Re: [Wiki Loves Monuments] Commons app WLM integration - help needed

2021-05-07 Thread Mārtiņš Bruņenieks
Hello!

Our (Latvian) monuments list includes a lot of ancient burial grounds which
are not well suited for WLM as usually there is nothing visible above the
ground. Initially, our heritage board was also concerned about publishing
the coordinates of these locations as there is a problem with illegal
digging to find the artefacts there.
We would appreciate a well-documented option to filter out these from the
results even if there is some manual work needed.

 Mārtiņš


On Fri, May 7, 2021 at 2:06 PM Rebecca O'Neill 
wrote:

> Hi Jean-Frédéric,
>
> We adopted them in the "post" Wikidata era, as there are many National
> Monuments in Ireland that have different numbering systems that are just
> numeric and thus are not a unique identifier in any reliable sense. Being
> able to curate what sites are offered to participants keeps them safe, and
> most importantly demonstrates to the government department that we are not
> being reckless in how we disseminate the information about what sites are
> safe and legal to visit. Using the P2186 identifier is central to
> maintaining trust with that government department which may allow us to
> build on this partnership in the future. I would strongly object to any
> deprecation of P2186.
>
> As Josephinehas also just pointed out, P1435 is too broad a designation to
> use, as it draws in too many sites that are not eligible for WLM. We have
> had issues with this in the 2020 WLM competition with "Tentative World
> Heritage Sites" (Q64414898 and Q917858) appearing on Monumental, which is
> not a designation we can stand over for WLM, and that is just 2 sites out
> of about 6500 that caused us a problem. Local organisers having control
> over what is and isn't offered to participants is central to being able to
> run WLM without additional work filtering out non-eligible submissions that
> were made in good faith.
>
> Kind Regards,
> Rebecca
>
> On Wed, 5 May 2021 at 16:51, Jean-Frédéric 
> wrote:
>
>> Hi Rebecca,
>>
>> Just to say I strongly disagree on your characterisation of the use of
>>> this ID as a "hack" and that it should be actively discouraged. Not all
>>> countries are as fortunate as others, and having comprehensive coverage of
>>> all the relevant sites on Wikipedia with unique identifiers from state
>>> bodies is a massive hurdle to overcome for many of us. Assuming that the
>>> whole world can meet you at the standard you are accustomed to is
>>> incredibly unfair, and dismisses the difficulties many countries face in
>>> this regard.
>>>
>>> In the Irish context we have found the P2186 a very neat solution to a
>>> systemic issue we face regarding data on Irish monuments and other listed
>>> buildings, and in the 4 years since we started using those IDs with
>>> Monumental, it has not only facilitated a huge amount of participation, but
>>> the Irish system has not been updated to a point that we have
>>> another system of unique IDs to fall back on. And this is an EU country
>>> with supposedly all the benefits that that entails. An alternative
>>> "scalable solution" has yet to otherwise manifest.
>>>
>>> Fair enough if you don't like a particular system of IDs, but I don't
>>> think completely writing it off is fair on those for whom it has been
>>> incredibly useful.
>>>
>>
>> I think that what Maarten meant to say is that, in the pre-Wikidata era,
>> the requirement to have a unique ID for monuments (in the Wikipedia lists,
>> in the UploadWizard and in all the rest of the tooling) led to the creation
>> of the "WLM ID" for such countries that indeed did not have an ID system.
>>
>> But now, any monument with a Wikidata item, whether or not it is part of
>> a national ID system, will have an identifier: the Q-ID itself − hence why
>> (at first glance at least) there is no need anymore for the legacy custom
>> ID. As I understand it, P2186 was only ever meant as a transitional
>> measure, so that whatever was relying on the previously assigned WLM-ID
>> could have a path to get to the Qids.
>>
>> I would also like to point out that just using "heritage designation" is
>>> potentially quite problematic. For example, here in Ireland a huge number
>>> of national monuments and listed structures are private property (in the
>>> case of the National Inventory of Architectural Heritage, private homes,
>>> hospitals, care homes) which may have Wikidata items, but for the purposes
>>> of Wiki Loves Monuments we cannot be seen to encourage people to trespass
>>> or infringe on people's property or privacy by including them on a map or
>>> other upload platform. Having the WLM ID allows for greater control over
>>> what structures are within the remit of the competition, and protects not
>>> only us as the organisers, but our participants as well.
>>>
>>
>> Ah, I see! You are effectively using P2186 as a way to curate a subset of
>> P1435. That’s a neat trick ; before deprecating P2186 we should indeed find
>> a suitable way to achieve the same result.
>>

Re: [Wiki Loves Monuments] Commons app WLM integration - help needed

2021-05-07 Thread Nicolas VIGNERON
Hi,

My 2 cents: if some countries want to keep using the "fake" ID of P2186 as
an easy-no-brainy solution, why not... But this feels in no way a desirable
long-term solution.
And sorry to be blunt but, it seems to me to be a very bad idea to force
P2186 when there is a real official external identifier.
Especially as most countries don't use P2186, it would be really a lot of
work to put a P2186 on everything eligible for WLM. For perspective: P2186
is used right now on only 134k items, for France alone we had 200k
buildings and 280k objects in WLM 2020, so you need to multiply almost by 5
the use of this property just to integrate France this doesn't seem
realistic.

I feel like that there is no choice for the app other to do a careful
selection of pairs of properties and values.
It should include but not be limited to P2186 and P1435 (not with all
possibles values, again selection is needed), all properties on
https://www.wikidata.org/wiki/Template:Cultural_heritage_properties should
be included and as Rebecca mentionned P5817 would be very useful too (sadly
this property is widely underused).

Cheers,
~nicolas

Le ven. 7 mai 2021 à 13:06, Rebecca O'Neill  a
écrit :

