I just happened to be out looking at this comet and what do you know,/. has an article about it when I came back inside.
I found it pretty quickly from the guide in Sky and Telescope, although the suggested spotting map will work just fine, too. I used a pair of 10x30 binocs and spotted it almost instantly. Took me another minute to make sure I wasn't looking at some other nebula. My non-techie wife was able to spot it just about as quickly - said it kind of pops out at you as you scan the heavens in the general area.
It's just a "dim fuzzy", and probably won't get much brighter, but it's cool to see. Set up my 6" telescope to see if the view was better. Kind of, but no further definition to speak of. Still definitely worth hauling out the binocs.
I have to agree with this response about enviro-friendly cars. Each part of the problem is somewhat separable: (1) manufacturing, (2) fuel consumption,and (3) eventual recycling of the discarded car.
BMW was working on that last item as long ago as 15 years ago, in that they were designing cars to be easily stripped down in the junkyard to allow differing materials to be separated (source: Roundel magazine circa 1989, long-discarded). Considering this was discussed in the context of their new 750il (12 cylinder with the name reminiscent of the 70's oil embargo), fuel efficiency was obviously not the main concern. This simply demonstrates one way that these three aspects can be addressed independently.
1) Sell insecure software products that is widely used because of market share 2) Observe as the world suffers under the impact of the insecure aspects 3) Patent new software that solves the problem original product caused???? 4) Profit!
I have an IBM T23 I've been nearly around the world with. I'm impressed with the battery life - I can get about 4 hours on a plane with it (low LCD brightness, no CD-ROM playing, MP3's on a separate device...)
But what matters almost more to me is how fast the batteries recharge. I haven't actually timed it, but I bet that one of the normal life T23 batteries would charge from 10% to 100% in little over an hour. Great for short layovers on x-country flights.
This is in stark contrast to the battery in a Toshiba laptop I had three years ago, which took at least twice as long to charge as it did to run down. Insane!
These are two trends in conflict: (1) mobility/portability of devices, (2) ubiquitous availability of same devices.
Personally, I can't see how ubiquity would ever replace mobility even if security were not an issue. Ubiquitous to me means "I have a computer available on any remote mountaintop or wheat field or desert" - yeah, right. Otherwise, you'll always need the mobile device.
On the other hand, perhaps there is a need for both. Still, I can't imagine waiting in line for a (rental) computer, or being chained to one location to do my computing, especially when I'm so accustomed to having a laptop.
It doesn't seem to be a worthwhile factorization of the problem. Concentrate on providing the wireless bandwidth everywhere, that's much easier!
It may come as a surprise to some, but there's quite a lot of interest in very small and very quiet PCs. Just check out this site as a for instance. And there's also this review on Tom's Hardware site.
I'm half chuckling with you, especially as I tend to be in airports a lot and have a hard time finding AC power adapters. Massport just installed Wi-Fi in Logan, for instance, but didn't help out with new outlets in the terminals. Likewise in Heathrow, they have a hotspot in one of the terminals, but no AC plugs to be seen. Had to sit on the hard windowsill in San Jose airport just to get close to the power. Seems like they want you to run your battery down before you get on the plane! (Can't count on seat power in the plane, either)
By the way, none of those places offer free Wi-Fi. And Logan's rates are especially ridiculous: $12 for 24 hours - as if!
I suppose I expected CDs to last longer than a few years. But the thing that shocked me the most was what you're supposed to do for maximum longevity: don't stick labels on them! Cripes! All those fancy labels are contributing to my bit rot!
Too bad there wasn't something called "microsoft" like in William Gibson's Neuromancer that Apple could invent and patent. Then we could have the equally disconcerting headline as this article!
Here are things that will affect you more than the shower curtain:
1) Those water filtering pitchers that live in your fridge (e.g., Brita filters). My family seemed to keep getting sick (colds, or sore throat) until we started taking real good care to clean the pitcher out regularly (dishwasher).
2) The pink stuff that can grow on your toothbrush (down at the bottom of the bristles). Yuck! I now *dry* my toothbrush off with a clean towel after use.
3) Razor blades! I used to get "shaving bubbles" under my chin and a rather irritated face until I dipped the double-edged razor in rubbing alcohol after every use.
I'm sure the shower scum isn't too healthy either, but heck, the easiest access microbes have to your body is through the mouth.
