All the houses are going to be just an empty room above ground with an extensive basement complex. You have the choice of the basement of the Sheriff from Eureka or Kevin Smith form the last Die Hard.
Exactly how often are you trying to get a diff? I don't think I've used that feature on a source control more than half a dozen times on any project I've ever been involved on. As for retrieving a copy of the repositiory I can't believe that there is any version control system that doesn't allow you to "get latest" for an entire source tree easily. That's one of the most important (and frequent) functions of a code repository.
Are you sure that the problem isn't somewhere in the chair/keyboard interface?
Do you actually think it's flamebait or do you just disagree with his opinion?
I think his analysis is well-reasoned, well-articulated and ultimately wrong. But there is no reason to attack him just because his opinion differ from yours.
I believe that while the coming recession will have some bad aeffects on open source software, I think that most of the bigger projects have too much momentum to survive. At the end of it all, there will still be a Linux and an Apache and a MySQL and dozens of other high profile projects. The projects that are going to suffer are the literally hundreds of borderline projects on SourceForge that most people have never heard of. Many of the authors of these projects are going to abandon working on them so that they can put more face time in at the office to avoid being the one who gets down-sized to help pay for the CEO's fourth yacht and second private jet. What little extra time they do have will be (hopefully) spent with their families or in other non-technical pursuits.
Sure, switching systems is expensive. But why switch? Most times I've seen people switch it was becasue they hired a senior tech who had a religious preference for one system over another, not because the new system would really create any real advantage (other than ditching SourceSafe, which is costly but highly recommended).
The truth is that despite the amount of invective on the subject, the choice of source control tools is not going to have any measurable impact on your project. Hell, most projects could easily run without a problem on a non-buggy version of MS-SourceSafe (if such an animal existed).
The biggest cost you're going to have with your source control package is the initial setup. The biggest benefit you're going to get from your source control package is going to be minimizing that cost. Choose any of the modern source control packages and just get on with what you're being paid to do: write code.
Isn't Kentucky where tobacco comes from? Why doesn't a judge in, say, New York state order the seizure of the name Kentucky for poisoning the good people of New York?
It probablyt doesn't really matter. The judge is going to leave office soon and seek a more public office, probably running for the Senate or state governor (this can't be anything but a publicity stunt) and the order will get overturned on appeal.
From what I've seen of Windows v. Linux shops, most shops that use Linux are still made up mostly of techies while Windows shops tend to be more of a mixed bag. When I've been in Windows shops where the majority of people are technical, the ratio of techs to users seem to be much higher: in fact, in the same ranges that you have quoted for Linux.
What I'd like to see is a study comparing similar situations: average number of techs for businesses that are mostly technial or average number of techs for businesses that are mostly business (banks, insurance companies, manufacturing companies, etc.)
If they did invent the hundred year car (and this is only an opinion) it would probably be much more modular than the current cars. Some of the things that you mentionned might be difficult to replace (new braking systems, airbags, etc.) but other things like a better stereo, better seats and even a new outer shell should be easily replacable. The market would move more to replacable parts.
As an example, I've owned a Jeep for a number of yers now. While it would be a major overhaul to replace any of the functional parts of the car, I can easily replace the doors, the roof, the sport/rollbars, the seats, the hood, the bumpers, the running boards, the fenders, the suspension, the wheels, the rear lights, the hood, etc. This has fed a huge add-in market for jeep parts. There is at least one store here in Toronto that actually specializes in parts for the Jeep.
As long as the basic chassis and the engine remain the same, and as long as the cars are built modularly, it should be a lot less expensive when you want to upgrade. Imagine buying what is essentially a Honda Accord for $40,000 and then every ten years for the rest of your life paying $5,000 - $10,000 to change everything that you can so that it's a completely new car. Assuming that you drive for 60 years: lifetime cost of buying Honda Accords (~$30,000 every 6 years): $300,000; lifetime cost of a long lasting modular car: $90,000. I think enough people are smart enough to respond to numbers like that.
What I'm saying is that comparing a hundred year car costing twice as much but lasting twelve times as long to a current car is the same as comparing a current car to a car costing half as much and lasting 1/12 of the time.
Most production cars today are made to last somewhere between four to eight years. That's less than your minimum of ten years and a far cry from thirty years.
Where did you get the metrics that say.net programs are buggy, have barely passable performance and poor user experience? Is there a Gartner study that I missed? And with the Mono project,.NET is almost as portable as Java (there are still some mainframes that Java runs on but.NET doesn't).
If you're getting a lot of hangs and crashes in Java, I would suggest that is more reflective of the quality of the programmer writing the code that you are running and not the quality of the Java VM.