> Hi Jean-Frédéric,
>
> We adopted them in the "post" Wikidata era, as there are many National
> Monuments in Ireland that have different numbering systems that are just
> numeric and thus are not a unique identifier in any reliable sense. Being
> able to curate what sites are offered to participants keeps them safe, and
> most importantly demonstrates to the government department that we are not
> being reckless in how we disseminate the information about what sites are
> safe and legal to visit. Using the P2186 identifier is central to
> maintaining trust with that government department which may allow us to
> build on this partnership in the future. I would strongly object to any
> deprecation of P2186.
>
> As Josephinehas also just pointed out, P1435 is too broad a designation to
> use, as it draws in too many sites that are not eligible for WLM. We have
> had issues with this in the 2020 WLM competition with "Tentative World
> Heritage Sites" (Q64414898 and Q917858) appearing on Monumental, which is
> not a designation we can stand over for WLM, and that is just 2 sites out
> of about 6500 that caused us a problem. Local organisers having control
> over what is and isn't offered to participants is central to being able to
> run WLM without additional work filtering out non-eligible submissions that
> were made in good faith.
>
> Kind Regards,
> Rebecca
>
> On Wed, 5 May 2021 at 16:51, Jean-Frédéric 
> wrote:
>
>> Hi Rebecca,
>>
>> Just to say I strongly disagree on your characterisation of the use of
>>> this ID as a "hack" and that it should be actively discouraged. Not all
>>> countries are as fortunate as others, and having comprehensive coverage of
>>> all the relevant sites on Wikipedia with unique identifiers from state
>>> bodies is a massive hurdle to overcome for many of us. Assuming that the
>>> whole world can meet you at the standard you are accustomed to is
>>> incredibly unfair, and dismisses the difficulties many countries face in
>>> this regard.
>>>
>>> In the Irish context we have found the P2186 a very neat solution to a
>>> systemic issue we face regarding data on Irish monuments and other listed
>>> buildings, and in the 4 years since we started using those IDs with
>>> Monumental, it has not only facilitated a huge amount of participation, but
>>> the Irish system has not been updated to a point that we have
>>> another system of unique IDs to fall back on. And this is an EU country
>>> with supposedly all the benefits that that entails. An alternative
>>> "scalable solution" has yet to otherwise manifest.
>>>
>>> Fair enough if you don't like a particular system of IDs, but I don't
>>> think completely writing it off is fair on those for whom it has been
>>> incredibly useful.
>>>
>>
>> I think that what Maarten meant to say is that, in the pre-Wikidata era,
>> the requirement to have a unique ID for monuments (in the Wikipedia lists,
>> in the UploadWizard and in all the rest of the tooling) led to the creation
>> of the "WLM ID" for such countries that indeed did not have an ID system.
>>
>> But now, any monument with a Wikidata item, whether or not it is part of
>> a national ID system, will have an identifier: the Q-ID itself − hence why
>> (at first glance at least) there is no need anymore for the legacy custom
>> ID. As I understand it, P2186 was only ever meant as a transitional
>> measure, so that whatever was relying on the previously assigned WLM-ID
>> could have a path to get to the Qids.
>>
>> I would also like to point out that just using "heritage designation" is
>>> potentially quite problematic. For example, here in Ireland a huge number
>>> of national monuments and listed structures are private property (in the
>>> case of the National Inventory of Architectural Heritage, private homes,
>>> 

Re: [Wiki Loves Monuments] Commons app WLM integration - help needed

2021-05-07 Thread Rebecca O'Neill
 Hi Jean-Frédéric,

We adopted them in the "post" Wikidata era, as there are many National
Monuments in Ireland that have different numbering systems that are just
numeric and thus are not a unique identifier in any reliable sense. Being
able to curate what sites are offered to participants keeps them safe, and
most importantly demonstrates to the government department that we are not
being reckless in how we disseminate the information about what sites are
safe and legal to visit. Using the P2186 identifier is central to
maintaining trust with that government department which may allow us to
build on this partnership in the future. I would strongly object to any
deprecation of P2186.

As Josephinehas also just pointed out, P1435 is too broad a designation to
use, as it draws in too many sites that are not eligible for WLM. We have
had issues with this in the 2020 WLM competition with "Tentative World
Heritage Sites" (Q64414898 and Q917858) appearing on Monumental, which is
not a designation we can stand over for WLM, and that is just 2 sites out
of about 6500 that caused us a problem. Local organisers having control
over what is and isn't offered to participants is central to being able to
run WLM without additional work filtering out non-eligible submissions that
were made in good faith.

Kind Regards,
Rebecca

On Wed, 5 May 2021 at 16:51, Jean-Frédéric 
wrote:

> Hi Rebecca,
>
> Just to say I strongly disagree on your characterisation of the use of
>> this ID as a "hack" and that it should be actively discouraged. Not all
>> countries are as fortunate as others, and having comprehensive coverage of
>> all the relevant sites on Wikipedia with unique identifiers from state
>> bodies is a massive hurdle to overcome for many of us. Assuming that the
>> whole world can meet you at the standard you are accustomed to is
>> incredibly unfair, and dismisses the difficulties many countries face in
>> this regard.
>>
>> In the Irish context we have found the P2186 a very neat solution to a
>> systemic issue we face regarding data on Irish monuments and other listed
>> buildings, and in the 4 years since we started using those IDs with
>> Monumental, it has not only facilitated a huge amount of participation, but
>> the Irish system has not been updated to a point that we have
>> another system of unique IDs to fall back on. And this is an EU country
>> with supposedly all the benefits that that entails. An alternative
>> "scalable solution" has yet to otherwise manifest.
>>
>> Fair enough if you don't like a particular system of IDs, but I don't
>> think completely writing it off is fair on those for whom it has been
>> incredibly useful.
>>
>
> I think that what Maarten meant to say is that, in the pre-Wikidata era,
> the requirement to have a unique ID for monuments (in the Wikipedia lists,
> in the UploadWizard and in all the rest of the tooling) led to the creation
> of the "WLM ID" for such countries that indeed did not have an ID system.
>
> But now, any monument with a Wikidata item, whether or not it is part of a
> national ID system, will have an identifier: the Q-ID itself − hence why
> (at first glance at least) there is no need anymore for the legacy custom
> ID. As I understand it, P2186 was only ever meant as a transitional
> measure, so that whatever was relying on the previously assigned WLM-ID
> could have a path to get to the Qids.
>
> I would also like to point out that just using "heritage designation" is
>> potentially quite problematic. For example, here in Ireland a huge number
>> of national monuments and listed structures are private property (in the
>> case of the National Inventory of Architectural Heritage, private homes,
>> hospitals, care homes) which may have Wikidata items, but for the purposes
>> of Wiki Loves Monuments we cannot be seen to encourage people to trespass
>> or infringe on people's property or privacy by including them on a map or
>> other upload platform. Having the WLM ID allows for greater control over
>> what structures are within the remit of the competition, and protects not
>> only us as the organisers, but our participants as well.
>>
>
> Ah, I see! You are effectively using P2186 as a way to curate a subset of
> P1435. That’s a neat trick ; before deprecating P2186 we should indeed find
> a suitable way to achieve the same result.
>
>
> @Josephine: Rebecca mentioned Monumental − indeed Monumental already had
> to solve the same problem (querying Wikidata for monuments, displaying them
> on a map, allowing upload with the right templates/etc.). Was there ever an
> issue with the way monumental does it? If not, I’d suggest using that
> methodology as starting point.
>
> I believe the relevant source code for it is:
> * For WLM-Monumental (ie https://maps.wikilovesmonuments.org/) :
> https://github.com/hatnote/monumental-wlm/blob/master/src/services/campaigns.service.js
> * For classic Monumental (ie https://monumental.toolforge.org/ ):
> 

Re: [Wiki Loves Monuments] Commons app WLM integration - help needed

2021-05-07 Thread Josephine Lim
Hi Jean-Frédéric,

Thanks for the links! I experimented with WLM-Monumental for a bit, and it
seems that it does indeed display natural heritage sites and natural
reserves as well (which is the concern that I mentioned in my previous
email). For instance, searching around my area, I get
https://www.wikidata.org/wiki/Q8023299 and
https://www.wikidata.org/wiki/Q65099475 in addition to other actual
monuments.

If this is intended behaviour, we are more than happy to go with P1435.

