Burgess COMMENTARY

Peter Burgess

Fwd: more important than it sounds Inbox x alan longley Jun 19 (2 days ago) to me, John John, Have you ever explored non-Euclidean geometries? It looks like SAP now owns the patent on all tree structures in geometries with negative curvature, which is more useful than it sounds. Alan Longley ---------- Forwarded message ---------- From: alan longley Date: Mon, Jun 19, 2017 at 11:42 AM Subject: more important than it sounds To: Olinga Ta'eed Cc: Bruce Longley , barbara.mellish@cceg.org.uk Dear Olinga, You wrote: << the UK£40m bid for the educational platform now not only includes Microsoft, but also SAP and Cisco >> SAP, I believe, ended up owning the patent for hyberbolic trees. This patent was originally owned by Xerox along with the patents for the mouse and windowing systems. I forgot the whole history of how mouse and windowing IP was transmitted to Microsoft and Apple, but I have long thought that both computers and the web would look much different if the hyperbolic tree patent had been used in conjunction with mouses and windows. It might be this patent that SAP now owns: US patent 5590250, Lamping; John O. & Rao; Ramana B., 'Layout of node-link structures in space with negative curvature', assigned to Xerox Corporation There aren't many good examples of hyperbolic trees on the web because it is patented. This link, for example, is probably a patent violation: http://ocsigen.org/js_of_ocaml/files/hyperbolic/index.html There you can get a just a glimpse of how important this structure is - but it is rather weakly presented. Notice the alphabetic text at node labels? Chinese characters carry vastly more information compactly. Hyperbolic trees would be more useful in China than they are here. I think SAP acquired the hyperbolic tree patent when it bought: Inxight StarTree A company that marketed hyperbolic tree data representation to governmental entities handling large data sets. In the era of big data the utility of hyperbolic trees is self evident. Here's something from the web; the edges of the circle extend to infinity. Any tiny thing on the edge can be brought towards the center where it will be big and details appear. A mouse could draw the node to the center or this could be accomplished algorithmically: The origin does not have to be in the center of the fisheye as written above. The focus can be changed by the user either by clicking on any visible point to bring it into focus at the center or by dragging any visble point interactively to any other position:
HypTree changinfocus.jpg The educational platform you are developing might benefit from this technology - but I am not here making a good case for it - just thought I would put out the idea seeing SAP on your roster. The exact center of the circle is in theory infinitely large, something of course impossible to accommodate practically. There used to be a good display of hyperbolic trees at this opensource (java) project (I have a copy some place - i think the download protocol from sourceforge is odd) but the site no longer displays hyperbolic trees & all links to example projects go no place useful): http://hypergraph.sourceforge.net From the title of the patent: 'Layout of node-link structures in space with negative curvature' It might well be that the patent generalizes to all non-Euclidean geometries with negative curvature, not only hyperbolic geometries- just a guess. If correct, the 'hyperbolic tree' moniker is inaccurate, not sufficiently general. Best, Alan Longley
===========================================================================
idea of the day Inbox x alan longley 3:45 PM (18 hours ago) to Olinga, Olinga, Bruce, barbara.mellish, me Dear Olinga, Frustrating day - a thousand places to push on strings, only a couple place to pull. I am trying to resurrect the day by asking one question and providing one idea. Question: I didn't see in the documentation a mining mechanism for the seatio coin. Am I correct in assuming it will be able to be mined using the open source ethereum methods - ethereum 'proof of work'? Idea: Let's say there is a proof of authority blockchain (as opposed to proof of work) containing authorized care providers, say care of elders. Those authorizing care providers require their own blockchain, also proof of authority. Let's call members of this blockchain auditors. For discussion purposes, let's say auditors are affiliated with churches, mosques, synagogues, or some humanitarian organization. Call them member organizations. To make things simple, assume auditors authorize care givers a certain number of hours per day during which they provide care (better allocation mechanisms doubtless exist). Member organizations start out with a certain balance of coins with which they allot to caregivers based on hours of care. Member organizations are therefore constrained in allotments to caregivers. But member organizations are also permitted to conduct mining on some coin in proportion to the care giver hours it authorizes, or perhaps in proportion just to number of care givers. One might design the system to achieve pareto equilibrium - so the initial balance stays the same. Or let it grow some. Perhaps reward auditors some way, perhaps through achievement of pareto equilibrium This is an idea in rough form Best, Alan
===========================================================================
Olinga Ta'eed 3:11 AM (6 hours ago) to alan, Bruce, barbara.mellish, me Yes it’s mined using Ethereum Proof-of-Stake (not Work). You know it’s remarkable how closely your ideas and ours are aligned. So before your email this is how we are preparing to do the same: 1. The Proof-of-[…] is based on our S/E ratio translation and captures the attributes of whatever organisation we are dealing with (as you say). 2. S/E is calculated for those organisations and they are rewarded in Microshares that they can issue as rewards 3. The reward is fully distributed and decentralised as they can use their own certification systems but S/E is the translator into the blockchain. 4. The only difference is we thought it would be too complex to have nested blockchains, so were going to ‘license’ S/E to these organisations. Very very close to your idea! From: alan longley [mailto:alanglance@gmail.com] Sent: 20 June 2017 20:46 To: Olinga Ta'eed; Olinga Taeed Cc: Bruce Longley; barbara.mellish@cceg.org.uk; Peter Burgess Subject: idea of the day
===========================================================================
alan longley 8:44 AM (1 hour ago) to me Dear Peter, It upsets me some to see two members from Microsoft on the CCEG's 40 million pound application for a computer aided education system for Britain. The addition of SAP to the scene provides some hope in that SAP embraces Linux. Another hope is that SAP would license the group to use the hyperbolic tree, as developed in a past email. It's hard for me to image anything but a clumsy rendition of hyperbolic trees in windows (though advances in windows usage of GPUs might have been made, dunno). I would hope that the bargaining position of SAP in the group would rise relative to microsoft because of the hyperbolic tree patent Idea: frame your accounting metrics for use as aspect of hyperbolic trees and SAP. Best, Alan addon Inbox x alan longley 8:52 AM (57 minutes ago) to me Peter, Half the 40 million pounds comes from taxpayers. Muster your British countrymen and ping the right people at HEFCE to attempt negotiations to include the hyperbolic tree and your accounting metrics!