W3C home > Mailing lists > Public > public-bpwg@w3.org > November 2008

New draft 2008-11-11 of mobileOK Scheme

From: Jo Rabin <jrabin@mtld.mobi>
Date: Tue, 11 Nov 2008 18:00:10 +0000
Message-ID: <4919C82A.1010500@mtld.mobi>
To: Phil Archer <phil@philarcher.org>
CC: Francois Daoust <fd@w3.org>, MWI BPWG Public <public-bpwg@w3.org>

Another day another version.

Hi Phil

I took your amendments and spread them around the docuument a bit. Hope 
it suits your points in 1.

Ref 2. I think we probably don't want people referring to the W3C copy 
of the PNG, at least that's what we say in the previous section ref 
visual indication.



On 11/11/2008 11:02, Phil Archer wrote:
> Jo, taking your comments on board...
> I've uploaded an edited version of the doc to [1] - obviously 
> temporarily. In it I have:
> 1. Amended section 2.2.1 on RDF to firstly correct the triple (we'd 
> missed off the 'Conformant' bit at the end) secondly, to add a line that 
> links that section to the following one on POWDER.
> Why include the RDF bit at all?
>  - because it's good practice to create a human readable document that 
> defines what terms in a vocabulary mean
>  - because the mobileOK vocabulary states that it is defined in this 
> document
>  - because triple stores (RDF data sets) can describe anything, whether 
> there is a link to that data or not, just as I can comment on, say, 
> Barrack Obama, even though we've not met.
>  - because the POWDER method described uses the vocabulary to make the 
> claim.
> 2. Amended the POWDER example to give the correct URI of the logo.
> 3. Amended the text talking about HTML link to a) update the rel type 
> (we're going to use describedby, not powder) and note that not all 
> versions of HTML support the profile attribute (HTML5 is bent on killing 
> it off).
> 4. Added a single line about HTTP link with an example and a link to the 
> current Internet Draft.
> [1] http://philarcher.org/mobileOK/20081111.html
> Jo Rabin wrote:
>> Thanks Phil
>>  > 1. No, there's no linkage to the RDF - the example triple is just a
>>  > triple that could occur in any RDF data set. It would be unlikely to
>>  > exist on its own and you wouldn't link to it as such.
>> OK so I am unclear why we are telling people this information in this 
>> document? Seems to be "If you want to make an abstract statement 
>> unconnected to your actual content not accessible by any means 
>> described in this document _and_ you want to use RDF, this is how you 
>> might do it?" Sorry if this seems flippant, but seems to be probably 
>> something that confuses rather than helps your average "Joe". Perhaps 
>> we should remove it.
>>  > 2. Do you really want to omit mention of HTTP Link here?
>>  > http://philarcher.org/powder/20081110.html#httplink. We plan to 
>> leave it
>>  > in the document even if we end up having to flag it as informative.
>> No, I guess we ought to say something, but what?
>> Jo
>> On 10/11/2008 20:40, Phil Archer wrote:
>>> Jo Rabin wrote:
>>>> OK folks, here is another shot at it. Machine readable claims back in.
>>>> Still needs to be aligned with the license. Some bits missing still 
>>>> too.
>>>> http://www.w3.org/2005/MWI/BPWG/Group/Drafts/mobileOK-Trustmark/20081110 
>>>> I think Phil wanted to have some aspects of the POWDER stuff 
>>>> changed, but regret that I can't find his email on the subject, 
>>>> despite frantic Google Desktop searching.
>>> Jo, let me put you out of your misery:
>>> The example in the version you've posted was written pre-TPAC. Now 
>>> that we have the URI of the logo, and now that I've put the relevant 
>>> POWDER doc through the spill chucker, the example at:
>>> http://philarcher.org/powder/20081110.html#eg5-3
>>> is correct (this version of the doc has now been handed over to our 
>>> team contact (Matt) for publication so this is very much a temporary 
>>> URI and SHOULD NOT normally be referenced!).
>>> Two more things:
>>> 1. No, there's no linkage to the RDF - the example triple is just a 
>>> triple that could occur in any RDF data set. It would be unlikely to 
