Data on the Web Best Practices Working Group Teleconference

<newton> Good morning

<deirdrelee> hi

<scribe> scribe: Caroline_

<phila> scribe: Caroline_

<phila> scribeNick: Caroline_

<phila> PROPOSED: Accept last week's minutes https://www.w3.org/2016/04/15-dwbp-minutes

+1

<BernadetteLoscio> +1

<PWinstanley_> +1

<newton> +1

<deirdrelee> 0

<annette_g> +1

<laufer> +1

<phila> +1

RESOLUTION: Accept last week's minutes https://www.w3.org/2016/04/15-dwbp-minutes

deirdrelee: we will focus on BP document this week

issues from recent comments

<phila> Issue list

BernadetteLoscio: we would like to thank everyone for the feedback and detailed review
... thank you a lot!
... w the editors discussed the comments and we created this table to track the comments and have proposals and resolutions for each comment https://www.w3.org/2013/dwbp/wiki/Comments_to_be_considered_before_publishing_the_last_working_draft
... some of the comments have a resolution and the corresponding commits are there
... we still have a lot of comments we need to address
... we have some questions for the group that will help us to continue updating it

deirdrelee: there were discussions around timeframe last week
... the idea is when these comments are closed we can freeze the document?

<Zakim> phila, you wanted to talk about CR/LC

BernadetteLoscio: yes, because we had the document frozen until Wednesday night

phila: we will vote on a last draft
... we should resolve our comments internally before publishing it

BernadetteLoscio: Carol and Newton if want to add something just interrupt me :)

<BernadetteLoscio> Should we write subtitles using the imperative mode?

BernadetteLoscio: our first comment is a question: Should we write subtitles using the imperative mode?

<BernadetteLoscio> https://docs.google.com/spreadsheets/d/1eSTt3A6kTfXYTcVMt5VGDardLIk8b7FnsgENpCuNRBA/edit#gid=0

https://docs.google.com/spreadsheets/d/1eSTt3A6kTfXYTcVMt5VGDardLIk8b7FnsgENpCuNRBA/edit#gid=0

scribe: this table has all the BPs and their subtitles
... if we have to change the subtitles using the imperative mode we will need help from the Native Speakers to review all of them

<Zakim> phila, you wanted to ask what "Metadata must be provided for both human users and computer applications" becomes

<phila> Metadata must be provided for both human users and computer applications.

phila: regarding this example above
... if you take out the word must

<annette_g> yes, that's correct

<phila> SO it becomes Provide metadata for both human users and computer applications.

BernadetteLoscio: it would be "provide metadata for both human users and computer applications
... some BPs will need to have something added so won't be the same as the title

phila: if that is the case my suggestion might be that if the title itself is already nice and short don't bother with the subtitle at all

BernadetteLoscio: or we could say provide data provenance for both humand and computer applications

annette_g: there is probably something we can say for each BP that will be useful for the users

BernadetteLoscio: I think the subtitles should be short
... because if it is long it would be almost as the "Why"

deirdrelee: all subtitles will be reviewd but they will be capt short

BernadetteLoscio: we need to have a Native Speaker review again
... before puting in the document

<annette_g> +1 for native speaker review. I'm willing to help.

<phila> annette_g - how about you and me split the doc and we do the new subtitles?

<annette_g> sure

BernadetteLoscio: please help us changin this table https://docs.google.com/spreadsheets/d/1eSTt3A6kTfXYTcVMt5VGDardLIk8b7FnsgENpCuNRBA/edit#gid=0

thank you annette_g :)

BernadetteLoscio: let's use this table and then when it is finished we put on the document

phila: annette_g and I will do it

thank you phila and annette_g :)

<BernadetteLoscio> http://w3c.github.io/dwbp/dwbp-example.html

BernadetteLoscio: about the human-readable example (should we include the example in the doc or keep it as a separate file?)

<BernadetteLoscio> http://w3c.github.io/dwbp/bp.html#DataLicense

BernadetteLoscio: when there is human readable example, for example in BP 5 we link to the example http://w3c.github.io/dwbp/dwbp-example.html

<phila> ACTION: phila to provide imperative subtitles for BPs 1-19 [recorded in http://www.w3.org/2016/04/22-dwbp-minutes.html#action01]

<trackbot> Created ACTION-273 - Provide imperative subtitles for bps 1-19 [on Phil Archer - due 2016-04-29].

