2020 10:14:58 GMT+01:00, Christian Quest
>>   a écrit :
>>
>> Here is a hort report on this experiment...
>>
>> I started a week ago by searching OSM France tile server logs
>> for referer and checked manually if the map on the refering
>> page was correctly attributed.
>>
>> This allowed me to create a short list of 20 entries of sites
>> using the french styled tiles and the humanitarian tiles
>> (yes, it is made by OSM France).
>>
>>
>> I then modified our nginx based proxy_cache configuration, to
>> redirect some tiles to an "attribution tile" only for the
>> domain in the list.
>>
>> For two of them, I tweeted about it... the most visible one
>> is the moroco yellow page service, generating a little less
>> than a million daily tile requests on our servers.
>>
>> https://twitter.com/cq94/status/1234516075695525888
>>
>> In less than 24 hours, the attribution appeared and I removed
>> them from the list.
>>
>> https://twitter.com/cq94/status/1234779931537739776
>>
>>
>> Then I included an email address in the attribution reminder
>> tile... and got emails back within a few hours.
>>
>> Some were asking how to do the attribution, others telling me
>> the attribution was now ok and asking how to remove the
>> reminder tiles.
>>
>> In my answers, I also remind that our tile service made by
>> volunteers on donated hardware is not unlimited and inviting
>> them to have a look at switch2osm to setup their own tile
>> server or use a commercial provider.
>>
>> Up to now, nobody complained :)
>>
>>
>> Yesterday, I've started automating attribution checking using
>> selenium. For each referer, a python script loads the page,
>> searches for tiles, then looks for attribution text or link.
>> The result is stored in a postgresql database which allows to
>> group referers by url, hostname and ip.
>>
>> The attribution percentage I currently see is around 70-80%
>> which is not that bad.
>>
>> My next major step is to use the same technique to remind
>> about tile usage policy...
>>
>>
>> To do something similar on osm.org , a first
>> step is to extract referers from the cache logs, then use the
>> automated attribution check to evaluate the situation.
>>
>>
>> Le 08/03/2020 à 01:52, Nuno Caldeira a écrit :
>>> That would be a good option for those that use third party
>>> providers of OSM. But to be 

Re: [OSM-talk] #AttributionIsNotOptional experiment on OSM France tile servers

[joost schouppe]

Simon,

I guess with small overlap you mean it's only about people who use osm.org
tiles, not people who use other services?
While that is true, the double whammy of both heavily using resources and
also not attributing does seem like a good subgroup to start with some
measures.

In the case of the OSM.org tiles, I suppose this is regulated by
https://wiki.osmfoundation.org/wiki/Terms_of_Use . At first glance I didn't
see anything providing people who do not respect those terms. Am I missing
something, or is this a naive approach to the problem?
Even if the ToU's could be lacking in detail, couldn't we simply change
them? The final section talks about changes, which we seem to be able to
just do when we want to.

I would think the biggest challenge on OSMF side would be the workload for
OWG/sysadmins.

Best,
Joost

Op zo 8 mrt. 2020 om 12:18 schreef Simon Poole :

