I DENY HAVING HAD SEX WITH BARNYARD ANIMALS!!! It's true that my barn door was open, but no way did anything get out. How do I know? I JUST KNOW. And anyway, it would depend on what your definition of "barnyard" is. It was a pet. The goat was JUST A PET... which opened with a key to a minibar. Hey, so this is what they mean by "entrapment." Oh, wait, it's not.
I'd love it if someone would do this in every state where someone agreed to buy Diebold voting machines.
Wrong actor, right technique. Based on security issues alone, we know Diebold is always the wrong choice. Just by a knee jerk methodology, we could keep the machines out of people's hands for another few months each time. It would generate some press, if nothing else.
LBJ wanted his opponent accused of having sex with barnyard animals. It wasn't that he thought the charge would stick -- he wanted people to hear the candidate deny it. In this case, the response will be "well, your software is a joke -- completely insecure." We'll get to hear Diebold deny the charge. Any suit brought to force reopening analysis before purchase of Diebold's stuff would mean that, once again, they'd have to say "No, our software isn't laughably insecure. No the fact the our code showed up on the Internet isn't a problem. No, our keys are not from a hotel minibar and orderable over the Internet, and no, they're not all the same. No, we didn't miscount this race in this way or that race in that way." If they deny it enough, everyone will know that it's true. Oddly, though, in this case it actually WILL be true.
So I think we should also allege that they have sex with barnyard animals.
Small shops should and do love JOAT folks, but they may have trouble hiring them.
If a shop is large enough to have an HR department (complete with screeners) it's hard to get a foot in. It's often difficult to get a position open to hire for with the designation Jack of All Trades. Uppers don't understand, and HR certainly doesn't. (Thats a common situation, but not a rule.)
If they do get that designation, starting salary for the job will be somewhat low, because that designation will be of lower status than an "expert".
And small shops tend to pay less than big ones in the first place.
This wouldn't apply at all for a shop consisting of just one or two people, but then, they tend to pay VERY low salaries.
For the record, I'm a JOAT, too, but have been involved in a bunch of hiring.
OK, I don't need to know this, but I'm an old assembler-head: I remember how much SMALLER DRDOS was than MSDOS. Microsoft makes bloated things.
I want to know how much memory ReactOS takes up versus WindowsXP. Has someone run it who can trivially answer? Did these guys make a smaller, lighter windows?
Ok, when SATA first came out, I was told the cables were really expensive because there was a patent on the connector, and tthere were a very limited number of cable makers able to make them, and they were ALL EXPENSIVE.
Same situation? Are these cables going to be $40 for a 2ft cable?
Is anyone here well versed on the patents on the original one and what happened to bring the prices down?
I'm sorry, but this is completely wrong. I don't like Suse's SLES or the company, based on their recent stupidity, but to suggest there's never a case where this expensive distro is to forget how bad Windows is for administration. If you have a hundred Suse boxes, that's gonna run you something like $35k a year. After spending that, you can (in some cases) get by with one $75k admin running the whole show. Now if they're doing a bunch of things, you will probably need another admin, but if it's a homogeneous group of machines doing something simple, it can definitely fall out that way.
A Windows admin will usually be a bit cheaper, but A) you still have to pay for server licenses, and B) There is NO WAY a single guy can run all those boxes. I'll defer to people who have been in this situation, but I suspect you'd need three guys to keep a hundred windows server farm from imploding.
You can stump for free distros (I very much believe that's the way to go) but a blanket statement like "Novell will NEVER save you money." is nutty and undermines your entire message.
Meanwhile parasitic? Hardly. These guys spend money on R&D. The money comes from corporations. Yes, they skim money from that process, but both Novell and RedHat add value which we all benefit from. Think of it this way: they get a cut as middlemen, and the service they provide is getting Ford and Chase and Shell Oil (and whoever else has more money than you or me) to PAY FOR LINUX DEVELOPMENT. That's doing well while doing good, and you should be all for it.
TCO calculations here are not because they pay for their distros, it's because they have no CLUE what they're doing. They didn't hire good Linux admins, so they lose. Probably they handed Linux boxes to Windows admins, and they automated almost nothing. TCO for Linux can be WAY lower than Windows, but if you run Linux like windows, and deal with it like windows, it won't be lower.
Bringing us to your last line "Stupidity is its own punishment." -- I agree. Both for HSBC and Novell, here. But as they blunder forward, they step on everything we've planted. We suffer too.
