Did anyone notice they said the max resolution on DVI was 1600x1200 and not 1920x1200, thereby either wasting space or haveing a 4:3 image stretched? DVI is really the only way to go with a high end monitor like that, especially at that resolution. Analog-ing it would almost be a waste.
I didn't see any special clarification on the LG website specs, I wonder if he got it wrong or that monitor is really crippled that way.
I've got a Samsung 213t 21 inch LCD, and I love it. I have an LG fridge, and I love that too.
There's nothing more obnoxious when a pretentious web designer asks you, in the web page, to change your resolution in order to view their web page.
Like I'm really going to mess up my icons, window placements, and so forth, as well view a non-optimial resolution on my LCD screen, just so I can better experience your badly designed and overly elaborate UI.
Well, while I don't use MySQL, this does look like 4.0 to me. According to the make file there is a package available via ftp and on the CD-ROM. I also doubt you bothered to apply all of the patches in the patches directory when you made it from source yourself. That's proably why it doesn't work. On a general note if you ask on misc@ they are going to tell you to just use the package and only to use the port if you need a special flavor (and it doesn't sound like you do).
They've just added MySQL 4.0 for OpenBSD 3.5. It doesn't exist on the 3.4 packages, and it wasn't in ports for 3.4.
If you got a snapshot on the 29th you probably didn't get 3.5. Instead you ended up with current, and not only that, but you got it durring the greatest period of change and instability....
Well, it booted up as 3.5, so if it wasn't 3.5 it was sure a neat trick:) Either way, it didn't behave any differently.
However, I was able to find a solution. 0racle mentioned that there was a long standing issue with MySQL and OpenBSD, so I took a look in the MySQL manual and found an entry regarding the same issue for OpenBSD 2.8. It hadn't come up in my numerous Google searches before, probably because I was searching for OpenBSD 3.x issues. I used the suggested fix, and SQL-bench was able to run. It just ran last night.
However, the results were pretty terrible for the insert operations. It took longer than even the bad FreeBSD 5.1 result, about 3 times longer than with Solaris, FreeBSD 5.2.1, or Linux on the same hardware. I'll publish an addendum to the article soon with the results.
It didn't crash the system because of an invalid XF86Config file, it crashed because he tried to start X one a Serial Console, you can't do that.
As I stated in the article, I figured it probably wouldn't run and die with an error. What I was surprised about was that it actually caused the operating system to crash.
Interesting, I hadn't seen that before. It's for 2.8, but it seems to be the same issue I had. When I get back from Hawaii I'll run MySQL with that tweak, see if that resolves the problem.
For example he tried to run the various X configurations utilities. The FAQ clearly states that there is a WORKING example configuration that you should start from in/usr/X11/README.
The FAQ does not state this at all. Which FAQ are you refering to? ftp://ftp.openbsd.org/pub/OpenBSD/doc/obsd-faq.txt
While it was odd that starting up X11 with a non-working config crashed the system, I was able to get X11 working, as you can see in the screenshot.
Furthermore the FAQ also states to not compile from source unless absolutely nessessary.
That is not what the FAQ says. In fact, it only references compiling when talking about ports versus packages, and not compiling in general. It says that packages are easier, and it's recommended to use them, but it seems more of trouble-saving and time-saving advice than anything else, not an admonishment of compiling.
If he had used packages, he might not have had the problems with the databases that he had. However there was a MySQL glitch in 3.4 (I think, it could have been 3.3) that was fixed in stable.
I've gotten several comments about using packages. For one, I'm using MySQL 4.0. There are no packages or even ports for MySQL 4.0. I've been using it with the other evaluations, and it is MySQL's latest.
I prefer to compile from source whenever possible. Packages and ports are a convienence, and can often contain patches and workarounds, especially on non-x86 distributions. It would seem silly to rely entirely on ports/packages.
