Re: comments on UCR

Hi Rachel & all,

On 06/02/15 15:08, Heaven, Rachel E. wrote:
> Frans, Alejandro
> Thank you for all your work on this. My comments on the UCR document:
>  
> 5.6 CRS definition, 5.8 Default CRS
> If the URI of the CRS must always be specified, then there won’t be any cases
> when a default CRS would be needed. Is 5.8 redundant now ? Or does it need to
> be re-worded to make it clearer in which cases this would be applicable ? It
> looks a bit contradictory to me as it is.
>  
> Missing requirements?:
> As noted by Phil below, I think a requirement along the lines of “It should be
> possible to request a subset or slice of a coverage dataset” is missing.
> Similar to 5.35 [Reference data chunks] but this is about getting hold of the
> data not just having an identifier for it.
> From use cases 4.18 , 4.45,  4.37 (and 4.2?)

couldn't agree more - it is a key property of coverages that they typically are
"too large to move", so extraction along space (and time, where there is a time
dimension) should be possible. Additionally, it should be possible to extract
bands ("channels", "variables") from an image - consider a 36-band image where
you are just interested in RGB, or the infrared channel.

Also the format question: someone wants to get a small JPEG / PNG just for
display, someone else wants to analyze data and needs NetCDF.

An advanced service might deliver a coverage (cutout) in the client's favourite
coordiante reference system. Example: Data are stored in Germany's old
Gauss-Krüger strip 5 (we have several strips!) and we want to see it in WGS84.

OGC WCS (soon also becoming ISO WCS and INSPIRE WCS) offers such functionality
in a protocol-independent manner (supporting GET/KVP, XML/POST, SOAP; REST is
practically finished and JSON under discussion).
See http://en.wikipedia.org/wiki/Web_Coverage_Service

>  
> I also think we are missing a requirement along the lines of “It should be
> possible to combine datasets, selecting the most appropriate where coverage
> overlaps”.  Or to break this requirement down we could extend the data
> subsetting requirement (and/or 5.35) so that a user can create and reference
> excluded subsets of a dataset ie the data chunk can have holes in it.
> From use cases 4.18, 4.32, 4.34, 4.35

Indeed combination / processing (whatever we want to call this) is relevant. Use
case: I want to get the vegetation index (NDVI) as one image, rather than
downloading red and near-infrared bands and computing NDVI on the client.

BTW, datasets to be combined might sit at different URLs. Use case: combine
weather data from ECMWF/UK with reprocessed Landsat of NCI/Australia (this is
what we are going to do in EarthServer-2).

In any case, it is a substantial advantage for the client to ship a query,
rather than a dataset. In the extreme case, only statistical summaries (use
case: percentage of crop vegetation) wander over the Web.

All this is the realm of the Web Coverage Processing Service (WCPS) standard.
See http://en.wikipedia.org/wiki/Web_Coverage_Processing_Service

cheers,
Peter

>  
> Use case cross references:
> Use case 4.26 could also relate to 5.4, 5.7
>  
> Typos:
> Use case 4.18
> New line and bullet point has not been transferred from the use case working
> document, needs inserting just before “to represent complex geology (folding,
> faulting, intrusions”
>  
> Use case 4.26
> s/tatic data/static data
>  
> 5.15, 5.16
> s/it should possible/it should be possible
>  
> 5.49 (Linda has already raised the other change on this line)
> s/It should be possbible for Coordinate/It should be possible for Coordinate
>  
> 5.56
> s/It should be possible to valdidate/It should be possible to validate
>  
>  
> Best wishes,
> Rachel
>  
> -----Original Message-----
> From: Phil Archer [mailto:phila@w3.org]
> Sent: 01 June 2015 23:12
> To: SDW WG Public List; Frans Knibbe | Geodan; Alejandro Llaves
> Subject: Editorial amendments to UCR
>  
> Frans, Alejandro,
>  
> I've been through the UCR today to make editorial changes of two types:
> - native speaker edits;
> - simplifying the spelling for our American friends (they do get so upset with
> metres and optimisations).
>  
> I've also added a skeleton Acknowledgements section - to which you may or may
> not choose to add specific names.
>  
> I've sent a pull request that you may or may not wish to accept.
>  
> In doing this, I also have some more substantive comments (below).
> *None* of these, IMO, should be a brake on publishing an FPWD of the doc,
> they're just links between UCs that came to mind as I read through them all
> (yes, I read the doc from start to finish!).
>  
> Not all the related requirements seem to show up. This might be a ReSpec
> thing, but it might be more serious. For example, Locating A Thing has
>  
> <p class="relatedRequirements"><a href="#TimeDependentCRS"></a></p>
>  
> But the text isn't being written into the hyperlink. Can you check these
> through please?
>  
> use Cases 4.7 and 4.8 (your two Frans) perhaps relate to Andrea's one on the
> GeoDCAT-AP as well?
>  
> 4.9 seems to relate to Ed's 4.6
>  
> 4.9 also seems to relate to 5.45 and 5.51
>  
> 4.10 refers to identifiers. The DWBP's BP Doc has a section on this - that I
> have an action item to improve in the coming 24 hours or so. That
> *might* be enough for SDW but time will tell.
>  
> 4.14 seems to call for things like ID management, privacy etc. ?
>  
>  
> 4.15 has this:
> * Agreed-upon vocabulary for metadata about spatial datasets
>  
> Which seems to relate to 4.7, 4.8 and Andrea's GeoDCAT one.
>  
> 4.16 seems to call for very similar issues as 4.15. might they be combined?
>  
> 4.18 has:
> * user can subset the data by x,y,z limits
>  
> which looks like Jeremy's UC (4.2) ? See also 4.35 and 4.37
>  
> 4.22 again looks like it relates to 4.7 and 4.8
>  
> 4.24 looks really interesting - but what's the spatial angle?
>  
> 4.28 looks very similar to Manolis's work on Greek forest fires - can they be
> combined do you think?
>  
> 4.33 Seems to call for detailed provenance info - might that be a new req?
>  
> 4.47 sounds like a whole new WG!
>  
> HTH
>  
> Phil.
>  
> -- 
>  
>  
> Phil Archer
> W3C Data Activity Lead
> http://www.w3.org/2013/data/
>  
> http://philarcher.org
> +44 (0)7887 767755
> @philarcher1
>  
>
> _  ________________________________  _
> This message (and any attachments) is for the recipient only. NERC is subject
> to the Freedom of Information Act 2000 and the contents of this email and any
> reply you make may be disclosed by NERC unless it is exempt from release under
> the Act. Any material supplied to NERC may be stored in an electronic records
> management system.

-- 
Dr. Peter Baumann
 - Professor of Computer Science, Jacobs University Bremen
   www.faculty.jacobs-university.de/pbaumann
   mail: p.baumann@jacobs-university.de
   tel: +49-421-200-3178, fax: +49-421-200-493178
 - Executive Director, rasdaman GmbH Bremen (HRB 26793)
   www.rasdaman.com, mail: baumann@rasdaman.com
   tel: 0800-rasdaman, fax: 0800-rasdafax, mobile: +49-173-5837882
"Si forte in alienas manus oberraverit hec peregrina epistola incertis ventis dimissa, sed Deo commendata, precamur ut ei reddatur cui soli destinata, nec preripiat quisquam non sibi parata." (mail disclaimer, AD 1083)