The person takes on some risk that it might become obvious the code wasn't theirs
So are you saying is that if someone can bluff their way through a piece of code then they're ok? I'm not sure that I want these kind of people on my team. I can't imagine how I'd ever think that somebody who would submit somebody else's code and then lie about it would make a valuable addition to my group. People who cheat before they're even hired are sure to be a problem later.
Although I generally disregard minor errors, there's something to be said about being picky. At the very minimum, I expect the candidate to walk through the code and explain what it's doing. Most candidates will find an error or two along the way. Even if they miss something. that's ok, as long as it isn't something critical. Misspelled variable names and missing semi-colons are one thing. Failure to allocate memory to a pointer variable is something completely different.
We've turned candidates down because they talked a good game and could write subroutines on the whiteboard, but their larger samples were a mess that we wouldn't want to maintain.
If you're just asking for functions then you're doing something wrong. You should be asking them to solve complete problems. To understand their ability to grasp higher-level concepts, ask them non-coding questions. You can easily cover a lot of ground in an hour.
References and recommendations are as big a waste of time as looking over a piece of code that you get from an interviewee. Even if I know that the reference works for the company that they say they do, how do I know that they have any clue that they know what they're talking about? As far as I'm concerned, there's no difference between someone who has great references and someone with a high GPA from a notable university. Just because you were able to make somebody else happy with your performance doesn't mean that you'll be able to do the work that I have for you to do. The only way for me to make that determination is to test you in some way.
Anytime somebody tries to show me a code sample, the first thing I ask them is where they downloaded it from. Seriously, any employer that asks for a code sample has no clue what they're doing. They should put you at a whiteboard with a pen and have you write something on the fly.
Toshiba is pretty good about selling parts too. But there are limitations. Toshiba CPU fans and hard drives don't last very long (about a year for each, I switch to Seagate for hard drives). What really sucks about Toshiba is when you have to replace something like a hinge for the display. They have so many combinations that finding the replacement can be difficult if not impossible.
"Every professional investor has biases he brings to the table." - Experience does play a big role in investing, and I don't disagree that an advisor will bring biases to the table. I guess the issue is the level to which these biases restrict the investment strategy. For example, an advisor may recommend stocks from companies whose name begins with the letter "M" (Merck, Microsoft, etc). Wouldn't you expect your advisor to reveal that bias, even if the strategy has been effective in the past?
I think that you're missing the point. When you bring a bias like "I like open source and want to support it" into the equation, you're placing restrictions on what you'll invest in. If you've disclosed that information up front then there are no issues. That happens all the time with funds that try to be socially responsible. Beyond that, the adviser has a duty to meet the investor's objectives. In many cases this is maximum return on the investment, but not always. Some investors, like myself, want a balance between return and safety. Regardless, the advisor is responsible to meet whatever objectives I set for my portfolio. To do anything different is at the very minimum unethical.
"it's presently ~12m under it's peak use" - A lot of this depends on how you use FF. I also believe that this is a bigger problem on Linux than it is on Windows. (Not sure of your platform.) I had FF leaking several hundred megs, to the point where the system was paging so hard that it was hard to kill it. No other OSS that I was running (LAMP, Asterisk) had any serious memory issues.
I find it interesting that you've adopted *exactly* the same attitude as the guys working on FF. Even though I found this problem documented in many places on the web (including/.), nothing was being done about the problem because those who were working on the code refused to recognize it as a serious issue. So back to my original point - if I've spent money on the software then I have every right to complain. But since I didn't pay anything for FF and I didn't want to spent the time to maintain a private build, I switched to (what for me is) a more reliable browser.
I know that you're being sarcastic here, but I'll reply anyway...
"It only took them THREE MONTHS to put a patch out for it." - So how does this stack up with OSS? Let's see - FF leaks memory for many, many months and the community keeps getting told it's not a problem. I've stopped using FF for the most part because I don't want to have to restart my browser on a regular basis. I lose too much work that way.
