The Compromises Of Raspberry Pi Hardware Documentation

[Rowan Patterson] informed us about a recent ticket he opened over at the Raspberry Pi Documentation GitHub repository. He asked about the the lack of updates to the Raspberry Pi 4’s USB-C power schematics for this board. You may recall that the USB-C power issue was covered by us back in July of 2019, yet the current official  Raspberry Pi 4 schematics still show the flawed implementation, with the shorted CC pins, nearly two years later.

[Alasdair Allan], responsible for the Raspberry Pi  documentation, mentioned that they’re in the process of moving their documentation from Markdown to AsciiDoc, and said that they wouldn’t have time for new changes until that was done. But then [James Hughes], Principal Software Engineer at Raspberry Pi,  mentioned that the schematics may not be updated even after this change due to a of lack of manpower.

As [James] emphasized, their hardware will probably never be open, due to NDAs signed with Broadcom. The compromise solution has always been to publish limited peripheral schematics. Yet now even those limited schematics may not keep up with board revisions.

An easy fix for the Raspberry Pi 4’s schematics would be for someone in the community to reverse-engineer the exact changes made to the Raspberry Pi 4 board layout and mark these up in a revised schematic. This should be little more than the addition of a second 5.1 kΩ resistor, so that CC1 and CC2 each are connected to ground via their own resistor, instead of being shorted together.

Still, you might wish that Raspberry Pi would update the schematics for you, especially since they have updated versions internally. But the NDAs force them to duplicate their efforts, and at least right now that means that their public schematics do not reflect the reality of their hardware.

86 thoughts on “The Compromises Of Raspberry Pi Hardware Documentation

  1. In my opinion, the raspberry pi foundation is unconcerned with anything beyond getting a product out the door. Their advertising is not true in all markets, and they refuse to address false advertising. Avoid them if you care about truth or fairness.

      1. In Japan, every product they have released via official sellers is 15+% more than the advertised price after currency adjustment. Something’s like the pi zero never got close. They were closer to regular pi prices. The pico is made in Japan, but still couldn’t match the advertised price. Every inquiry as to why has been ignored.

          1. isn’t Japan’s VAT rate 10%? – so, you’re now complaining about a 5+% variance. 1 JPY gets you 4% less USD than it did 12 months ago……….and, the value has dropped even more over the past 5 years. So, In summary, VAT, Exchange rate fluctuations lead to prices being 15%+ more expensive than advertised USD prices….that’s my belief, what’s your response?

          2. No. This has been a problem from the beginning, and tax is not a factor. I don’t want special treatment, I only want the same everyone else is getting. How can you justify cheaper prices after shipping over the Pacific than after shipping down the street? I’m talking about non brick and mortar internet retailers. How do you justify missing the advertised price, even by 5%? Do advertised prices not carry weight where you live? Is bait and switch the standard? I pity you.

          3. Do you believe the raspberry pi foundation has no control over their official resellers (the very same ones they link to from their website)? Is there no contract? How is that even possible in today’s world?

          4. Same could be said about every other product being sold in various countries. Prices differ from country to country and it’s always been this way. +/-15% is marginal, as Bleugh said VAT differences, exchange rate etc etc

          5. You guys need to reread my posts. I’ve already accounted for vat, exchange rates, blah blah blah. The 15% is on top of that. Clearly you don’t have to pay extra and you’re happy. I do, and I’m not. Your justifications won’t make me feel better, nor will they make me stay quite about this.

          6. Show us the numbers & source then. So far you’ve presented no facts, just keep complaining about the same thing over and over under every single RaspberryPi-related article on Hackaday.

          7. Can you please give some lists of resellers and prices? lots of speculation right now, and, honestly, I have zero clue, just trying to present you with sensible reasons as to why. Every country is subject to ‘xyz premium’ tax related to all sorts of stuff….I lived in Australia a decade, it’s particularly evident there.

