Have you tried looking up Catalyst Wireless here in Ann Arbor? They have (or had) an antenna on the 777 building at Eisenhower & State, so if you're within a few miles of that you should be able to get a 512Kbps connections.
If that doesn't work, a T1 will cost you about $125-250 per month plus internet access, or about $400 per month if you don't abuse it.
It's deductible, of course.:-)
Re:WRONG!!! Its missing Silicon! is fake recipe!
on
Homemade Silly Putty
·
· Score: 5, Interesting
You are correct. It is fake!
We made actual silly putty in high school about 20 years ago. The main chemical was highly reactive to water, including the moisture in the air. A small bottle of the stuff (Dimethyldichlorosilane IIRC) came packed in 5 layers of protection, including a small wood box, and could only be ordered with the aid of a helpful adult with access to dangerous chemicals.
When we withdrew the liquid from the bottle using a syringe in our lab's hood, a wet rag 4 feet away starting smoking.
The putty we created wasn't very good. It did bounce, but didn't have all of the properties of the real stuff. Once we realized that we had no good way of disposing of the chemical, we had to keep making putty until it was all gone!
Today, we'd probably all be arrested for doing this in a public high school, but it was a great learning experience.
What a boring, piece of crap, 1960's vintage airliner Boeing is designing.
Oooo....it's 10% more efficient. Who cares? Everybody hates flying now because the airlines have made the experience worse than going to the dentist.
How about they design a cool, comfortable airplane that people actually WANT to travel on?
The year is 2003. We should be flying at Mach 1 or 2 by now. I guarantee you that people would pay 10% more to save 2 hours on a 5 hour flight. Plus, you'd get more passengers per day on a single plane and each pilot would be able to do an extra flight a day too.
Boeing is taking the boring, safe, dull approach that all big companies take when they become risk averse. Unfortunately, the cost of entry into the airplane industry is so high it's unlikely that another company is going to step in with bold new designs.
So for the next 30 years - and probably the rest of my life - we'll be flying at the same speed in the same uncomfortable seats as we did 20 years ago. feh!
I've used Macs since 1984. MacOS X is slow. Period.
For comparison, I fired up my old 40MHz Quadra 840AV. It was slower than my 667MHz TiBook, but not even remotely close to 15x slower. My 180MHz 8500 running the BeOS is easily 3x more responsive than my TiBook.
Everyone has their own theories about why it is slow. Most are probably true, such as:
MachO is slower than CFM
Quartz is a memory hog due to its backing store design
Too many layers - APIs on top of APIs on top of APIs (this is my personal choice)
It really needs to be fixed, but Apple has little incentive to fix it when reviewers refuse to publicize the problems. Plus Apple wants you to buy new systems that run faster.
Seems to me that the server should send a set of MD5 hashes first and then the client could request only those blocks that have changed.
It would probably be much faster, especially to multiple clients, since you wouldn't have to send the entire image over the network multiple times.
You'd have to do the math to figure out the best block size to use without letting errors creap into the transfer, though. Seems like 512KByte blocks would work well.
There used to be a Mac utility that did this that used AppleTalk broadcast packets and then individual clients could request whatever data that they missed during the bulk transfer. This was really important when your network was 400Kbps.
(1) Everyone shares the same encryption key, just like 802.11b. This means that your data isn't secure from other customers of the service. You could put a bridge-level encryptor on both ends, but that would cost even more.
(2) Backbone connections to the internet are not inexpensive. If you do the math, you'll find it very difficult to make money selling T1-class service for less than $300/month without oversubscribing. And that's not even including the wireless hardware costs.
I'll gladly pay more money if they'll give me a static IP address, allow me to run servers and a VPN, and improve my upload speed (currently capped at 128kbps).
I'm thinking about dumping my cable modem for a real T1, but $450/month is a lot more than the $45 I'm paying Comcast right now.
The one big thing missing from all Linux and Palm handhelds is good handwriting recognition.