"The gorilla lice needed an empty ecological niche--pubic hair--that they could occupy in order to survive. If hominids had full-body hair, the lice that already lived on it might have been able to outcompete an invader."
Someone jump in and tell me how this could possibly happen. We lost our body hair, but not our crotch, underarm or head hair. So the lice we were carrying before losing that hair...
A. Hated our pubic reagions and head.
B. were unable to adapt to those different hairs, as we lost our hair.
C. Were shocked and died the day every darn human dropped 90% of their hair, leaving only the pubic and head places available for colonization.
No. We had lice before we lost that hair, and we had lice after. There's a mistake here. This louse must have out-competed or bred together with the one we had.
You're in high school. Now is NOT the time to think about what you're going to do as a career. People will tell you it is -- they'll all have technical undergraduate degrees. You are robbing yourself of a huge chunk of life if you approach the world this way at your stage of life.
Go to college. Pick a non-tech degree. THEN CHANGE IT. CHANGE IT AGAIN. FART AROUND. Find out what you like. Learn something about literature, math, physics, chemistry (in moderation), the social sciences and the humanities in general. THEN if you really want to major in some technical thing, you might decide to do that.
But if you're a computer guy, you'll be doing computer things with the other computer guys even while you study philosophy. You'll learn way more by doing and talking with people who take classes then by actually taking them. The pace of CS classes are aimed just a little ahead of the slowest people (the F's who'll leave) -- they're way behind someone who has a natural affinity and who's doing computer work, anyhow, because he loves it.
Make friends, make contacts, learn about girls (or boys or whatever) and grow for four years. Those contacts will help you later on. Nurture them.
Get an on-campus job in the computer vein. Aim for academic computing -- the central guys, not the computer store or desktop support. Establish good relationships with your employers and faculty members.
Then, when you're done, you'll find that you can get a job through your contacts, both on-campus and graduated. You won't need or benefit from the main application process at most companies -- you'll use your network of friends to get into job interviews with people who can really evaluate your skills (meaning you'll skip past the screeners who know nothing about the job, but look for a degree).
Yes, your first job will pay less because you don't have a technical degree. After changing jobs twice, it won't matter. You'll have caught up. You might go back for a masters, anywhere in this process, and that masters can be in a tech field. In most cases, this is not cost effective.
Get a real degree from a real school (aim high, aim for a name, transfer to a better school if need be) and you'll do just fine. It's about who you know and what you can do, and, most crucially, your EXPERIENCE, not your degree. A degree will bump you in salary, but has much less of an impact than your previous earnings history. One or two job changes (don't change without working somewhere for a year) will completely negate that, if you're good. And if you're not good, you'll be happier doing something else, anyhow.
You've asked a bunch of computer people how to approach life. Look at all the answers and imagine the people behind them. Who do you want to be? No, not what job do you want to have, WHO do you want to be? Take a year to decide on a first answer even to that.
(Oh, and don't get an A+, or Network+ or other easy to get cert, unless its a path to a hard to get cert which you'll actually aquire. Baby certs only 'prove' you don't have experience. Go ahead and get a CCNA, if you want, because it'll only take you a week of concerted studying to get, but don't put it on your resume until you get a Cisco cert one above it.)
NO. 10% is not enough. Moving for 10% means you'll take anything with a slightly higher salary -- which is what you're advocating. It makes sense if you're flipping burgers, but not if the comapny is investing anything in you. It screams "I will leave you the minute I find anythig slighly better -- I don't care about you as people or your projects." You need to set a standard "I won't change out of a job I'm happy with for less than X%" which you'll modify with things like commute, project coolness or benefits, but which you'll stick to in general.
Go to CMU. If you don't love it, then try to transfer to MIT or another school you like the name of better. For some schools it's easier to transfer in. Besides, you'll have recommendations form professors with recognized names at the University level, since they'll be CMU faculty.
Those recommendations and projects you work on at CMU will get you in many doors.
Then you'll have CMU AND another top-tier University on your resume.
Because students learn less when there's a computer in front of them. There's a place for computers, and computer education, and learning to use them as tools. It's not in most classrooms.
The article's uninteresting, but did you read the discussion? There are some people who spent a lot of time posting, who quote Microsoft documents, and keep steering the discussion back to Microsoft's talking point, and away from technical points, whenever they're raised.
I don't know the people involved, and I don't know where they're coming from. But I suspect something. That suspicion colors everything I read in it.
I cannot read a discussion of my peers and believe what I read today. Every peer is possibly specifically paid to market and lie. Therefore, I have no peers.