          8. Pico prices and links for Japan: It clearly claims, “The new, flexible $4 microcontroller from Raspberry Pi.” They are never $4, and rarely in stock:

            You click on that link and we come to the page: ” ” where it says, “Buy Now” click that button and we go to the bottom of the same page where we can choose “Japan” from the drop down list. There has been ZERO mention of the price being different than the so far advertised $4.

            Now we see 4 OFFICIAL RESELLERS:

            KSY – 528 yen with 10% tax included and an additional 550 yen for shipping. The yen is 110 yen to 1 dollar at the moment. So $4 should be 440 yen. 10% tax on top of that 440 yen comes to 484 yen. A difference of 44 yen compared to the 528 yen price tag placed by KSY, or an additional 10%. Zero stock.

            Baks – 490 yen BUT 10% tax is NOT included. So 440 yen as the base $4 makes a difference of 50 yen. So a mark up of 11% beyond the advertised $4 price. They have a flat fee of 1000 yen for shipping, which they include in their tax calculations, bringing tax to 149 yen, so the total order for one unit is a ridiculous 1639. You can figure out that markup if you like. It’s crazy. That’s not proper shipping costs.

            Next is OKDO – 454 yen advertised price with no way to add to a shopping cart or actually buy the thing. I’m stumped as to why they’re even listed as an Official Seller:

            Finally is Switch Science – 550 yen with the 10% tax included. This is the most expensive option so far until you consider the cheap shipping of an additional 200 yen. Still the base price is the highest, with a markup of 13%:

            Not a single OFFICIAL RESELLER has the base price of $4. Once you include tax and shipping, they can inflate the prices to many times that amount. In the US, tax is never included in the price because each county can have different tax percentages. If you consider that, then the $4 advertised price is pre-tax. That clearly means at today’s exchange rate, no Pico should cost more than 440 yen excluding tax. If they do, the Raspberry Pi Foundation is falsely advertising $4 Picos.

            I don’t want anything special. I want the advertised price of $4 or 440 yen at today’s exchange rate. Anything beyond that is false advertising and a scam. If you like, I can show you the differences in prices on the Pi 4B 8Gbs. $75 on Adafruit, but 9680 yen on KSY. Of course I’ve already given you links. You can check yourself.

        1. This doesn’t mean the advertising is false. I suspect the hardware cost is not what is changing – and I’ve never seen the Raspberry Pi foundation claim that the headline price includes extras like freight, import taxes or the costs/profit margin of retailers (but I could be wrong – if you have a link that proves otherwise, please share it!)

          1. The price has never been within a couple of percentage points close to $4. If that doesn’t mean false advertising to you, you’re under the spell of corporations and capitalism and there’s nothing I can say to convince you. The problem is that especially at release, the markup was already there. If the raspberry pi foundation has no contract with suppliers to at least TRY to stick to advertised prices, there is a problem with the foundation’s pricing and contracts. Then it would seem that they don’t care at all about their customers. These issues seem meager at pico prices, but the fact they are manufactured in Japan yet were unavailable and over priced is what made me incensed. That the situation is worse with regular pis and is why I started to complain in the first place. The pi zero was a disaster in Japan. Nothing ever got close to the $10 or so that they claimed. I imported from the UK and paid a ton on shipping. Maybe I should have bought from a scalper and gotten it faster. I’m done with the rpf, and I will continue to tell others about their apparent disdain for customers until something improves. This is a dear hobby of mine and I like their products despite their prices, but at 10,300 yen for a pi 4b 8gb as quotes below in the comments, I might as well buy a Jetson.

          2. Here in the UK you do pretty much always get within a stones throw of the advertised price, and then often pay P&P. The fact that shipping and import duties add cost to the rest of the world shouldn’t come as a surprise…

            You don’t want to know what import into the UK costs for supposed ‘cheap’ stuff from many other nations as an individual. Needed to order some more window tint films to control heat, light etc that matched the ones I’d already got, but the stock having to be ordered and shipped from the China factory directly this time as I needed a bigger roll they didn’t stock here anymore (not that the seller told me this) meant the import duties ended up about as much as the damn product, think they might actually have been more expensive..

