$25-35 Raspberry Pi Computer

RaspberryPi computer

RaspberryPi computer

For about $25 to $35 ( I’ve seen both stated, but not checked the order page. FAQ says Model A is $25 while B is $35) you can have a Linux computer for all those day to day tasks; like email, web surfing, spreadsheets, running and analyzing GIStemp ;-)

I’ll be getting one. As I already have all the peripheral bits it wants, all I need to do is “plug and go”. Nice.

Web site here:


FAQ shows a card layout. Looks like RCA or HDMI video out. ARM CPU with 256 MB memory (quite enough for Linux). LAN on board. USB for disk et all. SD card slot where I’d put the OS and related. (Swap on the disk, if needed).

Gotta love it. A “disposable” PC as needed…

Video of one set up, with close up of all the cables plugged into the little board… ;-)

Subscribe to feed

About E.M.Smith

A technical managerial sort interested in things from Stonehenge to computer science. My present "hot buttons' are the mythology of Climate Change and ancient metrology; but things change...
This entry was posted in Tech Bits and tagged , , , , . Bookmark the permalink.

18 Responses to $25-35 Raspberry Pi Computer

  1. R. de Haan says:

    E.M, I have been following this development since last year and ordered one last week.
    Still waiting for delivery.

    Not entirely unrelated but interesting:

    And a nice initiative to promote new business idea’s

  2. sandy mcclintock says:

    I checked this out some months ago but it appears it was ‘not ready yet’.
    Is it really available now ?

  3. Chuckles says:

    ‘I’ll be getting one.’
    Gotta love the optimists. A lot of people feel the same way you do E.M., but don’t hold your breath in anticipation. We’re talking ‘Good British Ingenuity’ here. It’s made by a ‘Charity’ for starters.

    You are guilty of thinking like an American – ‘Hey, neat product, I wouldn’t mind one of those. Here’s my money, sell me one.’

    Nope, not a chance of that. The product is so enmeshed with what looks like noble intentions, naive incompetence, and similar, that tis best to just shake the head and walk away till they hopefully sort things out
    Apart from wierd target markets, unanticipated demand, wrong networking components, failure to anticipate having to get FCC and CE Mark testing done, etc tec, they’re doing fine.

    Rasberry flavour –


  4. R. de Haan says:

    Chuckles says:
    22 April 2012 at 12:32 pm

    I think consumers all over the world think like Americans.
    They want a product ASAP if they like it or need it.

    Try to invoke a gasoline embargo in Iran and you will see burning fuel stations.
    If it’ true that the Raspberry is the brain child of a bunch of do gooders, well good luck with that.
    Better make one yourself.

    Products that are announced like the flying car or the gigantic list of so called green products able to replace oil but never make it to the market has become a trend.

    They do it to get their hands on subsidies and capital but also to put the consumer on the wrong leg crating the illusion oil can be replaced now if we accept certain legislation.

    So I am still waiting for that high capacity battery that fits the size of a fridge able to power an entire house including the washing machine and the dish washer and comes at a price of USD 2500,-.

    Powering one house at a time was their slogan but today we’re still waiting for the first house they power with their battery.

    Similar hubris from the field of carbon capture and cold fusion.
    Much of this silly talk comes from university laboratories and the best way to handle this is to ignore
    their “accomplishments” and go on with your life. Especially if the scientist involved are italian and the product development is called “revolutionary”.

  5. LG says:

    E.M. , You might want to also consider beagleboard.

  6. Chuckles says:

    R de H, Yes I agree, pity the RP Foundation doesn’t though. They have enormous interest and demand, people like all of us keen to get a feel for their product, 350 000 units ordered. But can they ship any? Mope. Instead they witter on about Silicon roundabouts and centres of tech excellence, and drinking wine with politicians. I think I’ll pass thanks, tired of waiting for this lot. Amateurs are too unpredictable..


  7. j ferguson says:

    I’ve been building gadgets with Arduino PIC computers. Arduinos cost $15 tor $40 depending on how many ports you need and seem very good for taking in signals from different devices and letting you display data on LCD’s which is what I’m interested in. You code in C or C++ . The IDE is very nice. I’m almost done with my shipboard annunciator project which will display present position (lat,lon), heading, depth, windspeed, and true wind direction, and bearing and distance to last fix. It will be installed over my head in the berth on the boat we live on. It has 4 buttons (turned out I only needed 3) which turn it on and off, set the LCD backlight level and let me reset the fix – useful in detecting dragging at anchor. next one will be for reporting engine data to the helm, and then maybe some others so far unthought of.

