My girlfriend's mother did...she said it did jack shit for grass and mud stains.
If this is true, why does OxyClean take up more shelf space at my local Wal-Mart than most of the other detergents? Isn't Wal-Mart the store that will only sell stuff that has something like 97% per second throughput? Would stuff sell like that if it did "jack shit"?
Why do PC Notebook components require 3 extra lbs!??
Considering the Pentium 4 in this Toshiba counter-top model draws three or four times the power of the G3 and G4 CPUs that Apple uses, they probably have an extra pound of battery plus an extra pound of heat sink in there somewhere.
Geez, at 10 lbs...airlines might need to charge for an extra ticket.
The problem with Javascript is that there are so many crappy programs out there that don't properly utilize the language, resorting to stupid 'Netscape' or 'IE' detection hacks rather than testing for the existance of functions.
Very good advice.
I wonder if there will be an "autoconf" program for JavaScript, one day...
Yup, CSS (two or three versions), Flash, JavaScript (multiple versions), XML (+123 XML-related acronyms), proprietary browser exensions, ActiveX, people still using Netscape 4.5beta, and IE-specific sites sure make the me feel all warm and fuzzy about web "innovation".
In about ten more years, after the dust settles, and all that's left is XML (+10 acronyms), perhaps, then, web developers can grow their hair back.
People who are driven by the "experience" of the web, where everything needs to look and feel like Fox News, need to be put into time-out to rethink thier plan.
Re:If anybody cares, the NT boot sector is O/S cod
on
Bill Gates On Linux
·
· Score: 1
That file is dated 1987, OS/2 1.0 joint code with Microsoft and IBM.
I don't know, but perhaps 1987 was the last time they really needed to write a 32-bit kernel loader. 1987 was probably the last time that anyone need to write a 32-bit loader (look in specific spot on disk, load a file, run it). Regardless, I'm suprised that the copyright didn't read "Regents of UC Berkeley" or something like that.
"Best Practices" is yet another ridiculous business term that is crossing over into the IT world.
It does have a little meaning. Best practices are basically what a person would do after spending some time thinking about what would be an elegant way to solve a problem.
Best practices are not achieved after one 30-minute meeting. Nor are they achievable by design-by-committee. They require people on a team to be honest, first, and political, last.
Unfortunately, this rarely occurs in the real world, especailly due to the technological religious fanaticism present among most people.
Like all ORA books, this one was typeset using FrameMaker.
Hmm..FrameMaker is the only program I remember using that rivaled TeX in the flexibility of page layout and quality of the output. Framemaker even accepts EPS for embedded graphics.
We should just be happy that O'Reilly chose a non-Microsoft application, because a Linux book written in Word, for example, would not only be hideous in appearance but leave OSS users seriously conflicted.
How many people out there use java in their web based apps that interface with a database server backend?
I would bet that nearly all the use of Java is for web-based apps that interface to a database.
What platforms do you use for development?
For Java, it would be J2SE from Sun (it includes JDBC).
What Dbs do you use?
Oracle, MySQL, PostgreSQL, DB2, etc.
What web server?
Apache, WebLogic, Sun ONE, etc.
These questions, at face value, mainly answer themselves. Perhaps a better question would be: what do Java programmers do to prevent database applications from turning into a quagmire of naive mistakes and bottlenecks making a well-intentioned application into a sick joke of engineering?
Database programming is one of those things that is simple enough, at first glance, to be attempted by every college sophomore on the planet, but complex enough, in reality, to make people who hired those college sophomores look foolish.
Neither Linux or Windows has died out yet so none of them are passing fads.
To make a biological analogy, Windows could be viewed as a lifeform already occupying 90% of its ecosystem with little room for growth and debatable ability to reuse the space behind it (i.e., Microsoft is its own largest competitive force in the market).
When an established ecosystem dies, the smaller players, e.g. Linux and others, come to fill in the area now fertilized by the decomposing life from before. The prior lifeform might still exist but only in a much smaller niche surrounded on all sides.
BG: . . . We will never have a price lower than Linux, in terms of just what you charge for the software. We compete on the basis of, if you look at the value you get out of the system and the overall cost that the system has that apply in our software. For any project, if you look at communications costs, hardware costs, personnel costs, all that, software licensing ranges -- the highest you'd ever find is, like, 3% of any IT-type project. And so the question is can that 3% [compensate], in terms of how quickly you get the system set up?