BernadetteLoscio: it was not clear that we were showing data license, or the information about location parameters

<phila> ACTION: annette to provide imperative subtitles for BPs 20-37 [recorded in http://www.w3.org/2016/04/22-dwbp-minutes.html#action02]

<trackbot> Created ACTION-274 - Provide imperative subtitles for bps 20-37 [on Annette Greiner - due 2016-04-29].

BernadetteLoscio: the suggestion was to include the human readable directly in the document
... I am not sure it is a good idea
... newton put links to each fragment
... if you want to see the structural metadata the link we will take you there
... we would split the HTML if we put them in the document
... do you the ink we should put just the HTML text in the doc?

annette_g: I like the metadata document
... there are a few lines that you can't find
... I would suggest that ??
... maybe put something bold in the BP doc itself

<newton> 1+

BernadetteLoscio: for the structure metadata we could keep in the HTML, for example

annette_g: when it is a bigger text you can keep the HTML

<Zakim> phila, you wanted to make a slightly diff suggestion

phila: I am happy with the HTML
... another alternative would be to have a table
... which BP is relevant in the HTML example
... that would be more work
... I am happy with your suggestion

BernadetteLoscio: the idea is to show how it would look after you finished

newton: I like phila suggestion

phila: newton if you want to try it you could do annotation, light box, etc
... it would take more time
... you would enjoy doing but it would take time

newton: I will try doing it and I will share with you
... we could point to the URL fragment the specifc texts

phila: sometimes that doesn't work

newton: I will try to change

annette_g: you can refer the link
... it might be more geek

BernadetteLoscio: talking about the description of the example
... Discuss basic example. Should we change to include other transito modes?
... we have to be careful
... I don't know how to rewrite it in a way that is short
... that we would have example considering the transit mode

annette_g: we just have to not limite it when we first describe it

BernadetteLoscio: are you talking about the description of the example?

annette_g: if the description is only abobut buses we will limit it

BernadetteLoscio: we will change the first line than
... instead of buses stops what do you suggest?

annette_g: we can say transit stops instead of bus stops

BernadetteLoscio: when you go to the example it is only for bus stops

<phila> make it 'Transit stops' not just 'bus stops' (all bus stops are transit stops)

BernadetteLoscio: the example for everything

annette_g: some examples need to be more general
... I would say "he is in charge of publishing data about the transit system"

<newton> yep

BernadetteLoscio: the editors will discuss that and see if could be more general
... What should we do with the BP about Content Negotiation? Shoul we keep it or move to BP 14?
... about providing multiple formats
... we think we should keep the BP about content negotiation

annette_g: as a BP I don't think it is clearly a BP to always use content negotiation

<newton> we could link them, instead of merging them

annette_g: among developers it is not at all

phila: what are the main objections?

annette_g: the main one is that it doesn't allow someone to share the link
... you have to tell them taht you use content negotiation

phila: you give people a directly serialization
... if you take of the file extension

<laufer> +1 to annette claim that it is an implementation issue

annette_g: if people want to do content negotiation let them avoid poiting URI
... can you point to any place that is actually doing that?

phila: yes, we do all the time
... our namespace doc will do it

newton: is very useful when you are requestiong the specific resource

<phila> -> https://www.w3.org/ns/org, also available https://www.w3.org/ns/org.rdf, https://www.w3.org/ns/org.ttl

newton: we can likn the BPs and put in the content negotiation
... if you can provide the specifc URL

<phila> And https://www.w3.org/ns/org.n3 if you want the triples

annette_g: my reading on the subject has not came up yet

<BernadetteLoscio> maybe we can do the other way around in BP14 we make a link to BP21

deirdrelee: annette_g do you want to complete remove the BP?

