Hi Nil,  Hi everyone,

Sorry, yes, its intentional,  got discussed in a meeting here last week,
and therefore "everybody knew about it" ... except, of course you :-)

To recap, this is https://github.com/opencog/opencog/pull/2641 which
resolves a circular dependency between opencog and the robotics code.   Its
now arranged so that all of the thinking/understanding/behavior is in the
main opencog repo, while all of the ROS nodes providing sensation (vision,
sound) and all motor movements (3D animations) remain in the
https://github.com/opencog/ros-behavior-scripting directory.

This allows, at least in principle, other, non-ROS sources of sensory
inputs, and also maybe someday a better, more advanced motor-control
interface; right now it remains closely tied to the existing blender
animations.

The moved code is in https://github.com/opencog/opencog/opencog/eva
directory. In a few days time, select portions of the chatbot will end up
there as well, although hopefully most of the nlp code will stay in the nlp
directory.

I've provisionally named the directory "eva" because, at least for now, it
is tightly tied to the eva/sophia blender animations.  Perhaps in the
future we will be able to refactor this into a more generic, disembodied
intelligence, with sensory input from wherever, and motor control over
something.  But for now, its not entirely clear how to split this up, and
so the directory is named "eva" instead of "embodiment".

One prominent subdirectory is "model" which contains a provisional "model
of the world" (which includes a model of the robot itself).  The code and
ideas here need a lot of elaboration.

Another directory is "behavior" which currently contains a mish-mash of
hand-authored stimulous-response type scripts, and the plumbing to enable
these.  This will get replaced/updated by a better authoring system that
Cassio is leading (a chatscript-like interpreter)

I am not quite sure where or how PLN will get wired in -- it will get
stapled into the language parts, somehow.  I won't be doing this work.

OpenPsi is attached to this stuff, but its sort of a dr-jekyll/mr-hyde type
system, uncertain as to its role in proper society.

I've started a refactoring of the space-time server code, but am running
out of time.

All of this code moving-about will probably generate a long list of
"buglets", breaking how things attach and talk to each other. I will
appreciate all help in diagnosing and fixing these.

--linas

On Fri, Mar 24, 2017 at 8:37 PM, 'Nil Geisweiller' via opencog <
[email protected]> wrote:

> Actually never mind, it comes from
>
> https://github.com/opencog/opencog/commit/3b2637e6c0c774ac06
> ecc779a265564fa616be0e
>
> which I suppose has been done purposely. It was looking very suspicious at
> first because it contains commits as old as 2y.
>
> Nil
>
> On 03/24/2017 02:29 PM, Nil Geisweiller wrote:
>
>> The github history on the opencog repo is seriously f-up. Please do not
>> commit anything before I figure out what went wrong.
>>
>> Nil
>>
>
> --
> You received this message because you are subscribed to the Google Groups
> "opencog" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to [email protected].
> To post to this group, send email to [email protected].
> Visit this group at https://groups.google.com/group/opencog.
> To view this discussion on the web visit https://groups.google.com/d/ms
> gid/opencog/f20bb1a4-6ea0-d6bc-56e0-22d07c9f5da2%40gmail.com.
>
> For more options, visit https://groups.google.com/d/optout.
>

-- 
You received this message because you are subscribed to the Google Groups 
"opencog" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To post to this group, send email to [email protected].
Visit this group at https://groups.google.com/group/opencog.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/opencog/CAHrUA342Dja3CNA0_i10J7znYy%3DFkNVfthUf4oE2pGWbc6b72w%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.

Reply via email to