The problem with comparing OS X with an intel server running windows is that companies buying windows servers are most likely tied to MS for some reason (.NET). I don't see the Xserve competing with a wintel box at all.
The Apple servers do compete directly with wintel servers. As time passes, there are more and more people looking to sever their dependencies on Microsoft software, and one outlet is the new Apple servers (or Linux/Intel, or Solaris/SPARC, etc.). Basically, anything that is "the way out" of the Microsoft-only business infrastructure is legitimate competition.
One substantial issue with consultants is they will completely milk a large contract for all it is worth regardless whether the project is a success. I remember one consultant on a project who was always 80% finished. "Oh, I finished this, but that will take another two weeks..." His work wasn't all that great, but it worked just well enough to keep him going and going. However, it never worked well enough that the end product was worth the investment.
Another substantial issue with consultants is that there are so many firms that choosing one almost seems arbitrary. Knowing that there are dozens of firms in a city and that many of them are incompetent or sleazy, how does a person dig through the crap without getting burned first?
Does anyone have any idea how Buran got transported?
Russia transported it on the back of a huge cargo jet. I don't remember the jet's designation, right now, but it is bigger than the U.S. Air Force's C-5 cargo jet. I think it held/holds a world record for its size.
The huge growth of the Internet coincided with a huge increase in Windows computers on the Internet.
Now, which OS is the favorite for automated distributed denial of service attacks on the Internet? Which OS is responsible for nearly all viruses and worms on the Internet?
Truth is, Windows was never ready to be connected to a public network. The public proved this. So, there absolutely is a correlation between the number of Windows computers and the amount of cracking on the Internet.
What about other operating systems? Well, UNIX, for example, has already had its public Internet shake-down. A good example would be the story in "Cuckoo's Egg" by Clifford Stoll. As a result, UNIX is the subject of a relatively small amount of current cracking activities.
This is good, but I ask that the Slashdot community choose a new fashionable catch phrase, such as "ad nauseum" or "until the cows come home", for this week.
Wouldn't it be really neat to have computer room floor tiles each with a tight grid of white LEDs under a clear plastic covering? Imagine a computer room with these floor tiles and a single IBM server standing in the middle...
For some reason top doesn't always give me an acurate readout of how much ram my system is using.
This is generally due to whether top is reporting memory used by shared libraries and other shared memory in addition to the regular text and data parts of a program. Also, the file system cache can make the total memory use look unusually high, depending on the version of your OS.
Under Solaris 8, there is a program, pmap, which outputs a dissection of a process and its libraries. I'm sure there is something similar for Linux that could help determine exactly what top is displaying.
This can go even further...I have KDE on a system with only 128MB. Running Konquerer, Netscape 4.7X (don't have Mozilla, yet), or the other KDE office programs still results in no swapping. Looking at `top` output, however, would lead me to suggest no less RAM than 128MB, since larger apps like Mozilla would definitely be too much.
For those of you who like to run Mozilla, Open Office, Emacs, etc. all a the same time...go for 512MB. You won't regret it.
OSS is only about standards. TCP/IP is one. POSIX is another. ASCII is another. X Windows is yet another. Even the licenses are standard: GPL, BSD, etc. It's just that these kinds of standards don't happen overnight.
The variety among open source software will only get better with time. KDE and GNOME are always getting better. Open Office and Mozilla are always getting better. The OS distributions are always getting better. Etc.
Computing is still young, and natural selection is keeping OSS stirred up, for now. It will stabilze, and the result will be really nice. We just need to be patient.
10. Done. (I use Emacs) 9. Not on my watch. (I'm the only thing installing software on my workstation) 8. Not needed. (I use Emacs) 7. Done. (well-configured Solaris and/or OpenBSD) 6. Not on my watch. (Nothing leaves my site.) 5. Not needed. (No M$ software, here) 4. Not needed. (see #5) 3. Not needed. (see #8) 2. Not needed. (see #8) 1. Can do. (see #7)
Don't feel discouraged by that magazine survey. The analysis in it is horrible and doesn't normalize the data based on cost-of-living. Also, the numbers posted are salary+benefits, which counters our salary-centric thinking. Trust me, inflation still needs a lot of time before the average American programmer is earning $90,000/year.
