16 November 2011

Boogie Board Rip - Review

Perhaps I should have titled it "The Tale Of Two Tablets".  A few months ago I picked up myself and Asus Eee Note, the main selling point to me was the integrated Wacom digitiser making me believe this could be a useful sketching device.  Unfortunately the software on it was a custom operating system based on Qt (framebuffer) and was pretty deficient from a sketching point of view.  So I offloaded the thing off on eBay....

Then I heard about the new version of the Boogie Board which is less of a tablet, more of a paper replacement, but it would have a saving capability for pictures.  All for the price of £100 + VAT.  The Boogie Board RIP has a 9.5 inch area for drawing on and can even capture those renderings. The way the display on previous boogie boards worked was impressive in that it only consumed power to blank the display, and putting pressure on the display caused an image to be formed.

I did wonder how on earth they would have taken that technology and been able to capture the images.  But the guys at Improv Electronics are much smarter than that, it looks like the way it works is there is a separate digitiser to the screen, so it is kind of like a graphics tablet where you can see the image you are also drawing.  You have to wake the digitiser to start using it and press the Save button to store the file.  The digitiser stores the file in a vector-based PDF leading to some quite nice images.

The primary reason to get hold of it was for sketching and note-taking.  Also, I wanted to actually be able to use the device on Linux, and since when you connect it, it just appears as a mass storage device meaning a win there compared to other connection mechanisms (like in the Eee Note).

The device itself is light and flimsy in the best possible way, making it easy to hold and also you don't worry too much about protecting the device rather than actually using it.  The battery apparently lasts for a week on a charge, but I have no idea the amount of charge my device is currently holding because it has no indicator.  This is not necessarily a problem because the device is ludicrously stripped down and simple, so I see it has having an incredibly simple UI.

The only criticism I have about the device is more subtle than most people would notice.  Because the digitiser and display are separate devices they seem to work at different resolutions.  The digitiser is highly accurate by the look of it generating cleaner and thinner lines, but the reactive display generates thicker less accurate lines.  This may also be due to the design of the pen nib.  The pen itself is required to capture input, and I found it comfortable to use.

Overall the Boogie Board Rip is a cool little device good at the job it was designed to do.  I'm still trying to work out the best way it will be my digital sketchpad but it is the device I have seen get closest to what I want (in fact if they did a really large A3 version that would be awesome).  The guys at Improv Electronics have done an excellent job.

You can buy them in the UK from Improv Electronics for £120 including VAT.

This is a picture comparing the PDF output to theactual display - you can see finer lines on the image compared to the screen where some lines bleed into one another.

And some quick two minute sketches of Rodimus and Megatron.

20 October 2011

OpenELEC - New Distro For Media Centres

I saw that a project called OpenELEC (Open Embedded Linux Entertainment Centre) got released released recently.  What is interesting about it is not so much the fact that it is a fast booting XBMC distribution, but the fact that it is an embedded Linux system for x86 platforms with minimal size and uses the latest components.

I can vouch for the most important thing when dealig with media is having the latest graphics dependencies on Linux.  You tend not to have long-lived stable releases and most issues and advances are done in the day-to-day development.

The source code is available on Github here.  Going through I can see it has Python, GTK+, and Python gobject bindings, Linux 3.1rc10 kernel and lots more.  Even though it is a cut-down minimal system it doesn't lack handy things like ssh and a text editor.

The final images are a mere 100MB or there about in size.  There are scripts to create the USB install key and it even has an automated update system.  There are optimised builds for ION and AMD Fusion as well as a few other platforms.

If you want an XBMC computer then I can think of more heavyweight ways to get it, but if you want it to boot in 10 seconds and you don't need to cusomise it too much then this would be a really good choice.

To me it looks like an ideal platform for trying out different media type applications (say kiosks or other media centres) with not a huge amount of code as the base system looks pretty solid.  I reckon something like Clutter's Media Explorer could be quite good on it.

26 September 2011

PyCon UK 2011 - Great Conference

Now with EuroPython being held outside the UK, the PyCon UK takes on more significance.  Run by the same guys and now relocated to Coventry this year was a triumph.  Something like 200 people attended for the weekend log conference which is pretty impressive.

The venue was excellent.  The most important aspect was all meals were supplied at the venue (and were all very tasty).  This meant that people tended not to drift off and hopefully that helped the mingling.  Some of the rooms were a little small but nobody seemed to forsee how well attended the weekend would be with lots of last minute tickets.

The topic that dominated the entire conference was computing and education.  Specifically the useless nature of ICT courses and the complete lack of the next generation of coders coming through.  The obvious technology on show and being discussed was RaspberryPi.  There was also the BBC CodeLab host attempting to evoke those old feelings of the BBC Model B.  Basically the conclusion comes down to there is no will in government for changing the nature of computing in schools and it will require hackers in number making the changes piece by piece themselves.

The Coding Dojo was a lot of fun using the task of Maze Generation.  We managed to lash enough together five minutes form the end and I think we surprised ourselves by it working.  Also the web development open forum was good, maybe I spoke too much in that....  and I managed to get a demo of Camvine in there.  There was also three thought-provoking and entertaining keynotes.

As always the people there made the conference.  The team who ran it all did a sterling job, even better than their previous EuroPythons.  Team 4 in the Code Dojo did a great bit of work, Matt, Safe, Adrian, and Menno.  I also got to speak to the keynote speaker Allison Randal at the meal, in fact the whole table was entertaining.

