My employer is kind of a hybrid, at least for those of us in I.T. We do a lot of work from home because that's where we happen to be after-hours. Even when we're in the office, we're spead between east and west coast ops. We use a lot of the same IRC and email and teleconferencing as described in the article. Works well for us.
Wasn't MySQL AB recently purchased by Oracle? Just curious...
This needs to be part of a multi-pronged approach, in which employers who exploit these super-low-wage workers are also brought to justice, and in which for those lucky illegals who've been here for so long there's no extracting them, making damn sure they pay taxes like the rest of us. And so on.
Cameras along our borders is a nifty idea, but in no way can be by itself a solution to the immigration issues in the U.S. Also, who is going to respond to alarm calls when illegal immigrants are spotted via these cameras? The National Guard? Where else could we get that kind of manpower?
"Peer to Patent" is a pretty lame euphemism. Sounds more like "Patent Invalidation Program" would be a better name. Or perhaps, "Finally Getting Around to Cleaning Up The FUD."
I have installed and operated several production Oracle db's on both RedHat and SuSE. SuSE is by far superior for Oracle. SuSE supports Oracle much better than RedHat does. It's much easier to install Oracle on SuSE, and SuSE has a very nice mailing list for Oracle dba's, with moderators from both companies. So in this sense, SuSE is a much more attractive acquisition target for Mr. Lawrence "Gotta Have It" Ellison.
All I know is: as soon as Verizon wires my neighborhood with fibre, I'm switching to it. Cable is too unreliable. I live in a densely packed neighboorhood, so I never get really fast download speeds with cable because I'm always sharing bandwidth with a hundred other people. And the cable company's infrastructure is crap. It's always having problems.
The Space Shuttle program is not a "clunker". It generated a tremendous number of technology spinoffs. It pushed the envelope of space travel. It was the envy of the free world for decades. I am extremely proud of America's accomplishments with the Space Shuttle Program.
So we have better ideas and better technology now. And we should move towards these better ideas and away from the old technology of the Space Shuttle. But keep in mind that we wouldn't have the better ideas and better technology today, were it not for the Space Shuttle Program.
Some db's need Oracle, others do not. For example a mission-critical OLTP db might need Oracle, for exactly the reasons you stated. But other db's, such as a internal reporting server that is not customer-facing, can get by with something else.
You are putting words in my mouth. I did not say that $18,000/CPU is a deal.
What I will say is this: if you don't need Oracle, then don't pay for Oracle. But if you do need Oracle, then nothing else will do. If you fall into the latter category, make sure you are getting the best deal possible.
And also keep in mind that named-user licenses might save you some money too. But be careful with named-user, because Oracle enforces a user-to-cpu ratio with named-user licenses. So named-user is really per-cpu in disguise.
$40K/CPU is full-boat retail. Anyone who pays full retail for Oracle licenses gets what they deserve. With only a little negotiation, you can get Oracle to come down 45% off retail. Or go to some vendor like CDW (I do not work for CDW), and they'll get you a nice discount.
And if you're negotiating with Oracle directly, something I do not recommend, then all you have to do is mention mySQL or PostgreSQL, and Oracle will drop their prices.
No company with a board of directors is going to bet the farm on a mission critical software product for which they cannot purchase support. You have to have someone to blame when everything falls over.
The real advantage to taking the business logic out of the db is: scalability. The db will always have much more limited scalability than the middle tier. It's much easier and less expensive to throw more app servers into production, than it is to throw more db servers in.
I've purchased many things on Ebay, including high-ticket items, and only got ripped off once, and disappointed one other time. The rip-off was for a $30 dress shirt that I never received, and the disappointment for was for a $60 "leather" desk chair that turned out to be pleather. Both sellers had one thing in common: 98% positive feedback.
Here are my rules for successful purchasing on Ebay:
1) Make sure the seller has at least 99% positive feedback.
2) Make sure the seller has at least one feedback for every dollar that you plan to spend.
For example, if the item is $100, but the seller only has 10 feedbacks, you're taking a risk. Even if he has 100% positive feedback. Similarly, if he has 30,000 feedbacks, but only 98% positive, that's also a risk. 2% of 30,000 is a lot of negative feedback.
Slashdot Mirror
Wasn't MySQL AB recently purchased by Oracle? Just curious...
Cameras along our borders is a nifty idea, but in no way can be by itself a solution to the immigration issues in the U.S. Also, who is going to respond to alarm calls when illegal immigrants are spotted via these cameras? The National Guard? Where else could we get that kind of manpower?
Why don't we just import the sugar cane itself, and process it here in the states?
As an American citizen, I have certain inalienable rights. Internet access is not one of them.
"Peer to Patent" is a pretty lame euphemism. Sounds more like "Patent Invalidation Program" would be a better name. Or perhaps, "Finally Getting Around to Cleaning Up The FUD."
I have installed and operated several production Oracle db's on both RedHat and SuSE. SuSE is by far superior for Oracle. SuSE supports Oracle much better than RedHat does. It's much easier to install Oracle on SuSE, and SuSE has a very nice mailing list for Oracle dba's, with moderators from both companies. So in this sense, SuSE is a much more attractive acquisition target for Mr. Lawrence "Gotta Have It" Ellison.
VZ FiOS has to be better. It just has to be.
So we have better ideas and better technology now. And we should move towards these better ideas and away from the old technology of the Space Shuttle. But keep in mind that we wouldn't have the better ideas and better technology today, were it not for the Space Shuttle Program.
Some db's need Oracle, others do not. For example a mission-critical OLTP db might need Oracle, for exactly the reasons you stated. But other db's, such as a internal reporting server that is not customer-facing, can get by with something else.
What I will say is this: if you don't need Oracle, then don't pay for Oracle. But if you do need Oracle, then nothing else will do. If you fall into the latter category, make sure you are getting the best deal possible.
And also keep in mind that named-user licenses might save you some money too. But be careful with named-user, because Oracle enforces a user-to-cpu ratio with named-user licenses. So named-user is really per-cpu in disguise.
And if you're negotiating with Oracle directly, something I do not recommend, then all you have to do is mention mySQL or PostgreSQL, and Oracle will drop their prices.
Pride in one's race is not racism. At least not according to m-w.com. Please explain.
SNL with guest Ed Asner, "Nuclear Plant Retiree". One of my favorites.
No company with a board of directors is going to bet the farm on a mission critical software product for which they cannot purchase support. You have to have someone to blame when everything falls over.
The real advantage to taking the business logic out of the db is: scalability. The db will always have much more limited scalability than the middle tier. It's much easier and less expensive to throw more app servers into production, than it is to throw more db servers in.
Here are my rules for successful purchasing on Ebay:
1) Make sure the seller has at least 99% positive feedback.
2) Make sure the seller has at least one feedback for every dollar that you plan to spend.
For example, if the item is $100, but the seller only has 10 feedbacks, you're taking a risk. Even if he has 100% positive feedback. Similarly, if he has 30,000 feedbacks, but only 98% positive, that's also a risk. 2% of 30,000 is a lot of negative feedback.