Oh yes, there is one more thing I wanted to add here before we all run off and trip over ourselves to get in line to work on core extraction tasks[1] ;-) ...
The plan for the extraction is moving ahead, and we are leaving the discussion of community fork of nupic (argued for by Fergal and Stewart in this thread) behind for now. I've talked with both Fergal and Stewart about this, and it seems we are all on the same page at this point, and ready to move ahead with the core extraction. The idea of a community fork (a "free fork", as Fergal calls it) is not out of the question. If the community demands a fork of NuPIC in the future (after we do some housecleaning with NuPIC), AND there is enough momentum behind that movement, AND the instigators of the fork provide proper technical justification for the fork, we (Numenta) will attempt to provide support for it as a part of the Numenta collection of machine intelligence projects. That being said, I believe that situation should evolve organically and be result of clear technical frustration from the OS community. I think we'll all realize this situation when it arises. The important thing is that the community remains cohesive and works together on common goals towards a new era of machine intelligence! [1] https://github.com/numenta/nupic/issues?labels=core+extraction --------- Matt Taylor OS Community Flag-Bearer Numenta On Wed, Jan 29, 2014 at 2:54 PM, Matthew Taylor <[email protected]> wrote: > Thanks everyone for your input. Further discussion and planning for the > extraction is continuing on the nupic-hackers list > (*http://lists.numenta.org/pipermail/nupic-hackers_lists.numenta.org/2014-January/000169.html > <http://lists.numenta.org/pipermail/nupic-hackers_lists.numenta.org/2014-January/000169.html>*). > Please subscribe to nupic-hackers here if interested in the implementation > of this work: > > http://lists.numenta.org/mailman/listinfo/nupic-hackers_lists.numenta.org > > --------- > Matt Taylor > OS Community Flag-Bearer > Numenta > > > On Tue, Jan 28, 2014 at 2:55 AM, David Ragazzi <[email protected]>wrote: > >> Hi, >> >> I haven't read all messages up to now, so I can talk some nonsense.. but >> Scott's vision is pretty similar than mine. >> >> I believe Numenta enginers are right on their concerns regarding >> stability and other issues as already exists workable products running from >> current repository. And I dont see any need for "reinvent the wheel" >> according to another post from Matt, because we just need a good and >> heavy refactoring. But as Scott and Matt said, this is a tricky job, we >> have do this in a delicate way.. so.. all we need be patiente... I would >> love work in a new clean house. Because this, I've made or supported >> several contributions related to CMake, repository structure, C++11, etc. >> >> I'm pretty sure that when we have a clean house with a separated core, >> we will can make better decisions, and contributions will be done according >> to community members, i.e. some will contribute with core, others only in >> grok, and so on... >> >> Anyway, I'm suspected to give opinion, my friends from OpenHTM know I'm >> addicted to clean and reorganize things (to better). :-) >> >> Best, David >> >> >> On 27 January 2014 06:22, stewart mackenzie <[email protected]> wrote: >> >>> On Mon, Jan 27, 2014 at 8:26 AM, Fergal Byrne < >>> [email protected]> wrote: >>>> >>>> I have a suggestion (really a phasing) which appears to match both main >>>> camps' arguments: >>>> >>> >>> >>> This is an interesting strategy that ensures the best for Grok camp and >>> the community camp. >>> I appreciate this point of view, but it isn't needed. >>> >>> Instead of finding a solution that satiates two camps. Lets blast the >>> differentiating illusion apart. >>> What unifies us is the creation of a technically superior software >>> implementation of the neocortex that remains true to biology. >>> >>> Might I remind everyone on this list we are nupic-hackers, not Grok >>> employees, nor husbands, wives etc, >>> Here our concern is technical superior sw neocortex with a high degree >>> of biomimicry. >>> Not Grok, not Google's intelligent index thingy, nor George's pet >>> project takes precedence. >>> In here there is no hierarchy we listen to other than what nature >>> dictates as revealed by our neuroscience comrades. >>> >>> Our aim is to deliver something that is efficient, clean and true >>> biomimetically. >>> >>> Let's keep things simple. Fork, strip off abstractions, delicately >>> remove core then make a decision upon a direction from there. >>> Grok and other clients will wait at NuPIC till we (nupic-hackers as a >>> united entity) determine a clean direction forward for nupic.core >>> a direction that has regression tests, builds on all platforms, and is >>> well commented and documented. >>> We aren't doing anything dramatic, it's a house cleaning, a >>> reorganization with slight renovations here and there. >>> >>> What we come up with will satisfy. >>> It'll be a grand re-opening under the new owners (nupic-hackers) >>> Grok and other guests in their lovely gowns will be seduced into living >>> in our house. >>> >>> Kind regards >>> Stewart >>> >>> >>> P.S </propaganda> >>> >>> _______________________________________________ >>> nupic mailing list >>> [email protected] >>> http://lists.numenta.org/mailman/listinfo/nupic_lists.numenta.org >>> >>> >> >> _______________________________________________ >> nupic mailing list >> [email protected] >> http://lists.numenta.org/mailman/listinfo/nupic_lists.numenta.org >> >> >
_______________________________________________ nupic mailing list [email protected] http://lists.numenta.org/mailman/listinfo/nupic_lists.numenta.org