Best regards,
Josephine


On Thu, 6 May 2021 at 01:51, Jean-Frédéric 
wrote:

> Hi Rebecca,
>
> Just to say I strongly disagree on your characterisation of the use of
>> this ID as a "hack" and that it should be actively discouraged. Not all
>> countries are as fortunate as others, and having comprehensive coverage of
>> all the relevant sites on Wikipedia with unique identifiers from state
>> bodies is a massive hurdle to overcome for many of us. Assuming that the
>> whole world can meet you at the standard you are accustomed to is
>> incredibly unfair, and dismisses the difficulties many countries face in
>> this regard.
>>
>> In the Irish context we have found the P2186 a very neat solution to a
>> systemic issue we face regarding data on Irish monuments and other listed
>> buildings, and in the 4 years since we started using those IDs with
>> Monumental, it has not only facilitated a huge amount of participation, but
>> the Irish system has not been updated to a point that we have
>> another system of unique IDs to fall back on. And this is an EU country
>> with supposedly all the benefits that that entails. An alternative
>> "scalable solution" has yet to otherwise manifest.
>>
>> Fair enough if you don't like a particular system of IDs, but I don't
>> think completely writing it off is fair on those for whom it has been
>> incredibly useful.
>>
>
> I think that what Maarten meant to say is that, in the pre-Wikidata era,
> the requirement to have a unique ID for monuments (in the Wikipedia lists,
> in the UploadWizard and in all the rest of the tooling) led to the creation
> of the "WLM ID" for such countries that indeed did not have an ID system.
>
> But now, any monument with a Wikidata item, whether or not it is part of a
> national ID system, will have an identifier: the Q-ID itself − hence why
> (at first glance at least) there is no need anymore for the legacy custom
> ID. As I understand it, P2186 was only ever meant as a transitional
> measure, so that whatever was relying on the previously assigned WLM-ID
> could have a path to get to the Qids.
>
> I would also like to point out that just using "heritage designation" is
>> potentially quite problematic. For example, here in Ireland a huge number
>> of national monuments and listed structures are private property (in the
>> case of the National Inventory of Architectural Heritage, private homes,
>> hospitals, care homes) which may have Wikidata items, but for the purposes
>> of Wiki Loves Monuments we cannot be seen to encourage people to trespass
>> or infringe on people's property or privacy by including them on a map or
>> other upload platform. Having the WLM ID allows for greater control over
>> what structures are within the remit of the competition, and protects not
>> only us as the organisers, but our participants as well.
>>
>
> Ah, I see! You are effectively using P2186 as a way to curate a subset of
> P1435. That’s a neat trick ; before deprecating P2186 we should indeed find
> a suitable way to achieve the same result.
>
>
> @Josephine: Rebecca mentioned Monumental − indeed Monumental already had
> to solve the same problem (querying Wikidata for monuments, displaying them
> on a map, allowing upload with the right templates/etc.). Was there ever an
> issue with the way monumental does it? If not, I’d suggest using that
> methodology as starting point.
>
> I believe the relevant source code for it is:
> * For WLM-Monumental (ie https://maps.wikilovesmonuments.org/) :
> https://github.com/hatnote/monumental-wlm/blob/master/src/services/campaigns.service.js
> * For classic Monumental (ie https://monumental.toolforge.org/ ):
> https://github.com/hatnote/monumental/blob/HEAD/src/components/main/dashboard/dashboard.js#L97-L123
>
> --
> Jean-Frédéric
> ___
> Wiki Loves Monuments mailing list
> WikiLovesMonuments@lists.wikimedia.org
> https://lists.wikimedia.org/mailman/listinfo/wikilovesmonuments
> http://www.wikilovesmonuments.org
>
___
Wiki Loves Monuments mailing list
WikiLovesMonuments@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/wikilovesmonuments
http://www.wikilovesmonuments.org


Re: [Wiki Loves Monuments] Commons app WLM integration - help needed

2021-05-05 Thread Jean-Frédéric
Hi Rebecca,

Just to say I strongly disagree on your characterisation of the use of this
> ID as a "hack" and that it should be actively discouraged. Not all
> countries are as fortunate as others, and having comprehensive coverage of
> all the relevant sites on Wikipedia with unique identifiers from state
> bodies is a massive hurdle to overcome for many of us. Assuming that the
> whole world can meet you at the standard you are accustomed to is
> incredibly unfair, and dismisses the difficulties many countries face in
> this regard.
>
> In the Irish context we have found the P2186 a very neat solution to a
> systemic issue we face regarding data on Irish monuments and other listed
> buildings, and in the 4 years since we started using those IDs with
> Monumental, it has not only facilitated a huge amount of participation, but
> the Irish system has not been updated to a point that we have
> another system of unique IDs to fall back on. And this is an EU country
> with supposedly all the benefits that that entails. An alternative
> "scalable solution" has yet to otherwise manifest.
>
> Fair enough if you don't like a particular system of IDs, but I don't
> think completely writing it off is fair on those for whom it has been
> incredibly useful.
>

I think that what Maarten meant to say is that, in the pre-Wikidata era,
the requirement to have a unique ID for monuments (in the Wikipedia lists,
in the UploadWizard and in all the rest of the tooling) led to the creation
of the "WLM ID" for such countries that indeed did not have an ID system.

But now, any monument with a Wikidata item, whether or not it is part of a
national ID system, will have an identifier: the Q-ID itself − hence why
(at first glance at least) there is no need anymore for the legacy custom
ID. As I understand it, P2186 was only ever meant as a transitional
measure, so that whatever was relying on the previously assigned WLM-ID
could have a path to get to the Qids.

I would also like to point out that just using "heritage designation" is
> potentially quite problematic. For example, here in Ireland a huge number
> of national monuments and listed structures are private property (in the
> case of the National Inventory of Architectural Heritage, private homes,
> hospitals, care homes) which may have Wikidata items, but for the purposes
> of Wiki Loves Monuments we cannot be seen to encourage people to trespass
> or infringe on people's property or privacy by including them on a map or
> other upload platform. Having the WLM ID allows for greater control over
> what structures are within the remit of the competition, and protects not
> only us as the organisers, but our participants as well.
>

Ah, I see! You are effectively using P2186 as a way to curate a subset of
P1435. That’s a neat trick ; before deprecating P2186 we should indeed find
a suitable way to achieve the same result.


@Josephine: Rebecca mentioned Monumental − indeed Monumental already had to
solve the same problem (querying Wikidata for monuments, displaying them on
a map, allowing upload with the right templates/etc.). Was there ever an
issue with the way monumental does it? If not, I’d suggest using that
methodology as starting point.

I believe the relevant source code for it is:
* For WLM-Monumental (ie https://maps.wikilovesmonuments.org/) :
https://github.com/hatnote/monumental-wlm/blob/master/src/services/campaigns.service.js
* For classic Monumental (ie https://monumental.toolforge.org/ ):
https://github.com/hatnote/monumental/blob/HEAD/src/components/main/dashboard/dashboard.js#L97-L123

-- 
Jean-Frédéric
___
Wiki Loves Monuments mailing list
WikiLovesMonuments@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/wikilovesmonuments
http://www.wikilovesmonuments.org


Re: [Wiki Loves Monuments] Commons app WLM integration - help needed

2021-05-05 Thread Josephine Lim
Hi all,

Thank you all for the input and suggestions! I just want to note that this
feature isn't a one-time thing that we intend to implement and then just
forget about - the intention is to get a minimum viable product up in time
for this year's WLM, and then we will see how it can be improved upon
for future iterations.

