Open Knowledge Foundation Activity Aggregator

For all you Twitter users out there, the OKF has now got an account... http://twitter.com/okfn J.

Hi all, After discussions with Cameron Neylon and Kaitlin Thaney, I've set up a new open science mailing list: http://lists.okfn.org/cgi-bin/mailman/listinfo/open-science I've blogged about this here: http://blog.okfn.org/2008/09/05/new-open-science-mailing-list/ We'd really like to get as much of the open science community on this as possible, so please blog and forward as appropriate! Jonathan

After discussions with Cameron Neylon of Open Wetware and Kaitlin Thaney of Science Commons we’ve set up an open science mailing list:

• http://lists.okfn.org/cgi-bin/mailman/listinfo/open-science

As far as we could tell, there wasn’t a general mailing list for people interested open science. Hence the new list aims cover this gap, and to strengthen and consolidate the open science community.

We hope it will be a relatively low volume list for relevant announcements, questions and notes. We also hope to get as full as possible representation from the open science community - so please forward this to anyone you think might be interested to join!

Share This

[factlet/edit][l]: add support for factlet editing, adding in template action support to factlet.

• factlet.py:
  • new update method providing factlet editing, previewing and saving
  • new template method to provide json based templating
• lib/json.py: add flush option to converter to ensure don't have to flush to db. Need because new templating work requires the ability to use converter (and create python Domain Objects) without modifying domain model (i.e. saving to db).
• layout.html: remove any default jspagecontroller.

Dave Crossland dixit: Thanks, that sounds great. bye, //mirabilos

Nature special on big data, including politics of environmental data sharing and Cory Doctorow on big data centres (including Sanger, archive.org,..)... http://www.nature.com/news/specials/bigdata/index.html Jonathan

2008/9/1 Rufus Pollock at >public.gmane.org>: Based on the discussion with Thorsten, I change my suggestion slightly: 1. Whose data is open as defined by the open knowledge definition (http://opendefinition.org/1.0/) with the exception that where the data is personal in nature the data need only be made available to the user (i.e. the owner of that account). 2. Whose source code is Free/Open Source, as defined by both the FSF Free Software Definition and OSI Open Source Definition and is available to the users of the service. (The FSF and OSI maintain example lists of licenses that meet their definitions)

2008/9/4 Thorsten Glaser at >public.gmane.org>: Hahaha yes you are quite right. How silly of me. Hmm. The problem with boolean-OR for the lists is that it means the FSF doesn't trust it, because its includes non-free licenses. Okay, again: :-) Definitions have ambiguity. When writing a definition, it is good to limit ambiguity. A definer typically writes an approved example list to reduce ambiguity. A list is finite, which is how they reduce ambiguity. Therefore when writing a definition that depends on one other definition, it is better to refer to list instead of the lists' parent definition. Second best is to refer to the list, and mention the definition in case the list isn't long enough. But if a definition depends on more than one other definition, lists have conflicts (licenses in one but not the other). Therefore when writing a definition that depends on two or more definitions, it is better to refer to the definitions with boolean ANDs, and mention the lists to so

Ive just been made aware of the Worldcat: http://www.worldcat.org/copyrightevidence/registry on one of the kid's literature lists I'm on which outlines copyright evidence that has been added to books. Looks nascent at the moment but might be useful for any searching the out of copyright or openness of a text and appears to be user generated. Iain -- Iain Emsley Blog: www.yatterings.com Mobile: 07942 259725

Dave Crossland dixit: This is, since AND means it must be approved by both. The natural language “and/or” would be what you want, I think. Maybe point to the OSI list, but add a list of “should be in the OSI list” licences to an explanatory page? //mirabilos

[datapkg/cli][s]: fix some bugs in cli/index found courtesy of robmyers.

[datapkg/misc][s]: fix README.txt to talk about datapkg man rather than datapkg info plus make ckanclient a requirement (now it is on pypi).

[ckanclient][xs]: bump version number to 0.1 for release to pypi.

2008/9/3 Thorsten Glaser at >public.gmane.org>: Sorry, I'll try again: Definitions have ambiguity. When writing a definition, it is good to limit ambiguity. Definers write approved example lists to reduce ambiguity. Lists are finite, which is how they reduce ambiguity. Therefore when writing definitions that depend on other definitions, it is better to refer to lists instead of the lists' parent definitions. Second best is to refer to the lists, and mention the definition in case the list isn't long enough. Any serious license will be submitted to FSF and OSI lists. Since the FSF list is also included, and they don't discriminate like that (AFAIK), if a less well known etc license is rejected unfairly by OSI it will still be listed by the FSF. Since the OSSD definition says "OSI boolean-AND FSF approved licenses" this isn't an issue.

[thread][m]: various misc fixes to thread user experience.

• Fix it so successful thread create now leads dirctly to thread update/edit page.
• Add docs to side of thread edit page
• Fix up template when no factlets in thread.

Removing milton directory at root as no longer used (now in trunk/).

fixed:

See r176.

Refactor to use SQLAlchemy rather than SQLObject

fixed:

Xapian based search implemented in r169, r191, r191 etc.

fixed:

See r187, r193 etc.