The cost-of-living index in the U.S. easily ranges from 0.9 to 2.5. Approximately 2.5 would be for cities like New York, approximately 2.0 would be for Chicago, while 0.9 would be for a rural town in the southern U.S. Literally, New York requires 2.5 times the salary to live well than the rural southern town. The magazine survey simply ignores this fact, leaving everyone who lives in rural southern towns look poor. This couldn't be further from the truth, since it is easy to argue that the quality of life outside the city is better--less stressful, friendlier people, etc.
In general, I think magazine surveys like this are published to look good rather than to be useful. Each time I see one, I feel bad for all the college students who will see it and question their worth at a time when they are already under stress. The publishers really should either not publish these articles or, at least, hire an statistician/economist to make the numbers realistic.
Please keep in mind that $63K in Chicago, for example, is equivalent to $40K to $45K in the southern USA, due to cost of living. These surveys and studies just aren't specific about who and where is earning what. These details really matter.
In think, on average, people are earning less than they are admitting to, when the real value of a dollar is taken into account.
Yes, but if you look at the quality of the analysis, there is no way at all to see who is actually making $90K.
Magazine surveys, such as this one, are crap. I have yet to see one that normalizes the data based on cost-of-living indexes, years experience, or class of position. Just that simple step would make the data infinitely more useful. But the magazine people are lazy, stupid, and don't care; they just want to put out the numbers, even if they are wrong.
$88K in the South...yeah, right. Who?? College graduate or senior manager??
...there is a blur between programming and engineering now in IT...
Absolutely. I saw a mediocre programmer become a mediocre project manager. As expected, the project floundered, was full of high-risk dependencies, and had a kindergarden-grade database schema.
My favorite quote at the first team meeting: "I want coding to begin next week [emphasis mine]." Looking back at this, I am amazed (and disappointed) that over million dollars was wasted due to this person.
Most factory jobs today require a 2 year college degree,...
Some, certainly. Most might be a stretch. A couple of the factories I have worked at would accept anyone with two arms and a leg. A brain was optional, since brains tend to grow tired of the routine.
It really depends on the caliber of work being done. The factories I cite above were small appliance assembly plants with hordes of laborers. Truly high-tech factories, such as specialized steel mills tend to have smaller workforces of more highly trained people. Which type of factory employs more people is something I really don't know.
It is genuinely irresponsible for journals to require Word or WordPerfect files.
TeX and its many add-ons provide a truly great and open resource for scientists to record their findings. It is widely available, text based, and non-proprietary. For those scientists who can't figure out a text editor, there are GUI front-ends to TeX. If there is too much resistence to using TeX, then use one of the SGML applications (e.g. Docbook, HTML). Just don't use Word, for cripes sake!
Very true. From what I've seen, it seems that (data == Oracle) recently when developing software for the goverment.
Let's see, a low volume website with 20 tables in the schema. Of course we need: two licenses each of Oracle and iPlanet, a DBA, a webmaster/sysadmin, seven programmers, and three levels of management.
One thing I've learned is that small projects just don't fly too well when funded by the goverment. There is a lot of red tape put in place for those billion-dollar contracts, and they try to apply it to sub-million-dollar contracts as well!
In short, if you have a Sun server, Solaris is the only OS that allows you to use it to its full potential.
I'm sure Compaq put forth the needed effort to adapt Linux to the Wildfire. Perhaps, Sun will be forced to do the same for its big servers. Then, Linux == Solaris.
If this does happen, IBM vs. Sun vs. Compaq vs. et. al., when all offer big-iron Linux, will be interesting. Even though I am a fan of Sun, these companies really do compete well with eachother. Only time will tell.
