I remember when I was at the University of Illinois in Urbana, someone built a robotic arm that used a webcam and some image processing algorythms to pick up a pop can that it had no other idea where was.
BTW, the main push of the project was that the robot used a learning algorythm to determine how to pick up the can. Basically all that was programmed was a way to interpret the images coming from the webcam, and a "desire" to pick up the can. This caused some interesting effects like when the arm would pass out of view of the camera, the robot "learned" to wave the arm at the camera to re-locate itself.
My plan would involve a computer, with a WiFi link to the hardware on the mower (possibly a micro-ATX or scavenged laptop, something with a serial port to communicate with a microcontroller). Connect the computer to a webcam that can see the entire yard. Roll-your-own image processing algorythms, and there you go. Plus, for safety reasons, you could have the thing shut off when the webcam recognizes a fast moving object (neighbor's brat kid that can't keep the football in his own yard). Or, if you think you can mount enough processing power on the mower itself, you could do away with the need for the WiFi link. This would definately require the webcam to be wireless, then.
Or, configure Linux box as a router. Plug in a second NIC. I used to put FreeBSD on 486's and early Pentium systems that I found in the garbage to be routers to babysit windows machines online. I've heard it's even easier to do in Linux, but have never tried.
If ECE stands for Electrical and Computer Engineering, you'd better plan on graduating fairly high in your class. I got a EE degree from a top 10 school (University of Illinois), but my grades weren't that great, so I didn't get any internships or job leads straight out of school. After graduation, I found that most any job that you'd want as an EE you would have had to basically be hired right out of college (think job fairs). On the "open" market, firms are looking for engineers with 3-5 years experience or more, and they can get that with no problem, because the job market is dry, and there's bunches of out-of-work engineers. Get on the internship->graduate->permanent job track, or you've wasted 4 years of your life. Or you could do like me, goof off, drive home to visit your girlfriend every weekend, burn out acedemically at about junior year, and be SOL. For the record, I am in a dead-end sysadmin job right now, and being laid off at the end of the month. Woo-hoo!!!
It is not completely illegal. If those metrics (especially gender) are part of the job description, then it is perfectly legal. For example, look at Hooter's. If I am a better waiter, quicker and more accurate with orders, and more fun to talk to than the 20-year-old coed with the nice rack, do you think that I could (or even should) be a Hooter's waiter. No, because the job description specifically demands a female.
There are age requirements for U.S. Congressmen, Senators, and even the President (although being elected officials, I'm not sure if this applies to the arguement.)
You are right in this case, however, that hiring SOLELY based on gender is illegal. So it just comes down to making up an official excuse as to why the equally-qualified males don't get hired.
The preferred scientific name for that little molecule of happiness is actually "Hydrogen Hydroxide". Most branches of chemistry classify it as such because one of the hydrogens connected to the oxygen forms a "hydroxide" group that you can find in many MANY other basic molecules.
Re:There's also O'Reilly's free Using Samba online
on
Samba 3 By Example
·
· Score: 1
How is this any different than normal. Walk through a crowded bar and you'll hear a hundred different conversations about a hundred different things happening in the space of a small mobile home. Same location, different worlds. Same thing's true on the highway at rush hour (although I will admit that cell phones are my biggest gripe here). Or if you remember large lecture halls at college, the guy beside you is relating the story that starts "last night I was so blitzed" to the guy beside him, the guy behind you is busy trying to get the girl next to him's phone number, and the guy on the other side of you is fast asleep while you are half-heartedly taking notes.
I'm of the opinion that there is only so much "awareness" that the human mind can take. After that, we block out the less important stuff. (Ever notice how your eyes are instinctively drawn to moving things? Wonder why the cursor on your screen blinks?) This is my problem with combination cell phones/pdas/gps recievers/web browsers/walkie-talkies/remote controls/... There's only a certain amount of information I can process at once. See, I've got the mobile web on my cell-phone-with-integrated-camera. Do I really need a wi-fi enabled, Belverde-equipped, camera-included PDA also? Do I need them to be able to talk to each-other, and my toaster for that matter. IMHO, quite a bit of this is just an engineering exercise.
