The biggest problem Apple would have is getting anyone else to give a damn. I can't see USB going away for a very long time yet.
Agreed, but is that why they've brought Intel in on this technology? If Intel develops/adopts Light Peak (and it works as advertised), then it could move along the same path as USB. In that respect it could be far more successful than Firewire.
So how would this work, in practice? If the interconnect is capable of handling lots of different signal types, do you just put a bunch of Light Peak ports on a machine, so that the user can use them in whatever configuration he or she wants?
Or does the spec allow multiple signal types on a single port simultaneously? So that, for instance, if you had a device that you wanted to make as small as possible, you could give it a single Light Peak port on an internal Light Peak bus which could connect to a hub/router of some sort and drive a variety of external devices i/o devices at once?
The problem with any new i/o protocol is how do you connect to all of the legacy equipment you already own? If I buy a new Mac I don't want to buy all new peripherals or a slew of little dongle adapters all over the place.
USB is still going to be useful for some time in my opinion. It is good for all low power, low bandwidth applications.
Ethernet is just too ubiquitous and inexpensive to be made obsolete by Light Peak.
Seems like a good alternative to HDMI though which is not that common on computers yet anyway and maybe Firewire since that protocol is becoming less popular. As we move toward SSD it might also be suitable as a bus for moving data between storage and memory. And as mentioned, the mobile device syncing seems likely, but I can't really agree with the notion that it will replace everything.
Nothing replaces everything, but we've seen many theoretically indispensable technologies vanish over the past 20 years alone, replaced gradually by other more useful technologies. It certainly can happen.
The problem with any new i/o protocol is how do you connect to all of the legacy equipment you already own? If I buy a new Mac I don't want to buy all new peripherals or a slew of little dongle adapters all over the place.
USB is still going to be useful for some time in my opinion. It is good for all low power, low bandwidth applications.
Ethernet is just too ubiquitous and inexpensive to be made obsolete by Light Peak.
Seems like a good alternative to HDMI though which is not that common on computers yet anyway and maybe Firewire since that protocol is becoming less popular. As we move toward SSD it might also be suitable as a bus for moving data between storage and memory. And as mentioned, the mobile device syncing seems likely, but I can't really agree with the notion that it will replace everything.
I agree. The elephant in the room in all this is what about the infrastructure. What good is having high speed connection if it is twiddling its thumbs because the infrastructure is still stuck in the stone age? This is really a problem in areas that have been relatively poor in the past.
"In addition, Intel said it's working on bundling the optical fiber with copper wire so Light Peak can be used to power devices plugged into the PC, he said."
I checked your reference and it is very interesting. The thing that I like about having a single fiber-based port that can handle all protocols is that it very much simplifies interfacing multiple devices to the computer. Of course it's likely that most laptops won't have enough of these for all the attached devices, so some sort of hub will be required. And all those legacy USB devices won't be going away. There are a couple of options. The hub can have all Light Peak optical ports, and users can run fiber to the USB device where an adapter converts optical to USB. Or the hub can serve as a USB hub married to a Light Peak hub to which existing devices connect with USB cables (the more likely option). For video output, I expect it would be more practical to use an optical cable directly from the computer or hub, and have an adapter to legacy video displays/TVs at the TV end.
Wonder how soon all of this could be cheap enough to reasonably replace firewire/USB/SCSI/all the video cable formats? I suspect that low bandwidth uses such as USB to printers, for example, would be the last to go, with high bandwidth such as HDMI video the first to convert.
I have a Time Capsule and can't figure out why Apple only included a USB port and not a Firewire port. Good practice is to back up regularly off-site. I plug a small portable USB drive in periodically and it takes forever to back up. That is quite inconvenient. If eventually all devices can plug into a single port natively or via inexpensive adapter, that would help all large data transfer tasks.
Of course it's likely that most laptops won't have enough of these for all the attached devices, so some sort of hub will be required.
You mean, like Firewire? Even SCSI didn't require a hub, and that's ancient tech now. Hubs are an artifact of USB. Something that replaces USB should improve on it, not duplicate its deficiencies, yes?
Perhaps if you are part of the race to the bottom crowd but I suspect that the high end will wand the speed long before the price get to where Joe Broke ComputerGeek will even look at it.
there is a long list of technically superior solutions that lost to cheaper good enough competitors. Beta was better than VHS and lost. a lot of people said LD was better than DVD because it wasn't compressed. FW was better than USB but Apple wanted too much money. there were a lot of technically better solutions than TCP/IP and Ethernet but they are history.
