One of the reasons NT is so expensive is the heavy duty testing that goes into the product. Are we really to beleive that MS didn't notice that they broke a major application? Software prices has nothing to do with testing costs. Testing is much less expensive per day than paying the C++ programmers who wrote the bug. Of course, fixing bugs is pretty expensive. Which means we can expect M$ to collapse of it own weight in a few years. Software pricing has nothing to do with the cost of producing software. Software pricing is determined by what the market will bear. For a PC (read inexpensive off the shelf hardware) people think that NT is the most reliable and cheapest platform for running their favorite Office Suite/web browser. BTW - If someone claims they know how much software costs to develop, be sure to not believe anything else they say.
Sun hardware is very good for moving alot of data very fast. The backplace is just fatter. I work with a data center backup product and the Sun boxes can move data an order of magnitude faster than any NT box I have seen. Ditto for high end SGI's.
X86 still has a much smaller pipe inside. You can only move data through PCI just so fast.
BTW, get back to me when you can add practically an unlimited number of drives and 64 processors that can be split into seperate logical domains.
NT/X86 has it's uses but people choose Sun hardware for more than the applications it will run. People write and optimize those apps for Sun because the hardware allows them to scale properly.
It is always fabulous when someone points out deficiencies in a piece of software. It gives everyone a goal to start shooting for. These benchmarks seemed pretty good, having the numbers and the methodology should allow the developers to fix the problems that may have cause Linux to come in second.
In my experience once and issue is identified, developers can fix it pretty quickly.
The article notes "At Novell, we just added a new title: distinguished engineer. To become a distinguished engineer, you have to get elected by your peers."
Has anyone actually seen this in practice?
I would be very concerned that this type of popularity contest might have a negative effect in the workplace. What if Joe/Josephine Friendly wins since he is always popular in the meetings, but you have been busy fixing all the bugs he/she wrote.
I would be more in favor of a source control sytem that allowed you to moderate and rate the last persons check-in.
Slashdot Mirror
This is no great loss to me.
The real loss was when Rainbow Magazine stopped publishing.
Some of you may remember this magazine which was geared towards the TRS-80 Color Computer.
I miss magazines that actually have machine language source code that you can type in yourself and run.
If you have the urge to fire up your old TRS-80 now, check out this website TRS-80 Homepage
One of the reasons NT is so expensive is the heavy duty testing that goes into the product. Are we really to beleive that MS didn't notice that they broke a major application? Software prices has nothing to do with testing costs. Testing is much less expensive per day than paying the C++ programmers who wrote the bug. Of course, fixing bugs is pretty expensive. Which means we can expect M$ to collapse of it own weight in a few years. Software pricing has nothing to do with the cost of producing software. Software pricing is determined by what the market will bear. For a PC (read inexpensive off the shelf hardware) people think that NT is the most reliable and cheapest platform for running their favorite Office Suite/web browser. BTW - If someone claims they know how much software costs to develop, be sure to not believe anything else they say.
I take issue with one part of your comment.
Sun hardware is very good for moving alot of data very fast. The backplace is just fatter. I work with a data center backup product and the Sun boxes can move data an order of magnitude faster than any NT box I have seen. Ditto for high end SGI's.
X86 still has a much smaller pipe inside. You can only move data through PCI just so fast.
BTW, get back to me when you can add practically an unlimited number of drives and 64 processors that can be split into seperate logical domains.
NT/X86 has it's uses but people choose Sun hardware for more than the applications it will run. People write and optimize those apps for Sun because the hardware allows them to scale properly.
IMHO.
It is always fabulous when someone points out deficiencies in a piece of software. It gives everyone a goal to start shooting for. These benchmarks seemed pretty good, having the numbers and the methodology should allow the developers to fix the problems that may have cause Linux to come in second.
In my experience once and issue is identified, developers can fix it pretty quickly.
timbu
The article notes "At Novell, we just added a new title: distinguished engineer. To become a distinguished engineer, you have to get elected by your peers."
Has anyone actually seen this in practice?
I would be very concerned that this type of popularity contest might have a negative effect in the workplace. What if Joe/Josephine Friendly wins since he is always popular in the meetings, but you have been busy fixing all the bugs he/she wrote.
I would be more in favor of a source control sytem that allowed you to moderate and rate the last persons check-in.