1) Brandon Eich wasn't executed. 2) Mozilla ensures its financial existence, while Salem Puritanism ceased to be considered a credible religious doctrine.
Yup. And it costs that Ph.D absolutely nothing to make generalizations using data which may be utterly invalid today, post 2006.
I believe the truth is somewhere between four results. Some people will go to college, and become financially successful. Some people today will try to complete their college degrees, in exchange for indentured servitude (to student debt) for most of their working lives. Some people won't go to college, strike it out on their own, and not be financially successful. A few people will skip college, strike it out on their own, and become financially successful. There is no real middle strategy; its just a matter of which choice you will make, and you will have to live with the result. My feeling is that you make the choice that best fits yourself, and have a strategy when it doesn't work out.
The only thing I'm pretty certain about: If you're poor, and can't get a subsidized ride into a very good four year college, you are definitely opting to be a wage slave if you take on a lot of student debt. Very few poor kids are going to take on debt in a profession like doctor or lawyer, and work his way out in his lifetime.
1) There are a lot of detractors of Gnome 3. Everyone is not happy with it.
2) What you don't grasp is what I admire about Canonical/Shuttleworth. What kept Linux (or Ubuntu) from dominating the desktop was an inferior GUI experience to Apple/Windows (and unequal driver support). You could chose Gnome or KDE, both too flawed and two unmotivated to compete for the general (dumb) consumer market, give up and go with some form of LXDE(?), or do what Canonical did. Canonical wanted to go after what it saw was the future, and instead of unsuccessfully negotiating with Gnome & KDE designer councils to implement what Canonical wanted, Canonical took its future into its own hands. Computing hardware was moving to tablets, phones & gear, not desktops, and they wanted a GUI that could bridge both worlds. That's why they went to Unity, and given the problems producing a compositor competitive with DirectX, they went to Wayland, then Mir.
3) Every megalomanic in Unixland operates like a bloodthirsty Bolshevik, and thinks Canonical owes them a living for failure or half a loaf.
Its pretty freaking obvious that Canonical's GUI strategy (Unity/Mir) does not involve depending on Gnome. Since neither Shuttleworth is the richest man on earth, or Canonical isn't making a huge profit, why should they be investing salaries into an ancillary GUI, which isn't even that popular within the industry?
I hope Canonical decides to include Gnome into the 14.04 LTS, but only if the company thinks they can afford to invest resources towards it.
The problem wasn't merely that the geeks had no fucking clue how to maintain security. The problem was that the management had no fucking clue how to properly operate a finance establishment. There is no way a competent financial institution does not become immediately aware that someone is stealing from their "vaults". That's because the financial institution should be validating their assets daily, if not hourly.
Here's the question: Do the major exchanges STILL in operation ALL run basic accounting and bank operation practices? Is the Bitcoin Foundation moving the industry towards enabling functioning "arbitrage" for the bitcoin system? What is the Bitcoin Foundation doing to incorporate real world consumer transactions? Until those three aspects come about, you will not have the necessary stability needed for people and businesses to treat bitcoin as "money".
If you don't care how long your programs take to solve a problem, avoid coding in cuda. If you want to keep your job, and your employer needs to run the app on nvidia cards as fast as possible, you're writing it in cuda.
No, AMD hardware outperforms Nvidia hardware, computational units/$1, for bitcoin mining. That's why bitcoin miners use OpenCL; its about the card. (When one cares about protein folding, programmers that need speed, are going to CUDA and nvidia cards.) Pot, meet Kettle.
I suggest that it is about time programmers started getting used to coding in assembly once again.
You can give up right there. The days of humans getting 100x more efficiency out of a CPU using assembler rather than a higher level language are over. Optimizing compilers are able to devise efficiencies at large scale/detail that a human can at this point. Enterprise level software requiring millions of lines of code are just too large to be optimizable by one human writing in assembler. Speed efficiencies with out of order execution, deep pipelines consumer CPUs will be better utilized by compilers able to make better predictive arrangement of code.
Don't get me wrong. You'll always be able to find ONE "John Henry" that will be able to outcode the "stream compiler". But you can't build a world economy on one programmer. And forget about finding COMPETENT assembler programmers. The people you need to extract these kind of efficiencies are like finding prima ballerinas. Sadly, the world's economy needs more mediocre programmers to generate more working code, and more higher-level, software engineers to implement new solutions for problems addressable by a computer.