Philip - I think a customizing table would be a great idea for experienced
contributors, however for new contributors I'm not sure if they even know
what a Wikidata property is, let alone which one to set. If we ask them
which country they are in and then automatically fill in the property for
them, that would work, however that is a big project all on its own and
unfortunately we do not have the time to get this additional enhancement
done in time for this year's WLM. We do however intend to have an option in
v3.1 (the same version that WLM integration is planned for) that allows
experienced users to manually modify the SPARQL query used by the app if
they wish, so that can be used as a workaround for the time being. :)

Magnus - I love this idea! I think the benefits would exceed the downsides,
from the app's perspective anyway. Unfortunately everyone on our core team
is only familiar with Android development and not tool development. If a
volunteer would like to take this up, we would be more than happy to
collaborate with them on it. Probably we cannot use it this year since we
have only two months left before we need to release this feature in beta,
but for next year I think it could be doable. This might be a good idea for
a GSoC project too.

Maarten - Yes, we are just trying to query Wikidata for items in the
competition, so we can show them on our map with a special designation. I'm
curious about P1435 - if it is used as the criteria, wouldn't you get a lot
of natural heritage areas that aren't monuments? For instance in Australia
the Great Barrier Reef has P1435 because it is a World Heritage Site, but
it's not a monument (or is it still considered a monument for WLM
purposes?). Also, it seems like P1435 also does not work in some countries
- e.g. Sweden and Ireland, based on the discussion above.

Regardless, we are happy to go with either P1435 or P2186 (or even both),
whichever the majority of the community prefers.

Best regards,
Josephine


On Wed, 5 May 2021 at 09:28, Rebecca O'Neill 
wrote:

> I would also like to point out that just using "heritage designation" is
> potentially quite problematic. For example, here in Ireland a huge number
> of national monuments and listed structures are private property (in the
> case of the National Inventory of Architectural Heritage, private homes,
> hospitals, care homes) which may have Wikidata items, but for the purposes
> of Wiki Loves Monuments we cannot be seen to encourage people to trespass
> or infringe on people's property or privacy by including them on a map or
> other upload platform. Having the WLM ID allows for greater control over
> what structures are within the remit of the competition, and protects not
> only us as the organisers, but our participants as well.
>
> Rebecca
>
> On Tue, 4 May 2021, 22:51 Rebecca O'Neill, 
> wrote:
>
>> Hi Maarten,
>>
>> Just to say I strongly disagree on your characterisation of the use of
>> this ID as a "hack" and that it should be actively discouraged. Not all
>> countries are as fortunate as others, and having comprehensive coverage of
>> all the relevant sites on Wikipedia with unique identifiers from state
>> bodies is a massive hurdle to overcome for many of us. Assuming that the
>> whole world can meet you at the standard you are accustomed to is
>> incredibly unfair, and dismisses the difficulties many countries face in
>> this regard.
>>
>> In the Irish context we have found the P2186 a very neat solution to a
>> systemic issue we face regarding data on Irish monuments and other listed
>> buildings, and in the 4 years since we started using those IDs with
>> Monumental, it has not only facilitated a huge amount of participation, but
>> the Irish system has not been updated to a point that we have
>> another system of unique IDs to fall back on. And this is an EU country
>> with supposedly all the benefits that that entails. An alternative
>> "scalable solution" has yet to otherwise manifest.
>>
>> Fair enough if you don't like a particular system of IDs, but I don't
>> think completely writing it off is fair on those for whom it has been
>> incredibly useful.
>>
>> Rebecca
>>
>> On Tue, 4 May 2021 at 22:40, Maarten Dammers  wrote:
>>
>>> Hi Josephine
>>>
>>> On 29-04-2021 11:27, Josephine Lim wrote:
>>> > Hi all,
>>> >
>>> > Hope you are all safe and well. After having finalized our plans for
>>> > WLM integration into the Commons app based on community feedback
>>> > (https://commons.wikimedia.org/wiki/Commons:Mobile_app/WLM , which
>>> was
>>> > posted to this mailing list a couple of months ago), we have started
>>> > work on implementation.
>>> Good to hear you're 

Re: [Wiki Loves Monuments] Commons app WLM integration - help needed

2021-05-04 Thread Rebecca O'Neill
I would also like to point out that just using "heritage designation" is
potentially quite problematic. For example, here in Ireland a huge number
of national monuments and listed structures are private property (in the
case of the National Inventory of Architectural Heritage, private homes,
hospitals, care homes) which may have Wikidata items, but for the purposes
of Wiki Loves Monuments we cannot be seen to encourage people to trespass
or infringe on people's property or privacy by including them on a map or
other upload platform. Having the WLM ID allows for greater control over
what structures are within the remit of the competition, and protects not
only us as the organisers, but our participants as well.

Rebecca

On Tue, 4 May 2021, 22:51 Rebecca O'Neill,  wrote:

> Hi Maarten,
>
> Just to say I strongly disagree on your characterisation of the use of
> this ID as a "hack" and that it should be actively discouraged. Not all
> countries are as fortunate as others, and having comprehensive coverage of
> all the relevant sites on Wikipedia with unique identifiers from state
> bodies is a massive hurdle to overcome for many of us. Assuming that the
> whole world can meet you at the standard you are accustomed to is
> incredibly unfair, and dismisses the difficulties many countries face in
> this regard.
>
> In the Irish context we have found the P2186 a very neat solution to a
> systemic issue we face regarding data on Irish monuments and other listed
> buildings, and in the 4 years since we started using those IDs with
> Monumental, it has not only facilitated a huge amount of participation, but
> the Irish system has not been updated to a point that we have
> another system of unique IDs to fall back on. And this is an EU country
> with supposedly all the benefits that that entails. An alternative
> "scalable solution" has yet to otherwise manifest.
>
> Fair enough if you don't like a particular system of IDs, but I don't
> think completely writing it off is fair on those for whom it has been
> incredibly useful.
>
> Rebecca
>
> On Tue, 4 May 2021 at 22:40, Maarten Dammers  wrote:
>
>> Hi Josephine
>>
>> On 29-04-2021 11:27, Josephine Lim wrote:
>> > Hi all,
>> >
>> > Hope you are all safe and well. After having finalized our plans for
>> > WLM integration into the Commons app based on community feedback
>> > (https://commons.wikimedia.org/wiki/Commons:Mobile_app/WLM , which was
>> > posted to this mailing list a couple of months ago), we have started
>> > work on implementation.
>> Good to hear you're working on this! One of things I see on that page is
>> the mention of the importance of the unique ID. That was always the case
>> before we had Wikidata. Now every Wikidata item has an unique qID so
>> that makes the unique identification a lot easier.
>> >
>> > We have run into a bit of an issue, however, that we are hoping to get
>> > help with. The initial plan (based on the discussion at
>> >
>> https://lists.wikimedia.org/pipermail/wikilovesmonuments/2020-July/009263.html
>> > and at the abovementioned wiki page) was to use Wikidata as our data
>> > source, displaying all Wikidata items with the identifier property of
>> > P2186 (Wiki Loves Monuments ID) on our map.
>> I'll say it a bit stronger than the other commenters: Don't ever use
>> P2186 (Wiki Loves Monuments ID). That's the horrible hack/work around
>> that got introduced for countries that didn't have a well documented ID.
>> >
>> > However we have now found out that some countries don't use P2186 at
>> > all. Someone kindly linked
>> > https://www.wikidata.org/wiki/Wikidata:WikiProject_WLM/Status#Table to
>> > me, and based on that it seems that there is a rather large variety of
>> > identifiers that different countries use.
>> >
>> > Realistically speaking, it would be really difficult for us to account
>> > for every possible identifier that every country uses, and it would
>> > also make stability and maintenance much more tricky. We can try, but
>> > it would be risky and tedious.
>> >
>> > On the other hand, it was suggested to me that if there is WLM
>> > community consensus that P2186 should be used as a universal
>> > identifier, countries might be willing to adopt P2186 for the sake of
>> > harmonization.
>> Relic from the past. Usage should be reduced, not expanded.
>> >
>> > Does anyone have any thoughts on this?
>>
>> Why do you need the ID again? What problem are you trying to solve?
>> Every in scope item should have heritage designation (P1435) on it. You
>> can just query SPARQL for that with distance or bounding box to get
>> anything nearby.
>> Also because most (all?) heritage properties need this P1435, it's
>> usually added as a constraint for which you can query, see
>> https://w.wiki/3HBM (and https://w.wiki/3HBW to check).
>>
>> For fun and to show what's possible, a query of all different types of
>> monuments around the tripoint where Belgium, German and the Netherlands
>> meet: https://w.wiki/3HBm
>>
>> I'm 