            You want to complain about that sort of thing then you have to argue for free and open, tariff and inspection free trade globally, which is an obviously stupid idea while things like Narcotics and Guns exist, and doesn’t work for local business while the imaginary thing that is the economy/money of your nation is worth so much more than in the sweatshops of the world – the local business can’t possibly survive if 1000’s of man hours can be imported for the cost of 2…

        2. It’s now too late here. I have to go to sleep. In the morning I will happily show you all the evidence you need. Until then, you’re welcome to go to the raspberry pi website and click on the official resellers list and choose Japan. Then click on any of the resellers listed there and check the prices. This isn’t a hidden conspiracy. It’s all available to the public. That’s why I’m upset about it. They advertise a price and then suggest you buy from people who charge above that price.

          1. I did what you said, checking RPi4B 4GB & Pico in Japan, UK and Germany and the prices are pretty much the same everywhere, not sure what you are talking about.
            If anything the prices in the US are a bit lower, but that’s the case with all the electronics there.

          2. If that is true, you should all be upset that the prices are not as advertised. See my response with the price breakdowns above this reply. The lack of reply depth on these comments is problematic.

        1. This was the case with the pi zero even in the us when first released. If you wanted one you had to buy a bundle with a bunch of options some people never wanted. That is also a scam.

      2. “What are they claiming that is not true?”

        “[T]he lack of programmable hardware for children – the sort of hardware we used to have in the 1980s – is undermining the supply of eighteen-year-olds who know how to program, so that’s a problem for universities, and then it’s undermining the supply of 21 year olds who know how to program, and that’s causing problems for industry.” ~ Co-founder Eben Upton in 2012

        And so, Mr. Upton says that this Raspberry Pi thing was created to solve problems for industry by educating children on how to program. Do programmers not benefit from some knowledge of the hardware they write programs for – does it not improve their creativity? Is it possible that the availability of an accurate schematic might be useful to a budding programmer? Maybe I’m different – but controlling the hardware was always what it was about for me.

        To be clear, I completely respect Mr. Upton’s drive, ambition and engineering talent. He’s now CEO of the Raspberry Pi Trading Ltd, – a subsidiary of The Foundation. But respecting his accomplishments doesn’t blind me to the possibility that he may be disingenuous wrt his motives.

        1. Personally I think they use Broadcom and all the NDA’s that come along with it because its where they started – many of them worked for Broadcom in the first place, and so got a good deal – thus making the Pi possible in the first place. Then there is that inertia, as they know how to work with Broadcom, making future better products easier..

          But if you look at the Pico it does point towards a trend towards stuff that can be documented fully, being entirely in house silicon. Don’t think you are wrong, being able to properly see and control the hardware because its well documented would be great. But there is only so much you can do on that front with NDA’s in place, and the Pi is vastly better at the important details and support than any other SBC I’m aware of… So its not like we really have a better choice yet.

    1. As a maintainer of a lesser known OS for microcontrollers I’ve noticed that many popular platforms aren’t as open as they seem on the surface. Want full schematics and datasheet for the raspberry Pi SoC? Out of luck. Want the source of the ESP32 WiFi stack so as to port it to anything but the stock provided FreeRTOS? Last time I checked it was a binary blob and you had to reverse engineer it.
      Don’t know why, but there are too many platforms out there that only have the aura of openness, but aren’t really open. We need a Stallman for open hardware.

      1. RPI and siblings never had that aura of openness!

        PIs are just cheap and omnipresent.
        That just gives them a Windows-like advantage.
        My surroundings just expect me to know enough about PIs to help them.
        I simply cannot ignore them.

      2. I’m under the impression that Espressif is under some sort of a licensing-system that prevents them from open-sourcing their radio-stuff — Bluetooth and WiFi — and it’s not really something they have much say in. I don’t remember where I read this and it could be false, but it’d make sense: most companies don’t develop their own radio-stuff from scratch, but instead license the designs, and so they legally have to follow the terms set in the agreement. Doesn’t even have to be the entire design, but instead just a small thing somewhere that then contaminates the whole thing.

