Does 802.11ac slow down in the presence of older networks like 802.11g slows down in the presence of 802.11b clients? All you needed was one old client present on the network for all transfers to slow down by 50%.
Yeah, you could set the access point to be 802.11g only but even then the 802.11b would keep interfering with the access point...
Actually, 802.11g only does NOTHING to prevent this. It's not the 802.11b client on the network, it's just an 802.11b client on the same channel.
802.11 in general is a protocol used to communicate data packets with some encoding over wireless. It dictates certain things like RTS/CTS clearance frames and CSMA/CA. It has a concept of "reservation timer" which means "even if you cannot understand the data, consider the media busy for N time".
802.11 actually cooperates, and clients on different networks actually do communicate with each other - just because a frame isn't destined for the network doesn't mean the client or AP doesn't react to it. (Interference does happen from the hidden node problem, though).
The thing with the 802.11g thing is easy - the 802.11 spec identifies that the preamble and header are sent at 1/2Mbps (802.11) speeds, then the body is sent at higher speeds (the reason for this is so every 802.11 client can consider the media busy). The trick with the 802.11g is that it could send the 802.11 headers at 802.11g speeds, but any 802.11b or lower client had a chance of interference so G-protection kicks in that forces headers to be sent at 802.11 speeds. This happens in the presence of ANY 802.11 client that appears, attached to that AP or not. After all, an 802.11b ad-hoc network could be using the airwaves as well, and interference just means no one's packets get through.
The reason you still get intereference is hidden node - you will have clients transmitting randomly because they can't hear the other transmitting node, and the receiver can hear both (it's alleviated with RTS/CTS - which would silence both transmitters). Just once you hit a certain density, there's enough that even that can't protect it due to more hidden node issues.
While the PS3 remains the best console for japanese games, the Xbox 360 is a much better console for western audiences. The Wii did sell better, but who is still playing their Wii? The console was just a gimmick.
Actually, I think the Xbox360 outsold the Wii sometime this year in total units (the Xbox360 has been outselling the Wii for the past 18 months, so it's been only a matter of time before it overtook the Wii). The PS3 has consistently been third.
And the primary reason people are developing for Xbox360 first is simple - it's a more "standard" platform - a triple core CPU (2 threads per core) plus a GPU, just both sharing 512MB of RAM. The PS3, when considered as a "standard" platform is inferior - dual core (two threads per core though) with 256MB of system RAM and 256MB of GPU memory. The only way to tap the power of the Cell is to get deep down and program the 6 SPUs, which requires a core to manage the data flows for them.
Expecting to plug in to AC is rather retarded on an aircraft as it would require large inverters to power a full aircraft, and then all the inverters are going to do is power your converter thats going to bounce it back down to essentially the same voltage as it started out as.
Actually, the plane sockets provide very little power - usually under 55W or so. For a small laptop, not a big deal, but a larger ones will often require entering special modes that disable battery charging (I believe the Apple one tells the laptop to not charge the battery, for other laptops, you should remove the battery). Otherwise there's a chance you can trip the power outlet and it won't provide any more power until it's reset.
They're not providing full 15A to the AC sockets. Just barely 0.5A. Though the airline sockets are the same, really - 28V @ 2A.
In my opinion the real challenge hasn't even started yet: marketing the final product. If they can't convince developers to create exclusive content as well as convince the average consumer to buy the device they probably go down soon after the release. If sony is failing to successfully move the Vita out of the shelves even though it has the playstation brand, is homebrew friendly(PS Mobile SDK), has hardware comparable to the latest flagship smartphones and over 100 real games(almost a 1000 if you also count PSP and PS1 downloadable titles) due (mostly) to the insufficient marketing that they have right now, then the OUYA better be prepared.
No, the real challenge is figuring out a way for purchasers to wade through the eventual morass that it will become.
Remember, it's an "open" console - you buy it, you get the SDK with it, and you can make games. Which means you're going to get a bunch of crap apps as everyone sees dollar signs and tries to scam a few bucks from people for some sudoku or crossword or weather or other app.
Since it runs Android, there's a ready set of developers, so convincing people to develop for it is easy. Their questions (and everyone else's) would be how can their apps be found? Will anyone browsing the store have to wade through pages of crap to get to the gem? If someone unknown makes a great game, are they going to be drowned out by the hundreds of other apps also released?
Apple, Google and Microsoft has tried, and failed horribly. So bad Apple actually bought Chomp to help with the problem (and searching the Play Store can be frustrating, ironic given what Google does...).
I wonder what would happen if US colleges (or even earlier in our educational system) let students have free reign, and really specialize. If over in India this man had been nurtured in college, and allowed to stay in math courses (or even better conduct his own lines of study), might he have had a more enjoyable or productive life? If we recognize genius and cultivate it, what might grow in that garden?
They do let students specialize - students are free to take any course they want (as long as the prerequisites are met). There's nothing limiting a student from taking all higher level math courses, other than perhaps wanting a piece of paper at the end.
And no, you don't have to do university in 4 years. You can do it in 5 or more years, taking all the classes you want as long as your finances hold out.
The only real reason they force all students to take courses in other departments is to round them out. An engineer is useless if he doesn't know how to communicate his ideas to others, and perhaps taking some liberal arts classes gives him the tools he needs to relate and communicate. Hell, what's the point of proving some hard theorem if the only people who can understand it are people just like you - and no "lesser" mathematician can comprehend it? And really, if you're such a hot shot, two semesters of that should make for an easy A.
Hell, the programming equivalent is "bad code". Only because the original programmer has failed to communicate to the next person who maintains it through the code. (Hell, a lot of crap code doing stuff stupidly could probably be resolved by communication).
I know your trying to defend Apples "lets gouge our customers policy" by limiting customer choice (and competition) to Apple on its (not your) phone...but to do do so I think you need to understand that on Android you have to actually go into the settings and *enable* his voluntary, and have to agree to a warning screen...Apple users are do desperate to have this functionality they "Jailbreak" Apples phone, even though Apple have attacked their customers for doing so.
And practically every US Android user has it checked.
Why? Amazon. Or if you want to install those cheap Humble Android Bundles.
Which means yes, it's a superior system, except it's broken because it's an all-or-nothing. Why should I have to enable it to install an app from Amazon? (Which to be honest is probably why AT&T was forced to put the option back in). Why can't I just enable it for Amazon? Or why can I just install the Humble Bundle apps?
