[okfn-discuss] Open Knowledge Foundation Strategy slides

Mon Jul 22 09:45:19 UTC 2013

Dear Peter,

I really appreciate what you are saying, and I agree that with regards to
many of the data/content processing tools you describe there is a lot of
benefit to such a systematic approach. But I think its a huge error to mix
up stand-alone software and software as a service in how the open movement
must approach them.

Your email reflects the assumption that F/OSS is not winning is because
there isn't the right software. In some cases, that may be true. But the
bigger concern now seems to be that we really don't know how to make F/OSS
as a service. The reason that LibreOffice is such a horrible alternative to
suggest to GD/O365 is not that it doesn't have nice icons - it's that it is
a piece of desktop software that attempts to compete with a set of
integrated services. It has lost the ball, and we look silly suggesting
otherwise.

The fact that services like Google Docs, Facebook, Skype etc. are operated
in a centralized mode gives them a number of benefits that we can't just
ignore. This begins with simple network effects - all your friends are
using them, so you should, too - but it also extends into technical and
economic benefits (imagine the CKAN team never had to spend an hour helping
to update client instances or that your setup of VIM/Emacs was available on
any computer you touch), and many opportunities for improvement based on
behavioral data-mining that decentralized systems just don't support. We
all love decentralization and autonomy for a variety of political, economic
and technical reasons, but we can't let that blind us to the fact that
there are just as many things it can't do (often that's what we like about
it).

OKFN (and especially Rufus) is not naive about this, that's one of the
things I like best about it. KForge, the DataHub, OpenSpending, etc. are
all examples of open services. With OpenSpending, we've been trying to
combine SaaS and F/OSS, following a model where many satellite sites use a
commonly operated resource. I wouldn't say we have a good recipe, though.
We don't really have a good sustainability model for any of this,
governance is WIP and decentralizing the actual code development has been
tough. ScraperWiki is another good example: I really enjoy the service and
admire the folks doing it, but its latest incarnation is impossible to
distinguish from any other commercial start-up like Heroku.

So, please, can we focus on open services rather than F/OSS? The world
around us has moved on. I wish Cory Doctorow all the best on his "general
computation" campaign, but chances are that we also have to prepare
ourselves for a world in which we can't even install software on many of
our devices.

As for great ways to decentralize open services, I'm really open to
whatever ideas there are. Let's not try to use political arguments to
defuse the fact that most of the current approaches suck, though, and that
it is a matter of economics and usability that open protocols are now eaten
up at a monthly basis (RSS, XMPP, soon SMTP).

Cheers,

 - Friedrich

On Mon, Jul 22, 2013 at 10:49 AM, Peter Murray-Rust <pm286 at cam.ac.uk> wrote:

