The credit card histories were selected from a pool of people that allowed the research firm to be copied on all credit card transactions for a given window of time. This should have set off alarm bells for anyone reading the report. By relying on only those willing to hand over a history of their financial transactions, they've already fundamentally broken any chance at having a both a truly random selection and a selection that is representative of the population at large.
Someone cracking a list of alphanumeric passwords where it is known that there is no requirement that the users include at least one numeric digit will (or at least should) assume that most users will be to lazy to include at least one numeric digit. Since this assumption will be true in the majority of cases, they've just reduced the time that it takes to them to use either brute force or a dictionary attack in most cases. Requiring all users to at use at least numeric digit means that the hacker will always fail if this assumption is made. Requiring at least one digit/or/ punctuation symbol is even better.
The GPL v2 doesn't explicitly address operating system kernel modules. Arguments can be drawn up on either side of the argument and it would take one or more court cases to figure out which interpretation was the correct one. The ambiguity is heightened because the intentions of the party that drew up the license (the FSF) are not necessarily the intentions of the creator of the product (the Linux kernel developers). Because of this situation, Linus has said (more than once) that if binary only kernel modules do violate the GPL, he is granting an exception for them.
His more recent statements simply continue this long-standing policy. It also gives more credence to the notion that the Linux kernel will probably not be moving to the GPL v3 when it is finally released. Linux is somewhat unique in the Free Software World in that it explicitly identifies itself as being released under v2 of the GPL rather than v2 or later.
You start with an ad hominem attack, then appeal to authority (your own), then offer your opinion as a factual assertion.
Further, your post reveals an astonishing arrogance combined with a complete lack of comprehension. I'm just not certain whether it was a complete lack of comprehension of my earlier post or a complete lack of comprehension of what a database actually is. My guess is that it's the former.
A database is no more (and no less) than a container for the storage of data. MySQL is no less of a ``real'' database than latest and greatest from Oracle and IBM. The variation lies only in the features needed for a given application. You're basically arguing that supporting certain types of enterprise class features makes databases from the big vendors more of a database than other databases. That simply isn't true. That's like saying that a Cray supercomputer is more of a computer than my venerable old Palm Pilot. The truth is that both are equally computers but that they were engineered to meet entirely different sets of needs. They vary in complexity and capacity but they don't vary in their computerness. The same is true of MySQL and Oracle.
I think it fair to say that Debian is the second most popular Linux distribution for web servers surveyed by Netcraft. It's also probably fair to extrapolate this to mean that Debian is the second most popular Linux distribution among all web servers. But I think it it quite the unwarranted stretch to take that to mean that Debian is the second most popular Linux in any market segment. I have no reason to believe that use of Linux in the business desktop, development database, database server, or application server follow the same trends. They may very well follow the same trends, but until someone studies the question and offers a cogent analysis that question has to remain unanswered.
It may not be an ACID compliant transactional DBMS. It certainly isn't a fully relational DBMS. But the definition of a database is pretty much a system for storing data which MySQL certainly qualifies as. And if we're going to start being pedantic, we also have to consider that none of the mainstream enterprise databases are truly relational. While they may have many relational features, none can consistently enforce proper relational behavior. In fact, any database that fully supports SQL cannot be a fully relational database. The only difference between Oracle, SQL Server, DB/2 and MySQL is one of extent, not of kind.
If a company holds that it can make a 50% ROI on one product line and a 25% ROI on onther product line (and all other things being held equal) they will put their resources into the line with the 50% ROI until such time as the law of diminishing returns brings the marginal ROI for additional resources being added to under 25%.
For example, when I was a kid a local pizza delivery chain started delivering breakfast pizzas. They made money hand over fist. But after a few months, the calculated that the additional cost of maintaining a third shift of workers and an expanded breakfast menu would bring in more money if put into opening additiona stores serving the traditional lunch, dinner, late night crowd with the normal pizzaria menu.
Most likely what is happening is that the MySQL corporation finds that if it spends the same number of dollars training a support tech, those dollars bring in more money if the tech is dedicated to Redhat and/or SuSE than if the tech is also trained on Debian. This doesn't mean that there is no market for Debian support. It means only that MySQL has a higher relative profit from supporting just two databases. The calculation may be different for another company that has a different resource pool. For example a company that already supports Debian Linux, may have a very low marginal cost for adding MySQL on Debian support and, consequently, have a far higher ROI for supporting MySQL on Debian.