This whole thign bugs me. Mostly because the telcos selling edge want to use it to extort money out of people. Do you think, if they got that money, they'd improve bandwidth to the edge? Heck no. These guys are some of the worst lying sacks of shit you will ever have the good fortune not to meet.
This "war game" was pure navel gazing. It's not necessarily wrong, but the methods they used to come up with this are crap. 250kbs for a VPN connection? Who is gonna set all those VPN connections up? How many companies would actually be able to use that? How far off the mark could these guys be? Of COURSE they're gonna use youtube. 90% of the 30% to 60% of workers at home will just be on phone and a little file shuffling at worst. But don't believe the numbers I pull out of MY ass, they're not gonna be better than the ones smelling of these Telco guys poop shoot.
We should have a REAL test. Make a wednesday next month stay home and work day. Let's see what really happens.
If anything, some people's crappy ISPs that over-allocate their bandwidth would be clogged...
That's called "oversubscription" and every single DSL connection and Cable Modem in the country is oversubscribed. EVERY SINGLE ONE. It's what telcos do. The phone network could not possibly handle the load if we all picked up our phones right now.
Classic oversubscription is about 8 to 1.
So TPC (The Phone Company) expects you to use about 1/8 of your bandwidth, on average. Really, they figure that you won't use it at all most of the time. Servers at home make them pee their pants. They HATE them. How do you oversubscribe with things like that going on? Oversubscription is what makes telecom so profitable. You sell something to eight customers. They all pay a "fair" price. You make 700% more than a "fair" price. Then you burn through that cash by being incredibly innefficient.
And yes, ISPs do the same thing.
Even when a company says that they don't over-subscribe, they only buy bandwidth when usage hits 80% or so. And since oversubscription WORKS, that usage amount will be roughly 1/8th of real capactity.
Mind you, I'm just responding to your "crappy ISPs" comment. I think this whole thing is silly, for reasons I'll post elsewhere.
New Employees need:
To know things (rules, passwords, techniques, etc.)
To feel things (comfort with the people around them, a small sense of 'at home')
If you get the second one, the first will take care of itself. Spend time with the person. Actual "Face" time. Teach them something (anything) about the company, so they get in the learning mode -- they'll pick up a lot in between the lines of any topic you pick. Make the rules of the company (dress codes and crap like that) as small a part of the conversation as possible, but provide them with a document on it. Play the rules down. You want them to feel free, so that they'll talk and learn and grow on their own.
Take 'em to lunch. Do it as a small group -- not necessarily the whole team, but more than just Boss and hire. Get them to be social and see people around them as lunch partners.
Then GIVE THEM A TASK WHICH WILL REQUIRE COMMUNICATION WITH THE GROUP. After they've got plugged in, the assimilation process will take care of itself.
The idea that you'd want to give them a movie or something suggests that you're not interested in spending time talking with people. I don't want to work at your company.
My first reaction to these things is almost always "Bad Study or Bad Article or Wishful Thinking." The sample size is small and focused. That in itself means that there's nothing to see here, yet. Wait for well-run studies.
But let's take it on face value and look for some explanations.
1. People who speak two languages are tolerated in dementia more than those who aren't, so their mean date of getting handed to mental health folk is higher (They seem to have controlled for this, but what they're really proving with the MMSE tests is not when demntia onsets, but rather how well the people function -- and thus, they may be showing that bilingual people cover their dementia better. But maybe the tests are well designed to factor this out.) 2. People who don't learn a second language don't learn it for a reason. Perhaps they hate outsiders. Perhaps it's anger, hatred and suspicion which is the real measure here. 3. People who speak two languages have contact with a larger germ pool. Perhaps its germ isolation which speeds dementia. 4. People who speak more than one language have more contact with a second group of people and thus have a more varied diet. Perhaps it's variety of food which slows dementia. 5. People who read more than one language see more reporting on medical studies than people who read only one language. Perhaps it's disgust with bad science in medicine or bad reporting in a sensationalist press which causes demintia. Now where did I leave my car keys... oh, maybe the lady who owns this house hates me and hid them. Again. Who is she, anyhow, and how did I get here? What does this submit button do?
I think this analysis really depends on whether these hyper-growing rabbits still have the eating requirements of their smaller cousins. If they can still survive (at their growth rate and ultimate size) on what's indigestible to humans, then clearly this is a big win. If they require big nutricious vegtables (the breeder says he currently feeds them kale) that would otherwise be edible by humans, then they aren't going to help -- they'll be less efficient than the humans eating the vegtables.