Re: [Wiki Loves Monuments] Commons app WLM integration - help needed

2021-05-04 Thread Rebecca O'Neill
Hi Maarten,

Just to say I strongly disagree on your characterisation of the use of this
ID as a "hack" and that it should be actively discouraged. Not all
countries are as fortunate as others, and having comprehensive coverage of
all the relevant sites on Wikipedia with unique identifiers from state
bodies is a massive hurdle to overcome for many of us. Assuming that the
whole world can meet you at the standard you are accustomed to is
incredibly unfair, and dismisses the difficulties many countries face in
this regard.

In the Irish context we have found the P2186 a very neat solution to a
systemic issue we face regarding data on Irish monuments and other listed
buildings, and in the 4 years since we started using those IDs with
Monumental, it has not only facilitated a huge amount of participation, but
the Irish system has not been updated to a point that we have
another system of unique IDs to fall back on. And this is an EU country
with supposedly all the benefits that that entails. An alternative
"scalable solution" has yet to otherwise manifest.

Fair enough if you don't like a particular system of IDs, but I don't think
completely writing it off is fair on those for whom it has been
incredibly useful.

Rebecca

On Tue, 4 May 2021 at 22:40, Maarten Dammers  wrote:

> Hi Josephine
>
> On 29-04-2021 11:27, Josephine Lim wrote:
> > Hi all,
> >
> > Hope you are all safe and well. After having finalized our plans for
> > WLM integration into the Commons app based on community feedback
> > (https://commons.wikimedia.org/wiki/Commons:Mobile_app/WLM , which was
> > posted to this mailing list a couple of months ago), we have started
> > work on implementation.
> Good to hear you're working on this! One of things I see on that page is
> the mention of the importance of the unique ID. That was always the case
> before we had Wikidata. Now every Wikidata item has an unique qID so
> that makes the unique identification a lot easier.
> >
> > We have run into a bit of an issue, however, that we are hoping to get
> > help with. The initial plan (based on the discussion at
> >
> https://lists.wikimedia.org/pipermail/wikilovesmonuments/2020-July/009263.html
> > and at the abovementioned wiki page) was to use Wikidata as our data
> > source, displaying all Wikidata items with the identifier property of
> > P2186 (Wiki Loves Monuments ID) on our map.
> I'll say it a bit stronger than the other commenters: Don't ever use
> P2186 (Wiki Loves Monuments ID). That's the horrible hack/work around
> that got introduced for countries that didn't have a well documented ID.
> >
> > However we have now found out that some countries don't use P2186 at
> > all. Someone kindly linked
> > https://www.wikidata.org/wiki/Wikidata:WikiProject_WLM/Status#Table to
> > me, and based on that it seems that there is a rather large variety of
> > identifiers that different countries use.
> >
> > Realistically speaking, it would be really difficult for us to account
> > for every possible identifier that every country uses, and it would
> > also make stability and maintenance much more tricky. We can try, but
> > it would be risky and tedious.
> >
> > On the other hand, it was suggested to me that if there is WLM
> > community consensus that P2186 should be used as a universal
> > identifier, countries might be willing to adopt P2186 for the sake of
> > harmonization.
> Relic from the past. Usage should be reduced, not expanded.
> >
> > Does anyone have any thoughts on this?
>
> Why do you need the ID again? What problem are you trying to solve?
> Every in scope item should have heritage designation (P1435) on it. You
> can just query SPARQL for that with distance or bounding box to get
> anything nearby.
> Also because most (all?) heritage properties need this P1435, it's
> usually added as a constraint for which you can query, see
> https://w.wiki/3HBM (and https://w.wiki/3HBW to check).
>
> For fun and to show what's possible, a query of all different types of
> monuments around the tripoint where Belgium, German and the Netherlands
> meet: https://w.wiki/3HBm
>
> I'm sure we can figure out a nice scalable solution.
>
> Maarten
>
>
> ___
> Wiki Loves Monuments mailing list
> WikiLovesMonuments@lists.wikimedia.org
> https://lists.wikimedia.org/mailman/listinfo/wikilovesmonuments
> http://www.wikilovesmonuments.org
>


-- 
PhD in Digital Media
Project Coordinator Wikimedia Community Ireland 
Vice Chair of Women in Technology and Science 
Secretary of the National Committee for Commemorative Plaques in Science
and Technology
She/Her
___
Wiki Loves Monuments mailing list
WikiLovesMonuments@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/wikilovesmonuments
http://www.wikilovesmonuments.org


Re: [Wiki Loves Monuments] Commons app WLM integration - help needed

2021-05-04 Thread Maarten Dammers

Hi Josephine

On 29-04-2021 11:27, Josephine Lim wrote:

Hi all,

Hope you are all safe and well. After having finalized our plans for 
WLM integration into the Commons app based on community feedback 
(https://commons.wikimedia.org/wiki/Commons:Mobile_app/WLM , which was 
posted to this mailing list a couple of months ago), we have started 
work on implementation.
Good to hear you're working on this! One of things I see on that page is 
the mention of the importance of the unique ID. That was always the case 
before we had Wikidata. Now every Wikidata item has an unique qID so 
that makes the unique identification a lot easier.


We have run into a bit of an issue, however, that we are hoping to get 
help with. The initial plan (based on the discussion at 
https://lists.wikimedia.org/pipermail/wikilovesmonuments/2020-July/009263.html 
and at the abovementioned wiki page) was to use Wikidata as our data 
source, displaying all Wikidata items with the identifier property of 
P2186 (Wiki Loves Monuments ID) on our map.
I'll say it a bit stronger than the other commenters: Don't ever use 
P2186 (Wiki Loves Monuments ID). That's the horrible hack/work around 
that got introduced for countries that didn't have a well documented ID.


