Some assemblage required

Assembling the heterogeneous elements for (digital) learning

Birds hiding in a bush

Designing a personal “memex” with Foam

For years I’ve been annoyed at the lack of structure in my approach to managing and leveraging information (it never really becomes knowledge). The process through which I Seek > Sense > Share information/knowledge has been ad hoc, disorganised and broken.

Last week @downes shared details about Foam. The first release of which was announced two weeks before I’m writing this.

The following is me figuring out how I might kludge Foam together with the other assemblages I use to help fix what is broken. Of course, it takes much more than a technology.

What is Foam? Why use it? Why not?

Foam is a tool intended to help individuals store, manage and leverage knowledge. It’s inspired by tools such as Roam and Notion, which are getting some attention. It also harks back to Smallest Federated Wiki and Wikity. Ideas that have promised to help with the problem I’ve identified above, but which I’ve never been sufficiently disciplined to effectively use. Hence no solution.

Foam has some established principles which resonate and which may help address some of the dissonance prior toools created with my personal practice. Like @downes I am attracted by the underpinning technology used by Foam – Visual Studio Code and GitHub. First, because I’m increasingly using those technologies in my work so my familiarity with them is increasing. Forming a virtuous circle, using Foam will help improve my familiarity with these tools and help my work. Second, the small pieces loosely joined approach enabled by both technologies resonates strongly with my prior work and assumptions about what works, or not. Though the fact that they still position “Build vs Assemble” as an either/or suggests that they haven’t made explicit the idea that out it is both/and.

Foam’s principles mention an intent to “allow users to operate in a decentralised” manner, which is also pleasing. I need to figure out how to integrate Foam into the assemblage of tools, practices and knowledge that will form my PKM processes. Inevitably my choices will be different than others. VSCode and GitHub are designed to more easily support this difference than some other choices. For example, some early examples of how the tools underpinning Foam might help inter-connect Foam with my existing assemblages, include:

And that list is without exploring further into the VSCode ecosystem (e.g. this citation picker for Zotero or the VSCode extension that allows typing TODO in a note to create a github issue). The presence of these ecosystems point to the strength of using technologies (VSCode/GitHub) that are widely used by a lot of people (who code). The list above illustrates that other people have already provided solutions to requirement I have. Solutions which are generally protean enough for me to customise to my context.

Why not use Foam?

It’s still early days. There remains a question whether this tool and its community will evolve productively. It’s also a bit rough around the edges and perhaps not fully feature complete. Not surprising since it is essentially two weeks since the release of a functional prototype.

Foam is not an integrated system. It’s a best-of-breed system. Each of these product models have plusses and minuses. A best-of-breed approach is more flexible and customisable. It’s also requires more knowledge to use effectively. Especially when the number of components that make up the best-of-breed system becomes quite large. As it can quickly become with Foam. Also, Foam is a best-of-breed system that you are responsible for fitting together.

The technical knowledge required to get going is relatively high. Rough around the edges means it’s probably best if you are familiar already with GitHub and VSCode. Markdown is the core authoring environment. It’s something else to learn.

The reliance on the filesystem to store data may be a limit on functionality and scalability. This type of note-taking approach (arguably) works best with a hypertext type structure. Not a hierarchical/categorical structure. The Web versus Gopher. A problem touched on by Berners-Lee talking about Bush. GitHub controls file structures. File structures are hierarchical.

What else might you use?

There’s a GitHub repository to answer that question. The Digital Gardners repo is described as “A collective of gardners publicly tending their digital notes on the interwebs”. It includes sections on Gardening Tools, How-tos, theory etc.

Interesting that @holden’s dLRN 2015 presentation gets a mention. First connection to my earlier dabbling and still one of the best descriptions of the importance/difference of this type of approach. In which he harks back to memex and explains why the following isn’t an entirely fitting name for what I might do with Foam here. e.g. contains my material and the materials of others. Links are associative. Links made by readers and writers.

Introducing Memex

My first quick test of Foam was labelled secondBrain. I wasn’t happy with that name and the following illustrates that I’ve moved further away from that term. To quote Bush (Bush, 1945)…”it needs a name”…

Consider a future device for individual use, which is a sort of mechanized private file and library. It needs a name, and to coin one at random, “memex’’ will do. A memex is a device in which an individual stores all his books, records, and communications, and which is mechanized so that it may be consulted with exceeding speed and flexibility. It is an enlarged intimate supplement to his memory.

