[ckan-discuss] s/package/data set/g

Jindřich Mynarz mynarzjindrich at gmail.com
Tue Jul 26 09:23:49 BST 2011


Michael,

I'm not aware of any study of CKAN users: my belief that using the
term "dataset" instead of "package" would confuse users stems only
from a personal experience, which tells me that "dataset" in the
context of CKAN is quite prone to misinterpretation. For instance,
when we announced that we have added over hundred packages to Czech
CKAN [1], there was a blog post saying that we have released
"datasets" themselves (instead of their mere description). Also, I
remember that I needed to explain a couple to times to various people
that CKAN does not contain the data, but only metadata.

I agree with Rufus that the distinction between a "dataset" and a
"pointer" to a dataset is getting a bit blurry, since there are links
to direct download in CKAN and also, in some cases, data are stored in
the CKAN web store [2]. I guess there is no "right" answer to this and
I think we need a pragmatic decision based on how the terms "dataset"
and "package" are being used in the context of CKAN. Do the users have
hard time grasping what "package" is? Would users be confused if CKAN
used "dataset" instead (would they expect to find data in CKAN instead
of metadata)? Would renaming from "package" to "dataset" be confusing?

Best,

Jindrich

[1] http://cz.ckan.net
[2] http://wiki.ckan.net/Webstore

On Tue, Jul 26, 2011 at 9:37 AM, Michael Hausenblas
<michael.hausenblas at deri.org> wrote:
>
> I'm still +1 re dataset but if the majority thinks that package is the
> 'better' term (for whatever 'better' means in this context) I guess I'll
> just shut up.
>
>>>> I'm certainly -1 on this because I think it might confuse the users.
>
> Jindra, that's an interesting question. Who *are* the users? Can anyone
> point me to a precise definition of what we consider a profile of a typical
> and maybe a not-so-typical (aka power user), please? Maybe I've missed this
> one ...
>
> Cheers,
>        Michael
> --
> Dr. Michael Hausenblas, Research Fellow
> LiDRC - Linked Data Research Centre
> DERI - Digital Enterprise Research Institute
> NUIG - National University of Ireland, Galway
> Ireland, Europe
> Tel. +353 91 495730
> http://linkeddata.deri.ie/
> http://sw-app.org/about.html
>
> On 26 Jul 2011, at 08:27, Tim McNamara wrote:
>
>> I personally prefer 'package' also. My thinking was that dataset would
>> be less confusing. From Jindřich's comments, that assumption may not
>> be valid and there are probably more important tickets to work on.
>>
>> Have we had any direct complaints? If not, I suggest that we focus on
>> other areas.
>>
>> 2011/7/26 Pablo Mendes <pablomendes at gmail.com>:
>>>
>>> Have you guys discussed what to do with "resources" when package becomes
>>> dataset? Datasets continue containing resources? Or datasets contain
>>> datasets?
>>>
>>> I didn't really have anything against package. Does "data package" please
>>> greeks and trojans?
>>>
>>> Cheers
>>> Pablo
>>>
>>> On Jul 25, 2011 10:45 AM, "Jindřich Mynarz" <mynarzjindrich at gmail.com>
>>> wrote:
>>>>
>>>> I'm certainly -1 on this because I think it might confuse the users.
>>>> The package is not the dataset, it's a *description* (metadata) of the
>>>> dataset [1]. Maybe it's not the best term to use, but it's better than
>>>> to lead users to expect they can find the datasets themselves at the
>>>> Data Hub.
>>>>
>>>> Best,
>>>>
>>>> Jindrich
>>>>
>>>> [1] http://en.wikipedia.org/wiki/File:MagrittePipe.jpg
>>>>
>>>> On Mon, Jul 25, 2011 at 10:10 AM, Rufus Pollock <rufus.pollock at okfn.org>
>>>> wrote:
>>>>>
>>>>> On 25 July 2011 05:17, Michael Hausenblas <michael.hausenblas at deri.org>
>>>>> wrote:
>>>>>>
>>>>>>> I think that the term "package" should be removed from thedatahub.org
>>>>>>> and replaced with "data set" or "dataset". While CKAN is to be the
>>>>>>> Debian of data, I think "dataset" would lead to fewer mental steps
>>>>>>> for
>>>>>>> new comers.
>>>>>>
>>>>>> +1
>>>>>
>>>>> This has already been discussed quite a lot (e.g. it's one of the
>>>>> items on the UX refactor page: <http://wiki.ckan.net/UX>) and I think
>>>>> there is general agreement on this. The issue here is that it is a
>>>>> (relatively) costly refactor so we are waiting for the right time to
>>>>> do it (we can try doing it slightly 'hackily' by just using i18n but I
>>>>> think this is a short term approach).
>>>>>
>>>>> Rufus
>>>>>
>>>>> _______________________________________________
>>>>> ckan-discuss mailing list
>>>>> ckan-discuss at lists.okfn.org
>>>>> http://lists.okfn.org/mailman/listinfo/ckan-discuss
>>>>>
>>>>
>>>> _______________________________________________
>>>> ckan-discuss mailing list
>>>> ckan-discuss at lists.okfn.org
>>>> http://lists.okfn.org/mailman/listinfo/ckan-discuss
>>>
>>> _______________________________________________
>>> ckan-discuss mailing list
>>> ckan-discuss at lists.okfn.org
>>> http://lists.okfn.org/mailman/listinfo/ckan-discuss
>>>
>>>
>>
>> _______________________________________________
>> ckan-discuss mailing list
>> ckan-discuss at lists.okfn.org
>> http://lists.okfn.org/mailman/listinfo/ckan-discuss
>
>
> _______________________________________________
> ckan-discuss mailing list
> ckan-discuss at lists.okfn.org
> http://lists.okfn.org/mailman/listinfo/ckan-discuss
>



More information about the ckan-discuss mailing list