However we have now found out that some countries don't use P2186 at 
all. Someone kindly linked 
https://www.wikidata.org/wiki/Wikidata:WikiProject_WLM/Status#Table to 
me, and based on that it seems that there is a rather large variety of 
identifiers that different countries use.


Realistically speaking, it would be really difficult for us to account 
for every possible identifier that every country uses, and it would 
also make stability and maintenance much more tricky. We can try, but 
it would be risky and tedious.


On the other hand, it was suggested to me that if there is WLM 
community consensus that P2186 should be used as a universal 
identifier, countries might be willing to adopt P2186 for the sake of 
harmonization.

Relic from the past. Usage should be reduced, not expanded.


Does anyone have any thoughts on this?


Why do you need the ID again? What problem are you trying to solve? 
Every in scope item should have heritage designation (P1435) on it. You 
can just query SPARQL for that with distance or bounding box to get 
anything nearby.
Also because most (all?) heritage properties need this P1435, it's 
usually added as a constraint for which you can query, see 
https://w.wiki/3HBM (and https://w.wiki/3HBW to check).


For fun and to show what's possible, a query of all different types of 
monuments around the tripoint where Belgium, German and the Netherlands 
meet: https://w.wiki/3HBm


I'm sure we can figure out a nice scalable solution.

Maarten


___
Wiki Loves Monuments mailing list
WikiLovesMonuments@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/wikilovesmonuments
http://www.wikilovesmonuments.org


Re: [Wiki Loves Monuments] Commons app WLM integration - help needed

2021-04-30 Thread Platonides
I think this would need a property mapping a contest with the used
identifiers (basically, that table but in wikidata form).
So for instance,"Wiki Loves Monuments 2021 Sweden" could be defined as
including the "listed building in Sweden (Q328070)" and the "listed
historical ship in Sweden (Q16501309)", which use P1260 and P2317
identifiers respectively (whereas a few years earlier, maybe only Q328070
were included).

Cheers
___
Wiki Loves Monuments mailing list
WikiLovesMonuments@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/wikilovesmonuments
http://www.wikilovesmonuments.org


Re: [Wiki Loves Monuments] Commons app WLM integration - help needed

2021-04-29 Thread Magnus Manske via WikiLovesMonuments
Maybe a solution would be a new tool acting as an interface? Use P2186 but
also the other "national" properties/combinations.
Upsides:
- we can define the API
- we can adapt to any variation in tracking items, now or later
- we can remove individual solutions as the community migrates to P2186
Downsides:
- work
- slower service than wikidata proper
- no direct SPARQL
- likely needs to maintain its own database (for performance reasons),
which needs to be kept in sync with Wikidata

Cheers,
Magnus

On Thu, Apr 29, 2021 at 12:08 PM Paulo Santos Perneta <
paulospern...@gmail.com> wrote:

> Correction: We are only using the PT prefix for IDs attributed for WMPT.
> When it's the monument ID, we give a prefix related to the ID used: SIPA,
> DGPC, etc.
> To descomplicate things, we are now simply using PT- QXXX (example:
> PT-Q99463700) for listed monuments not having a specific or clear monument
> ID.
>
> Best,
> Paulo
>
> Paulo Santos Perneta  escreveu no dia quinta,
> 29/04/2021 à(s) 12:56:
>
>> Hello,
>>
>> For the Portuguese monuments we are using P2186 with a value prefixed by
>> PT (to avoid ID conflicts) and followed by:
>> * A monument ID, when there is one
>> * An unique ID given by ourselves, when it's not practical to use the
>> monument ID, or when there is no monument ID at all
>> in addition to that, we have a number of specific monument IDs, but we
>> don't use them directly for WLM or WLE (and it was a mistake to even
>> attempt to use them initially, it only brought a lot of confusion)
>>
>> Personally, I support uniformization in P2186,  using country prefixes
>> or some other strategy, in addition to any specific monument that may exist.
>>
>> Best,
>> Paulo
>>
>> Josephine Lim  escreveu no dia quinta,
>> 29/04/2021 à(s) 10:28:
>>
>>> Hi all,
>>>
>>> Hope you are all safe and well. After having finalized our plans for WLM
>>> integration into the Commons app based on community feedback (
>>> https://commons.wikimedia.org/wiki/Commons:Mobile_app/WLM , which was
>>> posted to this mailing list a couple of months ago), we have started work
>>> on implementation.
>>>
>>> We have run into a bit of an issue, however, that we are hoping to get
>>> help with. The initial plan (based on the discussion at
>>> https://lists.wikimedia.org/pipermail/wikilovesmonuments/2020-July/009263.html
>>> and at the abovementioned wiki page) was to use Wikidata as our data
>>> source, displaying all Wikidata items with the identifier property of P2186
>>> (Wiki Loves Monuments ID) on our map.
>>>
>>> However we have now found out that some countries don't use P2186 at
>>> all. Someone kindly linked
>>> https://www.wikidata.org/wiki/Wikidata:WikiProject_WLM/Status#Table to
>>> me, and based on that it seems that there is a rather large variety of
>>> identifiers that different countries use.
>>>
>>> Realistically speaking, it would be really difficult for us to account
>>> for every possible identifier that every country uses, and it would also
>>> make stability and maintenance much more tricky. We can try, but it would
>>> be risky and tedious.
>>>
>>> On the other hand, it was suggested to me that if there is WLM community
>>> consensus that P2186 should be used as a universal identifier, countries
>>> might be willing to adopt P2186 for the sake of harmonization.
>>>
>>> Does anyone have any thoughts on this?
>>>
>>> Many thanks.
>>>
>>>
>>> Best regards,
>>> Josephine / User:misaochan, Commons app project lead
>>> ___
>>> Wiki Loves Monuments mailing list
>>> WikiLovesMonuments@lists.wikimedia.org
>>> https://lists.wikimedia.org/mailman/listinfo/wikilovesmonuments
>>> http://www.wikilovesmonuments.org
>>>
>> ___
> Wiki Loves Monuments mailing list
> WikiLovesMonuments@lists.wikimedia.org
> https://lists.wikimedia.org/mailman/listinfo/wikilovesmonuments
> http://www.wikilovesmonuments.org
>
___
Wiki Loves Monuments mailing list
WikiLovesMonuments@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/wikilovesmonuments
http://www.wikilovesmonuments.org


Re: [Wiki Loves Monuments] Commons app WLM integration - help needed

2021-04-29 Thread Paulo Santos Perneta
Correction: We are only using the PT prefix for IDs attributed for WMPT.
When it's the monument ID, we give a prefix related to the ID used: SIPA,
DGPC, etc.
To descomplicate things, we are now simply using PT- QXXX (example:
PT-Q99463700) for listed monuments not having a specific or clear monument
ID.

