[okfn-discuss] [open-science] Practical reproducible science; implications for data storage

[Matthew Brett]

Hi Peter,

On Wed, Apr 11, 2012 at 1:34 PM, Peter Murray-Rust <pm286 at cam.ac.uk> wrote:
>
>
> On Wed, Apr 11, 2012 at 8:57 PM, Patrick Anderson <agnucius at gmail.com>
> wrote:
>>
>> Matthew Brett wrote:
>> > any prospect of this kind of infrastructure
>> > being available as an open service?
>> > What about OpenStack?  But then,
>> > who will pay for the infrastructure?
>>
>> Hello Matthew and all,
>>
>> I have been working on a organizational solution to the lack of "Free
>> as in Freedom" hosting.
>>
>> I am very excited about my findings, but can find few others with
>> these same interests.
>>
>> Is this within the scope of OKFN,
>
>
> Almost certainly a big YES.
>
> The OKFN is pragmatic and has also asserted that their must be no technical
> barriers to Openness while recognising that computing is not zero cost. So
> if it costs (say) 10 dollars per year that should not be a barrier to
> openness while 1000 almost certainly is. My rough rule is "similar to the
> cost of posting it" - but that's just me.
>
> If they build a walled garden, however, then that's a different problem

I wasn't sure what you meant here.   Firing up a small number of
Amazon nodes is fairly cheap.  I think hosting public data is free
[1].  Do you think that using Amazon is within the bounds of openness
as OKFN would see it?

One motivation for open-access publication is that people with little
or no money can read the papers.  Does this also apply to computing do
you think?  For example, I could imagine that relatively wealthy
universities would get scientific benefit from sharing resources with
less wealthy institutions, such as schools or universities from poorer
countries.

See you,

Matthew

[1] http://aws.amazon.com/publicdatasets/