There is nothing inherently wrong with long polling, but when you consider the overhead of a small payload (let's say 4 bytes for fun) the HTTP request headers add significant overhead. WebSockets have very little overhead for a 4 byte payload (2-4 bytes depending on masking (4 for the spec)) ... and the latency is less than half. It's just more efficient.
For those who still sit 8+ hours a day: Work on the boss or setup a budget for an adjustable stand up desk. Having a fixed height requires that you adjust if you want to add a floor mat, wear different shoes, or want to throw a treadmill down. And sometimes, after standing for hours, you just want to sit down. Spending $800 dollars is justified for anything that is used 8 hours a day.
When arranging my trip to China, I was told not to even try to get my visa by going to the consulate. I was told to pay a service company to do it. It worked. This certainly helps those who are willing to do it themselves. However, if you plan to go to China, find a service to do it for you.
Well, all frames actually do have length in the header. It's just not the first byte... it's the lowest 7 bits of the second byte, and possibly more (check the spec).
So, if I can read 2 bytes from my buffer, I can get a good idea of the message size. If I only have 1 byte, I rewind and wait until I receive another.
The WebSocket protocol works for both small and large messages in a single frame (message based), and also small and large frames in multiple fragments (stream based)... It's capable of being used for both. It's a good idea to restrict frame sizes on your application if you know what your limits are.
I don't think the study makes that claim. They are claiming that 6% of men with cancer don't eat enough fruits and vegetables (i'm not sure what enough is). There is no way to infer whether or not the lack of fruits and vegetables was the root cause of cancer in those cases.
Webbit looks much more like embedded Jetty and plain servlets than Sinatra. If you really want something like Sinatra in you Java project, wire in Scala and try Scalatra.
They are not totally slipping, but they are slipping in the high end desktop realm. They have not been able to compete with Intel, in terms of high end desktop performance, since the Core 2 arrived. Their first quad core was a train wreck... and the much anticipated (by me at least) Bulldozer seems to be at the level of Intel's year old Sandy Bridge.
The opinions of Unity are definitely mixed. I personally think Unity is great, and I have been using desktop linux for a decade. So, I am also glad to see someone find Unity useful.
What do you mean by robust query capabilities? There are different types of values in redis, each with special operations, but you can't really query on the values. If you can model your data with hashes, simple values, and lists, then redis may suite you. I use it in conjunction with relational tools, but I can imagine using it solely for a simple project.
I don't think IBM targets the enthusiast crowd with the Power6, so there are not as many knobs. I don't even know if windows supports Power6, which is standard platform for hardware enthusiasts.
I am no economist... so this is me trying to understand the situation. Doesn't news of imminent economic failure impact the economy negatively? If enough people are afraid to spend, because they are worried about money, then they slowing the economy. Once the forward economic momentum slows and wallets get thinner, the economic situation degrades further. Am I way off base here? It seems a self-fulfilling prophecy.
I used the think the exact same thing. However, after spending some time with tools that do consider white space significant (coffeescript, haml, sass) I have come to prefer it. It's part noise reduction and part consistency between devs and environments that I prefer.
EventMachine uses an event loop like node, so threads are somewhat irrelevant here. Unless "workers" are threads and not actual processes, but I am pretty sure they are processes.
Us Netflix users should do everything in our power to show that Netflix reduces piracy. Then, the studios, which own all of the IP, will work out a reasonable deal and we will have a larger streaming library. I don't care if it's true or false, i'm just tired of the MPAA not getting with the program!