IT in Education

A theme of mine for a decade has been using IT more fully in education to make education more effective and efficient. Big points in the list include:

  • using FLOSS and GNU/Linux on desktops,
  • using thin clients,
  • using servers,
  • in-school databases and web applications to bring the strengths of the web to a school without clogging the connection to the Internet, and
  • judging all purchases and acquisitions by price/performance.

According to IDC, some of these points are connecting in Australia where, “The Education sector contributes 5.9% of ICT spending in the total Australia market, representing $2,782.2 million in 2012. IDC expects ICT spending within the education sector will grow to $3,162.4 million by 2015, or 2.1% (CAGR) between 2010 and 2015. In 2011 hardware continues to account for the majority of education sector spending (41.7%).

Additional survey results show that the top three organisational priorities within the Education sector are: migrating to new hardware/software platforms, aligning IT/IS with business direction and developing effective business cases for IT investment. These priorities reflect the focus across the sector on putting the right infrastructure and platforms in place to deliver a new kind of reality in the delivery of education.”

It’s all good. As long as they look at price/performance, a lot of good things will happen in education there. I have been in schools that did not even have a server on the LAN let alone realize what a server could do for them. Indexed collections of documents, good web applications to accept and to maintain collections of knowledge, lessons, data, and leveraging the same characteristics that make FaceBook popular on the web to education are too important to miss. Rather than have everything in the cloud accessed by expensive Internet connections, it makes sense to have one or more servers in a school to give those benefits locally. The cross-fertilization that takes place in an academic community is priceless when everyone contributes and is valued for their contributions.

I recommend Debian GNU/Linux for the infrastructure of IT in schools. Debian’s first priority is to provide software that works well. There is no profit motive or desire to lock-in to interfere with educational objectives. It’s all about IT that works.

About Robert Pogson

I am a retired teacher in Canada. I taught in the subject areas where I have worked for almost forty years: maths, physics, chemistry and computers. I love hunting, fishing, picking berries and mushrooms, too.
This entry was posted in Linux in Education, technology. Bookmark the permalink.

23 Responses to IT in Education

  1. Kozmcrae says:

    aardvark said:

    “Not something you seem to bother with, much, since you apparently know everything there is to know already.”

    Now why would you say a thing like that? Have I ever hinted at knowing everything there is to know? How about even a lot of what there is to know? How about a lot of what there is to know in IT? Why broach the subject of knowledge at all?

    Why bother with such a cheap insult? It doesn’t make you look good. You must be able to do better than that. Can you? I’m sure you would have if you could.

    Next time think a little before you type. You just may come out looking a little less feeble minded.

  2. aardvark says:

    It’s a request for information, Mr McCrae. I am extending to Mr Pogson the courtesy of admitting that he has far more experience in this field, and asking him to share it.

    Not something you seem to bother with, much, since you apparently know everything there is to know already.

  3. Kozmcrae says:

    aardvark said:

    “Since you’re an educator and a thin-client expert, and you’ve used K12 in the field, is there any possibility of you taking some time out and giving us a review of where it works, where it could do with improvement, and what it has perhaps been missing since (say) 2009?”

    Are you freakin’ serious?

    Of course you are. This is going to be fun. Stick around aardvark.

  4. aardvark wrote, “Mr Pogson, don’t you think that “full screen video” is a rather basic requirement for a teacher?”. Yes, that’s why we have large monitors/TVs/projectors in schools, but we rarely need full-screen per student. That might be useful in a lab for multimedia courses. It’s also a good idea to have one thick client in each classroom for the teacher and full-screen and multimedia stuff. Video is highly overrated for education, too. It is an opiate for the masses not the cornerstone of education. Students learn at their own rate, not a certain framerate. Students have a much higher bandwidth for learning from text/speech than video. Ask witnesses at a crime what they saw…

  5. aardvark says:

    Apart from the fact that I regard K12Linux and its cousins as a pathetic and obvious failure to address the need to sell into the educational market, Mr Pogson, don’t you think that “full screen video” is a rather basic requirement for a teacher?