In fact, why isn't the option more like "Allow installation of unsigned apps"? Then Amazon and Humble Bundle and AppsLib and other stores can provide their own certificates for the user to install so that store is "trusted" and they can't install some random APK from facebook because their friends call it cool?
TVs should be a display and that's it. Give it the brain to decode HDMI signals and the tuner for over the air digital but that should be it. Every time a manufacturer tries to put more then that in their TV it just makes it worse. I spent a day at best buy looking at TVs this month, not a single internet enabled had an interface that doesn't make you want to rip out all your hair. Added features shouldn't break the main ones.
The problem is the "brain" part is getting smarter. What used to take all of whatever it was to control the scaler and all that, barely takes any at all because the video processors are coming with GHz class ARM processors (because really, making them at 200MHz isn't much cheaper) so the control of the inputs and video scalers and switchers and all that, are leaving a lot of cycles free.
Enough so that while my first TV probably ran a very limited RTOS, my upgraded TV (it is "connected" in that it has a... web browser) runs Linux. And actually put my embedded platform I was working on to shame (my TV - 128MB of RAM, 128MB of flash, 533MHz ARM11 (this was 2009). The platform I used had 128MB of RAM, 64MB of flash, and was a 233MHz ARM9 and was doing GigE networking and all that). Today's smart TVs probably are coming with single core Cortex A8s or dual core A9s because the SoCs with them are so cheap.
Heck, many of them are becoming mixed-signal and handle HDMI(+HDCP), component, composite and S-Video natively just because the SoC guys were bored, and tossed in A8s/dual A9s because they also wanted to sell them as tablet SoCs. With camera interfaces and all that.
Basically, TV manufacturers are given newer and newer SoCs for their video processors with more and more capability because it's so cheap to do so. Unlike say, a calculator... which I'm sure for the BOM cost 10 years ago and all that , probably could sport A8s and A9s with impressive displays and memory capacity.
There seem to be more stories about tech companies suing each other than anything else on/. anymore.
That's because, in the US, that's about the only new technical development that's happening. Anything actually productive has fled to parts of the world where such things are still legal.;-)
Or because most of the other news articles generally deteriorate into flame wars.
Facebook announces something! Comments: Facebook sucks. It steals your information, blah blah blah.
Google announces something! Comments: Google's tracking you. It steals your information. This is why the cloud sucks. Damn you for monetizing. blah blah blah. android rules.
Microsoft announces something! blah blah monopolist blah blah windows 8 sucks blah blah window ssucks blah blah android rules blah blah linux rocks blah blah
Small company announces breakthrough! Comments: not commercially ready, too expensive, NDAs, patents, blah blah blah
New game reviewd! blah blah EA sucks blah blah Steam suchs blah blah DRM blah blah consoles suck blah blah PC mouse and keyboard blah blah.
About the only thoughtful commentary happens in the patent lawsuit threads, like explanations of the rounded corners design patent (+grid of icons +dock). Though they start degenerating soon into "android rules, apple sucks blah blah blah samsung's awesome blah blah htc sucks blah blah motorola locked crap blah blah"
Note that sales of AR-15 type rifles has zoomed way up since the Dems started talking about banning them.
Which is giving a nice boost to gunshop owners as well as gun manufacturers.
Hell, I don't own an AR-15 or analog, but I'm considering it, just because they might be banned next year.
In other words, the proposed gun bans are having the opposite of the intended effects already. Too funny (but predictable)....
Or, it's called "how we're getting the US economy restarted". Given the exact same thing happened the instant Obama was elected in 2008 - gun store owners called it the "Democrat Economic Recovery Plan" because everyone was so afraid of the new laws that weren't introduced that they went out to stock up. Which gave a very nice boost in spending.
Looks like the same thing is happening again. Perhaps all they need to do is raise gun taxes and gun ownership taxes and the whole debt would be paid off rather quickly.
Which leads to oddball situations like the democrats constantly getting budgets balanced because of the economic boosts from gun owners worried they would take the guns away, to the deficit budgets when the Republicans come in because the gun laws would be loosened so there's no reason to buy a gun now, but wait...
I pirate. Most of what's "e-gifted" is just supporting the entertainment industry; and I'm loathe to support them until they clean up their act with all the DRM crap, manipulating the market prices, and throwing people in jail for trivial crap, as well as co-opting our entire legal system and feeling entitled to profits.
So... why pirate? Why not ignore it and enjoy something else?
You do realize that you're just encouraging the problem, right? Because they create something you want, and instead of giving them what they want in exchange, you say it's worthless, and then expect them to take it lying down. I'm sure if someone pirated the Linux kernel (by say, not releasing the source code), you'd be upset. Yes, you can pirate Linux by ignoring the GPL and violating the de-facto copyright law. You can pirate most open-source that way because by not agreeing to the freer terms in the license, it reverts back to "all rights reserved" and what copyright grants.
You're better off ignoring what the entertainment industry produces and enjoy the works of the indie bands, moviemakers, game creators and what not. Because the industry won't change unless they see success can be had without draconian measures, and piracy is not the way to do it.
Hell, if you want to pirate, go buy the DVD and download the DRM-free version, but that requires supporting the industry. And while they may not be entitled to profits, you're not entitled to enjoy their works just because you don't agree to their terms.
There's plenty of free and open indie entertainment out there that's DRM free and cheap to enjoy. Why not encourage those creative people by enjoying their work?
No crap they sold out. Branding is now all they have left after losing their cases preventing competition http://en.wikipedia.org/wiki/The_Lego_Group#Trademark_and_patents The only way they can continue to compete with companies like MegaBlocks is to have exclusive rights to Star Wars or Harry Potter etc etc. Honestly how much justification can you have for $50 USD plastic blocks unless you are the only game in town with whats "cool".
Well, the funny thing is, I have a tons of Megabloks (because they have the Halo license), and I have to admit, the quality of the bricks... suck. I don't know if it's the plastic, but it certainly doesn't feel like the standard ABS that Lego uses. It feels cheaply made and constructed and while the outcome is good, it just seems a lot less satisfying.
Not that they're that cheap anymore - I've seen some kits for $200 that didn't look too complex (certainly not half as complex as the 8000-piece Millennium Falcon Lego has).
Funny - we know this to be true, yet American continues to export jobs and technology to China. Idiots in CEO suits, idiots in Board of Director suits, idiots in Consultant suits - the whole lot of them have read time and time again about the Chinese stealing IP. But, they decide amongst themselves that it's best to take advantage of all that slave labor in China. All that IP given away, in the quest for free or almost free labor.
