>Didn't anyone remember that text processing was bulky and expensive? Human understanding of binary formats is what's bulky and expensive.
By comparison, getting a nearly-universal data description format widely adopted is hard. The W3C committee has now developed use cases for binarization and compression that show there are different problems to be solved, and they can be solved differently. These problems can be solved by using computers and algorithms, and by the nature of things, the solutions will get better, faster, and smaller as time goes on.
On the other hand, trying to get legions of programmers to hand-code optimal binary formats, or use systems such as ASN.1, has been shown time and time again to be difficult and error-prone and (I would say) counter-productive, and the problems is not likely to get any better as it's a human issue, not a computer issue.
As another poster commented, MIME, FTP, SMTP, HTTP, POP, and IMAP are all wildly succesful text-based protocols. For wildly popular binary protocols all I can think of is MS Word binary format (which they're dumping for XML and compressed XML), SSH, and the binary part of HTTPS (which encapsulates text-based HTTP).
I think using building on what Dave Raggett calls the "extraordinary success of HTML" (and by extension HTTP and MIME) and continuing to produce human-readable protocols is indeed the best path to success, and that designing a few standard ways to achieve the use cases that the W3C has gathered is the way to solve the specific problems.
Some of the most important use cases are - 1. Generic compression (gzip/deflate is pretty good here)
- 2. Schema-specific optimal compression (a program processes the Schema or DTD and outputs a binary converter; standardizing this compression mechanism would then allow the DTD or Schema to be enough to specify the binary version to any parser, preprocessor, or generator in any language)
- 3. Binary inclusions -- what if you want a JPEG file inside your document; short answer is to use MIME just like email does but there may be some other solutions
- 4. All of the above, but retaining the ability to include content allowed but not specified by the schema (foreign namespaces, etc.) This is much harder but just because this problem exists doesn't mean that we have to give up those people whose applictions are perfectly happy with 1, 2, and 3 and go back to everyone chiseling out their own binary protocols.
>I'm confused as to why the hiptop is so popular when there are similarly-priced phones with better features and more open development platforms. As you said, the keyboard is the killer and the built-in applications are well-designed and easy to use. Plus the unlimited data and unlimited SMS plans make data use a no-brainer. They just need to fix the relationship between third-party application developers and the phone carriers. Unfortunately, they sold consumers on the functionality of the device and they sold the phone carriers (their customers) on the ability to control. Now they're stuck in the middle.
Wonko42 wrote: The source code for Danger's SSH client is included in the hiptop SDK. If you suspect it's doing something shady, why not sign up for a developer account at http://developer.danger.com...
Sadly, it appears that the developer program is being "revamped" to the point that you won't be able to compile your own SSH. I presume this was done at the behest of T-Mobile, as the business model calls for selling "applications" (games and ringtones).
So we're supposed to trust that T-Mobile has vetted all the applications offered for sale, and not be able to distribute any applications written directly for the platform except through T-Mobile's commercial process, and then we find out that their security has been compromised.
I've been a long-time hiptop user and written my fair share of applications, but with this recent double whammy, I am beginning to think that Danger and T-Mobile need to take steps to recover the promise of abundant, safe, applications for the platform. Their new "walled garden" approach is certainly leading to fewer applications, and now it looks like it's leading to less safe ones -- how could you trust a pre-compiled SSH client on the hiptop knowing what we know now?
> Instead I'm saying that, at sufficiently large distances the true model approaches my model and diverges f No, what you said was "This is stupid." Do you see why that's insulting?
Yes, you are quite right. I omitted the multiple bits per baud factor to simplify the discussion as you and others have noted.
However, it might be good to calculate the number of symbols and the required s/n ratio to fit 1 Gigabit into the 2-80 Mhz presently requested by the bad HF BPL people and see if it would be better to fit it in a microwave region with more bandwidth and a lower requires s/n ratio, even dealing with the attenuation aspects.
Yes, you are quite right. I omitted the discussion of multiple bits per baud to simplify the frequency argument, but my point was not one of proof, but of suggestion that the gigabit bandwidths might be done at microwave frequencies. No matter how you slice it it's going to be hard to get a gigabit out of 30 megahertz of RF. With a low s/n ratio and enough symbols, yes, you could do it, but it may turn out to be easier to do it at microwave frequencies, in allocated spectrum, with significant attenuation due to atmospheric effects, for example, and have a BPL system that might not suffer from EMC problems.
