"As argued by Eric Raymond in his watershed essay, The Cathedral and the Bazaar, rapid, decentralized development models are inherently superior to slow, centralized ones in software development."
What about the slow, decentralized models and the rapid centralized ones? Did he have anything to say about those?
One difference between Linux and proprietary OS's is that the former is really developing at least n software projects where n is the number of distros, so the collective power is diluted by "doing your own thing".
It's really about entertainment in digital form. Record companies and movie studios have made tremendous profits from the transition from analog to digital.
In particular, music companies were able to sell CDs that cost less to manufacture than vinyl disks and charge significantly more for them. They were also able to release CDs of older music that otherwise would not be repurchased.
In recent years they've suffered from the other consequences of digital media (e.g. the ease of copying). Yet on balance, digitization has been a net positive for their bottom line.
You make it sound like going to a bar is some kind of necessity like food or water. In any case, there are many countries with tougher laws where the majority of drinkers do make arrangements to be safe.
The problem is that no analogous research can be performed for software because it's not grounded in physical laws. Potentially everything in the building code could be captured in software and it would still represent a tiny fraction of the space of possible programs.
If it's the law, than you best follow it. If you believe the law should be changed, than lobby for it or if your state allows create an initiative. As long as the law is on the books, it's dumb to argue in court that you "weren't really drunk" at.08.
Call me petty, but it doesn't bother me that offenders have to pay through the nose even if the fines don't deter them. It's not justice since the fines can't compensate for the damage caused, but it's better than nothing.
And if there were rules of thumb for programming that were as simple as those used for constructing buildings, the average construction worker could do it.
The point is that you have no way of determining what training data is "correct" because you don't know anything about what the NN is "looking at".
There also no guarantee that the network will ever converge and if it does there's no way to know if it has converged to a local minimum which isn't the solution rather than a global minimum.
Slashdot Mirror
Unfortunately, there's no particular relationship between Capitalism's "Job" and an effective and stable civilization.
"They are in trouble because of their antiquated delivery mechanism and their less than timely delivery."
Sure, by the time newspapers published their story on Paris Hilton's arrest, she had already been released.
had two identical sisters.
PointCast anyone?
Come on, everyone knows that moderation is just a lazy way to say "Right On!", "You're Wrong!", or just "Fuck You!"
On the other hand, I do get excited when people mod my jokes as funny. What can I say? I'm human.
"As argued by Eric Raymond in his watershed essay, The Cathedral and the Bazaar, rapid, decentralized development models are inherently superior to slow, centralized ones in software development."
What about the slow, decentralized models and the rapid centralized ones? Did he have anything to say about those?
One difference between Linux and proprietary OS's is that the former is really developing at least n software projects where n is the number of distros, so the collective power is diluted by "doing your own thing".
Immediately file for Chapter 11 because you might as well get all of the reorganization done all at once.
It's really about entertainment in digital form. Record companies and movie studios have made tremendous profits from the transition from analog to digital.
In particular, music companies were able to sell CDs that cost less to manufacture than vinyl disks and charge significantly more for them. They were also able to release CDs of older music that otherwise would not be repurchased.
In recent years they've suffered from the other consequences of digital media (e.g. the ease of copying). Yet on balance, digitization has been a net positive for their bottom line.
Google will gave them coupons for 10% off on gmail.
But how much of those libraries exist to achieve Java's religious beliefs on abstraction?
Make that "all of our" instead of "all are". A mind is a terrible thing to waste.
Like the widely-held belief that sorting speed is related to the software license used.
"Well, it is the best place to go get laid."
So you hear.
Don't trust, verify.
You make it sound like going to a bar is some kind of necessity like food or water. In any case, there are many countries with tougher laws where the majority of drinkers do make arrangements to be safe.
The problem is that no analogous research can be performed for software because it's not grounded in physical laws. Potentially everything in the building code could be captured in software and it would still represent a tiny fraction of the space of possible programs.
There is no such thing as U.S. Industry Standards for software development.
Had you been drinking? If not, then you've clearly identified a problem. If you were, than who knows?
If it's the law, than you best follow it. If you believe the law should be changed, than lobby for it or if your state allows create an initiative. As long as the law is on the books, it's dumb to argue in court that you "weren't really drunk" at .08.
Call me petty, but it doesn't bother me that offenders have to pay through the nose even if the fines don't deter them. It's not justice since the fines can't compensate for the damage caused, but it's better than nothing.
Of course, you could just refrain from drinking and driving.
And if there were rules of thumb for programming that were as simple as those used for constructing buildings, the average construction worker could do it.
But laws are usually based on the measured value rather then unmeasurable and subjective determination of the driver's ability to function.
The point is that you have no way of determining what training data is "correct" because you don't know anything about what the NN is "looking at".
There also no guarantee that the network will ever converge and if it does there's no way to know if it has converged to a local minimum which isn't the solution rather than a global minimum.
see http://neil.fraser.name/writing/tank/