So what actually happens in that chamber? Anyone know? Here's my guess: they set up a high power standing wave at the fundamental frequency in the chamber. The low-pressure node gets cold; the high pressure node gets hot. Then they pump heat into the low pressure node and somehow it flows through the helium to the high pressure node and is conducted out of the system.
When I first saw a lecture on turbo codes 5 years ago, they suffered from an irreducible error rate phenomenon. Have they figured a way around this? Also, it seemed that choosing the proper interleaving was a black art.
MacOS has done something like this for years, right? I'd bet that's where BeOS got it. I'm sure BeOS improved it, but it was a good idea to begin with.
A look at Max OSX would be helpful. They somehow achieve a similar functionality on top of a Unix file system. I don't have the docs on hand, but I remember reading in their tutorial for SysAdmins about their package management, and how all a particular app's files were kept in one folder. Don't kow how it works since I don't have OSX installed yet...
To be honest, I think the NYT article is telling us just what we want to hear. The above referenced article reminds us that there was a lot of useless rail capacity built on speculation. Which way si ti going to happen? We'll be sure, it's going to be good for the rich people, and it might be good for us working stiffs.
I agree on your main points, but as a stickler for details I wanted to correct/clarify a few names/numbers.
4. 802.11, 802.11a, 802.11b, 802.11. 802.11 is the general IEEE group for Wireless LAN networking. 802.11b is the 11Mbps standard. 802.11a is the 5GHz 54Mbps standard (once they decide exactly what that standard is). TGe, which should translate to 802.11e will be the new security standard. There are others (including a standard for 22Mbps in the 2.4GHz band, which I _think_ is 802.11h), but I don't remember what most of them are.
The original 802.11 standard provided 3 different physical layers. The most common was direct sequence spread spectrum and it offered 1 and 2 Mbps airlink data rates.
802.11a came along, which eventually provided for data rates up to 54 Mbps in the UNII band somewhere between 5 and 6 GHz. It took a long time to get standardized, but it is, and there are *numerous* vendors out there all working on chipsets available "real soon now". It uses OFDM exclusively for data modulation.
While.11a was laboring away, some of the 2.4 GHz.11 people came up with a way to get up to 11 Mbps out of channels in that band. This came to be known as 802.11b, and it zoomed past the finish line before.11a. It uses "8 chip complementary code keying" (CCK). That's what has really made.11 take off recently.
There is currently another PHY layer being proposed under Task Group g. This will at least double the data rate for the 2.4 GHz band (22 Mbps or higher). Currently, there are two main factions: one led by Texas Instruments who champions another extension of CCK (I think) and another led by Intersil (I think) that wants an OFDM variant. The dispute is quite bitter, and lots of very political voting is going on. Mostly the fight is between the 2.4 GHz weenies that don't believe in 5 GHz, and the 5 GHz weenies that want to reuse most of their OFDM modem designs for a 2.4 GHz solution.
But they're all out in the rain because the real way to go is to build a combined.11a/b product. Just you wait and see.
All of these standards use the same MAC layer with different PHYs. That MAC layer has work going on in it too. 802.11e, or TGe was handling QoS and security. Now it's just QoS, and security was spun out to TGi. See a list of my recent posts to find articles on the.11 website discussing the flaws in WEP and the proposed improvements.
There's also TGh, which has to do with Transmit Power Control (TPC) and Dynamic Frequency Selection (DFS), both of which are important for.11 to be accepted in Europe, or something.
Then there's TGf, which has something to do with inter-AP protocols.
I would also add w.r.t. Wi-Fi that a lot of the 802.11 pointy-hairs are indignant that Wi-Fi is effectively renaming their precious "802.11" and supposedly getting all the credit. Waah, waah, waah.
Did some machine vision stuff for my Ph.D., so I feel an urge to comment...
So much of what we (I mean the media) call AI is really algorithms or heuristics for solving a problem. We can call it "AI," but once it's codified in software, does it sit up and say "hello?" No, you feed it some inputs and it produces an output. An image or signal is classified, a matching dataset is extracted, etc, etc.
Neural networks! It's hard enough already finding a global optimum in a well-chosen multivariate cost function, now we have to go modeling the same cost function with an arbitrary ANN model with way too many degrees of freedom. I once saw a paper written to show how they trained an ANN to compute a FFT. Big whoop. We already know how to compute an FFT, there's an algorithm for it. The AI article was balanced on the subject of ANNs. They're useful for solving some ill-posed problem. Still, if there could be a better way to solve the problem, I would spend a lot of time trying to find it.