What's so interesting is that this exact same argument is used by other companies against Microsoft. Sun does it with their servers and will be doing it with that Mad Hatter desktop. I'm sure Red Hat is doing it every day. Even IBM will come in with a jab or two about their servers.
I find the bit about "personnel costs" regarding Microsoft's offerings hilarious, by the way. Bill G. himself admitted two weeks per person per year lost due to Windows 9x. How much do to Outlook viruses? How much due to IIS worms? How much due to constant manual intervention for Windows servers?
There will be no more small developers in windows, which means MS will have no one to rip fresh ideas from!
Also, it seems MS' bread-n-butter is those small MCSE outfits that sell "solutions" to local marks--er, other small businesses. How will these outfits deal with both those businesses who haven't upgraded since 1996 and those who demand the latest and greatest? TCPA means that the recently upgraded business just might not be able to talk to the 1996-vintage one. Should the 1996-vintage one be required to pay thousands of dollars in upgrades, when all they need is a spreadsheet for payroll, for example?
so, what, should we stock up on Win2000, XP, and Linux OS's along with our CD and DVD burners?
Yes. Keeping a working late-90's or early-2000's vintage computer around the house is a very good idea. The late-90's was the era of the forever "fast enough" CPU. For example, at home, I have yet to own a computer faster than 300MHz. Mozilla launches considerably faster than the modem will dial-in, so this is a good threshold of usefulness for me.
It'll be made illegal as a "circumvention device" in short order if it actually turns out to be useful for any sort of multimedia applications.
Tell that to the secondary market for UNIX workstations. There are literally dozens of companies willing to sell used Sun workstations, for example, for well under $1,000. This isn't a bad deal for a workstation that is built like a tank (excluding Ultra 5 and 10, of course) and has a decent CPU, ECC memory, and SCSI disks.
I've also had decent luck with non-Sun-branded parts in Sun workstations, such as regular Seagate SCSI drives and SCSI-2 CD-RW drives.
The only catch is that the official Solaris right-to-use license for used hardware is approx. $100, and the optional media kit is also about $100 (potential total of $200). Still not so bad for official UNIX(R). Solaris 9 also comes with tools for CD burning, viewing MPEG files, StarOffice, and, recently, Netscape 7.
If Solaris isn't your cup of tea, Linux and the *BSDs also all run on UltraSPARC, so you can still feel Free, if you like.
Objective simplicity is a laudable goal, but it conflicts with the goals of flexibility, performance, and ability to meet the requirements of the subject matter (if the subject matter is complex).
With regard to performance and flexibility, simplicity can be good, given that, sometimes, simplicity is actually more difficult to attain initially than ad-hoc complexity, which the popular frameworks tend to encourage (i.e., simplicity often requires some thought and modesty, but the fruit of that effort is forever delicious).
For example, in the long-run, it is easier to maintain a hierarchy of HTML/XHTML and CGI pages or a multi-tier JavaScript/XML/JSP/Servlet/Struts/EJB/etc/etc application? In practice, honestly, which is more likely to perform well, work predictably across browsers, and withstand the tests of time? Which method can be picked up easily by new hires or relearned quickly enough to make incremental changes month-to-month practical?
For very complex content, warranting permanent staff, it seems plausible (not speaking from experience) that DocBook could make for a very good base format, and the maturing XML tools, like XSLT, could provide a fairly direct and simple means of generating web pages as needed. Hourly Perl scripts could even keep static content up-to-date without adding much complexity and removing much of the need for real-time dynamic content. Something like DocBook also means for approximately one amount of effort, a multitude of outputs are possible (HTML and PostScript, for example). While this method might not be totally practical today, due to the immaturity of XML, it does seem promising.
Either way, it seems that simple tools can be used to make a solid website, where a one-page chart is sufficient for a person to really understand how things fit together.
Using Netscape 7.1 for Windows (I am at work, ok?) I still can't vote in a poll properly (pop-up appears, but the vote data is loaded in main window, leaving pop-up blank).
I think Netscape 7.1's Good Taste algorithms are working just fine. Good CSS support and preventing people from voting in Wolf's damned sensationalistic and myopic polls is win-win:)
Why do most software packages continue to add features without actualy providing a subjective goal to strive for?
Buzzwords are the ends and the means, in the fashion industry...er, I mean IT industry.
When will the goal of a products feature-list finally be met?
It never will. The true end of the Earth will be when Microsoft's code-base becomes so massive that it forms a black hole destroying everything within 3 light-years.