It's true that unbalanced lines will radiate, but given the high data rates they're talking about it's quite possible that they could be doing this in the GHz range and attenuation could be significant there, or they could use licensed spectrum. For example, there is already spectrum allocated for MAN (metropolitan area networks) and if the market and the FCC decides to use leaky transmission lines instead of antennas, it's still the same spectrum.
My point is that we should not have knee-jerk reactions to BPL technologies, but should look at the technical details and evaluate them. So yes, anything that uses HF on the power lines is bad, but at microwave frequencies with significant attenuation or with licensed spectrum not taken from other users, it might be tenable. The technical aspects of the system must be analyzed first. Otherwise it turns into a meaningless political battle on both sides.
This may not be the same thing...data rates in excess of 1 Gigabit require bandwidth in excess of 2 GigaHertz. The BPL that is causing radio users (such as hams and public safety and other users) such fits uses the spectrum from roughly 2 to 70 MHz. That's 68MHz wide and can carry roughly 38 Megabits per power line.
If the power company solution used a frequency range that was entirely contained within the multi-GHz band, for example, there would be no interference in the critical "high frequency" 3-30Mhz spectrum that has special properties of world-wide propagation due to the ionosphere.
So let's not rush to judgement on all network technologies that could be deployed on power lines...those that use microwave or UHF frequencies might not have the same interference problems.
The ARRL does not have opposition to all technology -- just those that have been shown to be problematic and the problems swept under the rug by the FCC.
KFury Wrote: >If I wanted to break this 'record' I would simply replicate the experiemnt from a distance of 273 miles (half the distance) where I could pick up the signal with 1/4th the required signal strength...
No, you are wrong.
You are assuming an isotropic radiator and a line-of-sight path. At 80 meters wavelength and 546 miles, it's clearly ionospheric refraction which is subject to quite a few modes that don't fit into your model, and is not necessarily inverse squared with increasing great-circle distance point-to-point on the earth's surface.
At exactly half the distnace, the signal would have been completely absent, because that would be directly under the point in the ionosphere that was refracting the signal back down to the earth.
So yes, you can berate people for comparing miles/watt when they ought to be comparing meters^2/watt, and specifying the propagation path, and qualifying with the desired S/N for an achieved BER.
But no, your criticism rings false because it suffers from the same lack of understanding of the physics that you cite. The only question is the degree to which the mental model approximates reality. Miles per watt has a simplicity to it, and a greater utility over the distances it is used for (large fractions of the earth's surface). Your model lacks all these features.
More to the point, it seems to me that one of the big features of P2P programs is that they can use the response path to get around firewalls, home routers, and dynamic IP address problems by using a symmetric (or at least combined) protocol. (Of course, not everybody can be behind a firewall but at least some people can.)
TinyP2P appears to require that a server be able to accept incoming connections, rather than just make outgoing ones, in order to share bidirectionally. Fixing that would be a good project, but the XMLRPC libraries available might make it difficult; I don't know.
The 1000 miles per watt award is fairly easy to get. I exceeded it twice recently, when I worked ES5MC in Estonia from California with 4.5 watts with my Elecraft KX1 and a pack of AA batteries and a 28ft wire in a tree in central California, and OH9SCL in Santa Claus Land (Rovaniemi Finland, news, news) with the same radio from a parking lot by the San Francisco Bay.
>Can anyone else recall any Ham Radio enthusiasts who went onto bigger things in Tech?
Well, in a word plenty, but this is Slashdot, so let's be current and link to this month's Wired, which has an article about Mike Lazardis, who founded RIM and developed the Blackberry.
The paper copy of Wired (though not the online version) says he was a ham since childhood, but a recent issue of IEEE Spectrum magazine makes it clear that the development of the Blackberry came directly out of Mike's experience as a ham in Canada:
"The interest [in developing the Blackberry] came through his fascination with ham radio."
The Defend Air people don't give any data I can find on the dB-vs-MHz effectiveness of their product, but plenty of competitive technologies do. I wish they did, because I want to line my laundry room with the stuff! The washing machine makes a terrible racket in my radio.
