And sue shoecorp, the manufacturer of the shoes they were wearing, because they didn't implement Other Corp's Patented shoeprotect hammer-drop protection technology in the shoe, to protect against heavy objects being dropped on the foot.
For bonus points, buy an Other Corp protected shoe, and drop something on the foot so heavy that the protection fails, then sue Othercorp.
Indeed. the whole purpose of a patent is to protect novel inventions from being used by people other than the inventor.
The consumer knew or was responsible to know that "Saw Stop" was a patented technology, and any device not advertising the technology implicitly does not have a similar safety feature.
A patent holder has a right to deny use altogether, or charge exhorbitatnt fees to exercise their patent.
So the person suing the company suggests, if the patent holder only sold or licensed the technology to one or two companies, only those companies can manufacture and market saws, and also, only their models the tech. is incorporated in can be sold (despite robust consumer demands for older models or refurbished units that don't have the tech) ?
Because the technology is unique and novel.
Novel and unique inventions are by definition not standard and probably not contained in the competitor's products.
If the government wants to mandate a type of safety measure, that's one thing (but they will probably also have to buy out the patent holder, or provide an option everyone can implement without imposed royalties).
In this case, there was no law mandating the safety feature.
There is no standard of safety that is violated by not having the feature.
The typical lawsuit against a healthcare provider is not suing over any obvious error, it's about suing over an outcome, and claiming healthcare provider should have done some thing in addition to or other than what they did (even if that thing is not common, or standard process).
For example "They should have tested for extremely rare disease PQRS".
They didn't, the patient was diagnosed with something less severe, and they died, now their family sues.
So... they get these huge legal expenses.
But more importantly... now suddenly they will have to start testing all people for PQRS, even though it's a genetic disease that effects 1 out of 10 million.
Every single person will have to be tested, to avoid future lawsuits.
So the lawsuit itself increases costs, but so do lawsuit-created modifications to medical practice: providers now have to waste money and compel tests and practices that the average person would not want to pay for, just to avoid the possibility of lawsuits.
A lot of expenses have to be imposed in 'practice modifications', and paperwork.
Your typical example is also people suing over something that went poorly for them, even though normal medical procedures were followed.
For example, patient died of X or suffered Y, when they were being treated for Z.
Even if X was not a reasonably avoidable outcome, or X/Y would not be prevented by generally accepted medical practice, there is a chance that a lawsuit will still be generated.
People sue when bad things happen to them.
Of course all bad things are always someone else's fault.
And the medical provider must be in full control at all times of everything that happens right?
If you leave with condition Y, or (for one reason or another) they can't or fail to cure it, it must be their fault.
If they induce condition Y, but it's not any error on their part, or problem with their facilities, then their practitioners really must still be liable, right?
Even though they disobeyed your medical provider's direct instructions, the provider must be to blame, because they supposedly have the cash.
Really? Wikipedia defines it as "a global system of interconnected computer networks that use the standard Internet Protocol Suite (TCP/IP) to serve billions of users worldwide."
Wikipedia is probably not a very reliable source for that kind of topic. The sentence is descriptive of Internet, but not a complete definition for Internet.
Also, it's just a fact that there are billions of users interconnected to Internet.
"Serving billions of users worldwide" is not the definition of internet.
It could be used by fewer than 100,000 people, and still be called Internet, the number of users on a network doesn't effect whether it's called Internet or not.
But for one thing... if your government implements 'technological controls over what you can say', then your systems are no longer properly interconnected. Interconnected implies connectivity, which is exactly what Venezuela plans to destroy by preventing some messages from being exchanged between hosts, even though both end hosts intend for the message to be exchanged.
The Internet is a global open communications inter-network.
The Internet could switch from TCP/IP to some other protocol, and it would still be called Internet.
The issue is the removal of the end-to-end complete interconnectivity of hosts.
He can stifle the press as much as he wants.
He can make things people say on the internet as illegal to say,
and throw people in jail as much as he wants.
What the internet community should not tolerate is placing
physical controls to determine technologically what can and cannot be said.
Basically, any country that tries to impose technical controls of what can be said on internet, needs to be blacklisted by all other communities on the internet,
and disconnected.