While this use of Foam is certainly not going to be functionaly equivalent to Bush’s “memex” or later conceptualisations informed by the Web and greater hypertext thinking. The purpose certainly fits. I’m hoping the disciplined use of Foam will provide “an enlarged supplement” to my memory etc.

So welcome to my personal prototype memex. That will take you to the github pages representation of the memex. If you’re really keen you can access the github repository.

As it stands it is pretty limited. I still haven’t figured out completely how to use it.

How to use it?

One of the Foam principles is “Foam is not a philosophy”. i.e. how you use Foam is really up to you. Figuring how to use Foam in an effective and disciplined way is perhaps the hardest task. A task I’ve not been successful with to date with other tools. The following is some exploration and thinking about how I might turn this around.

The Foam recipes section on Worflow is currently empty. But Foam does talk mention Second Brain and Zettelkasten (which is not a new idea). There are others. The digital gardeners repo mentions some. Following explore Zettelkasten, SecondBrain and Seek > Sense > Share for ideas to inspire how I might use Foam.

Zettelkasten

This Medium post on Zettelkasten (a form of notetaking) describes the zettelkasten method as relying on a sequence of activities: Collect, Process, Cross-Reference and Use. It also lists the following requirements

  • Notes are atomic.
    i.e. as mentioned here, make them agnostic to a parent topic.
  • Each note has a home.
    But also points out the limitation of a topic based organisation as notes can belong to multiple topics. An emergent home.
  • Each note has a name.
  • Notes link to other notes.
  • Cross reference keywords.

A more detailed post on zettelkasten offers the following “tools for the zettelkasten method”

  • Inbox.
    Where rough ideas are captured for further processing.
  • Reference manager.
    How you cite references. Mentions Zotero. Rasing the question of how to cite references in Markdown/VSCode? Hey presto, there’s a VSCode package for that.
  • The Zettelkasten
    Recommneded that this be plain-text. This is where Foam most explicitly enters the picture.

It also offers the following process

  • Capture.
    Reading and taking literature notes. Literature notes are a collection of all the notes made on an article/book.
  • Elaborate.
    Turning notes into Zettels. Summarise contents into the Zettel title. Write for others. A structure including

    • Title/Zettel id
    • Elaboration
    • Highlights
  • Connect
    Add them into the network of Zettels. Tags. Links out and in. Structure notes become important here. Including indication of type of note in Zettel UID: L=literature note. L2=mid-level structure note. L1=top-level structure notes….perhaps too much.

    Tags. Use different types of tags ## for L1 notes.

Second Brain

SecondBrain appears to be a similar and more commercial process (e.g. there’s a company LinkedIn page). This description of SecondBrain identifies two core concepts: CODE and PARA.

CODE summarises “four universal steps”

  • Collect – a place to collect what’s interesting.
  • Organise – the structure it suggests is captured by PARA
  • Distill – notes themselves aren’t enough. Progressive summarisation.
  • Express – share with the world.

PARA suggests structuring content into

  • Projects – tasks liked to a goal/deadline
  • Areas – sphere of activity to be maintained
  • Resources – topic or theme of on-going interest
  • Archives – inactive items from the first three.

This sequence is meant to capture the flow between actionable tasks (Projects down) and non-actionable information.

Progressive summarisation appears to be the description of how to create and manipulate notes through the PARA/CODE concepts. Based on the idea of “note-first” rather than tag or category first. i.e. the primary focus is design of individual notes. Balancing discoverability (relying on compression) and understanding (relying on context) – touching on the reusability paradox.

Progress summarisation has layers of summarisation and the idea of opportunistic compression. You do it over time as needed. Not all resources go through all the layers. The layers are

  • Layer 0 – the original, full-length resource.
  • Layer 1 – initial notes/summary generated by note-taking program.
  • Layer 2 – bold passages.
    The “best parts” of the Layer 1 notes. The core of the idea.
  • Layer 3 – Highlighted passages
    The “best of the best”.
  • Layer 4 – Mini-summary.
    2 & 3 converted into an executive summary.
  • Layer 5 – Remix.
    Changes to how and where they are structured. Perhaps more linked with expressing the idea.