So my recommendation, even if you don't use Python much, is to go to PyCon UK next year because I am pretty sure you will learn something or meet some really interesting people.

28 August 2011

Raspberry Pi - The Trojan Horse?

The Raspberry Pi project has got a lot of coverage today thanks to the article in the Guardian written by John Naughton.

The fundamental problem has never been the lack of cheap hardware. Even in this day and age it is not the lack of quality freely available tools.  A kid who has the propensity to be a hacker nowadays would invariably have Linux installed with the myriad of tools available, and also doing that they would have the knowledge it should work fairly easy on other computers running roughly the same operating system.

The Raspberry Pi is attempting to invoke the feeling of the BBC Micro in an attempt to get kids programming again, because in the near future we are going to be in a very precarious situation with our engineering ability in this country.

First of all why did the BBC Micro succeed? It is very hard to compare to today with the abundance of computing hardware just in your pocket.  Part of it was almost every classroom across the land had one of these and the children who could work out how to operate them hunted 1s and 0s in packs.  There was a communal aspect to it as the children would have to share, listen, and take it in turns.  Also, there was the TV programme Micro Live, which was needed to get a chunk of publicity.

My first programming was done on BBC Micros at school in the East End of London and at friends houses as we couldn't afford one at home. So, I for one know their importance and probably a majority of my friends and colleagues had similar start.

When I was at University Linux was just starting to gain some traction and the work of the FSF got us GCC and the mix of other tools.  But on Windows I was still trying to use DJGPP and there was no chance of getting more complex things working with an open (free and Free) toolchain.  Children coming up in that age must consider it the computing Dark Ages.

So I come back to my first statement that cheap hardware has never been the problem.  The people getting excited about it are 30+ developers like me who think - "Oooooh, cheap commodity ARM platform, that I can hack in lots of ways".  My immediate thought was to make it into an immensely cheap network connected digital signage system.  And this is probably because ARM is still a minefield and a static and well supported platform would be desirable.

I am not trying to negate the good work done by Raspberry Pi as it could be amazing if it is the Trojan Horse.  What do I mean by that?  The hardware becomes the makeweight for all the real important stuff that needs to happen if we are truly going to educate the next generation of programmers.

What is needed?  Community, publicity, documentation and traction.  Community because this goes beyond the single classroom of my youth.  You need IRC, social networking and tools for more permanent sharing (like Github and Gist).  This then ties into documentation, it would need to be good up front but also allow for the people to evolve it, so much like a wiki.

Just social media publicity would not be enough, you would really need a TV programme to generate it, but then this is where the social media would tie in.  For the programme they would have to show the users best contributions so there is an incentive to produce and produce something good.  After all how many people sent a picture into Hartbeat in the hope of getting into the gallery? It's the closest computing could get to the X-Factor...

The most important part will always be changing the educational system to get children learning how computers to make computers work for them rather than simply going through the commercial Word Processor/Spreadsheet operating.  I suppose this is where the documentation does come in, but I think there will need to be more.  If someone could show educators a curriculum that they could prove worked then that might get the traction required.  That would be a tough proposition and also I am unsure about whether there are enough technical skills in the educational sector to deploy such a thing.  Maybe it would require people from industry coming in and mentoring...

In the end I am hoping that the Raspberry Pi hardware is the Trojan Horse for something much more fundamental.

26 August 2011

Linux at 20

If it didn't exist I somehow think computing would be a much less fun place and I probably wouldn't be able to do my job so well.

Linux Foundation
I'll be celebrating 20 years of Linux with
The Linux Foundation!

24 August 2011

Pycon UK 2011

Unfortunately I was unable to make it to EuroPython 2011 due to its location (no longer being in Birmingham).  The good news is that the same guys who ran Europython in Birmingham have announced Pycon UK in Coventry.

For the reasonable price of £95 you get two days of Python goodness.  As their website says:

Sessions will include testing, a code clinic, Django, PyPy, XML, Design and Graphical Programming. There will be an unscheduled room for dynamic talks and time allocated for sprints. The weekend will also feature the return of our ever boisterous conference dinner!
I've paid for my ticket already and Python events are always educational and entertaining.

23 August 2011

New Github Project - pyxlib-ctypes

I had written a pyrex based binding to some elements of Xlib, but I decided to make it pure python to avoid the compilation step.  Luckily a kind person had created pyxlib-ctypes http://code.google.com/p/pyxlib-ctypes/ which is a pure Python binding to the underlying X functions.

I needed a few more features due to XComposite for writing crazy desktop compositors in pure Python and also some XFixes functionality so we can really hide the mouse cursor.

As such I am publishing my amendments to it on Github - you can check it out here: https://github.com/garrybodsworth/pyxlib-ctypes it literally has the bare minimum I needed and could probably be extended for greater coverage.  It is used in production and seems to be working very happily.


It's aliiiiiive!

After a few years of hosting my own blog on http://programmerslog.com I have come back to Blogger.  This unfortunately was caused by losing the domain which is pretty sad....  I have a back up of the few hundred blog posts in a SQL dump but I doubt I will find the energy to extract them(!)

What will be on this blog now?  Any projects I publish on Github, lots of Python, Linux, coding, and probably my random artistic scribblings.

I must say Blogger is looking rather swish nowadays, part of the reason to use Wordpress was for the features, but now I can do pretty much what I want here.  Also I don't have to worry about renewing my domain subscriptions or doing software updates...

I've reconfigured all my feeds to point to this blog since it is the only active one.

Now I just have to actually write some content...