It's more fair to say that if FF developers drop all further support of the 1.x and 2.x lines and never add Win9x support to the FF 3.x tree, then they'll be no better than Microsoft developers. This isn't a case of the developers removing existing support for all future releases. This is a case of developers not having the resources to surmount a technical obstacle for the next generation of a product. And, unlike Microsoft's IE group, I don't see any indication that the FF project is going to stop maintenance of the 1.x and 2.x lines any time soon
Google has a very, very complex interface. You have to put a `+' character in front of common words such as `how' or `when' or else Google will filter them out of your search. You can use logical operators such as `OR.' You can include numeric ranges.
But the magic is that Google can put all of this in an interface so that people who don't know (and who largely don't care) that the features are there don't have to deal with them.
Where MSN and Yahoo trump Google isn't in complexity. It's in visual clutter.
When Apple announced Safari, Microsoft immedieately abandoned the Mac version of iE.
Once Apple previewed Keynote, the writing was on the wall for Office for OS X. When Pages debuted, that only cemented the decision. The only thing keeping new releases on the drawing board is the settlement between Apple and Microsoft that guaranteed development of Office for the Mac for a certain number of years. Once the clock runs out on that agreement, there will be no no Office for Macs.
As far as databases go, don't forget that it was Apple that produced Filemaker Pro. Look for a re-acquisition of Claris sometime in the next couple of years.
What will kill Apple, though, is the demise of Entourage. Once Microsoft is no longer providing an Exchange client, there will be a powerful incentive to not go Mac. Hopefully Apple will step up to the plate.
I work for a Microsoft shop but at home I only run OS X. I find Entourage to be far, far better Exchange client than Outlook 2003 which we've standardized on at work. To illustrate just one example, Entourage can connect to multiple Exchange servers where Outlook can only connect to one. Is that a brain-dead design decision or what? I can't even begin to conjecture what Microsoft had in mind when deciding that Outlook users would only ever desire to connect to a single Exchange server.
On a twenty server cluster of SPARC kit with tens of processors each, MySQL can't touch Oracle for distributive processing and load balancing. Second, the point of dropping constraints for a few months was to allow the the database design to reach maturity in a development environment. Before the project reached production, constraints were turned back on.
Itanic was certainly marketed as a next generation chip. But rather than argue whether it was or not, I'll just point out that aside from the example you yourself listed there are many others as listed in my subject line. Those are the ones I can think off the top of my head.
With Oracle and SQL Server you can turn off constraints for an indefinite period of time. It may be used mostly for single bulk transactions but I've seen it used for months at a time on Oracle.
The history of the general purpose CPU marketplace has shown time and again that sometimes next generation processors aren't all they're made out to be. Itanic, anyone?
Nothing Intel does with regards to marketing is different in kind than AMD's approach. They may differ in the extent to which they push a specific strategy, but they both have the same sorts of plans on the table.
I thought it did. I even argued that it did. But someone challenged me to go back and look at the discussions over Sun's patent deal with Microsoft and licensing of SCOSource. At least from PJ's corner, there really is no change in tone from then to now. There is a change in the volume of the comments that follow her lead. More people defended Sun at Groklaw in 2003 and 2004 than are currently defending Novell over yonder.
Personally I think it's all a tempest in a tea-pot. Ten years ago something like this may have mattered. Nowadays there are to many large companies with vested interests in Linux, Open Office dot Org, and other free software projects. Many of these companies have large software patent portfolios and some are willing to go to launch an a retalitory strike should Microsoft start suing Linux users or distributors over patent infringement.
Furthurmore, the solution is simple. People who are pissed at this deal should spend less time boycotting Novell and more time badgering their local legislators to invalidate all software patents. If I am wrong and the MS/Novell deal actually does matter, it will come to nought if software patents become unenforceable.
With the Athlon series, AMD went to CPU ratings because, clock for clock, an AMD part was significnatly faster than an Intel part at the same clock speed. They had a hard time getting people to reject the ``megahertz myth'' that a faster clock was always a fast chip. And, in fact, this very myth came back to bite Intel in the ass and is part of the reason (but certainly not the only reason) that Merced died a much deserved death.
