[EP-underground] Re: EPrints, DSpace or ESpace abou design and interface (fwd)

Evelyn, I have forwarded your posting to the eprints list rather than
the AmSci list. The AmSci is about archive *filling* policy, not about
suggestions for archiving software features. Please sign on to
eprints-unbderground is you wish to discuss eprints features.

    http://software.eprints.org/maillist.php

Cheers, Stevan Harnad

---------- Forwarded message ----------
Date: Wed, 22 Sep 2004 14:23:59 -0300 (BRT)
From: Evelyn Cristina Pinto <[email protected]>
To: American Scientist Open Access Forum
    <[email protected]>
Subject: Re: EPrints, DSpace or ESpace abou design and interface

I've been studying and thinking about design for digital libraries and I
have some comments and suggestions.

Libraries have important roles in society and they interact with many other
important institutions such as scholarly communication, professional
societies, legal system and higher education. Now we've been developing
digital library software and we have to be careful to reflect on
relationships, values and mechanisms that the library system has.

If we maintain the possible characteristics and bring some useful
innovation, probably we'll be successful, because good software design
adapts the software to the reality, not the opposite. A common trap is to
incorporate software patterns into the design. See more about it on an
interesting article by Phil Agre:
http://polaris.gseis.ucla.edu/pagre/dl.html.

I have installed Eprints software already and I've used it many times. I think
that there are many positive aspects like excellent documentation,
installation facilities and it works well (few bugs), but I think that the
design and interface could be improved much more.

For example, it's hard to find out how to deposit an article though the
home page. And the self-archiving process is not easy to do on Eprints.
I think that the tool to deposit comments about articles is great, but I
think that there could be more space to debate. The search and browse
system are good but they could also incorporate a little more on the way that
people do search on traditional libraries.

Digital libraries could be used for collective cognition processes
(Collective Intelligence: Mankind's Emerging World in Cyberspace by Pierre
Levy), but we need to incorporate incentives for information change. On
this kind of software, there could be tools for producing scientific
articles in a cooperative way like Wikipedia. It's amazing the quality
level of some articles there.

Other point that I want to emphasize are the methodology to develop
software for digital libraries. I think that there are two very good ways:
the bazaar method and the extreme programming method.

The bazaar method uses many developers, testers and users to produce the
software. So it can satisfy the users because users are also the software
programmers. And there's another advantage: there are many eyes to correct
the bugs. For more details, see
http://en.wikipedia.org/wiki/The_Cathedral_and_the_Bazaar.

Extreme Programming states that the users and developers must work on the
same physical place to make easier the communication among them. For more
details, see http://en.wikipedia.org/wiki/Extreme_programming.

Please, I want to know your comments about my comments.

Best regards,
Evelyn.
----------------
master student in Computer Science - IME-USP