Businesses are structured for the quarterly stock profit. For the suggestion your making, the management team would have to be in position to maintain the strategy for years, while taking a killing in profitability. Basically, AMD would have to be taken over by a Japanese company for even a *chance* of a comeback. AMD will have to hope that its pivot to gaming consoles gives it enough oxygen to sustain a radical, new CPU architecture. AMD is a dead man walking.
Marissa Meyer's tumblr purchase strategy isn't nuts, just the price ($1.1 billion?!?!)
Meyer wants to improve Yahoo's current products, and move Yahoo to a focused social media/portal platform. She's counting on Yahoo grabbing a piece of the mobile social media pie, which no big player has right now. (Google would be closest.) This is what will fuel Yahoo's "comeback" into relevance. The problem is that Yahoo has zero product presence in mobile. She's buying tumblr as an infrastructure purchase.
The next issue is pretty cool. Turns out, Facebook will probably not grow anymore. The tweens don't like Facebook. Kids don't like to treat their social media as maintenance work. And Facebook's zeal to grab eyeballs means parents have moved to Facebook, which makes it uncool. Kids want to maintain communication with their peers, so they're gravitating towards low maintenance social media, like twitter, tumblr, and instagram (which Facebook is borging to death). So Meyer buys tumblr to get more presence in social media, and deny another avenue Facebook can acquire to fix their mobile/tween growth. The problem is there's no way tumblr is worth $1.1 billion, unless Meyer is grabbing brains as well, and sees some sort of general social media on mobile framework she can build Yahoo on top of tumblr.
Apple, back in 1998-1999, was on the brink of bankruptcy. Even the early years of Jobs return, Apple was putting out colorful plastic, underpowered computers. It wasn't until the introduction of the Ipod, and Apple's redirection into the consumer device market, did Apple dig itself out of its 1990's stupor.
Did reality prove you wrong? Hasn't the Red Hat stock grown in multiples of its 1990's value? Did she sell it in the early 2000's?
A service pack is a form of configuration management. Think of every binary in the Windows operating system as a program with a version. Microsoft wants to encourage developers to support the latest version of their patched OS. That is, of course, feasibly impossible, especially when some developers are confronted with major behavioral change in one OS program update that their application is dependent upon. So having a "blessed" minimal collection of binary versions makes Microsoft only responsible for those versions. It then becomes incumbent for the developers to make sure their application works to SP1 versions of all those OS programs, and the developers cease to be responsible for making their app work with the original OS binary/daemon that was released with the Windows 7 rollout. (And yes, this is a descriptive simplification of the issue.)
There is more going on with a service pack than just throwing together the latest version of each OS binary. Yes, I wish Microsoft would put out an SP2 already, even if they want to commit corporate suicide by abandoning Windows 7 to get customers to move to Windows 8.
Have you taken a good look at how FIOS arranges its channels blocks? Fiber has so much bandwidth, they're broadcasting the same channel in SD, HD, and 1080i, simultaneously. (FIOS may charge extra to access the high end.) I seriously doubt the production company is compressing their signal; there's no advantage for them to do so. I suspect you're using a low-bandwidth channel to tape via TiVo. (In fact, I think that AMC only recently got a 1080i channel, here in the Northeast.)
Your question is vague and dependent on many circumstantial details. Generally, I don't see it possible to gain contract work without experience.
But the OP says he has years of VB/.NET coding experience. Then it boils down to what region of the country, what industries the programming work came from, what is his competence in what he is being asked to do, and how good are his professional networking skills.
You generally do not land "entry level" contract work from job websites from headhunters and HR departments. They will often ask for 5 years of experience in X, when the language didn't exist 5 years ago. (That's because they don't know how to properly do their job.)
But when trying to shove a foot in the door, you will get wind of a possible short-term work from old bosses, former coworkers, and headhunters (not through classifieds). The people who hire will be the (senior) project managers who only give a rats ass about getting a task completed. If you can sell *them*, and placate their HR, that's how you get contract work. But its not enough to make a living on without a previous programming jobs, and you can't be demanding the "going rate" until you develop a reputation as someone who delivers.
The OP said he was disinclined to go into project management. And MBAs now are a dime a dozen. Its certainly not worth spending a crap load of money to get a piece of paper that no one will hire.