If only this could become an export for them. However for both practical and political reasons, I doubt that will happen.
Thanks for the pointer to the Trilobite. I will remember this and drop you a line when/if I get one. My time frame for this will be beginning of next year -- they're in version 2, and I expect both bug fixes and prices to go down. If there's rumor of a version 3, I'll stall.
Your money back advice is excellent.
My wife is Korean and she insists on vacuuming every other day or so in a house where we have no dog or cat and leave our shoes at the door. On one hand, it's nice to know you can ALWAYS eat off the floor (she mops a couple of times a week, too.) but on the other hand, I'd like to avoid my part of it, and want the lost hours of my wife back. (Frankly, I'm surprised I haven't heard of his stuff coming from Korea, where so many hours are spent every week doing this.) Anyhow the real key for success on this is it removing a miniscule amount of dust and grit, and not missing any part of the floor, such that you never feel anything but smooth, clean floor with your bare feet.
BTW: I can recommend the best vac I've ever used (bought it at her insistence and it was a wise move) -- the upper-end Bosch cannister. The thing sucks like crazy and is very, VERY quiet, so you can vacuum at 3am.
"Not even close. This monstrosity gives you most of the things you think you want (subject to the limitations of today's technology). It's inelegant, expensive, complicated, and does an amazingly bad job. Despite all of its extra sensors, power, and mapping, it has almost all of the limitations of Roomba - at 5x the price."
Well crap, because I followed your link and Yeah, that's the puppy. That does most of what's needed. I think the choice of battery instead of capaciters is a problem (since it can go back and recharge and resume, low wh/kg, many cycle seems like a fine situation) but the caps won't be viable for a while, anyhow, and will be expensive. But this really looks like all the things needed.
But you say it does a crappy job of vacuuming. Worse than the roomba?
Is it really a vac? The rooma's more of a duster/brush system. Is this thing a vacuum cleaner? Amazon reviews are either one star or five. Not too helpful.
I'm aiming for buying something at the beginning of next year. Sounds like things are moving in the right direction. 5x the price of a roomba would be fine, since (if it works well) it gets me out of hours of work every week.
It's hard to move. And people like Intel for support engineering. Ok. Howabout we all agree to running 32bit code in emulation, and just cut huge chunk of old legacy processor support out of Intel CPUs? Smaller die. Lower cost. Faster clock (perhaps). 32 bit would be way slow, but once you've installed your OS, it'll scream for less $.
Your points are all good -- insightful even -- but look at what you're saing:
You might want several if you have more footage. (Mapping solves this. The device needs to navigate better, and not repeat the same areas over and over. With that, you'd only need one.)...having to manually empty it and clean the filter every time it runs. (Right -- it's not a well-designed vacuum cleaner. You're spot on. That needs fixing.)
Sure, it's not great for heavy duty cleaning (Right... Why accept a need for a manual step? It should be a better vacuum.)...has the added benefit of forcing a bit of discipline as far as just keeping the floor tidy. (Why? It's your house. The device should know what it has cleaned and what it hasn't and work on areas it can get to when it can get to them. It should tell what's not available because of mats, socks, etc, and navigate around, then later come back to those, when the floor clears.)
Right now the roomba is a novelty item. If the price doubled to that of a good upright (as you point out), the roomba could eliminate vacuuming by hand. Period. It should do that. To accept this silly toyness and apologize for it is to toss away a massive market. In a few years, all non-robotic vaccum cleans could be viewed as an absurd anacronism, like typewriters.
But they need to address all these issues you just apologized for.
I think late 2007 or early 2008, these things will actually get good.
Hoover or some other real vacuum cleaner company will come out with one which actually cleans well, and a damp mop clean technique version will emerge. Roomba will add native spacial intelligence on the bots it makes, meaning actual mapping of spaces.
And most access will be read (decryption) as opposed to writes (encryption) but I have to assume (please correct me) that the encryption phase is much slower than decryption.
And even if decryption isn't slow, it's an addition, unless there's compression involved, here too, in which case you're waiting for less IO. Is it a tangible slowdown, firstly, and is there compression, secondly?
Can you talk about any or all of these three points? I'm interested in knowing more.
Slashdot Mirror
I DENY HAVING HAD SEX WITH BARNYARD ANIMALS!!! It's true that my barn door was open, but no way did anything get out. How do I know? I JUST KNOW. And anyway, it would depend on what your definition of "barnyard" is. It was a pet. The goat was JUST A PET... which opened with a key to a minibar. Hey, so this is what they mean by "entrapment." Oh, wait, it's not.