> Just for the record:
>
> Enforcing attribution for services that you are providing directly (aka
> tiles in some form) only has a small overlap with the goals of the
> attribution guideline, and the avenues open to you depend on your ToUs /
> contracts with your users and the legal situation in the countries you are
> providing the service in.
>
> I would be very very wary of doing anything that deliberately defaces a
> web site without consulting with a local (to the country the web site is
> in) lawyer, particularly if the message implies wrong doing. The safe, I
> admit also the less fun, option, is to simply block access after giving any
> required notice.
>
> Simon
> Am 08.03.2020 um 11:04 schrieb Yves:
>
> This looks at first as a nuisance that could be perceived as a bad move,
> but the feedback you're receiving rather prove the contrary.
> Well done!
> Ps: would you share your nginx partial redirect, I may consider it for
> Opensnowmap tiles policy?
>
> Le 8 mars 2020 10:14:58 GMT+01:00, Christian Quest
>   a écrit :
>>
>> Here is a hort report on this experiment...
>>
>> I started a week ago by searching OSM France tile server logs for referer
>> and checked manually if the map on the refering page was correctly
>> attributed.
>>
>> This allowed me to create a short list of 20 entries of sites using the
>> french styled tiles and the humanitarian tiles (yes, it is made by OSM
>> France).
>>
>>
>> I then modified our nginx based proxy_cache configuration, to redirect
>> some tiles to an "attribution tile" only for the domain in the list.
>>
>> For two of them, I tweeted about it... the most visible one is the moroco
>> yellow page service, generating a little less than a million daily tile
>> requests on our servers.
>>
>> https://twitter.com/cq94/status/1234516075695525888
>>
>> In less than 24 hours, the attribution appeared and I removed them from
>> the list.
>>
>> https://twitter.com/cq94/status/1234779931537739776
>>
>>
>> Then I included an email address in the attribution reminder tile... and
>> got emails back within a few hours.
>>
>> Some were asking how to do the attribution, others telling me the
>> attribution was now ok and asking how to remove the reminder tiles.
>>
>> In my answers, I also remind that our tile service made by volunteers on
>> donated hardware is not unlimited and inviting them to have a look at
>> switch2osm to setup their own tile server or use a commercial provider.
>>
>> Up to now, nobody complained :)
>>
>>
>> Yesterday, I've started automating attribution checking using selenium.
>> For each referer, a python script loads the page, searches for tiles, then
>> looks for attribution text or link. The result is stored in a postgresql
>> database which allows to group referers by url, hostname and ip.
>>
>> The attribution percentage I currently see is around 70-80% which is not
>> that bad.
>>
>> My next major step is to use the same technique to remind about tile
>> usage policy...
>>
>>
>> To do something similar on osm.org, a first step is to extract referers
>> from the cache logs, then use the automated attribution check to evaluate
>> the situation.
>>
>>
>> Le 08/03/2020 à 01:52, Nuno Caldeira a écrit :
>>
>> That would be a good option for those that use third party providers of
>> OSM. But to be honest, from my experience I highly doubt that even
>> corporate members of OSMF, like Mapbox would do it, when their client
>> Facebook (also corporate member of OSMF) after one year and half, still has
>> maps with lack of attribution or attributed to HERE, when it's clearly OSM.
>>
>> On Sun, 8 Mar 2020, 00:46 Phil Wyatt,  wrote:
>>
>>> I am sure others may have seen this 'blacklist' implementation for
>>> showing a reminder about attribution.
>>>
>>> https://twitter.com/cq94/status/1234528717604577282
>>>
>>> Worthy of consideration for openstreetmap.org?
>>>
>>> Cheers - Phil
>>
>> --
>> Christian Quest - OpenStreetMap France
>>
>>
> ___
> talk mailing 
> listtalk@openstreetmap.orghttps://lists.openstreetmap.org/listinfo/talk
>
> 

Re: [OSM-talk] #AttributionIsNotOptional experiment on OSM France tile servers

[Mario Frasca]

On 08/03/2020 15:15, Pierre Béland via talk wrote:

semi transparent


the problem here is technical.

I'm rewording here something that was said to me here today:

I don't think you can reply with a composite object (tile + 
transparency) to a request for a single image object.  for this, you 
need to either compute the tile on the fly, for example using 
imagemagick, or you need to keep a parallel tiles repository, with the 
transparency on top of the original tile.


technical, and economical, in terms of server space and the human action 
required.




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

Re: [OSM-talk] #AttributionIsNotOptional experiment on OSM France tile servers

[Pierre Béland via talk]

 
Pierre 
 

   Mar.8, 2020 1byMateusz Konieczny 

> To be more clear: I fully support action like OSM France to actually enforce
> license requirements using methods like described here or by a legal action 
> like
> DMCA takedown notices against entities refusing to show a proper attribution.
  I also support such actions. To be less intrusive and more diplomatic, the 
box could be smaller and semi transparent, and the text rewrittten to be more 
specific What's about something like ? 

- WARNING : OSM Attribution is required for these tile products
- These OSM tiles are downloaded from openstreetmap-France tile server 
   without this website providing any visible attribution
- See https://www.openstreetmap.org/copyright
- Contact us at ti...@openstreetmap.fr

Mar 8, 2020, 20:26 by talk@openstreetmap.org:


Mar 8, 2020, 12:12 by si...@poole.ch:


I would be very very wary of doing anything that deliberately defaces a web 
site without consulting with a local (to the country the web site is in) 
lawyer, particularly if the message implies wrong doing.


Illegal use of OSM data and violating terms of use of service is a clear wrong 
doing.

I am not a lawyer, but showing message informing about violating license and 
terms of use of service seems 100% OK in case of website actually violating
OSM license.
And I fully support websites doing this.

To be more clear: I fully support action like OSM France to actually enforce
license requirements using methods like described here or by a legal action like
DMCA takedown notices against entities refusing to show a proper attribution.
 ___
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] #AttributionIsNotOptional experiment on OSM France tile servers

[Nuno Caldeira]

Hi highly doubt that's even "defacing" a website. Google does it
https://stackoverflow.com/questions/50977913/google-maps-shows-for-development-purposes-only
Attribute or leave it.