Best,
Paulo

Paulo Santos Perneta  escreveu no dia quinta,
29/04/2021 à(s) 12:56:

> Hello,
>
> For the Portuguese monuments we are using P2186 with a value prefixed by
> PT (to avoid ID conflicts) and followed by:
> * A monument ID, when there is one
> * An unique ID given by ourselves, when it's not practical to use the
> monument ID, or when there is no monument ID at all
> in addition to that, we have a number of specific monument IDs, but we
> don't use them directly for WLM or WLE (and it was a mistake to even
> attempt to use them initially, it only brought a lot of confusion)
>
> Personally, I support uniformization in P2186,  using country prefixes or
> some other strategy, in addition to any specific monument that may exist.
>
> Best,
> Paulo
>
> Josephine Lim  escreveu no dia quinta,
> 29/04/2021 à(s) 10:28:
>
>> Hi all,
>>
>> Hope you are all safe and well. After having finalized our plans for WLM
>> integration into the Commons app based on community feedback (
>> https://commons.wikimedia.org/wiki/Commons:Mobile_app/WLM , which was
>> posted to this mailing list a couple of months ago), we have started work
>> on implementation.
>>
>> We have run into a bit of an issue, however, that we are hoping to get
>> help with. The initial plan (based on the discussion at
>> https://lists.wikimedia.org/pipermail/wikilovesmonuments/2020-July/009263.html
>> and at the abovementioned wiki page) was to use Wikidata as our data
>> source, displaying all Wikidata items with the identifier property of P2186
>> (Wiki Loves Monuments ID) on our map.
>>
>> However we have now found out that some countries don't use P2186 at all.
>> Someone kindly linked
>> https://www.wikidata.org/wiki/Wikidata:WikiProject_WLM/Status#Table to
>> me, and based on that it seems that there is a rather large variety of
>> identifiers that different countries use.
>>
>> Realistically speaking, it would be really difficult for us to account
>> for every possible identifier that every country uses, and it would also
>> make stability and maintenance much more tricky. We can try, but it would
>> be risky and tedious.
>>
>> On the other hand, it was suggested to me that if there is WLM community
>> consensus that P2186 should be used as a universal identifier, countries
>> might be willing to adopt P2186 for the sake of harmonization.
>>
>> Does anyone have any thoughts on this?
>>
>> Many thanks.
>>
>>
>> Best regards,
>> Josephine / User:misaochan, Commons app project lead
>> ___
>> Wiki Loves Monuments mailing list
>> WikiLovesMonuments@lists.wikimedia.org
>> https://lists.wikimedia.org/mailman/listinfo/wikilovesmonuments
>> http://www.wikilovesmonuments.org
>>
>
___
Wiki Loves Monuments mailing list
WikiLovesMonuments@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/wikilovesmonuments
http://www.wikilovesmonuments.org


Re: [Wiki Loves Monuments] Commons app WLM integration - help needed

2021-04-29 Thread Paulo Santos Perneta
Hello,

For the Portuguese monuments we are using P2186 with a value prefixed by PT
(to avoid ID conflicts) and followed by:
* A monument ID, when there is one
* An unique ID given by ourselves, when it's not practical to use the
monument ID, or when there is no monument ID at all
in addition to that, we have a number of specific monument IDs, but we
don't use them directly for WLM or WLE (and it was a mistake to even
attempt to use them initially, it only brought a lot of confusion)

Personally, I support uniformization in P2186,  using country prefixes or
some other strategy, in addition to any specific monument that may exist.

Best,
Paulo

Josephine Lim  escreveu no dia quinta, 29/04/2021
à(s) 10:28:

> Hi all,
>
> Hope you are all safe and well. After having finalized our plans for WLM
> integration into the Commons app based on community feedback (
> https://commons.wikimedia.org/wiki/Commons:Mobile_app/WLM , which was
> posted to this mailing list a couple of months ago), we have started work
> on implementation.
>
> We have run into a bit of an issue, however, that we are hoping to get
> help with. The initial plan (based on the discussion at
> https://lists.wikimedia.org/pipermail/wikilovesmonuments/2020-July/009263.html
> and at the abovementioned wiki page) was to use Wikidata as our data
> source, displaying all Wikidata items with the identifier property of P2186
> (Wiki Loves Monuments ID) on our map.
>
> However we have now found out that some countries don't use P2186 at all.
> Someone kindly linked
> https://www.wikidata.org/wiki/Wikidata:WikiProject_WLM/Status#Table to
> me, and based on that it seems that there is a rather large variety of
> identifiers that different countries use.
>
> Realistically speaking, it would be really difficult for us to account for
> every possible identifier that every country uses, and it would also make
> stability and maintenance much more tricky. We can try, but it would be
> risky and tedious.
>
> On the other hand, it was suggested to me that if there is WLM community
> consensus that P2186 should be used as a universal identifier, countries
> might be willing to adopt P2186 for the sake of harmonization.
>
> Does anyone have any thoughts on this?
>
> Many thanks.
>
>
> Best regards,
> Josephine / User:misaochan, Commons app project lead
> ___
> Wiki Loves Monuments mailing list
> WikiLovesMonuments@lists.wikimedia.org
> https://lists.wikimedia.org/mailman/listinfo/wikilovesmonuments
> http://www.wikilovesmonuments.org
>
___
Wiki Loves Monuments mailing list
WikiLovesMonuments@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/wikilovesmonuments
http://www.wikilovesmonuments.org


Re: [Wiki Loves Monuments] Commons app WLM integration - help needed

2021-04-29 Thread Philip Kopetzky
Hi Josephine!

Would it not be possible to set up a customizing table that enables users
to set the property used in a specific country for monuments?

The WLM ID is only used by countries that don't have their own monument ID
and is not commonplace. Even if you started copying monument IDs to the WLM
ID, you'd have to run a daily job to check if those two IDs are still in
sync.

BR,
Philip

On Thu, 29 Apr 2021 at 11:05, Kimmo Virtanen 
wrote:

