Another person testing for memorization of language details.
Well, I did say that these people claimed to be "experienced". If someone needs to look up something that trivial then I wouldn't count them as experienced.
The correct answer is, of course: "It will not compile, since you forgot to provide headers".
Fair enough, and I'd give a small amount of bonus points to someone who said that before writing in the #include directives and asking the question again.
(Yes, I know the problem with gets - but smart-ass questions get smart-ass answers. And it actually does matter - who am I to say if you don't have your own version of gets?)
Actually, they don't need to know anything about gets() or puts() to answer the question correctly. If it was C++, then they would. All they need to know is how pointers are passed to functions, which is very very basic knowledge.
I've posted this before on Slashdot, so apologies for the dupe, but...
My first technical question in an interview is "what is wrong with this C code?"
void echo(void)
{
char *s;
gets(s);
puts(s);
}
Over 50% of the "experienced C coders" I interview fail to get the answer right, and this has been a constant for about the last five years. Scary, isn't it? What's even scarier is when an employer hires them after I've flagged this in the post-interview chat.
That only works properly with an intelligent manager. For example, I don't drink, and I'm not a twenty-something. I've got a full life outside of work. The boss has to understand that taking off at lunchtime to go to the pub is something that most people are cool with, not a universally acceptable trade-off for extra hours.
I should have included the words "and didn't go back to the office afterwards" there. Sure, anyone who wasn't a drinker could just head home at lunchtime instead. I'd probably have expressed a preference that they swung by the pub for a single coffee / soda / whatever on their way but it would have been by no means a requirement.
Your general point is good, though. Not everyone places the same value on things, so you can't cut the same deal with everyone.
Also, there is a difference between asking a young guy to come in for a full day on Sunday, when he would otherwise be watching a game in a bar or playing with a frisbee in the park, and asking an older guy to do the same, when -he- would otherwise be teaching a Sunday School class, or coaching his kid's softball team/
There's a balancing act though. Person A may not value those weekend hours as highly as person B, but they would be very unhappy if they saw that person B got a "better deal".
what do you get when you make a good geek mad enough to leave your company? A competitor!:-)
Heh, too true. Would say more, but someone might recognise me:)
2. You are not expected to work for more than 40 hours a week
Insert "an average of" in there, and I'd agree wholeheartedly. If you were in my field, I'd be sending my resume over now.
Re:beware of the "understanding friend" method.
on
Best Way to Manage Geeks?
·
· Score: 3, Insightful
We have names for employees like you - hourly wage earners. Someone who comes in at 7:30, punches the clock, does exactly as they're told, and goes home after they have 8 hours in, and is never expected to give anything more.
Honestly, there's very little use for those employees in an IT environment. I would make sure an employee with such an attitude was at the bottom of the pay scale, and would be constantly turned down for promotion, because it's obvious they have no motivation to better themselves.
Having been employee, manager, and business owner at various times in my career, I use the model that 40/week for a paycheck is the base deal. If one side wants something more, then they have to offer more in exchange.
For example, if I'm in a job that offers me nothing more than a paycheck, I would regard my boss asking me to work extra hours for any other reason than me screwing up as *exactly* equivalent to me saying to him "mind if I go take a few hundred from the petty cash tin?" Or: "You want me to work weekends? Then I get to telecommute when I don't need to be in for meetings."
When I was a manager I had the rule that "slack is a medium of exchange". Quiet times, everyone got off at lunchtime on Friday and went down the pub. Crunch times, we pulled crunch hours - and people were happy with that, because they accepted it as part of the trade.
When one side - employer or employee - acts as if they have a right to more than the base deal without offering anything in exchange, the other side will get very unhappy very fast. Even if circumstances force them to give what they're being asked for, the party getting screwed over will resent it happening, and that makes life worse for everyone concerned.
As far as the general question of how to manage geeks is concerned, my #1 rule was: "Happy people work harder."
Please note that I am not talking about accelerating someone with a high 80's average. I'm talking about those few kids that nail 100% time after time after time and don't bother to study becuase there is nothing to study.
I was fortunate enough to figure out that it meant that I could forget about studying for classes and spend all my time getting wasted and getting laid. Lucky I wasn't 8 years old though:)
But, as I see it, they are using tech developed, tested and spread by
American universities.
Well, they're also using a lot of fundamental tech developed by other nations as well. Software was invented in England by Ada Lovelace, the first Turing-complete computer was either German or English, etc. etc.
And if the Indians took back their invention of the digit zero, there wouldn't all that much you could send over IP:)
Do you really think letting international pettiness affect tech decisions would be a good thing? I'd call it asinine.