But you specifically mentioned iPhone in your post. So streaming large files is a non-issue since for the most part users send small files to the iPhone. There's no sustained writing there.
If you meant thumb drives in general, then perhaps you have a point but you specifically talked about iPhone usage of Light Peak.
And this is all assuming that flash technology will stagnate and not improve by the time Light Peak is out.
The price is a non-issue also since you're the one that brought up speed and not cost. If you're going to add more variables to the mix then, yes, the HDD is cheaper but will never use up anywhere near 1/10th of Light Peak's slowest standard.
if you transfer large files like movies then the performance sucks. i can transfer files in the several hundred MB to 1GB range faster over wifi between my windows 7 computers than to my 3GS
Apple will push technology forward yet again and if the PC manufacturers don't jump on quickly and Microsoft doesn't build support into Windows quickly, Macs are going to be way ahead of the game!
Mmm, I hope we don't end up so way ahead of everyone that we have to backtrack by buying all sorts of expensive adaptors for our Ligh Peak-only Macs. You know how Apple miniDisplayPorts us every now and then, while the rest of the industry bemusedly watches our antics.
I mean, would you expect all Cable Modem and ADSL Router manufacturers to produce cheap Light Peak-compatible units in, say, three months after Apple launches a LP-only Mac lineup?
I'm all for LP, but I want a smooth transition, and Apple's hardware transitions are usually painful for long time users, contrary to the party line.
FW was better than USB but Apple wanted too much money.
Apple initially charged $1.00 per port, but dropped even that negligible fee quickly. You could probably write a book about why some technologies become widely adopted and other don't (as you say, merit often has little to do with it), but I think in the case of Firewire vs. USB, the latter achieved broader adoption because it was developed and pushed by Intel.
As for Beta vs. VHS, that's a complicated story which is frequently over-simplified to make convenient points.
You mean, like Firewire? Even SCSI didn't require a hub, and that's ancient tech now. Hubs are an artifact of USB. Something that replaces USB should improve on it, not duplicate its deficiencies, yes?
Both hubs and daisychaining are valid ways to interconnect and group devices, logically and/or physically. Actually, a SCSI hub would have been a very nice thing to have back in the days.
Both hubs and daisychaining are valid ways to interconnect and group devices, logically and/or physically. Actually, a SCSI hub would have been a very nice thing to have back in the days.
I suppose ideally you get both options. Is that possible with optical? I'm wondering whether something less bulky like a splitter would be feasible.
The problem with any new i/o protocol is how do you connect to all of the legacy equipment you already own? If I buy a new Mac I don't want to buy all new peripherals or a slew of little dongle adapters all over the place.
USB is still going to be useful for some time in my opinion. It is good for all low power, low bandwidth applications.
Ethernet is just too ubiquitous and inexpensive to be made obsolete by Light Peak.
Seems like a good alternative to HDMI though which is not that common on computers yet anyway and maybe Firewire since that protocol is becoming less popular. As we move toward SSD it might also be suitable as a bus for moving data between storage and memory. And as mentioned, the mobile device syncing seems likely, but I can't really agree with the notion that it will replace everything.
When Apple dropped legacy i/o port support in favor of USB on the iMac, it didn't take very long for a great many peripherals to become available. There was some short term confusion and inconvenience (I remember a brief outcry that the iMac "couldn't use a printer" because it had no serial port).
Of course, that was an industry wide transition that had Intel putting their full weight behind USB, introduced into a market that really needed it. Whether or not Light Peak would follow a similar path or not is hard to say, but it wouldn't be unprecedented.
I could see the tech being initially introduced a couple of ways: First, as a docking solution for notebooks, requiring a single slender cable from a multi-port breakout box. Using a single port for multiple devices is always going to require some kind of a breakout box (or, as Dr. Millmoss suggests, a splitter) anyway, and multiple cables from multiple ports are no more elegant or compact, so such a set-up wouldn't really be a big kludge compared to what people are doing now. Of course, even better would be an array of devices that use the Light Peak standard and either a hub or a daisy chain (can't remember, has anything referenced whether or not this spec daisy chains?), but either way you still have to get discreet cables from the computer to discreet devices, at some point, no matter how capable your interconnect.