I too applause OSM FR attitude towards these license infringement. Seems,
judging by the previous board denial to cease Facebook rights under ODbL
that I requested last year, OSM FR is actually taking action instead of
delaying action like OSMF does (even towards their corporate members to
which we should be the firsts to show support to OSM by proudly
attributing).

On Sun, 8 Mar 2020, 19:28 Mateusz Konieczny via talk, <
talk@openstreetmap.org> wrote:

>
> Mar 8, 2020, 12:12 by si...@poole.ch:
>
> I would be very very wary of doing anything that deliberately defaces a
> web site without consulting with a local (to the country the web site is
> in) lawyer, particularly if the message implies wrong doing.
>
> Illegal use of OSM data and violating terms of use of service is a clear
> wrong doing.
>
> I am not a lawyer, but showing message informing about violating license
> and
> terms of use of service seems 100% OK in case of website actually violating
> OSM license.
> And I fully support websites doing this.
>
> Is there anything that would actually make it illegal, unethical or wrong
> in any way?
> If yes I would be happy to learn about it.
>
> It is not like they were serving tile images with lies or untrue claims.
> Or actually defacing
> website by serving tile images with shock content like gore, nudity or
> extreme statements.
> And I would not support doing this.
> ___
> 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] #AttributionIsNotOptional experiment on OSM France tile servers

[Mateusz Konieczny via talk]

Mar 8, 2020, 16:42 by ma...@anche.no:

>
> not to deface any web site relying on OSM.  or we  would put the 
> wrongdoing on our side.
>
>
Enforcing attribution requirements is not wrongdoing.

It is something that we should do more, and this specific method is very 
friendly.
Map is mostly functional and they have chance to fix the problem
and get back their free tiles fully functional.

And site relying on OSM is obligated to attribute OSM.

OSM France is not obligated to provide free tile servers,
especially to people violating OSM license.

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

Re: [OSM-talk] #AttributionIsNotOptional experiment on OSM France tile servers

[Mateusz Konieczny via talk]

Mar 8, 2020, 20:26 by talk@openstreetmap.org:

>
> Mar 8, 2020, 12:12 by si...@poole.ch:
>
>>
>> I would be very very wary of doing anything that deliberately  defaces a 
>> web site without consulting with a local (to the country  the web site 
>> is in) lawyer, particularly if the message implies  wrong doing.
>>
>>
> Illegal use of OSM data and violating terms of use of service is a clear 
> wrong doing.
>
> I am not a lawyer, but showing message informing about violating license and 
> terms of use of service seems 100% OK in case of website actually violating
> OSM license.
> And I fully support websites doing this.
>
To be more clear: I fully support action like OSM France to actually enforce
license requirements using methods like described here or by a legal action like
DMCA takedown notices against entities refusing to show a proper attribution.
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] #AttributionIsNotOptional experiment on OSM France tile servers

[Mateusz Konieczny via talk]

Mar 8, 2020, 12:12 by si...@poole.ch:

>
> I would be very very wary of doing anything that deliberately  defaces a 
> web site without consulting with a local (to the country  the web site is 
> in) lawyer, particularly if the message implies  wrong doing.
>
>
Illegal use of OSM data and violating terms of use of service is a clear wrong 
doing.

I am not a lawyer, but showing message informing about violating license and 
terms of use of service seems 100% OK in case of website actually violating
OSM license.
And I fully support websites doing this.

Is there anything that would actually make it illegal, unethical or wrong in 
any way?
If yes I would be happy to learn about it.

It is not like they were serving tile images with lies or untrue claims. Or 
actually defacing
website by serving tile images with shock content like gore, nudity or extreme 
statements.
And I would not support doing this.   
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] #AttributionIsNotOptional experiment on OSM France tile servers

[Greg Troxel]

Mario Frasca  writes:

> I would say that 1 in 25 is low enough as not to be considered
> "defacing" a web site.  what text have you used, concretely, which had
> the impact you describe?  in my opinion the shortest, the better, and
> I guess you did NOT use »It looks like this site forgot to put the
> required attribution in this map corner, so we added it for them. Thx
> for using OpenStreetMap !«, did you?  it was in French, wasn't it?

Also, I have seen on many sites google maps with an "API Key Required"
watermark.  This really seems like a similar situation.

So perhaps Christian could put a fairly loud attribution watermark over
the tiles, so it's basically functional but attributed.

Overall, I think this is a great experiment, and the combination of
failure to attribute and use of donated tile services is particularly
worthy of addressing.

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

Re: [OSM-talk] #AttributionIsNotOptional experiment on OSM France tile servers

[Mario Frasca]

On 08/03/2020 04:14, Christian Quest wrote:
the most visible one is the moroco yellow page service, generating a 
little less than a million daily tile requests on our servers.