As much as I hate to be dogmatically negative, I'd say he's too old to be retrained for coding jobs. 1) Look very carefully at what he said: he specialized in VB.NET. He never bothered to learn C#, which is a centerpiece language for.NET. If he had bothered to learn something outside his vocational concentration, he'd never be posting his question. 2) The only expense to him at this point should be a few books and his time. One shouldn't need to go back to college in to learn how to use C#; particularly if you were adept with computing concepts, algorithms, and development methodologies. If he thinks he needs to obtain paid certification to be "hireable" as a programmer, then he's not really cut out to be a programmer. If he feels the need to ask opinions over whether to make an effort to stay in the industry, then he's as good as done. A professional in his position would be learning C#, latest.NET techniques/additions, and scrounging around for C# contract work to put on his resume. The career life of a 40+ year old coder is not bright at this point, but he shouldn't have a problem finding low paying work, even when competing with 22 year olds and overseas coding shops. 3) Frankly, if he was competent at math, he wouldn't even need to retrain on C#. Retrain on F#, and get a job coding for quants, labs, and large organizations. 4) The only thing I can see him doing that is still computer related, and not programming, would be to recertify as a network administrator. That environment is unlikely to change radically, and there will always be networks needing to upgrade. But you'd have to ask network techs if that is worth bothering at this point. 5) He'd be better off (if living in an non-union state) to retrain as a plumber or electrician. I would imagine after a decade, he'd make more money doing that, than as a programmer.
Everything in Windoze 7 OS is built to the.NET API, and probably most of Vista. OS components were never meant to run exclusively off the CLR VM of.NET. All the huge legacy applications MS makes money on were never going to refactored to operate off of CLR. You don't have to run off of CLR to be.NET compatible, or run exclusively with.NET APIs..NET isn't great for OS code;.NET is great for ENTERPRISE APPLICATION development. A large corporation doesn't have to be limited to one language while running off the same APIs, and the previous codebase is still binary compatible.
Its not a matter that its "inhumane" to use ar-15 bullets on a deer. In many states, its *illegal* to use an ar-15 to hunt deer. The rationale for the law is that.223 bullets shot out of an ar-15 will not "drop" a deer dead. The bullet will fly through the deer, and the deer will run away and eventually bleed to death.
A FMJ.223 bullet shot out of an ar-15 is *not* "low power". It has enough energy to fly through a (sheet rock) wall, and through a car door. (Police and soldiers are taught to "take cover" behind the engine block of a car, not merely behind a car door.)
A military issue firearm is *not* (significantly) *deadlier* than a civilian version. The military issue "assault" rifles are capable of firing more than one bullet per trigger pull. Theoretically, it makes it more deadly than a civilian issue rifle, but NO (new) automatic rifles are allowed to be sold to civilians. The Aurora nutjob did not shoot up the theatre with a "military issue" assault rifle.
The only thing that currently makes a civilian copy of an "assault rifle" deadlier than a hunting rifle is that they can carry a 30-100 bullet magazine. I personally would not object to a law that banned selling firearm magazines with more than 10 bullets. Only police and gang bangers "need" to avoid reloading after shooting more than 10 bullets. I have no doubt that the body count would have been lower if Holmes didn't have a 100 bullet drum, and had to reload after every 10 shots.
1) The PPC code never gets inserted into other machine's architectures. So in that sense, it can't possibly "bloat" the kernel. Now there could be design issues with PPC that end up being carried into future Linux kernels, but those are much harder to root out without breaking something. 2) How else can Linux keep its reputation for being able to operate obsolete stuff, long after the commercial vendor has abandoned it? 3) Anything that's in the kernel (like PPC support), has an "active" maintainer for it. As long as there's an active maintainer, there is no reason to remove anything which was built into the kernel. When there's no more active maintainer, then the feature is deprecated. Eventually, a Cardinal in the LKML group gets motivated to eradicate obsolete stuff. Then it gets cut out of the kernel.
Slashdot Mirror
1) Brandon Eich wasn't executed.
2) Mozilla ensures its financial existence, while Salem Puritanism ceased to be considered a credible religious doctrine.
No, this is a capitalist culture. You exploit the gullible.
The overhead in $15K is not overhead; a huge chunk of it is taxes & the "payroll end" of taxes. If anything, the overhead is a little low.
