Archives

Weeknotes 18 March 2018

It took me a little while to get back into the swing of normal work after my trip to Washington. Not a huge amount got done in the first week back. This week just gone was better. I think that was partly jetlag, partly a cold weather winter slump (it hasn’t helped that the heating at home is broken). I’ve also been following my many former colleagues who are on strike in an effort to maintain their pensions. I wish them all the best.

I did some reviewing for PressEd, “A WordPress and Education, Pedagogy and Research Conference on Twitter”. I really appreciated the option for “include if space”, even though I knew there wouldn’t be space, because there were a huge number of really interesting presentations offered and suggesting that some were rejected due to lack of quality would not have done them justice.PressEd. March 29th on Twitter. pressedconf.org #pressedconf18. PessEd is happening on Twitter, on March 28th, follow @PressEdConf for more details. I know it will be good.

I spoke with Doug Belshaw about Project MoodleNet. There’s a lot in Doug’s thinking that aligns really well with where I would like to go next with ideas that came from projects around UKOER, Core Material, the Learning Registry, kritikos. It’s about discovering, curating and sharing resources and ideas, but through social interactions, not repositories.

EOCred Schema, the project for describing  Educational and Occupational Credentials in schema.org continues to progress. Soon I think we will be at the point where we can plan the mopping up of what it left, that is we’re nearing the beginning of the end. I expected this phase to last until about Easter, so that’s on target.

I finally fixed a problem with deja-dup / duplicity backup which had been failing with an error message about a GPGError. Turns out it was due to some corrupt, zero-length gpg files in the backup destination folder. Deleted these and backups started working again.

I’ve been learning how to do option settings for WordPress plugins so that my wikidata-powered custom taxonomies can be more flexible / less hard coded.

Cetis has a new project starting on the digital campus of the future; and it looks like we might actually get our new website going next week.

Weeknotes 1 March 2018

I am writing these notes in an airport hotel at Heathrow when I had planned to be back home in Edinburgh from Washington DC via Dublin. It’s nearly two weeks since my last weeknotes, here is what  has been happening.

For my CMALT I got some really useful advice and feedback from a tweet asking for comments.

 So, my advice to other CMALTers, ask for help as widely as you are comfortable with,  provide something for people to comment on, ask early, ask often. It looks like it will now be next week before I can finish and resubmit.

I blogged about PressBooks and ePub as an OER format and also about Using wikidata for linked data WordPress indexes. I went to wikidata meeting at Edinburgh Uni, run by Ewan McAndrew, and after which we met up with some of the other lovely Open Edinburgh / wikimedian people. It turned out useful to have all this WordPress, OER and wikidata stuff at the front of my mind when meeting people in DC.

For the work on educational and occupational credentials in schema.org, I think we settled on how to link the course descriptions in schema.org to the credential one would earn at the end of it. But the real highlight of this period was my visit to Washington for a meeting hosted by the Credential Engine, which is where this work started. In short the Credential Engine is based around open and transparent information about qualifications, degrees, and other such credentials, closely tied to information about who offers them and what learning opportunities are available that lead to them. This in a way that can be linked to a wider graph relating credentials to workforce development, personal career planning, recruitment, employability analysis, and so on. You can find out more about the credential engine on their website, or if you prefer concrete demonstrations of what they do, look at the credential finder, or you might want to look at the technologies used, especially the Credential Transparency Description Language and the Credential Engine Registry. I was lucky enough to be invited to a meeting with the team running the Credential Engine, and some of those leading the technical effort and strategic thinking, to come up with ideas and priorities for the next phases of its development.

Lincoln Memorial (with human figure for scale only) taken by Jeanne Kitchens .

The conversations were all stimulating. They are already sparking ideas, and I’m sure will continue to so. The hospitality was also great, thanks especially to Jeanne for the orientation tour of the mall area. The timing was also notable, it was glorious spring weather; somehow the world is small enough that I bumped into Andy Powell on a similar flying visit; but less serendipitously, the #BeastFromTheEast and #StormEmma conspired against my homeward travel plans.