        Pine64 – folks are trying to reverse – engineer the WiFi – stack for the Bouffalo Labs BL602, so there’s at least some hope that we’ll one day have a fully, wholly open – source software – stack for at least one WiFi – enabled MCU.

        1. I wonder if the rules set by regulatory agencies like the FCC also make it difficult for someone like Espressif to open source their radio stuff and still maintain compliance with the approvals set down by said regulators that allow their products to be legally sold.

          1. Your comment actually just reminded me of the FCC – ruling from 2015, where they require WiFi – devices to be locked down so as not to allow e.g. OpenWRT – users to use “non-compliant” settings. ( )

            So yeah, I’d hazard a guess that you’re quite correct. It’s far easier to just keep any such stuff closed – source and in a proprietary blob than try to fight courts over nuances on a global scale.

        2. The hardware and software Wifi IP in the ESP32 SoC I believe was created from scratch by them hence forms part of their core IP. This also includes the low level wifi driver. The CEO Teo Swee was a specialist in Wifi communication prior to founding Espressif. The Bluetooth hardware IP was sourced from a third party.

      3. No we just need a charismatic truth in advertising advocate that works world wide. I think we can do without another person as odious as Stallman in any case. He’s not charismatic, nor was he all that effective at evangelism for free software.

        FWIW, it wasn’t Pi’s openness that made it popular. It’s the price. No one in their right mind and has any clue of computing history would ever accuse any ARM system of being truly “open”. ARM Ltd was always a mother hen over its IP and the addition of GPUs and WIFI just made it less open.

      4. “We need a Stallman for open hardware.”

        Don’t be too sure… the kind of hardware development required for the SoC requires big money – capital. Capital is invested only when there’s an expected return. Software is different.

        But we’re talking about a schematic here – how off-the-shelf components are connected – not how the chips are made.

  2. What are the alternatives to RPi if I want opensource GPU drivers and mainline kernel?
    I’ve tried some other boards. There were all complete shit in this regard, stuck on specific kernel version OR not having GPU driver. + without long term availability they’re just playthings.

      1. Rockchip, Allwinner and i.MX SoCs are among the more supported ones. I don’t have any experience with i.MX or Rockchip, but Allwinner SoCs may be a little hit-or-miss wrt. what features have drivers in the mainline-kernel and what don’t.

        As for the GPU: most of the ARM SoCs use Mali GPUs and, again, those can be a little hit or miss. Also, hardware-encoding/-decoding of video isn’t a GPU-feature, but instead an SoC-feature, so even if the GPU was 100% fully supported and working, video-acceleration may not be.

        I’ll leave a couple of links for you to peruse in the hopes they may be of some help. If you want to grab a board based on one of these SoCs, I might recommend taking a look at Armbian for the OS: if Armbian doesn’t support the board and the vendor only offers some crappy throw-together with outdated kernel, it’s best to just skip the board.

        1. +1 for Allwinner/sunxi.

          i.MX 5.3 is not good about mainline support (no CPU freq scaling, no VGA output with mainline Linux kernel), but I hope i.MX 6 will be good because of Purism/Librem 5 upstream patches.

        2. No open GPU hardware is why I think it’d be a good idea to get funding together to buy the Rendition Verite intellectual property from Micron. They bought it, made claims they would finish a full OpenGL ICD for the 1000 and 2×00 GPUs, then there was total silence.

          Buy it, open source it. Improve it. As it existed when it was in production the Verite GPU was pretty kick ass for 800×600 or lower resolutions. Topped out at 4 megabytes video RAM – but those limitations should be possible to expand.

          Or buy the IP to lay hands on the designs for the 3300 and 4400 GPUs to use as a base to build on.

          But whom at Micron is the go-to person for asking how much $ they’d want?

