*slaps forehead* I feel dumb now.

I knew that, but I don't usually use a mac keyboard. I still remap the window and alt key so everything is the same position as on my macbook pro, though, so when I was responding I just looked down.

But yes, it's cmd not alt :)

Alt-` is what you're looking for, I think.

I think the apple way of doing this is to hit the expose key, then use the keyboard if you're looking for a specific window.

I really really love the dock concept -- why do I have to have a firefox window open to keep downloads running? Why do programs have to go out of their way to manage system tray entries? If a program has a system tray entry, why are there *two* ways to minimize that app's window? All of this is solved by the dock!

But, it's taken me forever to get used to the two level window switching that you're talking about. It still tends to feel clunky. I guess I could use the expose key, but it's not as conveniently placed where my left hand can reach it in the resting position. I can't help but think there is a better way.

Yeah, we a better discussion of this further down... my blanket statement here wasn't well justified.

I have of course had exactly that experience, on both sides, but never something that I felt was patentable. This might be that most of my programming work has been very tightly coupled to mathematics. I feel math shouldn't be patentable, so I never feel that my programming implementation of it shouldn't be, but I understand how someone could feel that way. A good counterexample to what I was saying would be the RSA patent... shorter than most, and very clear on what the novel algorithm is and what it's useful for right up front.

I should have qualified what I was saying and emphasized *most*. Granted I haven't looked at *that many* patents, but all the ones I've seen are just vague descriptions of what some system looks like at the interface level. There was one that patented... something to do with slot machines that were networked and shared data. They weren't networked in any novel way, the patent didn't describe anything particularly novel they were doing with the data, it just seemed to patent *the idea of having slot machines in a network*.

That sort of patent doesn't show you how to do something, it just makes a land grab. I think we're in agreement about the USPTO and probably about patents in general, but I shouldn't have made such a blanket statement without qualifying it.

Thank you for expressing my thoughts much better than I can :)

Oh, trust me, I read past the claims section.. these were two patents that caught my attention because they were on slashdot previously, and I wanted to see what the average software patent looked like. They always look like a laywer and a programmer sat down and translated every diagram and requirements document into legalese.

I mean look...

That's in the "DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS" section. That's like a back of the box product description. I know they go into more detail later, but it's always meaningless minutiae about what part of the website connects to what, or the fact that such and such field would be retrieved from a database. Well duh. Were these bad examples? Is this what a "design patent" looks like?

Even the figures are mostly grossly simplified networking diagrams and *screenshots* of the web interface. I mean, come on!

The second patent looks better, but the novel parts look to be in exactly how they tie information from different logged in users. It definitely seems *interesting*, but it sounds like it's interesting because their problem domain is unusual, not because the concept is amazingly novel independently of that.

Don't get me wrong, the flow chart comment aside, I think I qualify as a someone reasonably proficient in the field. Up until recently my job was primarily doing aeronatic and numerical processing code for the DoD and designing a library and application for easily integrating aeronautic flight models for the purposes of weapons effectiveness and defensive analysis in dogfights primarily for use at Top Gun and Fallon.

And these (well, mainly the first one) sound like nearly useless 50,000 views of a system filled mostly with information that you could garner just from *using* the system.

You know what seems like a straightforward patent? http://www.google.com/patents?vid=4405829 RSA is the most deserving software idea I can think of for being granted a patent... and it seemed to stall web commerce for a decade. It was absolutely a deserved patent... but the barrier to entry for software is just *so low*. It actually seems more valuable to keep the barriers to entry into software engineering low than what encouragement is gained even very deserving patents. Oh, and a very similar idea was invented 5 years earlier in Britain, but was kept secret due to national security concerns, but that's besides the point.

I'm not saying that you do it, but so many people I've heard of that work with patents view patents as something of a land grab. You know -- it doesn't matter if you thought of it too, I thought of it *first*. Rather than how they're intended, to help promote progress until the point that they no longer do, and then stop.

Hey, I never meant anything personal, sorry if I offended you.

