As somebody who presently lives 3 blocks from this, let me tell you, this "art" project is TACKY. It would work in Las Vegas, but the Bay Bridge is too dignified for this sort of crap. Can't wait for it to be removed.
Read the next line: "Compared to abstainers, those who drank 3 cups of caffeinated coffee daily were at increased risk of EG/EGS (RR = 1.66; 95% CI, 1.09–2.54; P trend = 0.02)."
Still, given that there were 120,000 data points in this study, and they don't seem to be correcting for multiple comparisons, I would think a P of 0.02 is pretty weak.
Spoken by someone who clearly doesn't use RPN. The speed advantage of RPN doesn't come from the number of keystrokes needed (even though RPN does take fewer), but with the ability to enter the equation from left to right, start to finish, no matter how many embedded factors there are. While with non-RPN you must evaluate the equation to even know where to start entering and then keep track of where you are in your head.
Spoken by someone who clearly doesn't use modern non-RPN calculators. The speed advantage of conventional notation calculators comes from the ability to enter calculations froms start to finish exactly as they are written down on paper. No mental gymnastics required to translate what's written to what you need to punch into the calculator. And you can put the entire equation into one line, again, greatly simplying things.
Now, back in the day when calculators had only one line displays that could only display numbers, RPN was a great advantage. But in the modern age (meaning post ~1989 or so) calculators (like the TI-81) have multiple lines of display and can display the full set of alpha/numeric/symbols. So no need to keep track of anything in your head, it's entered in as written on the page, and displayed right there on the screen.
I try not to make such blanket statements on/., but I simply do not believe this. You may have had a friend at MIT, but there isn't any way a non-RPN calculator is as fast as an PRN calculator on any equation more complex than 2 + 2.
Well first off, I'll assume you're talking about a RPN calculator, and not a PRN calculator. Although a PRN (or pr0n) calculator would be very cool, I think it would make getting any real work done very difficult.
So why don't we go back in time together to the mid-90's, and you and I can hammer through some hard-core electrical engineering problem sets together. You can have an HP-48, I'll take a TI-81, and we'll see who wins at the 'ole plug-and-chug. What you probably haven't taken into account, is that you can't submit your problem set written out in RPN. So the time and errors that transpire in that back and forth translataion between the problem set and your calculator will more than compensate for what ever advantage in theory RPN may have.
RPN is argueably faster, as you don't need to enter in parenthesis. But you end up having to press the enter key a lot, so the advantage quickly evaporates.
A friend of mine at MIT had an HP-48, and I had a TI-81, we used to do a lot of engineering problem sets together and would often race on entering calculations. Averaged over time the competition was a draw. Although the HP-48 definitely wins from a "cool" factor perspective (where cool=geek).
Speaking of the TI-81, I bought mine in 1991 for $82, and I'm still using it every day.
I'm not sure how many extraneous libraries GnuCash 1.9 relies on... but if all the libraries already have windows ports, it's really easy. As an example, the last GTK2 app I ported to windows took about 30 minutes of time.
And the native MacOSX GTK2 port seems to be coming along as well.
If by "won the UNIX battle" you mean; to be the last, straggling, dying company whose business model requires people grasping on to their particular proprietary form of UNIX, then yes, Sun has won the UNIX battle.
You can't really blame Fedora for the NVidia versus rhgb problem. I stumbled into that one to, and it's just one of the hits you gotta take for running an external binary kernel module on your system.
I gotta assume the audio stuff will be fixed, it seemed like it was just conflicts between ALSA and gstreamer.
Yeah, this tech isn't anything new. I remember seeing a building in Japan with this technology on the show "That's Incredible" around 20 years ago. I always assumed it wasn't economically feasible and that's why I haven't seen it since.
Re:Latest Fedora-development has gcc 4.0
on
GCC 4.0 Preview
·
· Score: 1
No, this time they're giving you beta packages of gcc so that you can test things out if you want. The packages are labelled gcc4, and the installed binary is gcc4, so there's no chance you'll compile things with the beta compiler unless you explicitely want to. The default compiler is still gcc 3.4.
What created the gcc 2.96 fiasco was that RedHat made it the default compiler on the system and expected everybody to use it, even though it wasn't ready for prime time yet (and wasn't even an official gcc release). In fact, they even provided a kgcc package (which was just gcc-2.95) because the kernel couldn't be compiled with their 2.96. As a developer who uses RedHat, it was a total nightmare. And if I remember right, Mandrake did the same thing.
