Google Classroom goes live

When Google first got involved with EdX in September last year, I guessed that a Google-based educational management service couldn’t be far off. Earlier this year, Classroom was announced and the public previews started in July to selected Beta users. Today, Google Classroom is live for all Google Apps for Education domains. Earlier in the week saw a staggered roll-out to selected domains.

Classroom is linked very closely with Google Drive which is used to manage the back-end of the assignment process. For those that have played with “Doctopus” – a plugin for Google Sheets that manages document distribution and privileges within Drive – will appreciate the improved simplicity of this task in Classroom.

Classroom is nowhere near as fully featured as Moodle, and probably intentionally so. I guess that Classroom will see a much quicker uptake in schools, not because it is more useful than Moodle, but because of its tight integration with the rest of the Google services – specifically Drive.

As with any Google service, the goal is to release early, and update frequently. This is also true of Classroom.

During the preview, one of the notable weaknesses of Classroom was the fact that there was no “static” content page. Instead, everything in Classroom had to form part of this timeline of activity and announcements which meant that course materials would drift away and out of the timeline, however, that feature was rolled out earlier this week.

Another frequently requested feature was the ability for teachers to review assignments in progress but which had not yet been submitted – in the aims of monitoring progress and providing formative comments to perhaps guide the pupil in their work.

I’m looking forward to showing our staff what can be done in Classroom and seeing how they make use of it.

Working with Google Apps School Directory Sync

At school we are moving full force into a Google Apps deployment. In order to make it manageable, we have elected to use the Google Apps School Directory Sync app, developed by the same guy as the Google Apps Directory Sync app – and largely to fulfil the same function, but this time using easily exported data from a school management software.

I have written a module into ADAM which updates the CSV files periodically and a scheduled task on the server performs an automatic synchronisation of the data up to Google Apps. This process also creates and suspends users (although it doesn’t have to) leaving us with the most accurate list of accounts possible.

SDS does allow one to create students and staff in containers, but those containers are not customisable. We make use of several settings that apply to specific age groups (e.g. turning Google+ off for the primary school pupils) and so having two uncustomisable containers is not helpful. To that extent, I’ve turned off the containers and have undertaken the task every once in a while to scan the root container for new users and move them into their respective folders. This is a reasonably easy task for us since the usernames start with the year of their matriculation and so finding the containers to put them in is easy.

The system is rather American-centric (which may or may not be fair) in that it assumes Google Apps domains will be based on districts and not individual schools. It also means that the naming of the classes is odd when faced with the South African education paradigm (SDS uses “Course, Period and Section” instead of “Subject, Grade and Class”). So while it is still possible for SDS to automatically create the distribution lists for each class, they seem to appear with strange names. But that is a minor consideration in the long run.

I did need to learn a thing or two about regular expressions and negative forward lookups. The groups that are created by SDS are all prefixed with the school’s name. I want SDS to be able to add and delete those groups to its heart’s content, but leave my other groups alone. This was achieved with an exclusion rule that uses a negative forward lookup regular expression.

As with GADS, exclusion rules are crucial. We have a number of dummy accounts that we use for testing things in Google which, of course, are not part of our “live” data and so we created a container for these users and excluded it from the synchronisation process.

There are fewer options in SDS than there are in the more established (and powerful) GADS, but with more schools moving away from local area networks, there are not always options when it comes to managing user accounts and, more importantly, the groups of pupils. SDS provides a useful solution, but perhaps one with a few shortcomings.

Google Keep

I’ve been looking for a simple note-taking app on Android. There are lots, I’m sure, but the default “SNote” that is included on the Samsung flavoured phones doesn’t meet the grade when it comes to synchronisation and general availability of the notes.

Google Keep allows you to quickly add notes and check lists. It is web based, so available everywhere your browser is, although there is an app for Android (not iOS, but you can make a shortcut on your home screen to the site).

Synchronisation is good and there are a few nifty features including OCR for uploaded images.

