Well worth watching :
Well worth watching :
Sure - but do any of the FTDI interfaces properly support serial break yet? Annoying as ****
I think people have been saying that every year since i can remember on slashdot. Not entirely convinced that this one is any different!
Its an electrical device. Unplugnit before working on it. Its the first thing you get taught. End of.
Absolutely right, but not an issue in this case - it's completely isolated.
Broadly, building an ARM linux target via yocto, some in-house tools (portable C), a couple of GCC variants for other targets. Pretty standard stuff but the main thing is being able to able to address any potentially critical bugs and recompiling in the same environment at any point during the long lifetime of the system.
I think that's a little optimistic - I'm not aware of any long-lived embedded systems where the environment is updated in sync with the outside world. In practice this is nearly impossible (or rather, financially infeasible). Look at embedded systems developed for aerospace / medical devices / safety critical control systems for example. Once the system is tested, you don't do any major changes to the development environment or code!
And yes that does end up with a support nightmare, hence the question
We're not talking about purely compiling an application but a complex environment for an embedded target that needs to be reproduced in a almost identical manner in order to avoid full system re-testing that presents a massive overhead / cost. (look at build systems such as Yocto for an example). It's easy enough to archive all the tools required to do the build and test installing them on a non-networked machine to check you have everything in the archive.
However, consider how you'll be able to exactly replicate this environment in 25 years time when you don't even know if you'll be able to find a working machine from the same era to run the host OS on and that's the problem in a nutshell.
News item from Leicestershire Police :
Interestingly apparently a 0.2% false positive rate. On attendance size of 90,000 that's 180 people mis-identified. This is apparently using NEC's NeoFace Watch system which they started trialling in the middle of 2014.
If you don't do what the police ask you to, you will not be able to run your event, it's as simple as that. This is often a financial issue for example. :
I could name several smaller independent festivals in the UK that have been made financially infeasible by large imposed policing costs and had to be subsequently cancelled. Many would argue that often the policing costs are un-warranted and that the police may use this as leverage to prevent events that they don't approve of but no one has demonstrated this successfully to date. Certainly however it's true that there is wild variation in the level to which different events are burdened by the police but of course this can also be attributed to other factors such as the quality of their pre-event planning / paperwork and the differences in approaches that different regional forces take.
Personally however, as both and event professional and an engineer, I'd like to know whether this was imposed on Download as a condition of their license, and if so what their justification for requiring it was. Secondarily I'd like to know how it is that they have been allowed to do this on the quiet. Certainly I believe that storing images of peoples faces would be covered under the data protection act.
Really? You're implying that slashdot readers are properly technically savvy? Those days are long gone. If that were true, all these arguments would be presenting reproducible and well explained cases (like proper bug reports!) not vague anecdotal stories that can't be proven... Just sayin!
Maybe we have a history of not much over here, but we've got hundreds and hundreds of years more of not much than our american cousins...
...in other news, kim jong un now reportedly threatening verizon customer services with ground attack after being on hold for 90 minutes...
Can someone who understands PV cells properly answer me this - it's my understanding that most standard PV cells get say low 20-something percent efficiencies and that the highest at the mo is around 40% acheived with significant solar amounts of concentration. Is this a genuine like-for-like beating of the best of the current crop or some fudging of the maths?