The latest PocketPCs have a very good recognizer called Transcriber which makes the devices infinitely easier to use. It's the first one I've used that compares favorably with the Newton 2000 one.
Unless Calligrapher or some other decent recognizer is available under Linux, Microsoft is going to win this battle.
Microsoft is doing what they always do -- incrementally improving and adding features to their operating system over time. If/when they add real printer drivers to WinCE, it'll actually be a useful OS. The UI needs some work, but it's Microsoft so you can't really expect that.
If you don't truly outperform other codecs, you probably won't make money with it. Codec performance is very tricky to measure, with everyone using the codec wanting something slightly different (compression speed, decompression CPU load, image quality, motion quality, etc).
One option is to provide the compressor/decompressor for free and then offer a higher quality version of the compressor for sale. It's a tough market, though. If it's truly revolutionary, you could try selling it to Apple, Microsoft, or Real. You'll need some really good side-by-side comparisons with their current codecs and MPEG4 to get their attention, though.
Personally, I would love to see it open sourced.
Can we download a decompressor and see some demos?
This is a great move for Palm. It will be trivial to support current PalmOS applications running on a StrongArm version of the BeOS.
The BeOS should scale downward nicely to StrongArm-based Palm devices and it's one of the few operating systems in the world that truly understands the power of pervasive threading. (Linux, Windows, and MacOS X threads are a joke compared to the BeOS)
If Palm had any balls, they would open source the BeOS for use on desktop-class machines just to piss off Microsoft. I'm sure a lot of users would like to run the same operating system on their handhelds as their desktops, especially as their handhelds become more powerful.
The only thing Palm really needs now is some good handwriting recognition software. They really need to buy Calligrapher in order to make their handhelds useful to a larger number of people.
A for-profit corporation cannot use volunteer labor without paying the minimum wage in the US. If Gracenote is using freely submitted information and they now consider it proprietary, they owe a lot of people money and the taxes on it.
All 3 major media players will stream over an http connection if you set up their preferences to do so.
This isn't http web downloading -- it is actual streaming that is encapsulated within http connections to fool your firewall.
As long as you can get web access from work, you should be able to view streaming Real/QT/WMS video and audio. It is theoretically possible that a firewall module could prevent this, but I doubt many people are doing it since it involves parsing the http data to see if the content is streaming media.
The quality of streaming media over http can sometimes be poor, however, because TCP will retry/resend data that you may no longer need, unlike UDP-based streams.
The Psychology "winner" was a very interesting study.
It showed that the more incompetent people get, the less they realize their incompetence.
This could have a profound impact on the user interface design of everything from software to medical equipment to road signs.
It also demonstrates why people at Microsoft might actually think their operating system has a good user interface. I used to think they were just lying, but it turns out they might've just been so incompetent that they didn't realize how bad it was!
It seems to me that a cable company could use the bandwidth of a single coax going into someone's home for an unlimited number of channels. All they have to do is use a switching network rather than a hub-based one.
Think about it -- no home ever uses more than 4-5 video channels at one time. So why not just broadcast those 5 channels into the house? The set-top box or built-in tuner could be a 2-way device that transmits a channel request out to the cable company's equipment and then tunes in to the frequency that the cable company's switch assigns to that channel for that household.
The only problems with this are: (1) The cable company will still need to use fiber or multiple coax cables into each neighborhood to handle lots of HDTV signals. (They might be able to use a wider frequency range on a single coax) (2) Initial equipment cost might be high. (3) Lack of standards
This sort of switching system might tie-in well with handling voice traffic on cable. It would help prevent your neighbors from listening in on your phone conversations too.
Hmmm...maybe I should patent this and get rich even though it's an obvious idea. Consider it in the public domain in case some asshole tries to patent it in the future.
The real problem lies with Motorola. As far back as I remember, Motorola has never delivered a CPU on-time and in quantity.
The only reason this is Apple's fault is because they should know better than to trust Motorola. The 68020 was amazingly late. The 68060 was so late that Apple abandoned the chip family. Their biggest mistake was not going with Intel when they had the chance.
