I agree with you 100%, the government should absolutely get out of the distribution business. Naturally, that means repealing all copyright statutes entirely.

File birth time is a fairly difficult concept, and only really useful on say a database file that's edited in-place. Any text file/source code you've written will have btime=ctime, since it was 'created' as a temporary file, then renamed over the original. That's one reason why people think ctime means 'creation' time, since for the types of files people hand-edit it really is.

It's almost like people have discovered that you can have modularity and shared functionality in a different way than artifically seperating storage layers and throwing away important data at each layer boundry.

It's an issue with any CoW filesystem being full - in order to delete a file, you need to make a new copy of the metadata that has the file removed, then a copy of the entire tree leading up to that node then finally copy the root - and once the root is committed, you can free up the no-longer in-use blocks. At least, as long as they're not still referenced by another snapshot.

The alternative is to rewrite the metadata in place and just cross your fingers and hope you don't suffer a power loss at the wrong time, in which case you end up with massive data corruption.

I've filled up large (for home use) BTRFS filesystems before - 6-10tb. The code does a fairly good job about refusing to create new files that would fill the last remaining bit so it leaves room for metadata CoW to delete. The problem may come from having a particularly large tree that requires more nodes to be allocated on a change then were reserved - in which case the reservation can be tuned.

BTRFS isn't considered 'done' by any means. It was only in the 3.9 kernel that the new raid5/6 code landed, and other major features (such as dedup) are still pending. It's actually very encouraging that a work-in-progress filesystem is as solid as it is already.

Self-reinforcing. Start with an empty 'related' database and a huge library of books. The first person browsing leaves a distinct fingerprint on the database, because the second person 'sees' the first's trail, and follows it. The joy of a traditional library is there is no path trod into the browsing experience, every discovery is fresh, no matter how many times it's been done before.

To get an idea of how it works out in practice, check out the completely bonkers amazon reccomendations for super-low-traffic items.

There's a lot behind this comment that is really important. Sure, you "can" do the same thing with an electronic view of recently returned books - but you have all sorts of crazy privacy implications. If you read on-site in a traditional library, you don't need a name associated with a book to browse the shelves and see what's related, you can anonymously put it in the returns without your identity ever being associated. And it's a lot more casual and discovery-oriented: A lot of times I would walk through a section dedicated to fiction, grab something at random and read a bit of it to see if I liked it. If I did, I might finish it there, or check it out. Automated delivery is very goal-oriented: I know I want THIS book, go fetch it for me and ignore everything else next to it. At best, it could have a 'related reading' list which rapidly becomes a self-reinforcing subset of the available books.

I think we should digitize everything - but keep traditional libraries traditional. Just don't fill an entire moon with books.

The thing is, with a modern design around the capabilites of video cards you have a much cleaner way of grabbing the individual window frabebuffers (textures? Not sure what they're called when compositing) and streaming them as an image remotely. You can couple that with some hints ("Scroll down 50 pixels") and get a very efficient network protocol - that the applications can safely ignore. As far as they're concerned, everything is happining locally, and they don't care that instead of getting scaled and wobbly they're being encoded and streamed across a lan. As long as you forward the same input events, what's the difference?

Not really. It's a much more painful to try to remote a browser vi X11 on a LAN than it is to VNC/RDP into windows. The main issue is that while bandwidth has improved greatly from the early days of X11 there's only so much you can do about latency, and nearly no apps are written to handle that network latency well. It's actually a much better idea to let them do all their tiny operations on a local framebuffer and stream the whole image as a big blob of pixels. Back when X11 was designed, it was basically unthinkable to throw pixel streams across the network, because bandwidth was such a scarse resource (even on a LAN, 10 MB shared collision domain? Ouch.). When your design constraints change so drastically it's only natural that the optimal solution is no longer the same.

Why, it's almost as if your taxes were based on multiple people working and you're looking only at your own salary. Did you pay more in taxes than you paid everyone in your company combined?

You're free to move to Somolia, where there's no taxes at all; but I guess you like stupid things like "Roads", "A functioning legal system", "Not being murdered", crap like that. Guess what? You get to pay for it, same as the rest of us.

Do you not know what escrow funding is? You get a percentage of the promised funds as you reach each milestone, with some portion being released immediately. It's not like a product where the escrow is held until you deliver.

If only there were a -1 WRONG button.

That's for Pwn2Own, which google is also not particpating in. Pwnium (what this is about) allows pre-written exploits.

I've got a few issues with this whole thing.

First, since the main complaint against megaupload was they were doing data-deduplication, at what granularity do you need to dedupe before you become infringing?
This is problematic because of #2: DMCA complaints lie. Like, the majority of the time. They keyword search and send takedown requests to files with names they don't like without ever seeing the content. So, if a service dedupes content like that all you need to do is upload something that's popular and name it "Britney Spears #1 hit poop" and get it DMCAd. Troll heaven.
Third, their "Paying pirates" is a bald-faced lie. Putting up content you own and getting paid for it was a way to make money for independants of all stripes - modmakers, musicians, amateur directors, etc. The fact that some people made money with things they didn't own (And don't forget that takedowns were honored, dispite the misleading wording of the indictment.) doesn't change the fact that it had a massively non-infringing use. As for their "evidence", try to find one ISP where employees don't talk about using their bandwidth to download not-entirely-legit content. It's considered a job perk of a fat pipe.

Lastly, The MU takedown had the intended effect. Filesonic died early this morning as well. If you don't own your own hosting, good fucking luck sharing anything you've created with the world. There's a lot of minecraft mods that are just gone now, for instance.

Add the following URL to AddblockPlus (with http, not hxxp).
hxxp://meta.wikimedia.org/w/index.php?title=Special:BannerLoader&banner=blackout*

Since you're needing this comment, you're already aware of SOPA and you want to use wikipedia today. Also, pass it on to anyone who complains. They're aware of SOPA and they get to learn about how they'll need to use obnoxious technological workarounds to legislative damage.

Also, today's a great day for creative wiki editing.

actually you picked a terrible example - Crysis 2 was written for consoles and ported to PC, which means it's didn't even support DX11 on release, and it certainly doesn't have a DX11 optimized workflow. The levels are smaller, more confined, the textures have to fit in console RAM, etc. You're not missing
out on anything by not having a new card.

Wait, I missed the announcement where nVidia dumped all their hardware internal documentation on the Xorg guys and said 'have fun'?

There _IS_ no competition for linux until that happens, there's ATI and some proprietary vendor I never think of when buying cards.