There is a port for PostgreSQL which I used and had the same bus error problems, although I didn't use a precompiled package. I'll give that a spin when I'm by the system, but right now I'm in Hawaii so I don't have access to the OpenBSD system.
Also, the 3.5 snapshot from the 29th had some serious problems (people were told not to use it). Doing a little more homework would have avoided these problems; it's all documented.
The 3.5 snapshot worked fine for me, and I didn't see any "do not use" messages anywhere. The only problem I had with it was it exhibited the same strange MySQL behavior that 3.4 did. For me, it worked exactly the same as 3.4. So I'm not sure what "problems avoided" you are referring to.
You'll note that I did solicit the help of the OpenBSD/sparc mailing list, and while I recieved some excellent suggestions, nothing was able to sort out the particular issue.
While Solaris can be downloaded for free, it cannot always be used for free. The Solaris Binary License has provisions that allow it for development use and educational use for free, but otherwise you've got to pay to play. No one seems to get that. Does Sun enforce those licenses? Not that I've ever heard, but it's still an issue of legality.
If you've got some old hardware, and you want to run some license-inencombered operating system, then the alternative operating systems are a great bet.
There a numerous other advantages as well, such as much more extended hardware support (Sun wants you to pay $400 for a FE card, where you can use a $10 off-the-shelf PCI card with FreeBSD, NetBSD, etc), access to the source code, perhaps a smaller footprint, access to security patches for applications that Sun might charge for (not all of Sun's patches are free).
While people shouldn't just abandon Solaris, I love it too, there are plenty of cases where the alternatives make more sense than Solaris.
While not using applications that are specifically written for 64-bit (I know of no open source application that is specifically written with 64-bit in mind) and testing limited (all benchmarks are inherintly limited), these are a start, and represent more than just the pure conjecture which seems to be pervasive. Of course these are test on Solaris 9 for SPARC, so other platforms will likely vary.
Just think how marketing has imprinted in our brain that "milk is good for you". In fact, those claims would have to be described as "unsubstantiated".
http://notmilk.com
There're plenty more where that came from. Imagine drinking cat's milk, or rat's milk, or even horse milk. Why then, is it not disgusting to drink cow's milk? Marketing.
With synthetic bovine hormones (illegal just about everywhere except the US), and rampant use of antibiotiocs, it's even more disgusting.
I'm a vegan geek who works 8 hours sitting down in an office everday, but I also run Marathons.
Running is a great way to keep in shape, but if running isn't your thing (and it's not everyones), then do walking. There are numerous reports on the health benefits of simply walking for an extended period of time a couple of times a week. It burns off caleries, keeps your body in motion, and it does a body good. Walk a few miles a few times a week, and you'll probably start to notice a difference.
If you reclaimed all the energy it took to vibrate a cell phone, it wouldn't vibrate. That's like an electric motor reclaiming it's power by running it through a generator.
I'll have to second any motions for Suze Orman. She's genuinely good person with the ability to help people boil down their complex money problems to the simple key issues.
Most financial books are crap. They talk about how to make money on the stock market, get rich quick, and appeal to the people "it's that easy" mentality.
There's also a class of books that really only relates to fairly rich people, but appeal to those who aren't rich, but wish to be. Kind of like reading magazines about Ferrari's when all you can afford is a Honda. It can be entertaining, but it's not going to help your financial siutation.
Suze Orman's books aren't going to get you rich in a hurry. She doesn't give stock tips. She doesn't give direct investment advice. She tells you the patterns.
Also, when talking to friends about money, there are two types: Friends that seem to have money and friends that seem to be always broke, regardless of income. The ones that are always broke seem to have a never ending supply of financial advice. Ignore them all.
Remember, the only way to get rich quick is to either be extremely lucky, be extremely well connected, or to be extremely good at what you do and lucky. If you're not extremely good or well connecting in the stock market, you're not going to get rich quick off of them unless you're extremely lucky.
The problem with Linux installs from any of the vendors is that it works great for some hardware installs, but awful for others.