    Imagine the scene in the golden age of education, say the 1950s. Thirty eager pupils await, sitting by their desks. An enthusiastic teacher bounds in.

    “Sorry, kids,” she begins, “We had hoped to get the modern technology of a 24″ by 48” blackboard up and ready for this semester, but there are technical problems that you don’t need to worry your little heads about.

    “It doesn’t matter, though! Thanks to a bulk purchase from the Free Beer Company (please support our sponsor by mentioning it to your father!), we’ve been able to equip every single desk with a clay tablet and a triangular stylus.

    “Today’s first lesson, children, and I’m sure we’re all going to find it both challenging and interesting, is How To Write Cuneiform!”

    Isn’t gonna work, is it? You need the right tool for the right job. As far as I can see, K12Linux (abandoned by its developers and “gifted” to Red Hat circa 2009, as is the way with FOSS projects) is basically the Cuneiform of the School Room.

    I’m sure you, or Mr Oiaohom (who is the original champion of this absurd waste of everybody’s time, as usual) have links from, say, 2010 to prove otherwise.

  6. The only thing LTSP does not do well is full-screen video. Youtube videos part-screen are OK. The problem with full-screen video is that every pixel has to be re-written at every change of the screen something LTSP has a hard time doing more than a few frames per second. If more than a few people do full-screen video at the same time the network may limit throughput. Folks just browsing/typing/viewing/reading take ~200 kilobits/s, so a 100 megabit/s NIC can deal with 5 simultaneous users doing full-screen video at a few frames/s or 50 doing simpler things. Typically, for a school I would want a gigabit/s NIC for every few dozen users because loads are spikey and unpredictable. Booting is typically the biggest load when PXE is used. Starting clients remotely using WakeOnLAN is quite useful just before office hours.

    The first time I ever used LTSP in a LAB I had 30 users boot up and log-in on a 100 gigabit/s NIC. I was horrified to see all of 1.5gB RAM get used but a lot was in cached files so the system flew. That was a single 1.8gHz 32bit CPU and users got much better service than from P4ish thick clients running that other OS.

    GNU/Linux is a truly networked OS. The heart of LTSP is X which was there from early on. In fact, one can install GNU/Linux on the old hard drive of a PC and insert X -query serverip in /etc/rc.local and get a usable thin client with no other modification. Then there is no NFS traffic on the LAN. You get the boot-time of the old machine and the performance of the newer machine used as server. It’s like magic seeing an old machine beat a new thick client at many tasks like logging in or opening an application.

  7. aardvark says:

    Hmm, interesting. Since you’re an educator and a thin-client expert, and you’ve used K12 in the field, is there any possibility of you taking some time out and giving us a review of where it works, where it could do with improvement, and what it has perhaps been missing since (say) 2009?

    I still don’t see it as a saleable* package, but it’s probably something that needs more publicity.

    * Mr McRae et al, please spare me your definitions of “saleable.” I’m not talking about a monetary transaction. I’m talking about “selling” it to an opinion former in the educational field … which might be almost anything. But it isn’t the random “thousands” quoted on the K12Linux website, no matter how useful those thousands find the system to be.

  8. aardvark wrote, “K12Linux (Mr Pogson will be able to pull up the relevant stats from DistroWatch) appears to be moribund from about 2007 or so, and this is probably because it doesn’t offer anything of the sort.”

    According to Distrowatch the last release was 2009 and based on Fedora 10 which was released in 2008. Since Fedora does updating and upgrading, I doubt that means that installation is now “moribund”. In fact, on their site one finds, “K12Linux EL6 is our long-term supported LTSP distribution that will receive updates until the year 2017. Read more about our progress at 2011 Fund Drive.” See more at and

    In any event, most of the functionality which made K12 LTSP useful has been moved from a single distro to LTSP packages widely available in GNU/Linux distros. see, for examples, Debian GNU/Linux, EdUbuntu etc.

