Yann Forget wrote:

Recently a photo of Alain-Fournier from 1904 was deleted on Commons because "no proof of PD" [1]. The photographer is unknown, and therefore his date of death is obviously unknown. I advocate that we should keep this kind of images. Seeing what was the life expectancy 100 years ago (about 50-55 years in USA / Europe [2] [3]), a limit of 100 years seems reasonable to me. The figures I found are actually lower than I expected (60 years). While we accept a lot of content which is much less safe than this, it seems unreasonable to me to refuse this kind of images. It is in the public domain in USA anyway.

You are assuming that the photgrapher lived no longer than 30 years after taking the photo. This is not a safe assumption. Also, life expectancy of someone who is already an adult is generally higher than the life expectancy of the population in general -- especially in times and places with high infant morbidity. Statistics are a dangerous thing.

Rocket000 said [4] "I think it's very safe to assume it's PD or can be treated like it is, but that's different than allowing it on Commons." That's exactly the point: if it is very safe to assume it's PD, why should we refuse them? Why setting different standards? This goes against our mission.

No. Our mission is to host *definitly* free content. Not probably usable content.

Now, Cecil made an interesting research on life expectancy [5]. Actually that's the only meaningful arguments I have seen so far in this discussion. I understand this argument, but I am not really convinced that we have to be so strict about this issue. For me, it is all a matter of interpretation anyway. If we adopt the POV advocating deletion, we should have a clear rule, so that this gets clear once and for all, and most important, this rule should be applied equally on all projects, not only on Commons. Some people have suggested a 120-years old rule, because of such a duration mentioned in US copyright law.

Yes, a celar rule would be good. I think anything less than 150 or 160 years is unsafe. If we take into consideration the fact that some countries grant copyright from the date of first publication of previously unpublished material, this might not even be long enough.

Lupo said: “The "100 years rule" at the German Wikipedia is a kind of EDP. They clearly acknowledge that they are not sure these images are free, but they consider the risk of getting into trouble over hosting such files low.” Actually that's exactly what I am suggesting. The legal risk for the Foundation is quite nil as the images are in the public domain in USA anyway.

This is not neccessarily true. Provisions of foreigh copyright might be enforced in the US under some conditions, as far as I know. Also, the *uploader* is also taking a considerable risk. And may, if push comes to shove, sue the foundation for compensation because of misleading instructions.

Where the content is hosted is not so terribly relevant. Who ists it may be relevant, but you also have to take into account who made it, who uploaded it, and who it is uploaded for.

Also, people *have* sued for the copyright of old material thought to be "probably" PD. They usually only do when money gets involved. Agin, consider this: someone takes a images that is labeled PD from commons and uses it as a bug cover. He gets sued and has to pay 100K$ -- and then turns on us (the foundation, the uploader, commons admins, whatever) for compensation. Fun fun fun.

But we also have repeatedley have had legal threads about material that clearly *was* PD by law. Copyfraud is very common.

But why should different rules on different projects? The fact that the image is used to illustrate a biography on Wikipedia or the author's page on Wikisource, or standing in its own right on Commons does not change at all its copyright status. I am therefore requesting input from copyright knowledgeable people, including Mike Godwin as counsel for the Foundation, on copyright rules we should apply to our projects.

There's a difference between what is *legal* to have on a site, and what is acceptable by *policy*. Commons' policy is very strict, it's basically the smallest common denominator of all projects. A while back, the foundation said that all projects SHOULD be as strict as commons. I think there was an option to allow fair use, but with a stong recommendation not to.

When deciding what is legal for a project to have and an uploader to provide, there are several jurisdictions to consider. Generally, the use must be legal in the US (to protect the foundation), in the uploader's country, in the country of origin of the work, and in the countries the respective project ist mostly aimed at. As a result, the german wikipedia may choose to allow more logos than commons, because german law explicitly requires a higher level of originality for this type of work in order to be copyrightable. I don't like that choice, because it's germany-centric, but that's how it is, and it seems to be accepted.

I personally think the german wikipedia (and the english wikipedia) should be just as strict as commons, to have only truely free contnet usable anywhere. And we ourselfs should be more strict in order to live up to that ideal. This means: no fair use, no assuming short lifespan, no assuming anonymous publication, etc.

I also think that this attitude of excessively strict copyright interpretation is the main reason why Commons is not more widely used by the different projects: they have no guarantee that the content they allow and they need will be kept on Commons.

If Commons was too tollerant about copyright, people could no longer be sure that content taken from there is really free. Which would also be a reason not to use commons as a legitimate resource. Sure, being lax about this would probably gain us some popularity with the "source: google" growd, just have a look at flickr & co. But that's not what we aim to be, is it?

-- Daniel

PS: In case this doesn't get through to foundation-l because i'm not subscribed there, please forward.