A little more offtopic. I saw somewhere that the fastest Belverde chip is said to be 625MHz. This is a chip designed for PDA's and smartphones. I am typing this currently sitting next to my company's 500MHz _production_ web server and VPN'ed in to my personal 600MHz web/file server + VPN endpoint. Why does a cellphone need that much friggin' power? (That's more a moral/ethical question than a engineering/design question, and the answer is that they can sell it, because 625 is bigger than 400 -- current Xscale speeds).
You can't sell the Brooklyn Bridge! My cousin owns it. Bought it from a nice chap back in '95 when he visited Manhattan. Title company is taking forever to get him the deed, though...
Did you take time to watch the video of the interview? The guy has obviously run scams before, but knows nothing about computers. He must think it's going to work because he got t-shirts, business cards, and liscence plate covers printed...
Ok, allow me to clarify. First of all, I didn't mean to make it sound like getting linux compatibility was hard, it is terribly easy. The "correct libraries" comment refers to the difficulty I had with getting Opera (circa '99) to build from sources (ya know, before it was available as a port/package). There were two libraries missing, can't remember which ones. One I found quite easily, the other I couldn't. Later that box was scrapped, and I never used X on BSD again. so no need to try opera. Many of the ports/packages themselves use linux compatibility and work as flawlessly as they would on Linux (I'm resonably sure).
Now, my comment about X on BSD may have been a touch unfounded. Getting X to physically work was mostly painless (most of my experience was pre FBSD4.0). Part of my frustration could have been in not choosing a good window manager, or that the machine was a touch slow (486 DX/66 w/32 MB RAM), but I found that I had to do a whole lot of editing of my WM's config file to get the interface to look like I wanted it. Later, when I tried Mandrake (new laptop, wanted to try something different, was at work and buddy had mandrake install disks) X looked so pretty, and I liked it. I only had to change one config file to get my screen to display at 1280x800 (widescreen laptop) and it was set up very close to how I liked it. Now, it's possible that 4 years of development has something to do with it, or that I was just in a good mood that day, or that I should have tried KDE on the BSD box, or that I'm a stupid noob; I'm fine with all of those things, but the guy wanted opinions, and I gave them.
I would never say that one is better than the other (especially in Slashdot forums) because doing so would get me tarred-and-feathered by one group or another.
On obsolete hardware, I've found that FreeBSD, anyway, is much easier than Linux. FreeBSD is perfectly happy installing off of a floppy to any 486 or better with a network connection. Most Linux installs I've tried either don't have the option to go from floppy, or want a large amout of RAM (for the hardware in question) to install. My worst installation experience was Mandrake 9.1 on a P150 laptop with no cd drive and a non-standard cardbus controller that refuses to work with the yenta-socket driver. If BSD would have recognized the PCMCIA NIC, it would have been wham, bam, thank-you-ma'am. IIRC, even if Mandrake would have seen the NIC, I would still have to export the install CD's from a NFS server somewhere, because they don't have a central FTP site to use. As it was, I ended up pulling the hard drive out and loading the install stuff on it from another computer, then booting an install disk and using the low-mem configuration (I only have 48 meg) to install from the hard drive.
Having used FreeBSD since 1999 or so for one reason or another, and having looked at several Linux distros, I'll throw in my 2cents.
FreeBSD looks like UNIX (oversimplification, albeit) down-and-dirty. I ran X on it for awhile (enlightenment or fvwm95 on a 486DX/66) and will never again. It really is not set up for a GUI, and you will do a ton of work getting it there. It will run Linux-compatible binaries provided you have the right libraries.
This is what I would use as the server because I am comfortable with it, and feel it is faster and more secure in this capacity. (I have little proof of the proceeding statement, but know there are thousands of benchmarks that prove me either right or wrong)
Linux distros I've tried range from pretty and trendy (Mandrake, Knoppix), to Windows clones (Lin---s, Licoris), to down-and-dirty UNIX type (Slackware). Mandrake would be my choice for the laptop and the development box because I just like the way it feels.
The FreeBSD live CD doesn't seem like anything more than an educational tool, because, IMHO, FreeBSD is supposed to be installed, customized, and left to what it does best, run server daemons. Check out linuxISO.org if you haven't already. It is a quick resource for information about a TON of different distros.
