[Taxacom] Data query

Richard Pyle deepreef at bishopmuseum.org
Tue Jun 25 05:51:37 CDT 2013

Ah!  I'm glad you brought up that example.  This is something I've been
harping on a LOT recently at various biodiversity informatics meetings:
separating the role of "identification" from "dereferencing metadata".

So, consider the link
"http://atbi.biosci.ohio-state.edu:210/hymenoptera/1234".  I treat "1234" as
the identifier, and " http://atbi.biosci.ohio-state.edu:210/hymenoptera/" as
the dereferencing metadata. 

Why is this important?  Because I may have 2000 cross-links to the
identifiers (1234, 1235, 1236, ... etc.), but each of those is linked to
only ONE record for the Identifier Domain (Hymenoptera Name Server).  One
field of that Identifier Domain table is "URLPrefix", which would be
"http://atbi.biosci.ohio-state.edu:210/hymenoptera/".  All the links to the
identifiers are built by concatenating the dereferencing metadata with the
identifier. Thus, when HNS changes its web service to
http://osuc.biosci.ohio-state.edu/hymenoptera/ , I change ONE field in ONE
record, and presto! All 2000 links are working again. (This exact thing
happened with all 60,000+ links to Catalog of fishes, and it took me all of
about 10 seconds to correct it for all 60,000 records by changing one

Now, it's probably a pretty simple matter for you to do a global search &
replace on "http://atbi.biosci.ohio-state.edu:210/hymenoptera/" swapped for
"http://osuc.biosci.ohio-state.edu/hymenoptera/", and you'll effectively
achieve the same thing. So even in your case, you can make the repair very
easily, and then all that work you painstakingly spent creating the links
will be remain intact, and serve the benefit of all mankind.

OK, definitely past my bedtime now....


> -----Original Message-----
> From: Dr Brian Taylor [mailto:dr.brian.taylor at ntlworld.com]
> Sent: Tuesday, June 25, 2013 12:40 AM
> To: Richard Pyle; 'Stephen Thorpe'; Tony.Rees at csiro.au; Bob Mesibov
> Cc: taxacom at mailman.nhm.ku.edu
> Subject: Re: [Taxacom] Data query
> A little point but critical to the ongoing discussion.  On each taxonomic
> of my website "Ants of Africa", I painstakingly gave a link to the
> Hymenoptera Name Server.
> The link for several years has been on
> http://atbi.biosci.ohio-state.edu:210/hymenoptera/
> Suddenly the location (server, I assume) has changed to
> http://osuc.biosci.ohio-state.edu/hymenoptera/
> So all my 2000 or so pages have a non-operational link.
> So, I might argue, what was the point of my spending time giving links?
> Brian
> On 25/06/2013 07:55, "Richard Pyle" <deepreef at bishopmuseum.org> wrote:
> > I slogged through this discussion, and found many things I agreed
> > with, and many things I disagreed with, from all of the major
> > contributors to the thread.  However, I did want to call attention to
> > one point raised by several different people, that I STRONGLY agree
> > with, and that I would like to see amplified repeatedly  and very
> >
> > Tony Rees said it this way:
> >
> >> What we really need to aspire to is a network of linked systems
> >> whereby changes and new content entered once can propagate
> >> automatically elsewhere as needed rather than requiring manual input
> >
> > And again this way:
> >
> >> Perhaps it is time to move the discussion on to related matters such
> >> as how to possibly chain these systems together in the future so that
> >> updates may more readily flow from the point of entry to potentially
> >> multiple points of re-use...
> >
> > And then Stephen followed with this gem:
> >
> >> Well, one thing that stands out here is that the only honourable
> >> thing to do, but is rarely done, is to provide links from a taxon
> >> page in one project to the corresponding taxon pages in other
> >> projects which may have better and more complete data on that taxon,
> >> so as to give the user a chance of finding what they want to know
> >> even if the project they have visited doesn't have the data.
> >> Conversely, doing so will also highlight how good the visited project
> >> is, if the linked projects have less data, so it is win-win, really!
> >
> > YES!  YES! YES!!
> >
> > About two years ago, it became very clear to me that a small feature I
> > added to my Taxonomer database back in the 1990's, and included in the
> > first version of ZooBank back in 2008 (cross-linking identifiers) will
> > almost certainly come to be known as one of the most important (if not
> > *the* most
> > important) contributions from GNUB/ZooBank to our community; which is
> > a generic mechanism for cross-linking records to each other.
> >
> > It's an incredibly simple data structure, and is now built in to the
> > ZooBank website for Authors, References and Taxon Names (but it can
> > easily be applied to any record in any table of the database).  When
> > you're logged in to ZooBank and have access to a particular record for
> > editing, you'll see a small link next to the "Edit" button for "add
> > external identifier". Click on this and you get a pop-up form that
> > lets you create cross-links to other external databases.  Just for fun,
I went
> to my old trusty example:
> > http://zoobank.org/8BDC0735-FEA4-4298-83FA-D04F67C3FBEC
> >
> > and added a link to the Wikispecies page (Identifier Doimain =
> > "WikiSpecies Page"; Identifier = "Chromis_abyssus"), and voilĂ !  A
> > link from the ZooBank record to the WikiSpecies page is born!
> >
> > Now... a few caveats...  there is a buglet that prevents the closing
> > of the pop-up form after you create such a link (just refresh the page
> > to make it go away).  Also, it's only visible if you have edit access
> > to a particular record in ZooBank.  Once we fix the buglet and beef up
> > the protection code (to minimize bogus entries), we'll make this
> > feature available to anyone with a login account for applying to any
> > record (whether or not the user has edit access to the record).
> >
> > Our plan is to develop a generic service around this cross-linking
> > mechanism such that anyone can harvest the cross-links for any record,
> > and display them on their local page.  The REALLY cool part is this:
> > Once I link the ZooBank record to the Wikispecies page, the WoRMS
> > page, the FishBase page, and so on....I've not only linked ZooBank to
> > these other sources, we've also linked these other sources TO EACH
> > OTHER. In the example I gave above (Chromis abyssus), I didn't just
> > link ZooBank to WikiSpecies; I also linked WikiSpecies to BOLD, Bill
> > Eschmeyer's Catalog of Fishes, Digimorph, EoL, FishBase, MorphBank,
> > NCBI/GenBank, and WoRMS.  And, I've linked each of these databases to
> > WikiSpecies (as well as each other).  The idea for the generic service
> > would be for WoRMS to fire off its identifier (in this case,
> > 345844) to the service and get back the links to all the other
> >
> > As Stephen says, the honorable data managers will see this as a
> > no-brainer win-win.
> >
> > I'll post more on this service once we get it functional.
> >
> > Aloha,
> > Rich
> >
> >
> > _______________________________________________
> > Taxacom Mailing List
> > Taxacom at mailman.nhm.ku.edu
> > http://mailman.nhm.ku.edu/mailman/listinfo/taxacom
> >
> > The Taxacom Archive back to 1992 may be searched with either of these
> methods:
> >
> > (1) by visiting http://taxacom.markmail.org
> >
> > (2) a Google search specified as:
> > site:mailman.nhm.ku.edu/pipermail/taxacom
> > your search terms here
> >
> > Celebrating 26 years of Taxacom in 2013.

More information about the Taxacom mailing list