Microsoft(R)'s Windows(TM) name isn't describing the project's code name, yet the product's retail names are somthing like 95, 98, NT, 2000, XP. Well, it is honest to say that the NT and 2000 products are similar, 95 and 98 products are similar, and the XP product isn't quite similar to 95 and 98 and 2000, but then there was the fiasco that all the 9x users went out to buy product Windows(TM) 2000 and found they had been tricked by Microsoft that 2000 would be like a 95 or 98. Anyone see my point?
You wouldn't last a day working at Microsoft's marketing department. They brand Windows deceptively because it is in their interest to do so! They are professional liars and theives. You shouldn't be suprised.
Does anyone think they should continue calling those products by their initial names AFTER the programming syntax and methodology becomes completly different or non-compatible than they were first designed?
Yes, in the sense that they are now branded and well-known. For example, a new Chevy Impala bears no resemblance to the two-ton behemoth from 1971 bearing the same name. Yet, it is still familiar, and it is a car from Chevrolet, accomplishing their goals of how to market this car. It is well-publicised that Perl 6 is vastly different than Perl 5, and it is also well-known that people who assume interoperability across major versions of software are foolish. It doesn't matter that Microsoft's marketing department exerts tremendous psychological control over their customers, those customers are still foolish.
Struts is a web application development framework (WADF), of which there are many.
While factually correct, "many" might be an understatement. It seems everyone and their uncle/brother/dog/etc. are inventing WADFs, which get haphazardly adopted by organizations throughout the world leaving web developers feeling their resume is somehow inadeqate.
Some of them seem to prefer WebWork, which is now part of OpenSymphony. Debate over Struts is raging.
This is what is so frustrating, where the turn-over in fashionable WADFs is rampant. I would cringe upon hearing "Struts is so, like, 2001." That was only two years ago! Ugh.
Rather than adopting a "framework" that is almost certain to fall short in its capacity, why don't web developers adopt something even better: objective simplicity. Frameworks can be a constant battling ground for new employees and old employees alike, when learning and re-learning the framework becomes burdensome. Isn't good software architecture supposed to make things intuitive, even to the average software developer?
I believe the multitude of frameworks are the product of severe NIH syndrome, rather than genuine well-intentioned common sense.
Well, certainly OS X wouldn't work where I'm at, where we deal with high-end CAD/CAM/CAE systems. They just plain don't write this stuff for Macs. And they never will.
I woudn't make this conclusion too soon. It is easier than ever before to port from UNIX to Mac OS, and I wouldn't be suprised if Apple's market share is compelling to CAD companies. Imagine, a highly-usable and well-known desktop computer that isn't garbage like that other well-known and useable desktop.
The trend now is to Windows...
This has always struck me as deeply sad. There are now many decent CAD applications totally and irrevocably stuck in the Windows mud-pit. Porting them to Mac OS, Linux, or anything that isn't non-operably joined to Microsoft is not feasible. Simply, if Microsoft sinks like a rock, a lot of companies will realize that they are tethered to that rock. Very sad.
...and (hopefully with the pending release of Pro/E for Linux) to Linux systems.
I think this is very encouraging. Pro/E has been coming down in cost quickly enough that the hardware costs are becoming larger parts of the equation. A dual Opteron system, a top-end graphics card (not a gamer's card but one of those $1,000+ ones), and a Linux distribution running Pro/E is very compelling relative to a Sun, IBM, SGI, or HP workstation.
but a migration of data from Oracle to Mysql along with the associated training and support (yeah support still ain't free) is likely to cost you more than you're saving.
Ha. Administering MySQL is child's-play relative to Oracle. Just mentioning the word "Oracle" implies a full-time-equivalent adminstrator on staff just to do everyday operations along with dealing with Oracle's support staff. If an established Oracle DBA can't--or won't--figure out MySQL in less than a week, they might be better replaced by someone who can.
Slashdot Mirror
My girlfriend's mother did...she said it did jack shit for grass and mud stains.
If this is true, why does OxyClean take up more shelf space at my local Wal-Mart than most of the other detergents? Isn't Wal-Mart the store that will only sell stuff that has something like 97% per second throughput? Would stuff sell like that if it did "jack shit"?
Why do PC Notebook components require 3 extra lbs!??
Considering the Pentium 4 in this Toshiba counter-top model draws three or four times the power of the G3 and G4 CPUs that Apple uses, they probably have an extra pound of battery plus an extra pound of heat sink in there somewhere.
