Robert Pogson

I like to save money. It lets me buy more stuff sooner or later…

I have long saved money in education by using GNU/Linux on PCs and on thin clients with zero licensing costs. I always chuckle when I read the anguish of some people trying to eke out similar savings with that other OS. Yes you can save money by using thin clients with that other OS because thin clients are cheaper and CALs are cheaper than full licences (just barely) but the maths is really simple with GNU/Linux. $0 beats all other licensing regimes of that other OS. No need to agonize over four plans each with negotiated prices to work things out. Install GNU/Linux and go.

The latest article I read on this topic suggests it can cost $thousands per PC to run that other OS and the price can be cut down to $hundreds by using RDP and thin clients. You save power and maintenance at the same time. In the table below, the black numbers are from the article using that other OS (adding capital cost, power and maintenance) and the green numbers are for the same hardware using GNU/Linux.

Costs for 1K PCs per annum

I would also argue that, going with a terminal server setup, there is no need for £300 thin clients but more like £100 or less so the last number comes down to £11,000. Whatever, you save a lot per annum, per PC with GNU/Linux. TFA gets many other things right, though. Thin clients last longer, use less power and are easier to maintain by far. The numbers above do not reflect the costs of the server but a GNU/Linux server is much cheaper than one burdened with a licensing fee and a mess of CALs as well.

The part about full virtualization being pricey is true, too. A virtual machine per user is a terrible waste of resources compared to a GNU/Linux terminal server sharing memory and cached files amongst users. That other OS needs that extra layer of complexity to keep the whole thing from falling down.

Read The Register – Desktop virtualisation: Yes, it’s cheaper

“OSS Watch
National Software Survey
2010
Jane Alexen Shuyska
&
OSS Watch
…
In terms of procurement policy we see an everincreasing awareness of the possibility of using open source software. There has been another big increase in the number of institutions that include the consideration of open source in their procurement policies, both in Higher Education (HE) and Further Education (FE) (figure 6). This will help creating a more level playing field for suppliers of open source software.
On the other hand, there is still a relatively large number of institutions that indicate they prefer closed source over open source (35% of FE and 15% of HE respondents, figure 5). We suspect this is based on a continued lack of understanding about open source that needs to be addressed.
…
when asked for the common reasons for rejecting open source software in procurement, most of the top criteria are not related to TCO. Issues that heavily influence TCO, such as migration costs, do not appear in the top 5.

Thus the total proportion of institutions using open source software to any extent has increased from 17% to 50% in the FE sector and from 38% to 59% in the HE sector. The proportion of institutions reporting to use all or almost all closed source software has correspondingly decreased from 83% to 50% for FE 62% to 41% for HE institutions.
…
Windows XP and Windows 7 are currently the most popular operating systems on desktop computers across the FE and HE sectors. The use of Mac operating systems has increased since 2008 and so has the use of Linux systems. Linux (Red Hat) is now used in 34% of HE institutions (13% in 2008), Linux (Ubuntu) is used in 16% of FE institutions (8% in 2008) and 31% of HE institutions (10% in 2008). Overall HE institutions are more likely to use open source operating systems on their desktop computers than are FE institutions.
…
The Mozilla Firefox browser is also very popular, especially in HE where it is being used by 85% of institutions. The use of Safari has increased since the 2008 survey in both the FE and the HE sectors (from 30% to 47% in FE and from 37% to 66% in HE institutions). Google Chrome was introduced since the last survey and has been taken up by a sizable proportion of institutions across FE and HE. The use of Matlab in HE has grown (from 17% to 42%). The popularity of OpenOffice has increased to a lesser extent from 30% to 37% in FE and from 23% to 34 % in HE institutions.
…
A large number of comments concern reasons for choosing closed source over open source solutions. A number of respondents state that closed source solutions suit their institutions because they perform adequately and furthermore carry education discounts, wherefore they are cost-efficient solutions. Conversely one respondent wrote about open source products not functioning adequately:
…
“Software procurement is (like anything else) subject to tendering processes – the University decides it needs something, asks suppliers for information (including licensing and support costs), invites tenders and chooses the best fitting products. There is no route through which Open Source software which is not provided and supported by a supplier can (nor arguably should) break into this competitive and evaluative process.”
…
“In FE colleges, IT teams are relatively small and we cannot rely too heavily on their own knowledge to support open source because of the danger of critical staff leaving for higher pay or other reasons. It is better to use proprietary software for which support can be purchased in an emergency.””

