I wonder how hard would it be to obtain or build such an ADS-B receiver. This is exactly the kind of thing I could install on my balcony

Unless you know these things already, this also happens to be a good "what specs should I actually (not) care about" quickguide article.
http://gizmodo.com/5669331/why ... -specs-are-total-bullshit

I understand that machine is a laptop, right? Laptops usually have a very bad cooling which can barely keep the hardware from melting and a bit of dust buildup in the fans can sometimes render it almost useless. 83° C when playing CS is a lot considering that your CPU is a 1.6 GHz Pentium M.
Try to take the machine apart and clean the fans. The X300 Mobility drivers are not listed on the AMD's website, but I guess that X300 drivers will work just fine with it. AMD has dropped X300 support about 2 years ago though, so the latest drivers I could find were Cats 10.2.

As long as it's a pure eye candy thing with no effect on the cars' performance, it's useless.

Quote from J@tko :

Definitely now failing my exam tomorrow - SO COOL! Thanks for sharing

True that, I am so screwed...

Seriously, this is incredible...

WINE is absolutely necessary to get LFS dedicated server working. What Linux distribution are you using?

The simplest way how to do this is

The server itself should run fine through WINE.

Running the server without X is a little trickier, you really want to read the wiki for some more info...

This looks a lot like a graphics driver problem.

I see. This would probably work, but there's a little catch. Logitech wheels indeed use a ramp force FFB effect to create the soft-stop (DFP definitely does, not sure about G2x/DFGT). However, when the wheel is turned past the soft-stop, it still transmits the axes position data. If LFS limited the wheel range just by FFB, turning the wheel beyond the soft-stop would turn the wheel in game, making LFS adjust the wheel's calibration which in turn would lead to all sorts of unwanted effects.
When the range is limited via Logitech driver, the driver takes care of this and clamps the X-axis range accordingly.

Scawen once commented on including the extended abilities of the Logitech wheels into LFS. He didn't like the way he'd have to include the Logitech API into LFS's code so he ditched the idea.
You can use 3rd party InSim apps to adjust the Logitech wheels' range from LFS.

Quote from flymike91 :

I thought it was silly that speedway would call an American stupid for not supporting the current regime in his country and thereby exercising free thinking and speech. How does voicing opposition Obama's policies prove every stereotype about dumb Americans? It's as stupid as the stereotype that all Germans are Nazis. That was my point.

As for patent law, if you invent something, why should other people be able to make money selling your idea? Say a Norwegian invents a new computer gadget and produces it in his country where wages are very high because it supports his people and national economy. Then someone from China copies it exactly and sells it for $1000 less, using exploited workers in appalling conditions. Obviously most companies outsource labor to China anyway, but when a company pays R&D for new product, they expect to make a return on that investment, but whoever rips off the idea has to pay nothing for R&D, and can sell the product far cheaper. I don't see a difference between real products and software because some real people invented it either way, and deserve to profit from it, as does the company they work for.

Getting rid of patents can happen someday, perhaps in a future where we can simply describe to a computer the product we want, and advancements would come naturally as products are constantly improved upon by the end users. We are not at that stage yet. In my opinion, expecting researchers and engineers to work hard to make great new products, only to give away their best ideas for free so other companies can profit is wrong. Who is to say some seemingly menial computer code wasn't the best idea some guy ever had?

The idea behind patents is perfectly fine and reasonable, problem is that the idea is a couple hundred years old and thus doesn't really work with today's state of technology. When Oracle can base its lawsuit against Google on a few lines of code (written in both cases by the same guy!) or when Apple can patent a finger swipe as a screen unlocking gesture, we're not using patents to protect the inventors anymore. Patents have become weapons in the war of competing companies to gain a little edge on the market. Take MS as an example, they're making considerably more money through Android than by Windows Phone sales - talk about protection of inventors.

Today you're allowed to patent almost anything no matter how stupid or simple, it's nearly impossible to create a competitive electronic device and put it on a market without the risk of violating someone's IP. It's usually not like the companies are copying from themselves, they just happen arrive to the same solutions because those solutions are the way to go.
Ten years ago you needed a team of scientists and engineers to come up with some technological advancements, now you need a battalion of lawyers just to make sure your product doesn't violate someone's patents...

Matroska is a container, not a codec, the patent mentioned above is about the h.264 video codec. But that's both off-topic.

It's almost poetic justice, but I'd much rather see the IP and software patent restrictions reviewed, banning Xbox 360 in US will only hurt the end customers and it won't resolve anything...