And what IP would that be? A GoPro isn't exactly an innovative thing. It's just a plain old 1080p30 camera in a rugged case with a bunch of mounts to hang it in various locations.
OK, fine, GoPro has some design patents, but we know how/. likes the "rounded corners" design patent idea. (Yes, I know, it's rounded corner slate with a grid of icons and a dock that's static as the grid's adjusted).
But other than that, it's just a lens, a sensor, and a processor wired together with some software to make it usable. Add a little expansion header for a display and you get the LCD backpack, a little SDIO header for wifi, etc.
The new Hero 3 has some neat features like 720p120 capture and 4Kp15 for high rest stuttery, but I'm sure that's just another feature of the chips they use, in an other wise bog standard repackaging.
Microsoft bought off all the national bodies of ISO when they ramrodded through their undocumented and impossible to implement "document standard". If ISO knew anything about business processes or standards development they could have prevented that panel-stuffing result. And yet one of the standards they set is business processes for just this situation. I don't trust them any more and I don't think you should either. They are too easily swayed by corporate interests.
Welcome to every standards body in the world where it's all about politics and not standards-making.
Take say, the 3GPP, responsible for a lot of mobile telephony standards. It's an old boy's club. Why do you think Apple's minor contribution created so much chaos? Basically the telecoms wanted it, RIM, Motorola, Nokia didn't. Basically just a political standoff for the most part - those who are making money backed Apple, those who were suffering because of it opposed it.
Or say, the IEEE. It's a process of getting your patent in so everyone implementing Wi-Fi or Ethernet will have to pay you.
Ditto every other standards body out there, whether it's MPEG, JPEG, DVD Forum, Blu-Ray Association, etc. Heck, I'm fairly certain the W3C isn't immune either (Google, Microsoft, Mozilla, Apple, etc., hell, remember trying to get video working and no one could agree on a codec?).
A standard is just a way for patent owners to try to finangle their way in, because getting FRAND money is well worth it. It's all about political connections - you scratch my back, I'll scratch yours, backroom deals, etc. And we all get stuck with the results.
Remember who populates these standards bodies - given the cost of entry is often in the 6 to 7 digits (the lower one is just to use the standard, if you want to participate in making them, it's $$$$ and always has been).
Seems to me that I should be able to let anybody contact me and I can opt in to people being charged a dollar to contact me. I don't want to make long lost friends pay to send me a message but I can see how some people might appreciate this.
Easy, you do it the old way. You send a friend request first and if they accept, you two can message all you want for free. Your long-lost buddy can go and send you a friend request first.
This service is only for two random people who aren't on each other's friend lists to message each other.
For someone like me, it's probably a way to break through the friend barrier - I don't accept friend requests willy-nilly, everyone on it has to be someone I know personally face to face. But there are people there who I haven't decided if I wanted them on as well and have stayed on my friend request list in perpetuity (usually people I don't know as well).
Opting out is more like what we have with email today - spammers just shoving crap down your inbox, except in this case, it's your facebook inbox. All facebook is proposing is that if you want to do that, you pay a "stamp" of $1 so you don't try to spam the nearly billion members with crap.
The main thing he says he will be working on is artificial intelligence that can understand "context". The goal is for Google search to be able to find pages etc based on what you mean rather then on word counts of what you type.
That's the user-facing aspect of it.
The other thing is, Google is acquiring massive amounts of information on everyone, so they need an AI to help sort through it and figure out what ads you're supposed to see.
After all, if Google can determine what you want from your searches, Google can also determine what ads to show you thare are most relevant.
But to do that, Google needs more information on who you are, so if you're say a climate change skeptic, searching for papers means Google will find papers relevant to being a skeptic, while the seame search terms for a believer will return papers relevant to a believer.
Basically the dataset Google has on people is too big to mechanically analyze, so an AI is needed to help keep relevance and filter out data that's noise (like say, when people try to deliberately screw up tracking). As a side effect, they can also place that on searches so that extra data can appear to be more useful as Google tracks you through the web.
One wonders what would happen if an ad-hoc, "name and shame" reputation network were to identify TSA agents everywhere they went. It's easy to imagine the near-universal environment of hate stares, extreme rudeness and occasional violence from victims of the TSA's Orwellian tactics putting direct pressure on TSA employees themselves to drastically reform their arrogant policies.
That will just weed out the thin-skinned ones and leave the psychopathic "I love the TSA!" types. The ones who relish the power given to them. And being named just makes them "famous" and even prouder.
It's just like police officers - it's easy to say "we need to increase the number of officers by 50", but quite a lot harder to actually do so (finding the right people is very difficult, and it's a rather thankless job that doesn't pay that well for the risks). So the good people don't generally go into policing, and since you need 50 officers, you lower the bar of entry until the bottom-of-the-barrel folks get in. Which is what we have now - people who'll gladly violate your privacy and screw you over because they've got the power to do so.
The more rational among them say "screw it" and quit, leaving a position open for someone else who wants to be "powerful" to join the ranks.
If your Dad is 90, and has the money for whatever gadgets he might want (as evidenced that he goes out and buys them on his own) then what he'd probably most want is to spend some quality time with you.
I know it sounds cheesy, but I run into the same problem trying to buy gifts for my grandmothers. They have everything they need or want, but what they can't buy is time with family.
And related equipment. How about getting a nice camcorder and tape him telling stories and such, then preserving it and his memories? Your gift will be basically ensuring he's remembered.
Don't go for a cheesy GoPro, get a NICE camcorder or dSLR with full high def and everything, and lots of memory cards. Record record record. After all, at that age, the best things would be to be remembered, and in case he used to do stuff he can't do anymore, be able to relive those moments.
And anything that'll let the rest of the family spend time with him - videoconference if necessary. It may mean having to buy iPads and iPhones and Macs so the less tech savvy part of your family can FaceTime him (he probably can do anything, but if you make it too hard on the rest of the family, they won't want to chat).
If only there were some kind of way for people to find out how to drive, say through widespread knowledge of a training and licensing program by the state.
Of course, people should be expected to understand and manage not just regular cookies, but the Flash database [also used for cookies], your browser local-storage [also used for cookies], and probably other places, and know how multiple websites may store/use one or more of these locations to track what you are doing when you visit a single web page. Of course, each browser handles each of these things differently.
At which point do we decide that people "need to know this to use a computer" and "you don't need to know this"?