I'd love it if someone would do this in every state where someone agreed to buy Diebold voting machines.
Wrong actor, right technique. Based on security issues alone, we know Diebold is always the wrong choice. Just by a knee jerk methodology, we could keep the machines out of people's hands for another few months each time. It would generate some press, if nothing else.
LBJ wanted his opponent accused of having sex with barnyard animals. It wasn't that he thought the charge would stick -- he wanted people to hear the candidate deny it. In this case, the response will be "well, your software is a joke -- completely insecure." We'll get to hear Diebold deny the charge. Any suit brought to force reopening analysis before purchase of Diebold's stuff would mean that, once again, they'd have to say "No, our software isn't laughably insecure. No the fact the our code showed up on the Internet isn't a problem. No, our keys are not from a hotel minibar and orderable over the Internet, and no, they're not all the same. No, we didn't miscount this race in this way or that race in that way." If they deny it enough, everyone will know that it's true. Oddly, though, in this case it actually WILL be true.
So I think we should also allege that they have sex with barnyard animals.
Small shops should and do love JOAT folks, but they may have trouble hiring them.
If a shop is large enough to have an HR department (complete with screeners) it's hard to get a foot in. It's often difficult to get a position open to hire for with the designation Jack of All Trades. Uppers don't understand, and HR certainly doesn't. (Thats a common situation, but not a rule.)
If they do get that designation, starting salary for the job will be somewhat low, because that designation will be of lower status than an "expert".
And small shops tend to pay less than big ones in the first place.
This wouldn't apply at all for a shop consisting of just one or two people, but then, they tend to pay VERY low salaries.
For the record, I'm a JOAT, too, but have been involved in a bunch of hiring.
OK, I don't need to know this, but I'm an old assembler-head: I remember how much SMALLER DRDOS was than MSDOS. Microsoft makes bloated things.
I want to know how much memory ReactOS takes up versus WindowsXP. Has someone run it who can trivially answer? Did these guys make a smaller, lighter windows?
Ok, when SATA first came out, I was told the cables were really expensive because there was a patent on the connector, and tthere were a very limited number of cable makers able to make them, and they were ALL EXPENSIVE.
Same situation? Are these cables going to be $40 for a 2ft cable?
Is anyone here well versed on the patents on the original one and what happened to bring the prices down?
"Novell will NEVER save you money."
I'm sorry, but this is completely wrong. I don't like Suse's SLES or the company, based on their recent stupidity, but to suggest there's never a case where this expensive distro is to forget how bad Windows is for administration. If you have a hundred Suse boxes, that's gonna run you something like $35k a year. After spending that, you can (in some cases) get by with one $75k admin running the whole show. Now if they're doing a bunch of things, you will probably need another admin, but if it's a homogeneous group of machines doing something simple, it can definitely fall out that way.
A Windows admin will usually be a bit cheaper, but A) you still have to pay for server licenses, and B) There is NO WAY a single guy can run all those boxes. I'll defer to people who have been in this situation, but I suspect you'd need three guys to keep a hundred windows server farm from imploding.
You can stump for free distros (I very much believe that's the way to go) but a blanket statement like "Novell will NEVER save you money." is nutty and undermines your entire message.
Meanwhile parasitic? Hardly. These guys spend money on R&D. The money comes from corporations. Yes, they skim money from that process, but both Novell and RedHat add value which we all benefit from. Think of it this way: they get a cut as middlemen, and the service they provide is getting Ford and Chase and Shell Oil (and whoever else has more money than you or me) to PAY FOR LINUX DEVELOPMENT. That's doing well while doing good, and you should be all for it.
TCO calculations here are not because they pay for their distros, it's because they have no CLUE what they're doing. They didn't hire good Linux admins, so they lose. Probably they handed Linux boxes to Windows admins, and they automated almost nothing. TCO for Linux can be WAY lower than Windows, but if you run Linux like windows, and deal with it like windows, it won't be lower.
Bringing us to your last line "Stupidity is its own punishment." -- I agree. Both for HSBC and Novell, here. But as they blunder forward, they step on everything we've planted. We suffer too.
"The gorilla lice needed an empty ecological niche--pubic hair--that they could occupy in order to survive. If hominids had full-body hair, the lice that already lived on it might have been able to outcompete an invader."
Someone jump in and tell me how this could possibly happen. We lost our body hair, but not our crotch, underarm or head hair. So the lice we were carrying before losing that hair...