Most patents I've read look like this:

http://www.google.com/patents/about?id=RX94AAAAEBAJ

or this:

http://www.google.com/patents?id=Ay99AAAAEBAJ&dq=7139761

Lots and lots and lots of description of the minutia of any complex system that really wouldn't help me unless I was trying to do something that was exactly the same, and in the exact same environment, and maybe not even then. And that is buried in tons of description of basic computer science or engineering concepts like what a WAN or LAN is and what eeprom is.

Maybe you can make sense of that and gain something from it. I guess I could too, in the sense that they're like very obfuscated requirements documents.

But, you caught me, I have not had enough patience to actually read an *entire* patent like these before.

And, of course, they're not all like this... but man are there a lot of them.

Oh, and I thought the European Patent Convention required (at least recently) that the invention make significant contribution in an otherwise patentable area, i.e. no *pure* software patents. I might be wrong on that.

But yeah, I can read a flow chart, on a good day, I can even *draw* a mean flow chart. And I'm sure you can too :)

I think that *Europe* is pretty civilized, and they specifically exclude business method and software patents still, as far as I know.

And like fritsd said, we don't see source code attached to patent applications, and you go read most software patent applications and tell me if they *help* you develop anything.

That sounds pretty ridiculous... what you would have created would most definitely be a derivative work, which would have copyright protection.

If you clean-room reverse engineered the program down to the last pixel, wouldn't there be a copyright claim to the layout and specific appearance of the program?

If the program looks different, is coded independently, but performs the same algorithms and functions, well that's called *interoperability*. If you don't understand why that's important, then we'll never ever agree.

I'm talking about a specific steering wheel design, you're talking about *the entire idea of a steering wheel*.

Somebody else came up with the idea of html, markup, tying a scripting language to said markup, asynchronous updates on the markup, and everything else we're using to communicate here, and there are probably patents covering most of it. Do you think that only very large companies with very large armies of lawyers should be writing web pages?

I'll give you that there are probably a select few areas where some patentable idea could be expressed in software, but the barrier to entry is so low in software development, that they all would have been implemented anyway. Patents are not a land grab or a right, they were only supposed to be used to increase production and knowledge. They don't, they waste huge amounts of energy, which is why people around here get so frustrated (and make silly arguments).

Of course, I'm not adding anything new and you probably knew all this, so I'll shut up now.

I don't know if you're having the same problem as me, but my problem was that the install wouldn't reboot after installing karmic. I *think* kernel 2.6.31 has some problem with Parallels (or vice versa) that keeps any disk from being recognized. I get dumped to busybox, and there are no /dev/sd* devices :/

If you're having the same problem, just boot with kernel 2.6.28 from Jaunty.

The two biggest problems I've had personally are https://bugs.launchpad.net/ubuntu/+source/xorg-server/+bug/403339 and https://bugs.launchpad.net/ubuntu/+source/cups/+bug/465916

Which in the grand scheme of things are pretty minor. The first one is really annoying if you play WoW in wine, because you have to manually turn off keyboard repeat :/

Other than that, I've upgraded 3 machines without problems. My parallels VM upgraded to karmic doesn't detect any drives unless I use the older kernel, but I'm 90% sure that that's not a bug in ubuntu.

But *man* that xorg bug is annoying.

See, changing bits in the linux kernel *and distributing it* *and claiming it is your own* is an ocean of difference. We're not talking about that.

When you sleep in places you're not supposed to, you're infringing on someone else's physical rights.

When you ride a train without a ticket, you're using up resources you're not paying for.

There is a good reason copyright, patents, trademarks are treated differently, and also much differently than physical property, so stop confusing the issue.

I know you've heard this argument before, because you use words like "freetard". Stop trolling.

What I'm doing might be legally sketchy, but I, and most of the world, certainly doesn't think of it is as morally sketchy. To me it's about the same as saying that playing a cd with too many people in your house so it counts as a public performance. Might be legally sketchy, but only because of *a failure of the legal system*.

You seem like a good, moral person... probably someone who makes a living selling creative works. I want you to keep on doing that, I'm glad you do. I'll make sure that if I copy your works, I compensate you fairly.

If you then tell me it's illegal for me to run your code through a debugger, or reverse engineer your file format... well, that sounds like it's *you* who is the one infringing on my rights.

