True enough, but I didn't buy Ubuntu, and I did buy Windows, directly or otherwise.

My current laptop came with Ubuntu with the Fluendo codec pack. I don't think I got a free upgrade for that, either.

I'm not seeing any Rails-specific flaws there. If it's the '..' stuff, that really looks like the webserver that's the issue here.

Oh, and it does scale, it just scales horizontally. But maybe you missed the part where this particular site was never intended to scale, and doesn't need to. Or the part where it was done in 15 hours -- that's certainly a reasonable amount of time even if this is only a prototype, so that the entire thing needs to be rewritten from scratch later.

I'm also a bit curious as to why that AC would use this vulnerability to antagonize Rails, rather than submit it to the relevant projects.

That's cool, and also practical, but also frightens me a bit. While I don't know of a better way of doing it, that is effectively training people that the way to code something is to find something similar and copy/paste.

About the only way I can think of also teaching DRY in a game is by giving the player a severely restricted environment either in terms of amount of source code or memory usage.

I suppose, but then, there's also more error using mouse-aiming and WASD rather than auto-aim on a rail. Which one is more fun? I'm not even sure that I'm slower when I play without any sort of auto-aiming than when I have the game effectively cheat for me.

It's JavaScript. You might need to learn a library, but presumably it'd teach you.

Here's one. Here's another.

Even so...

I've definitely run into bots like this in IRC who were somewhat spam-focused, and identified them within a few minutes. Reading some of the transcripts, it looks like there are people who never figured it out.

I suppose it's possible that these guys had no idea that bots exist?

I don't think that's the case.

First, it's not English specifically. It's being able to communicate your ideas well. There are plenty of native English speakers who have a harder time being understood than countries where they don't speak English. (Though for what it's worth, English is one of the two official languages of India.)

Also, even the work that's going to India, a lot of it is stuff like call centers, and the better ones do require English. Not just English, but heavily americanized English.

It's also worth mentioning, there are still plenty of jobs which are here specifically because communication is valued. Maybe you hire a consultant in India, but your core team is still here. And even when you hire that consultant, you probably want someone here to act as a go-between, or at least to talk to management about it.

The last job I had was working for a startup. I started out doing HD-DVD development with them, and I realized we were one of very few companies actually doing this, and one of exactly two who did anywhere near a decent job. But we were in the US, and our headquarters were officially in California (along with our CEO), while the other company was in Germany. Guess which one the studios wanted to talk to?

Sorry, did you just compare doing actual schoolwork with collecting welfare? Maybe I missed the part where collecting a welfare check requires writing papers or understanding advanced mathematics.

I don't buy everything RMS says, but I really think this is a misrepresentation:

So, who's forcing you to develop GPL'd code? No, it's people who build on GPL'd code and then distribute it. In other words, what he's saying is that you cannot build a proprietary product on GPL'd code.

I don't think that's any more outlandish a demand than, say, Apple's licensing requiring that certain pieces of Apple software specifically should not be used to make nuclear weapons. Or, for that matter, Microsoft demanding that if you want to even execute most of their software, you have to pay them.

I prefer BSD-licensed code when I can get it, and I tend to release BSD-licensed stuff myself when I can, but that doesn't mean that anyone is "forcing" me to do anything with GPL'd code.

I suppose you could interpret "Software Libre" that way, but this is about the freedom that you as a user and developer have with such software. In particular, you're generally free to do anything you want with Free Software other than distribute a proprietary derivative. You can take GPL'd programs and invoke them from a proprietary program, so long as you obey the GPL wrt. the actual GPL'd programs you're using. You can even develop a proprietary derivative and not distribute it, as Google does with Linux internally. For that matter, you can sell it for a price and only distribute the source to your actual customers.

That's what he's talking about. Not the freedom of the code itself, but of future users and developers.

Now, if your problem is that it limits the freedom of developers as well -- for example, it limits my freedom to build a proprietary product under a different license while including GPL'd code -- then you're right. For what it's worth, BSD-licensed code is also considered Free Software, and it has almost no restrictions on its use. Hell, public-domain code, like sqlite, is also Free Software.