      1. I don’t need an opensource hardware!
        I just need something that can run current Linux and not just one very specific and archaic version like most those BananaPi and other jokes run.

        Proprietary RPi does that. (And both my workstation and laptop do that too of course.)

        But it would be nice to know some other non-x86 SBC alternatives that can do that.

        1. Several Bananapi – boards seem to be well supported by Armbian with mainline – kernel. I guess you just did not search for alternative OS – images for your boards or simply didn’t find Armbian when doing so? Armbian is basically the go-to for these kinds of boards nowadays, so it may be worth at least taking a look.

          1. 1. I’m not looking for a ready-made OS. Mainline kernel means any mainline kernel… I do not require compiled and prepackaged mainline kernel and so I’m not limited to distribution channels of specific OSs.

            2. I’ve said that the boards that I had tried didn’t have mainline kernel available OR working GPU drivers.
            That does not mean that I haven’t tried boards with mainline kernel! I have.
            Notice that “OR”!

            Having mainline kernel for me is necessary but NOT sufficient. And last time I checked there were no working GPU drivers on mainline kernel for Bananas.

          2. To grasp below, I admit I haven’t tried, but Lima (mali400 from the bananas) was merged years ago and mesamatrix looks very promising. What was your problem exactly, and how long ago did you last try?

            An honest question, as your complaint does not match the available information …

          3. Grawp, you never said anything about not wanting an OS. Also, mainline kernel is mainline kernel — even if the distro did package a precompiled kernel, it doesn’t mean you can’t still compile and use your own one, so I don’t get what you’re complaining about there.

            Besides which, most of the Bananapi – family seem to be based on Allwinner SoCs with Mali 400 GPUs, so what you want is already there.

        2. I would disagree and say we all need open source everything. Remember what things were like before open source? See how much better they are now? Where would any of us be without it?

        3. I’m pretty sure some of the Rockchip-based boards are well supported with upstream kernels, etc – I feel like I’ve heard of them running pretty stock Debian.

  3. > the schematics may not be updated even after this change due to a of lack of manpower

    Seriously? Is it so hard to extract a single page from the PDF printout of the schematics? C’mon give this to the intern, this literally takes no skill to do and takes minutes.

    This is just a cheap excuse. They do not lack the manpower, they just lack any will to do it.

    1. You’re assuming they have a pdf printout of the schematics. I’d be surprised if they design and review it using paper these days.

      The circuit probably needs to be manually rebuilt in a different file.

      But even if they did have a pdf of the schematics to hand, it all takes time to get done and someone to check it (both for accuracy and that there’s no details on there they can’t release). And then to upload it probably involves someone from a different team. So you’ve now got a designer, a legal, and a marketing person involved.

      The reality is that most projects are only able to do barely more than urgent+important tasks. And this looks like non-urgent+non-important.

      1. I don’t buy those excuses.

        The schematics clearly would have to already be correct, since they are using them, so there’s nothing to check there — if there are any mistakes there, then they are there on the physical hardware as well and thus it’d still be accurate! As for them being available as PDF or not: it takes literally one minute to export a PDF out of any PCB-design software and yes, you can choose what part of the schematics you wish to export as well. All the software I’ve so far seen can actually export in dozens of different formats, including ones that are very easy to edit, if there’s any need to, and both Linux and Windows have plenty of ways of then spitting out a PDF out of the most common file-formats.

        Whether someone else would have to upload the file is a different matter, but….just how long does it take to upload a small file?

        1. I don’t buy it either. The pro pi people are really sounding like any google or apple fanboys at this point. Nothing the great pi foundation do can be wrong. Everything is wrong with the people who don’t see the emperor’s clothes.

        2. No one is going to quit RPi train because any of the complains. For every project where customers demand documentation there will be more customers explaining why you can’t get it and it is fine for you.

          1. I already have quit the rpi train due to their pricing lies. I didn’t even need to read about their lies with documentation. You underestimate the will of the spurned.

