The Invention of USB, 'The Port That Changed Everything' (fastcompany.com) 231
harrymcc shares a Fast Company article about "the generally gnarly process once required to hook up peripherals" in the late 1990s -- and one Intel engineer who saw the need for "one plug to rule them all."
In the olden days, plugging something into your computer -- a mouse, a printer, a hard drive -- required a zoo of cables. Maybe you needed a PS/2 connector or a serial port, the Apple Desktop Bus, or a DIN connector; maybe a parallel port or SCSI or Firewire cable. If you've never heard of those things, and if you have, thank USB.
When it was first released in 1996, the idea was right there in the first phrase: Universal Serial Bus. And to be universal, it had to just work. "The technology that we were replacing, like serial ports, parallel ports, the mouse and keyboard ports, they all required a fair amount of software support, and any time you installed a device, it required multiple reboots and sometimes even opening the box," says Ajay Bhatt, who retired from Intel in 2016. "Our goal was that when you get a device, you plug it in, and it works."
It was at Intel in Oregon where engineers made it work, at Intel where they drummed up the support of an industry that was eager to make PCs easier to use and ship more of them. But it was an initial skeptic that first popularized the standard: in a shock to many geeks in 1998, the Steve Jobs-led Apple released the groundbreaking first iMac as a USB-only machine. The faster speeds of USB 2.0 gave way to new easy-to-use peripherals too, like the flash drive, which helped kill the floppy disk and the Zip drive and CD-Rs. What followed was a parade of stuff you could plug in: disco balls, head massagers, security keys, an infinity of mobile phone chargers. There are now by one count six billion USB devices in the world.
The article includes a thorough oral history of USB's development, and points out there's now also a new reversible Type-C cable design. And USB4, coming later this year, "will be capable of achieving speeds upwards of 40Gbps, which is over 3,000 times faster than the highest speeds of the very first USB."
"Bhatt couldn't have imagined all of that when, as a young engineer at Intel in the early '90s, he was simply trying to install a multimedia card."
When it was first released in 1996, the idea was right there in the first phrase: Universal Serial Bus. And to be universal, it had to just work. "The technology that we were replacing, like serial ports, parallel ports, the mouse and keyboard ports, they all required a fair amount of software support, and any time you installed a device, it required multiple reboots and sometimes even opening the box," says Ajay Bhatt, who retired from Intel in 2016. "Our goal was that when you get a device, you plug it in, and it works."
It was at Intel in Oregon where engineers made it work, at Intel where they drummed up the support of an industry that was eager to make PCs easier to use and ship more of them. But it was an initial skeptic that first popularized the standard: in a shock to many geeks in 1998, the Steve Jobs-led Apple released the groundbreaking first iMac as a USB-only machine. The faster speeds of USB 2.0 gave way to new easy-to-use peripherals too, like the flash drive, which helped kill the floppy disk and the Zip drive and CD-Rs. What followed was a parade of stuff you could plug in: disco balls, head massagers, security keys, an infinity of mobile phone chargers. There are now by one count six billion USB devices in the world.
The article includes a thorough oral history of USB's development, and points out there's now also a new reversible Type-C cable design. And USB4, coming later this year, "will be capable of achieving speeds upwards of 40Gbps, which is over 3,000 times faster than the highest speeds of the very first USB."
"Bhatt couldn't have imagined all of that when, as a young engineer at Intel in the early '90s, he was simply trying to install a multimedia card."
Next up: USB-C (Score:2)
The port that FRIED everything!
Re: (Score:2)
And more revisionist history (Score:3, Funny)
USB did not "change everything". It made some things a bit more convenient. That is also why it took very long to take off. For a long time, it was known as the "Useless Serial Bus".
Re:And more revisionist history (Score:5, Insightful)
USB did "change everything", in that "everything" plugs into USB now. It's one port to rule them all. One does have to give Apple credit for its adoption, as well. PCs hardly had USB ports, and we hardly had anything to plug into them, When Apple made USB their standard peripheral interface. Of course, they were still also pushing IEEE1394 at the same time...