Don't get me wrong -- I much prefer the 68K architecture to anything Intel ever created. Especially the nice, clean instruction set. But Intel has consistently outperformed Motorola in getting chips out the door, at the fastest possible speed, and in quantity.
I don't think you can sue someone for using your trademark if it is functional. For example, many BIOS's used to have an IBM Copyright string in a special place so that DOS would run properly. This isn't considered a copyright or trademark violation.
Copyrights and trademarks are generally for non-functional protection, while patents are for functional protection.
(I am not a lawyer -- thank god)
>> I'm surprised AOL hasn't implemented a fairly easy method of stopping non-authorized clients. They could merely take a small (15x15 pixels or something) BMP of a trademarked logo (such as the AOL logo), and use it as a "key" to access the servers. Official AIM clients would transmit this logo to the servers for authentication, but Microsoft could not implement that in its client without being sued for trademark infringement.
Compaq's screen resolution patent might be one of their few serious ones. It was filed in 1982, but not approved until '86 so it still is valid for a few years.
I owned an original Compaq portable and it had the innovative ability to work in both IBM compatible video modes -- monochrome and CGA -- with a special keyboard combination to switch between the two. The key combination changed the scan rate of the monitor, not just the resolution. I'm not sure that they invented this, but I hadn't seen it before and it gave the Compaq a huge advantage over the IBM PC.
NEC came out with a multiscan monitor much later.
Most of the other patents are garbage IMHO. Very obvious solutions to anyone "trained in that field".
Slashdot Mirror
Taco,
:-)
Have you tried looking up Catalyst Wireless here in Ann Arbor? They have (or had) an antenna on the 777 building at Eisenhower & State, so if you're within a few miles of that you should be able to get a 512Kbps connections.
If that doesn't work, a T1 will cost you about $125-250 per month plus internet access, or about $400 per month if you don't abuse it.
It's deductible, of course.
You are correct. It is fake!
We made actual silly putty in high school about 20 years ago. The main chemical was highly reactive to water, including the moisture in the air. A small bottle of the stuff (Dimethyldichlorosilane IIRC) came packed in 5 layers of protection, including a small wood box, and could only be ordered with the aid of a helpful adult with access to dangerous chemicals.
When we withdrew the liquid from the bottle using a syringe in our lab's hood, a wet rag 4 feet away starting smoking.
The putty we created wasn't very good. It did bounce, but didn't have all of the properties of the real stuff. Once we realized that we had no good way of disposing of the chemical, we had to keep making putty until it was all gone!
Today, we'd probably all be arrested for doing this in a public high school, but it was a great learning experience.
What a boring, piece of crap, 1960's vintage airliner Boeing is designing.
Oooo....it's 10% more efficient. Who cares? Everybody hates flying now because the airlines have made the experience worse than going to the dentist.
How about they design a cool, comfortable airplane that people actually WANT to travel on?
The year is 2003. We should be flying at Mach 1 or 2 by now. I guarantee you that people would pay 10% more to save 2 hours on a 5 hour flight. Plus, you'd get more passengers per day on a single plane and each pilot would be able to do an extra flight a day too.
Boeing is taking the boring, safe, dull approach that all big companies take when they become risk averse. Unfortunately, the cost of entry into the airplane industry is so high it's unlikely that another company is going to step in with bold new designs.
So for the next 30 years - and probably the rest of my life - we'll be flying at the same speed in the same uncomfortable seats as we did 20 years ago. feh!
I've used Macs since 1984. MacOS X is slow. Period.
For comparison, I fired up my old 40MHz Quadra 840AV. It was slower than my 667MHz TiBook, but not even remotely close to 15x slower. My 180MHz 8500 running the BeOS is easily 3x more responsive than my TiBook.
Everyone has their own theories about why it is slow. Most are probably true, such as:
MachO is slower than CFM
Quartz is a memory hog due to its backing store design
Too many layers - APIs on top of APIs on top of APIs (this is my personal choice)
It really needs to be fixed, but Apple has little incentive to fix it when reviewers refuse to publicize the problems. Plus Apple wants you to buy new systems that run faster.