After all, you need a license to drive, so they teach you the laws regarding the rules of the road and the courses tell you how to maneuver the car along the roads. But its not required to know, say, how to change your tire, how to check the oil, how to change the oil, etc. So there's some line that has to be drawn somewhere.
Likewise, would scouring your hard drive for cookies be considered of "checking the oil" difficulty? Or a rule of the road? Hell, how many people even know about cookies?
Or should we say everything that an iPad allows you to do should be required knowledge?
Yes, users should know the basics, but what ARE the basics? Do we expect them to be techies?
I mean, given the popularity of smartphones, tablets and other locked-down walled gardens, perhaps "basics" has gotten to the point where everyone who drives needs to know how give their car a tuneup, change the air filter, change the oil, etc. (And yes, you can go through car ownership without doing any of those things by bringing your car in for service and having the mechanic do it).
I was just in Beijing for two weeks. I have access to two OpenVPN servers, one in New York another in California. These are personal servers so they aren't on the IP based blacklist. However, my connection from Beijing to either of the two would crap out after a day or two, and the only remedy was to change the OpenVPN server port.
It seems right now they update their blacklist every 24~48 hours. I did not test whether the amount of traffic (idle vs. busy) would affect the time it takes them to block you. Blacklists last longer than two weeks, as the original ports I used was still blocked by the time I left. SSH connections does not seem to be affected at this time.
Traffic analysis and classification require... traffic!
After all, one can open a TCP connection out or into China, and if you leave it idle (other than keep alive) no traffic analysis system can determine its purpose - VPN or otherwise because there's no traffic to classify.
You create a VPN tunnel and used it enough that the traffic classifier could determine it was a VPN connection. An SSH session is more idle - probably just to issue commands and maybe transfer the occasional file, so the traffic classifier couldn't really figure out what the traffic was because the samples available was much too small.
I don't blame anyone for worrying about liabilities, but Pandora's Box is open, there's no closing it now. The specs for many, many firearms parts are readily available, and anyone who wants to take the time to translate those designs to 3D, is going to be able to print them, and distribute the designs. I'm waiting for someone to notice they can print 3D magazines, of any capacity they want. Yes, this is another opportunity to learn that all we do for good, can and will be perverted to bad. Are we willing to throw out the whole 3D printing movement as a result?
Not to mention existing firearms regulations already cover the issue. It's a federal offense to produce firearms without a license (and there are varying classes of licenses) to begin with. This was always true before 3D printing, and I'm fairly certain there are ways to build a gun purely from stuff you can buy at Home Depot. Heck, it may even be a criminal offense to do this.
(And really, anyone who can buy a gun can always reverse-engineer it - guns aren't terribly complicated devices compared to say, a mechanical watch. There's a bit of "art" to gunsmithing in order to make it "feel right", but anyone with a suitably equipped machine shop ought to be able to replicate them fairly easily).
And heck, guns are patented, so the specs are out in the open (it's about the only way to protect them from knockoff manufacturers.)
And really, a Makerbot-produced gun part isn't that great to begin with - they already tried the 3D printed lower receivers and they failed spectacularly after 5 shots.
About the only real form of gun control is responsibility - as individuals and society. Individual gun owners need to be responsible for the proper maintenance, care and storage of their weapons, which also includes securing it from those less responsible (children, burglars, etc). Society needs to take responsibility in that guns are everywhere, so encourage responsible use of them - perhaps less encouragement of gun violence (or violence as a whole - real or virtual), less isolationism (get to know your neighbours) and more community-mindedness (ever notice how every disaster scenario has everyone fending for themselves rather than getting organized as a community to defend the group? Instead of defending your small stash against your neighbours, why not arrange things such that you and your neighbours are prepared together?).
All I can see is that future 3D printers will now have a stick on the unit that basically says "Use of this printer to reproduce firearms parts may require a license or be subject to fines and/or jail".
The fact that Boeing's customers treat their passengers as sacks of potatoes is purely coincidental.
Execpt that passengers are typically self-loading cargo, thus saving the airline from having to pay people to load cargo as they would with sacks of potatoes.
That doesn't make any sense whatsoever. No production isn't generally linear: it is generally *sub*linear.
Making 1000 of something, generally costs LESS than 4 times the price of 250 of the thing, not *more* than 4 times.
It's true that making 10000 in an efficient will likely take longer than making 250 - because the more efficient production requires a longer and more complicated setup-phase, but there's no problem putting this in the kickstarter-description: "The first batch will be 250 items, if we get more backers, then we'll make additional batches as required..... "
Yes, it's sub-linear. However, if the original plans were for 250, and you get 1000, you need to retool stuff a LOT. Once you get into mass production (250 can be considered a small run that's easily handmade), the jump in costs is huge and beyond most people's abilities
Look at the Occulus guys - they originally planned on having the prototype developer hardware to be assembled with hot glue and duct tape. Well, that's fine for 250 units hand-assembled. Not so much with all the support they got, so now they have to invest a lot more money into moulds (easily $30K each) so it can be assembled in China (the original plan I believe was hand assembly by the team).
The problem is, ramping up is way beyond most people's abilities - it's trivial to make one-offs of something, even small runs, but once you get to larger runs a lot of work needs to be done.
250 can be done by a small team of people in the kickstarter over two or three weeks. To do 1000 this way is untenable - do it in 4 batches and you'll find the team burnt out by the 3rd batch.
So now you have to investigate contract manufacturers, have to do trial runs, costs of shipping, customs, etc. What you planned to complete in a few months suddenly takes a few months before production can get started, period. More if there are production problems (like a small tricky assembly gets done wrong, so it doesn't work and you have yields below 90%) because you didn't design for manufacture (again, something most people don't need to do for small runs).
Or, perhaps you planned on using a 3D printer to make parts, great for small runs, even contracted out. But this too can be problematic - Apple has seen this when they made an iPhone 4 part in a machine designed for prototyping, which resulted in Foxconn having to buy 20 of the machines to build enough parts. Probably more to accomodate breakdowns, since these machines aren't designed to run 24/7 continually.
And most people on KS do NOT have the skills, contacts, or ability to handle it when it blows up. It makes sense - some Joe comes up with something cool and gets a KS done to raise money to build it. But once it breaks beyond a "backyard hacking" style of manufacture to mass production, most folks are way beyond their league.
Kickstarter has a problem in that the funding sources are unlimited.
