To be fair, in a lot of those "curvature of the earth" pictures, the apparent curvature comes from the wide-angle lens more than actual curvature of the earth.
Not really. Thunder clouds usually top out around 60.000 feet. There are some wind phenomena that could reach this high, but mostly I think it's calm at this altitude. The air is so thin you would have to have very strong winds for there to be a problem.
Aerodynamic lift (in Newtons) is proportional to airspeed and invesely proportional to air pressure. So is aerodynamic drag. This holds true up to the transsonic region; ~800 kph of true airspeed. As long as you're subsonic, it's sufficient to have the same amount of thrust as at low altitudes, and you'll keep flying at the speed required for lift >= aircraft weight.
Granted, this will require proportionally higher propeller RPM, but with electric engines this is unproblematic. You'd get in trouble with hydrocarbon engines though, since you need the same amount of oxygen to generate power. So at altitude, you'd need a turbocharger or similar machinery.
What do you think about point number five; graphics? Is graphics an important part in the next future game hit? Graphics is certainly not a priority in the social games that have been succesful so far. The beauty of many hit social games I feel is in that they're so easy to start to use and even my mom understands them, will AAA graphics add to that effectively (since it's so expensive)?
Disclaimer: I'm not a gamer or games programmer. But it seems to me that "good" graphics are not just about beauty, but style and personality. For example, South Park animation is crude but instantly recognizable as South Park. There would be no point in converting it to Pixar-level quality except as a joke.
Hahaha. Figure this: "Do you have a card?", "Not only have I got a card, I've got an NFC-enabled card. Put it against your phone and let magic happen." Awkward but hey I want one.
Great to see a nice overview of best practise in app store screenshots. As an app-developer I also would like to stress the benefit of good screenshots. To add on these best practises I think you shouldn't do a real life example of the app in use without having some serious equipment (camera/lightning) and possibly "models". The author mentions to do this with care but my 2cents is just don't do it. I've only seen a few decent tries on this and generally that would be an app from a well-funded company that obviously spent a lot of effort/cash on it. Those screens set the standard in that genre and for a small developer to compete with that is not effective, do a digital less-is-more screen with good copy instead.
What kind of risk/liability does these "employees" generally take? Like washing someone's unlocked car, is the employee liable for any damages or the company? If I'm do carpenting work distributed from TaskRabbit and hurt myself so badly I get disabled, do TaskRabbit have insurance for me? What is the standard for this? One reason why buying services is expensive is that the service provider takes risk and needs to insure their workers, I very much doubt that these TaskRabbit-type companies do that... Or am I wrong?
I hired a TaskRabbit to pick up a large amount of IKEA and deliver it to my house. The TaskRabbit proceeded to show up an hour late with the furniture...because she dropped the furniture on her foot, breaking multiple bones.
Once I heard this, I realized "Shoot, I contracted her. Am I legally liable for her injuries?" (Fortunately for me, she shrugged it off, and seemed to be covered under some insurance of her own.) And if she had not delivered the furniture due to her broken foot, would TaskRabbit make me look like the bad guy for asking for my money back? (Fortunately again for me, she made the delivery, amazingly.)
While things turned out OK in the end, suddenly the whole service felt much riskier, for both the employee and the consumer - and I felt irritated at TaskRabbit for subjecting me to that risk without some sort of a safety net in place. (Heck, I would have paid extra for one, some sort of "TaskRabbit Insurance.")
I have to say, as much as I dislike the idea of tight regulations on workers, and can see directly how much cheaper my services are removing those regulations...I think next time I'll pay the premium for a more legitimate furniture move, at least, until the law catches up to services like TaskRabbit.
This is a huge point. A single uninsured mistake with an expensive car could potentially wipe out carwashing earnings for months. (See http://news.ycombinator.com/item?id=4492619)
This could be mitigated by TaskRabbit cooperating with an insurance provider and selling "insured" work. There are great margins on insurance as we all know. In their TOS TaskRabbit basically says "TaskRabbit.com Only Provides a Venue" (check it out: http://www.taskrabbit.com/tos). Why don't some retail-insurance megacorp approach them? This could be TaskRabbits no1 source of income instead of their no1 problem...
