I've traced (to a limited extent) data from Ordnance Survey OpenData. Everything that I have added to the map derived from there is tagged either Some time ago it was said that those who contributed data from OpenData couldn't accept the new Contributor Terms. I think I've heard since that the Ordnance Survey decided to release OpenData under the new OSM license, is that correct? Seeing as in the next few days then I will be barred from editing if I don't accept the Contributor Terms, I'd like to ask: can I accept the new Contributor Terms if I've contributed data from Ordnance Survey OpenData? asked 15 Jun '11, 16:57 fluteflute |
It's currently unclear as to whether there are any incompatibilities between the OS OpenData license and the Contributor Terms. If you are waiting for an "official" decision then you are likely to be waiting in vain. There may be some guidance from the OSMF but they haven't ever previously made any kind of "official" statement on any similar matter. Some people believe it's compatible. Other people don't. I'm going to stick my neck out and try to be helpful and explain the situation as I see it. The current OS OpenData licence is based on the Open Government License (OGL) with variant terms. One of these extra terms is the requirement to ensure that their specific attribution notice is required by all sub-licenses. This appears to be the only sticking point, since the rest of the licence (and the OGL in general) is clearly compatible and even has clarifying statements to the same effect at the bottom. So in the end it appears to be only the cascading requirement for attribution that is under discussion. The Contributor Terms 1.2.4 Section 1a says that you may contribute data so long as the OSMF has the right to distribute the data "under our current licence terms". If you believe the OS OpenData to be compatible with the CC-BY-SA license and the ODbL, then this can be read as being sufficient. If the "current license terms" were to change in the future, Section 1b allows OSMF to remove the data, if needed. Section 2 (right of grants) and section 4 (the guarantee of attribution, regardless of section 3 relicensing) are very nearly word-for-word identical to the rights you are able to grant when using the OGL, and the only ambiguity is to whether the OSMF has to guarantee cascading attribution by the downstream users of the data. But that's very similar to the situation as CC-BY-SA in any case, since neither the contributor nor the OSMF has any control (or legal responsibility to control) whether or not a 3rd party attributes fully and correctly. (I'm not aware of a single use of OSM data that attributes all the contributors, for example, but this happens routinely). Finally, Section 8 can be read as reassurance that the Section 1 permission to contribute data compatible with the current license terms will still hold, even if there's a problem with other sections. As you can see, there is nothing fundamentally incompatible between the OS OpenData license and the Contributor Terms - they are substantially similar and attribution requirements are similar. Whether you sign the contributor terms or not there is always the risk that at some future point the legal situation may change and, for whatever reason, the data may be deemed incompatible and we have to remove it. So be it. But this is the situation faced by all non-original, non-PD contributions - regardless of the CTs being signed or not. For the record, I believe the OS OpenData attribution requirements are compatible with both CC-BY-SA and ODbL, that cascading attribution to third parties is a matter for the third party in any case and not something enforceable by either the contributor or OSMF, and therefore I'm comfortable signing the CTs and using OS OpenData. answered 16 Jun '11, 09:29 Andy Allan I'm afraid I have to completely disagree with your 5th paragraph. You state "Section 2 (right of grants) and section 4 ... are very nearly word-for-word identical to the rights you are able to grant when using the OG" . This is is irrelevant, since we are not concerned with the OGL, but with the OS Opendata licence. You state "the only ambiguity is to whether the OSMF has to guarantee cascading attribution by the downstream users of the data". Its not abmiguous, the Opendata licence states quite clearly that there must be "a requirement that any further sub-licences do the same"
(17 Jun '11, 19:09)
dmgroom
|
Michael Collinson (License Working Group) has sent a further email to the TalkGB mailing list, titled "OS OpenData and ODbL OK" (July 4th 2011), where he states:
Based on this email I believe it is now clear that the OP and other contributors can accept the new Contributor Terms and continue to use OS OpenData without any fear of licence issues. answered 05 Jul '11, 10:41 GrahamS You view would only hold true when / if the LWG take the action which they only proposed as a draft in their minutes of 21 June 2011. i.e. they state "Therefore, it is the formal LWG minuted position that if data is compatible with current license, there is no legal impediment to contributing it and accepting the contributor terms". Currently they have not yet done so.
(05 Jul '11, 11:54)
dmgroom
|
My understanding was yes, given that if the data at a later point is deemed to be incompatible with the new licence, it probably can and will need to be removed en masse - therefore, I have accepted the new terms under that rationale... Cheers, Chris answered 15 Jun '11, 20:02 c2r So you're basically saying that legally you can't accept, but in reality you might as well since there will need to be a big cleanup anyway? I'd prefer not to accept them until this issue is sorted out officially.