There's Hospital Quality shielding done with aluminum foil, and more serious shielding of both E and H fields for MRI machines. I won't even go into the RF-sealed doors...
If you're concerned about magnetic fields, then Mu Metal is the stuff. Just don't bend any of the Hydrogen-annealed variety. You can get sheet and tape in small quantifies from these folks, who by the way also offer "Personal Protection Devices (silver-impregnated fiber baseball caps, not tinfoil hats, please)...and their own copper paint which is startlingly more expensive than the DefendAir product...they even have the same window film that keeps out the sun, but also makes your cell phone not work.
For sealing over the gaps, don't forget 3M 1181 Copper Tape, which features electrically-conducting adhesive, but only specifies 80dB isolation at 30MHz-1GHz.
But my favorite so far is Metal Foam, which reminds me of the almost weightless foamed glass Aerogel that was a announced a few years ago... Foamed aluminum is available commercially in 2x2ft sheets from from Austria, Alulight. They claim 40dB isolation in the 2Ghz range and over 100dB to 140dB in the 10Mhz-1GHz range, plus excellent sound isolation, structural, and fire safety properties. What's not to like? Anybody know where the get this stuff in the US? Reade seems out of my range, but I'll ask them.
See also Flex-Radio which can transmit (if you have a license), has a similar price point (within 3dB;-), works with GnuRadio software, and comes with its own Visual Basic software (source).
As long as they don't remove that feature, super-anti-commercial people like me will still be happy. Oh, ok. I'm sure they'll keep that feature in for you when they put in the paid advertizements during the regular fast forward. Don't worry at all.
When there is lots of aurora, the ionospheric reflection of radio is also perturbed. Read this description of the week from the RF point of view. In summary, the High Frequencies (3-30MHz) lose out, but he Very High Frequencies (30-300Mhz) gain. In particular, the 6 meter (50MHz) ham band showed some interesting reflections. For reference, that's right around tv Channel 2 in the US. Imagine not only being able to see Aurora, but sense them with your own radio and talk to someone by bouncing radio waves off of the aurora!
I plan on following them later this morning... I read this Yum Update Upgrade link above, and it says "Before proceeding, please read Seth Vidal's post (Seth is the author of yum) to the fedora-test-list mailing list regarding upgrading from FC2 to FC3 using yum."
The message from the developer of Yum says (among other things):
* dev->udev upgrade makes it impossible to open a new terminal *... * I'm sure there will be other problems.not be the time to try this new mechanism, if you're looking to upgrade a working system. The summaries I saw seemed to recommend anaconda.
I've been a ham since I was 7, but was inactive from the college years until recently. There's a tremendous number of things to do, from building your own low-power and medium-power equipment to computer-connected stuff, to Microwave (10 GHz is popular, and the 3.5 GHz band is getting more interesting these days too) and VLF (how about a signal on 176 KHz?).
Personally, I've ejoyed the following lately:
PSK-31 -- a cheap soundcard-based text-to-text mode that uses only 31Hz of bandwidth and goes around the world on 5 watts
RPSK -- a TCP/IP based protocol for remote operation of a PSK station with a Java applet client and a hiptop client. (The antenna is not hooked up right now so don't expect the applet to work.)
HFPack -- portable and picnic table operation with HF radio; I talked to Estonia with an Elecraft KX1 and about 4.5 Watts
An RSS feed for APRS -- working with APRSWorld I developed an APRS to RSS converter to help HFPackers let people know where and when they are operating, so people can listen for them.
....csv is not good enough for an address book anymore. There are too many fields (many of which won't be used for all entries) so viewing your book as a spreadsheet becomes tedious. Plus, there is no way to store picture information in a.csv file....but I wish the three would get together and hammer at a way to have a really powerful address book standard
Sound like there is call for an XML-based standard; then it could be manipulated by programs with XSLT and displayed in Mozilla XForms or CSS.
It seems that XML VCard work is ongoing and there is XML VCard support in Perl and a proposal for Jabber as well.
Ukraine has a research station in Antarctica too
on
Exploring Antarctica
·
· Score: 1
Yesterday I talked to Paul Budanov at Akademik Vernadsky Station on Galindez Island in Antarctica. Paul is there for the year, and is an amateur radio operator in addition to his scientific duties. I was using 25 Watts from my house, but I heard a friend talk to to Paul from his bicycle in Redding, California.