A reminder that a few of us are meeting on the #okfn IRC channel at oftc.net at 1900 BST tonight! If anyone has anything open knowledge related they'd like to discuss, propose, or tinker away at - please pop in and say hi! As well as using an IRC client, you connect using the web-based service Mibbit at: http://ur1.ca/4fh There's a rough agenda at: http://okfn.org/wiki/meetings/2008-09-03 Please add your name to the wiki if you're planning on coming. It would be great to see people there! Jonathan

Dave Crossland dixit: 1) The OSI list of approved licences is not a definition. The OSD is. 2) You have to weigh between ambiguity and the OSI's _admitted_ and _willingly_ done discrimination against licences with less (or less well-known) "community" (users, licence writers, etc). I for one would advise to not follow OSI's discriminatory actions. //mirabilos

[templates/thread][s]: minor fixes to thread read view to improve instructions and provide link to geochrono view.

[js/factletdetails][m/l]: add support for inpage updating of factlet details in thread view courtesy of a new factletsdetailview and and a new threadreadpagecontroller.

• factletdetailsview works by pulling templated html directly from server via factlet/read_core. May want to change this to reduce coupling of js and python server side code.

Nah, the final straw is that all their customers hate them for producing something with very high costs and minimal incremental value. They've been dead for a while- it was just a matter of time before someone played the Haley Joel Osment role. There will be very few large companies left; the industry just got a lot smaller. Luis

Surprised it took this long. The textbook industry, like the Unix industry before it, has been ripe for a gutting by a low-cost-structure, peer-production competitor for a long time. Luis On Tue, Sep 2, 2008 at 7:39 AM, Iain Emsley

I wonder if the torrenting of text books might have been the final straw. I'll be curious to see how it pans out for both small and large companies. Iain Quoting Luis Villa at >public.gmane.org>: -- Iain Emsley Blog: www.yatterings.com Mobile: 07942 259725

2008/9/2 Thorsten Glaser at >public.gmane.org>: Referring to approved license lists is unambiguous. When writing a definition that depends on other definitions, it is good to limit ambiguity. So I think the OKFN's Service Definition ought to refer to license lists rather than other definitions. "Liberal input, conservative output" and all that.

Afternoon, I've just seen this article on Wired which I thought might be of interest: http://blog.wired.com/business/2008/09/open-source-tex.html Iain -- Iain Emsley Blog: www.yatterings.com Mobile: 07942 259725

Dave Crossland dixit: Sure, but even the OSI themselves say that being OSI approved is only a subclass of being OSD compliant Open Source (tm) software. They admit to willingly not approve Open Source (tm) licences. //mirabilos

[factlet][m]: fix up factlet read views to show all the new attributes.

2008/9/1 Rufus Pollock at >public.gmane.org>: No probs :-) Yes, Fedora confirmed there were 0 programs with these licenses in its licensing audit last year, as there was some efforts to get Fedora on the GNU Project's Recommended GNU/Linux Distributions list. I'm not sure where that is stuck. :-) Yes :-) Thanks! :-)

First up: thanks for these excellent comments Dave. [snip] Sounds reasonable -- plus those don't seem to be major licenses so them getting excluded isn't really going to make anyone unhappy I would imagine ... Having read your links and comments I agree. [snip] Interesting use cases. As you say the change to 'made available to its users' would address all of these I think. I'm happy with this suggested mod. For the moment it can get wrapped into the 'development version' and if there is no subsequent objection :) it will go into v1.1. ~rufus

I'm always happy to help! :-) Think we could usefully list the interactions you want to support, identify a core and make them work as well as possible. Your suggestion for the main page would be the lid on this pan. :-) J.

Thought these blog posts might be of interest if people hadn't already seen them: http://powerofinformation.wordpress.com/2008/08/19/packaging-data-for-reuse/ http://www.tom-watson.co.uk/2008/08/taggov/ Jonathan

I've been talking with various people about rehauling the main OKF site (okfn.org). Bits of it need pruning, and I think the whole thing could do with a bit of re-structuring to make it easier for people to understand what the OKF is about. I quite like the idea of a severely pared down main page - perhaps with a few basic icons (!) linking to main resources, a one line mission statement, and so on. It would also be great to have a few more useful things on the site. (E.g. an events calendar - which we've been talking about for a while..) What do people think? How should it look? Any suggestions for things to use? Anyone keen to help? Jonathan

I am going to start spending time on the #okfn IRC channel at oftc.net on Wednesday evenings at 1900 BST. If anyone has anything OKF related they'd like to discuss, propose, or tinker away at - please pop in and say hi! As well as using an IRC client, you connect using the web-based service Mibbit at: http://ur1.ca/4fh Possible things to start with this coming Wednesday: * OKF events - including events in London this autumn and OKCon 2009; * moving forward with CKAN; * any open knowledge projects (including OKF projects); * re-designing the main OKF site; * an open knowledge competition? I've put a tentative agenda at: http://okfn.org/wiki/meetings/2008-09-03 It would be great to see people there! Jonathan

2008/9/1 Thorsten Glaser at >public.gmane.org>: While I am no fan of the OSI, I think it is fair to say that they are very influential and most people refer to FOSS as OS. I think its important to be inclusive instead of divisive, while maintaining a strong position on freedom.

Dave Crossland dixit: This is wrong: FOSS is much more than the licences whose submitters’ noses are pleasant to the OSI. I don’t disagree with the AND, I disagree with the OSI. //mirabilos

[js/controllers][s]: fixes to get controllers working when 'console' not available.