If you look through configure scripts, you'll still find programs that check for 2.96, and then either refuse to compile or turn off all optomizations.
I know of absolutely nobody who runs KDE, almost everyone I know runs Fedora Core and Gnome.
For most people, it's simply a matter of what they started using first. That being said, I started with KDE, and switched over to GNOME years ago for the cleaner and more simple interface.
I've been a Redhat user since the 3.0 days, and have generally been very happy with the product. I've been mainly a Linux desktop user, although I occasionally run samba or httpd on some of my boxes.
When they announced the whole Fedora thing, I was serious considering jumping ship for two reasons. First, I didn't want to have to upgrade my machines every half year. Second, I was worried a huge portion of the user base was also jumping ship, and I didn't want to be the last one off the boat.
Anyway, I considered Debian for a really long time, but just couldn't stomach being that obsolete. So I just stuck with Fedora instead. Things haven't been bad. The release cycle is more like 9 months or even longer, so I'm upgrading only slight more often than I was with the old Redhat release cycle. And the user base doesn't seem to have vanished (just installed Fedora for PowerPC this weekend, cool).
In general, I've been really happy with Fedora. Would I recommend it to others? Sometimes. The main caveat is that you'll have to upgrade your box yearly. Yeah you can get RHES, but nobody's going to pay that much for their home desktop operating system, and most users don't need the support, they only want the security patches.
I think Redhat could use an offering for the home or casual user who is savy enough to not need technical support, but likes to upgrade on a 3 year cycle.
As somebody who was at MIT, and knows many of the people involved in the biological engineering department, I'm surprised that MIT is considering adding BE as an undergrad major.
I'm currently doing graduate school in a BE department, and I'm very glad that my undergrad major was not in BE. I was an EE major (6-1 for you MIT folks) in undergrad with a interest in BE, which as the previous poster mentioned, is the right way to do things.
Most of the BE majors I've seen in my graduate school life are simply not adequately prepared when they get to graduate school. Obviously, there's exceptions, but for the most part somebody with a BE undergrad major has only had a superficial survey of the different fields of engineering and is essentially a glorified biologist.
The very reason MIT didn't try to create a BE undergrad major back in 1998 when the department was formed, is that they were concerned that their graduates would not have a solid foundation in any of the branches of engineering, and would therefore suffer after graduation.
To any undergrads considering BE as a major, I would suggest that you instead pick the branch of engineering that most interests you (e.g. EE, ME, ChemE) and then add a BE minor. Biology is a lot easier to learn yourself after graduation than is engineering, and particularly math.
"You can't do that with Linux either. The name aside, the Linux you run on one arch is not the Linux on another arch."
What the heck are you talking about? Of the architectures I mentioned, Debian Linux is available for every one of them except for POWER.
Well, maybe you don't like Debian. Okay. Fedora Linux (or distributions based directly off of it like Yellowdog and RHEL) gives you support for all of the architectures I've mentioned, including POWER.
And with Debian (and Fedora), you never loose the source-based flexibility. If you've got time to burn, you're welcome to recompile all the code from the source packages.
Anyway, the point that I was trying to make, is that on a modern polyarchitectural network, you're better of using Linux because it actually has support for all the modern architectures.
Well, if you're running Linux, you can hop between your 486, athlon xp, sparcbook, and g4 as well. Plus, you can also hop onto your Intel Itanium cluster, your not yet antiquated HP PA-Risc workstation, and your IBM Power5 server. You can't do that with NetBSD.
The grandparent post does have an element of truth to it.
There's a fair amount of hardware that NetBSD supports, but by "support" they mean you can hook a dumb terminal to the serial port, and boot off of ethernet. I personally consider "supported" to mean that the machine can access and boot of its own harddrive, can use and display more then just text on the graphics card, and can use its own keyboard.
As examples, check out the amount of hardware support offered for the vax or hppa architectures. It can be really spotty.
Okay, I'll give them peer to peer file swapping. That's pretty insightful.
By 1990/1991, MPG encoding had already been developed, along with all the technology behind HDTV. These researchers where already connected to ethernet in the lab (with a fiber optic backbone going around MIT). So the paper you're refering to was simply extrapolating the technologies that were already achievable in the lab for home use.