Otherwise, they will never learn, and their tyrannical philosophies that are perfectly contrary to the openness of internet, are in danger of surviving and being adopted by others.
So on this matter, the community needs to be prepared to act swiftly and decisevly to counter Chavez' war on internet openness.
'The Internet cannot be something open where anything is said and done.
That's the definition of the internet.
An open communications network where anything can be said and done.
If someone implements controls, then by definition it's no longer the internet,
and we no longer really have any reason to connect our networks to Venezuela, if they start putting controls in place.
In other words, I say blacklist them, cut them off, blacklist them, if they start trying to make something the very opposite of the internet.
They are damage, cut them off, and route around them, before they manage to do any more serious damage, such as spreading their philosophy to other countries.
Um "auto-increment offsets" are a kludgy workaround for the fact MySQL doesn't support multi-master replication.
You can't perform for example, "UPDATE xytable SET x = 5 WHERE y = 2 And category=6"
on a server and get a usable result.
You cannot guarantee no additional (conflicting) row will be inserted into 'xytable' during your update.
MySQL "replication" is not synchronous, and therefore, you cannot determine insertion order, or even transaction order.
Due to the fact that your "SELECT FOR UPDATE" or "UPDATE" statement only locks the table for writes on the local mysqld.
There is no such thing as remote locking in MySQL.
Don't be fooled by the simpleness of that example though.
Integrity of the data and database transactions are a very serious issue.
MySQL has no multi-master replication mechanism that is suitable for use in an OLTP system, sorry.
A good DBA knows to back up his database via replication, not daily dumps.
No they don't. Replication is a load distribution technology. For large database, it is not a data backup strategy.
With replication, if your DB encounters a software error, or the hacker issues 'DELETE ALL'... there goes your web site.... You need point in time recovery. Your replication setup will faithfully replicate just the error that causes you to need a restore.
And there you go (in MySQL), right back to restoring a dumpfile, only now you have two or more servers to fix instead of one, problem magnification, oops!
MySQL does not support true replication or synchronous mirroring. Only asynchronous transaction duplication, by the way.
And MySQL's "replication" feature has some serious problems with it (granted PostgreSQL doesn't even include a replication feature in the current version, you pick between a few external add-ons or commercial solutions, some of which are fully synchronous).
Oh yeah, as for the problem's with MySQL's replication feature... it doesn't provide an assurance of transactional integrity.
It doesn't replicate all statements correctly
There can be errors in the replication.
An error in your application could cause it to make changes on the backup/replica server instead.
In this case, the replica and the source dataset won't match -- MySQL replication does not implement a multi-master functionality, all modifications need to be done on the source ("master") which propagate to the slave,
but MySQL doesn't help in any way for you to ensure a mistake doesn't foobar your replica.
Bringing replication into the mix just underscores more ways in which MySQL is lacking (in terms of data integrity), and could use some serious improvements.
MySQL brings in these binary log things, it's actually a mess since it keeps multiple sets of binary logs (master log VS relay log), but lest I digres...
The second statement follows from a different logical proposition.
The proposition is: all software has bugs, you just have to look hard enough.
And as far as we know, it is a true proposition, and the 2nd conclusion above follows from it.
What's not certain is that the acceleration issues come from (those) bugs, rather than a defect.
If the code burnt to the ROM/PROMs don't exactly match the code they developed and tested, there could still be a "software" issue in the final product.
An issue that can only be detected by analyzing and attempting to verify/validate actual units known to have the issue.
Based only on testing and design experience, this is not sound logic.
Agreed software commonly has bugs not exposed to testing, but the same is true of hardware manufacture, and firmware/ROM code too.
But also, there may also be an unanticipated interaction that would not be considered a software bug in and of itself, but a combination of two unanticipated inputs creates a serious problem.
The software does not exist in a vacuum, and might be susceptible from effects from hardware, especially as the unit ages -- such as corrupt memory area, multi-bit errors, memory flip errors due to gamma radiation, overheating, and all the same issues that plague desktop computers.
Could be a characteristic of the design of some portion of the software causes the system to handle an error condition ungracefully, that should be trapped, that safety hardware should be added to detect/deal with, or that software changes should occur (even if it's not strictly a software bug, sometimes you need to change the software to be more robust in the face of a hardware problem).