Apple's tech support rated higher than Dell in the most recent survey from PC World:
l e/ 0,aid,105854,00.asp
http://www.pcworld.com/resource/printable/artic
Seems to me that the server should send a set of MD5 hashes first and then the client could request only those blocks that have changed.
It would probably be much faster, especially to multiple clients, since you wouldn't have to send the entire image over the network multiple times.
You'd have to do the math to figure out the best block size to use without letting errors creap into the transfer, though. Seems like 512KByte blocks would work well.
There used to be a Mac utility that did this that used AppleTalk broadcast packets and then individual clients could request whatever data that they missed during the bulk transfer. This was really important when your network was 400Kbps.
Apple uses gcc. I think they moved to gcc 3.x for the PowerPC builds. I don't know what version they are using for Intel builds.
There are probably some projects at Apple that use CodeWarrior or even MPW, but in general MacOS X is built using gcc.
There was a cool hack at MacHack in June 2001 that did this.
H ac ks/Airport%20Radar%e2%84%a2/
A quick Google search turns up a copy:
href=http://blueg3.homeip.net:81/MacHack/The%20
Two problems:
(1) Everyone shares the same encryption key, just like 802.11b. This means that your data isn't secure from other customers of the service. You could put a bridge-level encryptor on both ends, but that would cost even more.
(2) Backbone connections to the internet are not inexpensive. If you do the math, you'll find it very difficult to make money selling T1-class service for less than $300/month without oversubscribing. And that's not even including the wireless hardware costs.
I'll gladly pay more money if they'll give me a static IP address, allow me to run servers and a VPN, and improve my upload speed (currently capped at 128kbps).
I'm thinking about dumping my cable modem for a real T1, but $450/month is a lot more than the $45 I'm paying Comcast right now.
The one big thing missing from all Linux and Palm handhelds is good handwriting recognition.
The latest PocketPCs have a very good recognizer called Transcriber which makes the devices infinitely easier to use. It's the first one I've used that compares favorably with the Newton 2000 one.
Unless Calligrapher or some other decent recognizer is available under Linux, Microsoft is going to win this battle.
Microsoft is doing what they always do -- incrementally improving and adding features to their operating system over time. If/when they add real printer drivers to WinCE, it'll actually be a useful OS. The UI needs some work, but it's Microsoft so you can't really expect that.
If you don't truly outperform other codecs, you probably won't make money with it. Codec performance is very tricky to measure, with everyone using the codec wanting something slightly different (compression speed, decompression CPU load, image quality, motion quality, etc).
One option is to provide the compressor/decompressor for free and then offer a higher quality version of the compressor for sale. It's a tough market, though. If it's truly revolutionary, you could try selling it to Apple, Microsoft, or Real. You'll need some really good side-by-side comparisons with their current codecs and MPEG4 to get their attention, though.
Personally, I would love to see it open sourced.
Can we download a decompressor and see some demos?
This is a great move for Palm. It will be trivial to support current PalmOS applications running on a StrongArm version of the BeOS.
The BeOS should scale downward nicely to StrongArm-based Palm devices and it's one of the few operating systems in the world that truly understands the power of pervasive threading. (Linux, Windows, and MacOS X threads are a joke compared to the BeOS)
If Palm had any balls, they would open source the BeOS for use on desktop-class machines just to piss off Microsoft. I'm sure a lot of users would like to run the same operating system on their handhelds as their desktops, especially as their handhelds become more powerful.
The only thing Palm really needs now is some good handwriting recognition software. They really need to buy Calligrapher in order to make their handhelds useful to a larger number of people.
Seems like the lawsuit filed against AOL for violating the Fair Labor Standards Act would work equally well against Gracenote.
8 9, 00.html
See:
http://www.techtv.com/print/story/0,23102,33102
A for-profit corporation cannot use volunteer labor without paying the minimum wage in the US. If Gracenote is using freely submitted information and they now consider it proprietary, they owe a lot of people money and the taxes on it.