        3. I work in an electronics company, it can genuinely take months if nobody has the time. And the fact is, one tiny detail like this is probably not that high up the priorities list.

      2. “someone to check it (both for accuracy and that there’s no details on there they can’t release)’

        You would hope that they could avoid changing any parts of the schematic that interface with the forbidden NDA bits. If Maya’s right, and it’s remove a resistor and add two to ground, it should be OK — no proprietary signals there.

        But imagine that they had a deal with Broadcomm that said B had to approve every publicly released schematic. Ugh.

        The fact that they can’t just flip over the schematic that they’re using for production is one of the hidden costs of closed-source design. But that’s the boat they’re in.

        1. And when you’ve got a million and one things to do, updating a schematic that can’t even be used as is to include a minor detail that the public already knows… geez, talk about low priority.

          They recently came out with a completely chip with massive amounts of high quality documentation and people were bitching about it like it killed their cat.

          1. Generating documentation such as a schematic is an output of the basic design process and necessary for design review, V&V, manufacturing and QC. If they cant even promise long term that it’ll be eventually publicly released I have to wonder about their internal design process and whether their documentation would meet industry compliance standards.

        2. What supprises me more, is that nobody shrugs and points tot the unofficial schematics …

          I mean, how long is it do desolder a board, and draw the lines? Turn it into a schematic, measure the values of the resistors, capacitors and inductors. Especially during covid, people where swimming in spare time sadly. Heck, I’m surprised no cheap Chinese reverse engineering company exist and does that for a 100 or 2.

          I guess nobody cares enough …

          As for the pi org, they don’t really care either. I mean, really, what propriatery information is there, when it is so trivial to reverse engineer …

          And what is broadcom afraid of? To have people learn their shitty hardware requires some really quricky designs? Come on. Again, anybody who actually cares has a schematic a week after release, for the cost of 35 + sandpaper, and some patience …

          This is typically why we can’t have nice things …

      3. The reality is that for the financial year ending 31 December 2019, The Foundation’s income was £45,790,187 ($64M+ USD). The reality is that they have spun off the Raspberry Pi (Trading) Ltd company as a subsidiary, and their finances are not subject to the same disclosures as The Foundation (the “charity”).

        If you’re a fan of their policies wrt the disclosure of a schematic, that’s fine – but let’s not be naive or disingenuous, please.

  4. I’ve been in hackerspaces around Europe and met some hardcore free software enthusiasts who didn’t know the ESP ecosystem is built around a binary blob. These platforms do have an openness aura, unfortunately.

  5. I’d happily update their docs for them for free if they’d just sell me their products at advertised prices. Until then I’ll never buy another one, and I’ll tell others how they don’t ensure that their advertised prices are honored by their official resellers. I don’t like being cheated by corporations, even non profit ones.

    1. John, you sound very ‘hard done by’ – Make yourself a little happier and spend an hour having a nose around the planet….knock up a list of prices for the Raspberry Pi product(s) you have issues with. Come back here and lets discuss further. It’s very rare for a $RRP product pricing to translate perfectly over the globe. has the Pi 4 4GB for 4800 JPY which is $43.29 USD……..that’s quite cheap…….PLEASE tell me where i’m going wrong?

        1. Amazon is not one of their official resellers.

          Did you try either of the 2 links they have in their website for JP resellers? I did not, just asking.

          These 2 are listed for “home” use


          These 2 for “Industry”


          Here in the US, I have never paid more than the advertised $35/pi when purchasing just the PI with nothing else.

          I don’t know what the stocks are recently as I have not ordered one since last year and it was purchased in a real brick and mortar store.

          1. lets look at your 110JPY comment – that’s Yesterday’s rate…it’s actually nearer 111 JPY right now to buy 1USD….Just this past 6 months, the currency has flucuated from 102.68 to 110.82…….that’s close to a 7% change. and, it’s a change for the worse for you… means, by default, anyhing priced in USD will be 7% more expensive! Here in the UK, the Pico is £3.60…that’s over $5USD!!!….do I feel ripped off enough to vehemently lambast a company for macroeconomic currency and laour issues?