What's an AI? To me, it would have to pass the Turing test for starters. After that we're in fantasyland. Self-directed, self-preserving, creative... hmmmm...
I believe you are talking about 802.1x, not 802.11x. The current work within 802.11 related to security is going on in Task Group I, or 802.11i. They do specify an enhanced security mode that uses.1x for authentication and blocking access, just as you say.
802.11i also specifies AES as the encryption algorithm to replace RC4, as well as many other improvements over WEP.
I don't think it's overstating the problem to say that RC4 was cracked. I agree that it applies to this usage of RC4, but there are other encryption techniques, like DES, that can be used with a nonsecret init vector, which do not provide a "crack" to the encryption, or at least haven't been discovered yet. One example: APCO-25, a digital radio standard for public safety use, uses DES as a PRNG initialized by a nonsecret IV. I am not an expert, but "cracking" this use of DES would be the equivalent of cracking DES used in electronic codebook mode (ECB) with a known plaintext. And that has not happened yet (brute force != crack).
...this attack doesn't necessarily mean that SSL sessions using RC4 can be read
The key word is "necessarily." If SSL uses a nonsecret IV of any kind, then we're hosed. I don't know if it does. On the other hand, it does mean that Netscape's 40-bit encryption is even easier to break because the way they get 40 bit keys is simply by making nonsecret the 128-40=88 remaining bits of a 128 bit key. That's like having an 88 bit IV!
Disclaimer: I'm not a security expert, but I read a lot of tech stuff on the subject, and I work in wireless LANs.
3DES is a total pig as far as computation goes. RSA is a public key scheme and is much better suited for key distribution rather than bulk encryption (it's slower and relatively weaker than good symmetric ciphers).
802.11, Task Group I is developing a replacement for WEP, based on 802.1X for station authentication, and AES-OCB for data encryption/authentication. I've already posted references about that - search for my username to find the posting.
The Shamir, et al, paper shows that it's RC4 itself that is weak, not just the bad IV generator. They're saying that because the same key is reused many times with a different IV (even a truly unpredictable one), one is able to figure out the key. This is a crack of RC4.
There are numerous other weaknesses in WEP. I recommend http://grouper.ieee.org/groups/802/11/Documents/Do cumentHolder/1-230.zip, which contains a very interesting Powerpoint presentation on the subject of cracking WEP. This latest bit of news is the nail in the coffin, though, because it's RC4 that's the problem now, not just the WEP protocol.
I believe it's still a concern for SSL if the session key is chosen once and used many times for the encrypted (SSL) conversation. It depends on how SSL maintains crypto sync at both ends of the conversation. Anyone care to answer that? Does SSL use something like an initialization vector with each new encrypted packet? Or does it rely on perfect synchronization between sender and receiver so that the RC4 PRNG can just keep on generating without being reinit'd between packets?
Slashdot Mirror
More detail and a picture over here
I just happened to be out looking at this comet and what do you know, /. has an article about it when I came back inside.
I found it pretty quickly from the guide in Sky and Telescope, although the suggested spotting map will work just fine, too. I used a pair of 10x30 binocs and spotted it almost instantly. Took me another minute to make sure I wasn't looking at some other nebula. My non-techie wife was able to spot it just about as quickly - said it kind of pops out at you as you scan the heavens in the general area.
It's just a "dim fuzzy", and probably won't get much brighter, but it's cool to see. Set up my 6" telescope to see if the view was better. Kind of, but no further definition to speak of. Still definitely worth hauling out the binocs.
I have to agree with this response about enviro-friendly cars. Each part of the problem is somewhat separable: (1) manufacturing, (2) fuel consumption,and (3) eventual recycling of the discarded car.
BMW was working on that last item as long ago as 15 years ago, in that they were designing cars to be easily stripped down in the junkyard to allow differing materials to be separated (source: Roundel magazine circa 1989, long-discarded). Considering this was discussed in the context of their new 750il (12 cylinder with the name reminiscent of the 70's oil embargo), fuel efficiency was obviously not the main concern. This simply demonstrates one way that these three aspects can be addressed independently.
Do you know if the iPod do gapless?
Hmmm, let's see if I can get this right....