Slashdot Mirror
>Didn't anyone remember that text processing was bulky and expensive?
Human understanding of binary formats is what's bulky and expensive.
By comparison, getting a nearly-universal data description format widely adopted is hard. The W3C committee has now developed use cases for binarization and compression that show there are different problems to be solved, and they can be solved differently. These problems can be solved by using computers and algorithms, and by the nature of things, the solutions will get better, faster, and smaller as time goes on.
On the other hand, trying to get legions of programmers to hand-code optimal binary formats, or use systems such as ASN.1, has been shown time and time again to be difficult and error-prone and (I would say) counter-productive, and the problems is not likely to get any better as it's a human issue, not a computer issue.
As another poster commented, MIME, FTP, SMTP, HTTP, POP, and IMAP are all wildly succesful text-based protocols. For wildly popular binary protocols all I can think of is MS Word binary format (which they're dumping for XML and compressed XML), SSH, and the binary part of HTTPS (which encapsulates text-based HTTP).
I think using building on what Dave Raggett calls the "extraordinary success of HTML" (and by extension HTTP and MIME) and continuing to produce human-readable protocols is indeed the best path to success, and that designing a few standard ways to achieve the use cases that the W3C has gathered is the way to solve the specific problems.
Some of the most important use cases are
- 1. Generic compression (gzip/deflate is pretty good here)
- 2. Schema-specific optimal compression (a program processes the Schema or DTD and outputs a binary converter; standardizing this compression mechanism would then allow the DTD or Schema to be enough to specify the binary version to any parser, preprocessor, or generator in any language)
- 3. Binary inclusions -- what if you want a JPEG file inside your document; short answer is to use MIME just like email does but there may be some other solutions
- 4. All of the above, but retaining the ability to include content allowed but not specified by the schema (foreign namespaces, etc.) This is much harder but just because this problem exists doesn't mean that we have to give up those people whose applictions are perfectly happy with 1, 2, and 3 and go back to everyone chiseling out their own binary protocols.
>I'm confused as to why the hiptop is so popular when there are similarly-priced phones with better features and more open development platforms.
As you said, the keyboard is the killer and the built-in applications are well-designed and easy to use. Plus the unlimited data and unlimited SMS plans make data use a no-brainer. They just need to fix the relationship between third-party application developers and the phone carriers. Unfortunately, they sold consumers on the functionality of the device and they sold the phone carriers (their customers) on the ability to control. Now they're stuck in the middle.
ostiguy wrote:
I don't believe Sidekicks have anything - aren't they just dump pop3/imap devices?
No, they're J2ME (almost) compliant devices with additional Java libraries.
The built-in applications use a data sync that communicates with the back end over an encrypted channel using blowfish.
If you write your own apps you have to handle the encryption yourself, as the ssh client does.
Of course, if you buy applications from T-Mobile you don't know for sure what's in them; you just have to trust them.
Wonko42 wrote:
The source code for Danger's SSH client is included in the hiptop SDK. If you suspect it's doing something shady, why not sign up for a developer account at http://developer.danger.com...
Sadly, it appears that the developer program is being "revamped" to the point that you won't be able to compile your own SSH. I presume this was done at the behest of T-Mobile, as the business model calls for selling "applications" (games and ringtones).
So we're supposed to trust that T-Mobile has vetted all the applications offered for sale, and not be able to distribute any applications written directly for the platform except through T-Mobile's commercial process, and then we find out that their security has been compromised.
I've been a long-time hiptop user and written my fair share of applications, but with this recent double whammy, I am beginning to think that Danger and T-Mobile need to take steps to recover the promise of abundant, safe, applications for the platform. Their new "walled garden" approach is certainly leading to fewer applications, and now it looks like it's leading to less safe ones -- how could you trust a pre-compiled SSH client on the hiptop knowing what we know now?
> Instead I'm saying that, at sufficiently large distances the true model approaches my model and diverges f
No, what you said was "This is stupid." Do you see why that's insulting?
Yes, you are quite right. I omitted the multiple bits per baud factor to simplify the discussion as you and others have noted.