I installed RedHat 8.0 on my system, and it found my AC97 audio, and a bunch of other stuff. Didn't work at all for my scanner, and spending days trying to get it to work was frustrating. Also, compiling in NTFS readability with a RedHat source kernel is not my idea of fun, especially when the compile fails.
Mandrake 9.1 found my scanner, but didnt' see my audio, so I had to go an manually compile the drivers, and if you switch kernels from the distro's kernel, lots of desktop issues can crop up (not so with a server).
RedHat 9.0 found my scanner, found my audio, but the install itself didn't like my video card. I've got an nvidia geforce ti 4200, and it tried to use the VESA driver with my LCD display and choked. No way around it, so I had to do a text install, and install the nvidia drivers via text. Not a big deal, but it's not something most consumers would be willing or able to do.
So when you say "it works great for you", remember there are millions of other computer users with various configurations. Linux desktop installs lack the flexibility to work right on too many systems.
Linux advocation needs to get rid of the myiopic view of "it works great for me" as the end-all argument.
For database servers, that's very useful. However, the added expense of hot-swapping is pretty useless when dealing with web servers. A couple of years ago Sun's idea of a entry-level web server was a Sun E250, which when loaded with dual 300 MHz processors and a Gig of RAM ran you around $25,000.
Of course, that's absolutely rediculous when considering it's just a web server, and an Intel/AMD -based system will run you around $2,000 to 4,000 when totally blown out to get the same thing.
There is an angular velocity to atoms that give them "spin". It was called spin because it was an angular velocity, although they obviously couldn't see it.
I believe they are suggesting that this might be the force behind the quantum angular velocity, or spin. Calculations and experiments have shown particles like electrons to have an angular velocity, perhaps this is a physical manifestation of that. Not quite sure. A neutron has no charge, but a spin of 1/2, although it's made up of two down quarks and one up quark, which do have angular velocity.
For the government has never abused it's power... such as the FBI performing surveliance on Dr Martin Luther King? Or J Edgar Hoover being a man that even presidents feared?
All because they were able to compile information on people with impunity. Many of the privacy laws we have today are a direct result of abuses in the past. Marginalizing those concerns as "luddite" is an insult to America and to true patriots.
Let's face it. Times are tough. Budgets and earnings aren't what it used to be. We need to find ways to save money and tigthen our belts.
In short, we've got to outsource upper management to off-shore countries.
There are plenty of well trained and highly educated people in foreign countries that can do excellent upper-management work: CEOs, CFOs, vice presidents... and they'll do them for pennies on the dollar of the exhorbatent prices we pay for CEOs now. 40 million dollar golden parachute? No more!
I think it's not a viable platform for Java. The threads issue is a huge hassle, and there's all sorts of problems that can arrise that are difficult to troubleshoot in Java even on a native platform.
I work at a site that used the 1.1.8 native release on FreeBSD for a while, and it went ok. Then we upgraded some Java code, and things went quickly downhill. We'd have hangups that we were never able to figure out, and each server would have one at least once a day, which was a nightmare for me. We used Java, Tomcat, and Apache 1.3.27.
Even though the company I work for is fairly FreeBSD fantatical, it wasn't a hard choice to switch to Linux and IBM's 1.4.1 JDK. We switched, plopped on our code, and boom, no more problems.
For a development environment, there's definately something to be said for using a viable and trusted environment. Not to say I don't trust FreeBSD, but I don't trust FreeBSD with either the not-so-tested ports or the Linux JVM under emulation. For all that trouble, might as well make the switch. Right tool for the right job, and all that.
SCO has effectively alienated the geek community, which means they are either incredibly stupid (similarly business decisions have been made by smarter companies before) or the comments about their ultior motives are definately true.
The geek community made Linux, it made FreeBSD, and it makes or breaks the companies involved. Alienating them is business suicide.
