cdt5050

And that gets really, really complicated when you have 1 teacher and 200 students to teach that week, each with a different hardware configuration, differing skill levels, devices in various states of repair, etc.

RPi is pretty much the cheapest way to get a brand new, standardized device into the hands of each of 200 students, then give the same curriculum to each student that will work the same, with predictable errors the teacher is going to be more ready to troubleshoot, instead of trying to figure out why some kid's PC segfaults because of an error on a Broadcom device that was only available on devices manufactured in May of 2014 (but sold in September of 2017, the ones manufactured 05/2014 but sold before 9/2017 are perfectly fine) when he tries to use any function using sin in Python (but the libc one works just fine).

Schools have historically paid a premium for this, like the Basic Stamp, which was really just a PIC microcontroller preflashed with some other stuff on the PCB to help it run, and sold at a crazy markup, but was well integrated into a standard curriculum that gave every student the same starting point and something that a teacher could easily help hundreds of students with. It's also kind of why Arduino became so popular -- did the same thing, but a fraction of the cost.

Nah, the lack of hardware is a real problem in developing countries, and is something that OLPC and Raspberry Pi are well suited to solve. But hardware is the easy part -- the much harder part of the problem to solve is the lack of human capital in computing education, and that's the part of the equation OLPC missed.

For example, they sent hundreds of thousands of laptops to Uruguay, but they hardly got used for education because the teachers just simply didn't know how to use them or integrate them into the curriculum. But Uruguay continued the OLPC model under the Ceibal project, and it's only just now, ~15 years later, after teacher training regarding technology is far more widespread and teachers have far more experience in utilizing computers in the classroom, that it's starting to show effectiveness. And the OLPC model is considered a success in Uruguay, but it took over a decade of persistence to work itself out.

So this is actually something that RPi is tackling head on -- the Raspberry Pi Foundation is actually working hard to develop curriculums meant to address challenges in developing countries, like poor internet connectivity (which means a curriculum designed to be self-sufficient), and then also putting a lot of effort into training educators. This is before they even try sending Raspberry Pi's out, and largely is designed around using the already limited computing devices they have (not RPis) because a lot of those governments are unwilling to invest in computer education (because they have bigger problems to worry about and consider things like every student having a computer to be a luxury when nutrition is a more pressing matter on the quality of education).

Frankly, even the idea of "most people there use x86 PCs or laptops from the used market or rescued from e-waste imports and repaired because they're much cheaper" is a fair bit of a luxury and lacking understanding of the challenges of computing education in developing countries once you leave the cities and start looking at rural populations. In Mexico, nationally, 45% of households have a computer and 56% have an internet connection. Once you go to rural areas, that's ~20%. And that's before you consider socioeconomic status and the disparity in access to computers and internet. Some stats put the lower-income Mexicans at 5% computer ownership nationally.

In Rural Mexico, it's normal for a computer lab of 30-50 computers to serve a school of 500-1,000 students, or for schools to have NO computers and instead send students to nearby schools occasionally to utilize their larger computer labs for learning computer skills, and that's the only computer access they realistically have. This is really a great use case for RPi's, but like I said -- the hardware is the easy part to solve. The fact of the matter is if you sent a crate of RPi's or any other computer to these schools, they'd sit in the boxes for years because they don't have the infrastructure or the human resources to start using them. And that's actually the part that RPi Foundation is focusing on first.

The Pi 500 does definitely swing in the direction of the computer.

IMO the target demographic of the Pi 500 are people who want, first and foremost, a desktop PC, but one that is well suited to hobbyist computing and/or computing education.

Yeah, if you JUST want to run a word processor, browser, etc. one of those N100 PC's might make more sense. If you open up an issue of MagPi magazine and feel inspired, then the Pi 500 might make more sense. And that's what it really comes down to. Is computing, itself, a hobby or interest to you, or not?

A huge reason SD cards have been unreliable in the past is because of a lack of wear leveling functionality, and architecture assuming sequential writes (e.g. large buffers that require larger blocks of data to be erased and rewritten even when writes are small as is typical in random writes, resulting in excessive write cycles).

Modern SD cards have much more intelligent controllers that perform wear leveling and work with the OS to reduce excessive write cycles.

Not that they're as reliable as modern SSDs, but I'd be willing to bet a modern SD card from a reputable brand is as reliable as SSDs were a while back when they were generally regarded as reliable enough.

Almost every time I've teased the info out of people with failed SD cards on Pi's, either it's either physically damaged, or it's an off-brand, or it's an ancient card.