> Hi,
>
> I don't  have an opinion on what it should be, but if I remember
> correctly the Monumental's method was to include all items with P625 and
> P1435 values. P1435 however doesn't work least in Sweden.
>
> Br,
> -- Kimmo Virtanen
>
> On Thu, Apr 29, 2021 at 12:37 PM Jernej Polajnar 
> wrote:
>
>> Hi,
>>
>> "Wiki Loves Monuments ID" is self-referential, I don't think it would be
>> a good replacement for IDs in national monument databases - for example,
>> those can be linked to external database entries, while P2186 can't be. If
>> anything, P2186 could be used an additional ID, with own set of values or
>> mirroring the national IDs. Although, own set of values would probably be a
>> better idea to avoid conflicts between countries.
>>
>> Jernej
>>
>>
>>
>> On 29. 04. 2021 11:27, Josephine Lim wrote:
>>
>> Hi all,
>>
>> Hope you are all safe and well. After having finalized our plans for WLM
>> integration into the Commons app based on community feedback (
>> https://commons.wikimedia.org/wiki/Commons:Mobile_app/WLM , which was
>> posted to this mailing list a couple of months ago), we have started work
>> on implementation.
>>
>> We have run into a bit of an issue, however, that we are hoping to get
>> help with. The initial plan (based on the discussion at
>> https://lists.wikimedia.org/pipermail/wikilovesmonuments/2020-July/009263.html
>> and at the abovementioned wiki page) was to use Wikidata as our data
>> source, displaying all Wikidata items with the identifier property of P2186
>> (Wiki Loves Monuments ID) on our map.
>>
>> However we have now found out that some countries don't use P2186 at all.
>> Someone kindly linked
>> https://www.wikidata.org/wiki/Wikidata:WikiProject_WLM/Status#Table to
>> me, and based on that it seems that there is a rather large variety of
>> identifiers that different countries use.
>>
>> Realistically speaking, it would be really difficult for us to account
>> for every possible identifier that every country uses, and it would also
>> make stability and maintenance much more tricky. We can try, but it would
>> be risky and tedious.
>>
>> On the other hand, it was suggested to me that if there is WLM community
>> consensus that P2186 should be used as a universal identifier, countries
>> might be willing to adopt P2186 for the sake of harmonization.
>>
>> Does anyone have any thoughts on this?
>>
>> Many thanks.
>>
>>
>> Best regards,
>> Josephine / User:misaochan, Commons app project lead
>>
>> ___
>> Wiki Loves Monuments mailing 
>> listWikiLovesMonuments@lists.wikimedia.orghttps://lists.wikimedia.org/mailman/listinfo/wikilovesmonumentshttp://www.wikilovesmonuments.org
>>
>>
>> ___
>> Wiki Loves Monuments mailing list
>> WikiLovesMonuments@lists.wikimedia.org
>> https://lists.wikimedia.org/mailman/listinfo/wikilovesmonuments
>> http://www.wikilovesmonuments.org
>>
> ___
> Wiki Loves Monuments mailing list
> WikiLovesMonuments@lists.wikimedia.org
> https://lists.wikimedia.org/mailman/listinfo/wikilovesmonuments
> http://www.wikilovesmonuments.org
>
___
Wiki Loves Monuments mailing list
WikiLovesMonuments@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/wikilovesmonuments
http://www.wikilovesmonuments.org


Re: [Wiki Loves Monuments] Commons app WLM integration - help needed

2021-04-29 Thread Kimmo Virtanen
Hi,

I don't  have an opinion on what it should be, but if I remember
correctly the Monumental's method was to include all items with P625 and
P1435 values. P1435 however doesn't work least in Sweden.

Br,
-- Kimmo Virtanen

On Thu, Apr 29, 2021 at 12:37 PM Jernej Polajnar 
wrote:

> Hi,
>
> "Wiki Loves Monuments ID" is self-referential, I don't think it would be a
> good replacement for IDs in national monument databases - for example,
> those can be linked to external database entries, while P2186 can't be. If
> anything, P2186 could be used an additional ID, with own set of values or
> mirroring the national IDs. Although, own set of values would probably be a
> better idea to avoid conflicts between countries.
>
> Jernej
>
>
>
> On 29. 04. 2021 11:27, Josephine Lim wrote:
>
> Hi all,
>
> Hope you are all safe and well. After having finalized our plans for WLM
> integration into the Commons app based on community feedback (
> https://commons.wikimedia.org/wiki/Commons:Mobile_app/WLM , which was
> posted to this mailing list a couple of months ago), we have started work
> on implementation.
>
> We have run into a bit of an issue, however, that we are hoping to get
> help with. The initial plan (based on the discussion at
> https://lists.wikimedia.org/pipermail/wikilovesmonuments/2020-July/009263.html
> and at the abovementioned wiki page) was to use Wikidata as our data
> source, displaying all Wikidata items with the identifier property of P2186
> (Wiki Loves Monuments ID) on our map.
>
> However we have now found out that some countries don't use P2186 at all.
> Someone kindly linked
> https://www.wikidata.org/wiki/Wikidata:WikiProject_WLM/Status#Table to
> me, and based on that it seems that there is a rather large variety of
> identifiers that different countries use.
>
> Realistically speaking, it would be really difficult for us to account for
> every possible identifier that every country uses, and it would also make
> stability and maintenance much more tricky. We can try, but it would be
> risky and tedious.
>
> On the other hand, it was suggested to me that if there is WLM community
> consensus that P2186 should be used as a universal identifier, countries
> might be willing to adopt P2186 for the sake of harmonization.
>
> Does anyone have any thoughts on this?
>
> Many thanks.
>
>
> Best regards,
> Josephine / User:misaochan, Commons app project lead
>
> ___
> Wiki Loves Monuments mailing 
> listWikiLovesMonuments@lists.wikimedia.orghttps://lists.wikimedia.org/mailman/listinfo/wikilovesmonumentshttp://www.wikilovesmonuments.org
>
>
> ___
> Wiki Loves Monuments mailing list
> WikiLovesMonuments@lists.wikimedia.org
> https://lists.wikimedia.org/mailman/listinfo/wikilovesmonuments
> http://www.wikilovesmonuments.org
>
___
Wiki Loves Monuments mailing list
WikiLovesMonuments@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/wikilovesmonuments
http://www.wikilovesmonuments.org


Re: [Wiki Loves Monuments] Commons app WLM integration - help needed

2021-04-29 Thread Jernej Polajnar

Hi,

"Wiki Loves Monuments ID" is self-referential, I don't think it would be 
a good replacement for IDs in national monument databases - for example, 
those can be linked to external database entries, while P2186 can't be. 
If anything, P2186 could be used an additional ID, with own set of 
values or mirroring the national IDs. Although, own set of values would 
probably be a better idea to avoid conflicts between countries.


Jernej



On 29. 04. 2021 11:27, Josephine Lim wrote:

Hi all,

Hope you are all safe and well. After having finalized our plans for 
WLM integration into the Commons app based on community feedback 
(https://commons.wikimedia.org/wiki/Commons:Mobile_app/WLM 
 , which 
was posted to this mailing list a couple of months ago), we have 
started work on implementation.


We have run into a bit of an issue, however, that we are hoping to get 
help with. The initial plan (based on the discussion at 
https://lists.wikimedia.org/pipermail/wikilovesmonuments/2020-July/009263.html 
 
and at the abovementioned wiki page) was to use Wikidata as our data 
source, displaying all Wikidata items with the identifier property of 
P2186 (Wiki Loves Monuments ID) on our map.


However we have now found out that some countries don't use P2186 at 
all. Someone kindly linked 
https://www.wikidata.org/wiki/Wikidata:WikiProject_WLM/Status#Table 
 
to me, and based on that it seems that there is a rather large variety 
of identifiers that different countries use.


Realistically speaking, it would be really difficult for us to account 
for every possible identifier that every country uses, and it would 
also make stability and maintenance much more tricky. We can try, but 
it would be risky and tedious.


On the other hand, it was suggested to me that if there is WLM 
community consensus that P2186 should be used as a universal 
identifier, countries might be willing to adopt P2186 for the sake of 
harmonization.


Does anyone have any thoughts on this?

Many thanks.


Best regards,
Josephine / User:misaochan, Commons app project lead

___
Wiki Loves Monuments mailing list
WikiLovesMonuments@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/wikilovesmonuments
http://www.wikilovesmonuments.org



___
Wiki Loves Monuments mailing list
WikiLovesMonuments@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/wikilovesmonuments
http://www.wikilovesmonuments.org