The same theory applies, but these are 9'x9' semi-transparent screens each with it's own projecter. The hardware is extremely impressive, and, when I was a student at the University of Illinois, although I never got to write anything for it, I did get to play a port of Quake II that someone had (unoficially) done for it.
From the article -- "The primary computer, named Cassatt... is a twelve-processor Silicon Graphics Onyx 2 Reality Monster."
There's one in a research lab in Tokyo that has 5 sides (4 walls and the floor). This one has only 4 sides (3 walls and the floor). A fully-immersive (6 sides) environment was in the works at the above-mentioned Japanese lab, but I haven't been looking into it much since about 2000.
I had a desktop unit that would randomly reboot when printing, or scanning (both USB). Or just whenever (loading webpage). Thing would also not reset (soft reboot, power not cut to the machine). After you reset it, it would just hang during POST. could be anything. Turns out the CPU (Athlon XP 1700+) was flaking out. We got a new mobo (originally thought it was this) and the chip seems to run OK on that with a 100MHz FSB and a multiplier of 10 (supposed to run at 133x11). if this is your problem, I've heard they're doing amazing things in new laptops these days. Maybe it's time for an *ahem* upgrade...
I worked at a small repair shop in the back of a store that sold used computers and the like. The latest version of TS(troubleshooter) that we used was 6.54. We only really used it to "test" machines that customers brought in to sell to us, and only so we had some paperwork to show the owner
The consensus in the store was that we just replace components when something wasn't working. We had a stock of "known good" parts, and swapped out till we discovered who wasn't playing nice. We also had a test machine that was "known good" to test components it. 99 times out of 100 (at least in home/home office grade hardware) replacing a defective (or even just flaky) component was the way to go. Then, if the faulty component is under warranty, we use the vendor-supplied diagnostics (mostly for hard drives) to prove the part was bad, and get it covered under warranty.
Excuse the rambling. I think the tools you use are going to depend on what the final outcome needs to be. (do you just replace the mobo if you know 2 of the PCI slots are bad, or do you need to try to fix them). Nothing beats good-old-fashioned-know-how. Get a good POST code card (I've heard a bit about a product called "PC Geiger" which contains a diagnostic PCI card and a readout panel). In my experience, most software based tools (with exception of a few, please don't flame me) are crap and just tell you what you already know. Just last week I had a SiSoft Sandra burn-in run just fine on a faulty Athlon XP 1700+.
He probably meant Megabit/s (Mb/s) which would mean 300/8=37.5 Megabyte/s (MB/s). Or, he could be thinking that the "150" on the marketing spec SATA150 referrs to 150 Mb/s, and doubled it because (theoretically, in a very simplified world) RAID 0 doubles the throughput of each separate disk when added together, and our friend here happens to have 2 disks. Either way, his ignorance is still showing.
Consequently, the 150 on the SATA spec really means nothing. It comes from the fact that we had ATA, then ATA66, which was theoretically twice as fast. Correct me if I'm wrong, but the "66" referrs to the speed of the controller's electronics in MHz. Then along came ATA100, then ATA133 (which wasn't necessary, because ATA100 provided more bandwidth than disks can fill.) To look better to Joe Q. User, SATA had to have a bigger number, hence SATA150. Because, DUH, 150 is faster than 133!
This is actually a very good idea. Seeing that, in the most generic case, installing software on most any open OS involves:
1.) Download source.
1a.) (not specifically necessary) verify it is correct and from a trusted party (PHP sig, whatnot)
2.) Apply disto-specific patches. (Yes, I am aware that BSD is not a Linux distro, but for this arguement, it fits the model)
3.) compile (using correct compiler switches for specific distro)
4.) install (again, there are some distro-specific things that go here)
All other installation methods (package systems) Just try to make these steps simpler, or transparent to the end user. To follow these steps automatically, either the software developers, or maintaners of the installer program, would have to cobble a distro-specific script together for all distros (that they want to support) for each OSS project out there. This seems like a herculean task, but quite a bit of this information is out there; the problem is there needs to be a centralized repository for it. I propose partnering with a site that archives this information already, and is widely used by the community regardless of the distro they have (I am a BSD geek, so I don't know, but sourceforge seems to fit this bill). Now, the OSS writers will "release" a packaged version of their software to this centralized distribution system by TARring up the source, and supplying a patch and install script to the system for any distro they care to support. If your (as a user) distro is not supported, you are VERY encouraged to supply the patch and install scripts (usually fairly trivial to write) to the system. The client part would then log in to the central server, and through the magic of PHP, be presented with a distro-specific download page where a single click will d/l the source, appropriate patches, scripts, etc. and install. The script that coordinates all the d/l and install would not be distro-specific, because all distros install software (at the most generic level) the same way. You could even encorporate a bittorrent like d/l mechanism to speed up the d/l of more popular apps. (security through encryption comes in here).