>>> exist on its own and you wouldn't link to it as such.
>>> 2. Do you really want to omit mention of HTTP Link here?
>>> http://philarcher.org/powder/20081110.html#httplink. We plan to leave 
>>> it in the document even if we end up having to flag it as informative.
>>> Apart from that, I'm happy.
>>> HTH
>>> Phil.
>>>> Ref Francois's comments:
>>>> 1. OK
>>>> 2. That's not clear to me from reading the license.
>>>> 3. Think we need to in view of section 2. reading "Claiming mobileOK 
>>>> ..."
>>>> 4. I still don't understand why a machine testable claim needs to be 
>>>> included to reveal a machine testable condition. Further, ref CT, 
>>>> the claim says that a mobileOK representation is available at the 
>>>> URI, not that this representation is mobileOK. So I am not clear 
>>>> that this is very useful in that context either.
>>>> 5. No that wasn't what I meant really. More that I think that the 
>>>> world could possibly be a better place if we had a way of sites 
>>>> expressing their site map and including where various sorts of 
>>>> mobile content are to be found. The point being that mobileOK 
>>>> content is just one sort of mobile friendly content, and you might 
>>>> be looking for a more advanced experience. Per discussion on CT list 
>>>> we need to develop a vocab to do that. Telling folks to put a 
>>>> machine readable mobileOK claim on and then later telling them to do 
>>>> it in a different way, possibly not that helpful.
>>>> Jo
>>>> On 07/11/2008 17:46, Francois Daoust wrote:
>>>>> A couple of "IMO" thoughts:
>>>>> 1. mobileOK Scheme is to be published as a Working Group Note, so 
>>>>> for once I would not worry too much about making a reference to a 
>>>>> not fully existing POWDER spec.
>>>>> 2. The license draft specifies the conditions that must be met to 
>>>>> be allowed to use the trademarked "mobileOK®" string and the 
>>>>> trademarked mobileOK logo to claim conformance to mobileOK. The 
>>>>> claim could be made on paper, on a bus, on some other page, 
>>>>> whatever. There may be other claims that don't make use of this 
>>>>> trademarked material.
>>>>> In particular, there is no trademark on the machine-readable claim, 
>>>>> I don't think there can be one, and I don't think that's necessary 
>>>>> to be able to go after someone mis-using the 
>>>>> "http://www.w3.org/2008/06/mobileOK#conformant" URI. But a legal 
>>>>> view on that could be helpful.
>>>>> 3. About using the "mobileOK®" string, I don't think it is required 
>>>>> that we define it in the mobileOK Scheme document, but equally 
>>>>> agree that it looks odd that all possibilities mentioned in the 
>>>>> license do not show up in the mobileOK Scheme doc.
>>>>> 4. About telling whether this will be used by anyone, I agree with 
>>>>> Phil. There were some use cases listed in previous drafts. Whether 
>>>>> any of them will actually be put into practice is difficult to say. 
>>>>> It could be useful, which is what I think is important. I'd say we 
>>>>> should stay silent on this and focus on defining ways to claim 
>>>>> mobileOK conformance. It could have a direct use in the Content 
>>>>> Transformation Guidelines. Plus people tend to enjoy saying they 
>>>>> conform to this and that, so we'd better specify means to make this 
>>>>> possible.
>>>>> 5. about wrapping the mobileOK claim into something of more general 
>>>>> utility, I understand it as referring to the "aspirational" level 
>>>>> we've been talking about. I don't remember: did we ever resolve to 
>>>>> drop it? I think it's, in any case, something that is distinct from 
>>>>> the "real" mobileOK claim, and that it would need a different logo, 
>>>>> string, and/or machine-readable assertion.
>>>>> In short, I second Phil's proposal: i.e. the same document as the 
>>>>> latest one completed with the previous sections 2.1, 2.2, 2.4 for 
>>>>> the RDF vocabulary class and POWDER reference.
>>>>> Francois.
>>>>> Jo Rabin wrote:
>>>>>> On 07/11/2008 10:59, Phil Archer wrote:
>>>>>>> Jo Rabin wrote:
>>>>>>> [..]