I don’t pretend that this is even nearly a replacement for Evernote, but it is a quick and easy tool to use.

Department of Basic Education bans JAVA in favour of Delphi, FOSS in favour of Microsoft

At least they won’t have to worry about Java Update notifications…

It is a dark time in South African education. Already South African features near the bottom of any study featuring Mathematics and Science ability amongst its learners (GITR 2013, World Economic Forum Report 2012, 2011/2012 World Competitiveness Report, PIRLS 2011 to name a few) and yet again we see the Department of Education making poor choices based on bad information. This time, however, it is towards the teaching of Information Technology (IT) and Computer Applications Technology (CAT) in the state education system.

Circular S9/2013 (DBE website is down at the time of writing:’s reaction also includes a link to the circular) was recently released stating that for the IT examination, the only permissible programming language was Delphi (a number of provinces have standardised on Java) and that the office suite required for the CAT examination was MS Office 2010 or MS Office 2013.

Computer Applications Technology

Let’s address the CAT issue first, because it is so preposterous, it’s comical.

In order to run MS Office 2010 or 2013, the minimum specifications for a computer are more than twice the processing power, RAM and storage space required for Office 2003 (see Microsoft’s Minimum Specifications for Office). While they don’t look challenging on the face of it (any computer released after 2003 should be able to cope), these are very real challenges to many state school’s IT infrastructure.

In addition to that, licensing a workstation to run Windows and a version of Office will cost money that many schools can ill-afford. While one might imagine that a once-off licence might be able to last a while, it should be noted that all those computers with Office 2003 (possibly the most successful Office yet?) cannot read native data files created in Office 2007 and up (the XML file types). This necessitates potential software and hardware upgrades for schools that can ill-afford to do so.

Given that the DBE is only supporting the last two versions of Office, how much longer will Office 2010 be supported for? Microsoft are rubbing their hands together in glee!

This decision also runs contrary to government’s Free and Open Source Policy (PDF download, or cached copy from Google) which was published in May 2006 in that this encourages vendor lock-in and perpetual licensing costs.

Information Technology

Some history: Computer Studies started in the late 1970s with a handful of schools offering the subject. Over time, the number of students taking the subject grew. The generally accepted teaching language of the day was Pascal, developed by Niklaus Wirth in the late 1960s specifically for education to teach structured programming and data structuring (see Wikipedia).

Delphi is syntactically equivalent to Pascal, but adds in Graphical User Interfaces (i.e. windows) and allows the rapid design of interfaces using drag and drop techniques. It should be said that Delphi also includes object orientation and so while is functionally equivalent to Java and many other programming languages, it remains as one of the less frequently used languages. The TIOBE Software Index ranks Delphi 14th in September 2013, having dropped two positions since this time last year.

Java originates from C, syntactically, and was designed – from that point of view, at least – to be a language better suited to education than, perhaps C or C++ were. Java was purely object oriented, and contained far fewer “low level issues” for programmers to worry about (technically speaking, I’m referring to pointers, garbage collection and other such issues that people love/hate C and C++ for). Java also has the advantage of being a universal language in that it will run on (nearly) any computer, be it Windows, Mac, Linux or anything in between. TIOBE ranks it 2nd, after C.

A language popularity website calculated that Delphi is a significantly less popular language than Java.

I suppose that while a language’s popularity is important, its educational roots and benefits must be looked at with more vigour. With Object Oriented Programming recognised as an industry leading paradigm (embodied by languages such as Java, C++, C#, Objective-C and many others), Java lends itself to this way of thinking much more easily than Delphi. In addition, Java is syntactically similar to C, C++, PHP and a number of other languages which potentially means quicker learning for graduates.

The argument for Delphi to be chosen as a single language for all schools is, therefore, flawed.