    At first i thought the raspberry would be fun to use as a base for a recordplayer – load it up with mp3’s and play them through our sound system, but then I saw the HDMI video. There is a bit of a sarcastic remark about VGA on their FAQ’s page, but what they have done is make their product impractical for applications requiring a small (inexpensive) screen and simple key input.

    There are other small computers which can be used for this purpose but are more expensive.

    i hark back to my Sinclair Z80 (was it?) which I built and programmed to be a rolodex which it did with it’s tactile keyboard and the little tv set for some years after. I wonder what I did with it.

  8. E.M.Smith says:


    Looks like a fun toy, too; especially for ‘network servers’. Didn’t see a lot of video / audio, though, so a bit rough to read the email or listen to tunes ;-)

    @Sandy Mcclintock:

    Reports are that it is now shipping (which is why I mentioned it at all. I don’t pay much attention to ‘vapor ware’ and we’ve got the videos from recent ‘unboxings’…


    I’ve ‘been around the block’ which is why there no mention from me until AFTER they were shipping units. So they are now “real”, it’s just a question of volume… and wait time. Personally, I’ll not be ordering one until after the “first rev”. I never order a product with a odd first rev number or an even minor digit. 1.0 is beta not good. 1.1 is slightly better, first bug fixes in. 2.0 is a major upgrade, likely usable now, but with new bugs. 2.1 is first bug fix of them… I’m especially fond of releases like 2.5 and 4.7 ;-)

    So “I’ll be getting one” ought not be construed as “My check is in the mail” but rather “it’s on the watch and shop when mature” list…

    As a “Linux Lover” I’m not bothered by “Noble reasons” or “Amateur standing”. The whole of the Free Software movement is rooted in that ( and Red Hat too. I worked at the precursor company making the compiler tool chain decades back… Yes a company with paid employees that had a free product… that worked very very well.)

    The general idea of the product is straightforward, the ARM chip is very well proven (we were investigating back when I was at Apple and it was impressive then. The design competitions won since then and the advancements are rather impressive). As the software is Linux (also known to work on ARM for some time) I’m not worried about that part. The hardware is all COTS chips and connectors. Frankly, I think it would be hard to make it NOT work.

    But I’ll still be waiting for the 2.1 release ;-)

    @J. Ferguson:

    I’ve got a Sinclare kicking around here somewhere too…. Maybe I’ll dig it out ;-)

    The reason I like the RaspberryPi is the same one you do not. Different needs. There are a load of ‘device controller’ boards for a cheap LCD and a couple of buttons. “Juicing one up” to run a screen, network, and disk is a PITA. So I’ve been waiting for years for someone to make an “embedded controller” like board, but with built in disk and video support. (Not watching closely though, so don’t know if these guys are first or not.)

    What I want is a ‘disposable computer’.

    Why? Well, first off, because computers die. I’d rather not care when it did.

    Secondly, as the “authorities” have begun making it SOP to steal your computers if they every don’t like you or suspect you might have information about someone else, I’m of the opinion that my data needs to live “out there” somewhere (encrypted, of course) and the “computer” be entirely “fryable”. Someone says they want my computer, I just “turn on the light switch” and they can have the charred remains… I’m out $35 and get about $100 worth of entertainment ;-) They leave, I open the box on the shelf and “build a new computer” in about 5 minutes…

    Frankly, leave a few miscellaneous odd network cards and interface cards on the “workbench” and you could likely hand them an empty $20 Ebay Laptop and they would never think that “some card” was your computer… Everyone KNOWS what a computer looks like ;-)

    There are also times I’d like to have “special purpose” computers. A DNS server, for example. Right now I do that with old White Box PCs that are a few cubic feet and with a noisy fan. I’d rather replace them with a small ‘cigar box’ of a couple of cards…

    Finally, I have a minor passion for Beowulf Clusters. It would be rather easy to put a dozen of these into a very tiny shoebox and have a neat cluster machine for stuff like, oh, climate models ;-) Have a network switch for the backbone, only plug in a monitor and USB keyboard / mouse for configuration. So just plug in the network and power and you are done.