As to your final point, the fact that you wrap your code in try/catch blocks and hide the errors means exactly what you say: you haven't seen any errors. And exception handlers aren't going to catch memory leaks, which are the main source of heisenbugs in a non-managed language (a heisenbug is a bug that seems to appear only randomly because they are highly dependent on machine state issues that are way outside of the scope of the code being debugged; they are particularly hard to fix because they rarely or never appear in a controlled environment where they can be studied).
As for you final point, that C++ can do everything that.NET does and more: you're perfectly correct. And the corresponding code will run faster. It may or may not be more reliable, chances are it won't be because there is a whole class of memory management bugs that can crop up. As for the positive user experience, that is entirely due to the quality of the team developing the software and has very little to do with the tool being used.
However, you didn't address my main point which is that the cost of development using.NET is a small fraction of the cost of development for the same program using C++.
I think you're right as long as the more reliable cars lasted no more than twice as long as the less reliable cars.
On the other hand, when you're talking about a hundred year car, the market will probably change.
Think of it this way: the cheapest car currently on the market is probably the Kia Rio or the Hyundai Accent, both of which run ~$14,000. Would you pay $7,000 for a car that you were going to throw away (that's throw away, not sell or trade in) in 6 months?
My point is that people don't think about the browser, their focus is on the site they're at. It's only technical people like us (and I assume that you are in some way technical from the way you write) that think about things like the browser and the operating system. To most people searching the internet, they are using the computer and they are using Google. They're not even aware that they are using Linux or Windows, wouldn't be able to tell you what a shell is, and are only vaguely aware what the browser is. Because IE is part of the background, they won't think to blame IE.
If MySpace started screwing around so that they didn't work in IE, they would have the same problem. IT would take longer because people have invested a lot of time in their MySpace pages and so will accept a little more pain to avoid switching but eventually more and more people would move to Facebook and MySpace would disappear.
The big car companies aren't going to ally with Tesla for the simple reason that they don't want electric cars to succeed. The problem is that electric engines last a lot longer than internal combustion engines. Combine that with a car made mostly of aluminum, fiberglass and plastic and you get a car that you can buy as soon as you learn to drive and leave to the grandkids in your will. The business model of the big car companies need people to replace their cars every few years. If cars were made to last a hundred years, the auto manufacturing segment would be made up of a few companies of about 1000 emplyees each.
Most people (and when I say most people I mean non-technical people) don't think about their choice of browser. When they want to search for something, they "Google" it, the don't "Internet Explorer" it. They'll only understand that "I went to Google and it didn't work. What else can I use?"
And Google is not stupid enough to alienate 80% of the people surfing the web, especially when they depend so much on their search business to drive their advertising revenue.
Sure, any modern language is going to be slower than the equivalent written in C/C++. The trade-off is that it's also going to cost a small fraction to write. No to mention that managed languages avoid a whole class of heisenbugs that have plagued C/C++ for decades.
Also I'm not sure what you mean by "more bloated and has worse performance than Java". In terms of performance, Java and.NET are pretty similar. As for bloated I'm going to assume that you mean the libraries. If that is the case, what should Microsoft from the libraries? Support for data access? Regular expressions? HTTP support? All of those libraries are there for a reason: somebody needs them.
From reading the article you cited, the reason for the lost e-mails had nothing to do with the choice of software and everything to do with boneheaded consultants moving thousands of files manually instead of automating the process to prevent human error.
That's assuming that you believe the story that Ars is quoting.
I think it's more likely that the white house deleted all those emails and then came up with a convenient excuse as to why they were gone because they didn't want anybody to examine the true record of what was heppening at the time.
Sun's drop from the playing field had nothing to do with Microsoft's monopoly. Microsoft was accused of forcing companies like Dell to install only Windows on the machines they sold. Sun sold it's own servers and didn't realize that the market was shifting away from high prices for perceived high value to more of a commodity pricing model. They lost market share to Microsoft for sound economic reasons that had nothing to do with Microsoft's monopoly. It was just that Microsoft Windows on an Intel box would do the job for less than half the price of the equivalent Sun server.
But there's more to TV than just a signal. Most people who want to watch TV need this large box with a weird glass insert called a "screen" in it. If you don't have that, it's kind of difficult to watch a television broadcast no matter how good an antenna you have.
Slashdot Mirror
The cynical me wonders when the Open Source community will abandon the OpenID standard now that Microsoft has committed to it.
Sounds great, you can pay for it.
I haven't seen any real estimates of the cost of moving to IPV6, but it's going to be substantial. How much do you have in your wallet?
All the houses are going to be just an empty room above ground with an extensive basement complex. You have the choice of the basement of the Sheriff from Eureka or Kevin Smith form the last Die Hard.
