[Dspace-general] DSpace development priorities: starting a discussion

[Christophe Dupriez]

Dear Dorothea, (to the DSpace Community)

Thank you so much for your long needed initiative.

I am using DSpace for different customers who are paying me to adapt
DSpace to their needs. I am very lucky to work with those institutions
who trust me and provide me interesting challenges. I am taking great
advantages of the DSpace project and return much too few contributions
to the community. I previously made suggestions to improve this
(conclusion of the following paper)
http://www.aepic.it/conf/viewpaper.php?id=197&cf=11

For me, big institutions, universities, research networks have the
resources (money, people, organisation) to get what they want from
DSpace source code. They can request what they need  from their
developers and may (or not) encourage their developers to take the time
to contribute back to the DSpace community. From my point of view, this
process is not very efficient and is rapidly seen as not profitable
enough by most managers.

If one desires that other institutions (less money, less people, less
organisation skills available) to be able to publish their intellectual
production using DSpace, I believe more **coordinated** efforts should
be allocated to create a "standard" DSpace  flagship that NO developer
have to customize locally. If we remove from our mind that local
institutions can "always" develop their adaptations, we would look at
the project more cautiously and possibly put back the users where they
must be: in the driver seat.

One cultural problem we may have: Open source developers enjoy freedom
and protect it by opposing the "free software principles" to any
criticism. "free software" means "somewhat free from the commercial
empire", not free for all! Top-down processes must be in a right balance
with bottom-up ones.

Establishing generic institutional needs (a DSpace product definition)
must be a structured project to succeed. A project begins when one
identifies:
1) a global need, objective
2) project sponsors who approve important decisions
3) a knowledgeable project leader who animate, coordinate, mandate

The proposal I would like to make:
1) Apply the 80/20 rule: Create an immediatley applicable DSpace package
which answers to 80% of the needs of 80% of the smaller institutions
which would be happy to not hire any developer (and keep their money to
hire a very good application manager) to have an enthusiastic result
**that the DSpace foundation would guarantee to sustain on the long
term, always providing an easy upgrade path from one version to the next**
2) The sponsors should be institutions gathering to provide resources
(money, people, organisation skills) to obtain this result in a
reasonable time frame (18 months). The Foundation would coordinate this
committe, animate the process with a democratic "1 participating
institution = 1 vote" decision process
3) The project leader would be chosen using a "Call for Tender" process,
with the final decision took by the sponsoring committee.

IMHO, this is much more important than most radical restructuring of
DSpace code base (like some of  the ones currently envisaged). But it
may trigger some other unforeseen radical technical decisions...

Let see how things will evolve!

Have a nice day!

Christophe

Dorothea Salo a écrit :
> Greetings, DSpace community,
>
> For some time, I've been concerned that the DSpace development process
> hasn't enjoyed as much input from the broader community as would be
> desirable. The voices of less-technical repository managers and other
> staff associated with DSpace repositories have been particularly
> difficult to attract to the discussion. I'm hoping to gather
> impressions and suggestions from this specific segment of the
> community (though others are welcome as well!) to pass on to DSpace
> developers. With any luck, this process will build a stronger
> connection between developers and repository managers going forward.
>
> The DSpace development-priority survey done in 2007 was valuable and
> worthwhile, and if possible, I'd like to revisit some of the questions
> raised there. I'd also like to start "in your own words" discussions
> about what repository managers want and need from DSpace that it isn't
> yet providing.
>
> We can certainly talk here, and I welcome that! More than one DSpace
> developer has agreed to monitor these discussions, and I will be
> summarizing them back to the development list. But I'm completely open
> to other venues as well -- IM, group chat, Web 2.0, Skype, out-of-band
> email -- depending on what people tell me they want.
>
> So. How would you like to do this? Once we've sorted out the process,
> we can get down to business. Feel free to contact me off-list if you
> prefer.
>
> Dorothea
>
>   


-------------- next part --------------
A non-text attachment was scrubbed...
Name: christophe_dupriez.vcf
Type: text/x-vcard
Size: 454 bytes
Desc: not available
Url : http://mailman.mit.edu/pipermail/dspace-general/attachments/20080807/0537d2ea/attachment.vcf