Lets say you did good and had everything planned for a production run of say, 250 units (probably a good crowdfunded estimate). Now you suddenly get people who order 1000 units total.
Problem is, production doesn't scale linearly - building 1000 takes a LOT more effort than building 250. First, your production method may work for 250, but fail for 1000 - if you planned on using a small contract manufacturer, they may not have the ability to deliver you 1000 units anymore - existing commitments may be the first 250 will be built as planned, but the excess depends on spare capacity. Depending on when you put it in, there may be none. So for example, if you have a Kickstarter that started in May and ends in June, get the money in July and start getting parts to build out August-September (assuming you can get the quantity required - again, your design may call for a special part available in small quantities, so 250 units, fine, 1000 units, well, Digikey has 700 in stock, the 300 have a 4 week lead time). Oh wait, your contract manufacturer can build in September, but October is fully booked for holiday production, so you can build 500 easily, but now you're stuck because your production line is at capacity.
So either you have to plan for it ahead of time by making sure you can line up the contract manufacturers if necessary (expensive), get the parts ahead of time (expensive), or let the schedule slip because well, building 1000 units is a lot harder than 250.
Perhaps it got popular and instead of building 250, suddenly you're saked to build 10,000. Now you've got a problem because at 250 units, Design for Manufacture isn't an issue. At 10,000, it really is because complex assemblies take time and cost more, and you may seriously have to consider making it in China, too.
The real reason is if you want to target 60, you have to aim higher because if you just take a bit too long, your framerate drops dramatically.
Target 30, and you can probably render everything in time and have time to spare. But target 60 and miss, and you'll stutter, visibly.
That's the real issue - it's also why PC gamers go for the fastest video card even though their monitors may only refresh at 60Hz or so - you need to be able to do 60+ fps constantly in order to hit 60 fps solidly. Dip below that and you stutter.
For this generation, it did make sense since a lot of people had SDTVs and such where 30p could be converted to 60i trivially (and there were still a lot of devices that only handled 1080i60, or 1080p30, which consumed the same bandwidth as 720p60).
Next gen, devs may target 30 purely because they're not sure how the hardware will handle things - throw a bit too much geometry in by accident and end up in a case of stutters, or try to achieve a more solid 30fps without stuttering because 30 is achievable, but 60 was not.
Please All you nay sayers how you cant sell open source or make money off it... PLEASE, tell me how she is a failure...
Well, she sells open source hardware. Not software, hardware. People are more willing to pay for atoms (and it's why the open-source hardware license does not allow for NC/ND - because they realize people will want to buy the things rather than go through the effort of building from scratch).
Plus being a one-stop shop for practically everything you need helps a lot (like the Amazon model - you just go to one place and get it all - it may be more expensive, but you'll probably save by not having to deal with a million and one stores).
And finally - she does interesting and useful content. Her blog entries are extremely useful and detail how she turns stuff into useful stuff that people want to hear about. More "here's a way to make our charger iPod/iPhone compatible" and less "Apple suxx0rs, buy Samsung, no walled garden!".
In other words, she sells stuff people are interested in playing with, sells practically everything related to it (if it isn't on the site, it probably doesn't exist), and provides content useful to people.
And there are plenty of open-source companies. Though, oddly, the reaction to some of them is negative - like CodeWeavers, who support the WINE project. Seems every time their CrossOver product gets mentioned (which is their bits added to WINE, plus support), people complain "Why should I pay when I can get it for FREEEEEEEE apt-get-whatever".
Slashdot Mirror
Actually, 802.11g only does NOTHING to prevent this. It's not the 802.11b client on the network, it's just an 802.11b client on the same channel.
802.11 in general is a protocol used to communicate data packets with some encoding over wireless. It dictates certain things like RTS/CTS clearance frames and CSMA/CA. It has a concept of "reservation timer" which means "even if you cannot understand the data, consider the media busy for N time".
802.11 actually cooperates, and clients on different networks actually do communicate with each other - just because a frame isn't destined for the network doesn't mean the client or AP doesn't react to it. (Interference does happen from the hidden node problem, though).
The thing with the 802.11g thing is easy - the 802.11 spec identifies that the preamble and header are sent at 1/2Mbps (802.11) speeds, then the body is sent at higher speeds (the reason for this is so every 802.11 client can consider the media busy). The trick with the 802.11g is that it could send the 802.11 headers at 802.11g speeds, but any 802.11b or lower client had a chance of interference so G-protection kicks in that forces headers to be sent at 802.11 speeds. This happens in the presence of ANY 802.11 client that appears, attached to that AP or not. After all, an 802.11b ad-hoc network could be using the airwaves as well, and interference just means no one's packets get through.
The reason you still get intereference is hidden node - you will have clients transmitting randomly because they can't hear the other transmitting node, and the receiver can hear both (it's alleviated with RTS/CTS - which would silence both transmitters). Just once you hit a certain density, there's enough that even that can't protect it due to more hidden node issues.
Actually, I think the Xbox360 outsold the Wii sometime this year in total units (the Xbox360 has been outselling the Wii for the past 18 months, so it's been only a matter of time before it overtook the Wii). The PS3 has consistently been third.
And the primary reason people are developing for Xbox360 first is simple - it's a more "standard" platform - a triple core CPU (2 threads per core) plus a GPU, just both sharing 512MB of RAM. The PS3, when considered as a "standard" platform is inferior - dual core (two threads per core though) with 256MB of system RAM and 256MB of GPU memory. The only way to tap the power of the Cell is to get deep down and program the 6 SPUs, which requires a core to manage the data flows for them.
Actually, the plane sockets provide very little power - usually under 55W or so. For a small laptop, not a big deal, but a larger ones will often require entering special modes that disable battery charging (I believe the Apple one tells the laptop to not charge the battery, for other laptops, you should remove the battery). Otherwise there's a chance you can trip the power outlet and it won't provide any more power until it's reset.
They're not providing full 15A to the AC sockets. Just barely 0.5A. Though the airline sockets are the same, really - 28V @ 2A.
No, the real challenge is figuring out a way for purchasers to wade through the eventual morass that it will become.
Remember, it's an "open" console - you buy it, you get the SDK with it, and you can make games. Which means you're going to get a bunch of crap apps as everyone sees dollar signs and tries to scam a few bucks from people for some sudoku or crossword or weather or other app.
Since it runs Android, there's a ready set of developers, so convincing people to develop for it is easy. Their questions (and everyone else's) would be how can their apps be found? Will anyone browsing the store have to wade through pages of crap to get to the gem? If someone unknown makes a great game, are they going to be drowned out by the hundreds of other apps also released?
