The question for me is: what is OAI-PMH actually for? And does it need
to continue to be supported? Can't we just *let it die*?
As a developer, if I have a choice of an Atom/RSS feed or OAI-PMH
harvest service to get metadata about resources in a repository, I'll
do the former.
Why? Because I can do that without:
- having to find the OAI endpoint in the first place
- having a lengthy conversation by email with the repository owner
about how they've implemented it
- end up rewriting my code for *each repository*
- worrying whether the data will even parse *at all* when I get it
- worrying whether the data will even have any URLs for the
resources(!!)
... all of which massively outweigh any advantages OAI might have in
terms of "flexibility" or features.
My short-range predictions:
- OAI-PMH will struggle on as it has, gradually diminishing as those
engaged in it move on to other work
- OAI-ORE will gain some popularity, and will to some extent replace
PMH, but:
- The majority of systems will treat it as regular Atom feeds and
ignore all the weird extra stuff
- for SWORD deposits, the whole METS and DC thing is so problematic
that Atom Feed documents become the de facto metadata for depositing
"not very complex object" packages
- All repositories provide Atom/RSS feeds, which becomes the primary
method for aggregating the content by external discovery services
- A "simple SWAP"-type spec, ignoring the FRBR stuff, is developed as
a very small Atom extension and actually gets used successfully though
on a small scale
Cheers, I'll go duck behind this wall now.
S
On 2 May 2008, at 17:11, Andy Powell wrote:
> Hmmm... interesting. My (admittedly very light) reading of that stuff
> is that it starts from the position that the OAI-PMH doesn't fit the
> Web
> Architecture and that to do something about it you have to retro-fit
> an
> RDF model on to it and then build a "gateway" that sits in-front of
> the
> protocol. Hardly an optimal position... but I must admit I need to
> read
> stuff in more detail to comment properly.
>
> Andy
> --
> Head of Development, Eduserv Foundation
> http://www.eduserv.org.uk/foundation/
> http://efoundations.typepad.com/
> [log in to unmask]
> +44 (0)1225 474319
>
>> -----Original Message-----
>> From: Jewel Ward [mailto:[log in to unmask]]
>> Sent: 02 May 2008 16:29
>> To: Andy Powell
>> Cc: [log in to unmask]
>> Subject: Re: Google, OAI and the IRs
>>
>>>
>>> On May 2, 2008, at 7:46 AM, Andy Powell wrote:
>>>> ...
>>>>
>>>> It highlights the fact that OAI will never be a mainstream Web
>>>> protocol, but so what... I think we spotted that anyway!
>>>>
>> i wouldn't be so sure. check out the links below. :-)
>>
>>>> There are technical reasons why OAI was always going to
>> struggle (I
>>>> say that only with the benefit of hindsight) because of
>> its poor fit
>>>> with the Web Architecture. Whilst I don't suppose that directly
>>>> factored into Google's thinking in any sense, I think it is worth
>>>> remembering.
>>>>
>>
>>
>> (yes, Andy, I openly admit to my biases!)
>>
>> this was posted yesterday by Herbert Van de Somple on OAI-
>> Implementers. the links are worth checking out, for those of
>> you interested in the contents of this thread.
>> *****
>> Date: Wed, 30 Apr 2008 14:00:11 -0600
>> From: Herbert Van de Sompel <[log in to unmask]>
>> Subject: [OAI-implementers] OAI2LOD
>> To: OAI-implementers <[log in to unmask]>
>> Message-ID: <[log in to unmask]>
>> Content-Type: text/plain; charset=ISO-8859-1; format=flowed
>>
>> hi all,
>>
>> I felt like drawing you attention to an OAI-PMH tool that has
>> been created by the Linked Data community:
>>
>> (*) The OAI2LOD server is described and available at:
>> http://www.mediaspaces.info/tools/oai2lod/
>>
>> (*) A paper was presented at the recent WWW Conference Linked Data
>> Workshop:
>> http://events.linkeddata.org/ldow2008/papers/03-haslhofer-schandl-
>> oai2lod-server.pdf
>>
>> (*) And Tim Berners Lee took quite some interest in this
>> development, as can be seen from:
>> http://lists.w3.org/Archives/Public/semantic-web/2008Apr/0183.html
>>
>> Thanks to Chris Bizer for the pointers.
>>
>> cheers
>>
>> Herbert
>> ****
>>
>> also check out further comments (see: 27 April 2008) at:
>> http:// lists.w3.org/Archives/Public/semantic-web/2008Apr/
>>
>> Regards,
>>
>> Jewel
>>
>>
>> ..
>> Jewel Ward, Ph.D. Student
>> The School of Information and Library Science Univ of North
>> Carolina @ Chapel Hill [log in to unmask] http://ils.unc.edu/~wardj/
>>
>>
>>
>>>
>>
|