I am replying to myself but according to www.mysql.com/doc/M/y/MySQL-PostgreSQL_features.ht ml stored procedures in multiple languages are due in version 4.1. Version 4.0 is supposed to be released next month according to the mailing list.
Also due are subselects (4.1), triggers (4.1) and foreign keys(4.0 and 4.1).
From mysql.com's todo page (http://www.mysql.com/development/todo.html) under the heading of "Things that have to be done sometime":
Stored procedures. This is currently not regarded to be very important as stored procedures are not very standardized yet. Another problem is that true stored procedures make it much harder for the optimizer and in many cases the result is slower than before We will, on the other hand, add a simple (atomic) update language that can be used to write loops and such in the MySQL server.
What are the chances that this patch might make it into an offical release of mySQL, I know that that they are reluctant to add features that are not created by themselves.
Yes I also know that I can/should recompile from source but most people will simply install the binaries.
Unfortunately when you buy software you are not actually buying the software, you are only buying a license to use the software.
If you were buying the software then various consumer protection laws come into effect. If you are licensing the software then the license can set it's create its own terms. Of course the software makers have put very favorable terms into the license.
If the courts were to decide that the licenses are a legal fiction and what really has taken place was a sale then we will see some action in in the form of lawsuits and as an improvement in software quality.
If the courts were to decide that these licenses were invalid, how would that effect the GPL? It's not a sale since no money changed hands, but if licenses are invalid the viral aspects of the license would be invalid. In essence you have taken the GPL and turned it into the BSD license. I don't know if i like that.
It basically says that Microsoft cannot be held responsible for anything, including the results of their own incompetence. Nor does it promise that the software will actually do anything useful. But if you try to steal it, it suddenly becomes extremely valuable and you are subject to prosecution to the full extent of the law.
After writing the first 3 volumes he became disgusted at the current state of the art in typesetting, so he decided to take a short break to write TEX solve that minor problem. Now many years later I assume he has solved his typesetting problem because vol 4 is becomming available. Volume 5 is scheduled to be done in 2009.
Again, untrue. Key Exhaustion may be in fact the most practical attack, but there are other attacks which involve gathering a lot of data but then analyzing the date in far less time than an exhaustive key search attack.
I would say that needing a lot of data is an understatement. The best Differential Cryptanalysis Attack would require 2^47 chosen plaintexts. Thats one million gigabityes of chosen data run through the private key. In theory that is a much easier attack than brute force but its so impractical as to be impossable outside of an acedemic context.
But I think the original question is a little odd, since anyone who hacks systems will tell you the way to go is NOT through the cryptography. It's important to have good encryption, but even single DES is too much of a pain to break as a cryptosystem: a cracker would try to find some other way to compromise the security.
Agreed, although someone with a few hundred thousand to spend can build a cracker for single DES, you are much better off trying find other weaknesses in the system or the people. If all else fails you can turn to rubber hose cryptography.
Remember it is not possible to prove that a method is secure, its only by showing it to be insecure that we learn something. So to a certian extent the security of a algorythm is based in part on trust in the methods and the designers. As it is analysed we become more (or less) confidant in that trust.
DES is by far the most analysed algorythm around and it has withstood everything that has been thrown at it. The key size is much too small but there is no known method of attack that is faster then key exhaustion.When it is extended to 3DES we have a equivalent key length of 112 bits (minimum, some research say 128+). It is not feasable to brute force 112+ bits of key.
In time the other algorythms may analysed enough to match the trust that 3DES has but until then I will stick with tried and true.
Its been known for many years (how many I don't know) that the best chess players rely on pattern matching rather than analysis to decide which move to make.
This is why a grandmaster can play dozens of people simultanously and win most of the matches, spending only a few seconds at each board before moving on to the board. In essence they are treating each turn on each board as a seperate puzzle, without considering what came before it.
What is interesting is that computers play chess much like amateurs do, by analysing each move and trying to think ahead. However computers are much faster than your average chess player and therefore can suceed at a higher level that a person can.
Slashdot Mirror
Does anyone actually support all of the SQL specification?
Also due are subselects (4.1), triggers (4.1) and foreign keys(4.0 and 4.1).
Yes I also know that I can/should recompile from source but most people will simply install the binaries.
If you were buying the software then various consumer protection laws come into effect. If you are licensing the software then the license can set it's create its own terms. Of course the software makers have put very favorable terms into the license.
If the courts were to decide that the licenses are a legal fiction and what really has taken place was a sale then we will see some action in in the form of lawsuits and as an improvement in software quality.
If the courts were to decide that these licenses were invalid, how would that effect the GPL? It's not a sale since no money changed hands, but if licenses are invalid the viral aspects of the license would be invalid. In essence you have taken the GPL and turned it into the BSD license. I don't know if i like that.
It basically says that Microsoft cannot be held responsible for anything, including the results of their own incompetence. Nor does it promise that the software will actually do anything useful. But if you try to steal it, it suddenly becomes extremely valuable and you are subject to prosecution to the full extent of the law.
EULAs are very one sided documents.
Another standard for Microsoft to "embrace and extend".
Will these people ever learn.
After writing the first 3 volumes he became disgusted at the current state of the art in typesetting, so he decided to take a short break to write TEX solve that minor problem. Now many years later I assume he has solved his typesetting problem because vol 4 is becomming available. Volume 5 is scheduled to be done in 2009.
I would say that needing a lot of data is an understatement. The best Differential Cryptanalysis Attack would require 2^47 chosen plaintexts. Thats one million gigabityes of chosen data run through the private key. In theory that is a much easier attack than brute force but its so impractical as to be impossable outside of an acedemic context.
But I think the original question is a little odd, since anyone who hacks systems will tell you the way to go is NOT through the cryptography. It's important to have good encryption, but even single DES is too much of a pain to break as a cryptosystem: a cracker would try to find some other way to compromise the security.
Agreed, although someone with a few hundred thousand to spend can build a cracker for single DES, you are much better off trying find other weaknesses in the system or the people. If all else fails you can turn to rubber hose cryptography.
DES is by far the most analysed algorythm around and it has withstood everything that has been thrown at it. The key size is much too small but there is no known method of attack that is faster then key exhaustion.When it is extended to 3DES we have a equivalent key length of 112 bits (minimum, some research say 128+). It is not feasable to brute force 112+ bits of key.
In time the other algorythms may analysed enough to match the trust that 3DES has but until then I will stick with tried and true.
This is why a grandmaster can play dozens of people simultanously and win most of the matches, spending only a few seconds at each board before moving on to the board. In essence they are treating each turn on each board as a seperate puzzle, without considering what came before it.
What is interesting is that computers play chess much like amateurs do, by analysing each move and trying to think ahead. However computers are much faster than your average chess player and therefore can suceed at a higher level that a person can.