The weather forecast suggests there might be some rain during quali and race, could that mean Schumi's big comeback?

Let me see if I understand this: You forgot your web password and the e-mail address the account is registered to and lost the receipt?

If I get this right, you want the devs to check what e-mail address you entered when you registered your LFS account, right? I'm afraid that the devs can't tell you that because of security reasons.
I sent you a PM and unless you switched that option off, you should receive an e-mail notification in the mailbox that's registered with your LFS account.

I see, but abusing the staticness of class members is generally considered a bad practice in any language. In C++ you might be better off doing it like this:

Option 1 - global variable (bad idea too)

Option 2 - shared variable (much better)

In that case, wouldn't it be better to make the server object global and access it like that or share a pointer to it?

"using namespace foo" isn't the same thing as including a header. When you do #include <list>, you tell the compiler to read the header file which contains the implementation of the STL List. "using namespace" is more of a convenience thing. The STL List is defined within std namespace, so if you don't have "using namespace std" if your code, you have to use "std::list" instead of "list".

BTW, is having a list of connections static a good idea here? I of course have no idea what your intentions are, but do you really want to have a common list of connections for all instances of class Server?

Quote from mapa17 :

I thought the udev rule was supposed to generate a link to the current event handler.

That's my understanding of it too, however, there's a catch. LTWC literally takes the G27 away from the kernel's control, does stuff with it and then returns it back. Logitech wheels are handled "properly" since kernel 3.2, so when the kernel gets the wheel back, it reinitializes it again which leads to new /dev/input entries being created. Perhaps that somehow confuses udev so the link it creates points to the old invalid entry.

Does it work when you remove the LTWC part from your udev rules?

As of kernel 3.2 you should not need LTWC at all. The wheel should switch to the native mode automatically when it's connected and your dmesg log suggests that it really happens.

I can't test this at the moment, but because of the way LTWC sets the wheel up, it's possible that it's fighting with the kernel driver over the wheel which leads to an invalid /dev/input/G27event entry.

I suggest you try this: Remove the LTWC entry from udev rules, plug the wheel in, let it calibrate, run

(the "xxxx" value might change, you can read the current value from dmesg)
and try the fftest again.

CZE - SWE 4:3, how about that?

(A pozajtra bude nad Tatrou sa blýskat )

Standard InSim port is 29999. You can enable InSim interface either by typing "/insim=port_number" in LFS chat or by launching LFS or LFS DEDI with "insim" parameter, for instance "LFS.exe /insim=29999".

Quote from frotto110 :

insim programing is hard, dont even try. use finished

It ain't that hard, especially with helper libraries like InSim.NET, but sufficient programming knowledge is necessary.

What are the reasons for not having an open beta? Is there like some sort of NDA regarding the Fanatec SDK? I'm not sure how many people are willing to go through the process of asking you directly for the app. Judging by the viewcount on YouTube people are certainly interested in this...

Quote from cargame.nl :

You are getting slightly annoying now.

Take a good look around on the attachment. What do you all see?

Little hint; processor usage (one core) and GPU load.

What does it tell you?

It tells me that
A) You don't know how to properly interpret CPU load in Windows Task Manager with a multicore CPU
B) You're probably using something like 8xSSAA and high AF settings

LFS utilizes one core at approx 80 % on a 2.66 GHz Core2 with full grid of AIs on BW GP. GPU of a sorry ass 9800M GTS is loaded at about 50 %, but it's worth noting that the GPU is clocked down to 400 MHz (from 600 MHz). That's with 8xMSAA and 16xAF.

Quote from edge3147 :

under 25% of cpu usage is hardly CPU Intensive. GPU does more work than the CPU with this game.

You have a quad core CPU by any chance?

Quote from cargame.nl :

Exactly, my i7 1.6 Ghz mobile on ONE core is producing 10% activity.

Browsing the internet is heavier. Where does this hard to kill legend story come from that a GPU is not important for LFS? You all crazy here, you know that?

Oh I can tell you. It stems from the fact that - apart from vertex shading - LFS uses a fixed rendering pipeline. There is basically nothing for the GPU to do, no advanced pixel shader programmes, no ultralarge textures to transform, no realtime lighting, no HDR. LFS runs just fine on i915 or even through VirtualBox!

It's not like it really matters, I just want to save those folks who are stuck with a Celeron D from buying a second-hand GPU only to find out that it makes no difference. (unless the card you're replacing is a GF 440MX).