RISC OS in schools

Are there any RISC OS machines used now in schools? ie 2020.

I believe Tom Williamson went into schools (before lockdown) to promote RISC OS and programming.

He had mixed results – a lot of kids engaged with BASIC, the speed, and ease of RISC OS. However, lack of key elements (mature Python, Scratch) make it unsuited to curriculum coding education etc.

He did make it abundantly clear that one area RISC OS could work would be in a “minimal RISC OS + Python” setup, with fast-as-possible boot. Like Pico, but better, basically. It would run headlessly, but with an optional desktop, and allow comprehensive gpio control for robots etc. In this way, it could be much superior to Raspbian due to fast boot on Pi-Zero etc.

It’s not the first time that this kind of area (fast performance on minimal devices) has been identified as a possibility for RISC OS, but it isn’t an area that would really appeal to the more traditional desktop RISC OS user.

We all know that the histories of Acorn and educational computing are deeply entwined. When Acorn triumphed with its BBC B microcomputer hardware was expensive; now it is as cheap as biros and cellotape, though I suspect that educational administrators have not adjusted to this yet. Indeed, that lag has been an opportunity for many a business to exploit. The real costs come from software development, and the fixation that kids need special software anyway. My six-year-old grandson is using Scratch. Looking further along the educational sausage-machine, academics have been wrestling with the question how much computer science should a student know? for the last half century, without coming to any useful answers. If they have, please let me know :).

Some educationists have a pretty good handle on the issues; some academics have. They’re often wrestling not so much with the questions as with some of the other educationists and academics, and even more with the politicians. Something I spent quite a few years in…see http://clive.semmens.org.uk/Education/CheerioClive.html – note especially the PostScript…

and the fixation that kids need special software anyway

I can understand the requirement for specific software in order to make the curriculum consistent.
However, it seems all too often that schools get stuck teaching how to use a software package instead of teaching principles, using this software.

There’s a subtle difference that was very nicely highlighted a few years back when the company I work for had a brief foray into Linux.
Those people who learned the principles of things like spreadsheets and word processors switched fairly effortlessly.
Meanwhile, those who didn’t really understand the principles and were instead taught how to use Microsoft Office were completely lost as things weren’t in the same places and didn’t behave in quite the same way.

The latter lot, I should add, were lost all over again when the company ditched Linux in favour of Windows and shortly after upgraded Office to a version with a different UI (probably the much maligned ribbon, but I wasn’t paying much attention as it isn’t my job).
Oh, and they got lost all over again when the company bought a number of Samsung tablets to replace the clipboards. They run Microsoft’s stuff, but it doesn’t look like it does on the big machine.

I wonder how many hours of productivity were lost because these people were taught a specific (at the time of their education) version of Office (or if older, maybe Word and Excel) rather than “how spreadsheets work (with Office/Excel as an example)”.

A rather nice suite that seems to have been lost to time is Microsoft Works. It was like a mini version of Office that did the sort of things that one might want in day to day use, without the complications of the fully blown packages.
In RISC OS terms, the same question arises – why was there a need for Easiwriter on a platform with Ovation and Impression? The answer is simple. There’s a market for a word processor that is better than Edit (!) but that isn’t a full DTP package.

It seems all too often that schools get stuck teaching how to use a software package instead of teaching principles, using this software.

Amen to that!

On the other hand, I’m not that bothered about the curriculum being consistent. I’d rather see far more variety in syllabuses, curricula and education in general. So it makes examiners’ lives hard. Well, tough. There are more important issues here than making life easy for examiners.

However, lack of key elements (mature Python, Scratch) make it unsuited to curriculum coding education etc.

MUG did some work with Tim Rowledge back in 2017/18 and even funded an updated DDE for Tim courtesy of one of its members to get Squeak
running which it does up to version 4.4 even on a high vector ROMs now.

You can even download the 4.6 updates to get it working on a slightly later version.

Sadly Scratch is still at version 1.4 and that as Andrew says hinders its use.

Hopefully Python is getting in to the modern era but it would be nice to get a 5.3 version of Squeak and a more modern version of Scratch to go along with it.

In RISC OS terms, the same question arises – why was there a need for Easiwriter on a platform with Ovation and Impression? The answer is simple. There’s a market for a word processor that is better than Edit (!) but that isn’t a full DTP package.

Your answer may be simple, but to me it sounds just wrong. It seems to suggest that EasiWriter/TechWriter were “simpler” than a “full DTP package”, which is far from the truth. EasiWriter and TechWriter have so many features that neither Ovation (Pro) nor Impression (II/Style/Publisher(Plus)) had. They are just different beyond a basic feature set like “hey, I can make my text have different fonts and styles”. Writing a clearly structured document is different to layouting pages.

The approach of a traditional “word processor” is so different from a “DTP package” that it is clear why there is a need for both. Of course today, we have many hybrid approaches (look at the latest Xara offering, you’ll find that they have included basically all Impression functionality into Artworks, simply speaking), so formerly clearly visible boundaries have become completely blurred. See “frames in MS Word” as an example.