I'd also be interested to know how any commercial software vendor ranks against vendors of other products. How long does it take for Detroit to issue a recall on a car? I don't know the number myself, but I'd be shocked if it's less than three months. A problem has to go on for a very long time before a recall is issued, even if it's for safety.
"How often we've had to knock a patch out ourselves? 1 time... in 5 years" - I don't get it, you complain that vendors aren't fast enough but then you say that only once in five years have you needed a patch quicker than the vendor can provide it?
I don't know about you, but many IT departments neither have the time nor care enough to want to fix other people's code. They are willing to work around the limitations of the software that they have, or live without a feature until the vendor makes it available. You may think that this is bone-headed, but it's reality. And F/OSS is doing little to address this issue. I applaud Ubuntu for putting out 6.06 LTS. It's a clear recognition that what matters most to the people making the decisions is stability, not features.
Why is it bad form? Maybe the author is already contributing to another open source project and doesn't have time to fix the bug. Maybe he spends his nights and weekends feeding the homeless in his community. Maybe he's just had his first child and needs to take care of his new family for a while. There are lots of good reasons to point out issues without making any further contribution.
I think that it's a mistake to assume that everyone should care about Mozilla in the same way. There are going to be lots of great coders out there that never contribute to the project, but are willing to test new builds. They offer a valuable service to the community. If you say that a person with coding skills should fix bugs then you'll drive off a lot of reporters.
What I find so interesting about this post is that it's exactly why companies are nervous about using open source for mission critical projects. You're absolutely right - since people pay nothing for the software then they can make no demands of it. I've even posted bug reports on open source projects that start out saying something like "I know that I haven't paid anything for this so I have no right to complain, but..." So give me one good reason, after reading this post, that any IT manager would want to bet their future or the future of their company on open source? At least with proprietary software you have the right to demand that things get fixed, and if you don't get what you want then you can find an alternative.
Not true. If the person creating the patch has identified both the problem and a solution, cleaning up the code is easy. I think that far too many bugs never see the light of day because nobody is willing to verify the bug, let alone identify a solution.
For those that are not able to figure it out for themselves, there is almost always professional advice. I don't disagree that choice can be overwhelming, but there are times when I'm willing to ask somebody else for their opinion. I, for one, am happy that there are many options when it comes to health care. If I need something as simple as pain reliever, I pick the one that works best for my symptoms (Advil for muscle pain, Tylenol for just about everything else). What you're suggesting is that we just have aspirin and nothing else. Sorry, but I like having options, and if I can't pick then I'll ask a friend or call a doctor.
I agree. I like how the author says that reboot could be gotten rid of and that you'd just flip the power switch. Obviously he has never remotely managed a machine.
Having taught in college myself, I never liked having student SSNs which was used as a student ID. In fact, I was pretty unhappy that my campus ID badge had my SSN printed boldly on the front. After someone lifted my credit card info and ran up a bill for $15K, I started my cleanup of private information. I sent an email to HR and asked them how I could get my SSN removed from my ID. They issued a new number and then I picked up my new ID at the campus security office. At some point they realized the risk and issued new IDs for everyone. It's just sad that anyone ever thought that an SSN would make a good ID.
What you provided isn't science, but rather a hunch. By how much do sea levels rise? What temperatures are necessary to create these conditions. By how many hectares does arable land decrease? Is the loss of arable land complete or is it just partial? How many people migrate? In what direction? Where do they settle? How is the "global civil war" started? Who is involved? Are all humans destroyed? If not, how many survive? If you expect your science to be respected then you have to start answering some pretty hard questions and backing it with some solid scientific data. Anything else is just a guess and isn't worthy of being called science.
Because science adheres to the method of changing when new evidence comes out.