        2. I have posted a comment above (assuming it gets past the moderator) showing pico prices and 1 source of Pi 4B 8GB prices. The site you have linked to here has the 8gb pi at 10,300 yen. That’s even more than what I quoted above. You’re not doing anything wrong, you’re just starting to get my point. Pi products are ridiculously priced in Japan. What really gets me angry is that the Pico is made in Japan by Sony and it’s cheaper after going across the pacific and then on to the customer than if they just ship it to me locally. Most Pi products were never officially released in Japan and had to be imported by a 3rd party.

          1. Eh? Did you actually convert that JPY price to USD? Using current exchange rates?
            Superficially it seems that stuff is quite a bit cheaper than Europe , UK and even some Australian prices?

          2. Ah, was too quick to type.

            Looks like you’ve done your research. BUT, you need to look GLOBAL here. Is ANY territory, other than the US selling at $4?

            Answer, doubt it. You’re discovering globalization and localisation.
            Prices are set with longer term currency trends in place.

  6. What I’d love to know is how there are still “unfixed” boards in the supply chain more than a year after the USB-C fix revision was released.

    I purchased a bunch of Pi 4s from Vilros earlier this year for a work project and was wondering why I had problems powering them up – they were old-rev boards with the USB-C bug and most of my USB-C cables are higher quality e-marked ones!

  7. My original issue I opened about this back in 2020 (after the updated version was known to be in the wild) also got closed recently due to their planned removal of things like schematics from the documentation ( so they have been aware of the lack of corrected reduced schematics for some time.

    Given there have been no schematics for the Pi 400 either (given it has changes to the gipo power setup that are not disclosed elsewhere they would have been usefull) I can see the scematics being silently dropped during this documentation rewrite.

      1. But we’re talking about a board schematic – not the fab details for the silicon. The firmware is not open-source. The “software” is Debian and the Linux kernel – the license agreement for both requires the modifications to it be open.

  8. I’m very happy to see how many comments this article has generated. I’ve felt for some time that “The Foundation” has cultivated an attitude more closely resembling a bureaucracy than a company dedicated to serving their customers. I found myself “banned for life” from their forum due to a [post I made]( in 2016. I gathered from this experience that external criticism is frowned upon by The Foundation… and yes, I do occasionally hold strong opinions :)

    I became interested in the nature of the Raspberry Pi organization several years ago. I did some research. and posted my findings and opinions on my [GitHub site]( in Feb 2020 if you’re interested. Two of the more interesting items from this research were the [UK Government’s Charity Commission website](, and the Wikipedia article. The Foundation is a far more complex organization than I imagined.

    This doesn’t explain the paucity of documentation of course, but calls into question what it means to be a “Charity” with a [*self-stated* mission of computing education]( I would like to hear from “The Foundation” on this point: “How do you square your self-stated mission of computing education under the charter of a Charitable Organization with your policies of closed-source hardware, firmware and withholding documentation?”

    And then there’s this from the Wikipedia article – apparently a quote from Eben Upton in 2012:

    “[T]he lack of programmable hardware for children – the sort of hardware we used to have in the 1980s – is undermining the supply of eighteen-year-olds who know how to program, so that’s a problem for universities, and then it’s undermining the supply of 21 year olds who know how to program, and that’s causing problems for industry.”

    As an accomplished engineer, and an employee of Broadcom, I wonder if Mr. Upton believes that the children he feels he is educating are best served by having no knowledge of the hardware they program?

    And I’ll close with this: It’s easy enough to understand the need to protect one’s “intellectual property”. But a schematic for a printed circuit board seems a particularly weak spot to draw a line. Does The Foundation imagine that withholding this schematic is effective protection?

  9. I have avoided Raspi for many years due to lack of fully open documentation and the whole Broadcom business. They could move to another chip vendor and a more open scheme if they wanted to. The current situation doesn’t meet my needs, but they seem to sell lots of units to other people.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.