My reference to “time is short” is that when trying to put this into

the context of societal pressures and make the point that today’s issues with 

monopolies and distorted markets is nothing new: we’ve been here before 

many times and dealt with these issues as a set of societies many times in

the past.  - that normally takes long than 20 minutes to go though.





The conclusion that the critical points of vulnerability where we could’ve 

changed the direction of momentum are indeed well in the past. Much of these

decisions were made unconsciously in the early to mid nineties in my view. 

For example, when the US Library of Congress passed over the opportunity to

provide search over its catalogue of material and passed the indexing function

over to the private sector thereby. Another example is the issue over whether

carriage paid content or not. The fight occurred in the mid-nineties when the

content model was foundering and representations were made to the ISPs saying

“well the only reason why you have paying customers is becuase of the content

we are providing. You, the IUSP, should fund us.” The ISPs pushed back hard

and the content folk were pushed to look directly at customers for revenue.

They solved their problem in a way the ISPs never dreamed possible, and at a scale that

the emerging ISP business never dreamt possible.



So yes, I think we are too late.



I can cite the stalled GENIE project in the US from 4 - 5 years ago. They

correctly guess that the infrastructure was entering into a world of abundance

of comms, computation and storage and they wanted to set up a highly elastic

testbed where US researchers could play within such a space at realistic scale.



It largely failed, and OpenFlow, SDN and similar memorialise the failure

of that effort. Why did it fail? Becuase the underlying assumption was that while

they assumed an abundance of comms, compute and storage they still assumed

that there was still some scarcity pressure that provided sufficient incentive for

sharing - so virtualisation and segmentation of a common resource would still

be relevant. So right answer, but just not enough scale in their starting assumptions!



But the abundance is so great and the level of centrality so overarching that

none of these actors have the slightest incentive to share any more. So they separately

design, construct and operate their own infrastructure - no sharing.



That’s way I feel it is just too late to bend this around now.



> On 11 Nov 2021, at 8:27 am, Adrian Farrel <[log in to unmask]> wrote:

> 

> Hi Geoff,

> 

> Thanks for finding a slot at crazy-o'clock to talk to us today.

> 

> Actually, I didn't find your message so depressing from a technology point

> of view (the social message is something else and it brings us to the debate

> between computer engineering and social engineering which is a whole

> different place).

> 

> One thing that made me think was that you started by noting that "time is

> short" and ended with "it's too late." Do you think we are on an inevitable

> path, already at the destination, or are we embarked on a voyage where we

> could pull back?

> 

> In the latter case (which I guess I hope for) is there stuff that the

> research community can do to help us understand the situation and maybe make

> the right choices. I guess that would not be totally a technical piece of

> work, but the engineering research might feed and strengthen the societal

> debate.

> 

> Best,

> Adrian

> 





########################################################################



To unsubscribe from the SARAH list, click the following link:

https://www.jiscmail.ac.uk/cgi-bin/WA-JISC.exe?SUBED1=SARAH&A=1



This message was issued to members of www.jiscmail.ac.uk/SARAH, a mailing list hosted by www.jiscmail.ac.uk, terms & conditions are available at https://www.jiscmail.ac.uk/policyandsecurity/