For example, some chip periodically has a minor (indetectable) manufacturing chip defect, or some minor variation.
Or it could be as simple as the code that gets burned to some chip isn't the same as it's supposed to be 100% of the time.
For instance, if there was a revision to one component, but the old code was still getting burnt to new units... things could still be being produced with an old fixed bug, maybe.
Or combination of components at different revision levels than what was tested.
Moreover, an unexpected input could have been found in the field to result in physical lockup of some microcontroller.
You can't test everything 100% it's arrogant and flat out wrong to claim you can.
You don't get it, apparently.
When I say you lost 6 hours of transactions, that means you restored to your most recent backup (6 hours or so old)
When you are running an E-Commerce site,
that could be thousands of dollars in lost orders.
Even worse... someone cancelled their account, but your database went 6 hours back in time, before they clicked 'Cancel account', so not only did the cancel not get through, but they thought it did, and they'll be pissed off when you bill next month.
There are a lot of applications where you can't afford to lose minutes of data, let-alone hours.
MySQL doesn't provide the level of data integrity of other DBMs.
Heck, it doesn't even really support an automated recovery procedure.
Its limited (and not very good), and not really usable on a large scale, point in time recovery capabilities are basically unknown to 90% of its users.
With PostgreSQL you get a lot of data integrity protections by default.
You don't have to make a decision to switch to InnoDB to get basic integrity protections and basic transactional support.
In MySQL you have to.
And InnoDB performance sucks, and generally can't be used for large DBs, they are almost always MyISAM.
The judge needs to nail SCO to the wall for abuse of process.
They're also trying to use the company's lawsuits as assets and to sell them to various SCO insiders so that the legal wranglings can continue even if nothing is left of SCO.
Two words: fraudulent transfer.
An attempt to evade justice, or the ruling of the courts, by fraudulently transferring property after the case was initiated against them.
This is just like trying to transfer your bank account balance and your house to another family member, after someone sued you for damages you caused them, or for not paying a huge debt.
If the transfer occured after the case started, the courts can generally recognize it as fraudulent, unwind the transfer, and penalize you.
It would definitely apply if you as an individual attempted to evade judgement in such a situation.
The courts need to stop treating corporations as above the law, and apply the same rules.
I think they have a more actually effective method though........ malware activates if determined to be unlicensed and being used "illegaly", turns into a trojan working on behalf of the the maker.
Not necessarily. It depends on the issue date of the patent.
In particular, if issuance of the patent was delayed, it may be issued for longer than 20 years.
For instance, if the patent was filed for in 1994 and issued in 2000, it could be 7 more years before it will expire, PLUS a number of additional years, according to the issuance delay.
Also, some special types of patents (as in gov' patents) can get terms much longer than 20 years.
Assuming is not good here, unless you have looked up the patents in question, you cannot really be sure they would have expired by now.
No DBM will save the day if you aren't very careful about how you design your usage of very large databases.
In the case of SQL, that would be things like Schema, choice of Index columns, views, stored procs, joins, SQL statement contents.
In some cases, the performance of a SQL statement can be horrible, but can be rewritten in a different way to answer the same question but have stellar performance.
Slashdot Mirror
"Big Banking Firms Barclay's Capital, Morgan Stanley and Merrill Lynch successfully obtained an injunction against theflyonthewall.com
and violated the "hot news" doctrine by "misappropriation" of the fact that "buy" or "sell" opinions had been expressed.
What is the "misappropriation" of the fact that an opinion has been expressed by someone?
Isn't that a form of free speech?
And sue shoecorp, the manufacturer of the shoes they were wearing, because they didn't implement Other Corp's Patented shoeprotect hammer-drop protection technology in the shoe, to protect against heavy objects being dropped on the foot.
For bonus points, buy an Other Corp protected shoe, and drop something on the foot so heavy that the protection fails, then sue Othercorp.
???
Profit
Indeed. the whole purpose of a patent is to protect novel inventions from being used by people other than the inventor.
The consumer knew or was responsible to know that "Saw Stop" was a patented technology, and any device not advertising the technology implicitly does not have a similar safety feature.