>>>>>>>> I anticipate some proposed text from the Member from Suffolk 
>>>>>>>> addressing re-insertion of references to machine readable claims. 
>>>>>>> See e-mail sent late last night: 
>>>>>>> http://lists.w3.org/Archives/Public/public-bpwg/2008Nov/0010.html
>>>>>> Thanks! (Makes mental note to read last night's email before 
>>>>>> sending out today's email)
>>>>>>> Both you and he
>>>>>>>> are welcome to join the little task force.
>>>>>>>> In this context I need to say that the organisation I represent 
>>>>>>>> has no plans to use the logo under license - so I think I have 
>>>>>>>> now reached the point where I need to understand
>>>>>>>> a) what the use cases are, as I mentioned above
>>>>>>>> b) that this is really going to get used in some way by content 
>>>>>>>> providers
>>>>>>>> c) that some search engine somewhere is planning to look for 
>>>>>>>> machine readable labels.
>>>>>>> Here's what one rather large search engine told me on a recent 
>>>>>>> visit to Silicon Valley: You create the data, make sure it's not 
>>>>>>> full of spam, and we'll use it. But don't expect us to make a 
>>>>>>> public statement on the issue to help you on your way.
>>>>>>> In other words, if you want a search engine or two to make a 
>>>>>>> statement about the usefulness of mobileOK, or any other 
>>>>>>> machine-readable label, 
>>>>>> I think there is a difference between mobileOK and general machine 
>>>>>> readable labels. Today (and remember that things have moved on 
>>>>>> since we started discussing this three and a half years ago) 
>>>>>> mobileOK contains only machine testable aspects. So a search 
>>>>>> engine that is truly interested in whether a site is mobile 
>>>>>> friendly is likely to test it. Ours (find.mobi) does at least.
>>>>>> To my mind this is rather different from using labels to express 
>>>>>> judgements that cannot be determined by a machine other than by 
>>>>>> reading the label. To me this is what labels are useful for and 
>>>>>> naturally I comment your and POWDER's work in this very important 
>>>>>> area.
>>>>>> If we still had a non machine testable aspect to mobileOK then I 
>>>>>> think the labelling stuff would be very useful. As things stand 
>>>>>> today I think labelling is at best moot in respect of mobileOK.
>>>>>> So far as it being used by browsers, well, from a mobile browser's 
>>>>>> perspective finding a label on content it has already retrieved is 
>>>>>> really all too late, isn't it?
>>>>>> However, on this theme, I also think that labels are going to be 
>>>>>> very useful indeed for labelling sites for the discovery of mobile 
>>>>>> friendly content - of various kinds, where mobileOK is just one 
>>>>>> very basic type of mobile friendly content. To my mind this is one 
>>>>>> of the major unresolved issues to come out of the CT work, to 
>>>>>> which it is also relevant. So perhaps it would be sensible to wrap 
>>>>>> the mobileOK claim into something of more general utility?
>>>>>> Jo
>>>>>>> before you decide to use it, you'll never decide to use it. If, 
>>>>>>> however, you really want search engines to have an easy way to 
>>>>>>> identify mobile-friendly content, and if you want to encourage 
>>>>>>> content providers along a route that ends up with more mobile 
>>>>>>> friendly content that they can advertise as such, then you need 
>>>>>>> to create the best platform possible for that to happen.
>>>>>>> It seems to me that, given the feelings expressed around this 
>>>>>>> issue and the fact that, for all our best efforts, POWDER is 
>>>>>>> going to be at PR, not Rec, next month, that including some 
>>>>>>> examples of what you MAY do to make mOK machine-readable in this 
>>>>>>> doc is a pretty basic step that we can take without upsetting the 
>>>>>>> apple cart too much or creating dependencies we could do without.
>>>>>>> Phil.
>>>>>>>> I suggest that if we are to continue this work then We really 
>>>>>>>> need to get to the bottom of this, with the idea that it all 
>>>>>>>> needs to be sorted out by Dec 1 which is the end of review for 