I just hope that the things that make UltraSPARC, Power-4, and Alpha great won't be forgotten. It would bad for Microsoft to get replaced by Intel as "ruler of the world".
CPU2000 zealots out there, who think the Pentium 4 is the Supreme Being, don't see the larger picture, which is that Sun hardware is typically very well-rounded and very well-engineered. They build their hardware from the CPU innards on out to be consistent and robust.
Frankly, I don't care if the Xeon or Athlon do better in certain benchmarks. These comparisons are like saying a Dodge V10 pickup truck is a better sports car than a BMW M3, because it has more horsepower. Why of course a pickup truck will get better lap times than the M3...what was I thinking???
As for server hardware, IBM's x440 is a competing product to some Sun hardware, so of course it will compare favorably in some respects. No suprises, here. However, it is not exactly comparible with Sun's 8 CPU Fire 3800, for example, since the 3800 is a much larger computer. The x440 is more comparable in size with the Sun Enterprise 3500, which is a few years old, now. A better comparison to the Fire 3800 would be one of IBM's p-series machines, which are themselves better than the x-series and do compete directly with Sun's latest offerings.
The full suite, as far as I am told is far in excess of $50,000.
That's a binder almost no one buys. Solaris comes with so much by default (<$100), that additional software is really on an if-needed basis. Even then, there are Free alternatives to much of it.
For example, Forte is a really good compiler for UltraSPARC processors, but GCC is workable and is free. iPlanet is a really good application server, but Apache, Tomcat, and jBoss are workable and are free.
Buy the commercial stuff when it is warranted; use the free stuff when you just need to get by.
In short, the only time I've seen someone spend $50,000 on software is for commercial databases, high-end CAD systems, commercial application servers, or (jokingly) all the stuff needed to make Windows useful.
Slashdot Mirror
The problem with comparing OS X with an intel server running windows is that companies buying windows servers are most likely tied to MS for some reason (.NET). I don't see the Xserve competing with a wintel box at all.
The Apple servers do compete directly with wintel servers. As time passes, there are more and more people looking to sever their dependencies on Microsoft software, and one outlet is the new Apple servers (or Linux/Intel, or Solaris/SPARC, etc.). Basically, anything that is "the way out" of the Microsoft-only business infrastructure is legitimate competition.
One substantial issue with consultants is they will completely milk a large contract for all it is worth regardless whether the project is a success. I remember one consultant on a project who was always 80% finished. "Oh, I finished this, but that will take another two weeks..." His work wasn't all that great, but it worked just well enough to keep him going and going. However, it never worked well enough that the end product was worth the investment.
Another substantial issue with consultants is that there are so many firms that choosing one almost seems arbitrary. Knowing that there are dozens of firms in a city and that many of them are incompetent or sleazy, how does a person dig through the crap without getting burned first?
Does anyone have any idea how Buran got transported?
Russia transported it on the back of a huge cargo jet. I don't remember the jet's designation, right now, but it is bigger than the U.S. Air Force's C-5 cargo jet. I think it held/holds a world record for its size.
The huge growth of the Internet coincided with a huge increase in Windows computers on the Internet.
Now, which OS is the favorite for automated distributed denial of service attacks on the Internet? Which OS is responsible for nearly all viruses and worms on the Internet?
Truth is, Windows was never ready to be connected to a public network. The public proved this. So, there absolutely is a correlation between the number of Windows computers and the amount of cracking on the Internet.
What about other operating systems? Well, UNIX, for example, has already had its public Internet shake-down. A good example would be the story in "Cuckoo's Egg" by Clifford Stoll. As a result, UNIX is the subject of a relatively small amount of current cracking activities.
dual celeron 533
A high-end low-end system???
I agree.
How about "Choice in Software: The Real Way Out".
Slashdot. Pot. Kettle. Black. Rinse. Repeat.
This is good, but I ask that the Slashdot community choose a new fashionable catch phrase, such as "ad nauseum" or "until the cows come home", for this week.
now, im not saying windows is crap
Why not?