Secondly I could really see this as a great phone/palmtop docking solution. Basically the same deal as a laptop dock, but bringing vastly more functionality to your iPhone/Touch. With the rapid improvements in hardware, it surely won't be very long before your "phone" could serve as the the brains of a perfectly capable desktop/mobile system, as long as you had easy access to full sized peripherals-- keyboard, monitor and storage and power.
Basically, it sounds as if Light Peak could enable an über-dock that would make your iPhone/Touch an actual laptop/desktop replacement.
You mean, like Firewire? Even SCSI didn't require a hub, and that's ancient tech now. Hubs are an artifact of USB. Something that replaces USB should improve on it, not duplicate its deficiencies, yes?
But in Apple's world, the "hub" is the monitor.
I would expect the firs use of this technology will simply be to replace all the ports on a MacBook with the single Light Peak cable which leads to the Apple Cinema Display that has a bunch of ports on the back as usual. Either that, or the single cable that leads from the monitor to the power brick will put the ports on the power brick.
Ergo: buyers of all-Apple gear will have their single cable "dock" solution and anyone who buys anything else will have to connect it to the back of the monitor or onto the brick under the desk.
there is a long list of technically superior solutions that lost to cheaper good enough competitors. Beta was better than VHS and lost. a lot of people said LD was better than DVD because it wasn't compressed. FW was better than USB but Apple wanted too much money. there were a lot of technically better solutions than TCP/IP and Ethernet but they are history.
Could you supple references to support your statement?
As far as I know optical fibres can't power up a device...so?
where's the power gonna come from intel?
(As for as cable length in the vid, give us all a break...)
Who said that?
But in case you were wondering, "In addition, Intel said it's working on bundling the optical fiber with copper wire so Light Peak can be used to power devices plugged into the PC…" as previously pointed out by badNameErr in post #14, and supported by his link http://news.cnet.com/8301-30685_3-10360047-264.html
Comments
The biggest problem Apple would have is getting anyone else to give a damn. I can't see USB going away for a very long time yet.
Agreed, but is that why they've brought Intel in on this technology? If Intel develops/adopts Light Peak (and it works as advertised), then it could move along the same path as USB. In that respect it could be far more successful than Firewire.
So how would this work, in practice? If the interconnect is capable of handling lots of different signal types, do you just put a bunch of Light Peak ports on a machine, so that the user can use them in whatever configuration he or she wants?
Or does the spec allow multiple signal types on a single port simultaneously? So that, for instance, if you had a device that you wanted to make as small as possible, you could give it a single Light Peak port on an internal Light Peak bus which could connect to a hub/router of some sort and drive a variety of external devices i/o devices at once?
The problem with any new i/o protocol is how do you connect to all of the legacy equipment you already own? If I buy a new Mac I don't want to buy all new peripherals or a slew of little dongle adapters all over the place.
USB is still going to be useful for some time in my opinion. It is good for all low power, low bandwidth applications.
Ethernet is just too ubiquitous and inexpensive to be made obsolete by Light Peak.
Seems like a good alternative to HDMI though which is not that common on computers yet anyway and maybe Firewire since that protocol is becoming less popular. As we move toward SSD it might also be suitable as a bus for moving data between storage and memory. And as mentioned, the mobile device syncing seems likely, but I can't really agree with the notion that it will replace everything.
The problem with any new i/o protocol is how do you connect to all of the legacy equipment you already own? If I buy a new Mac I don't want to buy all new peripherals or a slew of little dongle adapters all over the place.
USB is still going to be useful for some time in my opinion. It is good for all low power, low bandwidth applications.
Ethernet is just too ubiquitous and inexpensive to be made obsolete by Light Peak.
Seems like a good alternative to HDMI though which is not that common on computers yet anyway and maybe Firewire since that protocol is becoming less popular. As we move toward SSD it might also be suitable as a bus for moving data between storage and memory. And as mentioned, the mobile device syncing seems likely, but I can't really agree with the notion that it will replace everything.
I agree. The elephant in the room in all this is what about the infrastructure. What good is having high speed connection if it is twiddling its thumbs because the infrastructure is still stuck in the stone age? This is really a problem in areas that have been relatively poor in the past.
