Comments on: Repository architecture #83

By: Bookmarks about Repository

Bookmarks about Repository — Sat, 23 Aug 2008 09:01:09 +0000

[...] - bookmarked by 2 members originally found by rickenriquericky on 2008-08-09 Repository architecture #83 http://blog.paulwalk.net/2008/07/07/repository-architecture-83/ - bookmarked by 3 members [...]

By: paul

paul — Mon, 14 Jul 2008 15:17:22 +0000

Dorothea,
Be my guest

I’d be interested to see how you use it…..

Paul

By: Dorothea Salo

Dorothea Salo — Mon, 14 Jul 2008 14:12:49 +0000

Paul, may I borrow that screenshot (with credit, of course) for a presentation I am giving?

By: paul

paul — Tue, 08 Jul 2008 13:32:00 +0000

Regarding point 3 - as Jim Downing suggests, the best we might do is to move the complexity around. I suggest moving it away from the source repository - but the institutional repository may have to pick up much of this.

But yes - those expectations should be managed!

By: Steve Hitchcock

Steve Hitchcock — Tue, 08 Jul 2008 11:19:21 +0000

Three points arising from Paul’s presentation and the comments:

1 On subject vs institutional repositories, *some* of the content (papers) is in subject Rs, of course. But the real question is: what about all the rest? Where is that to become OA if 100% is the target?
2 Paul’s idea of ’source’ repositories is good if it means getting the repository interfaces closer to the end-users, perhaps school or dept. repositories, with a central services approach institutionally.
3 Complexity appears to be rife. Data, metadata, workflows, etc. - all need to be worked out to ensure value always exceeds cost. We cannot afford to burden IRs with unconstrained expectations, responsibilities, and costs, especially in these uncertain economic times.

By: paul

paul — Tue, 08 Jul 2008 10:23:19 +0000

Chris,
three quick responses:
First: I agree that Technorati does not always deliver…. I think the authority figure is utterly suspect - my own feeble ‘authority’ figure seems to change with the weather….
However, while the execution is not great, I still think the model is a good one. The problem which Technorati tries to solve is considerable - it allows the creation of content to be remain completely distributed, unlike Slideshare.

Second: I agree with you - I believe the institution is well placed manage this - we should see this activity grow as institutional workflows become established and smarter.

Three: yes - rather than giving overt incentives to ‘deposit in the repository’, give the incentive to ‘use our fabulous workflow tools’ - which happen to deposit resources as appropriate.

By: Chris Rusbridge

Chris Rusbridge — Tue, 08 Jul 2008 08:47:55 +0000

Three thoughts. I just don’t get the Technorati versus Slideshare reference. Technorati sucks! I don’t know if this is because it’s by reference, or just that its business model is so broken that it can’t afford sufficient grunt to make it work, but it loses things, known searches fail, multi-page search results loop… I wouldn’t use it at all except for the “authority” figure (which all of the above makes doubtful. Slideshare, OTOH, just works.

Second, I agree that most of the data that are in repositories are in subject repositories. Likewise most of the papers that are in repositories are in IRs (perhaps with the honourable exception of Arxiv). But most data are not in repositories at all. (Likewise most papers.) Most data cannot go into subject repositories because there are not enough of them, because the business model for subject repositories is fragile in the extreme. The business model for IRs, however, looks more robust the more content they can capture. So I think we need to find more ways of getting our data into sustainable IRs.

Thirdly, (I think I catch your support for this by implication but not explicitly) getting the stuff in involves both incentives (in making science easier) and integration with researcher workflow. If they have to come back and do it later, I believe it just won’t happen. It’s why I was suggesting the (possibly over-complicated) research repository system idea in my blog post…

By: Richard Akerman

Richard Akerman — Mon, 07 Jul 2008 18:34:41 +0000

I wouldn’t say reusability is over-rated as a virtue. I would say that reusability is really, really hard. In any case I think we frame the discussion differently these days. Originally reusability was so that you could maximize your ROI. Now we’re more concerned with flexibility, to maximize your… hmm… Change On Investment? So instead of “design & build an architecture that supports reusability”, let’s say “design & build an architecture that is flexible”. Flexibility Oriented Architecture?

By: Unilever Centre for Molecular Informatics, Cambridge - Jim Downing » Blog Archive » Repository architectures, leaky abstractions and Paul’s principles

Mon, 07 Jul 2008 16:15:53 +0000

[...] Walk kicked off the day by presenting his work on a next generation architecture for repositories. His presentation started off with a number of starting principles and moved on to some diagrams [...]

By: JISC meetings « Names Project Blog

JISC meetings « Names Project Blog — Mon, 07 Jul 2008 14:01:58 +0000

[...] the morning about how such an architecture might look - links to his slides are available from his blog entry about the day. As Paul points out, there was a fair amount of discussion about repository content [...]