So, in spite of all the naysayers here, FLOSS is alive and well on the desktop and making significant gains over the last few years. FLOSS is used on the desktop in more than twice as many institutions since 2008. There are obstacles to adoption but performance is not one of them. When people get around to thinking objectively about FLOSS they will choose it. The largest reason for not choosing FLOSS? “It’s not what users want.” As if IT is a democracy in education or anywhere else. Institutions should choose software that fits the current and future needs of the institution. Budgets and performance should be far more important than what the users want. Users are largely not knowledgeable about choices and only know what they have used previously, not a basis of rational thought.

In the largest migration I attempted, expressions by staff that the GNU/Linux system and applications were not what users wanted were extremely rare. People were glad to have IT all over the place that worked well and they did not give a damn that it was “different”. I suspect this supposed push-back is mostly in the minds of IT staff or their bosses.

The survey shows clearly how far FLOSS has come and how far it has yet to go in adoption on desktop and server but there is no sign that wider adoption will not happen. It’s just a matter of how fast. As we see the rapid adoption of FLOSS on mobile gadgets, I expect that FLOSS will rapidly expand on desktops. The world clearly accepts FLOSS if some institutions do not.

Government is all about IT. There is a lot of paper involved still but computers do more of the heavy lifting each year. After years of tightening IT budgets the need to optimize IT in price/performance is rising up the list.

“One of the biggest drivers for government IT spending is the need to become more efficient and reduce long-term costs, which was named by about 60 percent of respondents. Other big government IT spending influencers include compliance with government mandates or regulations, responding to the needs of citizens and staff, as well as the necessity of modernizing legacy systems, particularly those vulnerable to security threats.” Continue reading ‘Governmental Spending on IT’

A reader brought this to my attention. An organization has developed a prototype of a small cheap computer on a USB stick that could bring back the enthusiasm of young people for IT that they had in the 1980s with the Ohio Scientific Superboard, and other inexpensive PCs for ordinary people. In those days, there were many who could afford a PC that cost a few hundred dollars but not the commercial stuff costing $thousands. Today there are still children who can be stimulated by smaller cheaper computers.

This video describes the device and the goals. ICT is “Information and Communications Technology”, a current theme in education, basically “using a PC”. The speaker wants to encourage youth to create programmes and content rather than just using a PC.

The speaker makes the point that “using a PC” stifles the creativity of young people who are curious and want to understand how things work rather than what others want you them to do with a PC. A lot of the vitriol on this blog is about that. Some folks think it is just fine to use M$’s software and no other. Others think we should cast off such limitations. In education ICT is a big step forward in making use of ICT universal but it does not encourage the mastery I have seen in “Information Processing” courses where computer programming and servers and networking and other content creation is taught/learned. Many high school students are turned off by “keyboarding” when they could be stimulated by solution of real problems and understanding the basics of computer science and pushing the limits of what a PC can do.

One of my most interesting lessons involves getting students to do a measurable amount of computation to see how fast a PC is. It is amazing to them that an 8 year old PC can do millions of calculations per second and yet be sluggish with XP… I also love to teach what kinds of things brute force attacks can do, like cracking passwords or graphing or searching. Comparing a binary search to a linear search is mind-blowing for some students. One of the tasks I loved to do in a computer lab was taking some large project and dispatching parts of it to every PC in a lab so that it is done N times faster. Simple ideas like that are immediately useful for students who are not afraid to look under the hood and use the hardware for what it can do and not just what some business lets them do. GNU/Linux and ARM has a role to play in making this kind of technology more available.

see also
A 15 pound computer to inspire young programmers

Geek.com

RaspberryPi The Raspberry Pi Foundation is a UK registered charity (Registration Number 1129409) which exists to promote the study of computer science and related topics, especially at school level, and to put the fun back into learning computing.

Our first product is about the size of a USB key, and is designed to plug into a TV or be combined with a touch screen for a low cost tablet. The expected price is $25 for a fully-configured system.

Having simpler desktop/client machines and more complex servers saves money eventually even though M$ likes to “get value” all over the place:
“Microsoft, has estimated that the cost of a full VDI implementation will be roughly 11 per cent more than a distributed PC-based environment.“. VDI saves money by making clients simpler and putting all the configuration in one place where the IT guys can get at it easily, on the server. Visits to a PC are much less likely. Updates of software and installation of software and backups are all easier on a centralized system. Continue reading ‘Cheap Virtual Desktops’

For more than ten years I have been pushing for schools where I worked to have more computers in the classrooms. There are pros and cons:

• some teachers fear being “replaced”,
• some classrooms and furnishings just don’t fit a raft of desktops,
• old desktops are noisey, gather dust and release heat,
• some teachers don’t know how or the software they use does not help to manage students on computers,
• they cost money, and
• it’s more/different work or change to adopt computes in the classroom

Continue reading ‘Computers in the Classroom’