As someone who (tried to) use(d) Impression Style for university documents, but switched to LaTeX for the larger ones (because I couldn’t afford TechWriter), I feel qualified to make that distinction.

Impression Style certainly wouldn’t cut it, but Impression Publisher Plus was perfectly capable of producing university documents to Cambridge University Press standards – we did so, for years. (Probably still is, but I’m not in that business any longer.)

@ Rick:

However, it seems all too often that schools get stuck teaching how to use a software package instead of teaching principles, using this software.

Amen to that (to quote Clive!) and adding: and not just here in the UK unfortunately :(

@ Andrew:

It’s not the first time that this kind of area (fast performance on minimal devices) has been identified as a possibility for RISC OS, but it isn’t an area that would really appeal to the more traditional desktop RISC OS user.

Sure, but the IoT world can be a potential big market for RISC OS. Way bigger than desktop market can ever be these days, because even if by miracle RISC OS gets rewritten from scratch to be the best desktop ever, the end user market nowadays is more on mobile devices: tablets, phone and handled devices. That means, in best case, RISC OS Desktop would be a laptop OS that has to compete with macOS and Windows (both being ported to ARM laptops already).

Also, if business comes back to RISC OS then also developers will too. With that more free maintenance for the Kernel, Network Stack, WiFi, Bluetooth, Security, 64bits etc. Money for ROOL that seems interested into making consultancy and also for RISC OS Developments.

Let’s not forget that ARM DevStudio license costs a fortune per year, so a lot of IoT companies (especially start-ups) are forced to stick to Keil. ROOL DDE is good enough for IoT (it only requires more libs given the fast production timing in certain markets, but this is achievable through time, so not a show stop) and a license for let’s say 50GBP per year is super-competitive.

In theory this also could mean that the community here could focus on the Desktop and/or porting apps across the releases, which is less effort than now.

Anyway, just my 0.5c.

Your answer may be simple, but to me it sounds just wrong.

It is wrong, for all the reasons you outline. EasiWriter always felt like a “WYSIWYG LaTeX” in use, which is far from “simple”. It is for creating structured documents, where the formatting comes from the content.

Impression Style certainly wouldn’t cut it, but Impression Publisher Plus was perfectly capable of producing university documents to Cambridge University Press standards – we did so, for years.

Publisher was certainly good enough to create my Masters Dissertation Report for Edinburgh back in 1999, and the output from the Level 2 printer driver was good enough for the department’s expensive-looking laser printers to handle.

The corresponding Masters Project Report was done in Word 97 (or 98), because that way I could write it at my sponsor company and not at home in the evenings. Given that Word is a structured WP (like Easi or TechWriter), it probably did a better job than Impression because it could understand the document and apply things like keep-together. The project had far more tables and figures than the dissertation.

The layout of the two finished documents is indistinguishable, however, and both fully complied with the extremely strict formatting requirements of the department.

On the other hand, I’m not that bothered about the curriculum being consistent. I’d rather see far more variety in syllabuses, curricula and education in general. So it makes examiners’ lives hard. Well, tough. There are more important issues here than making life easy for examiners.

In my experience the problem was political. After the takeover of education by administrators, the curriculum, examinations, even the software was determined not by educational grounds (let alone by the needs of students) but administrative. Teach how to use this or that software package, and the information is out of date by graduation day, as others have noted. The trouble is that those running our institutions are themselves profoundly uneducated, uncultured people. I recommend the works of Dijkstra on the triangle: education, business, government. They are still relevant.

Yeah – structure wasn’t Publisher’s forte. Layout, brilliant; complicated math, chemistry and biology formulae, text and diagrams (imported from Draw or wherever) perfectly okay. Huge quantities of tables and figures, no problem. But you had to look after structure yourself – which was hardest for the Proceedings of the Physiological Society, because all the others had “papers” which occupied an integer number of pages and so could be treated as separate files; Proceedings could start and finish partway down a column, never mind at the top or bottom of columns or the top or bottom of pages. So very often the whole publication – 100 sides of A4 or more – was a single document. With dozens of tables, figure and in-text formulae – stuff like this: http://clive.semmens.org.uk/Photos/Assortment/JPhysiol4.jpg

Sure, but the IoT world can be a potential big market for RISC OS.

Not really no;

RISC OS doesn’t run on microcontrollers
RISC OS doesn’t have wifi (yet)

Let’s not forget that ARM DevStudio license costs a fortune per year

No, ARMs dev systems for IoT are free, download them today if you’d like;

https://developer.arm.com/tools-and-software/open-source-software/developer-tools

a more modern version of Scratch

So Scratch 3 uses HTML5/Javascript for the on line version so I thought I would try !Iris but that doesn’t work but you never know.

They seem to have also dropped Linux support for the offline version as well so it is only Windows/MACOS.

