enquiry: Host images in Wikimedia commons?

(Michael Maier) #1

#What is this thread about?

For the SSEDAS project, one of the requirements was that it should be possible to show an image for each initiative (POI).

As we are building a commons here with TransforMap, I proposed to add images (if they fit) to Wikimedia Commons, where they can be used beyond TransforMap too.

You can see my proposal for the SSEDAS import process here in this thread, and my proposal for how the web map collects its data - where Wikimedia Commons is one of the Media providers.

There were questions if Wikimedia Commons is the right place for images:

Siehe Post von User:Srittau im deutschen OSM-Forum: http://forum.openstreetmap.org/viewtopic.php?pid=522909#p522909

Alles, was einen “edukativen” Charakter hat, ist okay - sie müssen auch nicht den (strengeren) Wikipedia-Kriterien entsprechen.

Importing Images
ETL - Extract-Transform-Load hub for data aggregation
(Josef Kreitmayer) #2

Hello @species would you engage in a discussion with wikimedia commons on their native forum on that topic and find out, if that is also the case for our effort?

  • The example you linked (which seems very promisiong) seems to be very much of core geographic nature, as it is images of street signs or hiking path signs.
  • It would be great, if we can host our images there, and find an easy workflow to integrate them.

(Jon Richter) #3

I doubt they want to take all of it.

(Josef Kreitmayer) #4

Me too.

(Michael Maier) #5

As I see my primary focus on OSM stuff, not the Media DB - may I delegate this work to the Berlin community? @alabaeye, @TDoennebrink, @fraukehehl ?

The Berliners may also talk with Wikimedia DE, when they are there?

(Josef Kreitmayer) #6

I personally see it very difficult to try to come to agreements with already existing communities about data, that we do not even know yet, as we are in the very beginning of collecting it. Therefor I would not bother trying to find agreements on using Wikimedia Commons as Image-DB at this moment.

I think @almereyda agrees with me on that.

Nevertheless, getting in connection and having as much data replicated, or in the end linked/hosted with public domain publishers as Wikimedia, I think, is a very good idea. At the same time, I think we should currently focus on core developments we need for the upcoming months.

(Gualter Barbas Baptista) #7

Taking into account the following statement and the overall TransforMap as commons:

I would say we put as less energy as possible in this media aspect. Uploading to Wikimedia Commons might or might not be the simplest solution.

I personally think the simplest is just getting the images on a folder (wherever on our servers) and on the database an URL/URI to it.

Later on we can develop a proper Media DB (preferably distributed, e.g. Mediagoblin), if needed at all.

(Michael Maier) #8

+1 for that.
Links to images in OSM are the following format:

Either if we develop a Media DB or simply extract it to a folder to on of our webserves, the link format is the same for everything outside Wikimedia Commons.

(Michael Maier) #9

In the recent time, I hear the word ‘doubt’ much too often, in my opinion.

It sounds to me like: “I don’t wanna ask, we better invest time to build our own stuff”.

I would like to quote @almereyda here:
It is easier to ask for forgiveness afterwards, than for permission beforehand.

Maybe I have to clarify my questions:
Which type of images do you fear are not accepted by Wikimedia Commons? And don’t say: “I do not know yet” - to fear the unknown is not the way to explore where no one has gone before.

If there are really images that do not serve any “educational purpose”, do you want to have them in TransforMap?

(Josef Kreitmayer) #10

I will leave that conversation at this point, as I am not competent in database-questions.

(Gualter Barbas Baptista) #11

That seems simple enough to implement. I also think that all SSEDAS (and even TransforMap) related images fit well inside Wikimedia Commons, as they are relevant educational/information content. And since changing location is as simple as adding an absolute URL, then we can very quickly implement an own server folder as fallback without losing compatibility.

Therefore a +1 for the use of a very simple image field linking to Wikimedia Commons filenames.

We probably will need to prefix all files so that there are no overlaps with other Wikimedia content, right? transformap_ will probably be a good candidate.

(Michael Maier) #12

File naming is irrelevant in Wikimedia Commons, they just have to be unique.

But I think it would be nice if we add some categories to the images we upload, e.g. Category:TransforMap, or Category:Community_garden e.g.