I agree with you, but would point out that *all* sides involved are behaving as they do for reasons that are not purely technical.
From my memory of waving the legislation he mentions (Patents Act 1977) in front of an employer during contract negotiation time, it's not only ridiculous, it's wrong. As far as I can remember, the employer only owns the rights if the IP: (a) is produced on the employer's time, or (b) is produced using the employer's equipments, or (c) relates to the employer's business activities. If none of these are true, UK law says that the ownership of the IP is the employee's.
(Though it's seven years since I left the UK - UK-based folks should double check this yourselves).
Are you really this culturally ignorant? The British, Australians, etc., in fact any English speakers who got it from the Poms & not the Yanks spell it 'referer.'
Being a Pom, please allow me to inform you that those of us who can spell correctly, spell it "referrer". Check the OED.
I can still solder but surface mount is damned hard to do by hand.
Make sure whoever makes the board puts a solder mask on it. Then you just dump a big blob of solder over the pins and suck it back off again. Works a treat...
I agree with your comments about the value of a solid understanding of CS theory, but would add a caveat: CS theory is only useful after you've spent enough time in the "real world" to understand that CS theory isn't everything.
The Rojakpot guide provides fillrate and memory bandwidth so you can directly compare cards without doing the math. These are the two most import factors in the performance of the cards.
Maybe 10 years ago this was true, but nowadays there's a whole heap of other issues that are equally significant. Quality of the JIT compilers in the driver; cache architecture; number of silicon bugs that the driver has to work around to pass WHQL; how good the AA downsampling algorithm is at adapting to GPU workload; shader architecture and how well it translates to the DX shaders... and these are just the first few thoughts that go through my head after a night's solid drinking.
A simplistic overview like this one is vaguely interesting from a historical/evolutionary perspective, but isn't really much use in terms of describing performance. As is said with all other aspects of computer benchmarking, the only relevant question is: "How fast does it run *your* application?"
Go look at the stuff on Tom's Hardware. It's a pile of biased, paid-for, crispy dog-wank (and I know this from working in the industry), but is still a million times more useful than this.
Given that all the examples on my list are open source, you have no point.
Not exactly - that pretty much is my point. To expand on my hastily typed single sentence answer:
The cost of purchasing the current version of Apache is zero, so that is not a valid excuse for not upgrading, even if you don't need the new features.
The cost of purchasing the current version of Windows is non-zero, even if you simply want something that works correctly [1] and the new features have no value to you.
Therefore, there is a difference between the two cases: for the first, I can have bugfixes for free - for the second, I have to pay a second time in order for the software I have already purchased to work correctly. (Remember - all I want is bugfixes, not a picture of dog that wags its tail while searching for files).
If you really want to look into your logic, well most commercical sofware companies are guilty since they all drop support for aging versions of their software. When was the last time Apple released a patch for OS 8?
My comment does not only apply to Microsoft. I feel the same way about any company that supplies a faulty product, whether they are in the software business or not. If the user breaks something after the purchase, then it's the user's problem. If the manufacturer shipped a broken product, then they should either fix it for free or refund the purchase price. Is that really so unreasonable?
(Now, the legal status of modifying the contract after the purchase (i.e. the EULA) may affect the manufacturer's legal obligations, but that wasn't the point of my comment).
[1] For the purposes of this comment, I'm using "correctly" to mean: (a) stable, (b) behaves according to the manufacturer's documentation, and (c) complies with all standards that it claims to comply with - which I suppose is just a subset of (b).
Running Win98/Me at this point is like running ancient versions of Linux, OpenSSH, Apache, Samba, etc... and complaining if your system gets exploited.
Except that the manufacturer does not expect me to pay them to fix problems that were present in the product at the time I purchased it.
Actually, one of my big complaints about both emerge and apt-get is figuring out what the name of the package you want to install, or what it is listed by in the package availability database.
Slashdot Mirror
Shouldn't that be "gets(&s)"? And puts() appends a '\n' automatically :)
But yes, you'd pass :)
Next question would be the good old "write strrev()"
I must admit that I'm only just now entering Chapter 2 of "The C Programming Language" by Kernighan and Ritchie.
I can't remember the order of chapters in K+R, but skip ahead to the chapter on pointers and see if you still think that the code is OK.
Another person testing for memorization of language details.
Well, I did say that these people claimed to be "experienced". If someone needs to look up something that trivial then I wouldn't count them as experienced.
The correct answer is, of course: "It will not compile, since you forgot to provide headers".
Fair enough, and I'd give a small amount of bonus points to someone who said that before writing in the #include directives and asking the question again.