On 08/03/2020 10:42, Mario Frasca wrote:

what further steps does your team plan?


for example with the above heavy user?

just an attribution, that's still not enough, is it?

Mario


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

Re: [OSM-talk] #AttributionIsNotOptional experiment on OSM France tile servers

[Pierre Béland via talk]

You could use the navigator user language preference for the language to use 
for the message.

 
Pierre 
 

Le dimanche 8 mars 2020 11 h 40 min 47 s UTC−4, Christian Quest 
 a écrit :  
 
 Le 08/03/2020 à 16:00, Mario Frasca a écrit :
> well, it does look slightly invasive …


That's the goal... slightly invasive... and not time consuming for the 
tile server and myself.

Please, don't forget who's wrong here.


-- 
Christian Quest - OpenStreetMap France


___
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] #AttributionIsNotOptional experiment on OSM France tile servers

[Mario Frasca]

Hi Michael and Christian,

I don't know if there's any other way than going through imagemagick, if 
that's the only way, then sure it would cause extra work on (y)our 
servers.  if that's too much, let the server manager measure and decide.


my approach was trying to find a solution to the problem raised by Simon 
Poole, not to deface any web site relying on OSM.  or we would put the 
wrongdoing on our side.


but then, that was what just »I would probably do« … maybe after half an 
hour spent on trying to do it this way, I would fall back to Christian's 
way.


"whatever works"

btw, Christian, you are measuring the effects of your action, and the 
timings.  you started with 20 "abusers", you got immediate effect on 
some, then you added the email address in the tile, and got more 
results, how far are you now?  and what further steps does your team plan?


ciao,

Mario

On 08/03/2020 10:11, Michael Reichert wrote:

Hi Mario,

Am 08/03/2020 um 16.00 schrieb Mario Frasca:

well, it does look slightly invasive … I had imagined something like a
transparent text on top of the requested tile.  doing it the way you are
doing it, you are removing part of the underlying information.

A transparent overlay requires calling Imagemagick or a similar tool for
each tile and to cache the results. It makes the setup more complicated
and requires more processing power. Should we spend even more volunteer
time and donations on abusers? No.

Best regards

Michael


___
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] #AttributionIsNotOptional experiment on OSM France tile servers

[Christian Quest]

Le 08/03/2020 à 16:00, Mario Frasca a écrit :

well, it does look slightly invasive …



That's the goal... slightly invasive... and not time consuming for the 
tile server and myself.


Please, don't forget who's wrong here.


--
Christian Quest - OpenStreetMap France


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

Re: [OSM-talk] #AttributionIsNotOptional experiment on OSM France tile servers

[Michael Reichert]

Hi Mario,

Am 08/03/2020 um 16.00 schrieb Mario Frasca:
> well, it does look slightly invasive … I had imagined something like a
> transparent text on top of the requested tile.  doing it the way you are
> doing it, you are removing part of the underlying information.

A transparent overlay requires calling Imagemagick or a similar tool for
each tile and to cache the results. It makes the setup more complicated
and requires more processing power. Should we spend even more volunteer
time and donations on abusers? No.

Best regards

Michael



signature.asc
Description: OpenPGP digital signature
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] #AttributionIsNotOptional experiment on OSM France tile servers

[Mario Frasca]

well, it does look slightly invasive … I had imagined something like a 
transparent text on top of the requested tile.  doing it the way you are 
doing it, you are removing part of the underlying information.  it is 
true that the user can zoom in or out, and the chance that the same are 
is also hidden is, as you state, 4%, but still. … did you experiment 
with a transparency added on top of the requested tile?  that's how I 
would probably do it, and I would only use the lower text, not the top 
icon and bold text, using a smooth opaque mask to make sure the text 
stays well visible, separated from the background.


in practice, I would only add what they need to add themselves, and not 
remove anything.


On 08/03/2020 09:39, Christian Quest wrote:

Here are a few examples:

http://www.ardennes-neige.be/
http://autogas-network.co.uk/
http://mapa.guadalajara.gob.mx/basura
http://vivenda.hercesa.ro/
https://www.visitarnhem.com/routes/wandelroutes

The reminder tiles is in available english and french:

https://tilecache.openstreetmap.fr/attribution-en.png
https://tilecache.openstreetmap.fr/attribution-fr.png

I mix both at different locations on the maps.

Automatic attribution checking for one day of log files now takes 
around 20 minutes.


I'll clean the nginx config file and share it.


Le 08/03/2020 à 15:05, Mario Frasca a écrit :

On 08/03/2020 05:04, Yves wrote:
Ps: would you share your nginx partial redirect, I may consider it 
for Opensnowmap tiles policy? 