Back in 2001, Helvetica Bold Oblique was the big winner of the 73rd annual "Fonty Awards". Times New Roman 14 didn't even get an honorable mention! Look at how far a font can come in a mere 2.5 years!
A lot of the time I will go over to a friend's house and fix her computer, she'd have a bunch of her friends there, after I fixed it up we'd all party and hop in the hot tub.
You suck. I'll go back to my pathetic little life now....
First off, great site. Hillarious! There is one problem, however...
This theory, as with all theories about female behavior that cannot be backed by phisiological proof, are fundamentally flawed because they attempt to constrain female behavior to rigid, logically-derived boundaries. Those of us that have been on the "Friends" ladder more times than not know all too well that the female mind seems to function on a kind of anti-logic. That's how arguements with your GF that you seem to be winning, or know for a fact you can win, turn on you, and you end up losing. Because the female way of thinking doesn't require the boundaries of logic that men's minds seem to require.
Fixing a girls computer, or anything else for that matter, can get you on the "good" ladder, but you just have to play it right. Think of it as being at work. Use the same rules you would use when trying to bag the cute new intern up in HR, and you'll be fine. (Of course I say this with no successful experience.)
Slashdot Mirror
Offtopic, I know, but the line in your sig is actually a combination of two lines in that particular number:
A giddy little thrill at a reasonable price.
It's that little extra spice that makes existance extra nice.
Theatrically, most of The Simpsons' musical numbers are incredible. You could almost get a cast together and perform them on broadway.
I remember when I was at the University of Illinois in Urbana, someone built a robotic arm that used a webcam and some image processing algorythms to pick up a pop can that it had no other idea where was.
BTW, the main push of the project was that the robot used a learning algorythm to determine how to pick up the can. Basically all that was programmed was a way to interpret the images coming from the webcam, and a "desire" to pick up the can. This caused some interesting effects like when the arm would pass out of view of the camera, the robot "learned" to wave the arm at the camera to re-locate itself.
My plan would involve a computer, with a WiFi link to the hardware on the mower (possibly a micro-ATX or scavenged laptop, something with a serial port to communicate with a microcontroller). Connect the computer to a webcam that can see the entire yard. Roll-your-own image processing algorythms, and there you go. Plus, for safety reasons, you could have the thing shut off when the webcam recognizes a fast moving object (neighbor's brat kid that can't keep the football in his own yard). Or, if you think you can mount enough processing power on the mower itself, you could do away with the need for the WiFi link. This would definately require the webcam to be wireless, then.
AMEN BROTHER!!!
I currently have to support a terminal emulation package that requires anyone using it to have local administrator priviledges.
Or, configure Linux box as a router. Plug in a second NIC. I used to put FreeBSD on 486's and early Pentium systems that I found in the garbage to be routers to babysit windows machines online. I've heard it's even easier to do in Linux, but have never tried.
If ECE stands for Electrical and Computer Engineering, you'd better plan on graduating fairly high in your class. I got a EE degree from a top 10 school (University of Illinois), but my grades weren't that great, so I didn't get any internships or job leads straight out of school. After graduation, I found that most any job that you'd want as an EE you would have had to basically be hired right out of college (think job fairs). On the "open" market, firms are looking for engineers with 3-5 years experience or more, and they can get that with no problem, because the job market is dry, and there's bunches of out-of-work engineers. Get on the internship->graduate->permanent job track, or you've wasted 4 years of your life. Or you could do like me, goof off, drive home to visit your girlfriend every weekend, burn out acedemically at about junior year, and be SOL. For the record, I am in a dead-end sysadmin job right now, and being laid off at the end of the month. Woo-hoo!!!
I've had 5.2 running smooth for quite awhile now. Don't get me wrong, I loved the 4.X branch just as much as the next guy...