SecondBrain doesn’t (as of yet) capture the flow of random insights coming in???

Seek > Sense > Share

The Seek > Sense > Share Framework from @hjarche offers a conceptualisation that I’ve used personally and in my teaching for quite some time. It’s seen as a method for putting Personal Knowledge Mastery into practice. It offers more a way of thinking about what’s required rather than offering explicit guidance into the how – one of its strengths. It’s easy to see connections between Seek/Sense/Share, CODE and the other processes above.

What do I want to do with it?

Frameworks are useful, but how do I want to apply them. What do I need/want to do? Who am I? Who do I want to be?

I increasingly think of myself as a toolsmith in the sense expressed by Brooks (1996) whose delight “is to fasion powertools and amplifiers for minds” (p. 64). I’m a nerd who builds things that people want to use for pesonally productive ends. Mostly in the realm of (formal) educational technology. Consequently, it’s much more complex than just technology. Hence the current title of this blog – some assemblage required. Theory about people, learning, systems, organisations, teaching etc is required. Hence the tasks I need to undertake could include: reading, writing papers, developing software, thinking.

I’m not sure I see SecondBrain’s PARA approach fitting here. I won’t be using Foam for everything. Just tracking of notes. Zettelkasten seems a better fit for the purpose. Plus I’m worrred about too explicit a categorisation scheme and especially Foam’s reliance on the file structure and subsequent scalability issues. All notes in a single directory?

The purpose

The intent is that my memex will be where I store more notes which eventually become Zettels. The workflow will mirror Seek > Sense > Share. With memex being the core of Sense but having connection swith Seek and Share. Perhaps suggesting that Seek > Sense > Share become the structure.

Initial design

The home page could have a brief background, but then links to Seek, Sense and Share. Sections on the home page but also directories in the repo used in the following way

  • Seek.
    Contains pointers (and perhaps scripts) used to triage notes from seeking into memex. Layer 1 in SecondBrain concepts. A todo list of notes that need to be made sense of.
  • Sense.
    Where all the Zettels sit. As “seek notes” are summarised and connected they move into this directory. The sense page contains lists of stucture notes and unallocated notes. Structure notes represent major areas of interest I’m working on and may correspond to directories within the sense directory. Unallocated notes are those that have been “zettelised” but which aren’t yet connected.
  • Share.
    Contain an explanation of various means I’m using for sharing, but also explicit outcomes. e.g. papers, blog posts, github repos? Etc. Some of the outcomes would be works in progress, but others may be concrete artifacts that are used as part of my POSSE model.

To do

I now need to figure out if and how this plan can be best implemented with Foam’s nascent and evolving functionality and subsequently integrated into my assemblage of technologies and practices.

Foam questions

The whole directory/file question is probably the main one at the moment. If/how to do that and what are the trade-offs?

How do wiki links work and from there visualisation and back linking?

If and how can tags be simulated within Foam? Can I add “tags” to markdown files in a way that allows the automated genration of structure notes?

And I’m sure many others

Assemblage integration

For this to work I need to connect memex to the rest of the tools I use

Seek

  • Fix my Zotero installation.
  • Translate notes from various sources (Kindle, Zotero, Diigo, Hypothes.is, Wikity) into memex.
  • Put in place process to keep those notes flowing into this.

Sense

  • Work out the process, structure and technology to make sense of in-coming notes.
    • What’s the landing page into memex like?
    • Structure underneath.
  • Figure out how Zotero and VSCode can effectively work together
  • Figure out how to handle moving notes around the directory structure and connection with links (wiki and markdown)

Share

  • Explore POSSE style links from Foam to
    • my blog (wordpress).
      Posting and pages?
    • Twitter.
    • MS Teams?
  • How it figures into more formal writing.

And of course the simple task of trying to get all this to fit together and use in a disciplined and effective way.

References

Brooks, F. (1996). The Computer Scientist as Toolsmith II. Communications of the ACM, 39(3), 61–68.

Bush, V. (1945). As We May Think. The Atlantic Monthly.