(Waldir Pimenta) #13

I’m interested in hearing this as well. I am an administrator at Wikimedia Commons (albeit a somewhat reclusive one), and would be glad to help clarify any doubts about this. For starters, I can quote the page “What Commons is not”, particularly the section “Commons is not your personal free web host” which says:

Although we do host media and images on Commons, all content must be within our project’s scope,
which requires, among other things, that all media must be realistically useful for an educational purpose.

This notion is expanded in further detail in the Commons:Project Scope, which provides examples of what is not accepted:

Examples of files that are not realistically useful for an educational purpose:

  • Private image collections, e.g. private party photos, photos of yourself and your friends, your collection of holiday snaps and so on. There are plenty of other projects on the Internet you can use for such a purpose, such as Flickr.
  • Artwork without obvious educational use, including non-educational artwork uploaded to showcase the artist’s skills
  • Files apparently created and/or uploaded for the purpose of vandalism or attack.
  • Advertising or self-promotion.
  • Files that add nothing educationally distinct to the collection of images we already hold covering the same subject, especially if they are of poor or mediocre quality.

Does that help?

(Josef Kreitmayer) #14

@waldyrious, thank you, that is very helpful!

I think there will be mainly 3 types of images, of which

  1. definitely fits
  2. might fit, might not
  3. definitely does not fit:

  1. Clear / good quality Images, that fit Wikimedia quality and license criteria
    Images, that definitely fit the criteria for Wikimedia Commons, and very likely will be provided by the original owner under a public domain license.

  2. Not so clear / good quality images
    Images, that might or might not fit into the criteria of Wikimedia Commons, as they are not so clear / good in quality / to the point, what it is about in content, but do somehow have meaning for the initiative, that wants to use that image.
    I think that a substaintial part of the images will be of that kind. We will get data from more than 20 Partners, which they themselves get from initiatives, which they map as outside-partner third party.

  3. Logos of organizations (do not fit quality nor license criteria of Wikimedia Commons)

  • Quite many of the pictures will be logos of organizations and initiatives.
    For my organization, for example, that is an image, that I would use first.

  • Most organizations would not agree on having their logo published under an open license. It is something very personal. Speaking from the personal unserstanding of my organization, we could not agree on having our logo used for any purpose.

  • So we need to find storage space for the items fiting to the criteria of non publicly usable images of the 3rd category.
  • For example, a question with each image to publish could be, if it should be published under public domain, or is specific to the organization and cannot be published under an open license.
  • I do not know, how we integrate or simply not accept those images of the 2nd category.

(Thomas Kalka) #15

@species: kannst du bitte eine Einführung schreiben, worum es in diesem Thread geht ?

(Thomas Kalka) #16

@josefkreitmayer, warum können die einzelnen Initiativen die Bilder nicht selbst hosten ?
Das würde Fälle 2 und 3 abdecken.

Nebenbei: ein Usecase, für die du Logos imho gerne veröffentlichen würdest, ist die Nutzung in Werken, die über die ensprechenden Organisationen berichten und dort das Logo einbinden wollen. Logos, die durch das Markenrecht geschützt sind, könnten wohl ohne Probleme mit offenen Lizenzen veröffentlicht werden.

(Josef Kreitmayer) #17

as we get more and more non-germanspeaking contributors, please post small notes in english.

(Josef Kreitmayer) #18

@toka, please reformulate in english

sorry for being oncomfortable about the language, but it get´s really difficult to stay in the same conversation with people from various language backgrounds mixing languages. If a regional team keeps their conversation in their langage, that is fine, but having general conversations on discourse in german does not give credit to the international character of the project.

proposal: have english as main discourse.transformap.co language
(Michael Maier) #19

Ah - the top post is not quite clear, I will update that:
For the SSEDAS project, one of the requirements was that it should be possible to show an image for each initiative (POI).

As we are building a commons here with TransforMap, I proposed to add images (if they fit) to Wikimedia Commons, where they can be used beyond TransforMap too.

You can see my proposal for the SSEDAS import process here in this thread, and my proposal for how the web map collects its data - where Wikimedia Commons is one of the Media providers.

(Michael Maier) #20

It is technically possible, that they host the images themselves, yes.
If they send us a link or the image itself (and we upload it to a webserver), doesn’t make a difference in the data storage, see Post #8.

But: We think some of the partners don’t want to host the images themselves, and for SSEDAS TransforMap is a “service provider”, and this would include hosting images somewhere for them.