    The load on LTSP was too great so by creating a package easily ported to other distros, LTSP is now a bunch of scripts that exploit readily available packages in GNU/Linux. There is a chroot on the server and a system administrator can add/subtract software for the thin clients in the same way done for a regular system. LTSP5 does tend to add bloat to the clients, about 72MB is needed to run them but LTSP 4 works fine in less than 64MB. So, many old machines make good thin clients even with slow hard drives or crappy motherboards. Old machines may not have the gut to run SSH so I often switch to plain old X.

    So, LTSP and its derivatives is not dead but rather has evolved and spread over Earth like the dandelion.

    K12LTSP was the first distro I ever used as a computer teacher in a lab. It was wonderful. In Edubuntu, I used the same technology for a whole K-12 school. It works very well when folks are mostly reading, writing and clicking… We even made USB devices like storage, cameras and scanners available just by plugging in anywhere in the building although all applications ran on the servers. Audio worked, too, right from the thin clients but video was jerky. That’s why teachers should have thick clients.

  9. aardvark says:

    Mr Oioaohm:

    I see you have added a B.Ed to your list of glowing qualifications, which apparently include service on an Australian submarine, an extensive career in the law, and many other challenging areas of expertise.

    “Teachers order of requirement.
    Stability number 1.
    Repeat ability number 2.”

    I’m only from a teaching family; I have never been a teacher myself. My guess is that requirement number 1 is educational software. Your purported requirements are subsequent to that.

    “Number 2 is quite important so they don’t look like a twit in front of class. MS products are very good at making teachers at times look like twits because something works on one machine not others.”

    And I’m totally convinced that you have chapter, verse, and links to prove this rather sweeping assertion of yours.

    Feel free to share them with us.

  10. aardvark says:

    Mr Oiaohm:

    “aardvark asked for something that exists.”

    Well, possibly I mis-phrased the requirement, for which I apologise. From the rest of my post you should be able to glean the obvious fact that I am not looking for yet another thin-client terminal server. I am suggesting that a coherent and broad-based software solution is in order.

    K12Linux (Mr Pogson will be able to pull up the relevant stats from DistroWatch) appears to be moribund from about 2007 or so, and this is probably because it doesn’t offer anything of the sort. What it does offer is:

    * Educational Desktop Software
    * Typing tutors
    * Math, chemistry, physics
    * Basic office productivity
    * Web browsing
    * Other Educational Server capabilities:
    * Web content filtering
    * Moodle

    I think it’s fairly clear that this list (apart from being sheer puffery) does not meet my prospectus.

    Or, in other words, no, Mr Oiaohm: what I am asking for does not currently exist. It would be excellent if even the germ of it did, however.

  11. I tell teachers to “lead, follow, or get out of the way” when it comes to IT. Some students are very adept at learning using IT and put their teachers to shame. All a teacher has to do is point students at some resource with some objectives and it will work. That’s why I set up a “web” portal for schools with one-stop shopping for education: vocabulary for younger students, and hard data and search engine for older students. I like to have the content in-house to avoid the Internet connection being a bottleneck. I have been in schools where a teacher approached me with a CD loaded with exercises or resources from the appropriate department of education and I was legally able to put the contents of the CD on the server and have multiple students use it at once. GNU/Linux + Apache is wonderful for schools. Chuckle, one time, the CD had variable-case filenames, so I had to write a script to shift all of them. I gave them feedback… They had no clue that any PCs were case-sensitive. Fixed that. I e-mailed them the right HTML and directory structures.

  12. oiaohm says:

    Robert Pogson Multiseat helps repeatability. Less instances to have instance differences.

    Cost falls into administrators of schools not the teachers exactly. Teachers decide how or if the tech is used. You cannot deploy tech if the teachers are hating how it working for them. Because then the tech will sit there and do nothing.

    So 1 and 2 items I stated you could give every student a new computer and if the teachers did not find those computer passing 1 and 2 they would not exploit them.