A. Hated our pubic reagions and head.
B. were unable to adapt to those different hairs, as we lost our hair.
C. Were shocked and died the day every darn human dropped 90% of their hair, leaving only the pubic and head places available for colonization.
No. We had lice before we lost that hair, and we had lice after. There's a mistake here. This louse must have out-competed or bred together with the one we had.
"having car manufacturers artificially prevent the cars from going over 65mph"...
Using a rate limiter velcroed to the dashboard.
Which obscured the driver's vision.
And stopped the car completely when it turned onto non-toll roads.
You're in high school. Now is NOT the time to think about what you're going to do as a career. People will tell you it is -- they'll all have technical undergraduate degrees. You are robbing yourself of a huge chunk of life if you approach the world this way at your stage of life.
Go to college. Pick a non-tech degree. THEN CHANGE IT. CHANGE IT AGAIN. FART AROUND. Find out what you like. Learn something about literature, math, physics, chemistry (in moderation), the social sciences and the humanities in general. THEN if you really want to major in some technical thing, you might decide to do that.
But if you're a computer guy, you'll be doing computer things with the other computer guys even while you study philosophy. You'll learn way more by doing and talking with people who take classes then by actually taking them. The pace of CS classes are aimed just a little ahead of the slowest people (the F's who'll leave) -- they're way behind someone who has a natural affinity and who's doing computer work, anyhow, because he loves it.
Make friends, make contacts, learn about girls (or boys or whatever) and grow for four years. Those contacts will help you later on. Nurture them.
Get an on-campus job in the computer vein. Aim for academic computing -- the central guys, not the computer store or desktop support. Establish good relationships with your employers and faculty members.
Then, when you're done, you'll find that you can get a job through your contacts, both on-campus and graduated. You won't need or benefit from the main application process at most companies -- you'll use your network of friends to get into job interviews with people who can really evaluate your skills (meaning you'll skip past the screeners who know nothing about the job, but look for a degree).
Yes, your first job will pay less because you don't have a technical degree. After changing jobs twice, it won't matter. You'll have caught up. You might go back for a masters, anywhere in this process, and that masters can be in a tech field. In most cases, this is not cost effective.
Get a real degree from a real school (aim high, aim for a name, transfer to a better school if need be) and you'll do just fine. It's about who you know and what you can do, and, most crucially, your EXPERIENCE, not your degree. A degree will bump you in salary, but has much less of an impact than your previous earnings history. One or two job changes (don't change without working somewhere for a year) will completely negate that, if you're good. And if you're not good, you'll be happier doing something else, anyhow.
You've asked a bunch of computer people how to approach life. Look at all the answers and imagine the people behind them. Who do you want to be? No, not what job do you want to have, WHO do you want to be? Take a year to decide on a first answer even to that.
(Oh, and don't get an A+, or Network+ or other easy to get cert, unless its a path to a hard to get cert which you'll actually aquire. Baby certs only 'prove' you don't have experience. Go ahead and get a CCNA, if you want, because it'll only take you a week of concerted studying to get, but don't put it on your resume until you get a Cisco cert one above it.)
NO. 10% is not enough. Moving for 10% means you'll take anything with a slightly higher salary -- which is what you're advocating. It makes sense if you're flipping burgers, but not if the comapny is investing anything in you. It screams "I will leave you the minute I find anythig slighly better -- I don't care about you as people or your projects." You need to set a standard "I won't change out of a job I'm happy with for less than X%" which you'll modify with things like commute, project coolness or benefits, but which you'll stick to in general.
For most people, that number is closer to 20%.
Go to CMU. If you don't love it, then try to transfer to MIT or another school you like the name of better. For some schools it's easier to transfer in. Besides, you'll have recommendations form professors with recognized names at the University level, since they'll be CMU faculty.
Those recommendations and projects you work on at CMU will get you in many doors.
Then you'll have CMU AND another top-tier University on your resume.
"Alternatively can anyone say why this is silly?"
Because students learn less when there's a computer in front of them. There's a place for computers, and computer education, and learning to use them as tools. It's not in most classrooms.
The article's uninteresting, but did you read the discussion? There are some people who spent a lot of time posting, who quote Microsoft documents, and keep steering the discussion back to Microsoft's talking point, and away from technical points, whenever they're raised.
I don't know the people involved, and I don't know where they're coming from. But I suspect something. That suspicion colors everything I read in it.
I cannot read a discussion of my peers and believe what I read today. Every peer is possibly specifically paid to market and lie. Therefore, I have no peers.