But other than that, I pretty much agree. As has been said for many eyars, brand loyalty is a socially acceptable version of battered spouse syndrome. Products should always be objectively evaluated for suitability to a task. Sometimes one chip is better for a certain task. Sometimes another chip is better. Sometimes the CPU doesn't really matter at all.
HT also shines as memory sizes grow larger. One of the problems with the current generation of workstations is that few have enough (and fast enough) memory and memory bandwidth to keep a single core, let alone several, at 100% utilization all of the time. A dual core duo system with four cores, however, would benefit immensely with something like HT and multiple banks of memory.
First, pretty much every modern dictionary (Merriam-Webster, American Heritage, Random House, etc.) lists the word as ID. I'd check the OED too but my library card is expired and I can't log into the OED online anymore.
Second, an acronym is ``a word formed from the initial letters or groups of letters of words in a set phrase or series of words''. Consider RADAR. RAdio Detecting And Ranging. If ID were a contraction of identifier it would be spelled id' not ID.
Lastly, whether acronyms are all upper case or not is entirely a matter of convention. ID is typically always upper case to avoid confusion with the Freudian term id.
OS/2 v3 came in two flavors. The full pack had a blue spine and contained the Win-OS2 runtime than ran 16 bit Windows applications somewhat seemlessly. The version with a red spine called ``OS/2 For Windows'' (truly a most horrible name) came without the 16 bin Win-OS2 runtime and could not run Windows applications unless installed on a computer that already had Windows 3.1 or 3.1.1 installed. The salient point, however, was that because IBM was either shipping the runtime with OS/2 or marketing the runtime-less version as an add on to Windows, all development shops could depend on OS2 being able to run 16 bit Windows apps.
This strategy was given up in later versions. Warp Connect and OS/2 v4 both shipped only in the full pack flavor. But by this time, Windows 95 was also out and most people were only interested in 32 bit Windows applications which wouldn't run on any flavor of OS/2.
In either case, the problem with attracting developers was most likely much larger a function of the lack of click and drool development tools. IBM's Visual Age ran like a cow compared to Microsoft's Visual Studio and I don't think any other vendor was really in the visual space at the time. (This was the bad old days of Borland's 5.x compiler that sucked canal water for building GUI apps.) Then the nail in the coffin (developer-wise) were the changes to the OS/2 v4 APIs where some API calls that were somewhat common in v2 and v3 would either trigger kill the synchronos input queue (no more keyboard or mouse) or even trigger a seg fault in the kernel. IMO, IBM ought to have shipped the EMX version of GCC with every version of OS/2. If they had done that and supported XFree86 for OS/2, they might have had a chance. On the other hand, though, disk space wasn't nearly as cheap back then. But if they had done that, OS/2 would have gotten the attention of quite a few *nix programmers.
And blamed Christianity in all its forms for the not only the devestation of western Europe (the thirty years war) but also the dark ages of western europe. See Diderot's claim that humanity would not be free until the last king was strangled with the entrails of the last priest.
Constantinople, where the capital of the empire had been moved in the 4th century, didn't fall until the fifteenth century. By that point, the Byzantine empire had been reduced to pretty much a single city-state. But that's just the final end point. Most scholars of the present day generally consider the golden age of the Byzantines to be the ninth through eleventh century CE.
The tendency to disparage the Roman Empire of the East is mostly an artifact of the early modern historians that wanted to demonize everything that happened between the glory days of Rome and Greece and the modern era. This tendency culminates in Gibbon's Decline and Fall in which he has almost nothing good to say about the Byzantines.
Slashdot Mirror
The credit card histories were selected from a pool of people that allowed the research firm to be copied on all credit card transactions for a given window of time. This should have set off alarm bells for anyone reading the report. By relying on only those willing to hand over a history of their financial transactions, they've already fundamentally broken any chance at having a both a truly random selection and a selection that is representative of the population at large.
Someone cracking a list of alphanumeric passwords where it is known that there is no requirement that the users include at least one numeric digit will (or at least should) assume that most users will be to lazy to include at least one numeric digit. Since this assumption will be true in the majority of cases, they've just reduced the time that it takes to them to use either brute force or a dictionary attack in most cases. Requiring all users to at use at least numeric digit means that the hacker will always fail if this assumption is made. Requiring at least one digit /or/ punctuation symbol is even better.