    Poor mans “supercomputer”… I used to have a Beowulf of about 6 to 8 nodes (varied by date) but I’ve dumped a few of the older ones and only have 2 nodes left (and not enough space to really set it up anyway…) So sometimes I ponder a “mini-Beowulf”… Having a tiny shoebox would fit on my clutter desk. Put GHCN on one of the SD cards and I’m “good to go” for a variety of temperature inspections. Even a full blown climate model.

    If I want “animated video” out, either via the network, or plug in a dedicated monitor if needed.

    But, as you pointed out, for an “embedded controller” type applications, it’s quite wrong. Needs LCD support and a ‘button / pad’ input. A few software controlled switches would be nice too…

  9. E.M.Smith says:


    OK, read the review in your link. About what I’d expected. Some lack of particular bits of software ( use of hardware animation engine, Flash support). Typical for an early Linux device. The general conclusion at the end is they would recommend buying one for a person like me. Well, “I’m good with that.”

    Frankly, that the thing basically worked out of the box and even when they overclocked it is a surprise. I’d have expected a couple of more ‘bugs’ to have been surfaced. In essence, it works as advertised (though with some jerky web pages on high graphics loads, likely to be fixed when the onboard graphics hardware gets better software support and is actually used… ) For 90% of what I do, that would be “just fine” and as most of the time I find Flash a PITA (mostly shoving advertising tongue down my throat) having it turned off is mostly a feature to me.

    My estimate is that in about a year it will be a highly competent general purpose Linux based desktop “box”. Until then it would still do about 80% of the stuff I’d like to use it for, only being an “issue” on high graphs weight code, Flash, and stuff like GIMP that could be slow. All that is FINE with me.

  10. j ferguson says:

    I figured out how to make a jukebox with an arduino and it would hold the 16 gigs of mp3s i listen to with any frequency, and selections could be made off of a 4X20 lcd, but something like the raspberry would be a lot easier to set up. I may try the arduino jukebox anyway just to see what sorts of problems i run into.

    to expose some more of my ignorance, friend bought a big (BIG) Samsung 6400 series flatscreen. It is wi-fi enabled. His idea was to show slides from his Ipad on the Samsung via wi-fi.

    he couldn’t get it to do this, thought there might be an app, but couldn’t find one. he then discovered he could do it if he connected via a cable. my guess is that this meant the tv had become a monitor which is what I sometimes use our big Sony for. DVD’s play on the hp laptop and the movies show up on the screen which thinks it’s a monitor. but of course there’s a cable.

    so maybe the trick is a very small computer like the raspberry running a remote console program so that what shows on TV/monitor is what’s running on the Ipad. And Look Ma, no wires.

    I must say that I really like the idea of the little innocent looking wall-wart computers. There they sit plugged into the wall and each one with 32 gigs of encrypted data (all the real stuff). My guess is my stuff wouldn’t need anywhere near that much room.

    learning C has been fun. I used to be able to write unix scripts, dbase 2 “code” and Basic off the top of my head. I’m getting there with C for the limited things I’m trying to do. Writing and building hardware to run it on simultaneously is a new experience. If you hadn’t thought about it, you’d never guess how many ways a momentary contact switch can be programmed to turn on the lcd backlight and then adjust the brightness. after all sorts of chimping (Jamaican term meaning dragging your knuckles on the ground) I realized that all i needed was a couple of “whiles” to recognize that i was holding the button down and do the delays, then turn lcd on – full bright and step down the 4 levels at one second intervals and then turn backlight off. it would hold any setting live when i took finger off button. this may not seem like a big deal to anyone else, but you’d never guess all the other schemes i considered before coming to this one. I’ve still got two more buttons to sort out and a half dozen so far identified things to be selected. No wonder so many projects aren’t much different from the one before.

    One other discovery, maybe a bit less pleasant is that at 69, I may not be as sharp as i was 30 years ago. It looks like doing a “while not” requires a lot more concentration to avoid confusion than I remember. My code was always well (maybe not well, but extensively) commented but always drew snickers when someone who actually knew what he as doing looked at it. Almost always it could be condensed to 1/2 the number of lines I’d needed. And sometimes it didn’t work the way i thought even though it looked like it worked.

    you can buy little vga screens for less than $100. They seem to be offshoots of the auto market onboard systems. other low cost small pcs (mini-itx) types which a lot of people build car computers with cost over $100 for the board and more for the rest of the stuff. it’s hard to put a tiny system together for less than $400. of course i could buy an Ipod, but what would I learn?

