My main focus was working on the sample implementation for SRFI-167 and SRFI-168 along the continued work on a real implementation on top of WiredTiger using (Chez Akku) Arew Scheme. I managed to get together a release dubbed 0.1.1, that you can try using the following command:
The standardization process is kicking and more people got involved by sharing wishes, bug reports and in general valuable positive and negative feedbacks. My talk on this subject to was accepted. I would have liked write and submit a full paper about it, I mostly failed. I started writing a tutorial (book?) about the Ordered Key-Value Store that is called .
I got involved in some W3C mailling list related to RDF and AI.
At last, I got some energy to read things and listen to Noam Chomsky.
Here is a first quote about their "scientifical" approach to design systems:
Here is another about the relation between tools, individual accomplishments and how to achieve them:
... and allow effective learning.
Present and Future
The sample implementation for SRFI-167 in particular needs more love. It happens that I need it, for some future projects. Among others I use it in emacs-like editor and I might rely on it in the functional package manager. I have still things to figure. It seems like I am trying to shoehorn it, especially in the case of the functional package manager. In the case of the editor, it is less likely to be a mistake because indeed it is helpful and implements separation of concerns like Model-View-Controler. Having the same data-structure everywhere is helpful and it is more powerful than a hash-table or struct as it allow to easily do introspection to ease debugging.
While I was excited by datae as a demonstration of change-request mechanic over structured data that is bigger than memory. I was thinking that it might lead to making a living out-of-it with grants, community support or consulting. Getting together that software with 100% support of SPARQL is not impossible. The question is do I really want to invest my time in this particular project? At some point, "maybe" is not anymore an acceptable answer. In the meantime, I will just put it on "standby" mode.
I am discussing a possible relicencing of Arew into something dubbed business-friendly most likely Apache 2.0. This will lead to a split of the project into two components. The first will remain in Arew and will be dedicated to support and moar. The second repository will be prolly called babelia and will be licensed in a fork of the Affero GPL licence that is more humane, more sensible to the challenges faced by the anthorposcene.
I got a better idea about my Personal Knowledge Base project or if you prefer my Personal Assistant or Research Assistant. To quote Ronald J. Brachman:
Otherwise said, I need an application of the application (!) to be able to dogfood the idea. I think studying Artificial Intelligence and its history is a good subject of study, in particular I need to learn more about non-monotoic logic and ordinal number systems.