Pi 5 is a great everyday desktop and development machine.

Some websites are going to perform really badly, like a lot of news websites with lots of ads, but those tend to perform bad on my gaming PC, too, and I just think if a website performs bad on a Pi 5 it's really just not worth visiting.

The performance difference is practically unnoticeable in day-to-day desktop use. People don't really care about 15 second vs. 20 second boot time, and that first boot of Firefox is going to be another pretty meh comparison. Subsequent boots are going to be loading small enough data that the difference between will be meaningless.

The main appeal to an NVMe on the Pi isn't speed, it's cost effectiveness. SSDs are cheaper at larger capacities than SD cards. A 1tb SD card is 2x the cost of a cheaper 1tb SSD.

Storage speed is a bigger consideration on something like a gaming PC or a video editing machine, where you're going to be streaming gigabytes of textures/videos off of storage, not just trying to launch Firefox.

I find that communities like these are far too focused on "TECHNICALLY CORRECT". Like, yeah, there are some things that technically are faster and cheaper than a Raspberry Pi, like that ebay special or that Alibaba-only SBC that's a direct copy of orange pi, or whatever.

But when it comes to things you can't glean from spec sheets, they utterly fail. Like the most important thing: the human experience. Like when it comes to Pi competitors, nothing even comes close to the Pi in having such an ecosystem of learning resources. I personally really appreciate the MagPi magazine as it's a great source of inspiration.

The Raspberry Pi has three huge advantages over those other options:

1) GPIO, letting you do things like learning to flash LEDs or controlling robotic arms and getting ready to use a Pi Zero for robots or such.

2) Cool add-ons, like the Pi Camera or AI modules, and a product family such as compute modules, Pi Zero, etc. A strong third party market of Hats, etc.

3) SUPPORT. Pi 500 is available as part of a kit with the official Raspberry Pi Beginner’s Guide, which I know from experience in teaching people interested in learning this kind of stuff, is a really awesome getting started resource. But it doesn't stop there -- Raspberry Pi Press has a lot of freely available books that teach various topics, AND a monthly magazine to teach and inspire. Not to mention countless tutorials meant specifically for Pi and Pi OS.

This is really, really awesome.

Those ebay specials aren't going to satisfy any of these. And while direct RPi competitors do provide a response to the first point, they get weak on the second point, the RPi is pretty unique in the third point.

Upton talked about this in an interview. It's partly the interplay of manufacturing at this point, too.

When they have parts, they'd like to catch up with the backlog faster. Because the 5 doesn't rely on the robots that place through-hole components, its production can to a degree "overlap" with the Pi 4, basically increasing the total number of boards created per hour.

Having two SoCs that don't require the same machines also helps ease a bottleneck there.

Because I don't just want a PC, I want a learning/hacking/making platform, with the ecosystem and support to go along with it.

A used nuc or dumpster dived computer doesn't have two magazines, a couple dozen books, official support and documentation and guides, and an entire world-wide community dedicated to teaching and inspiring me.

The biggest reasons to prefer a Pi over a mini PC: the documentation, Raspberry Pi Press, the community, the ecosystem.

Let's put it this way: you pop in your SD card, set up a username/password, a few other settings, and you end up with a system that already has Geany and Thonny installed. Go to Help -> Bookshelf and find all the issues of MagPi and Hackspace for inspiration and learning, and a couple dozen books like "Essentials - Code Music with Sonic Pi" or "Get Started with MicroPython on Raspberry Pi Pico". Search "Raspberry Pi Hats" and get a lifetime worth of reading and learning there, hitting just about any domain you could care about.

The vision of the Raspberry Pi wasn't to create a cheap, tiny PC, it was to create a learning platform, of which the cheap, tiny PC is the basis. More than any other computing device I've seen, it really lives up to that vision.

It doesn't require a person to change how they think about programming from their college C or Java classes, but is a lot simpler than those two languages.

I was using a Pi 400 for my programming PC, basically testing out if I can sell my gaming PC after I got my steam deck. It was perfectly capable for that task, and I’d say I would be happy to use it as my primary desktop PC.

Just sold my Pi 400 after the announcement of the Pi 5, deciding the Pi 5 will be my primary desktop PC. (Also ordered a Pi 4 because I have some other plans for it, but don’t want it built into a keyboard).

Falling back onto my Pi 3 I never really used before — it’s surprisingly capable. Not as a general desktop but gets my programming done fine. Its bottleneck seems to be I/O and limited RAM.