A patent holder has a right to deny use altogether, or charge exhorbitatnt fees to exercise their patent. So the person suing the company suggests, if the patent holder only sold or licensed the technology to one or two companies, only those companies can manufacture and market saws, and also, only their models the tech. is incorporated in can be sold (despite robust consumer demands for older models or refurbished units that don't have the tech) ?
Because the technology is unique and novel.
Novel and unique inventions are by definition not standard and probably not contained in the competitor's products.
If the government wants to mandate a type of safety measure, that's one thing (but they will probably also have to buy out the patent holder, or provide an option everyone can implement without imposed royalties).
In this case, there was no law mandating the safety feature. There is no standard of safety that is violated by not having the feature.
Most medical malpractice suits are infact for medical errors and they don't even represent all of them.
Who said anything about malpractice suits?
You are obviously misinformed. I'm going to stop here, since you made it abundantly obvious that you have no fucking clue.
If Health care reform gets passed it will lead to thermonuclear war. :)
I leave figuring out how an exercise to the reader.
In many cases they are frivolous.
The typical lawsuit against a healthcare provider is not suing over any obvious error, it's about suing over an outcome, and claiming healthcare provider should have done some thing in addition to or other than what they did (even if that thing is not common, or standard process).
For example "They should have tested for extremely rare disease PQRS". They didn't, the patient was diagnosed with something less severe, and they died, now their family sues.
So... they get these huge legal expenses. But more importantly... now suddenly they will have to start testing all people for PQRS, even though it's a genetic disease that effects 1 out of 10 million. Every single person will have to be tested, to avoid future lawsuits. So the lawsuit itself increases costs, but so do lawsuit-created modifications to medical practice: providers now have to waste money and compel tests and practices that the average person would not want to pay for, just to avoid the possibility of lawsuits.
A lot of expenses have to be imposed in 'practice modifications', and paperwork.
Your typical example is also people suing over something that went poorly for them, even though normal medical procedures were followed.
For example, patient died of X or suffered Y, when they were being treated for Z.
Even if X was not a reasonably avoidable outcome, or X/Y would not be prevented by generally accepted medical practice, there is a chance that a lawsuit will still be generated.
People sue when bad things happen to them. Of course all bad things are always someone else's fault. And the medical provider must be in full control at all times of everything that happens right?
If you leave with condition Y, or (for one reason or another) they can't or fail to cure it, it must be their fault. If they induce condition Y, but it's not any error on their part, or problem with their facilities, then their practitioners really must still be liable, right?
Even though they disobeyed your medical provider's direct instructions, the provider must be to blame, because they supposedly have the cash.
In this case though, Fox has grossly understated or misunderstood the case. It's going to be 100x as bad as Fox says.
Translation: you had sex one night as a teen, when you were so drunk you don't remember anything that happened?
He didn't say every teenager remembers having sex when they had it :)
Sounds like a lab tech who should be fired, and humiliated.
(Although the testee see that day, from behind bars, since they matched....)
Really? Wikipedia defines it as "a global system of interconnected computer networks that use the standard Internet Protocol Suite (TCP/IP) to serve billions of users worldwide."
Wikipedia is probably not a very reliable source for that kind of topic. The sentence is descriptive of Internet, but not a complete definition for Internet.
Also, it's just a fact that there are billions of users interconnected to Internet. "Serving billions of users worldwide" is not the definition of internet. It could be used by fewer than 100,000 people, and still be called Internet, the number of users on a network doesn't effect whether it's called Internet or not.
But for one thing... if your government implements 'technological controls over what you can say', then your systems are no longer properly interconnected. Interconnected implies connectivity, which is exactly what Venezuela plans to destroy by preventing some messages from being exchanged between hosts, even though both end hosts intend for the message to be exchanged.
The Internet is a global open communications inter-network.
The Internet could switch from TCP/IP to some other protocol, and it would still be called Internet.
The issue is the removal of the end-to-end complete interconnectivity of hosts.
He can stifle the press as much as he wants. He can make things people say on the internet as illegal to say, and throw people in jail as much as he wants.
