Slashdot Mirror

USB 3.0 or *something faster* will be required for devices this large in portable storage capacity.. USB 2.0 is ~480Mbps (theoretical max) and it would take forever to transfer a terabyte over USB 2.0.
http://www.usb.org/usb30
http://www.eetimes.com/news/latest/showArticle.jhtml?articleID=201807389

You're right. If only there was a new, faster USB standard that would be able to take advantage of these new data rates. They could call it "USB 3.0", or "USB SuperSpeed" or something. Oh Wait...

The USB spec (page 178) specifies 500 milliamps of current. At 5VDC, that's 2.5 watts of power.

And, if they do like they did with USB 2.0, they will make a point of separating out the USB version from the speed. i.e. When referring to the standard, you call it "USB 1.1" or "USB 2.0", no matter what speed. And when referring to the speed, you call it "USB Low Speed", "USB Full Speed", or "USB Hi-Speed". I imagine this will continue to hold true, and it will be "USB 3.0" when referring to USB 3.0-compliant hardware, and "SuperSpeed USB" when referring to the new speed.

Where do I get the spec for USB, EHCI or a device class?

P.S. These things aren't that hard to find for yourself. You can almost always use Google and/or Wikipedia to find the Web site of the company or consortium that defines a specification, or a page that explains the licensing.

Umm, you do realize that USB 1.0 and 2.0 use the exact same cables and connectors, don't you?

Just asking, because you sound too serious to be joking.

Did they make a P2P version so that I don't need a computer to connect a camera to a hard drive and have it work?

Yes, they did. Several years ago, in fact. It's called USB On the Go

I checked the USB 2.0 spec, and it actually only requires only twist per 60-80 mm... much less than Ethernet CAT 6! So my ugly home-made cable is actually within spec or nearly so ;-)

They are both part of the standard though mini-B is being deprecated in favor of micro-A.

Protected CDs can't use the CDDA logo because some standalone players (including, ironically, my late model Sony ES) can't play them.

They are both part of the standard though mini-B is being deprecated in favor of micro-A.

Protected CDs can't use the CDDA logo because some standalone players (including, ironically, my late model Sony ES) can't play them.

Those things all understand more or less the same command set. There's a "USB Video Device Class Specification". The current revision is 1.1. This Linux driver effort isn't some huge collection of drivers, which would be silly and a maintenance headache. It's a generic driver.

A few years back, I wrote a generic Firewire camera driver for QNX. It's really not that hard. These things self-identify when they hot-plug, and you can read their configuration and capabilities over the bus interface. The USB camera spec, unlike the FireWire camera spec, has some annoying Microsoft-specific stuff in it, but it's not all that pervasive.

I would love to see a standard that would allow 100 ft cables for both video & USB.

Since the max USB cable length is 5 meters, that's a bit tricky. You'll need to string it along between 5 USB hubs -- that's 30m of cable, or about 98 feet. Each hub would need to be powered somehow (i.e., wall wart), which would ruin most of the convenience.

If you just want to hook up a DVD drive, try Firewire. You can have Firewire cable runs of 100 meters using cat-5.

Of course a simpler solution would be to just get a quieter computer, or try to silence the one you have.

HDMI claims cables of 15 meters, with longer cables to come. If you get a Firewire drive, HDMI is going to be the limiting factor.

Standard MIDI interfaces are slowly dying out, since the USB specification includes a MIDI device class. Any device implementing the MIDI profile can be used with any operating system that supports it. This works beautifully on OS X (as you'd expect, since most MIDI users went the Atari ST to Mac route and ignored the PC). Support is under development for Linux, but I don't know what the current status is.

Consider that you don't need a special driver for a particular brand of ATAPI CD-ROM drive, or for a particular USB Mass Storage Device. Heck, Windows has USB class drivers for Bluetooth devices, smart-card devices, hubs, HIDs (keyboards, mice, CueCats and such), mass storage devices, printers, PTP-protocol scanners and cameras, audio devices, modems and video devices. Linux has a variety of supported class drivers as well. There are, of course, more classes, and that's all just for USB devices.