    Some things fail in education because they don’t consider the market. Education is different to most other markets. There are particular requirements for Education.

  13. Those are good points, oiaohm but in there somewhere has to be number of seats at PCs. It’s really hard to do much for a whole class with 0-1 PCs in a classroom. In some schools I have set up a mini-lab with 7 seats. Then a teacher can divide the class into groups and on of three or four stations will be the PCs. That is actually affordable for some with Multiseat X (one PC with 7 monitors/keyboards/mice). Cost is perhaps $1200 per classroom and it is much more productive than a single whiteboard… GNU/Linux certainly helps. No licensing fees or CALs. Where I live, schools can get old PCs refurbished and delivered for $0. Of course GNU/Linux is stable and fast but even GNU/Linux is of limited use if students need to wait for inspiration for a visit to a computer lab. I often advocate that labs be closed down and PCs placed into classrooms if there are not enough PCs to go around.

  14. oiaohm says:

    aardvark asked for something that exists.

    This is Ubuntu problem the education market is a battle between Redhat and Debian or relations. Teachers are not after flashy they are after that it works.

    Ubuntu is the wrong class of OS mostly for education Ubuntu focuses in the latest and greatest. Heck even today you will find schools using MS Office 2003 to teach word-processing. Its not like teachers need the latest or greatest when they are teaching a concept.

    Teachers order of requirement.
    Stability number 1.
    Repeat ability number 2. Number 2 is quite important so they don’t look like a twit in front of class. MS products are very good at making teachers at times look like twits because something works on one machine not others.

  15. aardvark says:

    Thanks for updating me with the list, Mr OE: if I had a kid (or a nephew or a niece) at the moment, I’d probably try some of those.

    I’m not “trolling” on this one, or however you want to see it. I genuinely think that an “Educational” distro for Linux could get a pretty serious purchase on the market, if it was done right. (Clue: Canonical failed.)

    Imagine, for example, a start-up like RHEL was in the early days, selling the concept around University campuses. You could probably organise a pretty decent set of software teams, and produce some interesting and valuable stuff.

    I personally do not see it ever happening (surprise!) but I think it’s far more viable than expecting World Domination to fall into your lap.

    The problem with that list (and with every other Linux list I have ever seen) is that nobody much in the “opinion mover” area takes it up and runs with it. You can’t really build up an ecosystem in educational software simply by throwing neat little programs at it: you need a lot more organisation than that.

    And no, eight year olds do not count as “opinion movers.”

  16. TuxMath is a killer on a GNU/Linux terminal server. The “flaming” what-nots really hog the bandwidth. With each flicker of the flame the whole screen is redrawn… All the others are superb.

  17. oe says:

    Funny our 8-y.o. a Linux native certainly has found the following addictive:

    GCompris – Basic reading, math, visual-spatial skills
    Childsplay – ditto
    Kimo suite – Chock full of toddler and pres-school aged puzzles, games, math, reading
    TuxType – Typing tutor, games
    TuxMath – Math flash tutor that has been a breakthrough for him mastering speed addtion/subtraction
    Celestia – Incredibly powerful and addictive space simulator
    KGeography – Excellent geography drill tool.
    Even some of the games are very good—
    LinCity – A nice open-ended city simulator in the vein of the original Maxis offering

    This is not a comprehensive list by any means….

    Now if your looking for shrink-wrapped stuff with lot’s of polish made by high-paid consultants rather than advanced hobbyists (as a labor of love), then yes, the closed source world is where to go. But I’ve found these open source educational apps for more oriented towards synthesis (open-ended) skills as opposed to scripted skills spitback. For many kids failed by traditional school instruction, the FOSS educational stuff is a categorically better bet for stimulation of critical thinking and open-ended query(the typical commercial offering might be better for teaching to tests…)

    Some of the games are even pretty good:
    LinCity –
    SweetHome3D – Drawing

  18. Kozmcrae says:

    “Education is precisely the sort of niche that you would expect Linux on the desktop to target — but I see no evidence whatsoever that it does.”

    aardvark has his head stuck up his proprietary arse. Of course that’s the way you see it. There is no “Linux” to “target” anything. It’s not a single entity like a company. Yet itches get scratched. You want something done? Do it yourself. Nothing is holding you back except maybe a company that wants all the business for itself.