But think about what we mean when we talk about freedom in other contexts. We have freedom of speech in the US, but there are many well-documented limitations to that at pretty much every level. Certain kinds of speech -- "Fire!" in a crowded theater, false advertising or outright fraud... I don't think we would take these to mean that freedom of speech does not exist in this country, or that it's hypocritical of us to criticize the kind of censorship that China engages in, simply because we don't have absolute freedom of speech. Similarly, I don't think it's hypocritical to criticize proprietary software as being "non-Free" just because GPL'd software has some limitations.

For what it's worth, I don't generally use the terms Free Software, and while I prefer open source, I'll use proprietary software when it makes sense. I'm not a zealot. (Honest!) But the idea itself isn't inconsistent or hypocritical, and it does have some merit. It's just difficult to achieve in the real world.

Microsoft is phasing out XP support. Sooner or later, even the security patches will stop coming.

Linux is an 18-year-old OS in the same way that Windows 7 is an 18-year-old OS -- Win7 was based on NT, which was initially released in 1993. The difference is, XP is like running an old version of Debian, and even in the Linux world, I have to imagine that eventually no one will care about even 2.4 kernels.

So, do you actually use a version of Linux from 18 years ago? Do you even use, say, a major Debian release from then, with security patches only? Or are you using a modern Linux? Because if you're using a modern Linux, that's pretty much the equivalent of using Win7 instead of XP.

Really? Citation?

Because I was under the impression that it costs significantly more for some applications.

This.

I use Linux as my primary OS. I might consider paying for a service which operates via Flash, though I'd bitch about it. I refuse to pay for Netflix, because it requires Silverlight. While Silverlight is a theoretically open standard, the DRM Netflix uses with it is not, and only works on Windows.

That would be a bad reason, and I posted roughly the same to GP -- both nVidia and AMD provide H.264 hardware-accelerated decoders for all modern OSes, or at least for the ones for which they provide drivers.

However, WebM isn't "a lot worse." It's actually pretty close, particularly for the sorts of applications we want on the Web. The only significant difference is that H.264 is already everywhere, and in silicon -- WebM is just starting to get serious hardware acceleration.

And Linux is a symptom, not the actual problem. The actual problem is that if H.264 is used everywhere, even by Free-as-in-Freedom browsers, then those browsers will be significantly crippled on any platform which did not pay a licensing fee. For that matter, if I want to write software which reads, edits, or otherwise plays with video in such a way that the codec API isn't sufficient, I have to pay a licensing fee.

So this isn't really about the geeks running Linux. It's about stifling competition and raising the cost of entry to the market.

However, nVidia and AMD both provide H.264 codecs which are hardware-accelerated. Unless the entire framework is missing, installing video drivers is all you need -- and if you don't do that, then you're probably looking at pretty terrible video.

I don't know if any Intel cards have this, but this is still going to be a fair majority of desktops even before you get into versions of Windows which include a software H.264 decoder. Of course, there are free H.264 codecs available for download.

My Ubuntu actually came with these codecs -- I bought it from Dell, and Dell licensed the codecs. Additionally, I've got a laptop with an nVidia card and a desktop with AMD, so I've got the same story again -- in fact, the AMD card comes with VC-1 as well.

Again, if I for some reason have an Intel card, or otherwise have a hardware platform which doesn't include an H.264 decoder in hardware, there are codecs to be had. In this case, they are in a legal grey area, but they are available. And remember what happened with GIF? Patents do expire.

For video creation, similar things are true -- I think even the iPhone has an H.264 encoder in hardware.

I'd much rather have VP8/WebM, but the fact that I no longer own hardware which is capable of playing video and does not have native H.264 support (at the hardware/driver level) means I can't really bring myself to care. I mean, do I prefer proprietary codecs over open ones? No. But do I prefer proprietary codecs over proprietary plugins (like Flash)? Yes, a thousand times yes.

Then elaborate for those of us who are clueless. What is it about Ruby on Rails which automatically implies that anyone choosing it for anything doesn't have a clue?