Apple, Google and Microsoft has tried, and failed horribly. So bad Apple actually bought Chomp to help with the problem (and searching the Play Store can be frustrating, ironic given what Google does...).
They do let students specialize - students are free to take any course they want (as long as the prerequisites are met). There's nothing limiting a student from taking all higher level math courses, other than perhaps wanting a piece of paper at the end.
And no, you don't have to do university in 4 years. You can do it in 5 or more years, taking all the classes you want as long as your finances hold out.
The only real reason they force all students to take courses in other departments is to round them out. An engineer is useless if he doesn't know how to communicate his ideas to others, and perhaps taking some liberal arts classes gives him the tools he needs to relate and communicate. Hell, what's the point of proving some hard theorem if the only people who can understand it are people just like you - and no "lesser" mathematician can comprehend it? And really, if you're such a hot shot, two semesters of that should make for an easy A.
Hell, the programming equivalent is "bad code". Only because the original programmer has failed to communicate to the next person who maintains it through the code. (Hell, a lot of crap code doing stuff stupidly could probably be resolved by communication).
And practically every US Android user has it checked.
Why? Amazon. Or if you want to install those cheap Humble Android Bundles.
Which means yes, it's a superior system, except it's broken because it's an all-or-nothing. Why should I have to enable it to install an app from Amazon? (Which to be honest is probably why AT&T was forced to put the option back in). Why can't I just enable it for Amazon? Or why can I just install the Humble Bundle apps?
In fact, why isn't the option more like "Allow installation of unsigned apps"? Then Amazon and Humble Bundle and AppsLib and other stores can provide their own certificates for the user to install so that store is "trusted" and they can't install some random APK from facebook because their friends call it cool?
The problem is the "brain" part is getting smarter. What used to take all of whatever it was to control the scaler and all that, barely takes any at all because the video processors are coming with GHz class ARM processors (because really, making them at 200MHz isn't much cheaper) so the control of the inputs and video scalers and switchers and all that, are leaving a lot of cycles free.
Enough so that while my first TV probably ran a very limited RTOS, my upgraded TV (it is "connected" in that it has a... web browser) runs Linux. And actually put my embedded platform I was working on to shame (my TV - 128MB of RAM, 128MB of flash, 533MHz ARM11 (this was 2009). The platform I used had 128MB of RAM, 64MB of flash, and was a 233MHz ARM9 and was doing GigE networking and all that). Today's smart TVs probably are coming with single core Cortex A8s or dual core A9s because the SoCs with them are so cheap.
Heck, many of them are becoming mixed-signal and handle HDMI(+HDCP), component, composite and S-Video natively just because the SoC guys were bored, and tossed in A8s/dual A9s because they also wanted to sell them as tablet SoCs. With camera interfaces and all that.
Basically, TV manufacturers are given newer and newer SoCs for their video processors with more and more capability because it's so cheap to do so. Unlike say, a calculator... which I'm sure for the BOM cost 10 years ago and all that , probably could sport A8s and A9s with impressive displays and memory capacity.
Or because most of the other news articles generally deteriorate into flame wars.
Facebook announces something! Comments: Facebook sucks. It steals your information, blah blah blah.
Google announces something! Comments: Google's tracking you. It steals your information. This is why the cloud sucks. Damn you for monetizing. blah blah blah. android rules.
Apple announces something! Comments: blah blah walled garden blah blah walled garden blah blah android rules blah blah
Microsoft announces something! blah blah monopolist blah blah windows 8 sucks blah blah window ssucks blah blah android rules blah blah linux rocks blah blah
Small company announces breakthrough! Comments: not commercially ready, too expensive, NDAs, patents, blah blah blah
New game reviewd! blah blah EA sucks blah blah Steam suchs blah blah DRM blah blah consoles suck blah blah PC mouse and keyboard blah blah.
About the only thoughtful commentary happens in the patent lawsuit threads, like explanations of the rounded corners design patent (+grid of icons +dock). Though they start degenerating soon into "android rules, apple sucks blah blah blah samsung's awesome blah blah htc sucks blah blah motorola locked crap blah blah"
Then again, the others do generate ad views...
Or, it's called "how we're getting the US economy restarted". Given the exact same thing happened the instant Obama was elected in 2008 - gun store owners called it the "Democrat Economic Recovery Plan" because everyone was so afraid of the new laws that weren't introduced that they went out to stock up. Which gave a very nice boost in spending.
Looks like the same thing is happening again. Perhaps all they need to do is raise gun taxes and gun ownership taxes and the whole debt would be paid off rather quickly.
Which leads to oddball situations like the democrats constantly getting budgets balanced because of the economic boosts from gun owners worried they would take the guns away, to the deficit budgets when the Republicans come in because the gun laws would be loosened so there's no reason to buy a gun now, but wait...
So... why pirate? Why not ignore it and enjoy something else?
You do realize that you're just encouraging the problem, right? Because they create something you want, and instead of giving them what they want in exchange, you say it's worthless, and then expect them to take it lying down. I'm sure if someone pirated the Linux kernel (by say, not releasing the source code), you'd be upset. Yes, you can pirate Linux by ignoring the GPL and violating the de-facto copyright law. You can pirate most open-source that way because by not agreeing to the freer terms in the license, it reverts back to "all rights reserved" and what copyright grants.
You're better off ignoring what the entertainment industry produces and enjoy the works of the indie bands, moviemakers, game creators and what not. Because the industry won't change unless they see success can be had without draconian measures, and piracy is not the way to do it.
Hell, if you want to pirate, go buy the DVD and download the DRM-free version, but that requires supporting the industry. And while they may not be entitled to profits, you're not entitled to enjoy their works just because you don't agree to their terms.
There's plenty of free and open indie entertainment out there that's DRM free and cheap to enjoy. Why not encourage those creative people by enjoying their work?
Well, the funny thing is, I have a tons of Megabloks (because they have the Halo license), and I have to admit, the quality of the bricks... suck. I don't know if it's the plastic, but it certainly doesn't feel like the standard ABS that Lego uses. It feels cheaply made and constructed and while the outcome is good, it just seems a lot less satisfying.
Not that they're that cheap anymore - I've seen some kits for $200 that didn't look too complex (certainly not half as complex as the 8000-piece Millennium Falcon Lego has).