On 08/03/2020 06:12, Simon Poole wrote:

anything that deliberately defaces a web site

On 08/03/2020 07:13, Christian Quest wrote:

just 1 tile out of 25


very interesting experiment, and very amusing results.  bravo.

I would say that 1 in 25 is low enough as not to be considered 
"defacing" a web site.  what text have you used, concretely, which 
had the impact you describe?  in my opinion the shortest, the better, 
and I guess you did NOT use »It looks like this site forgot to put 
the required attribution in this map corner, so we added it for them. 
Thx for using OpenStreetMap !«, did you?  it was in French, wasn't it?


my best guess would be nothing else than the attribution text, and 
some help to solve their situation, or to get in contact with you.


I absolutely share the point of view "contact emails lead to no 
contact", and the "whatever works" policy.  this seem to work, so 
again, chapeau!


indeed, it would be interesting to see your nginx partial redirect.

MF


___
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] #AttributionIsNotOptional experiment on OSM France tile servers

[Christian Quest]

Here are a few examples:

http://www.ardennes-neige.be/
http://autogas-network.co.uk/
http://mapa.guadalajara.gob.mx/basura
http://vivenda.hercesa.ro/
https://www.visitarnhem.com/routes/wandelroutes

The reminder tiles is in available english and french:

https://tilecache.openstreetmap.fr/attribution-en.png
https://tilecache.openstreetmap.fr/attribution-fr.png

I mix both at different locations on the maps.

Automatic attribution checking for one day of log files now takes around 
20 minutes.


I'll clean the nginx config file and share it.


Le 08/03/2020 à 15:05, Mario Frasca a écrit :

On 08/03/2020 05:04, Yves wrote:
Ps: would you share your nginx partial redirect, I may consider it 
for Opensnowmap tiles policy? 

On 08/03/2020 06:12, Simon Poole wrote:

anything that deliberately defaces a web site

On 08/03/2020 07:13, Christian Quest wrote:

just 1 tile out of 25


very interesting experiment, and very amusing results.  bravo.

I would say that 1 in 25 is low enough as not to be considered 
"defacing" a web site.  what text have you used, concretely, which had 
the impact you describe?  in my opinion the shortest, the better, and 
I guess you did NOT use »It looks like this site forgot to put the 
required attribution in this map corner, so we added it for them. Thx 
for using OpenStreetMap !«, did you?  it was in French, wasn't it?


my best guess would be nothing else than the attribution text, and 
some help to solve their situation, or to get in contact with you.


I absolutely share the point of view "contact emails lead to no 
contact", and the "whatever works" policy.  this seem to work, so 
again, chapeau!


indeed, it would be interesting to see your nginx partial redirect.

MF


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


--
Christian Quest - OpenStreetMap France


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

Re: [OSM-talk] #AttributionIsNotOptional experiment on OSM France tile servers

[Mario Frasca]

On 08/03/2020 05:04, Yves wrote:
Ps: would you share your nginx partial redirect, I may consider it for 
Opensnowmap tiles policy? 

On 08/03/2020 06:12, Simon Poole wrote:

anything that deliberately defaces a web site

On 08/03/2020 07:13, Christian Quest wrote:

just 1 tile out of 25


very interesting experiment, and very amusing results.  bravo.

I would say that 1 in 25 is low enough as not to be considered 
"defacing" a web site.  what text have you used, concretely, which had 
the impact you describe?  in my opinion the shortest, the better, and I 
guess you did NOT use »It looks like this site forgot to put the 
required attribution in this map corner, so we added it for them. Thx 
for using OpenStreetMap !«, did you?  it was in French, wasn't it?


my best guess would be nothing else than the attribution text, and some 
help to solve their situation, or to get in contact with you.


I absolutely share the point of view "contact emails lead to no 
contact", and the "whatever works" policy.  this seem to work, so again, 
chapeau!


indeed, it would be interesting to see your nginx partial redirect.

MF


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

Re: [OSM-talk] #AttributionIsNotOptional experiment on OSM France tile servers

[Yves]

To me, the most important in attribution is to make potential contributors 
aware of the project. So the overlap is not that small in this regard. 