Weeknotes 18 February 2018

A pretty straightforward week. Looking back to my aims from the end of last week (x=done, s=started, p=in progress)

-[x] get a public demo of the wikidata taxonomy plugin up and write a blog post about it
-[s] figure out what to do about parts of the old cetis website that are not used and not maintained but shouldn't be deleted
-[p] conclude discussions on a couple of EOCred properties and start discussions about links to competencies
-[p] draft improvement of my CMALT & and ask for comments

I actually did the wikidata WordPress taxonomy plugin work exactly as planned: blog post here. Happy about that and the responses to it. I also learnt this morning that there is a similar plugin already existing, but have yet had time to look at that. I missed it before because I had mostly been looking in the WordPress plugin directory for wikidata, and it’s not there yet.

For the CMALT, I actually asked for comments first and then started on drafting improvements. Good reaction to that post as well, I had a suspicion that seeing and example of a CMALT portfolio that wasn’t quite good enough, with the feedback from the assessors would be as interesting as all the great CMALT portfolios that people who have passed have posted. Now hoping that I can back it up by showing the remediation based on the feedback.

Concluding discussions on the EOCred properties turned out harder than planned. Describing costs now seems to be done, but discussion is still going on about what we mean by the educational level of a credential. I think that’s near done, but not quite as I have hoped.  This ongoing discussion meant that we didn’t really get started on the proposal for how to show a link from a credential to learning opportunities that may lead to it.

We also had an LRMI call. The LRMI website will be relaunched soon, and with it some work that we essentially finished about a year ago, but which got caught up in various logistic issues around publishing it. We are beginning to talk about how to configure LRMI in order for it to do more work. Learnt about IEEE’s ICICLE, and SIGs that look relevant to various bits of Cetis work.

On Friday I made a start on looking at those bits of the Cetis website that we want to archive. Reminded  myself how I had used wget to get copies of all the files from some of my old blogs and wrote some python scripts to mod the html where necessary. If you’re interested, it starts with

$ wget --recursive --no-clobber --page-requisites --html-extension --convert-links --domains example.org.uk --no-parent example.org.uk/blog/

Weeknotes 11 February 2018

This week I spent a fair amount of time working on the wikidata taxonomy plugin for WordPress. I have improved the interface for the wikidata fetching part so that most of the parameters needed to get the data are passed to the methods rather than held as properties. I have also improved the SPARQL query so that rather just finding the immediate wikidata class for the item in question it will find find all super classes. This means that it works for subclasses as well as instances, and makes it easier to match items to their schema.org class (in that I only need to deal with the higher levels of the wikidata taxonomy). Together the improved interface and easier type matching make it easier to deal with new schema.org types in the wordpress taxonomy and to get suitable properties of them. The down side is that at some point I introduced a bug that affects adding and deleting terms without giving any meaningful error message. In related news the error trapping and reporting needs improving sometime.

The other wikidata-related bit of fun was on the anniversary of some women getting the vote for UK parliamentary elections, creating searches for women UK MPs who were descended from or ancestors of other MPs:

In the course of wikidataing, I used this Bob DuCharme web blog on The Wikidata data model and your SPARQL queries and book Learning SPARQL to learn a bit more about property relationships in SPARQL query triples. But probably the biggest help has been seeing wikidata queries shared by Ewan McAndrew and  Andrew Gray.

Another highlight of the week was a twitter chat for the #PressedConf18, which will be a twitter conference about WordPress in Education. Pgogy (aka Pat Lockley) and  Natalie Lafferty are doing great work in getting community interest going in that.

On Friday I went out to Heriot-Watt to meet friends. Nice to see progress being made there on supporting the Gradate Level Apprenticeships (Lisa will be putting in a presentation to #PressedConf18 about how they use WordPress as an open repository and, with H5P, development environment for the online resources used in their blended learning courses.) Also nice to talk to Tessa who will be teaching the Design for Online Learning course that I used to teach.

Got my CMALT appraisal back. The aspect that I thought weakest wasn’t quite strong enough. No matter, it has given me thoughts on being a bit more open in CMALT that just making a positive result available. More on that next week.

Made progress with the Educational Occupational Credentials in schema.org community group, discussing a couple more properties.

Cetis got some new funding confirmed for a project that I hope I will have some involvement in.