(Yes, I know the problem with gets - but smart-ass questions get smart-ass answers. And it actually does matter - who am I to say if you don't have your own version of gets?)
Actually, they don't need to know anything about gets() or puts() to answer the question correctly. If it was C++, then they would. All they need to know is how pointers are passed to functions, which is very very basic knowledge.
The problem is with lazy programmers.
I've posted this before on Slashdot, so apologies for the dupe, but...
My first technical question in an interview is "what is wrong with this C code?"
void echo(void) { char *s; gets(s); puts(s); }Over 50% of the "experienced C coders" I interview fail to get the answer right, and this has been a constant for about the last five years. Scary, isn't it? What's even scarier is when an employer hires them after I've flagged this in the post-interview chat.
That only works properly with an intelligent manager. For example, I don't drink, and I'm not a twenty-something. I've got a full life outside of work. The boss has to understand that taking off at lunchtime to go to the pub is something that most people are cool with, not a universally acceptable trade-off for extra hours.
I should have included the words "and didn't go back to the office afterwards" there. Sure, anyone who wasn't a drinker could just head home at lunchtime instead. I'd probably have expressed a preference that they swung by the pub for a single coffee / soda / whatever on their way but it would have been by no means a requirement.
Your general point is good, though. Not everyone places the same value on things, so you can't cut the same deal with everyone.
Also, there is a difference between asking a young guy to come in for a full day on Sunday, when he would otherwise be watching a game in a bar or playing with a frisbee in the park, and asking an older guy to do the same, when -he- would otherwise be teaching a Sunday School class, or coaching his kid's softball team/
There's a balancing act though. Person A may not value those weekend hours as highly as person B, but they would be very unhappy if they saw that person B got a "better deal".
what do you get when you make a good geek mad enough to leave your company? A competitor! :-)
Heh, too true. Would say more, but someone might recognise me :)
2. You are not expected to work for more than 40 hours a week
Insert "an average of" in there, and I'd agree wholeheartedly. If you were in my field, I'd be sending my resume over now.
We have names for employees like you - hourly wage earners. Someone who comes in at 7:30, punches the clock, does exactly as they're told, and goes home after they have 8 hours in, and is never expected to give anything more.
Honestly, there's very little use for those employees in an IT environment. I would make sure an employee with such an attitude was at the bottom of the pay scale, and would be constantly turned down for promotion, because it's obvious they have no motivation to better themselves.
Having been employee, manager, and business owner at various times in my career, I use the model that 40/week for a paycheck is the base deal. If one side wants something more, then they have to offer more in exchange.
For example, if I'm in a job that offers me nothing more than a paycheck, I would regard my boss asking me to work extra hours for any other reason than me screwing up as *exactly* equivalent to me saying to him "mind if I go take a few hundred from the petty cash tin?" Or: "You want me to work weekends? Then I get to telecommute when I don't need to be in for meetings."
When I was a manager I had the rule that "slack is a medium of exchange". Quiet times, everyone got off at lunchtime on Friday and went down the pub. Crunch times, we pulled crunch hours - and people were happy with that, because they accepted it as part of the trade.
When one side - employer or employee - acts as if they have a right to more than the base deal without offering anything in exchange, the other side will get very unhappy very fast. Even if circumstances force them to give what they're being asked for, the party getting screwed over will resent it happening, and that makes life worse for everyone concerned.
As far as the general question of how to manage geeks is concerned, my #1 rule was: "Happy people work harder."
Please note that I am not talking about accelerating someone with a high 80's average. I'm talking about those few kids that nail 100% time after time after time and don't bother to study becuase there is nothing to study.
I was fortunate enough to figure out that it meant that I could forget about studying for classes and spend all my time getting wasted and getting laid. Lucky I wasn't 8 years old though :)
First, the facts are uncopyrightable.
Ah, but what if I created a non-factual piece of data about myself and it got entered into their system ... I think I have a plan!
But, as I see it, they are using tech developed, tested and spread by American universities.
Well, they're also using a lot of fundamental tech developed by other nations as well. Software was invented in England by Ada Lovelace, the first Turing-complete computer was either German or English, etc. etc.
And if the Indians took back their invention of the digit zero, there wouldn't all that much you could send over IP :)
Do you really think letting international pettiness affect tech decisions would be a good thing? I'd call it asinine.
I agree with you, but would point out that *all* sides involved are behaving as they do for reasons that are not purely technical.
waa-aa-ay too much access to the big red 'special effects' button.
As a friend of mine put it after watching EP3 : "Industrial Light and Magic: the greatest turd polishers ever."