Why did it take a decade? Moore's law had to catch up. A workstation (and hard drive) to do DVR in 1990/1991 would of cost at least $20,000 (in 1990 dollars). A Tivo a decade later cost less than 1% of that (inflation adjusted).
I went to MIT in the mid 90's, and the real Media Lab was considered little more than a dog and pony show for corporate sponsorship. It was actually sort of an embarrassment to those doing real engineering and science at MIT, but it was tolerated because it brought in its own funding.
I knew people who worked at the Media lab, and I knew about a lot of research projects there. Absolutely none of them were advancing the fields of science or engineering (although I gotta imagine there are a couple of counter examples that I just don't know about). Their highest achievement was the development of "Lego Mindstorms", which is cool for sure, but is work that could easily be done in a purely corporate setting.
I gather, in Europe at least, corporations weren't as easiler suckered into shelling out money for research on gadgets and toys, especially when they could do the same research in house for less.
Okay, that's great, some paper out of the Media lab sucessfully extrapolated the obvious.
It's sort of like if I had written a paper in '90 or '91 predicting that processors would be running at over 1 GHz in 2000.
But imagine if that paper from the media lab had never been written, what sort of world would be living in right now? That's right, the same exact world. And processors are running at over 1 GHz even though I didn't write my famous paper in '90 or '91.
There's a big difference between spending your time predicting the future, and making it happen.
For crying out load, can't people read up on this?
McDonald's was knowingly superheating their coffee so that a brewed pot would stay fresher longer. They were heating their coffe to 180-190 degrees F. Your pot of coffee at home is more like 140 degrees. Even coffee at 155 degrees wouldn't give you serious burns.
Slashdot Mirror
As somebody who presently lives 3 blocks from this, let me tell you, this "art" project is TACKY. It would work in Las Vegas, but the Bay Bridge is too dignified for this sort of crap. Can't wait for it to be removed.
Read the next line: "Compared to abstainers, those who drank 3 cups of caffeinated coffee daily were at increased risk of EG/EGS (RR = 1.66; 95% CI, 1.09–2.54; P trend = 0.02)."
Still, given that there were 120,000 data points in this study, and they don't seem to be correcting for multiple comparisons, I would think a P of 0.02 is pretty weak.
And while the 3G may be listed as supported on iOS 4.2.1, it's so slow running this OS that realistically it's not supported...
Spoken by someone who clearly doesn't use RPN. The speed advantage of RPN doesn't come from the number of keystrokes needed (even though RPN does take fewer), but with the ability to enter the equation from left to right, start to finish, no matter how many embedded factors there are. While with non-RPN you must evaluate the equation to even know where to start entering and then keep track of where you are in your head.
/., but I simply do not believe this. You may have had a friend at MIT, but there isn't any way a non-RPN calculator is as fast as an PRN calculator on any equation more complex than 2 + 2.
Spoken by someone who clearly doesn't use modern non-RPN calculators. The speed advantage of conventional notation calculators comes from the ability to enter calculations froms start to finish exactly as they are written down on paper. No mental gymnastics required to translate what's written to what you need to punch into the calculator. And you can put the entire equation into one line, again, greatly simplying things.
Now, back in the day when calculators had only one line displays that could only display numbers, RPN was a great advantage. But in the modern age (meaning post ~1989 or so) calculators (like the TI-81) have multiple lines of display and can display the full set of alpha/numeric/symbols. So no need to keep track of anything in your head, it's entered in as written on the page, and displayed right there on the screen.
I try not to make such blanket statements on
Well first off, I'll assume you're talking about a RPN calculator, and not a PRN calculator. Although a PRN (or pr0n) calculator would be very cool, I think it would make getting any real work done very difficult.
So why don't we go back in time together to the mid-90's, and you and I can hammer through some hard-core electrical engineering problem sets together. You can have an HP-48, I'll take a TI-81, and we'll see who wins at the 'ole plug-and-chug. What you probably haven't taken into account, is that you can't submit your problem set written out in RPN. So the time and errors that transpire in that back and forth translataion between the problem set and your calculator will more than compensate for what ever advantage in theory RPN may have.
RPN is argueably faster, as you don't need to enter in parenthesis. But you end up having to press the enter key a lot, so the advantage quickly evaporates.