(16 Jun '11, 06:40)
fluteflute
2
No, I'm not saying that at all. If you read (1b) of the new contributor terms, it makes it quite clear that non-compatible data can and will be removed. Currently, the position is not clear in so far as I am aware it has not yet legally been determined whether or not the data is compatible with the new licence - however, I don't see how that prevents me from legally accepting the new contributor terms?
(16 Jun '11, 08:41)
c2r
I'm afraid you must have misread the CT's . If you re-read clause 1(b) you will see it does not say non-compatible data "will" be removed , its says it "may" be removed.
(17 Jun '11, 18:59)
dmgroom
|
Below is an email from Michael Collinson (a member of the License Working Group), sent today on the Talk-GB mailing list.
answered 16 Jun '11, 21:15 fluteflute SK53 ♦ 1
I've added Michael Collinson's follow up email as a new answer, which states that OS have explicitly approved ODbL.
(05 Jul '11, 10:43)
GrahamS
|
Personally I've taken the pragmatic approach of accepting the CTs, but making sure that everything I've traced or named from OS OpenData is suitably tagged with My view is that I'm not a lawyer. I believe that the licenses are compatible (based on my interpretation and the discussions I have read) and I am being open and honest about recording my sources in a way that allows proper attribution to be made, and that also allows OSM to withdraw my work easily if it is later determined that we are violating the OS agreement. That should be good enough IMHO. answered 18 Jun '11, 13:27 GrahamS |
You say "I think I've heard since that the Ordnance Survey decided to release OpenData under the new OSM license, is that correct?" No that is not correct, OpenData is released under the OS OpenData licence What appears a harder question to answer is "Can I accept the new Contributor Terms if I've contributed data from Ordnance Survey OpenData?" Many in OSM initially believed that OpenData had been issued under the UK OpenGovernment licence, but this was an incorrect understanding of the situtaion. The OS OpenData licence in its second paragraph clearly states it incorporates the OGL which is varied by some following terms. Two of those terms are: a) you must clearly attribute Ordnance Survey as the source of the data; (b) The same attribution statements must be contained in any sub-licences of the Information that you grant, together with a requirement that any further sub-licences do the same When you are asked to agree to the CT's, under Clause 2 you are asked to grant rights, which are stated only to be subject to two further clauses (clauses 3 & 4). These rights are very wide ranging, include "without limitation, the right to sub-licence", and crucially as far as atribution goes, only says that OSMF must agree to attribute the copyright owner. So to summarise, if you agree to the CT's you are in effect saying "OSMF can do pretty much what they like with the data as long as it is licenced under a free and open licence, and OSMF themselves attribute the sources", whilst the Ordnance Survey are saying "you can use our data as long as you, and anyone you give the data to, attributes Ordnance Survey. It is this point in bold which as far as I can see is crucial, because the CT's do not in any way state it. This point has been debated over and over, and those that say if you have used OS OpenData then you can still sign the CT's appear to do so on the belief that Clause 2 is also subject to Clause 1(b), when in fact if you read Clause 2 it only says it is subject to 3 & 4. Now I would argue that if whoever drafted the CT's felt you should assume that clause 2 was subject to other clauses without it being explicitly stated, then why did they feel it was necessary to specifically state that clause 2 was subject to clauses 3 & 4 (why could that not be assumed too). Indeed I would go further, given that clauses 2, 3 & 4 are all in the same subsection of the CT's ( the subsection headed "Rights Granted") whilst clause 1 lies outside this sub-section, then if Clause 2 is intended to be subject to clauses 1,3 & 4 then it is even more important that it is specifically stated that is is subject to clause 1, than it is to specifically state it is subject to clauses 3 & 4. I therefore cannot see how you could argue that Clause 2 is subject to Clause 1. But even if you believe that Clause 2 is subject to clause 1(b) then it is important to actually read what clause 1(b) guarantees. Because it guarantees nothing. It says OSMF may delete the data, but also leaves open the possibility that OSMF might not delete the data. I hope I don't paraphrase Michael Collinson too much when I say his email can be summarised as follows:
Ultimately therefore I'm afraid you'll have to make up your own mind as to what you do. Hopefully the above points may guide you. answered 17 Jun '11, 20:24 dmgroom 2
You could have paraphrased it as (i) we have legal advice (as requested by OS) which says it is OK, (ii) we have heard rumours the OS might not be happy with this, so (iii) the LWG are reluctant to issue a clear statement.
(18 Jun '11, 07:47)
EdLoach ♦
|