    I hope the raspberry people get to 2.2 really soon so you can tell us about it.

    have fun, john

  11. E.M.Smith says:

    @J Ferguson:

    NEVER feel bad because someone else can improve your code. Remember the Law Of Mutual Superiority:

    Anything you program, I can improve and anything I program, you can improve.

    Comments are your friend. I’ve written code with ‘way over the top’ comments and code with none or few. A decade later I’ve gone back to my code and found that the comments were very helpful to ME as that particular ‘trick’ I’d figured out was long gone from the top of my personal stack.

    I now tend to comment in ‘blocks’ just a bit away from the active code blocks, so there’s not a lot of interference between the comment bits and the active bits. So the “While” loop is commented just before it starts:

    C This bit of code counts down the timer until comment “keyup” happens
    While cdtimer do…

    BTW, I don’t think its being ‘less swift’ with age. There’s an interesting aspect of the brain. The theory is that anything not ‘touched on’ in memory inside about 5 years gets slowly degraded. That means you either need to use it to keep it fresh, or you get to relearn it if over 5 years.

    I “rediscovered” this doing the FORTRAN of GIStemp. Once upon a time I was a pretty good FORTRAN programmer. That was way longer than 5 years prior and about a dozen languages. (From Pascal to ADA to HP Business Basic – a Pascal like language – to a “few” database languages to Unix scripting – several – to C to…) When I first looked at the FORTRAN my thought was “What IS this?”… after a week or two of re-reading bits of the manual and practice, I was “back up to speed”. Compare that to 10 weeks for my first FORTRAN class. Much further much faster. Then the code itself was rather hideously complicated for no good reason at times. Still got it sorted out.

    So just realize that you need to make a particular tool to use it and it may take you a week or two to remake one you had a few decades back; and that is perfectly normal for folks of all ages. I tossed some Italian at my son, who took it in college as his language requirement. He informed me that Italian had not “stuck” very well… despite the A in it. It had been almost 5 years…

    Per “ways to measure a switch”: Yeah, know the problem…

    Or worse…

    Once I was writing a module for a Medical Accounts Receivable package. The product was already widely sold. I was hired to “do maintenance” on it. Turned out that included building the parts the first guys had not gotten around to doing. One of which was the “roll up and purge” old data. Also turns out a couple of customers were reaching absolute maximum size on the database on that particular platform. ( HP3000, Image database, Query report writer, HP Business Basic for the package) HP B.Basic is an odd language. BASIC written by a frustrated Pascal jockey. IF THEN DO ELSE DO; BEGIN END; and all… Including database calls.

    Well, long story short, it included recursive subroutines and function calls. I worked out a ‘way cool’ trick way of doing the search down account history, then rolling it all up, writing the archive, updating the database with the new summary record, and return from the end of the recursion chain… Worked perfectly on all the live client “test data” I was given.

    Went to the customer site for the OTHER live client who was approaching max size (and had about a month before he would be shut down in his medical practice from database lockup…)

    Installed the software. Fired it up. On about the 5 th customer / patient record, their computer started throwing “Out Of Memory” errors. Seems THEY had a Medicare rich practice with folks with long medical histories… So long, in fact (unlike ANY other customer or ANY of the test data I’d been given) that a recursive approach, holding each record as it went to discover the next, would run out of memory long before completion and return from recursion… We are talking years of weekly or daily doctor visits, sometimes with a dozen or three different procedure codes per visit.

    OK, now I’ve got maybe 3 weeks to rewrite something I’d spent a couple of months on. AND a very worried customer. AND a boss who doesn’t understand what I’m saying…

    I spent about a week of long days looking at the code and trying to completely re-figure my approach. About day 4 or 5 (its a bit of a blur) I figured a way to re-hack the existing code, take out the recursive aspect with some hideous use of BEGIN END and WHILE DO and a couple of GOTO and crap wrapped around the bulk of the code plus some strange added variables, and it worked. Everyone was quite happy. I then spent about another week redoing some the comments and adding some bits to explain WHY it had this crazy structure…

    Now IF you had asked me “Can it be done THIS way?” at the start, I’d likely have said “I don’t think so, and why in hades would anyone want to?”. But, in the moment, it was a Godsend to have that approach come to mind. I can only hope nobody EVER was asked to work on that bit of code later… And if they did, I’m sure they appreciated the very honest comments…

    There are hundreds of ways to do most things. The most efficient is often not the most clear or reliable, or least bug prone. The most elegant is sometimes the least robust (or in my case, least tolerant of variations in data or machine memory size). It’s choosing between them that matters.

