As someone who uses both octave and matlab and contributes to octave, I can tell you that octave is getting more and more compatiable with matlab all of the time. Try the 2.1.64 version and see..
Anyone who visits your site is a potential customer. There are many ways of generating a profit stream from these customers, from direct sales from your site, to sponsered links, etc. The point I was make is that click throughs from slashdot have a real set of eye, with a wallet about 70cm below, whereas as a DOS lacks the eyes and the wallet.
If you failed to generate a revenue stream from the slashdot exposure then that is your problem.
You're kidding,right? Any commercial site would love to get slashdotted, the people that click on those links on the slashdot's front page do so because they want to read the content, and in most cases will probably wait for the couple of minutes for the content to load (Oh the wonders of Tabbed Browsing).
This is a completely different situation than in a DOS where the sole reason to access the page is to occupy bandwidth uselessly.
I have no doubts that in certain places mesh network is the way to go. Anything with rapid deployment for example (military, emergency services). However, mesh networks can't compete against a combination of fixed acess points and existing infrastructure. The performance is worse, and always will be, since the limiting factor is the retransmission occupy channel capacity, thus resulting in an overall reduction in the total system capacity.
In a scenario of fixed APs and infrastructure, mesh routing will at best be used for a single hop to an AP as a temporary means to improve the range to a node that you couldn't get to otherwise.
But, if you have electrical power, then why not use this to also get the network connectivity. Its just a question of if the time expended to setup the additional infrastructure is justified.
I'll add another point, if the other nodes in the mesh offer a capability you can't get through the fixed infrastructure. Interesting network you show. I presume it is 802.11b with directive antennas to get the range you are showing (maybe 802.11g or 11a though). So I presume the aim is at least 11Mbps. What are the local ADSL and cable suppliers in the area offering in that area? I doubt it is 11Mbps.
The reason to use mesh routing is to extend the range of a transmission by hopping between devices. To make such a strategy make sense the mesh nodes should either be battery powered, so they can be used in an area where you can't get power or be for use in a rapidly deployable network, where it doesn't make sense to set up additional infrastructure
But 4W of power makes in a bit power hungry for set and forget type of installation. The addition of power-over-ethernet make it even more of a joke. Why extend the range of one AP effectively doubling the traffic over the air, if the mesh point is already plugged into the ethernet to get its power? You're better off with two seperate APs in that case, and use the ethernet infrastructure for routing between the APs.
So although this is a nice device, the I don't see much use for it.
Something like a Cisco Aironet 352 is perfect, with excellent drivers
Seems you haven't tried to use the Cisco LEAP protocol (which my work insists on using) with the open source drivers. Major pain in the arse and as unstable as all hell.
Its complete rubbish to consider that plasma antennas could be used for WiFi. All of the plasma antennas I've heard of need something to contain the gas in which the plasma is excited. So your normal antenna is replaced with a glass vessel of a size larger than the largest antenna you want to have.
Directivity comes at the cost of size. Sure you want 30dBi gain for your WiFi, I'll build it for you, but it'll be at least 10 wavelengths in size (1.25 metres for 2.4GHz WiFi). The only applications I see permitting that sort of antenna size are real infrastructure devices, not consumer WiFi devices.
The main advantage I see of plasma antennas is as a stealth technology. By their nature all antennas have a relatively large radar cross-section in at least one direction. This is true even of plasma antennas. However, the advantage of plasma antennas is that you can turn them off when you aren't using them. As long as the containment vessel for the plasma itself has a low radar cross section, you'll then be virtually undetectable.
A whip antenna (or rod as you call it) as you say is tuned to a specific frequency which relate to the standing wave that is generated at the resonant frequency of the antenna. However a parabolic antenna (or dish as you call it) is based on simple geometric optics and is entirely frequency independent (upto certain limits). It is the feed of a parabolic antenna that is frequency dependent, and I've personally been involved in design groups working on feeds covering 5 to 10 octaves of bandwidth.