Understanding (digital) education through workarounds and quality indicators

COVID-19 and the subsequent #pivotonline has higher education paying a lot more attention to the use of digital and online technology for learning and teaching (digital education). COVID-19 has made digital education necessary. COVID-19 has made any form of education – and just about anything else – more difficult. For everyone. COVID-19 and it’s impact is rewriting what higher education will be after. COVID-19 is raising hopes and fears that what will come after will be (positively?) transformative. Not beholden to previous conceptions and corporate mores.

Most of that’s beyond me. Too big to consider. Too far beyond my control and my personal limitations. Hence I’ll retreat to my limited experience, practices, and conceptions. Exploring those more familiar and possibly understandable landscapes in order to reveal something that might be useful for the brave new post-COVID-19 world of university digital education. A world that I’m not confident has any hope of being positively transformed. Regardless of what the experts, prognosticators, futurists and vendors are selling. But I’m well-known for being a pessimist.

Echoing Phipps and Lanclos (2019) I believe that making changes in digital education needs to be grounded in “an understanding of the practices that staff undertake and the challenges they face” (p. 68). Some colleagues and I have started identifying our practices and challenges by documenting the workarounds we’ve used and developed. Alter (2014) defines workarounds as

a goal-driven adaptation, improvisation, or other change to one or more aspects of an existing work system in order to overcome, bypass, or minimize the impact of obstacles, exceptions, anomalies, mishaps, established practices, management expectations, or structural constraints that are perceived as preventing that work system or its participants from achieving a desired level of efficiency, effectiveness, or other organizational or personal goals (p. 1044)

Workarounds are a useful lens because they highlight areas of disconnect between what is needed and what is provided. Alter (2014) suggests that this Theory of Workarounds could be used to understand these disconnects and leverage that understanding to drive re-design. Resonating with Biggs’ (2001) notion of quality feasibility, a practice that actively seeks to understand what the impediements to quality teaching are and to remove them.

The challenge I faced was whether I could remember a reasonable percentage of the workarounds I’ve used in 20+ years.

Enter the following list of eight Online Course Quality Indicators also available as a PDF download and tested in Joosten, Cusatis & Harness (2019) (HT: @plredmond and OLDaily).My interest here isn’t in the validity/value of this type of approach (of which I have my doubts). Instead, my interest in that the eight indicators offer a prompt for the type of considerations to which a conscientious teacher might pay attention. The type of considerations that will point out limitations within institutional support for (digital) education and generate workarounds.

Initial findings

So far I’ve remembered 53 workarounds. Detail provided below. The following table maps workarounds against the quality indicators. The biggest category is Doesn’t fit. i.e. workarounds that didn’t seem to fit the quality indicators. Perhaps suggesting that the quality indictors were designed to analyse the outcome of teacher work (online course), rather than provide insight into the practices teachers undertake to produce that outcome.

Peer interaction and content interaction are the indicators with the next highest number of workarounds. Though I have collapsed both content interaction and richness indicators into content interaction.

Quality Indicator

# of workarounds

Design

4

Organisation

6

Support

1

Clarity

3

Instructor interaction

7

Peer interaction

10

Content interaction / Richness

9

Doesn’t fit

13

53 is a fair number. But perhaps not surprising given my original discipline is information technology and part of my working life has been spent designing LMS-like functionality.

What’s disappointing is that a number of these workarounds are duplicates solving the same fundamental problem. The only difference being in the institutional and technological context. For example, a number of the workarounds are focused on helping with:

  1. Production and maintenance of well-designed, rich course content.
  2. Increasing the quanity and quality of what teachers know about students background and activity.

What does that say about higher education, digital education, and me?

Proper reflection and analysis will have to wait for another time. But evidence of a difficulties in at least two fundamental practices seems important. Or, perhaps it’s just showing how blinked and obsessive my interest is.

There are some questions about whether the following are actually workarounds. In particular, some of the fairly specific learning activities aren’t actually designed to change an existing part of the institutional context. There was no part of the institutional context that provided for the learning activities. Largely because the learning activities were so specific to the learning intent that the institution would never have been able to provide any support. However, most institutions now have lists of digital tools that have been approved for use in learning and teaching. Typically, the specificity of the learning need means that no appropriate tool has been added to the list.