>>>>>>>> mobileOK Basic Tests 1.0 [MOK] (sic). Otherwise probably the 
>>>>>>>> simplest thing is to drop the idea of a license and the logo 
>>>>>>>> till we are clearer on it, and publish the scheme document just 
>>>>>>>> as a way of linking together Best Practices, mobileOK Basic 
>>>>>>>> Tests 1.0 and the Checker.
>>>>>>>> Sorry if I seem to be a bit fed up with this topic. But I am.
>>>>>>>> Jo
>>>>>>>> On 07/11/2008 09:31, Francois Daoust wrote:
>>>>>>>>> I confess I'm a bit surprised to see that the section "Claiming 
>>>>>>>>> mobileOK Conformance using POWDER" was entirely removed.
>>>>>>>>> My recollection of our discussion was that:
>>>>>>>>>  1. it is not mandatory to use a machine-readable claim to 
>>>>>>>>> claim conformance to mobileOK
>>>>>>>>>  2. people may still use the machine-readable claim, and that's 
>>>>>>>>> something we still want to promote as a good practice.
>>>>>>>>> In short, I think the document should just provide several ways 
>>>>>>>>> to claim conformance to mobileOK:
>>>>>>>>>  - the logo
>>>>>>>>>  - POWDER
>>>>>>>>>  [ - and possibly RDFa although I'm not sure that's such a good 
>>>>>>>>> idea since there's no way to embed such a claim in a mobileOK 
>>>>>>>>> representation, leading to a pretty confusing message "Use RDFa 
>>>>>>>>> to claim you're mobileOK, but not in a mobileOK page. What ?!?". ]
>>>>>>>>> Did I miss something?
>>>>>>>>> Francois.
>>>>>>>>> Phil Archer wrote:
>>>>>>>>>> Whilst I understand that this draft reflects the resolutions 
>>>>>>>>>> taken at TPAC, I would like to propose an addition to the 
>>>>>>>>>> document that at least points to the option to make the 
>>>>>>>>>> mobileOK claim machine-readable as follows.
>>>>>>>>>> 2. Further Steps
>>>>>>>>>> As the previous section makes clear, the mobileOK trustmark is 
>>>>>>>>>> an icon that may be included on any Web page that conforms to 
>>>>>>>>>> mobileOK Basic Tests. However, it is possible to go further 
>>>>>>>>>> and make the claim machine-readable using any of a number of 
>>>>>>>>>> different methods, thus making mobileOK content more readily 
>>>>>>>>>> discoverable. The Protocol for Web Description Resources 
>>>>>>>>>> [@@POWDER] includes an example of how to do this in its 
>>>>>>>>>> documentation and alternatives include RDFa and microformats 
>>>>>>>>>> (@@ link to Jonathan’s work on this)
>>>>>>>>>> Link to http://www.w3.org/TR/powder-dr/#evidence (this is due 
>>>>>>>>>> to be updated w/c 10 November, see it now at 
>>>>>>>>>> http://philarcher.org/powder/20081104.html#evidence
>>>>>>>>>> WDYT?
>>>>>>>>>> Phil.
>>>>>>>>>> Jo Rabin wrote:
>>>>>>>>>>> Hi folks, I have updated the Editor's draft of mobileOK 
>>>>>>>>>>> scheme [1] in line with the resolutions taken at the F2F. I 
>>>>>>>>>>> did not put anything in the document about sticking a date in 
>>>>>>>>>>> the ALT test for the trustmark as I don't recall that 
>>>>>>>>>>> actually being a resolution. And anyway, I don't understand 
>>>>>>>>>>> what that is supposed to represent, how you'd encode it and 
>>>>>>>>>>> why it would be useful, what effect it would have on the 
>>>>>>>>>>> correct use of ALT and so on.
>>>>>>>>>>> I'd like to close out on this by the time mobileOK Basic 
>>>>>>>>>>> Tests goes to Rec (early Dec). A couple of  further things 
>>>>>>>>>>> need sorting out on this:
>>>>>>>>>>> a) ACTION-869 - Review the mobileOK license in more details 
>>>>>>>>>>> and send further questions to rigo [on Jo Rabin - due 
>>>>>>>>>>> 2008-10-27]
>>>>>>>>>>> b) RESOLUTION: Dom, Jo and Rigo to form a subcommittee to 
>>>>>>>>>>> come back with a final proposal to the group within 4 weeks 
>>>>>>>>>>> following on from Rigo's current proposal.
>>>>>>>>>>> I'm attending to these now. Meanwhile would appreciate 
>>>>>>>>>>> comments on the latest draft.
>>>>>>>>>>> thanks
>>>>>>>>>>> Jo
>>>>>>>>>>> [1] 
>>>>>>>>>>> http://www.w3.org/2005/MWI/BPWG/Group/Drafts/mobileOK-Trustmark/20081106 
Received on Tuesday, 11 November 2008 18:01:41 UTC

This archive was generated by hypermail 2.4.0 : Friday, 25 March 2022 10:09:52 UTC