1) Sell insecure software products that is widely used because of market share
2) Observe as the world suffers under the impact of the insecure aspects
3) Patent new software that solves the problem original product caused????
4) Profit!
Hi,
I have an IBM T23 I've been nearly around the world with. I'm impressed with the battery life - I can get about 4 hours on a plane with it (low LCD brightness, no CD-ROM playing, MP3's on a separate device...)
But what matters almost more to me is how fast the batteries recharge. I haven't actually timed it, but I bet that one of the normal life T23 batteries would charge from 10% to 100% in little over an hour. Great for short layovers on x-country flights.
This is in stark contrast to the battery in a Toshiba laptop I had three years ago, which took at least twice as long to charge as it did to run down. Insane!
These are two trends in conflict: (1) mobility/portability of devices, (2) ubiquitous availability of same devices.
Personally, I can't see how ubiquity would ever replace mobility even if security were not an issue. Ubiquitous to me means "I have a computer available on any remote mountaintop or wheat field or desert" - yeah, right. Otherwise, you'll always need the mobile device.
On the other hand, perhaps there is a need for both. Still, I can't imagine waiting in line for a (rental) computer, or being chained to one location to do my computing, especially when I'm so accustomed to having a laptop.
It doesn't seem to be a worthwhile factorization of the problem. Concentrate on providing the wireless bandwidth everywhere, that's much easier!
It may come as a surprise to some, but there's quite a lot of interest in very small and very quiet PCs. Just check out this site as a for instance. And there's also this review on Tom's Hardware site.
Sorry, don't think this is an attitude thruster - compare with this picture.
I'm half chuckling with you, especially as I tend to be in airports a lot and have a hard time finding AC power adapters. Massport just installed Wi-Fi in Logan, for instance, but didn't help out with new outlets in the terminals. Likewise in Heathrow, they have a hotspot in one of the terminals, but no AC plugs to be seen. Had to sit on the hard windowsill in San Jose airport just to get close to the power. Seems like they want you to run your battery down before you get on the plane! (Can't count on seat power in the plane, either)
By the way, none of those places offer free Wi-Fi. And Logan's rates are especially ridiculous: $12 for 24 hours - as if!
Well if they would just use emacs, they could just do "M-x upcase-region"!
:-)
Let the emacs flames begin...
I believe the final panel was this:
PHB: "You mean you'd stay at home and we'd just send you checks?"
Dogbert: "Actually, I was hoping for direct deposit" (little dogbert tail-wag...)
I suppose I expected CDs to last longer than a few years. But the thing that shocked me the most was what you're supposed to do for maximum longevity: don't stick labels on them! Cripes! All those fancy labels are contributing to my bit rot!
Too bad there wasn't something called "microsoft" like in William Gibson's Neuromancer that Apple could invent and patent. Then we could have the equally disconcerting headline as this article!
Here are things that will affect you more than the shower curtain:
1) Those water filtering pitchers that live in your fridge (e.g., Brita filters). My family seemed to keep getting sick (colds, or sore throat) until we started taking real good care to clean the pitcher out regularly (dishwasher).
2) The pink stuff that can grow on your toothbrush (down at the bottom of the bristles). Yuck! I now *dry* my toothbrush off with a clean towel after use.
3) Razor blades! I used to get "shaving bubbles" under my chin and a rather irritated face until I dipped the double-edged razor in rubbing alcohol after every use.
I'm sure the shower scum isn't too healthy either, but heck, the easiest access microbes have to your body is through the mouth.
Tired of chasing non-informative links...
So what actually happens in that chamber? Anyone know? Here's my guess: they set up a high power standing wave at the fundamental frequency in the chamber. The low-pressure node gets cold; the high pressure node gets hot. Then they pump heat into the low pressure node and somehow it flows through the helium to the high pressure node and is conducted out of the system.
Question for you with the big stack of papers:
When I first saw a lecture on turbo codes 5 years ago, they suffered from an irreducible error rate phenomenon. Have they figured a way around this? Also, it seemed that choosing the proper interleaving was a black art.
Would read the articles if I had the time...
MacOS has done something like this for years, right? I'd bet that's where BeOS got it. I'm sure BeOS improved it, but it was a good idea to begin with.
A look at Max OSX would be helpful. They somehow achieve a similar functionality on top of a Unix file system. I don't have the docs on hand, but I remember reading in their tutorial for SysAdmins about their package management, and how all a particular app's files were kept in one folder. Don't kow how it works since I don't have OSX installed yet...