Wouldn't it be really neat to have computer room floor tiles each with a tight grid of white LEDs under a clear plastic covering? Imagine a computer room with these floor tiles and a single IBM server standing in the middle...
For some reason top doesn't always give me an acurate readout of how much ram my system is using.
This is generally due to whether top is reporting memory used by shared libraries and other shared memory in addition to the regular text and data parts of a program. Also, the file system cache can make the total memory use look unusually high, depending on the version of your OS.
Under Solaris 8, there is a program, pmap, which outputs a dissection of a process and its libraries. I'm sure there is something similar for Linux that could help determine exactly what top is displaying.
This can go even further...I have KDE on a system with only 128MB. Running Konquerer, Netscape 4.7X (don't have Mozilla, yet), or the other KDE office programs still results in no swapping. Looking at `top` output, however, would lead me to suggest no less RAM than 128MB, since larger apps like Mozilla would definitely be too much.
For those of you who like to run Mozilla, Open Office, Emacs, etc. all a the same time...go for 512MB. You won't regret it.
OSS is only about standards. TCP/IP is one. POSIX is another. ASCII is another. X Windows is yet another. Even the licenses are standard: GPL, BSD, etc. It's just that these kinds of standards don't happen overnight.
The variety among open source software will only get better with time. KDE and GNOME are always getting better. Open Office and Mozilla are always getting better. The OS distributions are always getting better. Etc.
Computing is still young, and natural selection is keeping OSS stirred up, for now. It will stabilze, and the result will be really nice. We just need to be patient.
10. Done. (I use Emacs)
9. Not on my watch. (I'm the only thing installing software on my workstation)
8. Not needed. (I use Emacs)
7. Done. (well-configured Solaris and/or OpenBSD)
6. Not on my watch. (Nothing leaves my site.)
5. Not needed. (No M$ software, here)
4. Not needed. (see #5)
3. Not needed. (see #8)
2. Not needed. (see #8)
1. Can do. (see #7)
Don't feel discouraged by that magazine survey. The analysis in it is horrible and doesn't normalize the data based on cost-of-living. Also, the numbers posted are salary+benefits, which counters our salary-centric thinking. Trust me, inflation still needs a lot of time before the average American programmer is earning $90,000/year.
The cost-of-living index in the U.S. easily ranges from 0.9 to 2.5. Approximately 2.5 would be for cities like New York, approximately 2.0 would be for Chicago, while 0.9 would be for a rural town in the southern U.S. Literally, New York requires 2.5 times the salary to live well than the rural southern town. The magazine survey simply ignores this fact, leaving everyone who lives in rural southern towns look poor. This couldn't be further from the truth, since it is easy to argue that the quality of life outside the city is better--less stressful, friendlier people, etc.
In general, I think magazine surveys like this are published to look good rather than to be useful. Each time I see one, I feel bad for all the college students who will see it and question their worth at a time when they are already under stress. The publishers really should either not publish these articles or, at least, hire an statistician/economist to make the numbers realistic.
Please keep in mind that $63K in Chicago, for example, is equivalent to $40K to $45K in the southern USA, due to cost of living. These surveys and studies just aren't specific about who and where is earning what. These details really matter.
In think, on average, people are earning less than they are admitting to, when the real value of a dollar is taken into account.
Yes, but if you look at the quality of the analysis, there is no way at all to see who is actually making $90K.
Magazine surveys, such as this one, are crap. I have yet to see one that normalizes the data based on cost-of-living indexes, years experience, or class of position. Just that simple step would make the data infinitely more useful. But the magazine people are lazy, stupid, and don't care; they just want to put out the numbers, even if they are wrong.
$88K in the South...yeah, right. Who?? College graduate or senior manager??
...there is a blur between programming and engineering now in IT...
Absolutely. I saw a mediocre programmer become a mediocre project manager. As expected, the project floundered, was full of high-risk dependencies, and had a kindergarden-grade database schema.
