Group members: Craig, Ian, Iain, Jennifer, Richard (coordinator)

innovation.jpg

Oportet nos excitare, ad novationem

The INNOVATION point is the pivotal moment when talented and motivated people seek the opportunity to act on their ideas and dreams. - W. Arthur Porter

The Word on the Street is... innovation!

How would you teach the word "innovation" to a primary school kid?

Initial ideas circulated via email etc.

iainr:
I'd love to do something with telephony but that's not going to wash as IS won't allow us to hang non-approved stuff off the exchange.

Off the top of my head:

  • Start an IPv6 service.
  • See if we could run one of the touch panel machines off a raspberry pi.
  • Set up a pgp/gpg key exchange server.
  • Set up an openstep/openshift or GFS2 service .
  • Publish the output from the server component via RSS or a twitter like microblogging site (statusnet? or via irc. would ideally be something that could be ssl'd and had plenty of available clients.

idurkacz:

  • Another Raspberry Pi / graphics idea might be: put together a small purpose-specific device to drive a slideshow on either the drop-down screens in G.07, or the large 'cinema' screen in the Atrium.

    (That comes about from something I was asked to do for last Friday's (i.e. Friday Nov 16 2012) large 'Alumni' event held here, but couldn't easily arrange. Currently, if you want to display to either of those screens, you need to drive the displays from a laptop: and that's not good if the lecterns involved are being used for other things. You can imagine a Rasberry Pi that was set up so that, when a USB stick, was plugged in, it read a config file from that stick, and then displayed the slides in the same stick in the manner (i.e. timing, screen res, etc.) instructed by the config file.)

  • Redo the Forum touchscreens. These do work in their current form, but they rely on a third-party compiled X input driver, which will probably give maintenance problems in future. The web browser arrangement isn't great either. Maybe the entire setup can be redone in a better way?

From 22/01 meeting:

  • A web-frontend to a longjob filtering system - something that would decide whether a user's resource intensive long running process would be most appropriately served by eddie, hadoop, beowulf or a compute server.

  • Creation and provision of 'kickstart' images to run self managed linux via DICE virtualbox.

  • An easier way to access your DICE environment from home using the existing freenx software running on a dedicated server or pool of lab machines.

  • A web interface to the backup system, for user's home and self managed machines. Some kind of concept in the style of mac Time Machine.

From Operational meeting 23/01/13:

After discussion in the Operational meeting, it's been decided that we should proceed with the idea to drive the touchscreens with a Rasperry Pi device. Initial tasks include procuring a touchscreen device and RP device for testing.

From email discussion w/c 04/02/13:

From Ian Durkacz, with Iain Rae's comments interspersed:

> I have some comments about our stuff though, some of which I've
> already mentioned to Iain:
>
> 1. It turns out that there is a slew of cheap ARM-based dev boards
> coming off the conveyor belt just now: it seems to be in the in-thing.
> So, if we wanted, we would have (many) options for a platform other
> than the Raspberry Pi.

yes, there's also some arguments about whether we should be using arm or one of the low power intel boards and running Sl6/lcg on it.

> See e.g. Beagleboard, Wandboard, etc. etc. etc.
>
> (Btw there are seem deep technical arguments about the pros and cons
> of each of these, and a lot seem to be to do with the internal details
> of the USB hardware. That might very well matter in things like hifi
> applications, but probably doesn't worry us.)
>
> 2. I have recently spent some time converting the touchscreen systems
> here some SL5 to SL6 and, in the course of doing that, have managed to
> make the things run using stock Linux s/w. (The [previous SL5 versions
> used properietary binaries.)
>
> I have learned a bit along the away, and would now expect (only:
> expect!) that the Raspberry Pi (or similar) boards will probably 'just
> work' - perhaps, with the addition of a single kernel argument to do
> with the USB HID driver - when plugged into a touchscreen.

This is pretty much what I thought at the beginning, either it's going to be trivial in which case the project will be what can we do with the pi and touchscreen or the project will be just getting the pi to work with it. There was also your suggestion about having something in place to control the drop down screens in G07 and the screen in the atrium.

> 3. Assuming that's the case, (and it very well might not be) we
> should probably try to decide what it is we're trying to achieve as a
> final outcome.
>
how about 2 parts:

1. Confirming that we can replace the current touchscreen PCs with Pi or PI type hardware.

2. Do something with the g.07/atrium screens (run a presentation from a usbstick) 2a. touching on other groups stuff here, would a pi be able to take streaming video and display it on the G.07, atrium and touchscreens?

3 parts...

3. Have some way of monitoring/controlling what's happening on all the displays?

.....and a fanatical devotion to the (Edited to say now ex-pope) pope...

> 4. We do not have a spare touchscreen monitor to play with: if we
> want/need one, we will need to buy one. Of course, for very short term
> testing, we could plug our boxes into the touchscreen monitors mounted
> on the walls. But don't plan to 'borrow' one of those monitors: the
> USB cabling arrangement really means that they can't be moved.

The thing to do would be to see if it works, if it does so more or less out of the box then to a certain extent we don't really need one to develop on (other I guess than testing the touchscreen usability part of it)

Another thread to this is that we'd only have one (or two) pi's to develop on, someone could take qemu and configure a pi development environment for us to use. I think that there are enough pointers on the net that there'd be about 1/2 a man days worth of effort involved in setting that up.

I think this is open ended enough that we can just take it and run with it.....and in many ways it maybe works better if we report back that

"we can do this.....this....or this......this isn't possible with this hardware....."

From meeting in IF-1.16 12/02/13:

A meeting took place this morning, unfortunately Ian D could not make it but his email to the group last week was much appreciated. A number of conclusions were reached:

Considering everyone's time constraints, including Ian D's current annual leave, it was decided that Friday 15th Feb would be the first of our full day's Innovation (subject to Ian D's availability and approval). We will appropriate the hot-desking area outside Craig's office, and convene at 9am sharp for bacon rolls and coffee. (Presumably there is a budget for this sort of necessity, if not, Craig kindly agreed to buy us all breakfast - at least that's what was minuted when he wasn't paying attention.)

Various tasks were suggested, in no particular order:

  • Investigate whether a Raspberry Pi communicate effectively with the touchscreens?
  • Investigate whether a Raspberry Pi will have enough ooomph to handle live streaming?
  • Investigate whether we can we use QEMU to emulate a Raspberry Pi?
  • Investigate if there are other (better spec/price?) ARM based dev boards available?
  • Investigate whether we could expand use of the Raspberry Pi (or alternate) to drive the large screens outside the ballroom?
  • Investigate whether the Raspberry Pi (or alternate) can cope with PDF files and PowerPoint presentations of varying complexity and built on varying platforms (Win, MacOs etc)
  • Document the whole process, start to finish
  • Make a blog to record progress (this has been done thanks to Jennifer: http://blog.inf.ed.ac.uk/innovationgroup1/)

Day 1 of innovation commenced at 09:00 15/02/2013:

All of today's activities will be recorded here: http://blog.inf.ed.ac.uk/innovationgroup1/

-- IanDurkacz - 21 Jan 2013

Topic revision: r10 - 20 Feb 2013 - 09:07:28 - RichardBell
 
This site is powered by the TWiki collaboration platformCopyright © by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding TWiki? Send feedback
This Wiki uses Cookies