"In addition, Intel said it's working on bundling the optical fiber with copper wire so Light Peak can be used to power devices plugged into the PC, he said."
From: http://news.cnet.com/8301-30685_3-10360047-264.html
Bring it on!
I checked your reference and it is very interesting. The thing that I like about having a single fiber-based port that can handle all protocols is that it very much simplifies interfacing multiple devices to the computer. Of course it's likely that most laptops won't have enough of these for all the attached devices, so some sort of hub will be required. And all those legacy USB devices won't be going away. There are a couple of options. The hub can have all Light Peak optical ports, and users can run fiber to the USB device where an adapter converts optical to USB. Or the hub can serve as a USB hub married to a Light Peak hub to which existing devices connect with USB cables (the more likely option). For video output, I expect it would be more practical to use an optical cable directly from the computer or hub, and have an adapter to legacy video displays/TVs at the TV end.
Wonder how soon all of this could be cheap enough to reasonably replace firewire/USB/SCSI/all the video cable formats? I suspect that low bandwidth uses such as USB to printers, for example, would be the last to go, with high bandwidth such as HDMI video the first to convert.
I have a Time Capsule and can't figure out why Apple only included a USB port and not a Firewire port. Good practice is to back up regularly off-site. I plug a small portable USB drive in periodically and it takes forever to back up. That is quite inconvenient. If eventually all devices can plug into a single port natively or via inexpensive adapter, that would help all large data transfer tasks.
Of course it's likely that most laptops won't have enough of these for all the attached devices, so some sort of hub will be required.
You mean, like Firewire? Even SCSI didn't require a hub, and that's ancient tech now. Hubs are an artifact of USB. Something that replaces USB should improve on it, not duplicate its deficiencies, yes?
Only if it is cheap will people buy into it?
Perhaps if you are part of the race to the bottom crowd but I suspect that the high end will wand the speed long before the price get to where Joe Broke ComputerGeek will even look at it.
there is a long list of technically superior solutions that lost to cheaper good enough competitors. Beta was better than VHS and lost. a lot of people said LD was better than DVD because it wasn't compressed. FW was better than USB but Apple wanted too much money. there were a lot of technically better solutions than TCP/IP and Ethernet but they are history.
But you specifically mentioned iPhone in your post. So streaming large files is a non-issue since for the most part users send small files to the iPhone. There's no sustained writing there.
If you meant thumb drives in general, then perhaps you have a point but you specifically talked about iPhone usage of Light Peak.
And this is all assuming that flash technology will stagnate and not improve by the time Light Peak is out.
The price is a non-issue also since you're the one that brought up speed and not cost. If you're going to add more variables to the mix then, yes, the HDD is cheaper but will never use up anywhere near 1/10th of Light Peak's slowest standard.
if you transfer large files like movies then the performance sucks. i can transfer files in the several hundred MB to 1GB range faster over wifi between my windows 7 computers than to my 3GS
Apple will push technology forward yet again and if the PC manufacturers don't jump on quickly and Microsoft doesn't build support into Windows quickly, Macs are going to be way ahead of the game!
Mmm, I hope we don't end up so way ahead of everyone that we have to backtrack by buying all sorts of expensive adaptors for our Ligh Peak-only Macs. You know how Apple miniDisplayPorts us every now and then, while the rest of the industry bemusedly watches our antics.
I mean, would you expect all Cable Modem and ADSL Router manufacturers to produce cheap Light Peak-compatible units in, say, three months after Apple launches a LP-only Mac lineup?
I'm all for LP, but I want a smooth transition, and Apple's hardware transitions are usually painful for long time users, contrary to the party line.
FW was better than USB but Apple wanted too much money.
Apple initially charged $1.00 per port, but dropped even that negligible fee quickly. You could probably write a book about why some technologies become widely adopted and other don't (as you say, merit often has little to do with it), but I think in the case of Firewire vs. USB, the latter achieved broader adoption because it was developed and pushed by Intel.
As for Beta vs. VHS, that's a complicated story which is frequently over-simplified to make convenient points.
You mean, like Firewire? Even SCSI didn't require a hub, and that's ancient tech now. Hubs are an artifact of USB. Something that replaces USB should improve on it, not duplicate its deficiencies, yes?
Both hubs and daisychaining are valid ways to interconnect and group devices, logically and/or physically. Actually, a SCSI hub would have been a very nice thing to have back in the days.