And what IP would that be? A GoPro isn't exactly an innovative thing. It's just a plain old 1080p30 camera in a rugged case with a bunch of mounts to hang it in various locations.
OK, fine, GoPro has some design patents, but we know how /. likes the "rounded corners" design patent idea. (Yes, I know, it's rounded corner slate with a grid of icons and a dock that's static as the grid's adjusted).
But other than that, it's just a lens, a sensor, and a processor wired together with some software to make it usable. Add a little expansion header for a display and you get the LCD backpack, a little SDIO header for wifi, etc.
The new Hero 3 has some neat features like 720p120 capture and 4Kp15 for high rest stuttery, but I'm sure that's just another feature of the chips they use, in an other wise bog standard repackaging.
Welcome to every standards body in the world where it's all about politics and not standards-making.
Take say, the 3GPP, responsible for a lot of mobile telephony standards. It's an old boy's club. Why do you think Apple's minor contribution created so much chaos? Basically the telecoms wanted it, RIM, Motorola, Nokia didn't. Basically just a political standoff for the most part - those who are making money backed Apple, those who were suffering because of it opposed it.
Or say, the IEEE. It's a process of getting your patent in so everyone implementing Wi-Fi or Ethernet will have to pay you.
Ditto every other standards body out there, whether it's MPEG, JPEG, DVD Forum, Blu-Ray Association, etc. Heck, I'm fairly certain the W3C isn't immune either (Google, Microsoft, Mozilla, Apple, etc., hell, remember trying to get video working and no one could agree on a codec?).
A standard is just a way for patent owners to try to finangle their way in, because getting FRAND money is well worth it. It's all about political connections - you scratch my back, I'll scratch yours, backroom deals, etc. And we all get stuck with the results.
Remember who populates these standards bodies - given the cost of entry is often in the 6 to 7 digits (the lower one is just to use the standard, if you want to participate in making them, it's $$$$ and always has been).
Easy, you do it the old way. You send a friend request first and if they accept, you two can message all you want for free. Your long-lost buddy can go and send you a friend request first.
This service is only for two random people who aren't on each other's friend lists to message each other.
For someone like me, it's probably a way to break through the friend barrier - I don't accept friend requests willy-nilly, everyone on it has to be someone I know personally face to face. But there are people there who I haven't decided if I wanted them on as well and have stayed on my friend request list in perpetuity (usually people I don't know as well).
Opting out is more like what we have with email today - spammers just shoving crap down your inbox, except in this case, it's your facebook inbox. All facebook is proposing is that if you want to do that, you pay a "stamp" of $1 so you don't try to spam the nearly billion members with crap.
That's the user-facing aspect of it.
The other thing is, Google is acquiring massive amounts of information on everyone, so they need an AI to help sort through it and figure out what ads you're supposed to see.
After all, if Google can determine what you want from your searches, Google can also determine what ads to show you thare are most relevant.
But to do that, Google needs more information on who you are, so if you're say a climate change skeptic, searching for papers means Google will find papers relevant to being a skeptic, while the seame search terms for a believer will return papers relevant to a believer.
Basically the dataset Google has on people is too big to mechanically analyze, so an AI is needed to help keep relevance and filter out data that's noise (like say, when people try to deliberately screw up tracking). As a side effect, they can also place that on searches so that extra data can appear to be more useful as Google tracks you through the web.
That will just weed out the thin-skinned ones and leave the psychopathic "I love the TSA!" types. The ones who relish the power given to them. And being named just makes them "famous" and even prouder.
It's just like police officers - it's easy to say "we need to increase the number of officers by 50", but quite a lot harder to actually do so (finding the right people is very difficult, and it's a rather thankless job that doesn't pay that well for the risks). So the good people don't generally go into policing, and since you need 50 officers, you lower the bar of entry until the bottom-of-the-barrel folks get in. Which is what we have now - people who'll gladly violate your privacy and screw you over because they've got the power to do so.
The more rational among them say "screw it" and quit, leaving a position open for someone else who wants to be "powerful" to join the ranks.
And related equipment. How about getting a nice camcorder and tape him telling stories and such, then preserving it and his memories? Your gift will be basically ensuring he's remembered.
Don't go for a cheesy GoPro, get a NICE camcorder or dSLR with full high def and everything, and lots of memory cards. Record record record. After all, at that age, the best things would be to be remembered, and in case he used to do stuff he can't do anymore, be able to relive those moments.
And anything that'll let the rest of the family spend time with him - videoconference if necessary. It may mean having to buy iPads and iPhones and Macs so the less tech savvy part of your family can FaceTime him (he probably can do anything, but if you make it too hard on the rest of the family, they won't want to chat).
At which point do we decide that people "need to know this to use a computer" and "you don't need to know this"?
After all, you need a license to drive, so they teach you the laws regarding the rules of the road and the courses tell you how to maneuver the car along the roads. But its not required to know, say, how to change your tire, how to check the oil, how to change the oil, etc. So there's some line that has to be drawn somewhere.
Likewise, would scouring your hard drive for cookies be considered of "checking the oil" difficulty? Or a rule of the road? Hell, how many people even know about cookies?
Or should we say everything that an iPad allows you to do should be required knowledge?
Yes, users should know the basics, but what ARE the basics? Do we expect them to be techies?
I mean, given the popularity of smartphones, tablets and other locked-down walled gardens, perhaps "basics" has gotten to the point where everyone who drives needs to know how give their car a tuneup, change the air filter, change the oil, etc. (And yes, you can go through car ownership without doing any of those things by bringing your car in for service and having the mechanic do it).
Traffic analysis and classification require... traffic!
After all, one can open a TCP connection out or into China, and if you leave it idle (other than keep alive) no traffic analysis system can determine its purpose - VPN or otherwise because there's no traffic to classify.
You create a VPN tunnel and used it enough that the traffic classifier could determine it was a VPN connection. An SSH session is more idle - probably just to issue commands and maybe transfer the occasional file, so the traffic classifier couldn't really figure out what the traffic was because the samples available was much too small.
Not to mention existing firearms regulations already cover the issue. It's a federal offense to produce firearms without a license (and there are varying classes of licenses) to begin with. This was always true before 3D printing, and I'm fairly certain there are ways to build a gun purely from stuff you can buy at Home Depot. Heck, it may even be a criminal offense to do this.
(And really, anyone who can buy a gun can always reverse-engineer it - guns aren't terribly complicated devices compared to say, a mechanical watch. There's a bit of "art" to gunsmithing in order to make it "feel right", but anyone with a suitably equipped machine shop ought to be able to replicate them fairly easily).