Re: (Score:3)
Re:And more revisionist history (Score:4, Interesting)
It was Apple that went ALL USB - that was the shock.
Only the iMac G3 at CPU speeds of less than 400MHz had USB only. They clearly believed it to be a mistake at the time, since they did not repeat it. Every other mac has had at least two different kinds of peripheral connection.
Re: And more revisionist history (Score:5, Insightful)
The parallel printer port was a botched implementation of a Centronics Printer Port (which had been a standard since the 1950's). It was not actually "designed" to connect to a printer - more like bodged - and there were loads of variants - some of which did not work at all because they were implemented with known dud chips due to corruption. It was certainly not meant to connect to anything else than a printer. IBM in their "infinite wisdom" modified it to make it capable of bi-directional data transfer. People like me wrote software and/or designed FPGAs to hack it.
The reason USB succeeded was because it was actually designed - something tthat has happened all to rarely in the PC world.
Unfortunately the people that make the USB cables are utter numbskulls. It should have been obvious that if the connector is not reversible, the plastic handle should be visibly asymmetrical - like a D-shape in cross section. (Of course the dickheads that designed the D-connector set a precident by screwing that up too).
How do people that stupid end up in engineering design? (And just how sick do you have to be to design the SCART connector?)
Re: (Score:3, Interesting)
The plastic handle is asymmetrical. The USB logo is always on the top. At least it is on cables that follow the standard. Similarly the port is supposed to always be oriented the same way on the computer, so there should be no confusion.
Of course in practice the ports can be any way around, and the cables are mostly correct but not always and the logo imprint is often very hard to see at a glance.
The B type recognizes this as the device could potentially be upside down or on its side. But then USB flash dri
Re: And more revisionist history (Score:5, Interesting)
The plastic handle is asymmetrical. The USB logo is always on the top.
...a detail that is completely invisible in low light conditions (common with connectors in the back of a computer), or even in regular light, since there is no contrast with molded plugs. If I find used USB cables, I avoid getting any made from black plastic because I can color in the logo of any other color with a sharpie marker.
What was needed was a connector that you could identify the right way simply by touch. The molded logo is not sufficient to identify against the manufacturer logo on the other side, and not all plugs have molded logos.
As the saying goes, it takes three attempts to insert a USB A connector successfully.
Re: And more revisionist history (Score:5, Informative)
Circa 1997, most laptops had one USB port... maybe two... but the only things that really used them were mice and keyboards. I know Win95-OSR2 could use USB mice, and sort-of supported keyboards.
The catch w/keyboards was that native-usb keyboards had lag, so everyone used them as ps2 instead. Basically, it used D+ and D- as ps2's clock & data lines, detected it as a ps2-capable device at bootup, and transparently patched them through to the motherboard's ps2 interface. It was physically connected by "usb", but was electronically "ps/2". It caused lots of weird problems to users who didn't understand what was happening behind the scenes... if they hot-plugged the keyboard or connected it through (a/most) hub(s), it reconnected as USB, so suddenly it lagged in games. Connecting two keyboards made things randomly "interesting" as well.
There were also quirks involving ps/2-usb adapters with Model M keyboards. If the usb port worked in "legacy" (dual usb-ps2) mode, they'd usually work... IF the port could/would supply enough power. Ditto, if you had an "active" usb-ps2 adapter w/usb2.0 support (rare, until recently).
In general, Windows does a poor job of explaining WHY a USB device is flaky, esp. noticing when a device randomly connects + disconnects because the port refuses to let it draw as much power as it wants. Atmel's AVR programmers & debuggers (esp. the Dragon) were notorious for this if you connected them to a laptop usb port (as opposed to a slightly-dumb powered hub that didn't try very hard to limit power draw the way it was officially supposed to, so the device was happy).
Windows also had fucked-up handling of default drivers. If you plugged in a device that needed proper drivers without installing them first, Windows would install the default drivers, then subsequently refuse to allow you to replace them with the proper ones. Or it would install the proper drivers for ONE port, but fall back to the default drivers for OTHER ports, so the device would work on one/some port(s), but not others. And the only real way to fix it was to reinstall windows, because if you tried fixing it in safe mode, Windows would undo your work on the next boot.