Yup. And it costs that Ph.D absolutely nothing to make generalizations using data which may be utterly invalid today, post 2006.
I believe the truth is somewhere between four results. Some people will go to college, and become financially successful. Some people today will try to complete their college degrees, in exchange for indentured servitude (to student debt) for most of their working lives. Some people won't go to college, strike it out on their own, and not be financially successful. A few people will skip college, strike it out on their own, and become financially successful. There is no real middle strategy; its just a matter of which choice you will make, and you will have to live with the result. My feeling is that you make the choice that best fits yourself, and have a strategy when it doesn't work out.
The only thing I'm pretty certain about: If you're poor, and can't get a subsidized ride into a very good four year college, you are definitely opting to be a wage slave if you take on a lot of student debt. Very few poor kids are going to take on debt in a profession like doctor or lawyer, and work his way out in his lifetime.
1) There are a lot of detractors of Gnome 3. Everyone is not happy with it.
2) What you don't grasp is what I admire about Canonical/Shuttleworth. What kept Linux (or Ubuntu) from dominating the desktop was an inferior GUI experience to Apple/Windows (and unequal driver support). You could chose Gnome or KDE, both too flawed and two unmotivated to compete for the general (dumb) consumer market, give up and go with some form of LXDE(?), or do what Canonical did. Canonical wanted to go after what it saw was the future, and instead of unsuccessfully negotiating with Gnome & KDE designer councils to implement what Canonical wanted, Canonical took its future into its own hands. Computing hardware was moving to tablets, phones & gear, not desktops, and they wanted a GUI that could bridge both worlds. That's why they went to Unity, and given the problems producing a compositor competitive with DirectX, they went to Wayland, then Mir.
3) Every megalomanic in Unixland operates like a bloodthirsty Bolshevik, and thinks Canonical owes them a living for failure or half a loaf.
...I want a pony, Canonical. Pay up!
Its pretty freaking obvious that Canonical's GUI strategy (Unity/Mir) does not involve depending on Gnome. Since neither Shuttleworth is the richest man on earth, or Canonical isn't making a huge profit, why should they be investing salaries into an ancillary GUI, which isn't even that popular within the industry?
I hope Canonical decides to include Gnome into the 14.04 LTS, but only if the company thinks they can afford to invest resources towards it.
The problem wasn't merely that the geeks had no fucking clue how to maintain security. The problem was that the management had no fucking clue how to properly operate a finance establishment. There is no way a competent financial institution does not become immediately aware that someone is stealing from their "vaults". That's because the financial institution should be validating their assets daily, if not hourly.
Here's the question: Do the major exchanges STILL in operation ALL run basic accounting and bank operation practices? Is the Bitcoin Foundation moving the industry towards enabling functioning "arbitrage" for the bitcoin system? What is the Bitcoin Foundation doing to incorporate real world consumer transactions? Until those three aspects come about, you will not have the necessary stability needed for people and businesses to treat bitcoin as "money".
Lion can run on 32bit machines. Its Mountain Lion, OSX 10.8, that can't.
Are you telling me there are still people doing development on L4 kernels???
And don't bring up HURD, I know they've already given up on it.
Who "uses" Wayland?
Ssshhhhh.... Don't get the crazy man started...
If you don't care how long your programs take to solve a problem, avoid coding in cuda. If you want to keep your job, and your employer needs to run the app on nvidia cards as fast as possible, you're writing it in cuda.
No, AMD hardware outperforms Nvidia hardware, computational units/$1, for bitcoin mining. That's why bitcoin miners use OpenCL; its about the card. (When one cares about protein folding, programmers that need speed, are going to CUDA and nvidia cards.) Pot, meet Kettle.
I suggest that it is about time programmers started getting used to coding in assembly once again.
You can give up right there. The days of humans getting 100x more efficiency out of a CPU using assembler rather than a higher level language are over. Optimizing compilers are able to devise efficiencies at large scale/detail that a human can at this point. Enterprise level software requiring millions of lines of code are just too large to be optimizable by one human writing in assembler. Speed efficiencies with out of order execution, deep pipelines consumer CPUs will be better utilized by compilers able to make better predictive arrangement of code.