Plans for next week:

-[ ] get a public demo of the wikidata taxonomy plugin up and write a blog post about it
-[ ] figure out what to do about parts of the old cetis website that are not used and not maintained but shouldn't be deleted
-[ ] conclude discussions on a couple of EOCred properties and start discussions about links to competencies
-[ ] draft improvement of my CMALT & and ask for comments

 

Weeknotes 3 Feb 2018

If I am not careful these will become month notes. Anyway, the last three weeks I have been working mostly on two projects.

First, The work with the W3C Community Group on educational and occupational credentials in schema.org is going well. There was a Credential Engine working group call last week where I summarised progress so far.  I learnt never to describe meeting a requirement from a use case as ‘quick and easy’. Turns out that ‘wanting to know what something will cost’ depends a lot on what you read into ‘something’. Also, I was taken with a point made by Fritz Ray in a discussion about a property initially called credentialType, which became credentialCategory because, as he said, ‘credentialCategory is disarming to presumptuous software folk like me who “know what type and class means” and have a tendency not to read descriptions’.

Second, I got a nice tweet about something I did last year about getting wikidata into WordPress custom taxonomy.

This prompted me to go back and make a start on doing this properly. So far, I have a plugin built on two classes. One class will get data from data from Wikidata via the SPARQL endpoint and store it into taxonomy term metadata. The other class will create a custom taxonomy which can be used to tag posts with a schema.org relationship (currently hardwired as about, but could just as easily be mentions, or learningResourceType, or educationalAlignment). Unlike my previous effort at semantic WordPress, this works as a plugin alone, it doesn’t require a matching theme (though it can be improved by one). The minimal functionality is nearly ready for testing. I would like to go back over it and sort out the interfaces for the classes, make it a lot more fault tolerant, and generally improve things like security and internationalization to be closer to WordPress coding standards.

Omniana transcription on wiksource continues, as does putting content into the Pressbooks instance. That latter might soon be getting some wikidata-based semantic indexes 😉

We have also been working on updating the Cetis website. I think we are making progress.

Weeknotes 15 January 2018

It was a slow start to 2018, mostly because of the tail end of flu that I had over the holidays, but I did get a decent amount of work done on Educational and Occupational Credentials. We now have a prioritised list of use cases, from which requirements have been derived. The first requirement is to identify an object as being an Educational / Occupational Credential, and we had a week or so of discussion about that around a proposed resolution. To my mind it is still in the balance as to whether this type of object is going to end up under schema.org/CreativeWork and whether we shall create a more generic supertype of Credential.  What I think is not obvious to the general schema.org community is the practice in education established around open badges Badge Class and Assertion distinction. I can see why it’s not intuitive to have Credentials as CreativeWorks; as CreativeWorks go, Credentials are kind of abstract. I suspect that the CreativeWork type is somewhat more generic than was originally envisaged: Visual / literary work might have been a better name if that is what the creators of schema intended. Having Intangible as a distinct class from CreativeWork doesn’t really work for anyone who has met FRBR.

Highlight of the month so far has been being invited to a Credential Engine meeting in Washington DC at the end of February. So I have also been sorting out travel arrangements.

Also, made a fair amount of progress with Omniana transcription on wiksource, and putting content into the Pressbooks instance.

Weeknotes December 2017

OK, I slipped at the end of 2017 and stopped weeknoting. Just for the record, in December 2017 I mostly worked on a Cetis project for a client using SSMM (software system maturity model) to investigate potential solutions for MDM (master data management). It was pretty intense, the SSMM requires quite a lot of information that, at least for this category of software, wasn’t always easily findable. It was my first time using SSMM. I found it thorough. By design it is a breadth-first approach to finding software for a particular task, narrowing down a long list of potential solutions based on criteria such as sustainable development model, fit to business requirements and so on. The MDM side was also interesting, getting a better feel for who the commercial enterprise world approaches problems which in the library world are addressed by authority files and shared cataloguing principles.

I also worked on the Educational and Occupational Credentials W3C working group. We had a kick-off conference call just after the last weeknotes I did, and by the end of the December we had a list of use cases which were being prioritised.

Continued transcribing Omniana at Wikisource. Continued looking into PressBooks, including having a go at importing EPUB files, something which worked well and deserves its own blog post.