Re: (Score:3)
Circa 1997, most laptops had one USB port... maybe two
And now we've come full circle, with most laptops having two, maybe three USB ports.
Re: (Score:2)
Re: (Score:2)
At least they come with video out, dedicated power port, microsd reader and headphone jack. Wasn't a MacBook that came with none of these but just with ONE usb-c?
Re: (Score:3)
Those PCs were the Great Eastern of the computing world - the early steamship that had screws, paddlewheels and sails, because at the time nobody was sure which tech would win.
Re: (Score:2)
Re: (Score:2)
It is a very good analogy.
But, the comment that nobody was sure which technology would prevail is not quite correct. That ship was not built as a hedge against changing propulsion technologies but as a means to get from here to there on a huge scale. First, it was at that time by far the most massive ship ever built, and moving it needed power. Neither paddle nor sail alone could do it and still make economical speed. Second, the British trade routes to the far east were too far for most steamers to mak
Re: (Score:2)
Re:And more revisionist history (Score:5, Funny)
I was on the committee overseeing the design of the standard USB plug.
In addition to the 5V 1A power requirement, the necessity to tease and
frustrate users with a plug that, at best, took three tries to successfully
plug a device or cable in was the trickiest part of all. It took hundreds of
hours and dozens of designs iterations to meed the one most important
requirement of the draft -- but we succeeded where hundreds failed.
And for USB 2.0 and USB 3.x, we were able to prevail over common sense
and continue on with the same plug. Somehow the colour identification got
by us, but you can't win 'em all...
CAP === 'contact'
Re: (Score:3)
Re: (Score:2)
Re:And more revisionist history (Score:5, Interesting)
Every PC already had USB ports back then. The only reason why PC people didn't buy USB peripherals was because they were $40 more expensive than the parallel/serial port versions. I remember shopping for a flatbed scanner, only to find the USB version was TWICE the cost of the parallel version, so I got the parallel version, instead.
Mac people didn't have a choice but to buy the expensive versions, because Apple said so. Hence, if you wanted a floppy drive for your iMac, it cost $150, while PC floppy drives were aplenty for under $20. The same problem occurred with SATA CD-ROMs. They were often twice the cost of IDE CD-ROMs, too, which is why everybody continued to buy IDE versions for many, many years.
PC user eagerly adopted USB after a few years when the industry stopped being stupid and the prices came down. Apple had nothing to do with it, other than demand people pay top dollar, as usual.
Re: (Score:2)
The iPod was Apple's first big uptick for their comeback, and it used Firewire rather than USB as Firewire was a superior port at the time.
Firewire was also a superior way to rifle through your systems' memory, being a DMA-connected peer-to-peer bus. A firewire device pretending to be mass storage could read protected memory all the way up until the IOMMU became common, which was distressingly recently. And you also had to pay royalties to Apple to use it, although they were not very high.
Every PC in that era (ca. 2002) already had USB.
Yes, we all had USB, but most of us had nothing plugged into it because USB peripherals cost so much more than the devices which plugged into serial or paral
Re: (Score:2)
Apple was on its last dying gasps when USB came out. The iPod was Apple's first big uptick for their comeback
Actually the colorful all in one iMac was what saved them in the late 1990's. The iPod came later.
Re: (Score:2)
One of those legacy devices I had to leave behind was my early Garmin GPS, which connected with a serial port. I always detested having to transfer maps and tracks over the messy, uncertain serial connection, so I was hoping the forthcoming USB trackers would have a better interface.
Fortunately the iPhone came along before I had to make that buy. A better GPS implementation than I ever had from a single-purpose device. But I was still resigned to having to spring for the pricey navigation option built into
Re: (Score:2)
One of those legacy devices I had to leave behind was my early Garmin GPS, which connected with a serial port. I always detested having to transfer maps and tracks over the messy, uncertain serial connection, so I was hoping the forthcoming USB trackers would have a better interface.