Sure, there are a lot of corner cases and pathological hardware--I think video cards are the best example--but it's entirely possible and indeed desirable to support all kinds of devices in the kernel. Even if sometimes we have to say goodbye to one of them, it was worth it to have them around.

Then why didn't USB Implementers Forum standardize on something similar to the "X" notation popularized by CD-ROM, -R, and -RW? Here, "1X" is 1.5 Mbit/s, enough to transmit 150 KiB per second of mode 1 CD-ROM along with protocol overhead. This would make a "full speed" device up to 8X, while a "hi-speed" device goes up to 320X.

Is a wireless USB device still a USB device the answer is yes, except that instead of using a cable, it is a wireless conenction. The computer still utilizes the USB drivers/software stack except that at the lowest levels instead of cabling, its wireless. Existing USB devices and PCs can be retrofitted with a small USB wireless transcivever that plugs into a USB port, likewise, an existing USB device (Printer, camera, external disk, etc.) can also be adapted for use with USB Wireless. Moving forward, look for wireless USB to be built into PCs as well as other devices. In terms of performance, since wireless USB is based on USB 2.0 it has similar if not same performance and distance constraints/capabilties as cable based versions. Check out http://www.usb.org/developers/wusb for more info. Why wireless USB vs. say Bluetooth, as others have posted, bluetooth is very slow compared to USB and range limited. Why Wireless USB vs. say WiFi? The comparison should be along the lines of USB vs. Ethernet, and then, wireless USB vs. WiFi for IP. Hope that helps to fill in some of the blanks. Cheers GS

• USB Bit Whacker
• Arduino
• (you can buy the above two from SparkFun if you don't feel like DIY)
• LabJack
• various FTDI-based devices from many companies

This product is a UWB Wireless USB Hub, but I don't see any indication on the web site that this product complies with Certified Wireless USB.
There has been a long running battle between two opposing UWB groups that each have incompatible UWB implementations. If you're going to get any Wireless USB products, make sure to look for the product's compliance to "Certified Wireless USB". More information at Certified Wireless USB. Certified Wireless USB is built on top of the WiMedia UWB standard. The WiMedia UWB platform is an ECMA standard, and certification by other standards bodies has been in the works.
Disclaimer: Until this past April, I worked for one of the Certified Wireless USB developers & promoters, and I was deeply involved in this standardization effort.

So, Wireless USB is also using WiMedia's UWB technology. Does this bring them any closer to convergence, or just mean that they're two different protocols using the same bandwidth and the same spectrum-hopping technology?

USB 2.0 Spec download

Read 7.2 of usb_20.pdf
Devices default to low-power and as such can only pull one "unit load" (100mA.) If a device is configured to be High-power, it can draw five "unit loads" (500mA.) This is at 5V, so it will supply about 2.5W to a device (if you are lucky.)

IEEE 802.3af (better known as Power over Ethernet) would be a much better solution for the applications you mentioned (routers and hubs,) as well as others including IP phones, IP security cameras, and RFID tag readers. It provides power at 48V and around 15W. More info can be found here.

There is a standard. (see http://www.usb.org/developers/wusb) My company is one of many that is racing to implement it. I would not use this device at is unlikely to be compatible with WUSB.

Uhm. Last I checked, Apple *adopted* USB. They didn't invent it. If you'll kindly read http://www.usb.org/about, you'll note that Apple isn't even on the board of directors of the USB standards organization.

--S

Not really an important story. Just a filler technology until true wireless USB arrives in a couple of years. It will be built straight into the motherbaord and be far more secure.

http://www.usb.org/developers/wusb/

Look here:

http://www.usb.org/about/faq/ans5#q4

It describes the maximum length of cable and the next question describes what to do to extend the range.

