That's why this "chicken little" crap doesn't make sense. People predicted that Japan would kill us back in the 60's. They didn't. The fact is, that as a foreign economy steals jobs, it also adds consumers. This is an overall GOOD THING for the total world economy. And it's mainly the shittiest jobs getting outsourced anyway.
Interesting point, and also note that the Japanese car companies (notably Toyota) have been opening quite a few assembly lines in the US as of late. Lexus recently opened one in Canada. So I suspect the dynamics at work here and elsewhere are probably beyond the understanding of the typical doom-and-gloomer.
Even better are the problems with solutions like: "Oh you have widget.so.1 linked to widget.so.1.2, this will only work if it is linked to widget.so.1.1", you wind up with private lib versions and wrapper scripts that set linker variables.
As a developer that's the main thing that sucks about linux, frequently spec docs written by "product managers" who don't know any better don't even specify which distro should be supported, thinking all linuxes are created equal. They are not. Some of the issues can be worse than the windows "dll hell".
Linux vendors who tend to do specious things to the kernel and libraries (***hat) only make things worse. I wouldn't go so far as to say it was total balkanization, but at the very least it drives companies to only support X version of linux because they don't have time or money to do any more.
"Reuse is bad" and NIH syndrome are basically symptoms of immature software development process (and indeed immature software developers). Any complex piece of software is going to have to reuse proven libraries and strategies to be successful. Imagine having to rewrite the STL every time you needed a container class.
That brings me to a related point, and that is the irony of some developers' irrational fear of technology. For example, they might "know" C++ but refuse to use Java because they are not familiar (and therefore afraid) of it, even though it might be a better tool for the job. This problem is evident in the area of systems administration as well, for example, the guy who wants to rip out all the working PHP code and replace it with "X" vendor solution (or vice versa) because they're not familiar with the toolkit in question. That sort of thing only does a disservice to the organization because it invariably ends up costing money
The best approach is usually a degree of pragmatism -- try to work within the existing structure, and if you find yourself not understanding what you're looking at, use it as an opportunity to learn something new. Don't be arrogant and assume you're just going to rip everything out and start fresh, and somehow magically avoid the same pitfalls.
I used to be a CCSI on 3.x/4.x. "Phoneboy" was in my 3.x training course at Netrex, he was basically heckling (competing with?) the instructor the whole time. I thought it was quite amusing. (yes, the course, and the phoneboy).
You think it's bad now, you should have seen the early days of FW-1, like when they first started supporting "non Sun" platforms; NT, HP/UX, AIX. I don't know if they still support those "other" *IXes, but it sounds like business as usual with their half-ass tech support.
Don't remind me about indoor plumbing...I had a second-floor sewage pipe clog and the only way to fix it -- because the idiots who built the house did not put in a clean-out -- was to pull the pipe (6" cast iron thing) out of the ceiling, chop it in half, and pull the clog out.
Of course there was about 10-20 gallons of water behind the clog. WHOOOSH! Kitchen covered in poo.
We were bit by this one in a large multithreaded server project on Solaris. Customers with 4+ way SMP boxes were reporting crashes under heavy load. Luckily the behavior indicated heap corruption in a heavily concurrent area of the codebase -- that lead us to uncover the one rogue OS call (buried in a support library) that was causing the problem.
The developer responsible for the naughty non-reentrant call was summarily executed. (er, ok, poked with a sharp cluestick.)
Slashdot Mirror
This guy is quite a happy florist.
That's why this "chicken little" crap doesn't make sense. People predicted that Japan would kill us back in the 60's. They didn't. The fact is, that as a foreign economy steals jobs, it also adds consumers. This is an overall GOOD THING for the total world economy. And it's mainly the shittiest jobs getting outsourced anyway. Interesting point, and also note that the Japanese car companies (notably Toyota) have been opening quite a few assembly lines in the US as of late. Lexus recently opened one in Canada. So I suspect the dynamics at work here and elsewhere are probably beyond the understanding of the typical doom-and-gloomer.
Linux vendors who tend to do specious things to the kernel and libraries (***hat) only make things worse. I wouldn't go so far as to say it was total balkanization, but at the very least it drives companies to only support X version of linux because they don't have time or money to do any more.
That brings me to a related point, and that is the irony of some developers' irrational fear of technology. For example, they might "know" C++ but refuse to use Java because they are not familiar (and therefore afraid) of it, even though it might be a better tool for the job. This problem is evident in the area of systems administration as well, for example, the guy who wants to rip out all the working PHP code and replace it with "X" vendor solution (or vice versa) because they're not familiar with the toolkit in question. That sort of thing only does a disservice to the organization because it invariably ends up costing money
The best approach is usually a degree of pragmatism -- try to work within the existing structure, and if you find yourself not understanding what you're looking at, use it as an opportunity to learn something new. Don't be arrogant and assume you're just going to rip everything out and start fresh, and somehow magically avoid the same pitfalls.
Dang, they start them early these days. I was really tempted but went for this instead.
I used to be a CCSI on 3.x/4.x. "Phoneboy" was in my 3.x training course at Netrex, he was basically heckling (competing with?) the instructor the whole time. I thought it was quite amusing. (yes, the course, and the phoneboy).
You think it's bad now, you should have seen the early days of FW-1, like when they first started supporting "non Sun" platforms; NT, HP/UX, AIX. I don't know if they still support those "other" *IXes, but it sounds like business as usual with their half-ass tech support.
"NG" does not mean what you think it means.
that's a lot of m00se to get bit by...
"circus assclown" is a more accurate description.
Here's our retarded cousin larry, he's a CEO! Say hi, larry!
Usually you break things to find out what it isn't. ;-)
Don't remind me about indoor plumbing...I had a second-floor sewage pipe clog and the only way to fix it -- because the idiots who built the house did not put in a clean-out -- was to pull the pipe (6" cast iron thing) out of the ceiling, chop it in half, and pull the clog out.
Of course there was about 10-20 gallons of water behind the clog. WHOOOSH! Kitchen covered in poo.
Give me a printer on fire any day.
The OS sometimes helps you fall flat on your face as well. A textbook cause of hard to find bugs:
i th readed.html
http://developers.sun.com/solaris/articles/mult
We were bit by this one in a large multithreaded server project on Solaris. Customers with 4+ way SMP boxes were reporting crashes under heavy load. Luckily the behavior indicated heap corruption in a heavily concurrent area of the codebase -- that lead us to uncover the one rogue OS call (buried in a support library) that was causing the problem.
The developer responsible for the naughty non-reentrant call was summarily executed. (er, ok, poked with a sharp cluestick.)