Weeknotes 25 Nov 2017

Highlight of the week has been starting my first Cetis project as a partner. Scott, Simon and I are working on a data orchestration project for a client who wants to consolidate some of the information they hold. We will be using SSMM, the Software Sustainability Maturity Model to help find a solution. SSMM came from OSSWatch and Cetis has used to evaluate options for several other clients (and not just for open source options).  It has been interesting talking to Scott and Simon and seeing comparisons between library world “authority control” approaches and business world “master data management”.

Lowlight of the week has been a plague of petty computer problems. I am finding Ubuntu 17.10’s gnome desktop environment to be quite unstable. It freezes once or twice a day. I am having problems with sound device management with devices intermittently not being recognised.  Add modem crashes, batteries dying to that, and speaker cables coming loose, all in one day, and you can get frustrated.

Back to the plusses: I started tidying up the work I have been doing on pulling information in from wikidata to provide for linked data custom taxonomies in WordPress, putting it all together into a plug-in. I realised writing last week’s weeknotes that I was being a bit brain dead in pulling in all the wikidata about an entity via the raw JSON API, so I fixed that and am now sending SPARQL queries for just the bits I want.

I also spent a little time thinking about possible features for the “rockhopper” project Pat Lockley and I chatted about last week. It’ll be a great way of storing things, the greatest. It’ll rock.

Weeknotes 19 Nov 2017

Two weeks’ worth, but it’s all slow and incremental progress, so it doesn’t seem like much.

Spent a fair amount of time on the border between WordPress and Wikidata.  So now, for any person mentioned in Omniana, If I know the Wikidata ID, I can pull in relevant data from the Wikidata API, store it in a WordPress custom taxonomy, publish it as a index for a Pressbooks book in such a way that it can be picked up by linked data applications. The wikidata data model is kind of complex, it allows for several claims to be made about each property, along with qualifications of those claims. I just make a simple call for all the data about a particular entity, which leaves me needing to pick out a few values from deep in what can be a fairly large JSON object. I am wondering about whether it might be better to make more specific queries, perhaps via the wikidata SPARQL endpoint. The code button on the wikidata query editor even generates the PHP required to make a request.

I’ve fixed a date for the kick-off meeting for the educational and occupational credentials in schema.org working group. Also sent my first invoice for the work done so far on that project.

Also on the educational metadata front we had the monthly LRMI call. Mostly we talked about changes at DCMI and the implications for resourcing LRMI. I would really like to get on with wrapping up some of the technical work that is pretty much finished, and making sure the world knows that LRMI is active. But that very quickly leads to questions of how to publish that technical work, i.e. resources that are provided by DCMI.

Went to the Urban Big Data Centre Active Travel Data Challenge Day. Lots of tweets.

Talked to Pat Lockley (pgogy) about a possible joint venture, code name rockhopper.

Went to the monthly Edinburgh WordPress meetup. Good presentation about progressive enhancement / advancement (which I guess is a more positive slant that graceful degradation).

Weeknotes 5 Nov 2017

Some weeks are more boring to reflect on than others. This week was largely tying up loose ends from the progress made last week.

I wrote a blog post about kicking off the project on putting educational and occupational credentials into schema.org and I set up the W3C EOCred CG. There were some glitches in the W3C set up at first, but now it seems all in order, and has 15 members. Yay! If you would like to contribute to the work of the group (or even just lurk and follow what we do) please join it.  I also had a useful catch call with the people at the Credential Engine and SIUC who are paying for this work.

Most of the rest of the week was spent on a theme for the PressBooks version of Omniana. I have set up custom taxonomies that will allow me to index  the people, places, events and works mentioned in the articles. I quickly found out that sub themes for PressBooks are not straightforward. When did CSS get so complicated? Anyway, I have the custom taxonomies working now, and I am well on the way to finding them a place to display them. The long term aim is to explore linking the content to wikisource and the indexes to wikidata; using the indexes to build a graph of what is mentioned in the book.

I went to a presentation from Companies House on the duties of being a director.  I made a minimal website for PJJK Limited, in case people go looking for it.

[The cover image is based on Okkulte Erlebnisse 1924.jpg By H.-P.Haack (eigenes Foto, Antiquariat Dr. Haack Leipzig.) [CC BY 3.0], via Wikimedia Commons, adapted by the publisher and licensed CC:BY]