Well, the article does not even convince me that the benchmark was properly executed. When going from a 32-bit to a 64-bit architecture, you certainly need the code to be properly optimized for the new target architecture. For instance, if you do not use the new instructions, it is unlikely you will see a major performance improvement. If you normalize the benchmark result to clock speed and number of cores there is not much difference between the 2 processors.

So my guess is: they did not properly compile the benchmark.

I disagree with that. What we want is to have net neutrality in practice. Now competition in the ISP market is a tool that could lead to net neutrality. It is an indirect way of getting net neutrality. But I strongly believe that net neutrality is important enough that we want to have a direct regulation about it.

Now, more competition in the ISP market woud not hurt :)

This is slashdot. Nobody RTFA anyway (even the editors).

There is a similar law (in spirit) in North Carolina about public universities trying to hire a university professor from another public university. You can not make an offer more than xx% over what they currently have (I think it is 10%). The reasonning is similar, it cost more money to the state to get the same person they were already having at a similar position.

I already carry a tablet around. So if I need to have a watch AND a phone, then it is useless to me. I want it to replace my smartphone to keep only a watch and a tablet.

I want to be able to make a phone call with it. Not dick tracy style, but I could slide the watch to my hand and use it as a phone.
I want email/text message notification. (I don't care about composing if speech to text works fine.)
Appointment and todo list and reminders. (Once again modification using speech to text)
Time travel estimate to my next appointment (or home, work google now style)
Giving direction like a GPS.
Weather forecast.

ideally, battery life of more than 2 days. and something much smaller than what all smartwatch are. I don't want a half phone on my wrist. With modern screen resolution a quarter credit card is probably big enough.

Let's be honnest here. If anybody is willing to buy anything I did for $2B, I'd sell it without thinking about it (even if it is a cure for cancer). Cash in your $2B and go explore some other crazy ideas you have that you release for free (you'll probably find a second way to cure cancer). You no longer need money at that point: you can live with $10M/y for 200 years...

That is something I actually wondered. If you go slower then you need more boats and more crews. Also you'll need to store more food on the boat. (I guess you could fish, but let's not go there...) So there is a fixed overhead which prevents you from going arbitrarily slow.

According to [1], it takes about 10 days to cargo from the UK to the US (east coast). That boils to to roughly 26Km/h. I don't know much about boats, but that seems fairly slow to me.

Anybody knows more?

[1] http://www.searates.com/refere...

Well, that might be a little bit extreme. I am all for promoting local economies. But we need to make a good cost/reward analysis. Quite frankly, the cost of bringing an smartphone (or whatever electronic) from overseas is actually fairly small compared to the cost of producing that smartphone. (Even assuming you bundle in the cost of phone all derivative ecological cost.)

I am not an expert, but if you distribute the production line all over the world, then you might lose a significant economy of scale when producing the devices. Also I don't you might still need to transport prime (or refined) materials: you might need chip foobar, which is not produced in the USA and need to be brought from overseas. Producing everything everywhere is never a reasonnable choice.

Also shipping from california to florida might actually be worse than shipping from brazil. So you might want to produce in florida. But clearly putting one production site per state is certainly ridiculous. Some things might have to be transported over long distances. Producing mangos or strawberries in New York, might not be so easy...

I thi nk that what you are seeing is the difference between fat packages and slim packages. What I mean is that in your typical gnu/linux distribution, libs are installed on the system and applications depends on teh libs. That makes dependency issues a real nightmare.

But all other succesfull operating system take a different approaches. on windows, application typically deploy their own libraries. On macosX everything is typically in a fat binary. On android, all libraries are shipped in the APK. I assume IOS works the same. They all depend on a slim "operating system" and on shipping "complex" libraries when needed.

I wonder if that is the main problem with application deployement?

Wow! There are still people doing that? I was doing it 10 years ago and found it useful at the time. I no longer need that configuration, but I'll be glad to hear bout it. Could you tell more? Did you blog anything on that topic?

I use auto a lot. auto (or equivalent syntax) are used a lot in functional programming languages. Mostly in short functions where I do not really care what the proper typename is. It is clear how the variable behaves and that is I care about it. Often, I know I get some kind of iterator, but the actual type might not be easy to find. In particular, it might depend on a template parameter. So I guess I could add plenty of typedefs to get an easy to write type. But what is the point really?

Indeed! Where are concepts! These is the number 1 addition to C++ most of us need! I am sure that they were not added for a good reasons. But programming template is currently a nightmare because of the duck typing of the meta programming system.

Dear standardization committee, we NEED a solution to the template compile time debugging problem.

Dear Prof. Stroustrup,

Writing templated C++ code has become one of my favorite way to write abstract algorithms in an efficient (runtime wise) way mostly because it not alone allows to know types at compile time, but also some values. Yet programming templates is nightmare because the template system is compile-time duck-typed. How can that system be made better? Concepts were proposed for integration in the C++ norm but got pulled back. Yet it seems they are essentially what is required. Will template programming ever improve?

Best,

I am an unlimited 4g lte customer of t-mobile. And when I asked what unlimited meant, the seller told me exactly what it meant. unlimited up to 2GB per month (which is a lot, I never reached it), then throttled down to a slower speed which still allows you to check emails and navigate.

I even frequently use my phone as an internet acecss point for my computer. But I don't dump the web when I do so. So it never was a problem. The only people that reach the throtling are pretty much people that explicitely try to push the limit and know very well they are not supposed to. If you are smart enough to route P2P application through your phone network to use the "unlimited" internet, you are probably smart enough to know what unlimited actually means. So yeah, I get it, companies are misrepresenting, but does anybody actually get tricked by that?

I'd like to see the precise rules (but too lazy to RTFA now). There are many non-english words that can be highly confusing. In french "telephone" is "téléphone" which could be though as a way to trick users. Also turkish have a dotless i, I would not be surprised it appears in words with similar spelling in english.