However, it might be good to calculate the number of symbols and the required s/n ratio to fit 1 Gigabit into the 2-80 Mhz presently requested by the bad HF BPL people and see if it would be better to fit it in a microwave region with more bandwidth and a lower requires s/n ratio, even dealing with the attenuation aspects.
Yes, you are quite right. I omitted the discussion of multiple bits per baud to simplify the frequency argument, but my point was not one of proof, but of suggestion that the gigabit bandwidths might be done at microwave frequencies. No matter how you slice it it's going to be hard to get a gigabit out of 30 megahertz of RF. With a low s/n ratio and enough symbols, yes, you could do it, but it may turn out to be easier to do it at microwave frequencies, in allocated spectrum, with significant attenuation due to atmospheric effects, for example, and have a BPL system that might not suffer from EMC problems.
It's true that unbalanced lines will radiate, but given the high data rates they're talking about it's quite possible that they could be doing this in the GHz range and attenuation could be significant there, or they could use licensed spectrum. For example, there is already spectrum allocated for MAN (metropolitan area networks) and if the market and the FCC decides to use leaky transmission lines instead of antennas, it's still the same spectrum.
My point is that we should not have knee-jerk reactions to BPL technologies, but should look at the technical details and evaluate them. So yes, anything that uses HF on the power lines is bad, but at microwave frequencies with significant attenuation or with licensed spectrum not taken from other users, it might be tenable. The technical aspects of the system must be analyzed first. Otherwise it turns into a meaningless political battle on both sides.
This may not be the same thing...data rates in excess of 1 Gigabit require bandwidth in excess of 2 GigaHertz. The BPL that is causing radio users (such as hams and public safety and other users) such fits uses the spectrum from roughly 2 to 70 MHz. That's 68MHz wide and can carry roughly 38 Megabits per power line.
If the power company solution used a frequency range that was entirely contained within the multi-GHz band, for example, there would be no interference in the critical "high frequency" 3-30Mhz spectrum that has special properties of world-wide propagation due to the ionosphere.
So let's not rush to judgement on all network technologies that could be deployed on power lines...those that use microwave or UHF frequencies might not have the same interference problems.
The ARRL does not have opposition to all technology -- just those that have been shown to be problematic and the problems swept under the rug by the FCC.
KFury Wrote:
>If I wanted to break this 'record' I would simply replicate the experiemnt from a distance of 273 miles (half the distance) where I could pick up the signal with 1/4th the required signal strength...
No, you are wrong.
You are assuming an isotropic radiator and a line-of-sight path. At 80 meters wavelength and 546 miles, it's clearly ionospheric refraction which is subject to quite a few modes that don't fit into your model, and is not necessarily inverse squared with increasing great-circle distance point-to-point on the earth's surface.
At exactly half the distnace, the signal would have been completely absent, because that would be directly under the point in the ionosphere that was refracting the signal back down to the earth.
So yes, you can berate people for comparing miles/watt when they ought to be comparing meters^2/watt, and specifying the propagation path, and qualifying with the desired S/N for an achieved BER.
But no, your criticism rings false because it suffers from the same lack of understanding of the physics that you cite. The only question is the degree to which the mental model approximates reality. Miles per watt has a simplicity to it, and a greater utility over the distances it is used for (large fractions of the earth's surface). Your model lacks all these features.
It doesn't seem to work, but a working version by Richard Jones. Unfortunately, it's incompatible ;-).
More to the point, it seems to me that one of the big features of P2P programs is that they can use the response path to get around firewalls, home routers, and dynamic IP address problems by using a symmetric (or at least combined) protocol. (Of course, not everybody can be behind a firewall but at least some people can.)
TinyP2P appears to require that a server be able to accept incoming connections, rather than just make outgoing ones, in order to share bidirectionally. Fixing that would be a good project, but the XMLRPC libraries available might make it difficult; I don't know.
The 1000 miles per watt award is fairly easy to get. I exceeded it twice recently, when I worked ES5MC in Estonia from California with 4.5 watts with my Elecraft KX1 and a pack of AA batteries and a 28ft wire in a tree in central California, and OH9SCL in Santa Claus Land (Rovaniemi Finland, news, news) with the same radio from a parking lot by the San Francisco Bay.
Well, in a word plenty, but this is Slashdot, so let's be current and link to this month's Wired, which has an article about Mike Lazardis, who founded RIM and developed the Blackberry.