We need a law against astro-turfing.
This whole thign bugs me. Mostly because the telcos selling edge want to use it to extort money out of people. Do you think, if they got that money, they'd improve bandwidth to the edge? Heck no. These guys are some of the worst lying sacks of shit you will ever have the good fortune not to meet.
This "war game" was pure navel gazing. It's not necessarily wrong, but the methods they used to come up with this are crap. 250kbs for a VPN connection? Who is gonna set all those VPN connections up? How many companies would actually be able to use that? How far off the mark could these guys be? Of COURSE they're gonna use youtube. 90% of the 30% to 60% of workers at home will just be on phone and a little file shuffling at worst. But don't believe the numbers I pull out of MY ass, they're not gonna be better than the ones smelling of these Telco guys poop shoot.
We should have a REAL test. Make a wednesday next month stay home and work day. Let's see what really happens.
If anything, some people's crappy ISPs that over-allocate their bandwidth would be clogged...
That's called "oversubscription" and every single DSL connection and Cable Modem in the country is oversubscribed. EVERY SINGLE ONE. It's what telcos do. The phone network could not possibly handle the load if we all picked up our phones right now.
Classic oversubscription is about 8 to 1.
So TPC (The Phone Company) expects you to use about 1/8 of your bandwidth, on average. Really, they figure that you won't use it at all most of the time. Servers at home make them pee their pants. They HATE them. How do you oversubscribe with things like that going on? Oversubscription is what makes telecom so profitable. You sell something to eight customers. They all pay a "fair" price. You make 700% more than a "fair" price. Then you burn through that cash by being incredibly innefficient.
And yes, ISPs do the same thing.
Even when a company says that they don't over-subscribe, they only buy bandwidth when usage hits 80% or so. And since oversubscription WORKS, that usage amount will be roughly 1/8th of real capactity.
Mind you, I'm just responding to your "crappy ISPs" comment. I think this whole thing is silly, for reasons I'll post elsewhere.
New Employees need:
To know things (rules, passwords, techniques, etc.)
To feel things (comfort with the people around them, a small sense of 'at home')
If you get the second one, the first will take care of itself. Spend time with the person. Actual "Face" time. Teach them something (anything) about the company, so they get in the learning mode -- they'll pick up a lot in between the lines of any topic you pick. Make the rules of the company (dress codes and crap like that) as small a part of the conversation as possible, but provide them with a document on it. Play the rules down. You want them to feel free, so that they'll talk and learn and grow on their own.
Take 'em to lunch. Do it as a small group -- not necessarily the whole team, but more than just Boss and hire. Get them to be social and see people around them as lunch partners.
Then GIVE THEM A TASK WHICH WILL REQUIRE COMMUNICATION WITH THE GROUP. After they've got plugged in, the assimilation process will take care of itself.
The idea that you'd want to give them a movie or something suggests that you're not interested in spending time talking with people. I don't want to work at your company.
My first reaction to these things is almost always "Bad Study or Bad Article or Wishful Thinking." The sample size is small and focused. That in itself means that there's nothing to see here, yet. Wait for well-run studies.
But let's take it on face value and look for some explanations.
1. People who speak two languages are tolerated in dementia more than those who aren't, so their mean date of getting handed to mental health folk is higher (They seem to have controlled for this, but what they're really proving with the MMSE tests is not when demntia onsets, but rather how well the people function -- and thus, they may be showing that bilingual people cover their dementia better. But maybe the tests are well designed to factor this out.)
2. People who don't learn a second language don't learn it for a reason. Perhaps they hate outsiders. Perhaps it's anger, hatred and suspicion which is the real measure here.
3. People who speak two languages have contact with a larger germ pool. Perhaps its germ isolation which speeds dementia.
4. People who speak more than one language have more contact with a second group of people and thus have a more varied diet. Perhaps it's variety of food which slows dementia.
5. People who read more than one language see more reporting on medical studies than people who read only one language. Perhaps it's disgust with bad science in medicine or bad reporting in a sensationalist press which causes demintia. Now where did I leave my car keys... oh, maybe the lady who owns this house hates me and hid them. Again. Who is she, anyhow, and how did I get here? What does this submit button do?
I think this analysis really depends on whether these hyper-growing rabbits still have the eating requirements of their smaller cousins. If they can still survive (at their growth rate and ultimate size) on what's indigestible to humans, then clearly this is a big win. If they require big nutricious vegtables (the breeder says he currently feeds them kale) that would otherwise be edible by humans, then they aren't going to help -- they'll be less efficient than the humans eating the vegtables.