When the National Curriculum Statements (NCS) were released, it was said that the programming language could be anything that was “object orientated and netcentric”. While Delphi can do both, it was designed as neither. With the release of the CAPS curriculum, the programming language was changed to a “RAD-supported language” and, in its reading, was to be largely database centric. Thus the idea of teaching towards OOP was abandoned.

The IEB, thanks to a standardisation of the syllabus and that they were able to understand that the language is irrelevant to the concepts being taught, has been able to set a language-neutral paper for the last five years. There were certainly teething problems as the Delphi crowd slowly got used to vendor-neutral terminology already inherent in Java thanks to its multilingual origins. Thanks to successful training, have done remarkably well.

The move away from OOP is simply because key decision makers do not have any real-world programming experience and cannot therefore understand the importance and relevance of teaching computational thinking with Object Orientation as a logical starting point. Instead, attempts to introduce OOP in the syllabus were backwards – as an afterthought.

Certainly there are (good) reasons for standardisation on language: both teachers and pupils move between schools and a pupil who moves and has to change programming language because of that is disadvantaged (anecdotally they do not perform as well as their peers, but this is not well studied). Pupils are also disadvantaged if they get a teacher who is not proficient in the language.

Up to now, provinces have been able to choose which language they will standardise on thus avoiding most of the problems of migration. Inter-provincial migration is not common (anecdotal reports suggest). But now at least two provinces, the Western Cape included, will have to spend time and effort re-training teachers to use an obscure and obsolete programming language.

Sadly, our pupils are being taught by professionally trained teachers and not, in most instances, by competent computer scientists or real-world software developers. Thus this language switch is not just inconvenient – it is likely to be disastrous.

While I’m extremely disappointed by this decision, I do know that the idea of computational thinking has been lost a long time ago in this mire of politicking and point scoring. Computational thinking skills are needed to save a struggling software industry from being outsourced to countries like India. Countries like the UK have realised this and are now introducing computational thinking skills as early as primary school and have re-introduced a programming course at secondary (GCSE) level.

Other reactions

Samsung ATIV

A few months ago I attended a launch event hosted by Samsung for, amongst other things, their ATIV tablet range. I was reasonably impressed at the Windows 8 device then, but thought that it might be cumbersome to use: it is a heavy device.

I have been using one as a school work device for a month now and I must admit that the device is becoming a firm favourite. While Windows 8 desktop is a pain to use with fingers and the built-in track pad is, in a word, awful, the responsiveness and precision of the stylus is great.

I am now using it almost exclusively, with TeamViewer, as my classroom presentation tool of choice. That it is wireless means that the tablet can be passed around the class and pupils can also interact and participate in the learning that is going on in the classroom.

I must admit that I haven’t moved much past the “electronic whiteboard” functionality, but certainly having the power of a truly multitasking operating system puts my iPad at a distinct disadvantage. In addition, even with a stylus, the iPad is a poor substitute for handwritten notes.

Issues that I do have with the tablet include physical stability. Twice I’ve put down the tablet with a bit of a bump, and twice it’s given me the Windows 8 sad-face-blue-screen. I reckon there might be something loose inside? The stylus, pressure sensitive and all, has calibration issues at the screen’s edge, and while it’s usable, it would certainly benefit from some sensors in the bezel, just to help with the overall accuracy.

Decent inking software would also be nice. I’m relying on Word’s inking tools simply because they can be saved easily and distributed easily. However, they are huge. And Word starts getting unresponsive and touchy after a reasonably full page of inking. Even PDF-ing the resulting ink drops the size to between 25 and 30% of the original, but when a single page is coming in at 4MB, one has to wonder if it is really worth it. The iPad has some great software for this. It is probably out there for Windows too.

With the Pro version, the camera is limited which means that it makes a reasonably poor document scanner – one thing that I really like about my iPad (even iPhone). Again, this might boil down to software at some level, but the photos are not great – grainy and blurred in “indoor” light.

At any rate, I think Samsung have ticked a lot of boxes with this device. Their next steps are hardware refinement and costing.