This may have once been true, but it has not been this way for many years. Many theories are taken as fact even when the science behind them is suspect or there is evidence to the contrary. My favorite theory is that life exists in some form anywhere there is water present. There is absolutely no evidence to demonstrate this is true, and statistics suggests that this is not true. Yet many planetary scientists operate under this assumption. Would you not call that faith?
I take this to mean that you are predisposed to believe that the end result of any global warming is a catastrophic failure of our entire ecosystem. Do you have any evidence that suggests that this is going to be the case?
But you won't be able to convince people until the average yearly temperature about the world has passed the previously recorded high temperature.
I disagree. I believe that there are people out there like me who just want to see the science behind the theory. I've been saying for a long time now that it's not that I don't believe in global warming, but I'm very skeptical of the current science (or lack thereof) that even shows a reasonable, human contributed component. The global warming theories seemed to come out of nowhere, and they all predicted dire consequences if we didn't make changes to our behaviors right now. The problem with this approach is that it doesn't take into consideration natural consequences, in which case the correct course of action may be significantly different, like relocating huge numbers of people. If you were to prove that the causes are primarily natural and that changing our behaviors would have little or no impact, wasting our valuable resources on the wrong thing is a bad idea. In this case it's kind of like fighting a wildfire - you allow some homes to burn to save others.
I want to be clear that I'm not stating the global warming is just a natural condition over which we have no control. All I'm saying is that the scientific community has hurt itself in the debate by raising the alarm too early and not providing enough evidence to show that any warming experienced isn't part of a natural cycle.
there are ~14,000 bug reports that are unassigned to any particular developer or team, and nearly ten thousand who's status has not been changed from "unconfirmed"
If all 10,000 were "I'd like this to work better" then you don't have any problems. If they're "this doesn't work" then you do (assuming that they're not corner cases). When I logged bugs with Mandriva, they were for issues like "laptop suspend fails" and "MySQL init scripts fail". These are not corner cases if you expect to compete with commercial products. I'd agree that OSS probably gets more than it's share of bug reports that are user-specific, but then there has to be a way to work around that.
With the recent news surrounding tension on the Debian team combined with the concerns that I've heard about Ubuntu upgrades in this thread, I'm beginning to wonder about my choice to use OSS. I've been involved for about five years now and some things are just not changing. New features are high on the list and bug fixes are low. As a developer I know that it's a whole lot more fun writing a new feature than fixing an old one. But as a user I'd rather have fewer new features and a more stable platform. Say what you will about the commercial alternatives, but at least those companies have a greater incentive to fix bugs.
Slashdot Mirror
So are you saying is that if someone can bluff their way through a piece of code then they're ok? I'm not sure that I want these kind of people on my team. I can't imagine how I'd ever think that somebody who would submit somebody else's code and then lie about it would make a valuable addition to my group. People who cheat before they're even hired are sure to be a problem later.
Although I generally disregard minor errors, there's something to be said about being picky. At the very minimum, I expect the candidate to walk through the code and explain what it's doing. Most candidates will find an error or two along the way. Even if they miss something. that's ok, as long as it isn't something critical. Misspelled variable names and missing semi-colons are one thing. Failure to allocate memory to a pointer variable is something completely different.
If you're just asking for functions then you're doing something wrong. You should be asking them to solve complete problems. To understand their ability to grasp higher-level concepts, ask them non-coding questions. You can easily cover a lot of ground in an hour.
References and recommendations are as big a waste of time as looking over a piece of code that you get from an interviewee. Even if I know that the reference works for the company that they say they do, how do I know that they have any clue that they know what they're talking about? As far as I'm concerned, there's no difference between someone who has great references and someone with a high GPA from a notable university. Just because you were able to make somebody else happy with your performance doesn't mean that you'll be able to do the work that I have for you to do. The only way for me to make that determination is to test you in some way.
Anytime somebody tries to show me a code sample, the first thing I ask them is where they downloaded it from. Seriously, any employer that asks for a code sample has no clue what they're doing. They should put you at a whiteboard with a pen and have you write something on the fly.