No Jar Jar Binks. Enough said.
All 3 major media players will stream over an http connection if you set up their preferences to do so.
This isn't http web downloading -- it is actual streaming that is encapsulated within http connections to fool your firewall.
As long as you can get web access from work, you should be able to view streaming Real/QT/WMS video and audio. It is theoretically possible that a firewall module could prevent this, but I doubt many people are doing it since it involves parsing the http data to see if the content is streaming media.
The quality of streaming media over http can sometimes be poor, however, because TCP will retry/resend data that you may no longer need, unlike UDP-based streams.
The Psychology "winner" was a very interesting study.
It showed that the more incompetent people get, the less they realize their incompetence.
This could have a profound impact on the user interface design of everything from software to medical equipment to road signs.
It also demonstrates why people at Microsoft might actually think their operating system has a good user interface. I used to think they were just lying, but it turns out they might've just been so incompetent that they didn't realize how bad it was!
http://www.icab.de
It's a nice little browser, but doesn't support JavaScript or plugins yet. And it has some problems with a few popular sites.
For older Macs, it's the only way to go. And it's pretty quick on my PowerBook G3.
It seems to me that a cable company could use the bandwidth of a single coax going into someone's home for an unlimited number of channels. All they have to do is use a switching network rather than a hub-based one.
Think about it -- no home ever uses more than 4-5 video channels at one time. So why not just broadcast those 5 channels into the house? The set-top box or built-in tuner could be a 2-way device that transmits a channel request out to the cable company's equipment and then tunes in to the frequency that the cable company's switch assigns to that channel for that household.
The only problems with this are:
(1) The cable company will still need to use fiber or multiple coax cables into each neighborhood to handle lots of HDTV signals. (They might be able to use a wider frequency range on a single coax)
(2) Initial equipment cost might be high.
(3) Lack of standards
This sort of switching system might tie-in well with handling voice traffic on cable. It would help prevent your neighbors from listening in on your phone conversations too.
Hmmm...maybe I should patent this and get rich even though it's an obvious idea. Consider it in the public domain in case some asshole tries to patent it in the future.
This fiasco is only partially Apple's fault.
The real problem lies with Motorola. As far back as I remember, Motorola has never delivered a CPU on-time and in quantity.
The only reason this is Apple's fault is because they should know better than to trust Motorola. The 68020 was amazingly late. The 68060 was so late that Apple abandoned the chip family. Their biggest mistake was not going with Intel when they had the chance.
Don't get me wrong -- I much prefer the 68K architecture to anything Intel ever created. Especially the nice, clean instruction set. But Intel has consistently outperformed Motorola in getting chips out the door, at the fastest possible speed, and in quantity.
I don't think you can sue someone for using your trademark if it is functional. For example, many BIOS's used to have an IBM Copyright string in a special place so that DOS would run properly. This isn't considered a copyright or trademark violation.
Copyrights and trademarks are generally for non-functional protection, while patents are for functional protection.
(I am not a lawyer -- thank god)
>>
I'm surprised AOL hasn't implemented a fairly easy method of stopping non-authorized clients. They could merely take a small (15x15 pixels or something) BMP of a trademarked logo (such as the AOL logo), and use it as a "key" to access the servers. Official AIM clients would transmit this logo to the servers for authentication, but Microsoft could not implement that in its client without being sued for trademark infringement.
Compaq's screen resolution patent might be one of their few serious ones. It was filed in 1982, but not approved until '86 so it still is valid for a few years.
I owned an original Compaq portable and it had the innovative ability to work in both IBM compatible video modes -- monochrome and CGA -- with a special keyboard combination to switch between the two. The key combination changed the scan rate of the monitor, not just the resolution. I'm not sure that they invented this, but I hadn't seen it before and it gave the Compaq a huge advantage over the IBM PC.
NEC came out with a multiscan monitor much later.
Most of the other patents are garbage IMHO. Very obvious solutions to anyone "trained in that field".