My comments on the Best Practice document

Hello,

I thought this would be a good time to read the Best Practices document and
to share comments, see below. I am afraid it not a small list, but on the
plus side the numbers are not auto-generated. Anyway, I hope it helps.

Oh, and I should not forget my main comment: Lots of respect for the
editors! Finding and describing best practices for spatial data on the web
is a daunting task, but you are doing it anyway and it seems to be going
really well.

Regards,
Frans

1) (Glossary)
Coverage: I notice different definitions being used in the coverage
subgroup. Is  a coverage a function or a data structure? Why is a geometry
describing the shape of a thing not a coverage? Why are space and time so
special that they need a special data structure for multidimensional data
when one of them is involved?
CRS: the definition is about geography. It should be broader. Also include
the alternative term SRS (spatial reference system).
Extent: In common GIS parlance the extent of a spatial thing can be a
bounding box, which is not the same as the area covered by something.
Spatial data: A point is a type of spatial data, but does not have a
spatial extent.
Spatial thing: The same comment as above. Perhaps ‘spatial extent’ needs
its own glossary item? I used the term ‘spatial presence’ as something
similar. Perhaps such a phrase makes it easier to understand that
one-dimensional and two-dimensional thing are also spatial things? Also:
does having an extent means having a limit or boundary? Some spatial
phenomena are fields that can be considered to exist everywhere.
SDI: Misses the complete term (spatial data infrastructure). I wouldn’t
call it an ecosystem, because there is little biology involved. Perhaps
‘compound system’ is a better term?

2) The first three sections start with the comment “This section is
non-normative”. Does that mean the other sections are normative?

3)  (Introduction)The first sentences of the introduction are “Increasing
numbers of Web applications provide a means of accessing data. From simple
visualisations to sophisticated interactive tools, there is a growing
reliance on data.”. That seems to imply that the best practices are
intended for usage of data on the web in web applications, which probably
most people understand to be an application running in a web browser.
However, web data can be used outside of web browser applications, and it
is likely this type of usage will only increase if the quality and quantity
of web data increase.

4) (Introduction) “Location information, or spatial data, is often a common
thread running through such data; describing how things are positioned
relative to the Earth in terms of coordinates and/or topology.” For
contrast with what comes next it would be better to use the term
“geographic data”.

5) (Introduction) I really like the statement “The key problems we are
trying to solve in this document are discoverability, accessibility and
interoperability.” Those three things are very important. Perhaps the words
discoverability, accessibility and interoperability could be emphasized?

6)  (Scope) Typos:
“That describe how spatial data in commonly published and used on the Web”
 “Discussion of activities relating to rending spatial data as maps is
explicitly out of scope.”

7) (Spatial Things, Features and Geometry) “Looking more closely, it is
important to note that geometry is a property of a spatial thing.”: I would
rather say a geometry *can* be a property of a spatial thing. Geometries
can also exist by themselves, i.e. unrelated to spatial things.

8)  (Spatial Things, Features and Geometry) “We say that the spatial thing
is disjoint from the geometry object.” Disjoint is a specialist term from
set theory. Is its usage appropriate, considering the audience?

9) (Coverages: describing properties that vary with location (and time)):
This section could do with an explanation why multidimensional data
structures that have space or time in their domain justify a separate class
of data structure. Is it because agreed upon definitions of data types for
spatial and temporal data do not exist yet (that would be my guess)?

10) (Spatial relations): “Topological relations describe the relationships
between geometric objects…”. I think spatial things can have topological
relations too.

11) (Spatial relations): Perhaps add something about how one-dimensional
spatial relations differ (or do not differ) from temporal relations?

12) (How to use these best practices): This section suddenly takes the data
publisher perspective. Would it be an idea to separate sections for supply
and usage? Guidance on how data consumers (developers) could use the best
practices seems to be missing now.

13) (9.3 Who is your audience?) “It is likely that you will be able to
identify your intended “community of use”. In many cases this is not the
case. More and more data are published just for the sake of
sharing/sharing/transparency. Perhaps a minor detail, but I feel that a
phrase like ‘It could be that you have identified a “community of use”’ is
better.

14) (BP1) ‘spatial coverage’ could be confusing, because the word
‘coverage’ is used differently than elsewhere in the document. Change to
'spatial exent'?

15) (BP1) “which area of the world the data is about”: this seems to imply
geographical data only. The data could be about Mars or a different galaxy
too.

16) (BP1) Some spatial metadata could apply only to subsets or to certain
attributes. For example, if spatial things are described by different types
of geometries, some metadata only apply to those different types of
geometry. Some extra guidance seems needed.

17) (BP2, CRS note) The note seems to be only about geographical data.

18) (BP2, CRS note) “CRS can provide a map projection and also define the
transformations between different spatial reference systems” Very
confusing. Is CRS not the same as a SRS?

19)  (BP2, CRS note) 'SRID' is undefined.

20) (BP2, CRS note) “By default, most vector data should be shared in
un-projected geographic coordinates WGS 84– SRID 4326.“ This is a strange
guideline. It contradicts other guidelines, such as publishing data in
their original, unprocessed form, and the European guideline to use the
ETRS89 CRS for unprojected geographic coordinates.

21) (BP3) I think the CRS should always be specified. Numerical coordinate
data are rather useless without knowing the CRS.

22) (BP3) The question how data will be used (intended application) should
not be a decisive factor, see also section 9.3. Data on the web should be
made available in a generally good way. That way should best for foreseen
and unforeseen usage.

23) (BP3) I think the BP should say that coordinate data should at least be
made available in the CRS  in which the data where produced.

24) (BP3) WGS84 could be recommended as a extra CRS to use next to the
original CRS, but in that case temporal metadata can not be omitted. WGS84
is a versioned CRS.

25) (BP3) If coordinate data are made available in non-original CRSs, the
transformation procedure should be specified in the metadata.

26) (12.5 Spatial Data Vocabularies) “In this document there is no section
on formats for publishing spatial data on the web.”: I think this is an
area in which guidance is very much needed. If you only look at possible
formats for geographic data there is a lot to choose from. Many more
formats for spatial data exist outside the domain of geography. Having all
those different formats is very hurtful for interoperability, and
discoverability and accessibility too. Perhaps there is no best practice,
but some formats could be viewed as better than others. Of course it is a
very difficult issue, which hopefully our new spatial ontology will help
solving, but the world really needs a common data type for geometry. The BP
document should address the issue and provide as much guidance as possible,
I think.

27) A general remark about best practices: for some requirements there is
no good practice yet. Would it be worthwhile to identify gaps between what
is needed and what is available? Some awareness about the quality of
certain practices could caution practitioners  and help the world in making
work of improving those areas.