Geez, at 10 lbs...airlines might need to charge for an extra ticket.
The problem with Javascript is that there are so many crappy programs out there that don't properly utilize the language, resorting to stupid 'Netscape' or 'IE' detection hacks rather than testing for the existance of functions.
Very good advice.
I wonder if there will be an "autoconf" program for JavaScript, one day...
Flash is currently the best format for animated content on the internet.
Yes, but not for animated websites. Whole websites that are based on flash just make me feel sorry for the losers who paid for them.
CSS, Macromedia Flash...
Yup, CSS (two or three versions), Flash, JavaScript (multiple versions), XML (+123 XML-related acronyms), proprietary browser exensions, ActiveX, people still using Netscape 4.5beta, and IE-specific sites sure make the me feel all warm and fuzzy about web "innovation".
In about ten more years, after the dust settles, and all that's left is XML (+10 acronyms), perhaps, then, web developers can grow their hair back.
People who are driven by the "experience" of the web, where everything needs to look and feel like Fox News, need to be put into time-out to rethink thier plan.
That file is dated 1987, OS/2 1.0 joint code with Microsoft and IBM.
I don't know, but perhaps 1987 was the last time they really needed to write a 32-bit kernel loader. 1987 was probably the last time that anyone need to write a 32-bit loader (look in specific spot on disk, load a file, run it). Regardless, I'm suprised that the copyright didn't read "Regents of UC Berkeley" or something like that.
"Best Practices" is yet another ridiculous business term that is crossing over into the IT world.
It does have a little meaning. Best practices are basically what a person would do after spending some time thinking about what would be an elegant way to solve a problem.
Best practices are not achieved after one 30-minute meeting. Nor are they achievable by design-by-committee. They require people on a team to be honest, first, and political, last.
Unfortunately, this rarely occurs in the real world, especailly due to the technological religious fanaticism present among most people.
Like all ORA books, this one was typeset using FrameMaker.
Hmm..FrameMaker is the only program I remember using that rivaled TeX in the flexibility of page layout and quality of the output. Framemaker even accepts EPS for embedded graphics.
We should just be happy that O'Reilly chose a non-Microsoft application, because a Linux book written in Word, for example, would not only be hideous in appearance but leave OSS users seriously conflicted.
How many people out there use java in their web based apps that interface with a database server backend?
I would bet that nearly all the use of Java is for web-based apps that interface to a database.
What platforms do you use for development?
For Java, it would be J2SE from Sun (it includes JDBC).
What Dbs do you use?
Oracle, MySQL, PostgreSQL, DB2, etc.
What web server?
Apache, WebLogic, Sun ONE, etc.
These questions, at face value, mainly answer themselves. Perhaps a better question would be: what do Java programmers do to prevent database applications from turning into a quagmire of naive mistakes and bottlenecks making a well-intentioned application into a sick joke of engineering?
Database programming is one of those things that is simple enough, at first glance, to be attempted by every college sophomore on the planet, but complex enough, in reality, to make people who hired those college sophomores look foolish.
Neither Linux or Windows has died out yet so none of them are passing fads.
To make a biological analogy, Windows could be viewed as a lifeform already occupying 90% of its ecosystem with little room for growth and debatable ability to reuse the space behind it (i.e., Microsoft is its own largest competitive force in the market).
When an established ecosystem dies, the smaller players, e.g. Linux and others, come to fill in the area now fertilized by the decomposing life from before. The prior lifeform might still exist but only in a much smaller niche surrounded on all sides.
BG: . . . We will never have a price lower than Linux, in terms of just what you charge for the software. We compete on the basis of, if you look at the value you get out of the system and the overall cost that the system has that apply in our software. For any project, if you look at communications costs, hardware costs, personnel costs, all that, software licensing ranges -- the highest you'd ever find is, like, 3% of any IT-type project. And so the question is can that 3% [compensate], in terms of how quickly you get the system set up?
What's so interesting is that this exact same argument is used by other companies against Microsoft. Sun does it with their servers and will be doing it with that Mad Hatter desktop. I'm sure Red Hat is doing it every day. Even IBM will come in with a jab or two about their servers.
I find the bit about "personnel costs" regarding Microsoft's offerings hilarious, by the way. Bill G. himself admitted two weeks per person per year lost due to Windows 9x. How much do to Outlook viruses? How much due to IIS worms? How much due to constant manual intervention for Windows servers?
