I am not convinced AMD/ATI cares that much. Their driver situation is *still* bad enough that I'm using an integrated Intel chip that can barely run Google Earth until the Libre driver supports my ATI card. It doesn't say much for their regard for Unix that they haven't fixed a long-standing bug, 100% reproducible, affecting a standard Unix feature accessible on modern desktop environments in two clicks. It doesn't say much for their regard for openness that they haven't even acknowledged that it exists.

To be fair, the reason I got an ATI card in the first place is that they released their specs; I figured there would be passable Libre drivers eventually. I didn't anticipate how awful the Gratis ones would be.

I hate to feed the trolls, but last I checked you could run the VESA driver for pretty much any card. It offers a couple basic 2-D acceleration primitives that are in the VESA standard, which is usually enough for me.

Some people are working on a Libre ATI driver. It takes a lot of work, even given specs. It doesn't support my ATI board yet, so I'm not using it yet (I'm using my onboard Intel until then). When the Libre ATI driver is ready I'll put my ATI card back in and give it a whirl. It almost certainly won't have this stupid bug in it. And even if perf is a little worse than the gratis driver, it's worth it for that reason.

The difference is not much on Windows. ATI and Nvidia care enough to trace issues and fix them on Windows. They don't care enough to do it on Linux. They don't care enough to implement many X features in standard ways either.

The difference is that with a Libre driver when there's an annoying bug it will get fixed. With the gratis Linux drivers there are plenty of annoying bugs that will probably never get fixed. In another post on this story I mentioned ATI's fast-user-switching bug. It causes X to go down completely when you use it. A lot of KDE and Gnome devs probably want to fix this bug, and are perfectly capable. I want to fix this bug, and I'm probably capable. But none of us can.

The difference is that you'd get better quality in many respects with a Libre driver.

Because it's in the kernel of their operating system. Because the fact that the driver is not Libre prevents other desktop-related stuff from working because the one vendor doesn't care and nobody else can fix it. Here's an example:

Using the gratis ATI driver, running two X servers on the card crashes the driver and leaves X and the card in an unusable state (you have to ssh into the box to reboot it cleanly). This has apparently been a bug in the ATI driver for ages. And because multiple X servers are used to implement "fast user switching", ATI's crap driver blocks fast user switching.

This sort of bug would be fixed in a libre driver. It's 100% reproducible, incredibly annoying, and affects a feature in desktop environments with millions of users and thousands of developers. If I had the source code to ATI's driver I could probably fix this bug. But ATI doesn't care.

It's impossible for the Linux kernel team and X.org to design interfaces and a good model for how kernel drivers should interact with userspace X drivers to provide rendering in a way that fits in with X's model when the two biggest GPU makers will just ignore it, write their own kernel modules and their own interfaces. With a Libre driver new X.org standards and interfaces would be adopted much quicker and the drivers would fit into the system better. Nvidia and ATI care about this for Windows (to some degree) and so their drivers fit well there. On Linux they don't. But lots of other people do care, and non-Libre drivers prevent them from doing anything about it.

No Free Software license, GPL included, relies on the theory that you must get the permission of the copyright holder to use the software. The GPL only applies when you redistribute. The software is freely offered for any usage.

The theory that using the software requires a copy as defined by copyright law is used only by proprietary software makers to enforce EULAs. I think it's a pretty dubious theory (to be specific, I think the idea that copying the binary from system memory to RAM counts as a copy for the purposes of copyright law is absurd, but a copy from installation media to your hard drive should probably count), and apparently they're not too confident in it either -- they also do their best to make sure users agree to the license in one way or another.

Even if the proprietary vendors' theory is right it doesn't affect the GPL much. If copying to system memory constitutes a copy, you're now obligated to provide yourself with source code if you request it. Or is it that your hard drive is obligated to provide your RAM with source code?

It depends.

Say you watched it on YouTube or a similar site. YouTube keeps track of what videos are watched; the person that posted it keeps track of videos that are watched. The more people watch their videos the more money they make. So at the very least you've created demand for videos of monkeys attacking people. At some point if enough people love to watch hilarious monkey attacks people will stage them.

OK, that's a bit of a stretch. But that's a video made of an accident. Child porn is not an accident. If you download child porn, even if you don't pay the creator, the person you downloaded it from knows that it's popular and that he should try to get more. This will at some point probably lead to creators of child porn being paid.

Iunno, the Kindle makes a lot of sacrifices to publishers. Including incredibly stupid sacrifices like allowing them to disable text-to-speech for whatever books they want (I don't have a Kindle but I know some people that work for Amazon, one of whom does marketing for Kindle).