We've seen a similar article, different conclusion:
5 24 3&mode=thread
http://slashdot.org/article.pl?sid=01/06/25/163
To be honest, I think the NYT article is telling us just what we want to hear. The above referenced article reminds us that there was a lot of useless rail capacity built on speculation. Which way si ti going to happen? We'll be sure, it's going to be good for the rich people, and it might be good for us working stiffs.
4. 802.11, 802.11a, 802.11b, 802.11. 802.11 is the general IEEE group for Wireless LAN networking. 802.11b is the 11Mbps standard. 802.11a is the 5GHz 54Mbps standard (once they decide exactly what that standard is). TGe, which should translate to 802.11e will be the new security standard. There are others (including a standard for 22Mbps in the 2.4GHz band, which I _think_ is 802.11h), but I don't remember what most of them are. The original 802.11 standard provided 3 different physical layers. The most common was direct sequence spread spectrum and it offered 1 and 2 Mbps airlink data rates.
802.11a came along, which eventually provided for data rates up to 54 Mbps in the UNII band somewhere between 5 and 6 GHz. It took a long time to get standardized, but it is, and there are *numerous* vendors out there all working on chipsets available "real soon now". It uses OFDM exclusively for data modulation.
While .11a was laboring away, some of the 2.4 GHz .11 people came up with a way to get up to 11 Mbps out of channels in that band. This came to be known as 802.11b, and it zoomed past the finish line before .11a. It uses "8 chip complementary code keying" (CCK). That's what has really made .11 take off recently.
There is currently another PHY layer being proposed under Task Group g. This will at least double the data rate for the 2.4 GHz band (22 Mbps or higher). Currently, there are two main factions: one led by Texas Instruments who champions another extension of CCK (I think) and another led by Intersil (I think) that wants an OFDM variant. The dispute is quite bitter, and lots of very political voting is going on. Mostly the fight is between the 2.4 GHz weenies that don't believe in 5 GHz, and the 5 GHz weenies that want to reuse most of their OFDM modem designs for a 2.4 GHz solution.
But they're all out in the rain because the real way to go is to build a combined .11a/b product. Just you wait and see.
All of these standards use the same MAC layer with different PHYs. That MAC layer has work going on in it too. 802.11e, or TGe was handling QoS and security. Now it's just QoS, and security was spun out to TGi. See a list of my recent posts to find articles on the .11 website discussing the flaws in WEP and the proposed improvements.
There's also TGh, which has to do with Transmit Power Control (TPC) and Dynamic Frequency Selection (DFS), both of which are important for .11 to be accepted in Europe, or something.
Then there's TGf, which has something to do with inter-AP protocols.
I would also add w.r.t. Wi-Fi that a lot of the 802.11 pointy-hairs are indignant that Wi-Fi is effectively renaming their precious "802.11" and supposedly getting all the credit. Waah, waah, waah.
802.11i also specifies AES as the encryption algorithm to replace RC4, as well as many other improvements over WEP.
The key word is "necessarily." If SSL uses a nonsecret IV of any kind, then we're hosed. I don't know if it does. On the other hand, it does mean that Netscape's 40-bit encryption is even easier to break because the way they get 40 bit keys is simply by making nonsecret the 128-40=88 remaining bits of a 128 bit key. That's like having an 88 bit IV!
Disclaimer: I'm not a security expert, but I read a lot of tech stuff on the subject, and I work in wireless LANs.
802.11, Task Group I is developing a replacement for WEP, based on 802.1X for station authentication, and AES-OCB for data encryption/authentication. I've already posted references about that - search for my username to find the posting.
There are numerous other weaknesses in WEP. I recommend http://grouper.ieee.org/groups/802/11/Documents/Do cumentHolder/1-230.zip, which contains a very interesting Powerpoint presentation on the subject of cracking WEP. This latest bit of news is the nail in the coffin, though, because it's RC4 that's the problem now, not just the WEP protocol.
I believe it's still a concern for SSL if the session key is chosen once and used many times for the encrypted (SSL) conversation. It depends on how SSL maintains crypto sync at both ends of the conversation. Anyone care to answer that? Does SSL use something like an initialization vector with each new encrypted packet? Or does it rely on perfect synchronization between sender and receiver so that the RC4 PRNG can just keep on generating without being reinit'd between packets?