Exactly how often are you trying to get a diff? I don't think I've used that feature on a source control more than half a dozen times on any project I've ever been involved on. As for retrieving a copy of the repositiory I can't believe that there is any version control system that doesn't allow you to "get latest" for an entire source tree easily. That's one of the most important (and frequent) functions of a code repository.
Are you sure that the problem isn't somewhere in the chair/keyboard interface?
Do you actually think it's flamebait or do you just disagree with his opinion?
I think his analysis is well-reasoned, well-articulated and ultimately wrong. But there is no reason to attack him just because his opinion differ from yours.
I believe that while the coming recession will have some bad aeffects on open source software, I think that most of the bigger projects have too much momentum to survive. At the end of it all, there will still be a Linux and an Apache and a MySQL and dozens of other high profile projects. The projects that are going to suffer are the literally hundreds of borderline projects on SourceForge that most people have never heard of. Many of the authors of these projects are going to abandon working on them so that they can put more face time in at the office to avoid being the one who gets down-sized to help pay for the CEO's fourth yacht and second private jet. What little extra time they do have will be (hopefully) spent with their families or in other non-technical pursuits.
Sure, switching systems is expensive. But why switch? Most times I've seen people switch it was becasue they hired a senior tech who had a religious preference for one system over another, not because the new system would really create any real advantage (other than ditching SourceSafe, which is costly but highly recommended).
The truth is that despite the amount of invective on the subject, the choice of source control tools is not going to have any measurable impact on your project. Hell, most projects could easily run without a problem on a non-buggy version of MS-SourceSafe (if such an animal existed).
The biggest cost you're going to have with your source control package is the initial setup. The biggest benefit you're going to get from your source control package is going to be minimizing that cost. Choose any of the modern source control packages and just get on with what you're being paid to do: write code.
Isn't Kentucky where tobacco comes from? Why doesn't a judge in, say, New York state order the seizure of the name Kentucky for poisoning the good people of New York?
It probablyt doesn't really matter. The judge is going to leave office soon and seek a more public office, probably running for the Senate or state governor (this can't be anything but a publicity stunt) and the order will get overturned on appeal.
From what I've seen of Windows v. Linux shops, most shops that use Linux are still made up mostly of techies while Windows shops tend to be more of a mixed bag. When I've been in Windows shops where the majority of people are technical, the ratio of techs to users seem to be much higher: in fact, in the same ranges that you have quoted for Linux.
What I'd like to see is a study comparing similar situations: average number of techs for businesses that are mostly technial or average number of techs for businesses that are mostly business (banks, insurance companies, manufacturing companies, etc.)
Isn't 10:04 about when the effects of that third shot of scotch is starting to kick in.
in a study done by me, my 11 year old jeep is also worth $25 billion.
I wonder how much the Linux ecosystem would be worth if it were valued by an organization that didn't have a vested interest.
If they did invent the hundred year car (and this is only an opinion) it would probably be much more modular than the current cars. Some of the things that you mentionned might be difficult to replace (new braking systems, airbags, etc.) but other things like a better stereo, better seats and even a new outer shell should be easily replacable. The market would move more to replacable parts.
As an example, I've owned a Jeep for a number of yers now. While it would be a major overhaul to replace any of the functional parts of the car, I can easily replace the doors, the roof, the sport/rollbars, the seats, the hood, the bumpers, the running boards, the fenders, the suspension, the wheels, the rear lights, the hood, etc. This has fed a huge add-in market for jeep parts. There is at least one store here in Toronto that actually specializes in parts for the Jeep.
As long as the basic chassis and the engine remain the same, and as long as the cars are built modularly, it should be a lot less expensive when you want to upgrade. Imagine buying what is essentially a Honda Accord for $40,000 and then every ten years for the rest of your life paying $5,000 - $10,000 to change everything that you can so that it's a completely new car. Assuming that you drive for 60 years: lifetime cost of buying Honda Accords (~$30,000 every 6 years): $300,000; lifetime cost of a long lasting modular car: $90,000. I think enough people are smart enough to respond to numbers like that.
What I'm saying is that comparing a hundred year car costing twice as much but lasting twelve times as long to a current car is the same as comparing a current car to a car costing half as much and lasting 1/12 of the time.
Most production cars today are made to last somewhere between four to eight years. That's less than your minimum of ten years and a far cry from thirty years.
Where did you get the metrics that say .net programs are buggy, have barely passable performance and poor user experience? Is there a Gartner study that I missed? And with the Mono project, .NET is almost as portable as Java (there are still some mainframes that Java runs on but .NET doesn't).
If you're getting a lot of hangs and crashes in Java, I would suggest that is more reflective of the quality of the programmer writing the code that you are running and not the quality of the Java VM.