Just wait for the real thing. True Wireless USB. It is without the HUB Concept. Works on UWB(MBOA MAC). Devices are scheduled by Mid 2006. Just Wait !!!!!
http://www.usb.org/developers/wusb/

Point by point (skipping some that I don't have an argument with):

FAT isn't a camera standard. It's the filesystem that is used on most memory cards. It's a default standard because it's simple enough to fit within that small space and work in that price of hardware and it works well with Windows. It tells the camera nothing about how to communicate with the computer. Actually a lot of newer cameras are becoming standardized in how they communicate (showing up as a removable drive) which is good. There is no reason for any camera not to follow this standard for still photos. Streaming cameras could use some work on a standardized interface though.

FAT is a de facto standard now, which is how standards should be created IMHO. Even my 4.5 year old digital camera hooks up just fine as a USB storage device. There's nothing more that a camera needs to do in terms of PC interaction, and almost every decent camera can do this. And even if your camera doesn't, you should buy a camera that uses removable media so you can always pull the card out and use it in a standard reader instead.

VESA is a loose standard. It's more of a video API than a real driver. It doesn't support required modern features. Video cards need basically the same information to do what they do. Most software already uses standard API's such as DirectX or OpenGL to access these cards. There is no reason that the drivers can't be as standardized as these API's. Doing so might raise the price slightly by moving processing that goes on on the CPU to the card but I doubt it'd be much because the processors of modern video cards are already extremely powerful and flexible.

The problem is that drivers are the abstraction layer for the hardware. By moving to a standard like this, you're essentially saying that your abstraction layer needs to be built into the hardware itself (like nVidia already does to support their unified driver model, where a single driver binary will work for everything from an ancient TNT to the latest GeForce 78xx). Depending on the complexity, you're talking about a lot of added cost for an already expensive product (if you buy bleeding edge, anyway), and don't think for a second that the manufacturers won't pass that cost on to the consumer. At least for video cards, I think the model we already have (standard APIs like DirectX, SDL, and OpenGL to abstract away the hardware) is the right way to go. It may not be 100% perfect, but it's a damn sight better than the old DOS days where every video card had its own special interface.

USB has nothing to with joysticks other than being the generic method by which they connect. It doesn't tell the computer how to understand the joysticks input. Joysticks are fairly standardized but they are growing less standardized and that is bad for stability. Do you want your game to crash in the middle of a firefight because it has a non-standard driver that doesn't work with your video card? Do joysticks even need to be recognized as different from a mouse? Standard mouse drivers understand multiple axis at a high precision and many buttons. What else does a joystick need?

You missed three important letters -- HID, or Human Interface Device. Theoretically, all input devices (keyboards, mice, joysticks, even webcams) should adhere to the HID spec and at least provide basic functionality with a proper generic USB HID implementation. That they don't is a problem, but the standard already exists. Perhaps this is a case where the standard is "bad" (too strict, too vague, not extensible, etc).

Modems were mostly standardized until some moron invented the concept of a winmodem. Winmodems NEVER work well. They offload hardware processing to the CPU which has the mixed effect of slowing the computer and making the modem less reliable. Brilliant. T

there's really nothing "standard" about pre-installed drivers, they simply provide a basic instruction set.

If this basic instruction set for human interface devices is documented by the USB Implementers Forum, then it's as good as standard. The Xbox controller protocol, on the other hand, "embraces and extends" USB HID, incidentally making conforming USB HID device drivers incompatible.

imho, the new controller is sufficiently complicated that it definitely requires a new driver to be fully usable. MS has comitted to providing the necessary driver for Windows (obviously) for free.

So what will it take to run the driver in something like ReactOS?

• download images stored in your camera onto your computer (camera is slave)
  
• control a USB printer from the camera (camera is master)
  

I'm not sure what you mean:

- Sata Specifications

- USB

- Anyone can write a raid controller, we all understand the RAID levels and what they mean. What varies is how to program the chips to implement it (software interface).

- I will give you a little on the north bridge, the FSB is NOT standard and I think AMD's is different. But the rest of the north bridge contains well known PC bits: memory controllers, interrupt controllers, PCI(X|E) host bridges, power management, etc. etc. I could build one from scratch if I had a few million dollars laying around.

- The South Bridge could be implemented by anyone, it's got no secrets.

