"Ubuntu for Android" revisited?
Canonical tried this four years ago and gave up trying to partners two years ago.
https://en.wikipedia.org/wiki/Ubuntu_for_Android
Maybe this time it will work?
16 publicly visible posts • joined 29 Jan 2008
Wubi installs the linux image as a file on the Windows NTFS filesystem: This is hardly a fair "bare metal" comparison. On adequate hardware this shouldn't have much effect, but I found it significantly slower than installing on its own partition on an old system I experimented with.
"I'd probably consider buying one of these, but if they stick to their guns regards the whole "If we don't raise the full amount then we're not going to build it." thing, then it's pretty obvious that they're not going to be building any."
It would only take fifteen thousand people like you to decide to sponsor because they'd like the phone if it happens (although they expect to get their money back in practice) to make up the difference.
Over the three decades I've been developing software I've repeatedly worked on projects where the objectives were unclear, the targets changed and there was tension between funding, functionality and quality.
I've also repeatedly heard the sponsors, clients, or customers blamed for these problems. It is the developers that are at fault, change is a part of most problems that software solves, and the "soft" is in there for a reason - to support change. Change is normal, lack of clarity is normal, and in these cases to sell systems based on "you give me fixed, clear requirements and I'll give you a fixed price" is dishonest.
There are better ways.
"Nvidia's CUDA, for example, is tuned to its own GPUs, and Sutter admitted in a post-keynote Q&A that "if you want to get the absolute best performance from one vendor's GPU, you will hardly be able to do better than that vendor's GPU stack". Then there's open-source OpenCL – hardly a vendor-specific approach to GPGPU computing."
This is one of the delights of C++ - there are so many options to choose from. I prefer my code to be as portable as possible - across hardware, OS and compilers. Obviously, this cannot always be achieved but there's nothing about parallelism that requires Windows and Visual Studio. Selecting this technology would just make things harder to deploying on Linux,OSx, etc.
Time after time I've heard "we can't recruit good developers with X, Y or Z skills" - only to find that "we" are not paying enough to attract even trainee developers without any skills.
Accept that expertise costs (and that it delivers value to justify the cost).
While I currently need a full laptop most of the time I have fond memories of the Psion.
The keyboard was the thing - I could touchtype and it still fit in a jacket pocket. My (admittedly limited) survey of "modern" devices hasn't identified anything to match it but I continue to hope.
Galileo picked a fight with several cardinals and the pope on various grounds and notably by insisting that the burden of proof lay with those that maintained the established view that the Earth didn't move rather than those that supposed with Copernicus that it moves.
The Roman Catholic church took the very reasonable view that the established view fitted scripture and this was accepted as correct in the absence of proof of the movement of the Earth. Galileo had no such proof, and experiments of the time - such as attempts to measure stellar parallax - indicated he was wrong.
When proof was forthcoming, the church reinterpreted scripture to fit.
"Why is it accecptable to have to test in opera, safari and firefox and not in internet explorer? Given that IE has the biggest market share, should we not get it working in IE and then worry about the others?"
Because most (not all) problems seen in opera, safari and firefox are down to mistakes in the page that can be corrected by reference to the standards documentation. That allows me to correct the errors in my code.
Once my errors are eliminated, the remaining problems caused by browser bugs have to be worked around by searching the web for workarounds and experimentation.
Starting with IE is harder, because there are so many browser issues that the cause of each problem is less clear and the appropriate approach to solving it is less clear.
Attributing locations to this type of data is fairly routine in the area I used to work. Almost address validation software should get you most of the way, the postal address file the rest. (And there are plenty of companies that will do the job for you at a reasonable price - you might also get the advice you appear to need on storing spatial co-ordinates.)
But your project depends on this, and you should have addressed it early as an obvious risk.
But, as you note, there are other significant risks to your plan - depending on an unreleased product arriving on time and behaving as you expect. As others have noted, there are other products that exist now. (Even if they don't do everything, you can deliverer something.)
Simply allowing extra time is not usually a good risk management strategy. At best there is an opportunity cost of not delivering early, at worst a competitor takes the market while you are waiting.