align-toparrow-leftarrow-rightbackbellblockcalendarcamerachatcheckchevron-downchevron-leftchevron-rightchevron-small-downchevron-small-leftchevron-small-rightchevron-small-upchevron-upcircle-with-crosscrosseditfacebookglobegoogleimagesinstagramlocation-pinmagnifying-glassmailmoremuplabelShape 3 + Rectangle 1outlookpersonplusImported LayersImported LayersImported Layersshieldstartwitteryahoo

Meetup to talk about encapsule Project and the Schema HTML 5 application

Chris Russell will be giving a talk on his encapsulate project as well as the schema html5 application, details are shown below:

Encapsule Project:

The goal of this effort is to create an open source ecosystem of easily discoverable, easily re-usable, software intellectual property encapsulated in JSON-encoded Soft Circuit Description Language (SCDL).
SCDL represents encapsulated software IP using dataflow models (directed graphs essentially).
SCDL is an object model where objects represent chunks of "virtual circuitry".
SCDL models can be interpreted or compiled to executable runtime code.
SCDL objects may be composed via object containment and various object linking relationships to model large software systems.
SCDL "specifications" (a specific type of SCDL object) represent a large system.
SCDL "specifications" are transformed to runtime code.
This process can involve transformation, interpretation, compilation...
The point is that SCDL specifications are executable, or can be generically transformed into something that is executable.
SCDL is intended to be 100% portable:
No inherent coupling to hardware
No inherent dependency on any operating system
No inherent dependency on any programming programming language

Join or login to comment.

  • Chris R.


    I've gone off and given this a bunch more thought and done a _ton_ of work to ensure that my Javascript codebase elegantly supports the programmatic generations of SCDL specifications as would be required to leverage this system for building code generators, program transformation systems... (For example a program that takes as input a SCDL model set, and a graph and derives via traversal, a SCDL specification of a new program).

    The solution is a spin-off Javascript library I'll be releasing very soon called ONMjs (Object Namespace Manager). ONMjs is _generic_: it's not coupled with SCDL whatsoever. So you could, for example, declare your own data model and leverage ONMjs as an abstract class factory for your own data (in client code or via Node.js)

    I'm limited to 1000-chars in comments so here's a link to several blog posts that give more detail:

    If anyone is interested in messing around with this get in touch.

    Regards, Chris

    September 20, 2013

  • Chris R.

    Thanks to Saikat and Disney for hosting last evening as well as for everyone who came for the talk. I really appreciate the opportunity present my ideas regarding program composition and modeling.

    My notes from last evening's talk are posted here:

    In the short-term, my goal is to provide a means to create, edit, visualize, and share SCDL model sets using the Schema HTML 5 client app.

    The client app is modular and w/a small bit of extra work the core SCDL object factory and schema manager can be split off as a standalone JS lib allowing for programmatic manipulation of SCDL model sets.

    Programmatic manipulation of SCDL model sets enables many extremely interesting scenarios. For example, generating purpose-built programs bound to specific input data stores. Please follow Encapsule Project on Twitter for updates and thanks again for coming out. - Chris

    May 31, 2013

    • A former member
      A former member

      A theme arose that we continued talking on over a beer was how do we use SCDL model sets with Graphs and ML. Adam Coates PhD dissertation picks this up. In order to succeed, most learning algorithms are applied instead to "features" that represent higher-level concepts extracted from the raw data. These features, developed by expert practitioners in each field, encode important prior knowledge about the task that the learning algorithm would be unable to discover on its own from (often limited) labeled training examples. Unfortunately, engineering good feature representations for new applications is extremely difficult [kudos Chris Russell]

      In this work we study a class of algorithms that attempt to learn feature representations automatically from unlabeled data that is often easy to obtain in large quantities.


      1 · May 31, 2013

    • Chris R.

      Thanks for the link and citation. I had never actually considered SCDL as a means of engineering "feature representations" but the binding possibilities seem vast. I'm going to have to think some more about this one :) Regards, Chris

      May 31, 2013

  • Michael

    Good questions and discussion.

    May 31, 2013

  • Joseph S.

    Hate to miss this one.. a higher priority event interrupted.. Have fun..

    May 30, 2013

  • HK H.

    The idea of "Software ICs" has always been very appealing in terms of cost, productivity, and reliability. Wm. Wong wrote a book on "plug & play programming" ~20 years ago. Then there was the case made for "software factory/product line" about 10 years ago. Am looking forward to the latest take...

    1 · May 30, 2013

  • Silvia V.

    something has come up. sorry to miss it!

    May 28, 2013

  • Chris R.

    1 · May 21, 2013

  • Chris R.

    Project website is here:

    I would be very interested in the "questions that come to mind" when members of this group read the essay above.

    If you post your questions here I'll try to answer them now, or at least come to the Meetup prepared to discuss (I can imagine some seriously hard questions :)

    Thanks and I look forward to meeting you folks!

    Regards, Chris

    1 · May 3, 2013

21 went

Our Sponsors

People in this
Meetup are also in:

Sign up

Meetup members, Log in

By clicking "Sign up" or "Sign up using Facebook", you confirm that you accept our Terms of Service & Privacy Policy