What does this say about the reusability paradox and institutional approaches to digital education?

Workarounds and quality indicators

The following steps through each of the quality indicators and uses them as an inspiration to answer the above question. For each workaround links to additional detail is provided and initial thoughts on the workaround given.

Design

Systems Emergencies

One attempt at an authentic real world experience was the Systems Emergency assessment item for Systems Administration (Jones, 1995). Each student had to run a program on their computer. A program that would break their computer. Simulating an authentic error. The students had to draw on what they’d learned during the course to disagnose the problem, fix it and complete a report.

Is this a workaround? It’s so specific to a particular course and a particular pedagogical choice there is no institutional system that it is replacing.

Open Learning Computing Platform

Better example from the same course went by the acryonym OLCP (open learning computing platform) (Jones, 1994). The recommended computer systems almost all distance education students were using (Windows 3.1/95) was not up to the requirements of the course (Systems Administration). To workaround this limitation we distributed a version of Linux (Jones, 1996a), eventually relying on commercial distributions. Without Linux the course couldn’t be taught.

Personal Blogs, not ePortfolios

Arguably, my predilection for requiring students to use their choice of public blogging engines, rather than institutional ePotfolio tools was also driven by a desire for authenticity. Not to mention my skepticism about the value of institutional ePortfolio systems (which got me in trouble one time). Initially, individual student blogs were an extension of journaling (introduced in Sys Admin) and an encouragement to engage in open reflection and discussion. Intended to mirror good practice for IT professionals and first used in a Web Engineering course in 2002. Later evolving into the BAM and BIM tools to encourage reflection for assessment purposes and to encourage the development of a professional learning network.

Alignment and curriculum mapping

In terms of alignment of assessments and learning activities I’ve used and more often seen people use bespoke Word documents and spreadsheets to engage in mapping of courses and programs. Mainly because institutions did not have any practice of encouraging such practices, let alone systems to do it (e.g. this from 2009). There’s been a lot more attention paid and importance placed on mapping, but generally it remains an area of bespoke documents and spreadsheets. Perceived shortfalls that led to some design work on alternatives.

Organisation

Moodle Course design

Designing a well-organised course site that is easy to navigate with manageable sections and a logical and consistent form is no easy task given the nature of most LMS. My first foray into this (before 2012 I was using an LMS I developed) added the fikkiwubg design features using bits of HTML.

A “Jump-to: Topic” navigation bar to my Moodle course sites to avoid the scroll of death.

Addition of non-topic based navigation to the top of the Moodle site to provide a sensible grouping of resources (Course background & content) that didn’t fit with the default Moodle design.

Addition of topic-based photos to generate visual interest, perhaps a bit of dual coding with the topic, and encourage some further exploration.

A “Right now” section at the top manually updated each week (along with the banner image) of term to orient students to the current focus.

Moodle Activity Viewer

Since the in-built Moodle reports aren’t that good and because I really wanted to understand how students were engaging with the Moodle sites I designed the Moodle Activity Viewer scratched an itch. It enabled an analysis of student activity.

Evernote to search a course site?

On of the on-going challenges with using Moodle was the absence of a search engine, a fairly widespread and important part of navigating any website. I did consider a number of different options and ended up trying out a kludge with Evernote. But only for one offering.

Modifying the Moodle search book block

Hosting course content on a WordPress blog

In 2012 I took over a Masters course titled Network and Global Learning. Given the focus of the course, hosting the learning environment in a closed LMS site didn’t seem appropriate. Instead, I decided to try it as an open course. It ended up as a WordPress site and has since been taken over by another academic…at least for one offering. Looks like it probably ended up back in the LMS.

Diigo for course revision

Given NGL was hosted on a course blog, this raised questions about how to take notes about what wasn’t working and ponder options for re-design. In Word, this could be done with the comments feature. For the Web I used Diigo to produce annotations like the following.

Card Interface

Late 2018 saw me stepping backwards to Blackboard 9.1. A very flexible system for structuring a site, but incredibly hard to make look good without a lot of knowledge. How to enable lots of people organise their course sites effectively? Enter the Card Interface. Easily convert a standard Blackboard content page into a contemporary, visual user interface.