It is not completely illegal. If those metrics (especially gender) are part of the job description, then it is perfectly legal. For example, look at Hooter's. If I am a better waiter, quicker and more accurate with orders, and more fun to talk to than the 20-year-old coed with the nice rack, do you think that I could (or even should) be a Hooter's waiter. No, because the job description specifically demands a female.
There are age requirements for U.S. Congressmen, Senators, and even the President (although being elected officials, I'm not sure if this applies to the arguement.)
You are right in this case, however, that hiring SOLELY based on gender is illegal. So it just comes down to making up an official excuse as to why the equally-qualified males don't get hired.
The preferred scientific name for that little molecule of happiness is actually "Hydrogen Hydroxide". Most branches of chemistry classify it as such because one of the hydrogens connected to the oxygen forms a "hydroxide" group that you can find in many MANY other basic molecules.
Amen buddy! If I only had mod points...
How is this any different than normal. Walk through a crowded bar and you'll hear a hundred different conversations about a hundred different things happening in the space of a small mobile home. Same location, different worlds. Same thing's true on the highway at rush hour (although I will admit that cell phones are my biggest gripe here). Or if you remember large lecture halls at college, the guy beside you is relating the story that starts "last night I was so blitzed" to the guy beside him, the guy behind you is busy trying to get the girl next to him's phone number, and the guy on the other side of you is fast asleep while you are half-heartedly taking notes.
I'm of the opinion that there is only so much "awareness" that the human mind can take. After that, we block out the less important stuff. (Ever notice how your eyes are instinctively drawn to moving things? Wonder why the cursor on your screen blinks?) This is my problem with combination cell phones/pdas/gps recievers/web browsers/walkie-talkies/remote controls/... There's only a certain amount of information I can process at once. See, I've got the mobile web on my cell-phone-with-integrated-camera. Do I really need a wi-fi enabled, Belverde-equipped, camera-included PDA also? Do I need them to be able to talk to each-other, and my toaster for that matter. IMHO, quite a bit of this is just an engineering exercise.
A little more offtopic. I saw somewhere that the fastest Belverde chip is said to be 625MHz. This is a chip designed for PDA's and smartphones. I am typing this currently sitting next to my company's 500MHz _production_ web server and VPN'ed in to my personal 600MHz web/file server + VPN endpoint. Why does a cellphone need that much friggin' power? (That's more a moral/ethical question than a engineering/design question, and the answer is that they can sell it, because 625 is bigger than 400 -- current Xscale speeds).
You can't sell the Brooklyn Bridge! My cousin owns it. Bought it from a nice chap back in '95 when he visited Manhattan. Title company is taking forever to get him the deed, though...
Did you take time to watch the video of the interview? The guy has obviously run scams before, but knows nothing about computers. He must think it's going to work because he got t-shirts, business cards, and liscence plate covers printed...
Ok, allow me to clarify. First of all, I didn't mean to make it sound like getting linux compatibility was hard, it is terribly easy. The "correct libraries" comment refers to the difficulty I had with getting Opera (circa '99) to build from sources (ya know, before it was available as a port/package). There were two libraries missing, can't remember which ones. One I found quite easily, the other I couldn't. Later that box was scrapped, and I never used X on BSD again. so no need to try opera. Many of the ports/packages themselves use linux compatibility and work as flawlessly as they would on Linux (I'm resonably sure).
Now, my comment about X on BSD may have been a touch unfounded. Getting X to physically work was mostly painless (most of my experience was pre FBSD4.0). Part of my frustration could have been in not choosing a good window manager, or that the machine was a touch slow (486 DX/66 w/32 MB RAM), but I found that I had to do a whole lot of editing of my WM's config file to get the interface to look like I wanted it. Later, when I tried Mandrake (new laptop, wanted to try something different, was at work and buddy had mandrake install disks) X looked so pretty, and I liked it. I only had to change one config file to get my screen to display at 1280x800 (widescreen laptop) and it was set up very close to how I liked it. Now, it's possible that 4 years of development has something to do with it, or that I was just in a good mood that day, or that I should have tried KDE on the BSD box, or that I'm a stupid noob; I'm fine with all of those things, but the guy wanted opinions, and I gave them.