Don't get me wrong. You'll always be able to find ONE "John Henry" that will be able to outcode the "stream compiler". But you can't build a world economy on one programmer. And forget about finding COMPETENT assembler programmers. The people you need to extract these kind of efficiencies are like finding prima ballerinas. Sadly, the world's economy needs more mediocre programmers to generate more working code, and more higher-level, software engineers to implement new solutions for problems addressable by a computer.
Businesses are structured for the quarterly stock profit. For the suggestion your making, the management team would have to be in position to maintain the strategy for years, while taking a killing in profitability. Basically, AMD would have to be taken over by a Japanese company for even a *chance* of a comeback. AMD will have to hope that its pivot to gaming consoles gives it enough oxygen to sustain a radical, new CPU architecture. AMD is a dead man walking.
Marissa Meyer's tumblr purchase strategy isn't nuts, just the price ($1.1 billion?!?!)
Meyer wants to improve Yahoo's current products, and move Yahoo to a focused social media/portal platform. She's counting on Yahoo grabbing a piece of the mobile social media pie, which no big player has right now. (Google would be closest.) This is what will fuel Yahoo's "comeback" into relevance. The problem is that Yahoo has zero product presence in mobile. She's buying tumblr as an infrastructure purchase.
The next issue is pretty cool. Turns out, Facebook will probably not grow anymore. The tweens don't like Facebook. Kids don't like to treat their social media as maintenance work. And Facebook's zeal to grab eyeballs means parents have moved to Facebook, which makes it uncool. Kids want to maintain communication with their peers, so they're gravitating towards low maintenance social media, like twitter, tumblr, and instagram (which Facebook is borging to death). So Meyer buys tumblr to get more presence in social media, and deny another avenue Facebook can acquire to fix their mobile/tween growth. The problem is there's no way tumblr is worth $1.1 billion, unless Meyer is grabbing brains as well, and sees some sort of general social media on mobile framework she can build Yahoo on top of tumblr.
Apple, back in 1998-1999, was on the brink of bankruptcy. Even the early years of Jobs return, Apple was putting out colorful plastic, underpowered computers. It wasn't until the introduction of the Ipod, and Apple's redirection into the consumer device market, did Apple dig itself out of its 1990's stupor.
Did reality prove you wrong? Hasn't the Red Hat stock grown in multiples of its 1990's value? Did she sell it in the early 2000's?
A service pack is a form of configuration management. Think of every binary in the Windows operating system as a program with a version. Microsoft wants to encourage developers to support the latest version of their patched OS. That is, of course, feasibly impossible, especially when some developers are confronted with major behavioral change in one OS program update that their application is dependent upon. So having a "blessed" minimal collection of binary versions makes Microsoft only responsible for those versions. It then becomes incumbent for the developers to make sure their application works to SP1 versions of all those OS programs, and the developers cease to be responsible for making their app work with the original OS binary/daemon that was released with the Windows 7 rollout. (And yes, this is a descriptive simplification of the issue.)
There is more going on with a service pack than just throwing together the latest version of each OS binary. Yes, I wish Microsoft would put out an SP2 already, even if they want to commit corporate suicide by abandoning Windows 7 to get customers to move to Windows 8.
Of course, no one is going to notice, since everyone will be using cloud computing accounts.
Have you taken a good look at how FIOS arranges its channels blocks? Fiber has so much bandwidth, they're broadcasting the same channel in SD, HD, and 1080i, simultaneously. (FIOS may charge extra to access the high end.) I seriously doubt the production company is compressing their signal; there's no advantage for them to do so. I suspect you're using a low-bandwidth channel to tape via TiVo. (In fact, I think that AMC only recently got a 1080i channel, here in the Northeast.)
Your question is vague and dependent on many circumstantial details. Generally, I don't see it possible to gain contract work without experience.
But the OP says he has years of VB/.NET coding experience. Then it boils down to what region of the country, what industries the programming work came from, what is his competence in what he is being asked to do, and how good are his professional networking skills.
You generally do not land "entry level" contract work from job websites from headhunters and HR departments. They will often ask for 5 years of experience in X, when the language didn't exist 5 years ago. (That's because they don't know how to properly do their job.)
But when trying to shove a foot in the door, you will get wind of a possible short-term work from old bosses, former coworkers, and headhunters (not through classifieds). The people who hire will be the (senior) project managers who only give a rats ass about getting a task completed. If you can sell *them*, and placate their HR, that's how you get contract work. But its not enough to make a living on without a previous programming jobs, and you can't be demanding the "going rate" until you develop a reputation as someone who delivers.