> I am sympathetic to this concern. I don't think we can be universal, but I
> think it could be useful to detail and publish the different components.
> Questions that would be relevant:
>
> * is there an existing F/OSS solution?
> * if yes, is it readily usable by experts? If yes, how much training,
> tutorial is required for community use?
> * if no, it is high priority that one should be developed? if yes, is the
> OKF the right place (resources, community), to do this? If we can't can we
> still give the world a lead (e.g. by initial design and advocacy)
>
> We would need to answer these questions for a wide range of activities -
> some examples:
>  * slides
>  * editing/creating documents
>  * conferencing
>  * calendars
>  * data extraction and analysis
>  * repositories and knowledge bases
>  * graphics
>
> .. and probably > 20 more
>
> A summary of these, with the issues would be of enormous value to the
> world. We could get a good start in a hackday. Out of this could then come
> a list of the areas where we could make most impact (coding, advocacy,
> customisation, etc.). We might also develop an "Open Audit" for OKF ( do
> other organisations do this?).
>
> To give my own example, I am keen to develop Open content mining of the
> (STEM) literature. There are some closed commercial tools which I would
> never use ((lack of) functionality, extensibility, redistributability,
> money, opaqueness of algorithm, etc.) and anyway they don't do what we want
> to. So we have to develop our own. The good news is that:
>  * there are a number of F/OSS libraries (PDFReaders, image analysis,
> natural language processing).
>  * we are developing a critical mass of community.
>  * the result will have a big impact
>
> The major reason for doing this is to liberate the content that exists in
> the scientific literature. Many publishers will claim the content is
> "theirs". I and others in OKF challenge this. So the approach is manyfold:
>  * political (and Ross Mounce and I have spent much time in purely
> political issues).
>  * advocacy and growing the community.
>  * writing code
>  * creating Open corpora
>  * customising resources and software.
>
> We have very limited resources. So if I have to choose between developing
> software for content mining  and learning how to install and run an Open
> Word alternative I spend my time on the former. But I support those trying
> to promote Open document editors. And I will switch *when the Open
> community makes it simple and cost-effective to do so*. Similarly with
> Skype, Google-stuff, etc.
>
> Changing to F/OSS will take time and resources. I am sure that there are
> some easy transitions to F/OSS that we can make if they are identified *and
> have community support*.  I'd certainly come to a hackathon.
>
>
>
>
>
>
> On Mon, Jul 22, 2013 at 4:59 AM, Mr. Puneet Kishor <punk.kish at gmail.com>wrote:
>
>> Yes, thank you Diane. It is very clear, and I somewhat agree with you. I
>> agree that open communities should use open source *to the extent
>> practical*, and I don't think open source will disappear if open
>> communities do use non-open services.
>>
>> But, instead of getting into an argument about it, an argument that
>> generally has no end, I do want to say that I agree with the spirit of your
>> statement. To the extent practical, we *should* use and support open
>> source, because then others can build upon our work and we upon the the
>> works of others.
>>
>> Many thanks,
>>
>>
>> On Jul 21, 2013, at 5:59 PM, Diane Mercier <diane.mercier at gmail.com>
>> wrote:
>>
>> > Sorry for my poor english.
>> > Gitorious.org is completely open source.
>> > Open community should adopt open source like gitorious.org instead
>> privative SaaS (e.g. Github or Google).
>> > If Open community use privative SaaS (e.g. Github or Google, etc.) than
>> open source (e.g. gitorious.org, etc.) will disappear.
>> >
>> > Is this clear?
>> >
>> > Why OKFN choose Google Drive, Github, etc. instead contribute to open
>> source development and adoption.
>> >
>> > DM
>> >
>> >
>> >
>> >
>> > Le 2013-07-21 13:11, Mr. Puneet Kishor a écrit :
>> >> On Jul 21, 2013, at 9:04 AM, Diane Mercier <diane.mercier at gmail.com>
>> wrote:
>> >>
>> >>> Or join communities of developers until these Open tools disappear
>> purchased or removal by the SaaS of this world, for example gitorious.org
>> >> I want to understand the above sentence better. Are  you saying that
>> gitorious.org *has* either disappeared, purchased or removed, or are you
>> saying that gitourious.org is an "SaaS of this world" that will purchase
>> or remove "Open tools"?
>> >>
>> >> Afaict, gitorious is completely open source (offered under a GNU
>> Affero GPL).
>> >>
>> >>
>> >> --
>> >> Puneet Kishor
>> >
>>
>>
>> _______________________________________________
>> okfn-discuss mailing list
>> okfn-discuss at lists.okfn.org
>> http://lists.okfn.org/mailman/listinfo/okfn-discuss
>> Unsubscribe: http://lists.okfn.org/mailman/options/okfn-discuss
>>
>
>
>
> --
> Peter Murray-Rust
> Reader in Molecular Informatics
> Unilever Centre, Dep. Of Chemistry
> University of Cambridge
> CB2 1EW, UK
> +44-1223-763069
>
> _______________________________________________
> okfn-discuss mailing list
> okfn-discuss at lists.okfn.org
> http://lists.okfn.org/mailman/listinfo/okfn-discuss
> Unsubscribe: http://lists.okfn.org/mailman/options/okfn-discuss
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.okfn.org/pipermail/okfn-discuss/attachments/20130722/dd0e72b3/attachment.html>