Oh, I understand the issues just fine. It's DJB who clearly demonstrates he doesn't understand the issues. His argument derails in the second paragraph, where he appears to go by the assumption that at some point we flip a switch (he calls it "the magic moment") to go from ipv4 to ipv6. ipv6 was designed with the idea of it running alongside ipv4 and in fact all mainstream operatings systems and enterprise networking equipment are perfectly capable of this right now.
Yes, transition will be messy in the upcoming few years, but that is for the most part not because of poor design choices in ipv6. That is caused by almost everyone worldwide putting their fingers in their ears and singing lalalalalala whenever the topic is brought up (some think NAT is the solution, others say we should just force some of the holders of/8 blocks to give those up). Had everyone gotten their acts together 5 years ago, we would be visiting most websites over ipv6 by now.
You clearly haven't worked with Fx3 yet. At first I was skeptical about the new smart bookmarks and address bar, but now that I've used it for a month or two (starting with beta 2) I'm quite pleased with it. The address bar basically does exactly the same as it always has, except now it actually does an on the fly substring search in your history instead of just giving an arbitrarily sorted list of previously visited sites.
I would dare say this is one of the biggest improvements to the web browsing experience in the last several years.
I don't understand how this cash injection serves Microsoft's interests. What this will do, is keep the dark cloud of doubt over linux a little while longer. With this new money, SCO can appeal the august 11 court ruling and drag the nonsense on for a few more years. More doubt over linux means a few more chances for microsoft to maintain its stranglehold on the market. In the end, I think it's inevitable that microsoft's position in the market is going to change. This is just going to delay it a bit.
Nowadays, it isn't too hard anymore to create binaries that will work on most recent mainstream distros. Or they could open-source most of the steam engine and leave the authentication mechanism locked in a closed source library. I'd think there's lots of options to offer their products on many linux distros without giving up info on the authentication.
Also, if your main user base is business users, most of them will have some sort of software distribution mechanism. Whatever you decide upon, make very sure it'll work with any form of packaging. Also keep in mind that it's often much easier for the packager to be able to enter just a single license key for all installations, so don't make the software check license keys on the local network either. If you insist on counting concurrent users, you could consider adding a simple licensing server to be installed to the local network where each running copy can fetch their license (and fail to run if the license count is exceeded)
Slashdot Mirror
Oh, I understand the issues just fine. It's DJB who clearly demonstrates he doesn't understand the issues. His argument derails in the second paragraph, where he appears to go by the assumption that at some point we flip a switch (he calls it "the magic moment") to go from ipv4 to ipv6. ipv6 was designed with the idea of it running alongside ipv4 and in fact all mainstream operatings systems and enterprise networking equipment are perfectly capable of this right now. Yes, transition will be messy in the upcoming few years, but that is for the most part not because of poor design choices in ipv6. That is caused by almost everyone worldwide putting their fingers in their ears and singing lalalalalala whenever the topic is brought up (some think NAT is the solution, others say we should just force some of the holders of /8 blocks to give those up). Had everyone gotten their acts together 5 years ago, we would be visiting most websites over ipv6 by now.
http://cr.yp.to/djbdns/ipv6mess.html
I'm amazed at how much misinformation and outright bullshit someone can put into one single webpage...
I doubt that any of those tunneling services would offer me 80mbps up/down for free.
Actually, from what I've read, the Hurricane tunnel (http://tunnelbroker.net) gave someone their full 100mbps through the tunnel.
You clearly haven't worked with Fx3 yet. At first I was skeptical about the new smart bookmarks and address bar, but now that I've used it for a month or two (starting with beta 2) I'm quite pleased with it. The address bar basically does exactly the same as it always has, except now it actually does an on the fly substring search in your history instead of just giving an arbitrarily sorted list of previously visited sites. I would dare say this is one of the biggest improvements to the web browsing experience in the last several years.
Actually, since he mentions Burning Crusade, he probably meant Blood Elves.
Nowadays, it isn't too hard anymore to create binaries that will work on most recent mainstream distros. Or they could open-source most of the steam engine and leave the authentication mechanism locked in a closed source library. I'd think there's lots of options to offer their products on many linux distros without giving up info on the authentication.
Also, if your main user base is business users, most of them will have some sort of software distribution mechanism. Whatever you decide upon, make very sure it'll work with any form of packaging. Also keep in mind that it's often much easier for the packager to be able to enter just a single license key for all installations, so don't make the software check license keys on the local network either. If you insist on counting concurrent users, you could consider adding a simple licensing server to be installed to the local network where each running copy can fetch their license (and fail to run if the license count is exceeded)