Amazon's model seems to be centered around itself foremost, and clearly it must balance customer and publisher demands. If this other company wants to cater more to publisher demands I guess they can try it that way. But don't tell the public that!

TBH little we have today is that close to Craig's view of a "town". When looking for apartments in Chicago there's a big difference between Hyde Park and Uptown unless all your major commitments are downtown. If you work in Lakeview one is a couple stops down on the L, the other requires at least an hour and at least a transfer.

When a couple of my co-workers were looking for an apartment they wrote a Python script to go over Craigslist RSS feeds for apartments, plug the given street names into Google Maps, get walking distances to various important landmarks (the office, their favorite little Italian deli, Red Line stops), and weigh them against each of their personal criteria. They wound up finding a great place. Even if Craigslist were sorted by neighborhood, Chicago neighborhood boundaries are fluid enough that it wouldn't help that much.

Dude, I thought this was the 21st century! If I can't use a cell phone in a bathroom stall why'd I bother getting the damn thing? Oh, right, I didn't get a cell phone. But if I did I'd probably use it in bathroom stalls, this being the 21st century and all.

Craigslist leverages the Internet to provide a hell of a lot of service to a hell of a lot of people without doing much work at all. They skim a little money from some of those people and say that's enough.

Microsoft creates a lot of work for themselves by making lots of new features and then convincing people that they need them. It's how they leverage their advantage as the world's largest software company, and the rest of the industry (and lots of people doing OSS) fall for it.

Google is probably pretty much the same these days. The point is that these companies are worried about shareholder value first, they're worried about winning. That's why they make all this work for themselves. Craigslist just provides the service. Take it or leave it.

The first thing that came to mind when I thought of C, network-facing services, and security was buffer overruns. They aren't hard to avoid, but they aren't hard to write either. Encouraging programmers to write ad-hoc scripts to process input in C sounds like a remarkably bad idea to me.

CRT monitors beat the crap out of LCDs in similar price ranges in just about every aspect of picture and video quality, it's true. But LCDs are light and small; that's a lot more important to me. I used to have a 19" CRT. It was a beast. Took up way too much desk space, and space in the car every time I had to move it. I finally decided to get rid of it when moving away from California and none of my friends would take it for free. My current (rather old, obtained for free) 19" LCD has a tiny footprint.

The one thing I can't stand is widescreen monitors. Less visible area for bigger desk footprint. Stupid dimensions for doing anything but watching movies (but why would you watch movies on a crappy LCD computer monitor?).

Well, and that if they're working all day they can't get an education. Meanwhile there are lots of unemployed adults in many of these places (source: this guy on Colbert a bit ago) -- if they can't work, and the employers would rather hire kids (apparently this is the case in many places), well, that's just going to create a cycle of poor education and child labor, generation after generation.

I know it's basically impossible to get an accurate idea of the conditions of labor in these places without going there. They're probably not that good, though most are probably better than the horror-story sweatshops that are often portrayed.

1. Voice commands will work really poorly in crowded places. I'm trying to picture how I'd use voice commands reliably in Denver airport.

2. Any sort of holographic projection thing is going to require something to project on. For privacy reasons you'll often want something you can hold up to your face, and there are many surfaces that would be impolite to use for a holographic keyboard. If you have something you're holding up to your face, now you have to hold the watch steady relative to it... you'd be better off with a detachable LCD screen. At which point you might as well put the phone hardware in the LCD screen instead.

3. Again, ergonomic factors dictate that a watch is not more convenient than a phone for making calls. My watch has one button on the front and four out of the way on the sides. More buttons on the front would make it harder to use (it is typically pressed blind, sometimes while wearing gloves, usually while very tired and out of breath, and for my purposes I need fairly precise timing). A phone needs... a numeric keypad! A device with an effective keypad and be easy to talk into, something good enough that I'd prefer it to my landline, would have to be big. You could have a Bluetooth thing that looks like a phone or a headset or something... but, again, you might as well just make that the phone instead. If you get a call on your watch-phone and don't have a headset with you, what can you do about it? You just get a ring, and maybe you can listen to the message.

I think the biggest issues here are human factors.

There is no way a device small enough that I will want to wear it out running will be comfortable to use for phone conversations. And no way a device comfortable to use for phone conversations will have a large enough screen to read, browse the web fully, and write code.

It is at least plausible that cell phones will one day have good enough sound quality and ergonomics that I'll want to replace my land-line. Cameras are another like this -- I have no doubt the best phone-cameras are good enough today that my weak photography skills need nothing better.

Phones may replace some of these functions for some people, but they won't replace all of them for everyone.