    You think that because it doesn’t happen like you expect it to happen, it isn’t going to happen. That’s not how Microsoft became what it is today, doing things the way they’ve always been done.

    It should be clear to you by now that Microsoft is not on the leading edge anymore. They are the way things used to be. And if they don’t change their ways, they will be the way things were. People want what FLOSS gives them. It’s just taking time to get out from under Microsoft’s heavy hand. But time is the only constraint.

  19. aardvark says:

    As a small service, I found NotaBene whilst I was trawling around the other week for academic software. It sounds really rather good. It’s been thirty years since I was an academic, but even I started to salivate.

    Don’t worry, it’s not TOOS-only. You can get it on Macs.

    Eventually, of course, it will be available on God’s Own Operating System.

  20. aardvark says:

    “I like MediaWiki, phpBB, Gallery, WordPress on LAMP along with a good search engine like SWISH-e and databases like MySQL or PostgreSQL…”

    I’m glad you like them, Mr Pogson, but that alone does not make them educational software. In fact, for grade/high-school kids, it’s actually hard to think of anything less suitable as educational software than PostGreSQL (which I happen to like).

    I think it’s fairly clear that the sort of software available on Linux is not the sort of software that will appeal to teachers or, indeed, kids. And before you claim that this is an unfair comparison because of Microsoft’s marketing muscle, etc, let me remind you that Apple had a pretty good thing going with educational software in the 1980s and 1990s and even into the 2000s. The thing that killed Apple educational software was the price of the hardware and the price of the software.

    These are obviously not impediments to the successful promotion of Linux educational software. Education is precisely the sort of niche that you would expect Linux on the desktop to target — but I see no evidence whatsoever that it does.

    Of course, given the state of X-Windows and Pulse Audio, you would mostly be targeting schools for blind deaf-mutes, but it’s still a valuable niche, and you would at least have Helen Keller on your side.

  21. aardvark wrote, of educational software, “I don’t think there’s a single useful piece of software out there that is sold on Linux.”

    Not sold, perhaps but shared. You can buy Blackboard licences but Moodle is better and free as in $0.

    GCompris is excellent for youngsters. So is ChildsPlay. The normal GNU/Linux desktop works very well for older students and GNU/Linux includes many wonderful web applications a teacher can run form a normal PC or a server. I like MediaWiki, phpBB, Gallery, WordPress on LAMP along with a good search engine like SWISH-e and databases like MySQL or PostgreSQL for teachers and students to amass data in a usable form. Many schools still do too much on paper. GNU/Linux has all the tools to be paperless if they can get enough seats and seats are twice as affordable with GNU/Linux.

    The think I like about GNU/Linux is that one does not need a “toy” application designed for schools but you can use the normal “grown-up” things configured for education and they work very well. For example, it is simple to have a snapshot of Wikipedia running on the LAN but devoid of politics, drugs, sex and violence or whatever a school feels is inappropriate for students and teachers, student and parents can all interact with the body of knowledge that grows in the schools.

  22. aardvark says:

    Single useful piece of educational software, I should correct myself.

  23. aardvark says:

    It’s a nice dream, Mr Pogson, but not one founded in reality.

    For one thing, I don’t think there’s a single useful piece of software out there that is sold on Linux.

    For another, a high school or a grade school is not “academia.” This is not an elitist point: simply that the needs of a grade/high school student are utterly different from those of a college student.

    What I’ve seen of your argument suggests that you favour hacking something together locally, either on a per-school basis or a per-school authority basis. This isn’t likely to scale. In fact, when the Mr Pogson of the future falls under a bus (God forbid, I hasten to add), it isn’t even likely to survive very well.

Leave a Reply