C-band is acceptabe for weather radar, but even then, you would be more worried about weather radar obstructing broadband connections...not the other way around.
Some people care what the weather is like. For your information, the use of polarmetric C-band weather radar, was the major reason that Japan had only 100MHz of 5GHz bandwidth for their systems. They have since opened some spectrum up at 4.9GHz.
I imagine there are a lot of geek linux users out there who hide behind a proxy that fakes that they are connecting from MSIE to stop certain web develeopers telling them what they can and can't render with their browser..... So your proof means exactly nothing.....
And I hope nobody sees me as some kind of Microsoft evangelist for my comments, because that's about as far from truth as you can get:-)
Ok, I'll let you off after this, and won't acuse you you being an evangelist. But I'd like to confirm that like many linux users my UserAgent string lies. I sit behind a proxy that changes it to avoid sloppy web designers trying to take away my choice of whether my browser can display their crappy pages or not. If I was really radical, I suppose I should just not visit such websites.
Hah, I was thinking exactly the same thing. Though you beat me to saying it first on slashdot:-)
One probably difference, although I like Piers Anthony and have read quite a bit of his stuff, character development is not what you'd call one of his strong points. This review leading me to believe the opossite is true in Blind Lake.
So, I wait anxiously for the response to your question from someone who has read both.
I work with the 802.11 and 802.15 standards quite a bit. In fact the group I work with had a PHY proposal for the 802.15.3 standard, that got beaten by TI at the last minute, ours was 5GHz.
The fact is that even TI is no longer supporting their own 802.15.3 standard. There are only a few start-ups that are. So I don't expect to see too many 802.15.3 products out there.
What is now happening is that the 802.15.3a group is working on a replacement PHY using the 802.15.3 MAC using UWB technology. In fact there is a vote in the second week of July about which standard will be used, though I have doubts this will be conclusive and will probably be put over till the Singapore meeting 2 months later.
Ironically, the group in the 802.15.3a with the most momentum at the moment have the most complex PHY (multi-band UWB system where 3.1 to 10.6GHz band is divided into 15 bands). So if a decision is made quickly on 802.15.3a it will take longer to get products out.
BTW, 802.15.3a will have a minimum of 110Mbs data-rate with possibly upto 500Mbs depending on the proposal.
I work as an RF systems design in a research lab of a major semiconductor manufacturer. And from the inside I have to ask "What is a WiFi standard?". It may seem a stupid question, but consider the IEEE 802.11 and 802.15 standards process
The IEEE has a voting system where votes are assigned to individually that have attended 3 consecutive meeting (held about every 2 months). This is supposed to make the standards process more egalitarian. But what really happens is that it is only the large corporations that can afford to send someone to a meeting every 2 months. Lots of the people in this meeting just come, sign the book, get out their laptop and start working on something else. So the standards are strongly corporate driven, and the votes are therefore usually driven by issues other than technical merit.
The "down-selection" process of the IEEE then forces these disparate
industrial players to come to some sort of compromise. This either
takes the shape of one large block of companies getting behind a
single standard and blocking other proposals, or all the standards
being wrapped up as options of a single standard. Neither of these
will necessarily have any relationship to technical merit, with the
second option being a sort of "non-standard" Standard.
As you see, I rather sympathize with the original article, mainly because I don't like the standards process as it stands. The thing is I don't think many people do, but I'm not sure I see how it could be done better.
Hum, good point. But this is not really a problem with the motherboard but with the panel itself. LVDS is just a low-voltage signalling connected (I believe for my board) to the northbridge. The complications come in knowing how to use this with a particular panels desired inputs.
So, the answer to your quest ion "inside knowledge for the panel":-)
As the Windows and Mac OS GUIs increase in sophistication, we have seen that they have been able to add in "network transparency" to an extent (ok more like "remote viewing") with things like VNC, and other implementations, that exist entirely seperate from the GUI proper - they basically implement a very very basic bitmap-copying protocol.
Is there a case where THIS IS NOT SUFFICIENT? Is it really that much of a win to burden the entire architecture with a feature that in its common use can be implemented completely seperately and still solve 90% of the problem?