His more recent statements simply continue this long-standing policy. It also gives more credence to the notion that the Linux kernel will probably not be moving to the GPL v3 when it is finally released. Linux is somewhat unique in the Free Software World in that it explicitly identifies itself as being released under v2 of the GPL rather than v2 or later.
Further, your post reveals an astonishing arrogance combined with a complete lack of comprehension. I'm just not certain whether it was a complete lack of comprehension of my earlier post or a complete lack of comprehension of what a database actually is. My guess is that it's the former.
A database is no more (and no less) than a container for the storage of data. MySQL is no less of a ``real'' database than latest and greatest from Oracle and IBM. The variation lies only in the features needed for a given application. You're basically arguing that supporting certain types of enterprise class features makes databases from the big vendors more of a database than other databases. That simply isn't true. That's like saying that a Cray supercomputer is more of a computer than my venerable old Palm Pilot. The truth is that both are equally computers but that they were engineered to meet entirely different sets of needs. They vary in complexity and capacity but they don't vary in their computerness. The same is true of MySQL and Oracle.
I think it fair to say that Debian is the second most popular Linux distribution for web servers surveyed by Netcraft. It's also probably fair to extrapolate this to mean that Debian is the second most popular Linux distribution among all web servers. But I think it it quite the unwarranted stretch to take that to mean that Debian is the second most popular Linux in any market segment. I have no reason to believe that use of Linux in the business desktop, development database, database server, or application server follow the same trends. They may very well follow the same trends, but until someone studies the question and offers a cogent analysis that question has to remain unanswered.
It may not be an ACID compliant transactional DBMS. It certainly isn't a fully relational DBMS. But the definition of a database is pretty much a system for storing data which MySQL certainly qualifies as. And if we're going to start being pedantic, we also have to consider that none of the mainstream enterprise databases are truly relational. While they may have many relational features, none can consistently enforce proper relational behavior. In fact, any database that fully supports SQL cannot be a fully relational database. The only difference between Oracle, SQL Server, DB/2 and MySQL is one of extent, not of kind.
For example, when I was a kid a local pizza delivery chain started delivering breakfast pizzas. They made money hand over fist. But after a few months, the calculated that the additional cost of maintaining a third shift of workers and an expanded breakfast menu would bring in more money if put into opening additiona stores serving the traditional lunch, dinner, late night crowd with the normal pizzaria menu.
Most likely what is happening is that the MySQL corporation finds that if it spends the same number of dollars training a support tech, those dollars bring in more money if the tech is dedicated to Redhat and/or SuSE than if the tech is also trained on Debian. This doesn't mean that there is no market for Debian support. It means only that MySQL has a higher relative profit from supporting just two databases. The calculation may be different for another company that has a different resource pool. For example a company that already supports Debian Linux, may have a very low marginal cost for adding MySQL on Debian support and, consequently, have a far higher ROI for supporting MySQL on Debian.
It's more fair to say that if FF developers drop all further support of the 1.x and 2.x lines and never add Win9x support to the FF 3.x tree, then they'll be no better than Microsoft developers. This isn't a case of the developers removing existing support for all future releases. This is a case of developers not having the resources to surmount a technical obstacle for the next generation of a product. And, unlike Microsoft's IE group, I don't see any indication that the FF project is going to stop maintenance of the 1.x and 2.x lines any time soon
But the magic is that Google can put all of this in an interface so that people who don't know (and who largely don't care) that the features are there don't have to deal with them.
Where MSN and Yahoo trump Google isn't in complexity. It's in visual clutter.
Once Apple previewed Keynote, the writing was on the wall for Office for OS X. When Pages debuted, that only cemented the decision. The only thing keeping new releases on the drawing board is the settlement between Apple and Microsoft that guaranteed development of Office for the Mac for a certain number of years. Once the clock runs out on that agreement, there will be no no Office for Macs.
As far as databases go, don't forget that it was Apple that produced Filemaker Pro. Look for a re-acquisition of Claris sometime in the next couple of years.
What will kill Apple, though, is the demise of Entourage. Once Microsoft is no longer providing an Exchange client, there will be a powerful incentive to not go Mac. Hopefully Apple will step up to the plate.