- As for Video, the AGP specification was well documented and freely available (I downloaded it when I was in college for free, I can't recally where). Going forward PCI-Express will be used, they simply use x8/x16 connectors. I think Apple uses the same.

Or did you mean software interfaces are not standardized? It's not as bad as you imply, but it is getting worse. Getting back on topic, Apple could take PC hardware as it stands and require vendors to support a common apple interface (they seem opposed to the device driver model, and I agree with them on that). But then us Linux types could just assume that interface and copy it, without paying the Apple tithe.

I find it more likely they will build a Mac using mostly commodity components (processors, etc.) to get their costs down and leverage that market, while changing the north bridge, which is what makes a PC a PC, and without which we can't really write much software. That combined with new microcode will make a totally incompatible, totally Apple system, just at a lower price. I for one hope they do, only because it may make PCs better.

UWB is the physically layer and has been designed to take multiple link layers- USB, IEEE1394 and IP. More information can be found at
http://www.multibandofdm.org/
http://www.usb.org/wusb/home

Yes, 31200bps really is limiting. Or rather, it would be if there wasn't MIDI over USB (PDF) and MIDI over IEEE-1394 (Firewire) (PDF).

Actually, there is, its called "USB On-the-go"- see http://www.usb.org/developers/onthego/ which is what this device is an example of, allowing usb devices to negotiate connections without the use of a pc

The USB folk's naming and packaging recommendations actually discourage the use of "USB 2.0" since it is confusing as heck ... but I agree with the parent that they kinda created this monster by saying that there is a "Lo-speed" and "Full-speed" USB 2.0 that are the same speed as USB1.x ... so most consumers (myself included) see USB 2.0 and unless we look carefully for "Hi-Speed", then things aren't any faster than 1.x ... which is an issue for still photography and a BIG issue for video.

BTW, have we ever seen a first post that has been so informative - mod the parent to +10 - nice work roman mir

Will Rockbox be able to give the H1xx series USB on-the-go like the H3xx series have, I wonder?

Does anyone know if you need special USB hardware to support USB host operations (not found in the H1xx series), or is it just in the driver?

I believe that "ultra-fast" is a technical term.

I think you're thinking of the "Hi-Speed" label for USB.

Speed: The USB 2.0 Hi-Speed FAQ tells us that the maximum speed of USB 2.0 is 480 Mbit/s. The maximum speed of parallel ATA is 133 MByte/s = 1064 MBit/s, plus it does not have the "ATA over USB" protocol overhead. Serial ATA does 150 MByte/s = 1200 MBit/s, IIRC. The ethernet interface of the device supports 100 MBit/s. Modern harddisks can not deliver 133 MByte/s = 1064 MByte/s, but they become faster every day. Flash memory can be that fast, at least for reading.

Power: Each USB port must be able to deliver 5V @ 0.5A.

Now do the maths: You can see that already a single USB device can deliver more data than the ethernet port could transport. The CPU (according to http://www.batbox.org/nslu2-linux.html) is an XScale CPU with 131.48 BogoMIPS, roughly comparable in Performance to a slow Pentium II. I'm sure it can't handle much more than 100 MBit/s Ethernet and two USB 2.0 ports.

Adding a second USB port is convenient to copy data directly between USB devices, e.g. for backup or upgrade purposes. But adding more USB ports costs 0.5A per port for the PSU, making it much more inefficient for each added port.

Tux2000

Don't forget that the main point of USB 2.0 is dun-na-na-na!: Marketing!!!

No, the main point of USB2.0 is speed .
And you, sir, are a troll and an irritant, like all your mac fan-boy ilk. Go away.

Hm. Apple PowerMac G5:

* HyperTransport
* PCI-X / AGP
* DDR SDRAM
* S-ATA
* Gigabit Ethernet
* IEEE 1394b a.k.a. Firewire 800
* USB 2.0

So, tell me, which of these, which will be the only interfaces that you can sanely use, is proprietary?

In the PC world, anything other than an Opteron machine can compare in specs.