I'm not sure about the US legal situation but here in Sweden there is so much legislation to protect workers that a TaskRabbit kind of service becomes hard to achieve.
IANAL, but I'm almost positive they don't have to do this in the US.
In the US, hiring contractors exempts you from many of the taxes and liabilities of hiring employees, and they are also exempt from most benefits (their right to litigation against the company is also limited).
US readers will certainly be familiar with Vector (ie, Cutco knives) - they get away without providing any benefits/insurance, without reimbursing for expenses such as gas/phone bills, and with paying what often amounts to minimum wage because all of their salesmen are college students who are hired as 'contractors'[1].
Don't think I'm saying that contracting is bad - the same laws allow others to make very large sums of money working as contractors. It really just depends on what the setup with the company is.
[1] I may be confusing some of the details with other similar companies, but I know Vector engages in at least some of these practices, all of which are commonly practiced.
Thanks for your reply. Talked with some more informed friends about the situation in Sweden and their conclusion is that work like TaskRabbit probably could exist in some kind of grey area (especially when it's one private individual working for another private individual for a small amount). However when the service grows the unions will try everything to make your workers "unionized" implying that either the service company (e.g. TaskRabbit) or the buyer needs to pay insurance and minimum-wage for the work being performed - basically hiring the individual.
I guess this is not interesting for must of you US readers but it could be telling of what could happen to your TaskRabbiters.
legally they operate as entrepreneurs providing services for customers sourced through a website.
This exempts them from many of the worker protections.
However, at least in Belgium this kind of micro-entrepreneurship is actively discouraged.
Contractors require an accounting-certificate or academic degree in order to setup a company (or sole tradership).
Skipping the basic step of registering as a company can leave you or the customer liable to all sorts of fines, depending if a court thinks the customer or the worker is at fault. The issue is sensitive because it involves manual labour, and there is a huge incentive to utilize black labour as labour costs are very high.
This renders a service like TaskRabbit a legal hazard for customers and workers, but the service itself can operate legally.
Thanks! Good starting point. I'd like a section on using django on cloud services, e.g. AWS Elastic beanstalk. Another best practise I'd want is scheduling scripts with https://docs.djangoproject.com/en/dev/howto/custom-managemen... and Cron, or using Celery. There is much confusion on those topics I think.
I'm not giving best practises as I'm myself fairly new to Django, this is just a couple of packages to take a look at.
I've been deploying two projects to AWS very recently and we've been using django-pipeline and django-storages (with s3 boto storage) for asset management. ./manage.py collectstatic and all your static files are up on S3. With a bit of finagling around[1] you can even have user uploads hit there seamlessly as well.
EDIT: Pipeline isn't necessary for storing static files on S3, but if you want to compile SASS/LESS/cs files or any transforms really it works really well.
What's confusing you? (I'm trying to help, not being snarky)
If you want to invoke some Django-based logic at regular intervals without having to install Celery (and monitoring and a decent queue) you'll opt for a management command. The link you posted should help you out here. Invoking the script yourself or telling cron to invoke it for you shouldn't be hard if you know about cron.
With regards to Celery: I think the tutorial and docs are pretty clear on how to use it and how to set it up.
Thanks for your comment =) Nothing is confusing me there, but try to search for "django script cron" and you'll see people suggesting setting up urls to start the script, having an external script that import settings and many other complex things. The craziest thing I've seen (but probably useful for some cases) is to have the regular requests from google-bots invoke scripts... That's why I think it's a good thing to suggest managment commands when they're so easy and integrates nicely with your apps. Celery however is good for more complex usecases. I was merely making a suggestion for best-practise.
I imagine most of those icky suggestions are coming from people who used to run php on hosts with a lot of restrictions. You're right, this should definitely be a part of a best practices guide.