Slashdot Mirror
Did anyone notice they said the max resolution on DVI was 1600x1200 and not 1920x1200, thereby either wasting space or haveing a 4:3 image stretched? DVI is really the only way to go with a high end monitor like that, especially at that resolution. Analog-ing it would almost be a waste.
I didn't see any special clarification on the LG website specs, I wonder if he got it wrong or that monitor is really crippled that way.
I've got a Samsung 213t 21 inch LCD, and I love it. I have an LG fridge, and I love that too.
A friend of mine a while ago would make neck chains out of old HyperSPARC and SuperSPARC processor modules, ala Flavah Flav from Public Enemy.
There's nothing more obnoxious when a pretentious web designer asks you, in the web page, to change your resolution in order to view their web page.
:)
Like I'm really going to mess up my icons, window placements, and so forth, as well view a non-optimial resolution on my LCD screen, just so I can better experience your badly designed and overly elaborate UI.
End rant
They've just added MySQL 4.0 for OpenBSD 3.5. It doesn't exist on the 3.4 packages, and it wasn't in ports for 3.4.
If you got a snapshot on the 29th you probably didn't get 3.5. Instead you ended up with current, and not only that, but you got it durring the greatest period of change and instability....
Well, it booted up as 3.5, so if it wasn't 3.5 it was sure a neat trick :) Either way, it didn't behave any differently.
However, I was able to find a solution. 0racle mentioned that there was a long standing issue with MySQL and OpenBSD, so I took a look in the MySQL manual and found an entry regarding the same issue for OpenBSD 2.8. It hadn't come up in my numerous Google searches before, probably because I was searching for OpenBSD 3.x issues. I used the suggested fix, and SQL-bench was able to run. It just ran last night.
However, the results were pretty terrible for the insert operations. It took longer than even the bad FreeBSD 5.1 result, about 3 times longer than with Solaris, FreeBSD 5.2.1, or Linux on the same hardware. I'll publish an addendum to the article soon with the results.
It didn't crash the system because of an invalid XF86Config file, it crashed because he tried to start X one a Serial Console, you can't do that.
As I stated in the article, I figured it probably wouldn't run and die with an error. What I was surprised about was that it actually caused the operating system to crash.
You were of great help :)
I just checked the manual, and found this:
OpenBSD 2.8 problem
Interesting, I hadn't seen that before. It's for 2.8, but it seems to be the same issue I had. When I get back from Hawaii I'll run MySQL with that tweak, see if that resolves the problem.
For example he tried to run the various X configurations utilities. The FAQ clearly states that there is a WORKING example configuration that you should start from in /usr/X11/README .
t
The FAQ does not state this at all. Which FAQ are you refering to? ftp://ftp.openbsd.org/pub/OpenBSD/doc/obsd-faq.tx
While it was odd that starting up X11 with a non-working config crashed the system, I was able to get X11 working, as you can see in the screenshot.
Furthermore the FAQ also states to not compile from source unless absolutely nessessary.
That is not what the FAQ says. In fact, it only references compiling when talking about ports versus packages, and not compiling in general. It says that packages are easier, and it's recommended to use them, but it seems more of trouble-saving and time-saving advice than anything else, not an admonishment of compiling.
Ports Versus Packages
If he had used packages, he might not have had the problems with the databases that he had. However there was a MySQL glitch in 3.4 (I think, it could have been 3.3) that was fixed in stable.
I've gotten several comments about using packages. For one, I'm using MySQL 4.0. There are no packages or even ports for MySQL 4.0. I've been using it with the other evaluations, and it is MySQL's latest.
I prefer to compile from source whenever possible. Packages and ports are a convienence, and can often contain patches and workarounds, especially on non-x86 distributions. It would seem silly to rely entirely on ports/packages.
There is a port for PostgreSQL which I used and had the same bus error problems, although I didn't use a precompiled package. I'll give that a spin when I'm by the system, but right now I'm in Hawaii so I don't have access to the OpenBSD system.