Hm. Apple PowerMac G5:

* HyperTransport
* PCI-X / AGP
* DDR SDRAM
* S-ATA
* Gigabit Ethernet
* IEEE 1394b a.k.a. Firewire 800
* USB 2.0

So, tell me, which of these, which will be the only interfaces that you can sanely use, is proprietary?

In the PC world, anything other than an Opteron machine can compare in specs.

Um. Your numbers are a bit misleading, although that of course doesn't change the value of your story. USB 1.1, which I assume you were using since you're talking about "the past" and USB 2.0 is rather recent, has a max bandwidth of 12 megabits/second. That's 1.5 megabytes/second, of course. PCI, on the other hand, starts out as a 33 MHz bus that is 32 bits wide, for a maximum theoretical bandwidth of 133 megabytes/second. Thus, "raw PCI" is roughly 90 times faster than USB 1.1. Just wanted to point that out, since factual errors of this nature tend to annou the anal geek within. ;^)

You've got it backwards... make sure you get a "hi speed" reader, since 2.0 doesn't imply 480Mbps--it's just a version number. However, "hi speed" does imply 2.0, since USB 1.x didn't support 480Mbps transfer rates. See the USB Implementor's Forum's article on nomenclature.

The problems arise (in terms of USB sharing) when you stick a 1.1 device into a 2.0 chain, which does throttle down the whole bus. Even 1.1 devices have had the ability to share the [1.1] bus at the same time, since I've used a 1.1 keyboard, 1.1 mouse and 1.1 scanner at the same time without difficulty.

Realistically, this isn't an issue of USB 2.0 products needing labelling as low, full or high speed since the consumer is going to tell the difference pretty quickly on devices that could really use the extra speed (hard drives, CD-ROMs, MP3 jukeboxes, etc.) and return their hardware if it isn't satisfactory. The issue is that USB 1.1 stock has been retroactively called/labeled USB 2 on the flimsy theory that since the new spec is backward compatible, USB is USB (somewhat like 16-bit Windows 3.1 software being labeled "for Windows 95" since it can indeed run on that OS).

Unfortunately, this flimsy theory is probably supportable in court and I doubt anything is going to come of the problem except that retailers and manufacturers will be able to clear out their old stock and henceforth produce and sell "real" USB 2.0 devices.

USB supports isochronous data streams. It has since version 1.0. See this.

-h-

Look at them here. The USB High Speed devices have a logo that says "High Speed". In order to use that logo, the device MUST support USB High Speed.

The fact that USB 2.0 supports the two slower speeds doesn't mean that manufacturers are being misleading when they label a mouse as being USB compatible. It doesn't mean that when you put that mouse on your USB 2.0 bus that all of your devices will run at USB Low Speed. All it means is that the USBIF didn't do as robust a job in defining the marketing parameters of the specification.

All in all, there's nothing to see here, move along.

-h-

The USB-IF is a nonprofit industry organization. We do not and cannot control how manufacturers label their products.

The USB-IF is a nonprofit industry organization. We do not and cannot control how manufacturers label their products.

The USB-IF is a nonprofit industry organization. We do not and cannot control how manufacturers label their products.

It's obvious that Microsoft's month for secuirty wasn't enough (after 20 years of feature creep, we only get a month for?).

I've read the TCPA specs and it's not a bad idea in a commercial & some home environments as long as you can turn it off so you can develop code or run someone elses' if you choose to (as opposed to sneak-ware like Gator). I have two computers at work & wouldn't mind if the one running email were "locked down" to keep corporate IS from losing their minds for every MSBlaster/Fizzer/MSwormoftheweek as long as they leave my "programming" box alone (where's the checkbox for USB compliance suite on their audit checklist???).

At home, it would be an advantage to have two copies of Windows installed- one that lets me play & one that I run video editing on & only runs trusted code.

By mixing DRM in, Intel, Microsoft, h-p & others are guaranteeing that

1) there will be strong opposition to the tech in the form of boycotts (see the anti-RFID flak) & cracks

2) the tech will be weakened to serve its new DRM masters by complexity not needed for simple rogue code protection