A friend of mine at MIT had an HP-48, and I had a TI-81, we used to do a lot of engineering problem sets together and would often race on entering calculations. Averaged over time the competition was a draw. Although the HP-48 definitely wins from a "cool" factor perspective (where cool=geek).
Speaking of the TI-81, I bought mine in 1991 for $82, and I'm still using it every day.
Anyone have a mirror of the .torrent file? Look like even that's been slashdotted...
Well, GnuCash is a GTK app, not a GNOME app.
0 /win32/dependencies/
But all of GTK2 and most of GNOME has already been ported to windows anyway. You can get prebuilt binaries for most of the libraries at: ftp://ftp.gnome.org/pub/gnome/platform/2.12/2.12.
I'm not sure how many extraneous libraries GnuCash 1.9 relies on... but if all the libraries already have windows ports, it's really easy. As an example, the last GTK2 app I ported to windows took about 30 minutes of time.
And the native MacOSX GTK2 port seems to be coming along as well.
You guys are such light weights.
I keep all my music as FLAC compressed files (lossless). So you see, my music collection is much much bigger then yours.
160kbps? You pansy. Try 750kbps.
If by "won the UNIX battle" you mean; to be the last, straggling, dying company whose business model requires people grasping on to their particular proprietary form of UNIX, then yes, Sun has won the UNIX battle.
You can't really blame Fedora for the NVidia versus rhgb problem. I stumbled into that one to, and it's just one of the hits you gotta take for running an external binary kernel module on your system.
I gotta assume the audio stuff will be fixed, it seemed like it was just conflicts between ALSA and gstreamer.
Yeah, this tech isn't anything new. I remember seeing a building in Japan with this technology on the show "That's Incredible" around 20 years ago. I always assumed it wasn't economically feasible and that's why I haven't seen it since.
No, this time they're giving you beta packages of gcc so that you can test things out if you want. The packages are labelled gcc4, and the installed binary is gcc4, so there's no chance you'll compile things with the beta compiler unless you explicitely want to. The default compiler is still gcc 3.4.
What created the gcc 2.96 fiasco was that RedHat made it the default compiler on the system and expected everybody to use it, even though it wasn't ready for prime time yet (and wasn't even an official gcc release). In fact, they even provided a kgcc package (which was just gcc-2.95) because the kernel couldn't be compiled with their 2.96. As a developer who uses RedHat, it was a total nightmare. And if I remember right, Mandrake did the same thing.
If you look through configure scripts, you'll still find programs that check for 2.96, and then either refuse to compile or turn off all optomizations.
Really? I find Mac OS X completely unusable on the desktop. It's all gloss and little substance.
Meanwhile, back at the point, I'm typing this from a shiny 12" Powerbook, running Fedora Core 3.
I know of absolutely nobody who runs KDE, almost everyone I know runs Fedora Core and Gnome.
For most people, it's simply a matter of what they started using first. That being said, I started with KDE, and switched over to GNOME years ago for the cleaner and more simple interface.
I've been a Redhat user since the 3.0 days, and have generally been very happy with the product. I've been mainly a Linux desktop user, although I occasionally run samba or httpd on some of my boxes.
When they announced the whole Fedora thing, I was serious considering jumping ship for two reasons. First, I didn't want to have to upgrade my machines every half year. Second, I was worried a huge portion of the user base was also jumping ship, and I didn't want to be the last one off the boat.
Anyway, I considered Debian for a really long time, but just couldn't stomach being that obsolete. So I just stuck with Fedora instead. Things haven't been bad. The release cycle is more like 9 months or even longer, so I'm upgrading only slight more often than I was with the old Redhat release cycle. And the user base doesn't seem to have vanished (just installed Fedora for PowerPC this weekend, cool).
In general, I've been really happy with Fedora. Would I recommend it to others? Sometimes. The main caveat is that you'll have to upgrade your box yearly. Yeah you can get RHES, but nobody's going to pay that much for their home desktop operating system, and most users don't need the support, they only want the security patches.
I think Redhat could use an offering for the home or casual user who is savy enough to not need technical support, but likes to upgrade on a 3 year cycle.
I've upgraded a server Redhat 8 -> FC1 -> FC2, all done remotely. I haven't seen this box in two years.
Yum's pretty cool.
As somebody who was at MIT, and knows many of the people involved in the biological engineering department, I'm surprised that MIT is considering adding BE as an undergrad major.