Also, the 3.5 snapshot from the 29th had some serious problems (people were told not to use it). Doing a little more homework would have avoided these problems; it's all documented.
The 3.5 snapshot worked fine for me, and I didn't see any "do not use" messages anywhere. The only problem I had with it was it exhibited the same strange MySQL behavior that 3.4 did. For me, it worked exactly the same as 3.4. So I'm not sure what "problems avoided" you are referring to.
You'll note that I did solicit the help of the OpenBSD/sparc mailing list, and while I recieved some excellent suggestions, nothing was able to sort out the particular issue.
"Hydrogen is the most abundant element in the atmosphere. I believe it is the fuel of the future," said Kamaruzzaman.
While Solaris can be downloaded for free, it cannot always be used for free. The Solaris Binary License has provisions that allow it for development use and educational use for free, but otherwise you've got to pay to play. No one seems to get that. Does Sun enforce those licenses? Not that I've ever heard, but it's still an issue of legality.
If you've got some old hardware, and you want to run some license-inencombered operating system, then the alternative operating systems are a great bet.
There a numerous other advantages as well, such as much more extended hardware support (Sun wants you to pay $400 for a FE card, where you can use a $10 off-the-shelf PCI card with FreeBSD, NetBSD, etc), access to the source code, perhaps a smaller footprint, access to security patches for applications that Sun might charge for (not all of Sun's patches are free).
While people shouldn't just abandon Solaris, I love it too, there are plenty of cases where the alternatives make more sense than Solaris.
64-bit versus 32-bit
To correct, in an obnoxious, arrogant, and belligerent manner, what could easily be just a brain fart on the part of the submitter?
The answer is around 50 so far.
The SPARC platform is expensive, and you can get a much better value by using x86 hardware.
Just think how marketing has imprinted in our brain that "milk is good for you". In fact, those claims would have to be described as "unsubstantiated".
http://notmilk.com
There're plenty more where that came from. Imagine drinking cat's milk, or rat's milk, or even horse milk. Why then, is it not disgusting to drink cow's milk? Marketing.
With synthetic bovine hormones (illegal just about everywhere except the US), and rampant use of antibiotiocs, it's even more disgusting.
I'm a vegan geek who works 8 hours sitting down in an office everday, but I also run Marathons.
Running is a great way to keep in shape, but if running isn't your thing (and it's not everyones), then do walking. There are numerous reports on the health benefits of simply walking for an extended period of time a couple of times a week. It burns off caleries, keeps your body in motion, and it does a body good. Walk a few miles a few times a week, and you'll probably start to notice a difference.
I just got my notice in the mail. I'm giving mine up to the public domain after 14.
If you reclaimed all the energy it took to vibrate a cell phone, it wouldn't vibrate. That's like an electric motor reclaiming it's power by running it through a generator.
I'll have to second any motions for Suze Orman. She's genuinely good person with the ability to help people boil down their complex money problems to the simple key issues.
Most financial books are crap. They talk about how to make money on the stock market, get rich quick, and appeal to the people "it's that easy" mentality.
There's also a class of books that really only relates to fairly rich people, but appeal to those who aren't rich, but wish to be. Kind of like reading magazines about Ferrari's when all you can afford is a Honda. It can be entertaining, but it's not going to help your financial siutation.
Suze Orman's books aren't going to get you rich in a hurry. She doesn't give stock tips. She doesn't give direct investment advice. She tells you the patterns.
Also, when talking to friends about money, there are two types: Friends that seem to have money and friends that seem to be always broke, regardless of income. The ones that are always broke seem to have a never ending supply of financial advice. Ignore them all.
Remember, the only way to get rich quick is to either be extremely lucky, be extremely well connected, or to be extremely good at what you do and lucky. If you're not extremely good or well connecting in the stock market, you're not going to get rich quick off of them unless you're extremely lucky.
The problem with Linux installs from any of the vendors is that it works great for some hardware installs, but awful for others.
