[ckan-discuss] CKAN feedback

Mon Aug 30 09:59:43 BST 2010

Many thanks for all your time on this Chris! ;-)

Carbon copying this to ckan-discuss list...

All: any way someone could go through Chris's comments with thoughts /
tickets where relevant?

On Mon, Aug 23, 2010 at 5:03 PM, Christopher Gutteridge
<cjg at ecs.soton.ac.uk> wrote:
> Hi, and sure.
>
> I really don't want to join another mailing list so here's a brain dump.
> Sorry if I sound a bit blunt in some of it but I'm doing this in between
> other jobs!
>
> My key issue with CKAN is that I can't see a usecase for it. Once I've
> solved a problem in my life using it then I'll value it and maybe
> contribute. I know that's harsh, but maybe I've just not seen the bit of the
> future you can see (I said wiki's would never take off....)
>
> The other key issue is that I tend to get links to it in my inbox now and
> then, but not to the homepage. Landing on this page:
> http://ckan.net/group/bibliographic
> ..makes no sense. It's like looking at an RPM repository!
>
> Looking at this page;
> http://ckan.net/package/rkb-explorer-southampton
> is also very very confusing.
>
> OK. Here's some suggestions. Some general template ideas are muddled in as
> I'm doing this in under an hour as I'm afraid that's all I can spare. If I
> was doing this properly I'd sort my points a little better.
>
> Why is there no "download our entire database" button?
>
> Group page
>
> Add some text in the body of the page (not the right sidebar). Just one
> sentence to say "This page describes a number of pacakges of data available
> on the internet with the common theme of 'Bibliographic Data'. You can log
> in to add to or update this list."
> The closed/open door is not very clear as a visual aid.
> I have no idea what the blue bar to the left of some items means.
> Nor the green down arrow!
> There's plenty of space in the right sidebar for an icon key!
> The Administrators list is bulky and ugly. I guess it's a list of URIs? Not
> user friendly, change it to peoples name and affiliation.
> Your XHTML and CSS links in the footer look like options, not validator
> buttons (there are standard w3c icons)
> Your XHTML and CSS don't actually validate! (my sites often drift after I
> create them perfectly on day one, by a year later those buttons fail for me
> too, so not a big smugness on my part, just info.)
> The first thing I looked for on this page was a download-as link or links.
> Why not link into the API directly from the page to provide JSON,RDF,XML,CSV
> of the data. Instantly more useful.
> What is the URI for each group?
> Provide <link rel='alternate' > links to the API pages too.
> There's no about link to http://ckan.net/about -- people often land on
> subpages but now I've hunted around there *is* an about page and userguide,
> but only linked from the frontpage sidebar!
> Consider removing the full title of the site from the <title> as the title
> is way too long "CKAN - Package - foo" might be nicer.
>
> Package Page
>
> Again, needs a single sentence "This page describes a package of data
> available on the Internet. " or somesuch.
> Empty details are shown which makes most of your pages look a bit sketchy.
> You repeat the metadata in the main flow and the sidebar.
> I suggest removing any "unknown" information from the sidebar.
> Don't show "Tags" or "Groups" if there's no data.
> Don't show License, Maintainer or Version if not available.
> Add a section with a call for someone to improve the missing fields. "Add a
> tag", "Do you know who the maintainer is?" etc.
> This package: in the submenu in the title bar is a contextual title BEFORE
> the title below which gives you the context. It should appear below the
> title of the package, or the "this package" should be replaced with the
> package name.
> The URL column does not contain a URL.
> The HASH column does not contain any hashes. (maybe it sometimes does, but
> how about hiding it when irrelevant?)
> "Alternative metadata formats" is good -- should be expanded, but I can't
> read the JSON as it's white on light green. I had to look close at my screen
> and missed it until now I'm actually studying the page closely.
> Why no XML, CSV or RDF?
> The mimetype is not as interesting as the human readable name. I suggest
> moving the Description column to be first and making the URL col. last.
> Rename URL to "Link" and change the text "Download" to an icon.
> Consider adding an invitation to contribute into the top level bit with the
> CKAN logo.
>
> http://ckan.net/group/lod
>
> &raquo; appears to be a typo!
>
> On 16/08/10 17:58, Jonathan Gray wrote:
>
> On Mon, Aug 16, 2010 at 6:08 PM, Christopher Gutteridge
> <cjg at ecs.soton.ac.uk> wrote:
>
>
> I find CKAN pretty impenetrable. I have no idea what I am expected to do to
> contribute to that page or how to put it to use.
>
>
>
> [slightly off topic -- but cc'ing to list for posterity!]
>
> Uh oh! Sounds like we aren't doing something right! Thanks for letting
> us know! ;-)
>
> Chris: I know its a pain, but would you mind sitting down with someone
> on our end and going through some of the things you find problematic /
> difficult to understand with CKAN, so we can try to address them and
> make it more usable?
>
> Can either contact me or Friedrich (in cc) or ping our public CKAN discuss
> list:
>
> http://lists.okfn.org/mailman/listinfo/ckan-discuss
>
> We've also just discussed having better feedback mechanism so its
> easier for people to say what they find difficult / non-obvious
> (anticipate that some of this could be addressed in FAQ, linked to
> from relevant places...).
>
> J.
>
>
>
> On 16/08/10 16:33, Jonathan Gray wrote:
>
>
> If I'm understanding you correctly, this is *exactly* what CKAN is for:
>
>   http://ckan.net/
>   http://ckan.net/group/bibliographic
>
> CKAN is an open source registry of open data/open content 'packages'
> (as in software packages). Medium to long term idea is something like
> apt-get for open data, with support for automating lots of stuff.
>
> Focus is on material that is open as in opendefinition.org (of which
> PD material and *some* CC licensed content is a subset).
>
> Does that help? Would love to have any feedback on how we can improve
> CKAN for bibliographic material.
>
> All the best,
>
> Jonathan
>
> On Mon, Aug 16, 2010 at 4:52 PM, John Wilkin<jpwilkin at umich.edu>  wrote:
>
>
>
> All,
> Of course I can think of a bucket-load of reasons why this would be
> impossibly hard to assemble and maintain, but I'm still curious:  has any
> organization tried to create a database of essentially "open" bibliographic
> resources?  In this case, I'm interested in something broad enough to
> include CC, PD, etc.--i.e., resources that can be used (at least in
> scholarship and teaching) without fees paid to the maintainer of the
> resource?
> _______________________________________________
> open-bibliography mailing list
> open-bibliography at lists.okfn.org
> http://lists.okfn.org/mailman/listinfo/open-bibliography
>
>
>
>
>
>
> --
> Christopher Gutteridge -- http://id.ecs.soton.ac.uk/person/1248
>
> / Lead Developer, EPrints Project, http://eprints.org/
> / Web Projects Manager, ECS, University of Southampton,
> http://www.ecs.soton.ac.uk/
> / Webmaster, Web Science Trust, http://www.webscience.org/
>
>
> _______________________________________________
> open-bibliography mailing list
> open-bibliography at lists.okfn.org
> http://lists.okfn.org/mailman/listinfo/open-bibliography
>
>
>
>
>
> --
> Christopher Gutteridge -- http://id.ecs.soton.ac.uk/person/1248
>
> / Lead Developer, EPrints Project, http://eprints.org/
> / Web Projects Manager, ECS, University of Southampton,
> http://www.ecs.soton.ac.uk/
> / Webmaster, Web Science Trust, http://www.webscience.org/
>

-- 
Jonathan Gray

Community Coordinator
The Open Knowledge Foundation
http://blog.okfn.org

http://twitter.com/jwyg
http://identi.ca/jwyg