Raspberry Pi Classroom Workgroup Project

Ever since the CrowdStrike debacle brought the United States economy to its knees, people have been talking about our overdependence upon Microsoft Windows, and our need to get back to basics.

Have you heard about the Raspberry Pi? It's a computer roughly the size of an 3x5 index card; it was designed to introduce school-aged children to electronics and computing with an inexpensive (under $50) product that school districts can afford to buy in large quantities.

Most of these school districts DID buy dozens of Raspberry Pis; but finding a use for them is an ongoing challenge.

Have you heard about Open Source? An open source operating system such as Linux makes a marvellous living textbook on how to write software because the people who wrote Linux are mostly experts in their field and they have placed their work in the public domain so that others can learn, as they did, by studying what works, and extending it.

Linux also makes a great platform upon which to build because Linux is easily modified to conform to local requirements. The Linux project leads the way in internationalization and localization and has contributors around the world, making it a premier example of what the human race can do when we set profits aside and all work together towards common goals.

When we personally installed Linux, and FreeBSD, on the 'Pi - to take it for a test drive, as it were - we found that even the lowly Raspberry Pi v1 was capable of hosting a small Minecraft instance - nontrivial, as Minecraft is written in Java, we think, and requires a JVM, which category of application tends to be something of a resource hog.

The newer Pis are 64-bit multiprocessors with four or more cores. These are powerful multitasking computers; not toys. We were very impressed.

However, teachers - attracted to the Raspberry Pi by its promise of simplicity - find its complexity unmanageable. The operating system of choice is Linux; and you almost need to be an IT professional to put it all together.

This complexity has made the Raspberry Pi unusable for all but the smallest groups of the brightest students, ones who can be taught to administer their own computers and install their own software; the ones who already have computers at home. We want to reach outside this group.

In many school districts, for lack of a steady supply of such students, the Raspberry Pis fall into disuse and end up boxed up in the storerooms instead of being in use, introducing students to the intellectual pleasure of a well-designed algorithm or some nicely written source code that does Something Useful.

Let us imagine you are a teacher at a school that wishes to deploy these Raspberry Pi computers in the classroom so that you can introduce your students to computer programming and, perhaps, some sort of very fundamental make-a-red-LED-blink-on-and-off electronics.

Even if you, the teacher, are fresh from college, and a Linux enthusiast, as well, you will be overwhelmed:

Let's do the math:

A typical inner city computer lab classroom may contain as many as 30 students and the school schedule may allow for up to 8 classes per day.

If you have 30 students sitting at 30 computers and you want your students to be able to log in at any computer, you will need to, in a worst case scenario, manually create 30 accounts on each of 30 computers, or a total of 900 accounts. If you have 8 classes you will need to do that 8 more times, or create 7200 accounts ... just for that one classroom ... just for that one semester.

Very quickly, the teacher, inspired to teach his or her students the skills they will need to flourish in the 21st century, is bogged down in administrivia and has no time to teach, grade, or even prepare lesson plans; and another ambitious plan to get children to code, bites the dust, overwhelmed by the permutations inherent in large populations of users and computers.

What this teacher, or school, needs is some good software, to automate these tasks, so the teacher can teach.

Information technology professionals have been deploying and administering large populations of UNIX and Linux computers for decades. They compile customized kernels and customized software, package it and push it out, in parallel, to hundreds or thousands of computers, without a single defect, repeatedly. Some do so on a daily or even an hourly basis - "continuous integration", they call it.

Open source tools for orchestrating large populations of computers AND large populations of users from a single desktop computer have been readily available for decades; some of us wrote our own software tools to do the same thing, even before these more popular tools were introduced.

What is required is one or more people with prior experience building and automating these sorts of infrastructures, to put it all together, test it, release it, and support it.

A systems integrator- that's us.

We are a professional UNIX/Linux system, network, database and firewall administrator, analyst, and architect. We have been designing, building and managing UNIX- and Linux-based infrastructures for nearly forty years.

We have four decades of experience at building and managing industrial strenth, networked, clustered, heterogeneous workgroups and specialized computing infrastructures - for hardware vendors, software vendors, financial institutions, brokerages, dot-coms, hospitals, and schools, too.

In 2004 we designed and built a Linux-based workgroup product for Russ Button; see https://www.redwoodhodling.com/Exhibits/index.html#2004Button for more information. This is not the first time we have tried to bring this specific concept to fruition.

(See, also, https://www.redwoodhodling.com/Exhibits/index.html#Raspilab for some background on this Raspberry-Pi-specific project, and additional professional and biographical information about the prospective recipient of your donation.)

We estimate $4,000 per month for twelve months, or, roughly, $50,000, total, would give us the funding required to rent a one-room office in the San Francisco Bay Area, pay utilities, and acquire a few dozen different models of the Raspberry Pi to use in building a testbed for the software. We will also need at least a hundred microSD cards.