I installed RedHat 8.0 on my system, and it found my AC97 audio, and a bunch of other stuff. Didn't work at all for my scanner, and spending days trying to get it to work was frustrating. Also, compiling in NTFS readability with a RedHat source kernel is not my idea of fun, especially when the compile fails.
Mandrake 9.1 found my scanner, but didnt' see my audio, so I had to go an manually compile the drivers, and if you switch kernels from the distro's kernel, lots of desktop issues can crop up (not so with a server).
RedHat 9.0 found my scanner, found my audio, but the install itself didn't like my video card. I've got an nvidia geforce ti 4200, and it tried to use the VESA driver with my LCD display and choked. No way around it, so I had to do a text install, and install the nvidia drivers via text. Not a big deal, but it's not something most consumers would be willing or able to do.
So when you say "it works great for you", remember there are millions of other computer users with various configurations. Linux desktop installs lack the flexibility to work right on too many systems.
Linux advocation needs to get rid of the myiopic view of "it works great for me" as the end-all argument.
I was referring to 5 years ago. Also, Sun still sells the Sun E250. Lamer.
For database servers, that's very useful. However, the added expense of hot-swapping is pretty useless when dealing with web servers. A couple of years ago Sun's idea of a entry-level web server was a Sun E250, which when loaded with dual 300 MHz processors and a Gig of RAM ran you around $25,000.
Of course, that's absolutely rediculous when considering it's just a web server, and an Intel/AMD -based system will run you around $2,000 to 4,000 when totally blown out to get the same thing.
There is an angular velocity to atoms that give them "spin". It was called spin because it was an angular velocity, although they obviously couldn't see it.
I believe they are suggesting that this might be the force behind the quantum angular velocity, or spin. Calculations and experiments have shown particles like electrons to have an angular velocity, perhaps this is a physical manifestation of that. Not quite sure. A neutron has no charge, but a spin of 1/2, although it's made up of two down quarks and one up quark, which do have angular velocity.
Not sure here.
For the government has never abused it's power... such as the FBI performing surveliance on Dr Martin Luther King? Or J Edgar Hoover being a man that even presidents feared?
All because they were able to compile information on people with impunity. Many of the privacy laws we have today are a direct result of abuses in the past. Marginalizing those concerns as "luddite" is an insult to America and to true patriots.
Let's face it. Times are tough. Budgets and earnings aren't what it used to be. We need to find ways to save money and tigthen our belts.
In short, we've got to outsource upper management to off-shore countries.
There are plenty of well trained and highly educated people in foreign countries that can do excellent upper-management work: CEOs, CFOs, vice presidents... and they'll do them for pennies on the dollar of the exhorbatent prices we pay for CEOs now. 40 million dollar golden parachute? No more!
BUT
I think it's not a viable platform for Java. The threads issue is a huge hassle, and there's all sorts of problems that can arrise that are difficult to troubleshoot in Java even on a native platform.
I work at a site that used the 1.1.8 native release on FreeBSD for a while, and it went ok. Then we upgraded some Java code, and things went quickly downhill. We'd have hangups that we were never able to figure out, and each server would have one at least once a day, which was a nightmare for me. We used Java, Tomcat, and Apache 1.3.27.
Even though the company I work for is fairly FreeBSD fantatical, it wasn't a hard choice to switch to Linux and IBM's 1.4.1 JDK. We switched, plopped on our code, and boom, no more problems.
For a development environment, there's definately something to be said for using a viable and trusted environment. Not to say I don't trust FreeBSD, but I don't trust FreeBSD with either the not-so-tested ports or the Linux JVM under emulation. For all that trouble, might as well make the switch. Right tool for the right job, and all that.
SCO has effectively alienated the geek community, which means they are either incredibly stupid (similarly business decisions have been made by smarter companies before) or the comments about their ultior motives are definately true.
The geek community made Linux, it made FreeBSD, and it makes or breaks the companies involved. Alienating them is business suicide.