The OP said he was disinclined to go into project management. And MBAs now are a dime a dozen. Its certainly not worth spending a crap load of money to get a piece of paper that no one will hire.
As much as I hate to be dogmatically negative, I'd say he's too old to be retrained for coding jobs. .NET. If he had bothered to learn something outside his vocational concentration, he'd never be posting his question. .NET techniques/additions, and scrounging around for C# contract work to put on his resume. The career life of a 40+ year old coder is not bright at this point, but he shouldn't have a problem finding low paying work, even when competing with 22 year olds and overseas coding shops.
1) Look very carefully at what he said: he specialized in VB.NET. He never bothered to learn C#, which is a centerpiece language for
2) The only expense to him at this point should be a few books and his time. One shouldn't need to go back to college in to learn how to use C#; particularly if you were adept with computing concepts, algorithms, and development methodologies. If he thinks he needs to obtain paid certification to be "hireable" as a programmer, then he's not really cut out to be a programmer. If he feels the need to ask opinions over whether to make an effort to stay in the industry, then he's as good as done. A professional in his position would be learning C#, latest
3) Frankly, if he was competent at math, he wouldn't even need to retrain on C#. Retrain on F#, and get a job coding for quants, labs, and large organizations.
4) The only thing I can see him doing that is still computer related, and not programming, would be to recertify as a network administrator. That environment is unlikely to change radically, and there will always be networks needing to upgrade. But you'd have to ask network techs if that is worth bothering at this point.
5) He'd be better off (if living in an non-union state) to retrain as a plumber or electrician. I would imagine after a decade, he'd make more money doing that, than as a programmer.
Everything in Windoze 7 OS is built to the .NET API, and probably most of Vista. OS components were never meant to run exclusively off the CLR VM of .NET. All the huge legacy applications MS makes money on were never going to refactored to operate off of CLR. You don't have to run off of CLR to be .NET compatible, or run exclusively with .NET APIs. .NET isn't great for OS code; .NET is great for ENTERPRISE APPLICATION development. A large corporation doesn't have to be limited to one language while running off the same APIs, and the previous codebase is still binary compatible.
Common sense; mosquitoes need weeks to gestate. The RV will have driven off in a few days.
Its not a matter that its "inhumane" to use ar-15 bullets on a deer. In many states, its *illegal* to use an ar-15 to hunt deer. The rationale for the law is that .223 bullets shot out of an ar-15 will not "drop" a deer dead. The bullet will fly through the deer, and the deer will run away and eventually bleed to death.
A FMJ .223 bullet shot out of an ar-15 is *not* "low power". It has enough energy to fly through a (sheet rock) wall, and through a car door. (Police and soldiers are taught to "take cover" behind the engine block of a car, not merely behind a car door.)
A military issue firearm is *not* (significantly) *deadlier* than a civilian version. The military issue "assault" rifles are capable of firing more than one bullet per trigger pull. Theoretically, it makes it more deadly than a civilian issue rifle, but NO (new) automatic rifles are allowed to be sold to civilians. The Aurora nutjob did not shoot up the theatre with a "military issue" assault rifle.
The only thing that currently makes a civilian copy of an "assault rifle" deadlier than a hunting rifle is that they can carry a 30-100 bullet magazine. I personally would not object to a law that banned selling firearm magazines with more than 10 bullets. Only police and gang bangers "need" to avoid reloading after shooting more than 10 bullets. I have no doubt that the body count would have been lower if Holmes didn't have a 100 bullet drum, and had to reload after every 10 shots.
1) The PPC code never gets inserted into other machine's architectures. So in that sense, it can't possibly "bloat" the kernel. Now there could be design issues with PPC that end up being carried into future Linux kernels, but those are much harder to root out without breaking something.
2) How else can Linux keep its reputation for being able to operate obsolete stuff, long after the commercial vendor has abandoned it?
3) Anything that's in the kernel (like PPC support), has an "active" maintainer for it. As long as there's an active maintainer, there is no reason to remove anything which was built into the kernel. When there's no more active maintainer, then the feature is deprecated. Eventually, a Cardinal in the LKML group gets motivated to eradicate obsolete stuff. Then it gets cut out of the kernel.