Fwd: Spec review request: CSV on the Web

Hi all,

For the record here was my reply to the call for feedback.
The call from Dan can be read when scrolling down ;-)

Cheers,
Christophe

---------------------------------------------------------------------------------------------------

# http://www.w3.org/TR/2015/WD-csv2rdf-20150416/

* 4.4 -> what about pointing to a DCAT resource instead ? That resource
could in turn be associated to a distribution file. Would surely be more
complex that the proposed solution but would sound more integrated with
other vocabs / BPs.

* If 4.2 is normative I'm somewhat uncomfortable with "4.6.1 In standard
mode only, establish a new blank node R which represents the current row"
because if I would like not to use a BNode because I want dereferencable
URIs I would not be compliant. Otherwise, if I'm free to use a BNode or
something else everything is fine :-)

* 4.3 -> as for JSON it could be useful to define a resource for JSON-LD
too, just in case it's used in the tables.

* The notion of a 'standard' and 'minimal' modes sounds to me like an
unnecessary added level of conceptual complexity. Maybe you switch that to
something closer to what JSON-LD do: the minimal rendering is Web tabular
data, the standard rendering is Web tabular too just a bit more detailled.
This approach could also be taken for presenting the document, starting
with some simple examples using minimal conversation and then moving to
more elaborated one with the standard descriptions.

# www.w3.org/TR/2015/WD-csv2json-20150416/

* A bit surprising to find "5. JSON-LD to JSON" here as the abstract reads
"This document defines the procedures and rules to be applied when
converting tabular data into JSON.". But "6. JSON-LD to RDF" in
http://www.w3.org/TR/2015/WD-csv2rdf-20150416/ is also slightly
miss-located. At least, the documents those sections are in are not the
first ones I would consult to find this conversion algorithms. Considering
the fact that the two drafts csv2rdf and csv2json share a lot of concepts
and contain more that their title announces I would suggest making one main
document "X to Y" pointing to several specific docs. That main document
would list the conventions and have a table like :

     | CSV       | JSON-LD
--------------------------------
RDF  | rdf->csv  | rdf->json-ld
JSON | json->csv | json->json-ld

with each of the cell pointing to a specific document with the relevant
content and an extra set of columns/rows for XML. That table itself could
even be published using CSVW ;-)

# http://www.w3.org/TR/2015/WD-tabular-metadata-20150416/

* 5.1.2 : Example 6 alone is not equivalent to example 7, it's the
combination of 5 and 6 that leads to 7

* In 5.1.3 the naming "http://example.org/example.csv#row.1" is different
than the, eg, "#row=2" used in
http://www.w3.org/TR/2015/WD-csv2rdf-20150416/ . Not an issue as this is an
example showing what can be done (right ?) but this could be confusing.

* 5.11, as said earlier maybe good idea to add a datatype for json-ld ?

# http://www.w3.org/TR/2015/WD-tabular-data-model-20150416/

* Just generally wonder about counting things starting from 1 and using
arrays which naturally count from 0. Starting all the counters for row
reference etc from 0 could be more consistent.

* A.2.3 "Excel provides more control when importing CSV files into Excel.".
Excel surely does not control much things when importing CSV into
LibreOffice ;-) so "Excel provides more control when importing CSV files"
would suffice. Speaking of LibreOffice(/OpenOffice) I think it would
deserve a A.5 as it is used by several public administrations as an open
source alternative to Excel.









On 20 April 2015 at 13:42, Dan Brickley <danbri@google.com> wrote:

> The CSV on the Web Working Group [1] has just published a new set of
> Working Drafts, which we consider feature complete and implementable.
> We particularly
> seek reviews from Web Security, Privacy, Internationalization and
> Accessibility
> perspectives at this time. A request has also been sent to the TAG
> [7]. We request
> review now rather than later since we are following W3C's revised Process
> in
> which there is no distinct Last Call; we prefer to invite reviews now
> rather than wait for a formal Candidate Recommendation.
>
> The drafts are:
>
> Model for Tabular Data and Metadata on the Web [2]
> - an abstract model for tabular data, and how to locate metadata that
> enables
> users to better understand what the data holds; this specification also
> contains non-normative guidance on how to parse CSV files.
>
> Metadata Vocabulary for Tabular Data [3]
> - a JSON-based format for expressing metadata about tabular data to inform
> validation, conversion, display and data entry for tabular data
>
> Generating JSON from Tabular Data on the Web [4]
> - how to convert tabular data into JSON
>
> Generating RDF from Tabular Data on the Web [5]
> - how to convert tabular data into RDF
>
> We are keen to get comments on these specifications, either as issues on
> our GitHub repository [6]  or by
> posting to public-csv-wg-comments@w3.org.
>
> We would also like to invite people to start implementing these
> specifications and to donate their test cases into our test suite.
> Building this test suite, as well as responding to comments,
> will be our focus over the next couple of months.
>
>
> Dan
>
>
>
> [1] http://www.w3.org/2013/csvw/wiki/Main_Page
> [2] http://www.w3.org/TR/2015/WD-tabular-data-model-20150416/
> [3] http://www.w3.org/TR/2015/WD-tabular-metadata-20150416/
> [4] http://www.w3.org/TR/2015/WD-csv2json-20150416/
> [5] http://www.w3.org/TR/2015/WD-csv2rdf-20150416/
> [6] https://github.com/w3c/csvw/issues
> [7] https://lists.w3.org/Archives/Public/www-tag/2015Apr/0028.html
>
>


-- 
Onderzoeker
DANS, Anna van Saksenlaan 51, 2593 HW Den Haag
+31(0)6 14576494
christophe.gueret@dans.knaw.nl


*Data Archiving and Networked Services (DANS/KNAW)*[image:
http://dans.knaw.nl] <http://dans.knaw.nl>

*e-Humanities Group (KNAW)*
[image: eHumanities] <http://www.ehumanities.nl/>

*World Wide Semantic Web community*
http://worldwidesemanticweb.org/





-- 
Onderzoeker
DANS, Anna van Saksenlaan 51, 2593 HW Den Haag
+31(0)6 14576494
christophe.gueret@dans.knaw.nl


*Data Archiving and Networked Services (DANS/KNAW)*[image:
http://dans.knaw.nl] <http://dans.knaw.nl>

*e-Humanities Group (KNAW)*
[image: eHumanities] <http://www.ehumanities.nl/>

*World Wide Semantic Web community*
http://worldwidesemanticweb.org/