But would you have modded it "Funny" or "Insightful"?
Toshiba is pretty good about selling parts too. But there are limitations. Toshiba CPU fans and hard drives don't last very long (about a year for each, I switch to Seagate for hard drives). What really sucks about Toshiba is when you have to replace something like a hinge for the display. They have so many combinations that finding the replacement can be difficult if not impossible.
"Every professional investor has biases he brings to the table." - Experience does play a big role in investing, and I don't disagree that an advisor will bring biases to the table. I guess the issue is the level to which these biases restrict the investment strategy. For example, an advisor may recommend stocks from companies whose name begins with the letter "M" (Merck, Microsoft, etc). Wouldn't you expect your advisor to reveal that bias, even if the strategy has been effective in the past?
I think that you're missing the point. When you bring a bias like "I like open source and want to support it" into the equation, you're placing restrictions on what you'll invest in. If you've disclosed that information up front then there are no issues. That happens all the time with funds that try to be socially responsible. Beyond that, the adviser has a duty to meet the investor's objectives. In many cases this is maximum return on the investment, but not always. Some investors, like myself, want a balance between return and safety. Regardless, the advisor is responsible to meet whatever objectives I set for my portfolio. To do anything different is at the very minimum unethical.
I find it interesting that you've adopted *exactly* the same attitude as the guys working on FF. Even though I found this problem documented in many places on the web (including /.), nothing was being done about the problem because those who were working on the code refused to recognize it as a serious issue. So back to my original point - if I've spent money on the software then I have every right to complain. But since I didn't pay anything for FF and I didn't want to spent the time to maintain a private build, I switched to (what for me is) a more reliable browser.
"It only took them THREE MONTHS to put a patch out for it." - So how does this stack up with OSS? Let's see - FF leaks memory for many, many months and the community keeps getting told it's not a problem. I've stopped using FF for the most part because I don't want to have to restart my browser on a regular basis. I lose too much work that way.
I'd also be interested to know how any commercial software vendor ranks against vendors of other products. How long does it take for Detroit to issue a recall on a car? I don't know the number myself, but I'd be shocked if it's less than three months. A problem has to go on for a very long time before a recall is issued, even if it's for safety.
"How often we've had to knock a patch out ourselves? 1 time... in 5 years" - I don't get it, you complain that vendors aren't fast enough but then you say that only once in five years have you needed a patch quicker than the vendor can provide it?
I don't know about you, but many IT departments neither have the time nor care enough to want to fix other people's code. They are willing to work around the limitations of the software that they have, or live without a feature until the vendor makes it available. You may think that this is bone-headed, but it's reality. And F/OSS is doing little to address this issue. I applaud Ubuntu for putting out 6.06 LTS. It's a clear recognition that what matters most to the people making the decisions is stability, not features.
I think that it's a mistake to assume that everyone should care about Mozilla in the same way. There are going to be lots of great coders out there that never contribute to the project, but are willing to test new builds. They offer a valuable service to the community. If you say that a person with coding skills should fix bugs then you'll drive off a lot of reporters.
What I find so interesting about this post is that it's exactly why companies are nervous about using open source for mission critical projects. You're absolutely right - since people pay nothing for the software then they can make no demands of it. I've even posted bug reports on open source projects that start out saying something like "I know that I haven't paid anything for this so I have no right to complain, but..." So give me one good reason, after reading this post, that any IT manager would want to bet their future or the future of their company on open source? At least with proprietary software you have the right to demand that things get fixed, and if you don't get what you want then you can find an alternative.
Not true. If the person creating the patch has identified both the problem and a solution, cleaning up the code is easy. I think that far too many bugs never see the light of day because nobody is willing to verify the bug, let alone identify a solution.