My favorite quote at the first team meeting: "I want coding to begin next week [emphasis mine]." Looking back at this, I am amazed (and disappointed) that over million dollars was wasted due to this person.
Most factory jobs today require a 2 year college degree, ...
Some, certainly. Most might be a stretch. A couple of the factories I have worked at would accept anyone with two arms and a leg. A brain was optional, since brains tend to grow tired of the routine.
It really depends on the caliber of work being done. The factories I cite above were small appliance assembly plants with hordes of laborers. Truly high-tech factories, such as specialized steel mills tend to have smaller workforces of more highly trained people. Which type of factory employs more people is something I really don't know.
It is genuinely irresponsible for journals to require Word or WordPerfect files.
TeX and its many add-ons provide a truly great and open resource for scientists to record their findings. It is widely available, text based, and non-proprietary. For those scientists who can't figure out a text editor, there are GUI front-ends to TeX. If there is too much resistence to using TeX, then use one of the SGML applications (e.g. Docbook, HTML). Just don't use Word, for cripes sake!
What happens if, after I get my retina-encoded license, I lose both my eyes and my fingertips in an accicent?
Will I be able to get a new license with something else on it, like my toe prints?
Very true. From what I've seen, it seems that (data == Oracle) recently when developing software for the goverment.
Let's see, a low volume website with 20 tables in the schema. Of course we need: two licenses each of Oracle and iPlanet, a DBA, a webmaster/sysadmin, seven programmers, and three levels of management.
One thing I've learned is that small projects just don't fly too well when funded by the goverment. There is a lot of red tape put in place for those billion-dollar contracts, and they try to apply it to sub-million-dollar contracts as well!
Linux will drive that box to its full potential.
Only if IBM has added what is necessary to Linux to make this possible. They may very well have done this.
In short, if you have a Sun server, Solaris is the only OS that allows you to use it to its full potential.
I'm sure Compaq put forth the needed effort to adapt Linux to the Wildfire. Perhaps, Sun will be forced to do the same for its big servers. Then, Linux == Solaris.
If this does happen, IBM vs. Sun vs. Compaq vs. et. al., when all offer big-iron Linux, will be interesting. Even though I am a fan of Sun, these companies really do compete well with eachother. Only time will tell.
I just hope that the things that make UltraSPARC, Power-4, and Alpha great won't be forgotten. It would bad for Microsoft to get replaced by Intel as "ruler of the world".
Please reread my post. For example,
CPU2000 zealots out there, who think the Pentium 4 is the Supreme Being, don't see the larger picture, which is that Sun hardware is typically very well-rounded and very well-engineered. They build their hardware from the CPU innards on out to be consistent and robust.
Frankly, I don't care if the Xeon or Athlon do better in certain benchmarks. These comparisons are like saying a Dodge V10 pickup truck is a better sports car than a BMW M3, because it has more horsepower. Why of course a pickup truck will get better lap times than the M3...what was I thinking???
As for server hardware, IBM's x440 is a competing product to some Sun hardware, so of course it will compare favorably in some respects. No suprises, here. However, it is not exactly comparible with Sun's 8 CPU Fire 3800, for example, since the 3800 is a much larger computer. The x440 is more comparable in size with the Sun Enterprise 3500, which is a few years old, now. A better comparison to the Fire 3800 would be one of IBM's p-series machines, which are themselves better than the x-series and do compete directly with Sun's latest offerings.
The full suite, as far as I am told is far in excess of $50,000.
That's a binder almost no one buys. Solaris comes with so much by default (<$100), that additional software is really on an if-needed basis. Even then, there are Free alternatives to much of it.
For example, Forte is a really good compiler for UltraSPARC processors, but GCC is workable and is free. iPlanet is a really good application server, but Apache, Tomcat, and jBoss are workable and are free.
Buy the commercial stuff when it is warranted; use the free stuff when you just need to get by.
In short, the only time I've seen someone spend $50,000 on software is for commercial databases, high-end CAD systems, commercial application servers, or (jokingly) all the stuff needed to make Windows useful.