The BAC interpreter – which remained incomplete last year – received the introduction of file input/output, memory/forms management, string operations and terminal interactivity (the latter to enable a REPL “read-evaluate-print loop”, implemented using the ncurses programming library).

A range of XML tools were developed: one for extracting namespace URIs and another to then automatically apply corresponding XSL stylesheets to the input file. Also, the C++ StAX library was enhanced to recognize and report XML namespaces.

Some simple dashboard generators were produced. These can load the Peeragogy Project’s “Peeragogical Action Reviews” from the GitHub repository as well as video links from a YouTube playlist in order to fill obtained entries into the available slots.

Within the pattern catalog management server Web package, it’s now possible to configure slider controls for the data input forms. Typical use might be for polling/voting, but otherwise generally allows the selection of a value/option from a pre-defined range of choices.

A tool was developed to pivot a XML element hierarchy. Later, the method was expanded to transform an entire XML tree into a multidimensional graph structure, so it becomes navigatable by the existing user interfaces as well as with a new dashboard alternative, too.

An independent experimental immitation of a NoFlo runner/environment in C++ was attempted. It is also able to invoke the capabilities written in Java and could potentially be adopted for broader workflow orchestration. Unsurprisingly, reacting to results reported from the execution of a node for subsequent processing proved non-trivial.

The old XSL stylesheet for converting a WordPress RSS feed to HTML was replaced by a proper dedicated tool. This makes it easier to automatically generate the usual target formats from a WordPress instance.

In the field of augmented reality, the earlier Progressive Web App geolocation prototype was expanded by adding device orientation (“compass”) in order to show or hide data based on the direction the device is pointed towards – essentially clipping a limited viewport from the 360° of all the data that’s anchored near to and around the user’s current position.

The “Text Reckoner” (or “Tero” for short) was created, which is a tool written in C++ to perform pattern recognition and filtering operations on input text as controlled by formal language grammars. Additionally, streaming library interface implementations were produced for C++ and Java. Furthermore, using Tero, a simple Web “browser”/viewer was started. To help with grammar debugging, the step-through “Interactive Tero” was made, but was neither elevated into a full editor nor supports a break-point mechanism yet.

A downloader for retrieving the profiles referenced in the RSS feed of Derek Siver’s /now page list was realized.

An exploration was started with the goal to clone Mark Carranza’s “mx” (“memory experiment” conducted since 1984 of him taking notes and connecting textually similar entries into a web/graph, thus increasing re-discovery) as an opportunity to improve on the use of the ncurses library, to grow the graph tooling and to enter into interactive TUI programming. The initiative encountered a blockade and therefore was halted.

For Frode Hegland’s “Visual-Meta” (a scheme to print bibliographic metadata together with pointers about structural organization in the form of BibTeX at the end of a document for later OCR extraction), a rudimentary workflow was set up to add the escaped BibTeX to the LaTeX source so it ends up in the PDF, as well as another one for adding such generated special pages to already existing PDF documents retroactively. While OCR extraction remained inaccurate, the BibTeX was obtainable again from the plain text of PDF files, given the original digital files stay available. Common BibTeX tooling then did the format conversion to aid the collection, visualization and navigation of such metadata, and be it for print/PDF export from there once more to complete the full round-trip/cycle.

In the wider context of Jerry Michalski’s “Open Global Mind” group, several attempts were made to develop some “Emergent Event Sensemaking” practices, but ultimately were abandoned for lack of interest despite showing promising results. Topics covered include the GameStop stock trading short-positions squeeze, the COVID-19 Delta variant surge and the exit of US troops from Afghanistan.

For the TheBrain product, the graph navigation interface was generalized and adapted to independently browse an online graph using their Web API. Separately, a native TheBrain downloader was developed, which crawls the entire graph of a Web TheBrain online instance and converts the nodes to local standalone HTML pages for offline browsing.

Copyright (C) 2022 Stephan Kreutzer. This text is licensed under the GNU Affero General Public License 3 + any later version and/or under the Creative Commons Attribution-ShareAlike 4.0 International.