[AMF-talk] Re: software descriptions in ReDIF

[baum]

Thomas,

I certainly agree with your last point, that the divisions into template 
types are artificial and sometimes clumsy, and that we should move toward 
(more like a cataloging record) a metadata format that encompasses all of 
the possible fields that might appear in describing any object, only some 
of which will be relevant for a certain type of object. Then an entry form 
for a given object type (or the validation of those metadata) would merely 
select from the identifiers relevant for that specific kind of object. That 
would make great sense, and would encourage the development of smart entry 
forms that would present only the relevant fields for the specified object 
type.

But such a unified approach makes it even more difficult to understand why 
there should be such a yawning chasm between templates e-documents and 
templates describing software. After all, in the current scheme, there are 
very few differences between a software-template and a paper-template. If 
need be we can recast all of the software-templates into paper-templates to 
prevent their exclusion from the new system's reach, but that would seem 
subooptimal compared with having an expanded list of identifiers, some of 
which would (only) be relevant for describing software, in the same sense 
that library catalogers have special fields to describe media (including 
those containing software and/or data files) rather than print materials.

How do the maintainers of RePEc services feel about doing away with the 
distinction of working papers and software modules in order to deal with 
these limitations of AMF?

Thanks
Kit

--On Saturday, September 21, 2002 9:56 -0500 Thomas Krichel 
<krichel@openlib.org> wrote:

>
>
>   Kit Baum for the RePEc Team writes, regarding software
>   descriptions in RePEc
>
>> Can we rectify that by providing/drafting a metadata format?
>
>   Of course... but this is really not top of priority for me, I will
>   hope to get someone to do it some time.
>
>> After all I do not imagine there is any software description for
>> statistical data sets either, but we have those included in the new
>> system. All the advantages of being able to do OAI will be of
>> limited value if we can only handle certain of the existing template
>> types in an OAI repository.
>
>   Sure. But I doubt that we should shoehorn software descriptions
>   into AMF. I have seen a software description proposal submitted
>   to the W3C by some guys from Microsoft in 1997. It was never
>   adopted as a standard.
>
>> Christian was here Wed. for a seminar, and we had a chance to chat about
>> directions for various aspects of RePEc. One thing that we he suggests
>> that  he will be able to deal with on his new hardware are book and
>> chapter  templates, which will be useful.
>
>   I happen to believe that the split, within ReDIF, into various
>   text template types is a design mistake. It considerably complicates
>   the maintenance of the format and data encoded in the format
>   unnecessarily complicated. People do not care what type of the text
>   is, they mostly care about some fundamental characterstics of the
>   text such as authore and subject, and wether the fulltext is
>   available online.
>
>
>   Cheers,
>
>   Thomas Krichel
> mailto:krichel@openlib.org
> http://openlib.org/home/krichel
> RePEc:per:1965-06-05:thomas_krichel
>



________________________________________________________________________
Christopher F Baum, Boston College Economics, Chestnut Hill MA 02467 USA
baum@bc.edu                       http://fmwww.bc.edu/ec-v/baum.fac.html