I'm currently doing graduate school in a BE department, and I'm very glad that my undergrad major was not in BE. I was an EE major (6-1 for you MIT folks) in undergrad with a interest in BE, which as the previous poster mentioned, is the right way to do things.
Most of the BE majors I've seen in my graduate school life are simply not adequately prepared when they get to graduate school. Obviously, there's exceptions, but for the most part somebody with a BE undergrad major has only had a superficial survey of the different fields of engineering and is essentially a glorified biologist.
The very reason MIT didn't try to create a BE undergrad major back in 1998 when the department was formed, is that they were concerned that their graduates would not have a solid foundation in any of the branches of engineering, and would therefore suffer after graduation.
To any undergrads considering BE as a major, I would suggest that you instead pick the branch of engineering that most interests you (e.g. EE, ME, ChemE) and then add a BE minor. Biology is a lot easier to learn yourself after graduation than is engineering, and particularly math.
"You can't do that with Linux either. The name aside, the Linux you run on one arch is not the Linux on another arch."
What the heck are you talking about? Of the architectures I mentioned, Debian Linux is available for every one of them except for POWER.
Well, maybe you don't like Debian. Okay. Fedora Linux (or distributions based directly off of it like Yellowdog and RHEL) gives you support for all of the architectures I've mentioned, including POWER.
And with Debian (and Fedora), you never loose the source-based flexibility. If you've got time to burn, you're welcome to recompile all the code from the source packages.
Anyway, the point that I was trying to make, is that on a modern polyarchitectural network, you're better of using Linux because it actually has support for all the modern architectures.
Well, if you're running Linux, you can hop between your 486, athlon xp, sparcbook, and g4 as well. Plus, you can also hop onto your Intel Itanium cluster, your not yet antiquated HP PA-Risc workstation, and your IBM Power5 server. You can't do that with NetBSD.
The grandparent post does have an element of truth to it.
There's a fair amount of hardware that NetBSD supports, but by "support" they mean you can hook a dumb terminal to the serial port, and boot off of ethernet. I personally consider "supported" to mean that the machine can access and boot of its own harddrive, can use and display more then just text on the graphics card, and can use its own keyboard.
As examples, check out the amount of hardware support offered for the vax or hppa architectures. It can be really spotty.
PSA was bought by USAir, not Southwest.
Okay, I'll give them peer to peer file swapping. That's pretty insightful.
By 1990/1991, MPG encoding had already been developed, along with all the technology behind HDTV. These researchers where already connected to ethernet in the lab (with a fiber optic backbone going around MIT). So the paper you're refering to was simply extrapolating the technologies that were already achievable in the lab for home use.
Why did it take a decade? Moore's law had to catch up. A workstation (and hard drive) to do DVR in 1990/1991 would of cost at least $20,000 (in 1990 dollars). A Tivo a decade later cost less than 1% of that (inflation adjusted).
I went to MIT in the mid 90's, and the real Media Lab was considered little more than a dog and pony show for corporate sponsorship. It was actually sort of an embarrassment to those doing real engineering and science at MIT, but it was tolerated because it brought in its own funding.
I knew people who worked at the Media lab, and I knew about a lot of research projects there. Absolutely none of them were advancing the fields of science or engineering (although I gotta imagine there are a couple of counter examples that I just don't know about). Their highest achievement was the development of "Lego Mindstorms", which is cool for sure, but is work that could easily be done in a purely corporate setting.
I gather, in Europe at least, corporations weren't as easiler suckered into shelling out money for research on gadgets and toys, especially when they could do the same research in house for less.
Okay, that's great, some paper out of the Media lab sucessfully extrapolated the obvious.
It's sort of like if I had written a paper in '90 or '91 predicting that processors would be running at over 1 GHz in 2000.
But imagine if that paper from the media lab had never been written, what sort of world would be living in right now? That's right, the same exact world. And processors are running at over 1 GHz even though I didn't write my famous paper in '90 or '91.
There's a big difference between spending your time predicting the future, and making it happen.
For crying out load, can't people read up on this?
McDonald's was knowingly superheating their coffee so that a brewed pot would stay fresher longer. They were heating their coffe to 180-190 degrees F. Your pot of coffee at home is more like 140 degrees. Even coffee at 155 degrees wouldn't give you serious burns.