Anyone know how to get rid of it?
Try creating a new directory and copy the contents into that. Then nuke the old one.
Your argument implies corruption at every level of government. On what facts are you basing this assertion?
My guess would be the last ~10,000 years of human history.
His biography of William Blake is also a very good read.
Storing someone's DNA (or any information about them) does not deprive them of life, liberty, or property.
Given that my body is my property, how can you obtain my DNA without depriving me of (some of) my property?
I found the whole IP thing completely ridiculous.
From my memory of waving the legislation he mentions (Patents Act 1977) in front of an employer during contract negotiation time, it's not only ridiculous, it's wrong. As far as I can remember, the employer only owns the rights if the IP: (a) is produced on the employer's time, or (b) is produced using the employer's equipments, or (c) relates to the employer's business activities. If none of these are true, UK law says that the ownership of the IP is the employee's.
(Though it's seven years since I left the UK - UK-based folks should double check this yourselves).
Tell me what they AREN'T changing.
The unpleasant nature of their licensing policies?
Aerobic exercise especially running is the best thing you can do for your back.
Actually, I'd suggest Tai Chi as the best thing you can do for your back, but running is a pretty good alternative.
Are you really this culturally ignorant? The British, Australians, etc., in fact any English speakers who got it from the Poms & not the Yanks spell it 'referer.'
Being a Pom, please allow me to inform you that those of us who can spell correctly, spell it "referrer". Check the OED.
I can still solder but surface mount is damned hard to do by hand.
Make sure whoever makes the board puts a solder mask on it. Then you just dump a big blob of solder over the pins and suck it back off again. Works a treat...
I agree with your comments about the value of a solid understanding of CS theory, but would add a caveat: CS theory is only useful after you've spent enough time in the "real world" to understand that CS theory isn't everything.
The Rojakpot guide provides fillrate and memory bandwidth so you can directly compare cards without doing the math. These are the two most import factors in the performance of the cards.
Maybe 10 years ago this was true, but nowadays there's a whole heap of other issues that are equally significant. Quality of the JIT compilers in the driver; cache architecture; number of silicon bugs that the driver has to work around to pass WHQL; how good the AA downsampling algorithm is at adapting to GPU workload; shader architecture and how well it translates to the DX shaders... and these are just the first few thoughts that go through my head after a night's solid drinking.
A simplistic overview like this one is vaguely interesting from a historical/evolutionary perspective, but isn't really much use in terms of describing performance. As is said with all other aspects of computer benchmarking, the only relevant question is: "How fast does it run *your* application?"
Go look at the stuff on Tom's Hardware. It's a pile of biased, paid-for, crispy dog-wank (and I know this from working in the industry), but is still a million times more useful than this.
Given that all the examples on my list are open source, you have no point.
Not exactly - that pretty much is my point. To expand on my hastily typed single sentence answer:
The cost of purchasing the current version of Apache is zero, so that is not a valid excuse for not upgrading, even if you don't need the new features.
The cost of purchasing the current version of Windows is non-zero, even if you simply want something that works correctly [1] and the new features have no value to you.
Therefore, there is a difference between the two cases: for the first, I can have bugfixes for free - for the second, I have to pay a second time in order for the software I have already purchased to work correctly. (Remember - all I want is bugfixes, not a picture of dog that wags its tail while searching for files).
If you really want to look into your logic, well most commercical sofware companies are guilty since they all drop support for aging versions of their software. When was the last time Apple released a patch for OS 8?
My comment does not only apply to Microsoft. I feel the same way about any company that supplies a faulty product, whether they are in the software business or not. If the user breaks something after the purchase, then it's the user's problem. If the manufacturer shipped a broken product, then they should either fix it for free or refund the purchase price. Is that really so unreasonable?
(Now, the legal status of modifying the contract after the purchase (i.e. the EULA) may affect the manufacturer's legal obligations, but that wasn't the point of my comment).
[1] For the purposes of this comment, I'm using "correctly" to mean: (a) stable, (b) behaves according to the manufacturer's documentation, and (c) complies with all standards that it claims to comply with - which I suppose is just a subset of (b).
Running Win98/Me at this point is like running ancient versions of Linux, OpenSSH, Apache, Samba, etc... and complaining if your system gets exploited.
Except that the manufacturer does not expect me to pay them to fix problems that were present in the product at the time I purchased it.
More like:
What is this "Daylight" you speak of?
Actually, one of my big complaints about both emerge and apt-get is figuring out what the name of the package you want to install, or what it is listed by in the package availability database.
apt-cache search foo