Of course, we don't know, because the law in this area is so vague and uncertain that we're all violating it somehow. If you develop software, congratulations, you've violated someone's patent.

I know you've heard this before, just know that you can be a "freetard" and not a thief at the same time, okay?

Why the animosity here? Modding is not something which infringes anyone else's rights. We're not talking about breaking copyright, we're talking about *changing bits in a piece of hardware that you own*

You may very well be right about the legality of such a thing, but there is a reason you can't sign a contract to take away certain inalienable rights. EULAs are a one sided, non-negotiable contract, but IANAL... I'd just *like* them not to be enforceable.

However, there are very few that I've seen claim such a strong moral high ground against *changing bits in a physical device you own*. I have a lot of problem seeing that as any different morally from rewiring your car.

You think very highly of ownership of land and property, how is this any different? How would you feel if I wanted to sell you a piece of land that with an agreement that says the specific way trees are arranged on this land is my personal intellectual property, and any change to the arrangement of such trees is a violation of our agreement, at which time you will relinquish the rights to the land.

I own a wii... and I've put the homebrew channel on my wii. I've *never* pirated a wii game, but I have set up the homebrew wii SDK and am currently writing software for it, because I think it's an awesome piece of hardware and I'm glad I gave Nintendo my money.

I suppose I'm an evil, contract breaking vagrant peeing on poor Nintendo's property now.

  I guess I just can't wrap my brain around the concept that wanting to tinker with something to create something new is *selfish*.

In the US, facts are not copyrightable (I guess that's what the OP is talking about).

The data in question would presumably be US government secrets, which are facts. IANAL, so besides unlawful use of a computer system... I guess they'd charge him with unlawfully obtaining secret or confidential data?

I have no idea how that works. Having held clearance, it is really really *really* illegal to *reveal* government secrets when you have a clearance. I have no idea how that applies to learning government secrets outside of that, especially when you're not a US citizen and do it outside of US borders.

Copyright doesn't really come into play, I think. Besides, US documents aren't generally copyrightable anyway (except for crazy contractor situations).

I'm not saying it's impossible, just that it's *much* harder to understand than simply writing it recursively.

Let me put it this way... suppose in C++ you have a template instantiation defined dimension matrix...

In order to do operations on the matrix, you have to iterate through an arbitrary number of dimensions, then loop over (a subset of) the length of the matrix in that dimension. It's like a for loop inside a for loop, but the number of for loops deep has to be *runtime defined* (well, in my case template defined, but you get the idea). Without recursion you end up either writing gotos, or some other funky math where you determine the overall size of the n-dimensional matrix (dimensions multiplied together) and then loop through that index, and figure out the actual n-dimensional coordinates as you go (messy).

Here it is in code...

        template<typename OpType>
        inline Array& eval(OpType op, IndexType index = IndexType(),
                        size_t dim = 0) {
                for(size_t i = 0; i < extent(dim); ++i) {
                        index[dim] = i;
                        if(dim == Dimension - 1)
                                operator()(index) = op(index);
                        else
                                eval(op, index, dim+1);
                }
                return *this;
        }

where IndexType is some vector unsigneds of size Dimension, extent is a member function that returns the size of the array in that dimension, and operator() retrieves the value at the given location.

This is a (modified for clarity) member function of a class I have called Array, that has template defined dimension. I'm filling the Array by calling the passed in function on every location in the Array, and assigning the result to the location.

You *can* write it without recursion, but it's even more confusing than what I wrote.

There are probably better examples of exponential time algorithms that are clearer, but they make my head hurt.

By the way, at my old job, our simulation had to interpolate lots and lots and lots of experimentally gathered data in big tables (like an 11 dimensional table for infrared signature data). Writing an arbitrary dimensional interpolation library was the *simplest* option :)

How would you implement an algorithm that is O(2^n) then?

I wrote a template library for arbitrary dimensional matrix math and interpolation, and since many operations were O(2^n) for matrix rank (like linear interpolation), the *only* sane way to implement them was using dual-recursion.

Anything else, though, and I'd probably agree with you. It's easier probably easier to read a loop than it is to read a tail recursive algorithm for anything but purely mathematical problems.