Ok I'll give it a shot. In the past I've used applications with a graphical front-end that had to run from a supercomputer. The only way to use this application was run it with the graphics diplayed over the network with X11
Ok, you can argue that with a proper client/server architecture the graphics front-end could be seperated from the core application and run locally. You could even run both on the same machine to simulate the case of local graphics. So maybe this is just an issue of lazy programmers who don't want to write their own network protocols for their applications. But the network transparency of X11 is certainly a big convenience..
D.
Re:The problem of rewriting/forking XFree
on
XFree86 Politics
·
· Score: 1
What is X + ExtensionY? Is that still X? A application relying on ExtensionY will no longer run on plain X, will it?
X is X, Y is an extension. If the application needs and extension Y then yes it must be installed. If no one uses it then it dies and can be removed. I don't think its easy to question this as a model for an extensible system; consider plugin in other programs as an example. I don't expect to be able to run those nifty^H^H^H^H^Hcrap websites using Flash without having the right plugin installed.
The importance here is that the core software is well thought-out and maintained, etc. Do the gains in replacing entirely the X11 protocol with something else justify the downside of loss of compatibility. Up till now the overwhelm market response is No.
D.
Re:XI Accelerated-X
on
XFree86 Politics
·
· Score: 2, Interesting
This is a great laugh.... If you had any idea of the bad blood between XFree and Thomas Roell and XI.
Thomas Roell wrote the original X386 code base which he contributed to the X consortium. He then happily started his own business to sell x86 X-servers. Imagine how pissed off he was when other people took his code base and started undercutting his commerical offering
Until XFree 4.x was released XI's server was arguably the better offering. But I don't think you could make that case now...
Err, yes but have you visted and compared the differences between www.xfree86.org and www.x.org? Yes I know the X consortium is less important now than they were, but for a very long time they were the big boys and XFree was a side issue. So what is the difference to that case and having two versions of XFree?
Slashdot Mirror
Try looking at the link, if you want a symbolic toolbox for octave..
Cheers
D.
As someone who uses both octave and matlab and contributes to octave, I can tell you that octave is getting more and more compatiable with matlab all of the time. Try the 2.1.64 version and see..
Cheers
D.
Anyone who visits your site is a potential customer. There are many ways of generating a profit stream from these customers, from direct sales from your site, to sponsered links, etc. The point I was make is that click throughs from slashdot have a real set of eye, with a wallet about 70cm below, whereas as a DOS lacks the eyes and the wallet.
If you failed to generate a revenue stream from the slashdot exposure then that is your problem.
D.
You're kidding,right? Any commercial site would love to get slashdotted, the people that click on those links on the slashdot's front page do so because they want to read the content, and in most cases will probably wait for the couple of minutes for the content to load (Oh the wonders of Tabbed Browsing).
This is a completely different situation than in a DOS where the sole reason to access the page is to occupy bandwidth uselessly.
D.
I have no doubts that in certain places mesh network is the way to go. Anything with rapid deployment for example (military, emergency services). However, mesh networks can't compete against a combination of fixed acess points and existing infrastructure. The performance is worse, and always will be, since the limiting factor is the retransmission occupy channel capacity, thus resulting in an overall reduction in the total system capacity.
In a scenario of fixed APs and infrastructure, mesh routing will at best be used for a single hop to an AP as a temporary means to improve the range to a node that you couldn't get to otherwise.
D.
But, if you have electrical power, then why not use this to also get the network connectivity. Its just a question of if the time expended to setup the additional infrastructure is justified.
I'll add another point, if the other nodes in the mesh offer a capability you can't get through the fixed infrastructure. Interesting network you show. I presume it is 802.11b with directive antennas to get the range you are showing (maybe 802.11g or 11a though). So I presume the aim is at least 11Mbps. What are the local ADSL and cable suppliers in the area offering in that area? I doubt it is 11Mbps.
D.
The reason to use mesh routing is to extend the range of a transmission by hopping between devices. To make such a strategy make sense the mesh nodes should either be battery powered, so they can be used in an area where you can't get power or be for use in a rapidly deployable network, where it doesn't make sense to set up additional infrastructure
But 4W of power makes in a bit power hungry for set and forget type of installation. The addition of power-over-ethernet make it even more of a joke. Why extend the range of one AP effectively doubling the traffic over the air, if the mesh point is already plugged into the ethernet to get its power? You're better off with two seperate APs in that case, and use the ethernet infrastructure for routing between the APs.
So although this is a nice device, the I don't see much use for it.
D.
Something like a Cisco Aironet 352 is perfect, with excellent drivers
Seems you haven't tried to use the Cisco LEAP protocol (which my work insists on using) with the open source drivers. Major pain in the arse and as unstable as all hell.
Its complete rubbish to consider that plasma antennas could be used for WiFi. All of the plasma antennas I've heard of need something to contain the gas in which the plasma is excited. So your normal antenna is replaced with a glass vessel of a size larger than the largest antenna you want to have.
Directivity comes at the cost of size. Sure you want 30dBi gain for your WiFi, I'll build it for you, but it'll be at least 10 wavelengths in size (1.25 metres for 2.4GHz WiFi). The only applications I see permitting that sort of antenna size are real infrastructure devices, not consumer WiFi devices.
The main advantage I see of plasma antennas is as a stealth technology. By their nature all antennas have a relatively large radar cross-section in at least one direction. This is true even of plasma antennas. However, the advantage of plasma antennas is that you can turn them off when you aren't using them. As long as the containment vessel for the plasma itself has a low radar cross section, you'll then be virtually undetectable.
D.
A whip antenna (or rod as you call it) as you say is tuned to a specific frequency which relate to the standing wave that is generated at the resonant frequency of the antenna. However a parabolic antenna (or dish as you call it) is based on simple geometric optics and is entirely frequency independent (upto certain limits). It is the feed of a parabolic antenna that is frequency dependent, and I've personally been involved in design groups working on feeds covering 5 to 10 octaves of bandwidth.
D.
C-band is acceptabe for weather radar, but even then, you would be more worried about weather radar obstructing broadband connections...not the other way around.
Some people care what the weather is like. For your information, the use of polarmetric C-band weather radar, was the major reason that Japan had only 100MHz of 5GHz bandwidth for their systems. They have since opened some spectrum up at 4.9GHz.
D.
I imagine there are a lot of geek linux users out there who hide behind a proxy that fakes that they are connecting from MSIE to stop certain web develeopers telling them what they can and can't render with their browser..... So your proof means exactly nothing.....
D.
We should start saving up and buy our own Slashdot senator or congressperson.
I thought the correct term was congress-critter
D.
Ok, I'll let you off after this, and won't acuse you you being an evangelist. But I'd like to confirm that like many linux users my UserAgent string lies. I sit behind a proxy that changes it to avoid sloppy web designers trying to take away my choice of whether my browser can display their crappy pages or not. If I was really radical, I suppose I should just not visit such websites.
cheers
D.
Hah, I was thinking exactly the same thing. Though you beat me to saying it first on slashdot :-)
One probably difference, although I like Piers Anthony and have read quite a bit of his stuff, character development is not what you'd call one of his strong points. This review leading me to believe the opossite is true in Blind Lake.
So, I wait anxiously for the response to your question from someone who has read both.
Cheers
D.
since some spornsorships are coming from microsoft
Some people just have sex on the brain...
D.
Beep Wrong, the french say "McDo", and pronounce it something like McDoh....
D.
I work with the 802.11 and 802.15 standards quite a bit. In fact the group I work with had a PHY proposal for the 802.15.3 standard, that got beaten by TI at the last minute, ours was 5GHz.
The fact is that even TI is no longer supporting their own 802.15.3 standard. There are only a few start-ups that are. So I don't expect to see too many 802.15.3 products out there.
What is now happening is that the 802.15.3a group is working on a replacement PHY using the 802.15.3 MAC using UWB technology. In fact there is a vote in the second week of July about which standard will be used, though I have doubts this will be conclusive and will probably be put over till the Singapore meeting 2 months later.
Ironically, the group in the 802.15.3a with the most momentum at the moment have the most complex PHY (multi-band UWB system where 3.1 to 10.6GHz band is divided into 15 bands). So if a decision is made quickly on 802.15.3a it will take longer to get products out.
BTW, 802.15.3a will have a minimum of 110Mbs data-rate with possibly upto 500Mbs depending on the proposal.
Cheers
D.
The IEEE has a voting system where votes are assigned to individually that have attended 3 consecutive meeting (held about every 2 months). This is supposed to make the standards process more egalitarian. But what really happens is that it is only the large corporations that can afford to send someone to a meeting every 2 months. Lots of the people in this meeting just come, sign the book, get out their laptop and start working on something else. So the standards are strongly corporate driven, and the votes are therefore usually driven by issues other than technical merit.
The "down-selection" process of the IEEE then forces these disparate industrial players to come to some sort of compromise. This either takes the shape of one large block of companies getting behind a single standard and blocking other proposals, or all the standards being wrapped up as options of a single standard. Neither of these will necessarily have any relationship to technical merit, with the second option being a sort of "non-standard" Standard.
As you see, I rather sympathize with the original article, mainly because I don't like the standards process as it stands. The thing is I don't think many people do, but I'm not sure I see how it could be done better.
Hum, good point. But this is not really a problem with the motherboard but with the panel itself. LVDS is just a low-voltage signalling connected (I believe for my board) to the northbridge. The complications come in knowing how to use this with a particular panels desired inputs.
:-)
So, the answer to your quest ion "inside knowledge for the panel"
D.
Err, my cheap Via EPIA-M motherboard has an LVDS connector right on the motherboard... This for 115 Euro, with CPU, etc included.
Cheers
David
As the Windows and Mac OS GUIs increase in sophistication, we have seen that they have been able to add in "network transparency" to an extent (ok more like "remote viewing") with things like VNC, and other implementations, that exist entirely seperate from the GUI proper - they basically implement a very very basic bitmap-copying protocol.
Is there a case where THIS IS NOT SUFFICIENT? Is it really that much of a win to burden the entire architecture with a feature that in its common use can be implemented completely seperately and still solve 90% of the problem?
Ok I'll give it a shot. In the past I've used applications with a graphical front-end that had to run from a supercomputer. The only way to use this application was run it with the graphics diplayed over the network with X11
Ok, you can argue that with a proper client/server architecture the graphics front-end could be seperated from the core application and run locally. You could even run both on the same machine to simulate the case of local graphics. So maybe this is just an issue of lazy programmers who don't want to write their own network protocols for their applications. But the network transparency of X11 is certainly a big convenience..
D.
X is X, Y is an extension. If the application needs and extension Y then yes it must be installed. If no one uses it then it dies and can be removed. I don't think its easy to question this as a model for an extensible system; consider plugin in other programs as an example. I don't expect to be able to run those nifty^H^H^H^H^Hcrap websites using Flash without having the right plugin installed.
The importance here is that the core software is well thought-out and maintained, etc. Do the gains in replacing entirely the X11 protocol with something else justify the downside of loss of compatibility. Up till now the overwhelm market response is No.
D.
This is a great laugh.... If you had any idea of the bad blood between XFree and Thomas Roell and XI.
Thomas Roell wrote the original X386 code base which he contributed to the X consortium. He then happily started his own business to sell x86 X-servers. Imagine how pissed off he was when other people took his code base and started undercutting his commerical offering
Until XFree 4.x was released XI's server was arguably the better offering. But I don't think you could make that case now...
D.
Err, yes but have you visted and compared the differences between www.xfree86.org and www.x.org? Yes I know the X consortium is less important now than they were, but for a very long time they were the big boys and XFree was a side issue. So what is the difference to that case and having two versions of XFree?
D.