Yves 

Le 8 mars 2020 12:12:48 GMT+01:00, Simon Poole  a écrit :
>Just for the record:
>
>Enforcing attribution for services that you are providing directly (aka
>tiles in some form) only has a small overlap with the goals of the
>attribution guideline, and the avenues open to you depend on your ToUs
>/
>contracts with your users and the legal situation in the countries you
>are providing the service in.
>
>I would be very very wary of doing anything that deliberately defaces a
>web site without consulting with a local (to the country the web site
>is
>in) lawyer, particularly if the message implies wrong doing. The safe,
>I
>admit also the less fun, option, is to simply block access after giving
>any required notice.
>
>Simon 
>
>Am 08.03.2020 um 11:04 schrieb Yves:
>> This looks at first as a nuisance that could be perceived as a bad
>> move, but the feedback you're receiving rather prove the contrary.
>> Well done!
>> Ps: would you share your nginx partial redirect, I may consider it
>for
>> Opensnowmap tiles policy?
>>
>> Le 8 mars 2020 10:14:58 GMT+01:00, Christian Quest
>>  a écrit :
>>
>> Here is a hort report on this experiment...
>>
>> I started a week ago by searching OSM France tile server logs for
>> referer and checked manually if the map on the refering page was
>> correctly attributed.
>>
>> This allowed me to create a short list of 20 entries of sites
>> using the french styled tiles and the humanitarian tiles (yes, it
>> is made by OSM France).
>>
>>
>> I then modified our nginx based proxy_cache configuration, to
>> redirect some tiles to an "attribution tile" only for the domain
>> in the list.
>>
>> For two of them, I tweeted about it... the most visible one is
>the
>> moroco yellow page service, generating a little less than a
>> million daily tile requests on our servers.
>>
>> https://twitter.com/cq94/status/1234516075695525888
>>
>> In less than 24 hours, the attribution appeared and I removed
>them
>> from the list.
>>
>> https://twitter.com/cq94/status/1234779931537739776
>>
>>
>> Then I included an email address in the attribution reminder
>> tile... and got emails back within a few hours.
>>
>> Some were asking how to do the attribution, others telling me the
>> attribution was now ok and asking how to remove the reminder
>tiles.
>>
>> In my answers, I also remind that our tile service made by
>> volunteers on donated hardware is not unlimited and inviting them
>> to have a look at switch2osm to setup their own tile server or
>use
>> a commercial provider.
>>
>> Up to now, nobody complained :)
>>
>>
>> Yesterday, I've started automating attribution checking using
>> selenium. For each referer, a python script loads the page,
>> searches for tiles, then looks for attribution text or link. The
>> result is stored in a postgresql database which allows to group
>> referers by url, hostname and ip.
>>
>> The attribution percentage I currently see is around 70-80% which
>> is not that bad.
>>
>> My next major step is to use the same technique to remind about
>> tile usage policy...
>>
>>
>> To do something similar on osm.org, a first step is to extract
>> referers from the cache logs, then use the automated attribution
>> check to evaluate the situation.
>>
>>
>> Le 08/03/2020 à 01:52, Nuno Caldeira a écrit :
>>> That would be a good option for those that use third party
>>> providers of OSM. But to be honest, from my experience I highly
>>> doubt that even corporate members of OSMF, like Mapbox would do
>>> it, when their client Facebook (also corporate member of OSMF)
>>> after one year and half, still has maps with lack of attribution
>>> or attributed to HERE, when it's clearly OSM. 
>>>
>>> On Sun, 8 Mar 2020, 00:46 Phil Wyatt, >> > wrote:
>>>
>>> I am sure others may have seen this 'blacklist'
>>> implementation for showing a reminder about attribution.
>>>
>>> https://twitter.com/cq94/status/1234528717604577282
>>>
>>> Worthy of consideration for openstreetmap.org
>>> ?
>>>
>>> Cheers - Phil
>>>
>> -- 
>> Christian Quest - OpenStreetMap France
>>
>>
>> ___
>> 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] #AttributionIsNotOptional experiment on OSM France tile servers

[Christian Quest]

Le 08/03/2020 à 12:12, Simon Poole a écrit :


Just for the record:

Enforcing attribution for services that you are providing directly 
(aka tiles in some form) only has a small overlap with the goals of 
the attribution guideline, and the avenues open to you depend on your 
ToUs / contracts with your users and the legal situation in the 
countries you are providing the service in.


It simply demonstrates that as a tile provider, you can technically 
detect lack of attribution and enforce it.



I would be very very wary of doing anything that deliberately defaces 
a web site without consulting with a local (to the country the web 
site is in) lawyer, particularly if the message implies wrong doing. 
The safe, I admit also the less fun, option, is to simply block access 
after giving any required notice.


Giving notice is ususally difficult. Contact emails are not read or 
people reading them do not know how to handle it, who to forward it to, 
etc...


These attribution reminder tiles are a (highly) visible notice and just 
1 tile out of 25... 96% of the basemap is there (now partially attributed).