And heck, guns are patented, so the specs are out in the open (it's about the only way to protect them from knockoff manufacturers.)
And really, a Makerbot-produced gun part isn't that great to begin with - they already tried the 3D printed lower receivers and they failed spectacularly after 5 shots.
About the only real form of gun control is responsibility - as individuals and society. Individual gun owners need to be responsible for the proper maintenance, care and storage of their weapons, which also includes securing it from those less responsible (children, burglars, etc). Society needs to take responsibility in that guns are everywhere, so encourage responsible use of them - perhaps less encouragement of gun violence (or violence as a whole - real or virtual), less isolationism (get to know your neighbours) and more community-mindedness (ever notice how every disaster scenario has everyone fending for themselves rather than getting organized as a community to defend the group? Instead of defending your small stash against your neighbours, why not arrange things such that you and your neighbours are prepared together?).
All I can see is that future 3D printers will now have a stick on the unit that basically says "Use of this printer to reproduce firearms parts may require a license or be subject to fines and/or jail".
Execpt that passengers are typically self-loading cargo, thus saving the airline from having to pay people to load cargo as they would with sacks of potatoes.
Yes, it's sub-linear. However, if the original plans were for 250, and you get 1000, you need to retool stuff a LOT. Once you get into mass production (250 can be considered a small run that's easily handmade), the jump in costs is huge and beyond most people's abilities
Look at the Occulus guys - they originally planned on having the prototype developer hardware to be assembled with hot glue and duct tape. Well, that's fine for 250 units hand-assembled. Not so much with all the support they got, so now they have to invest a lot more money into moulds (easily $30K each) so it can be assembled in China (the original plan I believe was hand assembly by the team).
The problem is, ramping up is way beyond most people's abilities - it's trivial to make one-offs of something, even small runs, but once you get to larger runs a lot of work needs to be done.
250 can be done by a small team of people in the kickstarter over two or three weeks. To do 1000 this way is untenable - do it in 4 batches and you'll find the team burnt out by the 3rd batch.
So now you have to investigate contract manufacturers, have to do trial runs, costs of shipping, customs, etc. What you planned to complete in a few months suddenly takes a few months before production can get started, period. More if there are production problems (like a small tricky assembly gets done wrong, so it doesn't work and you have yields below 90%) because you didn't design for manufacture (again, something most people don't need to do for small runs).
Or, perhaps you planned on using a 3D printer to make parts, great for small runs, even contracted out. But this too can be problematic - Apple has seen this when they made an iPhone 4 part in a machine designed for prototyping, which resulted in Foxconn having to buy 20 of the machines to build enough parts. Probably more to accomodate breakdowns, since these machines aren't designed to run 24/7 continually.
And most people on KS do NOT have the skills, contacts, or ability to handle it when it blows up. It makes sense - some Joe comes up with something cool and gets a KS done to raise money to build it. But once it breaks beyond a "backyard hacking" style of manufacture to mass production, most folks are way beyond their league.
Kickstarter has a problem in that the funding sources are unlimited.
Lets say you did good and had everything planned for a production run of say, 250 units (probably a good crowdfunded estimate). Now you suddenly get people who order 1000 units total.
Problem is, production doesn't scale linearly - building 1000 takes a LOT more effort than building 250. First, your production method may work for 250, but fail for 1000 - if you planned on using a small contract manufacturer, they may not have the ability to deliver you 1000 units anymore - existing commitments may be the first 250 will be built as planned, but the excess depends on spare capacity. Depending on when you put it in, there may be none. So for example, if you have a Kickstarter that started in May and ends in June, get the money in July and start getting parts to build out August-September (assuming you can get the quantity required - again, your design may call for a special part available in small quantities, so 250 units, fine, 1000 units, well, Digikey has 700 in stock, the 300 have a 4 week lead time). Oh wait, your contract manufacturer can build in September, but October is fully booked for holiday production, so you can build 500 easily, but now you're stuck because your production line is at capacity.
So either you have to plan for it ahead of time by making sure you can line up the contract manufacturers if necessary (expensive), get the parts ahead of time (expensive), or let the schedule slip because well, building 1000 units is a lot harder than 250.
Perhaps it got popular and instead of building 250, suddenly you're saked to build 10,000. Now you've got a problem because at 250 units, Design for Manufacture isn't an issue. At 10,000, it really is because complex assemblies take time and cost more, and you may seriously have to consider making it in China, too.
The real reason is if you want to target 60, you have to aim higher because if you just take a bit too long, your framerate drops dramatically.
Target 30, and you can probably render everything in time and have time to spare. But target 60 and miss, and you'll stutter, visibly.
That's the real issue - it's also why PC gamers go for the fastest video card even though their monitors may only refresh at 60Hz or so - you need to be able to do 60+ fps constantly in order to hit 60 fps solidly. Dip below that and you stutter.
For this generation, it did make sense since a lot of people had SDTVs and such where 30p could be converted to 60i trivially (and there were still a lot of devices that only handled 1080i60, or 1080p30, which consumed the same bandwidth as 720p60).
Next gen, devs may target 30 purely because they're not sure how the hardware will handle things - throw a bit too much geometry in by accident and end up in a case of stutters, or try to achieve a more solid 30fps without stuttering because 30 is achievable, but 60 was not.
Well, she sells open source hardware. Not software, hardware. People are more willing to pay for atoms (and it's why the open-source hardware license does not allow for NC/ND - because they realize people will want to buy the things rather than go through the effort of building from scratch).
Plus being a one-stop shop for practically everything you need helps a lot (like the Amazon model - you just go to one place and get it all - it may be more expensive, but you'll probably save by not having to deal with a million and one stores).
And finally - she does interesting and useful content. Her blog entries are extremely useful and detail how she turns stuff into useful stuff that people want to hear about. More "here's a way to make our charger iPod/iPhone compatible" and less "Apple suxx0rs, buy Samsung, no walled garden!".
In other words, she sells stuff people are interested in playing with, sells practically everything related to it (if it isn't on the site, it probably doesn't exist), and provides content useful to people.
And there are plenty of open-source companies. Though, oddly, the reaction to some of them is negative - like CodeWeavers, who support the WINE project. Seems every time their CrossOver product gets mentioned (which is their bits added to WINE, plus support), people complain "Why should I pay when I can get it for FREEEEEEEE apt-get-whatever".