I think you forgot the metric choices. :-)
For those that are not able to figure it out for themselves, there is almost always professional advice. I don't disagree that choice can be overwhelming, but there are times when I'm willing to ask somebody else for their opinion. I, for one, am happy that there are many options when it comes to health care. If I need something as simple as pain reliever, I pick the one that works best for my symptoms (Advil for muscle pain, Tylenol for just about everything else). What you're suggesting is that we just have aspirin and nothing else. Sorry, but I like having options, and if I can't pick then I'll ask a friend or call a doctor.
I agree. I like how the author says that reboot could be gotten rid of and that you'd just flip the power switch. Obviously he has never remotely managed a machine.
Having taught in college myself, I never liked having student SSNs which was used as a student ID. In fact, I was pretty unhappy that my campus ID badge had my SSN printed boldly on the front. After someone lifted my credit card info and ran up a bill for $15K, I started my cleanup of private information. I sent an email to HR and asked them how I could get my SSN removed from my ID. They issued a new number and then I picked up my new ID at the campus security office. At some point they realized the risk and issued new IDs for everyone. It's just sad that anyone ever thought that an SSN would make a good ID.
And if we don't wait for science, we may take ineffective measures and exhaust our resources chasing the wrong problem.
What you provided isn't science, but rather a hunch. By how much do sea levels rise? What temperatures are necessary to create these conditions. By how many hectares does arable land decrease? Is the loss of arable land complete or is it just partial? How many people migrate? In what direction? Where do they settle? How is the "global civil war" started? Who is involved? Are all humans destroyed? If not, how many survive? If you expect your science to be respected then you have to start answering some pretty hard questions and backing it with some solid scientific data. Anything else is just a guess and isn't worthy of being called science.
This may have once been true, but it has not been this way for many years. Many theories are taken as fact even when the science behind them is suspect or there is evidence to the contrary. My favorite theory is that life exists in some form anywhere there is water present. There is absolutely no evidence to demonstrate this is true, and statistics suggests that this is not true. Yet many planetary scientists operate under this assumption. Would you not call that faith?
Humans die out based on what data? Please provide the science to back your statement.
I take this to mean that you are predisposed to believe that the end result of any global warming is a catastrophic failure of our entire ecosystem. Do you have any evidence that suggests that this is going to be the case?
I disagree. I believe that there are people out there like me who just want to see the science behind the theory. I've been saying for a long time now that it's not that I don't believe in global warming, but I'm very skeptical of the current science (or lack thereof) that even shows a reasonable, human contributed component. The global warming theories seemed to come out of nowhere, and they all predicted dire consequences if we didn't make changes to our behaviors right now. The problem with this approach is that it doesn't take into consideration natural consequences, in which case the correct course of action may be significantly different, like relocating huge numbers of people. If you were to prove that the causes are primarily natural and that changing our behaviors would have little or no impact, wasting our valuable resources on the wrong thing is a bad idea. In this case it's kind of like fighting a wildfire - you allow some homes to burn to save others.
I want to be clear that I'm not stating the global warming is just a natural condition over which we have no control. All I'm saying is that the scientific community has hurt itself in the debate by raising the alarm too early and not providing enough evidence to show that any warming experienced isn't part of a natural cycle.
If all 10,000 were "I'd like this to work better" then you don't have any problems. If they're "this doesn't work" then you do (assuming that they're not corner cases). When I logged bugs with Mandriva, they were for issues like "laptop suspend fails" and "MySQL init scripts fail". These are not corner cases if you expect to compete with commercial products. I'd agree that OSS probably gets more than it's share of bug reports that are user-specific, but then there has to be a way to work around that.
With the recent news surrounding tension on the Debian team combined with the concerns that I've heard about Ubuntu upgrades in this thread, I'm beginning to wonder about my choice to use OSS. I've been involved for about five years now and some things are just not changing. New features are high on the list and bug fixes are low. As a developer I know that it's a whole lot more fun writing a new feature than fixing an old one. But as a user I'd rather have fewer new features and a more stable platform. Say what you will about the commercial alternatives, but at least those companies have a greater incentive to fix bugs.