educational release in 2013?

The Guy Who wrote the article (@Peter Zotov) is taking a smashing by the community.

http://whitequark.org/blog/2012/09/25/why-raspberry-pi-is-unsuitable-for-education/

Fergus Byrne wrote:

 

The Guy Who wrote the article (@Peter Zotov) is taking a smashing by the community.

http://whitequark.org/blog/2012/09/25/why-raspberry-pi-is-unsuitable-for-education/

I read that article a few days ago and my objection to it isn't the usual "Leave the Foundation alone" witterings, but rather that I just don't buy into the whole "everything should be open" argument. It just doesn't follow that having a completely open platform means that well designed, reliable applications will drop from the sky in a timely manner, like unicorn tears. If anything it takes the pressure off software developers because any criticism of a particular piece of software is usually greeted by howls of "If you don't like it then write a better one..." from elements within the community, which does nothing to get things fixed...

In an ideal world everything would be open and properly documented, but it isn't an ideal world and we've coped up 'til now. Anyway, Broadcom couldn't release all the specs under the hood even if they wanted to, because there will be third-party stuff in there subject to NDA. The closed model is fine if market forces are allowed to play out - this brings pressure to bear on the manufacturer to  e.g. write drivers that work properly. Unfortunately, when fanboys are willing to put up with half finished products the model breaks. Hundreds of thousands of Pi shipped and still no sign of a decent USB driver. I guess that we only have ourselves to blame...

Anyway, I think that we (naturally) tend to overthink the hardware part in the context of the educational environment. Surely the most important component is the curriculum itself and the support that the teachers will receive in order to enable them to do a great job of getting it across? The Pi may be closed, but it really doesn't matter in this context. It's not as if secondary school students are likely to be writing GPU drivers (although some will doubtless want to try, just as some will be busy hacking in to the ineveitable networks that admins will set up for their classrooms full of Pi - I can't see many admins wanting to keep storage local only. All those SD cards...)

Anyway, a 2013 educational rollout - 11 months 'til the start of next school year. Best get cracking!

Surely the most important component is the curriculum itself and the support that the teachers will receive in order to enable them to do a great job of getting it across?

Yes, but where is this curriculum ?   I see no signs that it exists or is being created.

It's not as if secondary school students are likely to be writing GPU drivers

That's the problem though isn't it. They should be. 

Back in the 70's and 80's the secondary school students were learning how to code on the systems available at the time, they've since become todays engineers and software developers. Why shouldn't todays secondary school students be doing the same ?

For me a large part of the goal would seem to be to stop our society (i.e. the 'grown-ups') from turning our kids into mindless consumer zombies. So it's as much about changing our attitudes as it is getting the kids into programming.  It's up to us to offer them the chance, if we don't do so it's us that are to blame.

Instead of assuming our kids are incapable of writing a GPU driver we should instead encourage them to take an interest and go try, or at least we should give them access to the same sort of opportunities as we had.

I do agree on the open vs closed stuff.. The Pi is open enough to be useful which is where the opportunity lies to get the kids interested.

@ selsinork: With the best will in the world I think that starting from scratch to writing a proper, grown-up  ARM GPU driver in two school years is probably a bit optimistic, it'll possibly take a whole term to get to grips with (or at least learn to tolerate the idiosyncrasies of) Linux! All these modern doo-dads are many orders of magnitude more complex than the Z80 based boxes that I battled with at college - if you took the lid off one of those you could almost see it working. You could certainly imagine the chatter through the i/o and empathise with the addressing and data paths and have a reasonable idea of what was going on in that Zilog chip. There was far less abstraction then and with hindsight it was almost too simple (although it didn't feel like it at the time, especially machine code...) The concept of 8 bits as a magnitude range could be grasped by the human imagination, but 32bit? And don't get me started on modern i/o like USB, HDMI, I2C and the like - you just have to accept that it works,,,

I do have a bit of a grin at how some other old timers do the "in my day" speech. I feel that we had it easy compared to today's teens! While the principles of creating an application from the ground up are exactly the same as ever they were (with the majority of the time spent not writing code!), the act of closing one's eyes and imagining those lines of instructions doing their thing will be far more abstract today as it's to a great extent happening inside an assembly of little black boxes.

I also imagine that given the Pi's price point we're looking at classrooms full of them, rather than just a few for the elite kids, so it's likely to be a solid grounding in development, coding and debugging, rather than esoterics. This is where the GPIO will be a winner as it allows stuff that can really fire the imagination of children who aren't already into being creative with computers, maths, design, problem solving, teamwork and all the other stuff that this could inspire.

Jonathan Garrish wrote:

 

I do have a bit of a grin at how some other old timers do the "in my day" speech. I feel that we had it easy compared to today's teens!

How true, how true.  My favorite example is the PDP-11 Peripherals Handbook "Programming" chapter which describes both wait loops and interrupt-driven I/O in just eight pages, including assembly language examples.  And if you bricked your PDP-8, all you had to do was toggle in the RIM loader which was about a dozen 12-bit words, IIRC.  Proper PDP-8 nerds could toggle in a RIM loader using muscle memory.

In contrast, GNU/Linux is like programming a mainframe, with mainframe-like software layers between you and the bare metal.

Jonathan, I really do agree with a lot of what you say.  I just feel that it's a permanently downward spiral if we continue along these lines.  We can't just keep assuming it's too hard for people to learn or we'll soon be in a position where nobody understands how anything works anymore. Give the kids a chance, they might just surprise us.

And as stuff like i2c is almost as old as those 8bit systems, you don't get to use it as an argument

It's funny though, I share your view that GPIO is the thing that will grab the kids imagination, but lots of others seem to think that GPIO will be a minority thing. It'll be interesting to see how it all plays out..

I do wonder if this is still too late though, and in effect this means 2014 before schools wil start getting them in any quantity ?

I would have thought that budgets for next year will start being finalised soon, and I don't think i'd commit budget to something that may or may not occur? This term is already started, so I doubt you could build this in to anything before at least September of 2013 ?

Steve