The paper copy of Wired (though not the online version) says he was a ham since childhood, but a recent issue of IEEE Spectrum magazine makes it clear that the development of the Blackberry came directly out of Mike's experience as a ham in Canada:
The Defend Air people don't give any data I can find on the dB-vs-MHz effectiveness of their product, but plenty of competitive technologies do. I wish they did, because I want to line my laundry room with the stuff! The washing machine makes a terrible racket in my radio.
...and their own copper paint which is startlingly more expensive than the DefendAir product...they even have the same window film that keeps out the sun, but also makes your cell phone not work.
There's Hospital Quality shielding done with aluminum foil, and more serious shielding of both E and H fields for MRI machines. I won't even go into the RF-sealed doors...
If you're concerned about magnetic fields, then Mu Metal is the stuff. Just don't bend any of the Hydrogen-annealed variety. You can get sheet and tape in small quantifies from these folks, who by the way also offer "Personal Protection Devices (silver-impregnated fiber baseball caps, not tinfoil hats, please)
For sealing over the gaps, don't forget 3M 1181 Copper Tape, which features electrically-conducting adhesive, but only specifies 80dB isolation at 30MHz-1GHz.
But my favorite so far is Metal Foam, which reminds me of the almost weightless foamed glass Aerogel that was a announced a few years ago... Foamed aluminum is available commercially in 2x2ft sheets from from Austria, Alulight. They claim 40dB isolation in the 2Ghz range and over 100dB to 140dB in the 10Mhz-1GHz range, plus excellent sound isolation, structural, and fire safety properties. What's not to like? Anybody know where the get this stuff in the US? Reade seems out of my range, but I'll ask them.
See also Flex-Radio which can transmit (if you have a license), has a similar price point (within 3dB ;-), works with GnuRadio software, and comes with its own Visual Basic software (source).
As long as they don't remove that feature, super-anti-commercial people like me will still be happy.
Oh, ok. I'm sure they'll keep that feature in for you when they put in the paid advertizements during the regular fast forward. Don't worry at all.
When there is lots of aurora, the ionospheric reflection of radio is also perturbed. Read this description of the week from the RF point of view. In summary, the High Frequencies (3-30MHz) lose out, but he Very High Frequencies (30-300Mhz) gain. In particular, the 6 meter (50MHz) ham band showed some interesting reflections. For reference, that's right around tv Channel 2 in the US. Imagine not only being able to see Aurora, but sense them with your own radio and talk to someone by bouncing radio waves off of the aurora!
VA2VYZ has some nice aurora pictures from Quebec.
I have a GlowBug 40 (new) and an HW-16 (old). Both are fun. Vacuum tubes are still quite important for high power applications.
I plan on following them later this morning ...
...
I read this Yum Update Upgrade link above, and it says
"Before proceeding, please read Seth Vidal's post (Seth is the author of yum) to the fedora-test-list mailing list regarding upgrading from FC2 to FC3 using yum."
The message from the developer of Yum says (among other things):
* dev->udev upgrade makes it impossible to open a new terminal
*
* I'm sure there will be other problems.not be the time to try this new mechanism, if you're looking to upgrade a working system. The summaries I saw seemed to recommend anaconda.
I toy I absolutely loved as a kid was the 200-in-1 Electronics Lab.
Come join us over here
Weren't we all promised flying cars and hovercrafts when we grew up?
I bought one for $3 from Edmund Scientific, but my 3-year-old cousin stepped on it.
Well, they still make them, probably from the same vacuum mold.
I got a 3-pack, since the shipping was the same. Buy it now.
Personally, I've ejoyed the following lately:
Check it out and take a look at my Ham Web Log for more stuff.
Sound like there is call for an XML-based standard; then it could be manipulated by programs with XSLT and displayed in Mozilla XForms or CSS.
It seems that XML VCard work is ongoing and there is XML VCard support in Perl and a proposal for Jabber as well.
Yesterday I talked to Paul Budanov at Akademik Vernadsky Station on Galindez Island in Antarctica. Paul is there for the year, and is an amateur radio operator in addition to his scientific duties. I was using 25 Watts from my house, but I heard a friend talk to to Paul from his bicycle in Redding, California.
I wish it had XForms support...