Maybe changing the wording to "It looks like this site forgot to put the 
required attribution in this map corner, so we added it for them. Thx 
for using OpenStreetMap !" ? ;)



--
Christian Quest - OpenStreetMap France


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

Re: [OSM-talk] #AttributionIsNotOptional experiment on OSM France tile servers

[Simon Poole]

Just for the record:

Enforcing attribution for services that you are providing directly (aka
tiles in some form) only has a small overlap with the goals of the
attribution guideline, and the avenues open to you depend on your ToUs /
contracts with your users and the legal situation in the countries you
are providing the service in.

I would be very very wary of doing anything that deliberately defaces a
web site without consulting with a local (to the country the web site is
in) lawyer, particularly if the message implies wrong doing. The safe, I
admit also the less fun, option, is to simply block access after giving
any required notice.

Simon 

Am 08.03.2020 um 11:04 schrieb Yves:
> This looks at first as a nuisance that could be perceived as a bad
> move, but the feedback you're receiving rather prove the contrary.
> Well done!
> Ps: would you share your nginx partial redirect, I may consider it for
> Opensnowmap tiles policy?
>
> Le 8 mars 2020 10:14:58 GMT+01:00, Christian Quest
>  a écrit :
>
> Here is a hort report on this experiment...
>
> I started a week ago by searching OSM France tile server logs for
> referer and checked manually if the map on the refering page was
> correctly attributed.
>
> This allowed me to create a short list of 20 entries of sites
> using the french styled tiles and the humanitarian tiles (yes, it
> is made by OSM France).
>
>
> I then modified our nginx based proxy_cache configuration, to
> redirect some tiles to an "attribution tile" only for the domain
> in the list.
>
> For two of them, I tweeted about it... the most visible one is the
> moroco yellow page service, generating a little less than a
> million daily tile requests on our servers.
>
> https://twitter.com/cq94/status/1234516075695525888
>
> In less than 24 hours, the attribution appeared and I removed them
> from the list.
>
> https://twitter.com/cq94/status/1234779931537739776
>
>
> Then I included an email address in the attribution reminder
> tile... and got emails back within a few hours.
>
> Some were asking how to do the attribution, others telling me the
> attribution was now ok and asking how to remove the reminder tiles.
>
> In my answers, I also remind that our tile service made by
> volunteers on donated hardware is not unlimited and inviting them
> to have a look at switch2osm to setup their own tile server or use
> a commercial provider.
>
> Up to now, nobody complained :)
>
>
> Yesterday, I've started automating attribution checking using
> selenium. For each referer, a python script loads the page,
> searches for tiles, then looks for attribution text or link. The
> result is stored in a postgresql database which allows to group
> referers by url, hostname and ip.
>
> The attribution percentage I currently see is around 70-80% which
> is not that bad.
>
> My next major step is to use the same technique to remind about
> tile usage policy...
>
>
> To do something similar on osm.org, a first step is to extract
> referers from the cache logs, then use the automated attribution
> check to evaluate the situation.
>
>
> Le 08/03/2020 à 01:52, Nuno Caldeira a écrit :
>> That would be a good option for those that use third party
>> providers of OSM. But to be honest, from my experience I highly
>> doubt that even corporate members of OSMF, like Mapbox would do
>> it, when their client Facebook (also corporate member of OSMF)
>> after one year and half, still has maps with lack of attribution
>> or attributed to HERE, when it's clearly OSM. 
>>
>> On Sun, 8 Mar 2020, 00:46 Phil Wyatt, > > wrote:
>>
>> I am sure others may have seen this 'blacklist'
>> implementation for showing a reminder about attribution.
>>
>> https://twitter.com/cq94/status/1234528717604577282
>>
>> Worthy of consideration for openstreetmap.org
>> ?
>>
>> Cheers - Phil
>>
> -- 
> Christian Quest - OpenStreetMap France
>
>
> ___
> talk mailing list
> talk@openstreetmap.org
> https://lists.openstreetmap.org/listinfo/talk


signature.asc
Description: OpenPGP digital signature
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

Re: [OSM-talk] #AttributionIsNotOptional experiment on OSM France tile servers

[Yves]

This looks at first as a nuisance that could be perceived as a bad move, but 
the feedback you're receiving rather prove the contrary.
Well done!
Ps: would you share your nginx partial redirect, I may consider it for 
Opensnowmap tiles policy? 