Both hubs and daisychaining are valid ways to interconnect and group devices, logically and/or physically. Actually, a SCSI hub would have been a very nice thing to have back in the days.
I suppose ideally you get both options. Is that possible with optical? I'm wondering whether something less bulky like a splitter would be feasible.
The problem with any new i/o protocol is how do you connect to all of the legacy equipment you already own? If I buy a new Mac I don't want to buy all new peripherals or a slew of little dongle adapters all over the place.
USB is still going to be useful for some time in my opinion. It is good for all low power, low bandwidth applications.
Ethernet is just too ubiquitous and inexpensive to be made obsolete by Light Peak.
Seems like a good alternative to HDMI though which is not that common on computers yet anyway and maybe Firewire since that protocol is becoming less popular. As we move toward SSD it might also be suitable as a bus for moving data between storage and memory. And as mentioned, the mobile device syncing seems likely, but I can't really agree with the notion that it will replace everything.
When Apple dropped legacy i/o port support in favor of USB on the iMac, it didn't take very long for a great many peripherals to become available. There was some short term confusion and inconvenience (I remember a brief outcry that the iMac "couldn't use a printer" because it had no serial port).
Of course, that was an industry wide transition that had Intel putting their full weight behind USB, introduced into a market that really needed it. Whether or not Light Peak would follow a similar path or not is hard to say, but it wouldn't be unprecedented.
I could see the tech being initially introduced a couple of ways: First, as a docking solution for notebooks, requiring a single slender cable from a multi-port breakout box. Using a single port for multiple devices is always going to require some kind of a breakout box (or, as Dr. Millmoss suggests, a splitter) anyway, and multiple cables from multiple ports are no more elegant or compact, so such a set-up wouldn't really be a big kludge compared to what people are doing now. Of course, even better would be an array of devices that use the Light Peak standard and either a hub or a daisy chain (can't remember, has anything referenced whether or not this spec daisy chains?), but either way you still have to get discreet cables from the computer to discreet devices, at some point, no matter how capable your interconnect.
Secondly I could really see this as a great phone/palmtop docking solution. Basically the same deal as a laptop dock, but bringing vastly more functionality to your iPhone/Touch. With the rapid improvements in hardware, it surely won't be very long before your "phone" could serve as the the brains of a perfectly capable desktop/mobile system, as long as you had easy access to full sized peripherals-- keyboard, monitor and storage and power.
Basically, it sounds as if Light Peak could enable an über-dock that would make your iPhone/Touch an actual laptop/desktop replacement.
You mean, like Firewire? Even SCSI didn't require a hub, and that's ancient tech now. Hubs are an artifact of USB. Something that replaces USB should improve on it, not duplicate its deficiencies, yes?
But in Apple's world, the "hub" is the monitor.
I would expect the firs use of this technology will simply be to replace all the ports on a MacBook with the single Light Peak cable which leads to the Apple Cinema Display that has a bunch of ports on the back as usual. Either that, or the single cable that leads from the monitor to the power brick will put the ports on the power brick.
Ergo: buyers of all-Apple gear will have their single cable "dock" solution and anyone who buys anything else will have to connect it to the back of the monitor or onto the brick under the desk.
where is the power gonna come from?
As far as I know optical fibres can't power up a device...so?
where's the power gonna come from intel?
(As for as cable length in the vid, give us all a break...)
there is a long list of technically superior solutions that lost to cheaper good enough competitors. Beta was better than VHS and lost. a lot of people said LD was better than DVD because it wasn't compressed. FW was better than USB but Apple wanted too much money. there were a lot of technically better solutions than TCP/IP and Ethernet but they are history.
Could you supple references to support your statement?
Thank you.
VGA
Mini VGA
DVI
Mini DVI
Micro DVI
Display Port
I may have missed some.
How could anyone forget ADC?
like other people said:
where is the power gonna come from?
As far as I know optical fibres can't power up a device...so?
where's the power gonna come from intel?
(As for as cable length in the vid, give us all a break...)
Who said that?
But in case you were wondering, "In addition, Intel said it's working on bundling the optical fiber with copper wire so Light Peak can be used to power devices plugged into the PC…" as previously pointed out by badNameErr in post #14, and supported by his link http://news.cnet.com/8301-30685_3-10360047-264.html