As to your final point, the fact that you wrap your code in try/catch blocks and hide the errors means exactly what you say: you haven't seen any errors. And exception handlers aren't going to catch memory leaks, which are the main source of heisenbugs in a non-managed language (a heisenbug is a bug that seems to appear only randomly because they are highly dependent on machine state issues that are way outside of the scope of the code being debugged; they are particularly hard to fix because they rarely or never appear in a controlled environment where they can be studied).
As for you final point, that C++ can do everything that .NET does and more: you're perfectly correct. And the corresponding code will run faster. It may or may not be more reliable, chances are it won't be because there is a whole class of memory management bugs that can crop up. As for the positive user experience, that is entirely due to the quality of the team developing the software and has very little to do with the tool being used.
However, you didn't address my main point which is that the cost of development using .NET is a small fraction of the cost of development for the same program using C++.
I think you're right as long as the more reliable cars lasted no more than twice as long as the less reliable cars.
On the other hand, when you're talking about a hundred year car, the market will probably change.
Think of it this way: the cheapest car currently on the market is probably the Kia Rio or the Hyundai Accent, both of which run ~$14,000. Would you pay $7,000 for a car that you were going to throw away (that's throw away, not sell or trade in) in 6 months?
Which is exactly my point.
My point is that people don't think about the browser, their focus is on the site they're at. It's only technical people like us (and I assume that you are in some way technical from the way you write) that think about things like the browser and the operating system. To most people searching the internet, they are using the computer and they are using Google. They're not even aware that they are using Linux or Windows, wouldn't be able to tell you what a shell is, and are only vaguely aware what the browser is. Because IE is part of the background, they won't think to blame IE.
If MySpace started screwing around so that they didn't work in IE, they would have the same problem. IT would take longer because people have invested a lot of time in their MySpace pages and so will accept a little more pain to avoid switching but eventually more and more people would move to Facebook and MySpace would disappear.
The big car companies aren't going to ally with Tesla for the simple reason that they don't want electric cars to succeed. The problem is that electric engines last a lot longer than internal combustion engines. Combine that with a car made mostly of aluminum, fiberglass and plastic and you get a car that you can buy as soon as you learn to drive and leave to the grandkids in your will. The business model of the big car companies need people to replace their cars every few years. If cars were made to last a hundred years, the auto manufacturing segment would be made up of a few companies of about 1000 emplyees each.
Most people (and when I say most people I mean non-technical people) don't think about their choice of browser. When they want to search for something, they "Google" it, the don't "Internet Explorer" it. They'll only understand that "I went to Google and it didn't work. What else can I use?"
And Google is not stupid enough to alienate 80% of the people surfing the web, especially when they depend so much on their search business to drive their advertising revenue.
Sure, any modern language is going to be slower than the equivalent written in C/C++. The trade-off is that it's also going to cost a small fraction to write. No to mention that managed languages avoid a whole class of heisenbugs that have plagued C/C++ for decades.
Also I'm not sure what you mean by "more bloated and has worse performance than Java". In terms of performance, Java and .NET are pretty similar. As for bloated I'm going to assume that you mean the libraries. If that is the case, what should Microsoft from the libraries? Support for data access? Regular expressions? HTTP support? All of those libraries are there for a reason: somebody needs them.
.NET is gay? The best you can do is a school yard taunt worthy of an 8 year old?
If you're an example of what I could expect if I were to move into the open source world I'll gladly stay where I am with the adults.
I think it would be a great boon for LiveSearch.
Let's face it, the average user isn't going to think to blame their browser, they're going to blame Google.
From reading the article you cited, the reason for the lost e-mails had nothing to do with the choice of software and everything to do with boneheaded consultants moving thousands of files manually instead of automating the process to prevent human error.
That's assuming that you believe the story that Ars is quoting.
I think it's more likely that the white house deleted all those emails and then came up with a convenient excuse as to why they were gone because they didn't want anybody to examine the true record of what was heppening at the time.
Sun's drop from the playing field had nothing to do with Microsoft's monopoly. Microsoft was accused of forcing companies like Dell to install only Windows on the machines they sold. Sun sold it's own servers and didn't realize that the market was shifting away from high prices for perceived high value to more of a commodity pricing model. They lost market share to Microsoft for sound economic reasons that had nothing to do with Microsoft's monopoly. It was just that Microsoft Windows on an Intel box would do the job for less than half the price of the equivalent Sun server.
But there's more to TV than just a signal. Most people who want to watch TV need this large box with a weird glass insert called a "screen" in it. If you don't have that, it's kind of difficult to watch a television broadcast no matter how good an antenna you have.