What the internet community should not tolerate is placing physical controls to determine technologically what can and cannot be said.
Basically, any country that tries to impose technical controls of what can be said on internet, needs to be blacklisted by all other communities on the internet, and disconnected.
Otherwise, they will never learn, and their tyrannical philosophies that are perfectly contrary to the openness of internet, are in danger of surviving and being adopted by others.
So on this matter, the community needs to be prepared to act swiftly and decisevly to counter Chavez' war on internet openness.
'The Internet cannot be something open where anything is said and done.
That's the definition of the internet. An open communications network where anything can be said and done.
If someone implements controls, then by definition it's no longer the internet, and we no longer really have any reason to connect our networks to Venezuela, if they start putting controls in place.
In other words, I say blacklist them, cut them off, blacklist them, if they start trying to make something the very opposite of the internet.
They are damage, cut them off, and route around them, before they manage to do any more serious damage, such as spreading their philosophy to other countries.
Ok, well, CLAP is free sort of, seems a decent choice.
It's an abbreviation for cleft lip and palate, a physical deformity, but I doubt that there could be any confusion.
No more so than there could be confusion between LAMP and Lamps you plug in to produce light with.
Um "auto-increment offsets" are a kludgy workaround for the fact MySQL doesn't support multi-master replication.
You can't perform for example, "UPDATE xytable SET x = 5 WHERE y = 2 And category=6" on a server and get a usable result. You cannot guarantee no additional (conflicting) row will be inserted into 'xytable' during your update.
MySQL "replication" is not synchronous, and therefore, you cannot determine insertion order, or even transaction order.
Due to the fact that your "SELECT FOR UPDATE" or "UPDATE" statement only locks the table for writes on the local mysqld.
There is no such thing as remote locking in MySQL.
Don't be fooled by the simpleness of that example though. Integrity of the data and database transactions are a very serious issue. MySQL has no multi-master replication mechanism that is suitable for use in an OLTP system, sorry.
A good DBA knows to back up his database via replication, not daily dumps.
No they don't. Replication is a load distribution technology. For large database, it is not a data backup strategy.
With replication, if your DB encounters a software error, or the hacker issues 'DELETE ALL'... there goes your web site.... You need point in time recovery. Your replication setup will faithfully replicate just the error that causes you to need a restore.
And there you go (in MySQL), right back to restoring a dumpfile, only now you have two or more servers to fix instead of one, problem magnification, oops!
MySQL does not support true replication or synchronous mirroring. Only asynchronous transaction duplication, by the way.
And MySQL's "replication" feature has some serious problems with it (granted PostgreSQL doesn't even include a replication feature in the current version, you pick between a few external add-ons or commercial solutions, some of which are fully synchronous).
Oh yeah, as for the problem's with MySQL's replication feature... it doesn't provide an assurance of transactional integrity. It doesn't replicate all statements correctly
There can be errors in the replication.
An error in your application could cause it to make changes on the backup/replica server instead. In this case, the replica and the source dataset won't match -- MySQL replication does not implement a multi-master functionality, all modifications need to be done on the source ("master") which propagate to the slave, but MySQL doesn't help in any way for you to ensure a mistake doesn't foobar your replica.
Bringing replication into the mix just underscores more ways in which MySQL is lacking (in terms of data integrity), and could use some serious improvements.
MySQL brings in these binary log things, it's actually a mess since it keeps multiple sets of binary logs (master log VS relay log), but lest I digres...
The second statement follows from a different logical proposition.
The proposition is: all software has bugs, you just have to look hard enough.
And as far as we know, it is a true proposition, and the 2nd conclusion above follows from it.
What's not certain is that the acceleration issues come from (those) bugs, rather than a defect.
If the code burnt to the ROM/PROMs don't exactly match the code they developed and tested, there could still be a "software" issue in the final product.
An issue that can only be detected by analyzing and attempting to verify/validate actual units known to have the issue.
Wow... I'll make it a point to never turn on my left turn signal.
Thanks!
Based only on testing and design experience, this is not sound logic.
Agreed software commonly has bugs not exposed to testing, but the same is true of hardware manufacture, and firmware/ROM code too.