Le 8 mars 2020 10:14:58 GMT+01:00, Christian Quest  a 
écrit :
>Here is a hort report on this experiment...
>
>I started a week ago by searching OSM France tile server logs for 
>referer and checked manually if the map on the refering page was 
>correctly attributed.
>
>This allowed me to create a short list of 20 entries of sites using the
>
>french styled tiles and the humanitarian tiles (yes, it is made by OSM 
>France).
>
>
>I then modified our nginx based proxy_cache configuration, to redirect 
>some tiles to an "attribution tile" only for the domain in the list.
>
>For two of them, I tweeted about it... the most visible one is the 
>moroco yellow page service, generating a little less than a million 
>daily tile requests on our servers.
>
>https://twitter.com/cq94/status/1234516075695525888
>
>In less than 24 hours, the attribution appeared and I removed them from
>
>the list.
>
>https://twitter.com/cq94/status/1234779931537739776
>
>
>Then I included an email address in the attribution reminder tile...
>and 
>got emails back within a few hours.
>
>Some were asking how to do the attribution, others telling me the 
>attribution was now ok and asking how to remove the reminder tiles.
>
>In my answers, I also remind that our tile service made by volunteers
>on 
>donated hardware is not unlimited and inviting them to have a look at 
>switch2osm to setup their own tile server or use a commercial provider.
>
>Up to now, nobody complained :)
>
>
>Yesterday, I've started automating attribution checking using selenium.
>
>For each referer, a python script loads the page, searches for tiles, 
>then looks for attribution text or link. The result is stored in a 
>postgresql database which allows to group referers by url, hostname and
>ip.
>
>The attribution percentage I currently see is around 70-80% which is
>not 
>that bad.
>
>My next major step is to use the same technique to remind about tile 
>usage policy...
>
>
>To do something similar on osm.org, a first step is to extract referers
>
>from the cache logs, then use the automated attribution check to 
>evaluate the situation.
>
>
>Le 08/03/2020 à 01:52, Nuno Caldeira a écrit :
>> That would be a good option for those that use third party providers 
>> of OSM. But to be honest, from my experience I highly doubt that even
>
>> corporate members of OSMF, like Mapbox would do it, when their client
>
>> Facebook (also corporate member of OSMF) after one year and half, 
>> still has maps with lack of attribution or attributed to HERE, when 
>> it's clearly OSM.
>>
>> On Sun, 8 Mar 2020, 00:46 Phil Wyatt, > > wrote:
>>
>> I am sure others may have seen this 'blacklist' implementation
>for
>> showing a reminder about attribution.
>>
>> https://twitter.com/cq94/status/1234528717604577282
>>
>> Worthy of consideration for openstreetmap.org
>> ?
>>
>> Cheers - Phil
>>
>-- 
>Christian Quest - OpenStreetMap France
___
talk mailing list
talk@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk

[OSM-talk] #AttributionIsNotOptional experiment on OSM France tile servers

[Christian Quest]

Here is a hort report on this experiment...

I started a week ago by searching OSM France tile server logs for 
referer and checked manually if the map on the refering page was 
correctly attributed.


This allowed me to create a short list of 20 entries of sites using the 
french styled tiles and the humanitarian tiles (yes, it is made by OSM 
France).



I then modified our nginx based proxy_cache configuration, to redirect 
some tiles to an "attribution tile" only for the domain in the list.


For two of them, I tweeted about it... the most visible one is the 
moroco yellow page service, generating a little less than a million 
daily tile requests on our servers.


https://twitter.com/cq94/status/1234516075695525888

In less than 24 hours, the attribution appeared and I removed them from 
the list.


https://twitter.com/cq94/status/1234779931537739776


Then I included an email address in the attribution reminder tile... and 
got emails back within a few hours.


Some were asking how to do the attribution, others telling me the 
attribution was now ok and asking how to remove the reminder tiles.


In my answers, I also remind that our tile service made by volunteers on 
donated hardware is not unlimited and inviting them to have a look at 
switch2osm to setup their own tile server or use a commercial provider.


Up to now, nobody complained :)


Yesterday, I've started automating attribution checking using selenium. 
For each referer, a python script loads the page, searches for tiles, 
then looks for attribution text or link. The result is stored in a 
postgresql database which allows to group referers by url, hostname and ip.


The attribution percentage I currently see is around 70-80% which is not 
that bad.


My next major step is to use the same technique to remind about tile 
usage policy...



To do something similar on osm.org, a first step is to extract referers 
from the cache logs, then use the automated attribution check to 
evaluate the situation.



Le 08/03/2020 à 01:52, Nuno Caldeira a écrit :
That would be a good option for those that use third party providers 
of OSM. But to be honest, from my experience I highly doubt that even 
corporate members of OSMF, like Mapbox would do it, when their client 
Facebook (also corporate member of OSMF) after one year and half, 
still has maps with lack of attribution or attributed to HERE, when 
it's clearly OSM.


On Sun, 8 Mar 2020, 00:46 Phil Wyatt, > wrote:


I am sure others may have seen this 'blacklist' implementation for
showing a reminder about attribution.

https://twitter.com/cq94/status/1234528717604577282

Worthy of consideration for openstreetmap.org
?

Cheers - Phil


--
Christian Quest - OpenStreetMap France

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