They run on IBM's big iron mainframes, they use IBM's WebSphere (JSP and the whole shebang), and they use OS/2 on their desktops (with Netscape 4).
It would be interesting if IBM created a porting layer for Linux, so eventually those OS/2 apps could run on IBM-branded Linux desktops.
America and the West taking away the freedom of all computer users, and the Chinese coming to the rescue and restoring our freedom.
Isn't it ironic?
There will be no more small developers in windows, which means MS will have no one to rip fresh ideas from!
Also, it seems MS' bread-n-butter is those small MCSE outfits that sell "solutions" to local marks--er, other small businesses. How will these outfits deal with both those businesses who haven't upgraded since 1996 and those who demand the latest and greatest? TCPA means that the recently upgraded business just might not be able to talk to the 1996-vintage one. Should the 1996-vintage one be required to pay thousands of dollars in upgrades, when all they need is a spreadsheet for payroll, for example?
so, what, should we stock up on Win2000, XP, and Linux OS's along with our CD and DVD burners?
Yes. Keeping a working late-90's or early-2000's vintage computer around the house is a very good idea. The late-90's was the era of the forever "fast enough" CPU. For example, at home, I have yet to own a computer faster than 300MHz. Mozilla launches considerably faster than the modem will dial-in, so this is a good threshold of usefulness for me.
It'll be made illegal as a "circumvention device" in short order if it actually turns out to be useful for any sort of multimedia applications.
Tell that to the secondary market for UNIX workstations. There are literally dozens of companies willing to sell used Sun workstations, for example, for well under $1,000. This isn't a bad deal for a workstation that is built like a tank (excluding Ultra 5 and 10, of course) and has a decent CPU, ECC memory, and SCSI disks.
I've also had decent luck with non-Sun-branded parts in Sun workstations, such as regular Seagate SCSI drives and SCSI-2 CD-RW drives.
The only catch is that the official Solaris right-to-use license for used hardware is approx. $100, and the optional media kit is also about $100 (potential total of $200). Still not so bad for official UNIX(R). Solaris 9 also comes with tools for CD burning, viewing MPEG files, StarOffice, and, recently, Netscape 7.
If Solaris isn't your cup of tea, Linux and the *BSDs also all run on UltraSPARC, so you can still feel Free, if you like.
NTFS supports reparse points. No UNIX required.
Are "reparse points" equivalent to symbolic links in UNIX? Symbolic links link pathnames rather than inodes and can span filesystems.
Objective simplicity is a laudable goal, but it conflicts with the goals of flexibility, performance, and ability to meet the requirements of the subject matter (if the subject matter is complex).
With regard to performance and flexibility, simplicity can be good, given that, sometimes, simplicity is actually more difficult to attain initially than ad-hoc complexity, which the popular frameworks tend to encourage (i.e., simplicity often requires some thought and modesty, but the fruit of that effort is forever delicious).
For example, in the long-run, it is easier to maintain a hierarchy of HTML/XHTML and CGI pages or a multi-tier JavaScript/XML/JSP/Servlet/Struts/EJB/etc/etc application? In practice, honestly, which is more likely to perform well, work predictably across browsers, and withstand the tests of time? Which method can be picked up easily by new hires or relearned quickly enough to make incremental changes month-to-month practical?
For very complex content, warranting permanent staff, it seems plausible (not speaking from experience) that DocBook could make for a very good base format, and the maturing XML tools, like XSLT, could provide a fairly direct and simple means of generating web pages as needed. Hourly Perl scripts could even keep static content up-to-date without adding much complexity and removing much of the need for real-time dynamic content. Something like DocBook also means for approximately one amount of effort, a multitude of outputs are possible (HTML and PostScript, for example). While this method might not be totally practical today, due to the immaturity of XML, it does seem promising.
Either way, it seems that simple tools can be used to make a solid website, where a one-page chart is sufficient for a person to really understand how things fit together.
where Samba, NFS, and UNIX symbolic linking just might save the day.
If possible the tool would have the ability to filter files based on an exclude and an include list, file size, and last access date.
Awk! Awk! Awk! (clears throat) excuse me, uh, Perl or sh scripts might be useful, here.
a better way to accomplish this?
UNIX?
www.cnn.com
:)
Using Netscape 7.1 for Windows (I am at work, ok?) I still can't vote in a poll properly (pop-up appears, but the vote data is loaded in main window, leaving pop-up blank).