This amount would also provide us with a modest compensation, to cover modest living expenses in the very expensive Northern California region - and to insulate us from being evicted, in the turbulent Northern California real estate marketplace, while we do this valuable work.

According to Wikipedia, there are 3,242 counties in the United States. $50,000 divided by 3,242 county school districts works out to $15 and change, that's, what, lunch for two at McDonald's? All we need is one generous person in each school district quietly investing $20 in their school district's future - or, perhaps, fifty generous individuals, in fifty states, donating $1,000, each - and we're done fundraising, and can get to work.

The names of the project funders at the time of release will be embedded in the software release notes (or not, according to preference), in order of amount and date of donation.

$50,000 would, we think, pay for AT LEAST an entire year of software development; that should pay for a complete development cycle, including beta-testing by teachers and students, before the end of the 2024-2025 school year, so that teachers can spend the summer playing with the new software.

Regrettably, Silicon Valley is no longer a cost-effective place to develop software that does not produce significant quarterly profits, such as educational software. But it's where our children are going to school; so our mobility is limited.

(We would pay for this software development out of our own pocket, if we could ... but two decades of being outsourced, humiliated by the 'gig' economy, and replaced by H1-Bs, has left us homeless, and penniless - while Oracle Corporation's notorious executive gang leader, Larry Ellison, can buy entire Hawaiian islands and still have money left over. Just sayan'.)

We are prepared to set up a website for this project; publish up-to-date status reports of our progress; and provide a monthly accounting of funds spent ('burn rate'), if it seems necessary.

Here is our draft list of priorities, as of now:

1) Build and release a standalone proof-of-concept Raspberry Pi bootable image that teachers can download and show to school staff and other teachers (already done, just need to work on packaging and platform testing)

2) Sketch out our software stack - probably DHCP, a firewall, a networked filesystem of some sort, and some sort of central authentication mechanism, perhaps LDAP - and evaluate different operating systems (Raspian, Debian, Redhat, FreeBSD, etc) for their suitability as a development platform (already done, pending some R&D to make sure)

3) Accumulate a small collection of Rapberry Pis - v1s, v2s, v3s, v4s, maybe some other ARM-based SOCs - and use them to test our releases against small pseudo-classroom-sized infrastructures to see how they scale under load

4) Design, build, and test an administrative interface for teachers' computers; first, command-line shellscripts, then some sort of graphic user interface (GUI)

5) Research some sort of support mechanism; ideally the software will be self-documenting and so easy to use and so reliable to use that support will be unnecessary; but school districts will want such an option before they commit. Online manual pages, of course. A wiki, at the very least. A bug-tracking database. Email. A pager, perhaps.

6) Circle back to the standalone proof-of-concept and polish that up and see if we can release something aimed at home schooling families

7) Assemble a list of books and available PDFs that can be used as textbooks or syllabi for teaching Linux-related skills for aspiring teachers and other autodidacts (already exists in a draft format)

We think this project could serve as a platform for teaching a wide range of classroom subjects besides computer skills, computer programming and basic circuit theory.

With the assistance of the Internet and a web browser, reading, writing, mathematics, and the entire Wikipedia, Khan Academy, and content from most of the world's top universities is also available.

With the assistance of a firewall, those URLs that are not permitted, are denied; guaranteeing that your students remain undistracted by YouTube, or worse.

Because the entire project is based on open source software the cost to school districts for this software would be zero.

There would be no licensing fees; no "per-seat" fees; no "per-CPU" or "per-core" fees. The software would be free.

Based on design goals and prior experience we estimate that teachers will be able to build (or rebuild) a Raspberry Pi workstation in just a few minutes, with no loss of student data.

We think that almost every single school district in North America has a box or two of Raspberry Pis in a closet somewhere, gathering dust.

We want to correct that and put those assets to work, teaching American kids how to think sequentially and how to solve complex problems by untangling them into smaller, simpler tasks.

If you are worried about your grandchildren, or your children, or your students, or even the future of the United States, please consider contributing to this cost-effective effort to convey enthusiasm for computer science to our children, and help us to help you raise the next generation to be producers - not just consumers - of information technology.

No more CrowdStrikes! Never again.

Thank you for reading.

The hardware:

The operating system candidates (most of them):

Some of these operating system releases are application-specific, such as Kali, or Tails. Others may not choose to include the ARM processor amongst the hardware they support, or their support for the Raspberry Pi is inadequate. Some operating system teams may not include some open source software amongst those they have precompiled as packages. Some precompiled open source software may not work as advertised. Selecting the right operating system to use as a platform to build upon has enormous influence over the success or failure of the project in the long term.

Our professional history, including the history of this project:

Why we didn't do this twenty or thirty years ago: