I believe the project is born dead. Let me explain why.
I believe the signal bandwidth as well as it's transmission distance corresponds to the group of 3-5 standard Soviet-era buildings, 80 apartments each. The single transformer serves this group - and also insulates the group from the 10-kV backbone network. It means that we either should place the routers on each transformer - or the capacitor bridges will passively route the data, overloading the network as a whole - or the groups will be insulated and still need a group router somewhere.Don't forget that all the equipment will be special (read: expensive due to limited production size).
Now let us consider the nearest competing technology - UTP Ethernet. You can easily buy routers, bridges, switches, cables, protectors and have no problems with mounting the cables between the buildings since there are usually NO roads between them; you can either throw the cable from roof to roof or use the existing phone tubes. You can use 802.11, leased lines or laser links as longlinks between networks. Such networks exist in post-Soviet territory and they work, mostly being supported by enthusiasts.
Moreover, now I take part in designing of time signals transmission system using the power lines, and I know that it's problematic to obtain the transmission speed comparable not even to Ethernet, but to V.32. The special chips I have access to have about 900 bits/sec. Anything giving the higher speeds will produce a lot of RFI, and it cannot be avoided.
Then, the second factor. Using the modem I can contact with any other modem; using the V.90 modem I have some troubles:-( having 56k with my friends; with DSL I cannot use the modem for anything except the Internet access. The same may be true for the electric networks, but not for Ethernet.
Since the power line modem cannot be used as a traditional modem ( 3-5 buildings zone - not more) it will be promoted by the ISP's in the monopoly way only.
So I believe the project will fail leaving a lot of expensive equipment on consumers' hands - as it has already been with some DSL companies. You have been warned.
I am not a crypto wizard to understand all the article. But I believe the main purpose of Passport was the necessity either to keep a lot of different passwords or to have a center that authentifies a person using the single password.
I believe that it's possible to keep a lot of different random passwords in any hardware device attached to computer thus avoiding most problems. The device may be the special keyboard (for instance, with magnetic, chip or proximity card reader), the standalone card reader or, preferably, USB key attached to the USB port that is present everywhere now (For USB challenged people there is a FDD). Of course, the device should be supported with some device-independent open-source protocol (Or we shall not trust it).
But I believe that Microsoft needs Passport NOT for our benefits, but for benefits of Billy's pockets and so Password will be pushed into our throats leaving all non-M$ aside.
Never buy brands. Buy a computer in a standard (ATX?) case. If you do so you will be able to replace the parts that become obsolete and computer will still work. After some replacement you will have enough parts for a second computer that presumably will be a server, router or, say, base computer for interface cards design where your slightest error fries the m/b.
This trick will not help with brandnames. You will throw away everything at once - case, m/b, video, processor - all, and spend a fortune to buy a new computer at once.
Apparently he types faster than he can decide which adjective to use.
1. Use expletives instead of adjectives. They are chosen automatically so no thinking process takes part in it:-)
2. I believe that there should be something in keyboard hardware that can store the favourite expletives, programming language keywords a.s.o. Use of special drivers is inadequate since requires the homogenous human-computer interface which doesn't exist since most of us use both the M$ and Unix-like OSes.
I've written that I specially study the keyboard problem in order to make some progress in that area. And seeing the half-QWERTY keyboard I have immediately found the prototype dated 08-Mar-1997:
here.
The IBM has produced the research IMHO exclusively for use in handhelds and other devices where the user must manipulate the computer with one hand. I fear there is no more application where such a keyboard can find it's place. Since it uses one hand only, the input speed should be theoretically decreased a half. The amount of fingers movement is the same which means that the errors caused by imprecise movements is the same. The keyboard does not divide the load evenly meaning the twofold increase of dreaded carpal tunnel syndrome.
And even in the field of PDAs there are a lot of more compact and efficient designs, from
Alphagrip which does not require tricks with a space bar since it's not chording - to lots of chording keyboards - BAT, Twiddler/clones, Wlonk, DataEgg/MicroWriter/Agenda, Monomanus a.s.o.
I have studied the problem specially (for desktop though) and I see no good method of text input at all.
Firstly, forget about direct neural interface and speech input. The neural interface is terribly slow, the speech input cannot deal with programming language texts.
Then, forget about a stylus and touchscreen. Stylus is simply too slow, touchscreen will cover your screen with grease from your hands.
There is a nice product that looks promising - Alphagrip - but AFAIK it's still not in production, and I fear it has a lot of other problems, for instance, I cannot imagine how it can be used as a game device as claimed by inventors, or how the index finger can press UP.
Datahand looks the best and is ergonomic enough but is simply too big to fit in a laptop.Twiddler IMHO requires too non-ergonomic fingers movement severely limiting the input speed.
The following methods look promising: Keybowl that can be emulated with a pair of analog joysticks taken from Sony PS Doubleshock joysticks, Wlonk - a 10-key macroprogrammable chording keyboard (You should design the mouse, driver and formfactor yourself),
and variations of TheBAT, DataEgg, 7KEY a.s.o (7-key 1-hand chording keyboard. Are you going to study the chords?)
The idea looks promising (I've thought about something similar - TheBAT movable like a mouse).
But really, of all well-known deficiencies of standard keyboard it addresses only the problem of mouse-keyboard switch, and generates some new problems, the first of which is the problem of loss of bimanual motor memory. Let me discuss it in depth:
If your standard keyboard lies on your table, then you place one hand on "f" and know that your right hand will find "j" on it's place due to your motor memory. If you use the split keyboard, you adapt to more or less constant split. When the part of keyboard is movable then you should use some visual or tactile feedback.
I understand that the right hand lies on the mouse and has an anchor point so. But the proper keying requires your wrist to be ABOVE the keyboard. It's WRONG to type when your hand is supported (I do so, but it doesn't mean that I am correct; and when I type fast, I rise my hands)
Maybe the design should use the adaptive brakes that allow movement in mouse mode only.
About my own projects: I try to invent something that addresses the bigger part of known key/mouse problem and fully eliminates any hands movement except fingers. It will be joystick-like in form.
In Russian part of Fidonet there was already published the story about MICROS~1 that has found a big company that has no registered Windows. MICROS~1 requested them to buy appropriate number of Windows, but got an answer "We don't use such a [] as IBM PC and Windows, we use MACs".
No, I have no RSI, CTS or other keyboard-induced sickness. I have 2 other problems: ALFA-LOCK and mouse switching, both giving terrible amounts of errors.
Imagine that your keyboard has a separate capslock for every Window. Now, you begin typing a phrase and see it in capital letters, though you printed in other window with small ones. You switch to small letters and press BS, then try to print - and see that menu is called instead. Terrible? Yes. But it's so: localized Windows uses ALT/SHIFT as Alphabet lock
Mouse switching is not easier: when Xerox, Apple and MICROS~1 produced their "Desktop metaphor" they forgot that we have no separate hand to keep a mouse. So, every time I switch from mouse to keyboard I must reposition hands looking at the keyboard (Or creeping with my fingers for F-J-5 tactile mark).
Both problems aren't solved with any "ergonomic" keyboard, possibly except Datahand. But Datahand's mouse looks problematic.
I've tried to find an alternative input methodology in Internet - and failed. Everything - Datahand, BAT, Wlonk, Twiddler, Keybowl, Alphagrip, DataEgg, etc - has specific deficiencies and - moreover - the interface is so based on the standard keyboard that any modifications not directly mapping the alternative keyboards to standard ones become extremely complex. And they don't solve the ALFA LOCK problems, since they are invented in English-speaking countries.
Excuse me, but what is the _AIM_ of all this project? I hope, it's either _catch_ the vandals (yourself or with a help of police), or _save_ your *fishes. If you want to catch them, your webcam and cellphone are absolutely irrelevant. What will you do with a lot of best quality pictures of absolutely unknown vandals? Go to police? They have a lot of work even without your *fishery.
If you want to be there and immediately catch them - you are out of luck. You are one, they are many - and they have enough time to swim away. What are you going to do then? Boat races a-la agent 007?
So, I advice either: making there a room that is automatically closed and locked when entered so that only you can open it (There is a lot of such tricks in old knight films), or the guard (Including the man with at least a baton or a hungry dog fed with vandals only). I cannot advice leaving somewhere vodka with cyanide since it's extremely dangerous and highly illegal.
I don't think you are the only *fisher there. Make a treaty with your neighbours about common defense so that everybody seeing vandals informs the others and then you catch them together.
And, if you are a programmer, and you can attach your *fishery to the Internet, you can simply move your workplace to your boat.
Situation is the following:
AOL client must send back a derivative of itself. So it's impossible to send back a proper response if you have not a legal copy of AOL client.
Now let us imagine the following:
Let us create an open-source message protocol having no central server (As Gnutella does). And let us create a lot of gate servers located somewhere outside the US jurisdiction that interfaces the AOL world and the world of our protocol. There the gate may contain the valid *.exe file and produce valid responses.
Of course, the gate should behave exactly like the
official client or traditional proxy server (I am more than sure than AOL can work via proxy since it will othervise lose the unreasonably big number of Intranet clients) and so be untraceable from the AOL side.
Since the servers know each other they can request the valid response from their neighbours and so it's impossible to prove that any given server contains a *.exe file, so the violation of *.exe copyright restriction is unprovable from the new protocol side.
In the best case the new protocol will
soon replace the AOL one and sent it to oblivion, and in the worst case the clients will at least internetwork with AOL reasonably unpunishably.
(IANAL) AOL can object to gate servers that are unreachable but not to authors of gate and client since they don't include *.exe to their distributions.
And BTW: You can include in your protocol everything you like, including the much higher security and the other useful features.
The idea of controlling something with head movements is not new. At least there are rumors that we Russians have successfully implemented it to control the rockets by means of magnet in pilot's helmet and magnetometer above.
But the pilots simply have no other means since they control the fighter with hands and feet, they are paid really high and they have no enough rockets to turn their heads every day.
We can imagine the pain in the neck which is much more terrible than any RSI. The problem of ergonomic control is being studied by us and we see no satisfactory methods for now. Even the
AlphaGrip - the last cry of technofashion - seems inadequate.
We are Admin of FreeBSD. Windows is irrelevant. You will be assimilated.
1. Imagine you are walking around with such a thing. You should have an instant Internet access - or all this system becomes meaningless (If you don't reread something you have read before, of course). So, you should either have a reasonably powerful radio modem to connect with your own ground-based access point, or the cellular modem (read: pay a lot to your cellular provider), or the 802.11 card (read: pay a lot to your 802.11 network owner) or Bluetooth (Tens of meters from the access point or a lot of money to the support network owner).
So I believe it will be only be limited to campuses where either the administration supports the efforts or the student self-organization does the same.
Unfortunately, the authors have not mentioned any network access mechanism on their page.
The second problem is the absence (at least visible) of any keyboard input. So, it becomes the browsing ONLY device - NOT the organizer, NOT the electronic notepad, NOT the programming tool.
The hated MICROS~1 OS, lack of open-source OS, the terrible price, possible eye injury are IMHO secondary compared with these two.
Slashdot Mirror
I believe the project is born dead. Let me explain why.
:-( having 56k with my friends; with DSL I cannot use the modem for anything except the Internet access. The same may be true for the electric networks, but not for Ethernet.
I believe the signal bandwidth as well as it's transmission distance corresponds to the group of 3-5 standard Soviet-era buildings, 80 apartments each. The single transformer serves this group - and also insulates the group from the 10-kV backbone network. It means that we either should place the routers on each transformer - or the capacitor bridges will passively route the data, overloading the network as a whole - or the groups will be insulated and still need a group router somewhere.Don't forget that all the equipment will be special (read: expensive due to limited production size).
Now let us consider the nearest competing technology - UTP Ethernet. You can easily buy routers, bridges, switches, cables, protectors and have no problems with mounting the cables between the buildings since there are usually NO roads between them; you can either throw the cable from roof to roof or use the existing phone tubes. You can use 802.11, leased lines or laser links as longlinks between networks. Such networks exist in post-Soviet territory and they work, mostly being supported by enthusiasts.
Moreover, now I take part in designing of time signals transmission system using the power lines, and I know that it's problematic to obtain the transmission speed comparable not even to Ethernet, but to V.32. The special chips I have access to have about 900 bits/sec. Anything giving the higher speeds will produce a lot of RFI, and it cannot be avoided.
Then, the second factor. Using the modem I can contact with any other modem; using the V.90 modem I have some troubles
Since the power line modem cannot be used as a traditional modem ( 3-5 buildings zone - not more) it will be promoted by the ISP's in the monopoly way only.
So I believe the project will fail leaving a lot of expensive equipment on consumers' hands - as it has already been with some DSL companies. You have been warned.
I am not a crypto wizard to understand all the article. But I believe the main purpose of Passport was the necessity either to keep a lot of different passwords or to have a center that authentifies a person using the single password.
I believe that it's possible to keep a lot of different random passwords in any hardware device attached to computer thus avoiding most problems. The device may be the special keyboard (for instance, with magnetic, chip or proximity card reader), the standalone card reader or, preferably, USB key attached to the USB port that is present everywhere now (For USB challenged people there is a FDD). Of course, the device should be supported with some device-independent open-source protocol (Or we shall not trust it).
But I believe that Microsoft needs Passport NOT for our benefits, but for benefits of Billy's pockets and so Password will be pushed into our throats leaving all non-M$ aside.
Never buy brands. Buy a computer in a standard (ATX?) case. If you do so you will be able to replace the parts that become obsolete and computer will still work. After some replacement you will have enough parts for a second computer that presumably will be a server, router or, say, base computer for interface cards design where your slightest error fries the m/b.
This trick will not help with brandnames. You will throw away everything at once - case, m/b, video, processor - all, and spend a fortune to buy a new computer at once.
Delete all SPAM from my address to answer.
Apparently he types faster than he can decide which adjective to use. :-)
1. Use expletives instead of adjectives. They are chosen automatically so no thinking process takes part in it
2. I believe that there should be something in keyboard hardware that can store the favourite expletives, programming language keywords a.s.o. Use of special drivers is inadequate since requires the homogenous human-computer interface which doesn't exist since most of us use both the M$ and Unix-like OSes.
I've written that I specially study the keyboard problem in order to make some progress in that area. And seeing the half-QWERTY keyboard I have immediately found the prototype dated 08-Mar-1997: here.
The IBM has produced the research IMHO exclusively for use in handhelds and other devices where the user must manipulate the computer with one hand. I fear there is no more application where such a keyboard can find it's place. Since it uses one hand only, the input speed should be theoretically decreased a half. The amount of fingers movement is the same which means that the errors caused by imprecise movements is the same. The keyboard does not divide the load evenly meaning the twofold increase of dreaded carpal tunnel syndrome.
And even in the field of PDAs there are a lot of more compact and efficient designs, from Alphagrip which does not require tricks with a space bar since it's not chording - to lots of chording keyboards - BAT, Twiddler/clones, Wlonk, DataEgg/MicroWriter/Agenda, Monomanus a.s.o.
Remove CARPAL TUNNEL from my addressto answer.
I have studied the problem specially (for desktop though) and I see no good method of text input at all.
Firstly, forget about direct neural interface and speech input. The neural interface is terribly slow, the speech input cannot deal with programming language texts.
Then, forget about a stylus and touchscreen. Stylus is simply too slow, touchscreen will cover your screen with grease from your hands.
There is a nice product that looks promising - Alphagrip - but AFAIK it's still not in production, and I fear it has a lot of other problems, for instance, I cannot imagine how it can be used as a game device as claimed by inventors, or how the index finger can press UP.
Datahand looks the best and is ergonomic enough but is simply too big to fit in a laptop.Twiddler IMHO requires too non-ergonomic fingers movement severely limiting the input speed.
The following methods look promising:
Keybowl that can be emulated with a pair of analog joysticks taken from Sony PS Doubleshock joysticks,
Wlonk - a 10-key macroprogrammable chording keyboard (You should design the mouse, driver and formfactor yourself),
and variations of TheBAT, DataEgg, 7KEY a.s.o (7-key 1-hand chording keyboard. Are you going to study the chords?)
Remove all SPAM from my email to answer.
The idea looks promising (I've thought about something similar - TheBAT movable like a mouse).
But really, of all well-known deficiencies of standard keyboard it addresses only the problem of mouse-keyboard switch, and generates some new problems, the first of which is the problem of loss of bimanual motor memory. Let me discuss it in depth:
If your standard keyboard lies on your table, then you place one hand on "f" and know that your right hand will find "j" on it's place due to your motor memory. If you use the split keyboard, you adapt to more or less constant split. When the part of keyboard is movable then you should use some visual or tactile feedback.
I understand that the right hand lies on the mouse and has an anchor point so. But the proper keying requires your wrist to be ABOVE the keyboard. It's WRONG to type when your hand is supported (I do so, but it doesn't mean that I am correct; and when I type fast, I rise my hands)
Maybe the design should use the adaptive brakes that allow movement in mouse mode only.
About my own projects: I try to invent something that addresses the bigger part of known key/mouse problem and fully eliminates any hands movement except fingers. It will be joystick-like in form.
Please remove all SPAM to answer
In Russian part of Fidonet there was already published the story about MICROS~1 that has found a big company that has no registered Windows. MICROS~1 requested them to buy appropriate number of Windows, but got an answer "We don't use such a [] as IBM PC and Windows, we use MACs".
No, I have no RSI, CTS or other keyboard-induced sickness. I have 2 other problems: ALFA-LOCK and mouse switching, both giving terrible amounts of errors.
Imagine that your keyboard has a separate capslock for every Window. Now, you begin typing a phrase and see it in capital letters, though you printed in other window with small ones. You switch to small letters and press BS, then try to print - and see that menu is called instead. Terrible? Yes. But it's so: localized Windows uses ALT/SHIFT as Alphabet lock
Mouse switching is not easier: when Xerox, Apple and MICROS~1 produced their "Desktop metaphor" they forgot that we have no separate hand to keep a mouse. So, every time I switch from mouse to keyboard I must reposition hands looking at the keyboard (Or creeping with my fingers for F-J-5 tactile mark).
Both problems aren't solved with any "ergonomic" keyboard, possibly except Datahand. But Datahand's mouse looks problematic.
I've tried to find an alternative input methodology in Internet - and failed. Everything - Datahand, BAT, Wlonk, Twiddler, Keybowl, Alphagrip, DataEgg, etc - has specific deficiencies and - moreover - the interface is so based on the standard keyboard that any modifications not directly mapping the alternative keyboards to standard ones become extremely complex. And they don't solve the ALFA LOCK problems, since they are invented in English-speaking countries.
Excuse me, but what is the _AIM_ of all this project? I hope, it's either _catch_ the vandals (yourself or with a help of police), or _save_ your *fishes. If you want to catch them, your webcam and cellphone are absolutely irrelevant. What will you do with a lot of best quality pictures of absolutely unknown vandals? Go to police? They have a lot of work even without your *fishery.
If you want to be there and immediately catch them - you are out of luck. You are one, they are many - and they have enough time to swim away. What are you going to do then? Boat races a-la agent 007?
So, I advice either: making there a room that is automatically closed and locked when entered so that only you can open it (There is a lot of such tricks in old knight films), or the guard (Including the man with at least a baton or a hungry dog fed with vandals only). I cannot advice leaving somewhere vodka with cyanide since it's extremely dangerous and highly illegal.
I don't think you are the only *fisher there. Make a treaty with your neighbours about common defense so that everybody seeing vandals informs the others and then you catch them together.
And, if you are a programmer, and you can attach your *fishery to the Internet, you can simply move your workplace to your boat.
Situation is the following:
AOL client must send back a derivative of itself. So it's impossible to send back a proper response if you have not a legal copy of AOL client.
Now let us imagine the following:
Let us create an open-source message protocol having no central server (As Gnutella does). And let us create a lot of gate servers located somewhere outside the US jurisdiction that interfaces the AOL world and the world of our protocol. There the gate may contain the valid *.exe file and produce valid responses. Of course, the gate should behave exactly like the official client or traditional proxy server (I am more than sure than AOL can work via proxy since it will othervise lose the unreasonably big number of Intranet clients) and so be untraceable from the AOL side.
Since the servers know each other they can request the valid response from their neighbours and so it's impossible to prove that any given server contains a *.exe file, so the violation of *.exe copyright restriction is unprovable from the new protocol side.
In the best case the new protocol will soon replace the AOL one and sent it to oblivion, and in the worst case the clients will at least internetwork with AOL reasonably unpunishably.
(IANAL) AOL can object to gate servers that are unreachable but not to authors of gate and client since they don't include *.exe to their distributions.
And BTW: You can include in your protocol everything you like, including the much higher security and the other useful features.
The idea of controlling something with head movements is not new. At least there are rumors that we Russians have successfully implemented it to control the rockets by means of magnet in pilot's helmet and magnetometer above.
But the pilots simply have no other means since they control the fighter with hands and feet, they are paid really high and they have no enough rockets to turn their heads every day.
We can imagine the pain in the neck which is much more terrible than any RSI. The problem of ergonomic control is being studied by us and we see no satisfactory methods for now. Even the AlphaGrip - the last cry of technofashion - seems inadequate.
We are Admin of FreeBSD. Windows is irrelevant. You will be assimilated.
1. Imagine you are walking around with such a thing. You should have an instant Internet access - or all this system becomes meaningless (If you don't reread something you have read before, of course). So, you should either have a reasonably powerful radio modem to connect with your own ground-based access point, or the cellular modem (read: pay a lot to your cellular provider), or the 802.11 card (read: pay a lot to your 802.11 network owner) or Bluetooth (Tens of meters from the access point or a lot of money to the support network owner).
So I believe it will be only be limited to campuses where either the administration supports the efforts or the student self-organization does the same.
Unfortunately, the authors have not mentioned any network access mechanism on their page.
The second problem is the absence (at least visible) of any keyboard input. So, it becomes the browsing ONLY device - NOT the organizer, NOT the electronic notepad, NOT the programming tool.
The hated MICROS~1 OS, lack of open-source OS, the terrible price, possible eye injury are IMHO secondary compared with these two.