I think Netscape 7.1's Good Taste algorithms are working just fine. Good CSS support and preventing people from voting in Wolf's damned sensationalistic and myopic polls is win-win
Why do most software packages continue to add features without actualy providing a subjective goal to strive for?
Buzzwords are the ends and the means, in the fashion industry...er, I mean IT industry.
When will the goal of a products feature-list finally be met?
It never will. The true end of the Earth will be when Microsoft's code-base becomes so massive that it forms a black hole destroying everything within 3 light-years.
Microsoft(R)'s Windows(TM) name isn't describing the project's code name, yet the product's retail names are somthing like 95, 98, NT, 2000, XP. Well, it is honest to say that the NT and 2000 products are similar, 95 and 98 products are similar, and the XP product isn't quite similar to 95 and 98 and 2000, but then there was the fiasco that all the 9x users went out to buy product Windows(TM) 2000 and found they had been tricked by Microsoft that 2000 would be like a 95 or 98. Anyone see my point?
You wouldn't last a day working at Microsoft's marketing department. They brand Windows deceptively because it is in their interest to do so! They are professional liars and theives. You shouldn't be suprised.
Does anyone think they should continue calling those products by their initial names AFTER the programming syntax and methodology becomes completly different or non-compatible than they were first designed?
Yes, in the sense that they are now branded and well-known. For example, a new Chevy Impala bears no resemblance to the two-ton behemoth from 1971 bearing the same name. Yet, it is still familiar, and it is a car from Chevrolet, accomplishing their goals of how to market this car. It is well-publicised that Perl 6 is vastly different than Perl 5, and it is also well-known that people who assume interoperability across major versions of software are foolish. It doesn't matter that Microsoft's marketing department exerts tremendous psychological control over their customers, those customers are still foolish.
Yes.
Struts is a web application development framework (WADF), of which there are many.
While factually correct, "many" might be an understatement. It seems everyone and their uncle/brother/dog/etc. are inventing WADFs, which get haphazardly adopted by organizations throughout the world leaving web developers feeling their resume is somehow inadeqate.
Some of them seem to prefer WebWork, which is now part of OpenSymphony. Debate over Struts is raging.
This is what is so frustrating, where the turn-over in fashionable WADFs is rampant. I would cringe upon hearing "Struts is so, like, 2001." That was only two years ago! Ugh.
Rather than adopting a "framework" that is almost certain to fall short in its capacity, why don't web developers adopt something even better: objective simplicity. Frameworks can be a constant battling ground for new employees and old employees alike, when learning and re-learning the framework becomes burdensome. Isn't good software architecture supposed to make things intuitive, even to the average software developer?
I believe the multitude of frameworks are the product of severe NIH syndrome, rather than genuine well-intentioned common sense.
Well, certainly OS X wouldn't work where I'm at, where we deal with high-end CAD/CAM/CAE systems. They just plain don't write this stuff for Macs. And they never will.
...and (hopefully with the pending release of Pro/E for Linux) to Linux systems.
I woudn't make this conclusion too soon. It is easier than ever before to port from UNIX to Mac OS, and I wouldn't be suprised if Apple's market share is compelling to CAD companies. Imagine, a highly-usable and well-known desktop computer that isn't garbage like that other well-known and useable desktop.
The trend now is to Windows...
This has always struck me as deeply sad. There are now many decent CAD applications totally and irrevocably stuck in the Windows mud-pit. Porting them to Mac OS, Linux, or anything that isn't non-operably joined to Microsoft is not feasible. Simply, if Microsoft sinks like a rock, a lot of companies will realize that they are tethered to that rock. Very sad.
I think this is very encouraging. Pro/E has been coming down in cost quickly enough that the hardware costs are becoming larger parts of the equation. A dual Opteron system, a top-end graphics card (not a gamer's card but one of those $1,000+ ones), and a Linux distribution running Pro/E is very compelling relative to a Sun, IBM, SGI, or HP workstation.
but a migration of data from Oracle to Mysql along with the associated training and support (yeah support still ain't free) is likely to cost you more than you're saving.
Ha. Administering MySQL is child's-play relative to Oracle. Just mentioning the word "Oracle" implies a full-time-equivalent adminstrator on staff just to do everyday operations along with dealing with Oracle's support staff. If an established Oracle DBA can't--or won't--figure out MySQL in less than a week, they might be better replaced by someone who can.