Pretty much single AAA title the last years has come with a driver update requirement. Try to go pay BF1 or whatever the latest title is - you'll be required to get the latest nvidia or AMD driver. The game will tell you. Not sure if this is the case for many store games but I assume that for any AAA titles it would be the same. That's why this is a convenience for those cases. The driver is more or less PART of the game because the necessary driver changes were done in collaboration with the studio.

The excuse in this case is that the article is either a misunderstanding or deliberately misleading. Btw who posts without reading the comments, half of which are already about the fact that the article is crap

Of course there is no way of buying a Volvo that could somehow self drive but fail to detect humans while doing so because that would be an optional extra. TFA makes it sound like that is the case (it isn't) and I'm starting to think its deliberately misleading because no one could believe that honestly.

Either that, or TFA links to an article compketely misunderstanding what's actually going on in the video. Hint: no self driving going on, and pedestrian detection isn't related to parking at all. ignore TFA.

Get a different source, the independent article and the summary on /. makes no sense. The guy is manually driving, and hoping to test the pedestrian detection radar (which only works when driving forward at city speeds).

You just misunderstood what the auto parking and pedestrian detection systems are. During auto parking, all obstacles close by are detected. The pedestrian radar isn't involved in that. The pedestrian detection is a radar used when driving manually at city speed in order to auto-brake for kids running into the street. Even without the long range radar you can still avoid any obstacle, including humans, while parking. The title of TFA is crap eve for /.

Parking had nothing to do with it. It wasn't a demo of the parking system, and the problem wasn't with the parking sensors. The demo showed someone try to use the pedestrian detection (city speed long range obstacle radar) when there wasn't one on the car.

There are so many misunderstandings here, so let me clarify some things. There are two different technical (hardware) systems we are talking about: a "parking assistance system" which consists of cameras and ultrasound sensors, which work at short range during parking. They also provide input for autp-parking.

Second, there is the "pedestrian detection system". This is a radar-based long-range detection system used when driving in the city (for auto alerting the diver and/or auto-breaking if a child runs out in the street, for example).

In order to do self-parking, only the parking assistance hardware is involved. The parking assstance avoids all obstacles, and of course it would never automatically move the car if it detected an obstacle. The pedestrian radar is an optional package simply because the hardware is still quite expensive. Of course there will always be optional extras on cars. Volvo is probabl class leading when it comes to having the safety features made standard as soon as possible, but this piece of kit is just too expensive yet

So: 1) Volvo does not "charge extra" for enabling some feature on hardware already included. 2) There is no "pedestrian detection" that can be enabled or disabled that relates to parking .It's a city driving pedestrian safety option. 3) Other cars with parking assist or automatic parking have anything other than the sensors (cameras/ultrasound) that Volvo use.

Targeted surveillance is exactly what *should* be used, because it can self-regulate. There is a cost associated with each target, so there must also be a benefit otherwise it won't be done. So widespread strong crypto sounds perfect: it takes surveillance/intelligence ops back to the physical world where you pay per target and not per system of mass surveillance. And think of all the emissions saved at the datacenters!

• - Structured programming.
• - Programming paradigms and their differences and benefits (functional/procedural/declarative/imperative/object oriented), how to choose language/paradigm depending on problem, and useful patterns and strategies for each.
• - Type systems (weak/strong/dynamic/strict). Drawbacks and benefits
• - Basic development methodology (version control, testing, error handling, debugging, ALM, documentation).
• - "Data structures and algorithms" i.e. basic discrete math, complexity theory, data structures and their algorithms.
• - Some low level knowledge: basic understanding of how a computer works, how memory/cache/file systems and OS:es work. What's two's complement? What's epsilon for IEEE 32 bit floats?
• - Basic computer security, encryption and hashing.

I think databases, "The unix philosophy" etc. are more controversial and should not be on the essential shortlist.

If you work on games, desktop applications, web frontend etc. the database is by no means essential. I have programmed for many years, mostly heavy desktop applications, without having to use relational databases (of course there is *data* but usually structured in binary, text or xml). I can see your point about structures and pointers being somewhat analogous to tables and relations *but* unless you are converting a DBA into a developer, wouldn't most developers see it the other way around? That is, when they see a relational db table for the first time (having coded for a while) they see the similarity to a struct or object?

I think "tarball" makes very few people enthusiastic. Most people run windows (honestly people who know what "tarball" is are a rounding error) and prefer a zip or installer for applications, and a zip or package (nuget/npm/gem/etc) for source. When running Linux I don't care whether I get a git repo url or a tgz, they are about equally cumbersome.

Well during the rise of Linux servers on both ends of the spectrum, microsofts revenues from its server business has been very good.

People think of microsoft as making Windows, Office while failing with mobile and games. But you need to look a bit wider to see the whole picture. They have moved in on servers, making Windows Server quite a large player where mainframe systems used to rule. They have successfully moved in (through aquisitions) on the business system area, taking a large chunk out of the revenue of companies like SAP/Oracle. If you include Business systems, Databases, Servers etc. you will see that not only are they either enjoying there monopoly OR failing, they are actually quite successful in areas where they never had a monopoly.

Leaving more often is convenient, but to calculate the cost per passenger journey you still need to look at the total cost of all the trips over the lifespan of the system. The convenience of fast and regular departures may be what lures people from the roads to the hyperloop (that wouldn't otherwise have taken the train), which is a good point. But still: one order of magnitude cheaper than HST is no big deal if the total passenger throughput is an order of magnitude lower!