<phila> And elsewhere data.ordnancesurvey.co.uk/doc/postcodeunit/IP83PX (add .ttl, xml, .json)

annette_g: I think we could leave it there, I just want to know if there are people doing it

deirdrelee: maybe puting more examples

<laufer> it is a way of providing multiple formats

newton: a lot of RDF stars like virtuoso and Jena Fusek use content negotiation

deirdrelee: even the extension for DCAT

<PWinstanley> http://statistics.gov.scot/data/social-work-staffing.ttl

newton: I think it is a good practice for the machine

<laufer> it is an implementation issue...

deirdrelee: the human readble it is also used

annette_g: let's go then

BernadetteLoscio: we will keep the BP then
... we will make a link from BP 14 to BP 21
... then I have two questions related
... should we say the multiple access should be availvabe..

annette_g: no

BernadetteLoscio: in our BPs we say that we should have bulk donwload and APIs
... aren't they multiple access mechanisms?
... when I have a dataset I will have just one way to access the dataset?

<laufer> we have two things here... what and how

annette_g: I need to be convinced that I need to take to trouble to do both ways

<laufer> what to get... and how to get...

annette_g: the reason to do multiple formats is that people will have more than one format
... you have that trhough the API

BernadetteLoscio: I am not saying this is a BP
... it is just the description

annette_g: then you don't have to use the word should

BernadetteLoscio: the problem is with should?

<BernadetteLoscio> multiple dataset access mechanisms can be available

BernadetteLoscio: if we say datasets and mechanisms "can" be available

laufer: I think we are talking about 2 things
... how to get and what
... when we have bulk donwload someone has to give some time of data
... or how will you provide this?
... 2 things: what to get and how to get

<Zakim> phila, you wanted to talk about timing

deirdrelee: are you satisfied with annette_g's proposal sayhing that the word will change for can instead of should?

<laufer> i a having problems with audio... sorry

<laufer> its ok for me

phila: we just began the few questions
... how are we going to handle the other discussions?
... because we are having a great discussion

deirdrelee: if we continue the discussion next week we might have infinite looping
... we should make a timeframe to address them

BernadetteLoscio: we have several questions to annette_g so maybe we can have a chat with her today or monday to clarify these questions
... also we need help to rewrite some BPs
... and maybe this is a task to annette_g because a lot of questions were from her about data access
... since annette_g is the one who really knows about data access I would like to know if she is okay with dealing with her comments on that
... it would be more usefull if you may change

annette_g: I gave you some comments, do you want to talk on skype?

BernadetteLoscio: yes, thank you
... some comments we can do the update
... but for the data access would be more productive if you could work on your comments since you know that subject better
... we will talk later then :)
... What should we do with the BP Use Standardized Terms and Reuse Vocabularies? Should we merge? (see Antoine's message)
... Antoine mentioned we could merge these BPs
... I am not sure if would be clear if we keep as it is
... if we decide to merge them we would have to rewrite

<deirdrelee> http://w3c.github.io/dwbp/bp.html#MetadataStandardized

BernadetteLoscio: BPs 15 and 16

<deirdrelee> http://w3c.github.io/dwbp/bp.html#ReuseVocabularies

<BernadetteLoscio> http://w3c.github.io/dwbp/bp.html#MetadataStandardized

BernadetteLoscio: that we are talking about
... we need Antoine's feedback to know if he agrees with the BPs
... I am affraid that this might not be enough to clarify the difference between the 2 BPs
... instead of writing a new BP

annette_g: mostly keeping BP 15 and adding BP 16 to be part of that

phila: it makes sense to me

<laufer> +1

phila: maybe Antoine has something else to say

annette_g: the sense I get from reading Antoine's email is taht he agrees we could merge them

BernadetteLoscio: we need his feedback

phila: if you are editors able to spend time with annette_g and Antoine I might join and we see how far we can get before next week

BernadetteLoscio: we will work on this BP and try to talk with Antoine as well

phila: this discussion is very important
... we are having a detail discussion

<annette_g> +1 to phila, I'm really excited to see things shaping up1

BernadetteLoscio: the group is very mature and it is easier to talk about these things
... deirdrelee, for us editors is really hard to give you a dealine before talking with annette_g and doing more things

deirdrelee: we can have another talk next week
... I will be more available this week as well to help with the native English

BernadetteLoscio: we will try to finish everything before Wednsday

<annette_g> many thanks to the editors for all their hard work!!!

<BernadetteLoscio> :):):)

<laufer> bye all...