    Since then, I’ve written a LOT of pedantic inelegant code … that is very robust, very bug proof and just works. Often I have a very elegant alternative approach, that pondering a bit shows to be vulnerable to some odd edge case of data size, machine issue, whatever.

    Part of my complaint about a lot of the current approach to programming is that it is not nearly as interested in reliable operation, and far too interested in “cool” and “new”. I’d rather have deterministic operation, than cool but sometimes breaks. Yet the trend for the last few decades (basically after MicroSoft showed you could get away with it) is to toss in lots of semi-functional ‘features’ and then debug on the clients dime in future releases… Never really stable. Never really robust and reliable.

    So don’t fret about any particular bit of code being a bit “inelegant”. It most likely means you are avoiding some very ‘elegant’ and hard to figure out bugs, too.

    (OTOH, I occasionally will glory in some particular ‘way trick’ bit of code… The old roomy who worked for HP wrote diagnostic microcode for the processors. You never knew what hardware worked when it was run, as it was run when SOMETHING was broken, by definition. One CPU had 64 registers. They had a ‘step by’ feature to the program counter. He wrote a bit of code that you could enter at the top, it would ‘increment by step 1’ testing each of 32 registers, then increment by ‘step -1’ and run BACKWARDS through the same lines of code testing the next set of 32 registers. Cut the code size in half (so half as many chances it would land in broken hardware and not execute). But Wait! There’s more! You could enter the code at the bottom and run upward by step -1 first, then it would turn around and run back down… So you could enter this block of code from either end and it would run, then branch to the next operation.

    Now THAT’S elegant. Runs in either direction, entered from top or bottom, takes 1/2 the space (and can be loaded in either half). Now you can load in top, and test, and if it fails, load in bottom and test… all with the SAME bit of code.

    Or run from top to bottom, note where the fail happens, then run from bottom to top and note where the fail happens, and have your diagnostic… even if more than one register is broken.

    He was very proud of that bit of code…

  12. E.M.Smith says:


    The wiki on Beagleboard answers one of my concerns. Looks like they have a version with ethernet:

    A modified version of the BeagleBoard called the BeagleBoard-xM started shipping on August 27, 2010. The BeagleBoard-xM measures in at 82.55 by 82.55 mm and has a faster CPU core (clocked at 1 GHz compared to the 720 MHz of the BeagleBoard), more RAM (512 MB compared to 256 MB), onboard Ethernet jack, and 4 port USB hub. The BeagleBoard-xM lacks the on board NAND and requires the memory and OS to be stored on to a microSD card. The addition of the Camera port to the -xM provides a simple way of importing video via Leopard Board cameras.

    At $89 each, though, it’s a mite more expensive… Though it looks to have been out longer and be more stable already.

  13. Sandy McClintock says:

    I tried to place an order. I got the following back :( ….
    Thank you for joining the Raspberry Pi revolution and registering your interest in Raspberry Pi’s Model B board from RS Components.

    We have received extraordinary levels of demand for this product. To help ensure as many people as possible can experience the Raspberry Pi concept, we are initially limiting boards to one per customer, and we will send you regular updates on availability. As boards arrive into stock, they will be allocated on a first-come first-serve basis, in order of when requests were received.

    Thank you for your patience; we will be in touch as soon as possible with more details.

  14. A. C. Osborn says:

    I think this was developed by a UK Teacher to teach his pupils how computers work, rather than just pressing the computer’s keys.

  15. adolfogiurfa says:

    @E.M. It will be also a HIT to make a computer WITHOUT programs, like in the good old days. Then it was a good thing for school kids to make them develop their logical thinking. Nowadays computers are a bad thing for schools.

  16. R. de Haan says:

    That’s the idea behind this little computer adolfo, education

  17. E.M.Smith says:

    @Sandy Mcclintock:

    Nice to know. I’d kind of figured they would be ‘underwater’ for a while from a flood of folks waiting “for it to ship” before placing orders.


    One of the benefits of a Linux box is the full code development environment. A friend is teaching robotics… he’s looking at a related board for teaching programming… but liked this one too.

Comments are closed.