I would never say that one is better than the other (especially in Slashdot forums) because doing so would get me tarred-and-feathered by one group or another.
On obsolete hardware, I've found that FreeBSD, anyway, is much easier than Linux. FreeBSD is perfectly happy installing off of a floppy to any 486 or better with a network connection. Most Linux installs I've tried either don't have the option to go from floppy, or want a large amout of RAM (for the hardware in question) to install. My worst installation experience was Mandrake 9.1 on a P150 laptop with no cd drive and a non-standard cardbus controller that refuses to work with the yenta-socket driver. If BSD would have recognized the PCMCIA NIC, it would have been wham, bam, thank-you-ma'am. IIRC, even if Mandrake would have seen the NIC, I would still have to export the install CD's from a NFS server somewhere, because they don't have a central FTP site to use. As it was, I ended up pulling the hard drive out and loading the install stuff on it from another computer, then booting an install disk and using the low-mem configuration (I only have 48 meg) to install from the hard drive.
Having used FreeBSD since 1999 or so for one reason or another, and having looked at several Linux distros, I'll throw in my 2cents.
FreeBSD looks like UNIX (oversimplification, albeit) down-and-dirty. I ran X on it for awhile (enlightenment or fvwm95 on a 486DX/66) and will never again. It really is not set up for a GUI, and you will do a ton of work getting it there. It will run Linux-compatible binaries provided you have the right libraries.
This is what I would use as the server because I am comfortable with it, and feel it is faster and more secure in this capacity. (I have little proof of the proceeding statement, but know there are thousands of benchmarks that prove me either right or wrong)
Linux distros I've tried range from pretty and trendy (Mandrake, Knoppix), to Windows clones (Lin---s, Licoris), to down-and-dirty UNIX type (Slackware). Mandrake would be my choice for the laptop and the development box because I just like the way it feels.
The FreeBSD live CD doesn't seem like anything more than an educational tool, because, IMHO, FreeBSD is supposed to be installed, customized, and left to what it does best, run server daemons. Check out linuxISO.org if you haven't already. It is a quick resource for information about a TON of different distros.
Ahh.. That would be what John Connor used to "hack" the ATM and the CyberDyne security system in Terminator 2!
Ever heard of the CAVE?
... is a twelve-processor Silicon Graphics Onyx 2 Reality Monster."
The same theory applies, but these are 9'x9' semi-transparent screens each with it's own projecter. The hardware is extremely impressive, and, when I was a student at the University of Illinois, although I never got to write anything for it, I did get to play a port of Quake II that someone had (unoficially) done for it.
From the article -- "The primary computer, named Cassatt
There's one in a research lab in Tokyo that has 5 sides (4 walls and the floor). This one has only 4 sides (3 walls and the floor). A fully-immersive (6 sides) environment was in the works at the above-mentioned Japanese lab, but I haven't been looking into it much since about 2000.
I had a desktop unit that would randomly reboot when printing, or scanning (both USB). Or just whenever (loading webpage). Thing would also not reset (soft reboot, power not cut to the machine). After you reset it, it would just hang during POST. could be anything. Turns out the CPU (Athlon XP 1700+) was flaking out. We got a new mobo (originally thought it was this) and the chip seems to run OK on that with a 100MHz FSB and a multiplier of 10 (supposed to run at 133x11). if this is your problem, I've heard they're doing amazing things in new laptops these days. Maybe it's time for an *ahem* upgrade...
I worked at a small repair shop in the back of a store that sold used computers and the like. The latest version of TS(troubleshooter) that we used was 6.54. We only really used it to "test" machines that customers brought in to sell to us, and only so we had some paperwork to show the owner
The consensus in the store was that we just replace components when something wasn't working. We had a stock of "known good" parts, and swapped out till we discovered who wasn't playing nice. We also had a test machine that was "known good" to test components it. 99 times out of 100 (at least in home/home office grade hardware) replacing a defective (or even just flaky) component was the way to go. Then, if the faulty component is under warranty, we use the vendor-supplied diagnostics (mostly for hard drives) to prove the part was bad, and get it covered under warranty.