When I started teaching (1963) a colleague devised a special double-width type-writer to accomodate mathematical symbols that would otherwise have been put in by hand. Nearly twenty years later the department bought an Atari so that they could use Signum, a program written by a German mathematician out of desperation at the lack of word-processing facilities for mathematics. Donald Knuth had started writing his TeX document typesetting system already in 1977, but it took him eight years to complete it. It only ran on work stations that were far too expensive for the average UK mathematics department. TeX, and then, in 1985, LaTeX, became the almost universally adopted means of producing scientific documents. A postgraduate’s main transferrable skill was picking up how to use LaTeX to write her thesis. The Euromath project, which spent millions of euros, was overtaken by events on the ground before it even started, though some of its ideas on SGML may have been taken up later by the WWW consortium. The font technology for TeX started out pretty crudely – bitmaps at different sizes took up huge amounts of memory – the RISC OS system was much better. Things have changed, of course.
But TeX’s really big idea, to keep structure separate from content, is still what makes it practical.

Impression Style certainly wouldn’t cut it, but Impression Publisher Plus was perfectly capable of producing university documents to Cambridge University Press standards – we did so, for years.

You could also do such documents in !Draw, but this misses the point.

There is really very little difference between Style and Publisher Plus – mainly the CMYK separation stuff, irregular shaped frames and smart quotes support. As well as some finer points regarding kerning. So I don’t really understand why Style “certainly” would have failed where Publisher Plus was OK.

Other than that, the same weaknesses regarding proper document structuring applied to Publisher Plus as well as to Style – no sensible automatic support for TOC creation, index creation, citations, footnotes, endnotes…even simple things like nested numerical lists were sometimes fiddly. And table handling via the bundled version of TableMate were disgusting, as well as Thesaurus features via WordWorks. And not to forget Equasor, which was really horrible even for simple equations.

some finer points regarding kerning

Rather vital for doing the kind of thing required for academic journal publishing.

I wrote our own table handling app. The beauty of RISCOS was – and still is – how simple it is to write one’s own little apps. Thesaurus? Who the hell needs a thesaurus for publishing scientific journals? We didn’t use Equasor either, again, I wrote an app that we used for more complicated formulae and equations than you can imagine, which produced DDF (Document Description Format) files that imported straight into Publisher (but which Style wouldn’t have understood at all – all that kerning stuff…)

I actually did this stuff for years, running an office where we produced tens of thousands of pages of academic papers. I know of what I speak…I wish you luck trying to do it using Style.

Sure, autogeneration of TOC, auto footnotes and endnotes would have been nice, and no, we didn’t have it and had to do all that structure stuff ourselves. I didn’t have any software to do anything like that until I was doing Tech Pubs at ARM, where we used FrameMaker on PCs.

As for

You could also do such documents in !Draw

The point of the operation was that we were able to run the whole operation cheaper than paying CUP to do it, while matching the quality so well CUP themselves couldn’t tell which pages we’d done and which had been done by their colleagues on Macs in the first few issues after we started, when we were only doing a few sections of each issue. Productivity using Publisher Plus (with a handful of homespun apps to assist) was very good; using !Draw we’d still be working on the first issue of J Physiol…oh, we did use !Draw for diagrams, of course.

However, it seems all too often that schools get stuck teaching how to use a software package instead of teaching principles, using this software.

I remember in one of my classes, we only had a handful of people (8?) and the instructor divided us into two groups. Each group was given three PCs with nothing installed. One group was then given Windows client and server install CDs, and instructions detailing what to click on. The other group was given Linux install CDs, and Linux-style documentation (which tended more towards principles).

Each group was then told to set up a server with x, and to connect the other two clients to it. The Linux group succeeded a couple of hours before the Windows group did.

and not just here in the UK unfortunately :(

I live and work in France… :-)

I live and work in France… :-)

Nice! been there in the past for some projects and customers, love Paris, Bone soiree Rick then :)

The one thing that surprises me about Easi/Techwriter is that it doesn’t have orphan/widow control, although the headings do appear to keep with next even though you can’t actually set it as a property. It is very LaTeX-like, although—on other platforms—there are tools which are more like LaTeX, notably LyX—which requires a LaTeX installation to work.

When I was working in academia (1992–2014) LaTeX was the way to produce academic papers and the like. Most conferences/journals would supply a LaTeX style file and a Word template—.the latter for Mac users only (although LaTeX is available for Macs and has seen a revival there since it gained the ability to output PDF directly).

TechWriter does make it easier to align a set of equations than anything else I’ve tried.

I loved getting LaTeX files at the Journal of Physiology – beautifully structured files. The conversion to Impression Document Description Format was the easiest section of our import app to write, and produced the cleanest Impression documents. Fab.

One of the advantages of LaTeX was that there were many mathematicians contributing new libraries for it for every kind of speciality. This was particularly the case for commutative diagrams, exact sequences and the like. It is possible to typeset these with TechWriter, but it is a chore, because TechWriter does not offer the abstraction facilities of LaTeX.