search

w3c / tdm-reservation-protocol

Repository of the Text and Data Mining Reservation Protocol Community Group
https://www.w3.org/community/tdmrep/
Other
9 stars 8 forks source link

A general consideration on the way towards our discussion is aimed #17

Closed claudiotubertini closed 3 years ago

claudiotubertini commented 3 years ago

I understand that simplicity is a basic requirement, but what worries me is TDM exception might become something similar to GDPR cookie declaration: GDPR means more than you have to click on the "ok" button of a popup and forget, it is not a simple nuisance designed by Brussel bureaucrats, as is interpreted by all website publishers. As regards TDM, if a publisher wants to opt out, it should be gently pushed towards offering a licence not just to send an http header with TDMa=1. If you were a publisher which route would you follow?

Command Cost Δ Earnings
TDMa=0 -1 0
TDMa=1 -1 0
TDMa=2 -2 > 0

Δ Earnings are respect to present situation without machine reading proposal. There is always a cost in implementing the proposal. No one will never follow the route TDMa=0. Among those publishers that decide to do something to avoid the exception everybody will choose TDMa=1 if there is no gain in TDMa=2.

llemeurfr commented 3 years ago

When you're listing Earnings, are you speaking about rightsholder or TDM actor earnings?

Because for rightsholders I would rather say:

Command Δ Earnings Reason
TDMa=0 > 0 by stating that its content is open to data mining worldwide, the rightsholder avoids spending time responding to TDM actors outside EU
TDMa=1 > 0 by stating that he is opting-out from the EU TDM exception, the rightsholder applies the freedom expressed in the Article 4 of the DSM
TDMa=2 > 0 by stating that he is opting-out from the EU TDM exception but that a license can be acquired, the rightsholder is in position to get some money from commercial TDM actors

Respectively, for TDM Actors :

Command Δ Earnings Reason
TDMa=0 > 0 A TDM Actor knows that he can mine the content, without having to spend time contacting the rightsholder
TDMa=1 > 0 A TDM Actor knows that he must refrain to mine the content if he's not a European Reserch org; no time lost
TDMa=2 > 0 A TDM Actor gets a direct contact with the rightsholder; no time lost
llemeurfr commented 3 years ago

Besides that I agree that the spec can gently push rightsholders to add a policy to their TDM choice, i.e. it is up to us to promote TDMa=2 rather than TDMa=1. For that, the policy must be simple enough to manage.

claudiotubertini commented 3 years ago

The reason behind my comment is that I would like to stress the importance of licensing. In the last speach I heard during the TDM meeting someone spoke about separating the opting out from the possibility to offer a license. I think that would be a big mistake.

giuliamarangoni commented 3 years ago

Dear Claudio and all, happy to further discuss this at our meeting today. However, let me anticipate one thought. I think that it is fundamental that all the technical solutions envisaged will be neutral with respect to rightsholders business models and decisions. We are doing our best to provide technology solutions that are as easy as possible to adopt by all rightolders (and miners), to enable effective communication of TMD rights and licensing information; however, technical solutions are intended to serve users, without setting any constraints or their decisions.

claudiotubertini commented 3 years ago

I agree, of course. Technical solutions should be technical and usually do not imply any moral obligations. But the solution we try to define has the purpose to clarify to readers (that is the miners) in which sense and which are the details that put a restriction on the freedom allowed by the law. If we devise such a simple automatic procedure that makes everybody opt-out to the exception we obtain the same result as GDPR, a simple annoyance before accessing a web site. What we can do is simply require that tdm-b is always present.

Instead of:

TDM-b SHOULD be set only if the value of TDM-a is 2.

The absence of TDM-b if the value of TDM-a is 2 is considered a protocol error. In such a case TDM Agents MUST consider that the value of TDM-a is 1.

We can have:

TDM-b SHOULD be always set.

The absence of TDM-b is considered a protocol error. In such a case TDM Agents MUST consider that the value of TDM-a is 0. See you in less than an hour. Claudio

On Tue, 15 Jun 2021 at 15:03, giuliamarangoni @.***> wrote:

Dear Claudio and all, happy to further discuss this at our meeting today. However, let me anticipate one thought. I think that it is fundamental that all the technical solutions envisaged will be neutral with respect to rightsholders business models and decisions. We are doing our best to provide technology solutions that are as easy as possible to adopt by all rightolders (and miners), to enable effective communication of TMD rights and licensing information; however, technical solutions are intended to serve users, without setting any constraints or their decisions.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/w3c/tdm-reservation-protocol/issues/17#issuecomment-861479087, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAZP5W4AFIT4H4HCRB7HXI3TS5FRTANCNFSM46EJXPXA .

llemeurfr commented 3 years ago

Closed after a call of the community group.

Promoting the use of TDM Policies will essentially be realized via communication to the public; this is a huge opportunity for both TDM Actors and publishers, publishers should embrace it.