Excuse the rambling. I think the tools you use are going to depend on what the final outcome needs to be. (do you just replace the mobo if you know 2 of the PCI slots are bad, or do you need to try to fix them). Nothing beats good-old-fashioned-know-how. Get a good POST code card (I've heard a bit about a product called "PC Geiger" which contains a diagnostic PCI card and a readout panel). In my experience, most software based tools (with exception of a few, please don't flame me) are crap and just tell you what you already know. Just last week I had a SiSoft Sandra burn-in run just fine on a faulty Athlon XP 1700+.
As always, your milage may vary
They pulled the damn picture!!! Bastards!!!
it was a pic of MS Flight Simulator set up to look like it was displaying on 15 or so monitors. Google didn't cache it either.
anything less than this just isn't worth doing
He probably meant Megabit/s (Mb/s) which would mean 300/8=37.5 Megabyte/s (MB/s). Or, he could be thinking that the "150" on the marketing spec SATA150 referrs to 150 Mb/s, and doubled it because (theoretically, in a very simplified world) RAID 0 doubles the throughput of each separate disk when added together, and our friend here happens to have 2 disks. Either way, his ignorance is still showing.
Consequently, the 150 on the SATA spec really means nothing. It comes from the fact that we had ATA, then ATA66, which was theoretically twice as fast. Correct me if I'm wrong, but the "66" referrs to the speed of the controller's electronics in MHz. Then along came ATA100, then ATA133 (which wasn't necessary, because ATA100 provided more bandwidth than disks can fill.) To look better to Joe Q. User, SATA had to have a bigger number, hence SATA150. Because, DUH, 150 is faster than 133!
This is actually a very good idea. Seeing that, in the most generic case, installing software on most any open OS involves:
1.) Download source.
1a.) (not specifically necessary) verify it is correct and from a trusted party (PHP sig, whatnot)
2.) Apply disto-specific patches. (Yes, I am aware that BSD is not a Linux distro, but for this arguement, it fits the model)
3.) compile (using correct compiler switches for specific distro)
4.) install (again, there are some distro-specific things that go here)
All other installation methods (package systems) Just try to make these steps simpler, or transparent to the end user. To follow these steps automatically, either the software developers, or maintaners of the installer program, would have to cobble a distro-specific script together for all distros (that they want to support) for each OSS project out there. This seems like a herculean task, but quite a bit of this information is out there; the problem is there needs to be a centralized repository for it. I propose partnering with a site that archives this information already, and is widely used by the community regardless of the distro they have (I am a BSD geek, so I don't know, but sourceforge seems to fit this bill). Now, the OSS writers will "release" a packaged version of their software to this centralized distribution system by TARring up the source, and supplying a patch and install script to the system for any distro they care to support. If your (as a user) distro is not supported, you are VERY encouraged to supply the patch and install scripts (usually fairly trivial to write) to the system. The client part would then log in to the central server, and through the magic of PHP, be presented with a distro-specific download page where a single click will d/l the source, appropriate patches, scripts, etc. and install. The script that coordinates all the d/l and install would not be distro-specific, because all distros install software (at the most generic level) the same way. You could even encorporate a bittorrent like d/l mechanism to speed up the d/l of more popular apps. (security through encryption comes in here).
Back in 2001, Helvetica Bold Oblique was the big winner of the 73rd annual "Fonty Awards". Times New Roman 14 didn't even get an honorable mention! Look at how far a font can come in a mere 2.5 years!
A lot of the time I will go over to a friend's house and fix her computer, she'd have a bunch of her friends there, after I fixed it up we'd all party and hop in the hot tub.
You suck. I'll go back to my pathetic little life now....
First off, great site. Hillarious! There is one problem, however...
This theory, as with all theories about female behavior that cannot be backed by phisiological proof, are fundamentally flawed because they attempt to constrain female behavior to rigid, logically-derived boundaries. Those of us that have been on the "Friends" ladder more times than not know all too well that the female mind seems to function on a kind of anti-logic. That's how arguements with your GF that you seem to be winning, or know for a fact you can win, turn on you, and you end up losing. Because the female way of thinking doesn't require the boundaries of logic that men's minds seem to require.
Fixing a girls computer, or anything else for that matter, can get you on the "good" ladder, but you just have to play it right. Think of it as being at work. Use the same rules you would use when trying to bag the cute new intern up in HR, and you'll be fine. (Of course I say this with no successful experience.)