If only this could become an export for them. However for both practical and political reasons, I doubt that will happen.
No specs. That's what does it. Period. No design in the first place.
Its possible to iterate forward by making chunks that do parts, but in that case you need the customer to understand that. They usually don't.
Failing that you need specs. ANd the customers never have them.
Thanks for the pointer to the Trilobite. I will remember this and drop you a line when/if I get one. My time frame for this will be beginning of next year -- they're in version 2, and I expect both bug fixes and prices to go down. If there's rumor of a version 3, I'll stall.
Your money back advice is excellent.
My wife is Korean and she insists on vacuuming every other day or so in a house where we have no dog or cat and leave our shoes at the door. On one hand, it's nice to know you can ALWAYS eat off the floor (she mops a couple of times a week, too.) but on the other hand, I'd like to avoid my part of it, and want the lost hours of my wife back. (Frankly, I'm surprised I haven't heard of his stuff coming from Korea, where so many hours are spent every week doing this.) Anyhow the real key for success on this is it removing a miniscule amount of dust and grit, and not missing any part of the floor, such that you never feel anything but smooth, clean floor with your bare feet.
BTW: I can recommend the best vac I've ever used (bought it at her insistence and it was a wise move) -- the upper-end Bosch cannister. The thing sucks like crazy and is very, VERY quiet, so you can vacuum at 3am.
"Not even close. This monstrosity gives you most of the things you think you want (subject to the limitations of today's technology). It's inelegant, expensive, complicated, and does an amazingly bad job. Despite all of its extra sensors, power, and mapping, it has almost all of the limitations of Roomba - at 5x the price."
Well crap, because I followed your link and Yeah, that's the puppy. That does most of what's needed. I think the choice of battery instead of capaciters is a problem (since it can go back and recharge and resume, low wh/kg, many cycle seems like a fine situation) but the caps won't be viable for a while, anyhow, and will be expensive. But this really looks like all the things needed.
But you say it does a crappy job of vacuuming. Worse than the roomba?
Is it really a vac? The rooma's more of a duster/brush system. Is this thing a vacuum cleaner? Amazon reviews are either one star or five. Not too helpful.
I'm aiming for buying something at the beginning of next year. Sounds like things are moving in the right direction. 5x the price of a roomba would be fine, since (if it works well) it gets me out of hours of work every week.
It's hard to move. And people like Intel for support engineering. Ok. Howabout we all agree to running 32bit code in emulation, and just cut huge chunk of old legacy processor support out of Intel CPUs? Smaller die. Lower cost. Faster clock (perhaps). 32 bit would be way slow, but once you've installed your OS, it'll scream for less $.
How awful would that be?
Your points are all good -- insightful even -- but look at what you're saing:
...having to manually empty it and clean the filter every time it runs. (Right -- it's not a well-designed vacuum cleaner. You're spot on. That needs fixing.)
...has the added benefit of forcing a bit of discipline as far as just keeping the floor tidy. (Why? It's your house. The device should know what it has cleaned and what it hasn't and work on areas it can get to when it can get to them. It should tell what's not available because of mats, socks, etc, and navigate around, then later come back to those, when the floor clears.)
You might want several if you have more footage. (Mapping solves this. The device needs to navigate better, and not repeat the same areas over and over. With that, you'd only need one.)
Sure, it's not great for heavy duty cleaning (Right... Why accept a need for a manual step? It should be a better vacuum.)
Right now the roomba is a novelty item. If the price doubled to that of a good upright (as you point out), the roomba could eliminate vacuuming by hand. Period. It should do that. To accept this silly toyness and apologize for it is to toss away a massive market. In a few years, all non-robotic vaccum cleans could be viewed as an absurd anacronism, like typewriters.
But they need to address all these issues you just apologized for.
I think late 2007 or early 2008, these things will actually get good.
Hoover or some other real vacuum cleaner company will come out with one which actually cleans well, and a damp mop clean technique version will emerge. Roomba will add native spacial intelligence on the bots it makes, meaning actual mapping of spaces.
Ok, mid 2008.
And most access will be read (decryption) as opposed to writes (encryption) but I have to assume (please correct me) that the encryption phase is much slower than decryption.
And even if decryption isn't slow, it's an addition, unless there's compression involved, here too, in which case you're waiting for less IO. Is it a tangible slowdown, firstly, and is there compression, secondly?
Can you talk about any or all of these three points? I'm interested in knowing more.