[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Handle-info] https for hdl.handle.net
Dne 11/20/2015 v 11:48 AM Gerhard Gonter napsal(a):
> According to [1] your repository ask to cite the permanent identifier
> in the form of a resolver proxy url. IMHO, using an URL, even or
> especially a proxy url, would contradict the Handle RFCs (e.g. [2]).
> I assume, your organization has discussed this internally and thus
> came up with your citation guide. Did they consider the more basic
> forms of representing Handles, that is:
>
> * 11234/1-1548 (the handle value itself)
> * hdl:11234/1-1548 (with a schema indication)
>
> Well, just wondering what the arguments in your (internal?) discussion were...
Dear Gerhard,
your comments and question is very good and important indeed. And yes,
we had been discussing the
advantages and disadvantages and decided for proxied PIDs or whatever
you want to call them.
I will briefly? describe our reasoning below.
First of all, our background is academic and mainly related to data
stored digitally and services around them.
According to our experience, only very few (academic) people know what
hdl is.
Because we expect the citations to be used in papers and papers are
often available in the digital but also in non digital format, we think
that URLs are much
more understandable (and expected) than "hdl:11234/1-1548" or
"11234/1-1548".
Here you might point to the usage of DOIs in papers but from the
*social* point of view, DOIs are unique; moreover, see the examples of
other repositories below.
Showing hdl:... and having the underlying link pointing to the proxied
PID does not solve anything. There are other issues like
crawlers/parsers understanding
"http" protocol but not hdl when there is no associated link. And how
many browsers do support the hdl protocol?
Having this in mind together with being pragmatic leads to the fact that
we want to provide URLs as the reference.
So what are the work flows for resolution from a citation?
1. from a paper paper ;)
- retype the URL or the PID to your search engine
2. from web
- click on the link (or use 1. with copy and paste)
What can go wrong with our approach (well, with any other mentioned
approach in fact)?
hdl.handle.net stops working. I do not want to get into the details but
unless you know what handles are, you will
have the same problems with "hdl:11234/1-1548". When you know what
handles are, you can easily spot the PID in the URL
or just go to 1. and the search engine gladly returns our landing page.
In other words, the PID itself is (for our use case) useless unless you
have the service that can resolve it.
I have to add one more question and answer because people might be
wondering why handles then and not some sort of PURLs?
If you really do not know the answer, you should probably use PURLs then
(and google will be much more happy and maybe provide
you with proper back links points too).
And finally, we are not the only one - look at the following digital
repositories and search for citation
http://dx.doi.org/10.6084/m9.figshare.1608704
http://dx.doi.org/10.17632/6n2tfgxcgb.1
even
http://dx.doi.org/10.5281/zenodo.32481 (select e.g., acm-siggraph citation)
However, I do not know their reasoning...
Best,
Jozef
>
> References:
> * [1] https://lindat.mff.cuni.cz/repository/xmlui/page/cite
> * [2] http://tools.ietf.org/html/rfc3650#section-6.1
>
> GG
_______________________________________________
Handle-Info mailing list
Handle-Info@cnri.reston.va.us
http://www.handle.net/mailman/listinfo/handle-info