But also, there may also be an unanticipated interaction that would not be considered a software bug in and of itself, but a combination of two unanticipated inputs creates a serious problem.
The software does not exist in a vacuum, and might be susceptible from effects from hardware, especially as the unit ages -- such as corrupt memory area, multi-bit errors, memory flip errors due to gamma radiation, overheating, and all the same issues that plague desktop computers.
Could be a characteristic of the design of some portion of the software causes the system to handle an error condition ungracefully, that should be trapped, that safety hardware should be added to detect/deal with, or that software changes should occur (even if it's not strictly a software bug, sometimes you need to change the software to be more robust in the face of a hardware problem).
For example, some chip periodically has a minor (indetectable) manufacturing chip defect, or some minor variation.
Or it could be as simple as the code that gets burned to some chip isn't the same as it's supposed to be 100% of the time.
For instance, if there was a revision to one component, but the old code was still getting burnt to new units... things could still be being produced with an old fixed bug, maybe.
Or combination of components at different revision levels than what was tested.
Moreover, an unexpected input could have been found in the field to result in physical lockup of some microcontroller.
You can't test everything 100% it's arrogant and flat out wrong to claim you can.
You don't get it, apparently. When I say you lost 6 hours of transactions, that means you restored to your most recent backup (6 hours or so old)
When you are running an E-Commerce site, that could be thousands of dollars in lost orders.
Even worse... someone cancelled their account, but your database went 6 hours back in time, before they clicked 'Cancel account', so not only did the cancel not get through, but they thought it did, and they'll be pissed off when you bill next month.
There are a lot of applications where you can't afford to lose minutes of data, let-alone hours.
MySQL doesn't provide the level of data integrity of other DBMs.
Heck, it doesn't even really support an automated recovery procedure.
Its limited (and not very good), and not really usable on a large scale, point in time recovery capabilities are basically unknown to 90% of its users.
With PostgreSQL you get a lot of data integrity protections by default.
You don't have to make a decision to switch to InnoDB to get basic integrity protections and basic transactional support.
In MySQL you have to. And InnoDB performance sucks, and generally can't be used for large DBs, they are almost always MyISAM.
The judge needs to nail SCO to the wall for abuse of process.
They're also trying to use the company's lawsuits as assets and to sell them to various SCO insiders so that the legal wranglings can continue even if nothing is left of SCO.
Two words: fraudulent transfer.
An attempt to evade justice, or the ruling of the courts, by fraudulently transferring property after the case was initiated against them.
This is just like trying to transfer your bank account balance and your house to another family member, after someone sued you for damages you caused them, or for not paying a huge debt.
If the transfer occured after the case started, the courts can generally recognize it as fraudulent, unwind the transfer, and penalize you.
It would definitely apply if you as an individual attempted to evade judgement in such a situation.
The courts need to stop treating corporations as above the law, and apply the same rules.
I think they have a more actually effective method though........ malware activates if determined to be unlicensed and being used "illegaly", turns into a trojan working on behalf of the the maker.
Not necessarily. It depends on the issue date of the patent.
In particular, if issuance of the patent was delayed, it may be issued for longer than 20 years.
For instance, if the patent was filed for in 1994 and issued in 2000, it could be 7 more years before it will expire, PLUS a number of additional years, according to the issuance delay.
Also, some special types of patents (as in gov' patents) can get terms much longer than 20 years.
Assuming is not good here, unless you have looked up the patents in question, you cannot really be sure they would have expired by now.
No DBM will save the day if you aren't very careful about how you design your usage of very large databases.
In the case of SQL, that would be things like Schema, choice of Index columns, views, stored procs, joins, SQL statement contents.
In some cases, the performance of a SQL statement can be horrible, but can be rewritten in a different way to answer the same question but have stellar performance.
Sorry, LACP is already taken. You Linux folken can't have it, it belongs to the network itself, (IEEE 802.3ad) :)
LAMP was meant to mean Linux Apache Mysql and Perl though.
A bad policy when dealing with your data.
Once it's broke, it is way too late.
You can't un-LOSE the past 6 hours of transactions or table referential integrity that MySQL trashed, due to an unclean shutdown.
MySQL's great until it comes up to bite you in the arse.