I work for a Microsoft shop but at home I only run OS X. I find Entourage to be far, far better Exchange client than Outlook 2003 which we've standardized on at work. To illustrate just one example, Entourage can connect to multiple Exchange servers where Outlook can only connect to one. Is that a brain-dead design decision or what? I can't even begin to conjecture what Microsoft had in mind when deciding that Outlook users would only ever desire to connect to a single Exchange server.
On a twenty server cluster of SPARC kit with tens of processors each, MySQL can't touch Oracle for distributive processing and load balancing. Second, the point of dropping constraints for a few months was to allow the the database design to reach maturity in a development environment. Before the project reached production, constraints were turned back on.
Itanic was certainly marketed as a next generation chip. But rather than argue whether it was or not, I'll just point out that aside from the example you yourself listed there are many others as listed in my subject line. Those are the ones I can think off the top of my head.
With Oracle and SQL Server you can turn off constraints for an indefinite period of time. It may be used mostly for single bulk transactions but I've seen it used for months at a time on Oracle.
The history of the general purpose CPU marketplace has shown time and again that sometimes next generation processors aren't all they're made out to be. Itanic, anyone?
Nothing Intel does with regards to marketing is different in kind than AMD's approach. They may differ in the extent to which they push a specific strategy, but they both have the same sorts of plans on the table.
Personally I think it's all a tempest in a tea-pot. Ten years ago something like this may have mattered. Nowadays there are to many large companies with vested interests in Linux, Open Office dot Org, and other free software projects. Many of these companies have large software patent portfolios and some are willing to go to launch an a retalitory strike should Microsoft start suing Linux users or distributors over patent infringement.
Furthurmore, the solution is simple. People who are pissed at this deal should spend less time boycotting Novell and more time badgering their local legislators to invalidate all software patents. If I am wrong and the MS/Novell deal actually does matter, it will come to nought if software patents become unenforceable.
I'd be surprised if DB/2 didn't also have it.
But other than that, I pretty much agree. As has been said for many eyars, brand loyalty is a socially acceptable version of battered spouse syndrome. Products should always be objectively evaluated for suitability to a task. Sometimes one chip is better for a certain task. Sometimes another chip is better. Sometimes the CPU doesn't really matter at all.
HT also shines as memory sizes grow larger. One of the problems with the current generation of workstations is that few have enough (and fast enough) memory and memory bandwidth to keep a single core, let alone several, at 100% utilization all of the time. A dual core duo system with four cores, however, would benefit immensely with something like HT and multiple banks of memory.
Second, an acronym is ``a word formed from the initial letters or groups of letters of words in a set phrase or series of words''. Consider RADAR. RAdio Detecting And Ranging. If ID were a contraction of identifier it would be spelled id' not ID.
Lastly, whether acronyms are all upper case or not is entirely a matter of convention. ID is typically always upper case to avoid confusion with the Freudian term id.
This strategy was given up in later versions. Warp Connect and OS/2 v4 both shipped only in the full pack flavor. But by this time, Windows 95 was also out and most people were only interested in 32 bit Windows applications which wouldn't run on any flavor of OS/2.
In either case, the problem with attracting developers was most likely much larger a function of the lack of click and drool development tools. IBM's Visual Age ran like a cow compared to Microsoft's Visual Studio and I don't think any other vendor was really in the visual space at the time. (This was the bad old days of Borland's 5.x compiler that sucked canal water for building GUI apps.) Then the nail in the coffin (developer-wise) were the changes to the OS/2 v4 APIs where some API calls that were somewhat common in v2 and v3 would either trigger kill the synchronos input queue (no more keyboard or mouse) or even trigger a seg fault in the kernel. IMO, IBM ought to have shipped the EMX version of GCC with every version of OS/2. If they had done that and supported XFree86 for OS/2, they might have had a chance. On the other hand, though, disk space wasn't nearly as cheap back then. But if they had done that, OS/2 would have gotten the attention of quite a few *nix programmers.
The Windows 9x codebase used this very strategy.
And blamed Christianity in all its forms for the not only the devestation of western Europe (the thirty years war) but also the dark ages of western europe. See Diderot's claim that humanity would not be free until the last king was strangled with the entrails of the last priest.
The tendency to disparage the Roman Empire of the East is mostly an artifact of the early modern historians that wanted to demonize everything that happened between the glory days of Rome and Greece and the modern era. This tendency culminates in Gibbon's Decline and Fall in which he has almost nothing good to say about the Byzantines.