I still have a Garmin Nuvi, 1950 I think. It's a wince device and the PC software has traditionally been garbage. Just having another SD card device on your system was enough to screw it up. This remained true through a couple of complete software changes (not sure if they were full rewrites but garmin actually changed the name and whole behavior of the sync software) but it seems to actually work these days. But yes, the truth is that managing the maps on the device always sucked, even over USB2.
Sadly, Gar
Re:And more revisionist history (Score:5, Insightful)
USB did not "change everything".
Oh sweet Summer child, you have no idea.
The USB port was a godsend, a true breakthrough, and your ignorance flows like a mountain stream in the spring time.
The advent of USB ports most certainly DID change everything. You obviously never dealt with IRQs, serial ports, DMA, and trying to deconflict hardware that just HAD to live on IRQ 7.
Re: And more revisionist history (Score:5, Insightful)
The big problem with IRQs wasn't that you had to set them... it's that there were maybe 5 or 6 usable ones on a typical PC, and 97% of cards could only use 2/9, 3, 4, 5, and/or 7... and usually
only a subset of 2 or 4 of them due to using 1 or 2 jumpers to select it. So you'd have 4 or 5 cards that needed IRQs... one or two (maybe) could use IRQ 10+. Between your serial port & winmodem, IRQ 3 & 4 were taken. The printer port needed IRQ 7 unless you wanted to endure a world of pain configuring every app you used, which left IRQ 5 for the sound card. Then... you bought a scanner with shitty cut-down almost-SCSI card. If you were LUCKY, it could use 2/9. If it needed port 5, your life sucked at that point.
If every card had been long enough & had 4 jumpers to select any IRQ up to 15, it wouldn't have been so bad. It was the limited subset supported by each card that made life hell.
Re: (Score:2)
"The big problem with IRQs wasn't that you had to set them... Between your serial port & winmodem, IRQ 3 & 4 were taken."
The problem was that some of us had a serial mouse, internal modem, sound card (never a winmodem) and an external serial device. Some of the graphics adapters also stole an IRQ. Making them all function simultaneously was quite the PITA.
Until we discovered the STB4COM, 4 serial ports, fully configurable, and it was able to share IRQs. After I bought one of those, I never played "
Re:And more revisionist history (Score:5, Insightful)
The advent of USB ports most certainly DID change everything. You obviously never dealt with IRQs, serial ports, DMA, and trying to deconflict hardware that just HAD to live on IRQ 7.
Most especially, he must never have had to deal with SCSI. You had just spent a month income on a new 10 megabyte external hard disk that sounded like a jet engine as it started up, and SCSI was the only way to hook one up. Squinting through the forty-page Chinglish instruction booklet, you stuck pins in the enclosed voodoo doll and buried a vial of blood from a newborn goat at the crossroads at midnight, just to be the first nerd in your office to be able to type in:
C:> dir d:\
and hope that this time it would work and you would have access to more free space than you had ever imagined could exist.
Re: (Score:2)
A) I am a winter child
B) I never had any problems configuring serial or parallel ports. Even connected self-designed hardware on a couple of occasions. In fact, I still use the serial board on my MBs for remote debugging and debugging my headless systems. USB-serial? Somewhat usable, if you do not mind first scanning all possible IDs this thing may have gotten _this_ time.
Re: (Score:3)
Okay prodigal son. Just because *you* had the magic power doesn't mean USB didn't change the lives of the other 7billlion people on the planet for the better.
In fact, I still use the serial board on my MBs for remote debugging and debugging my headless systems.
Indeed. But that serial port is no longer a conflicting clusterfuck of hardware management to set up. Pop quiz: How many COM ports were part of the original specification, and what did you do when you needed more than your OS supported*? The correct answer is "I don't know" and "I don't know" because no one did. Hardware in the past before the advent o
Re: (Score:3)
The other big improvement was that USB devices could identify themselves enough for the OS to pick the right driver and for software to find them.
There was no standard way to figure out what was connected to a parallel or serial port. Software had to rely on the user telling it what port the device was on and that it was plugged in and ready to go.
Re: (Score:3)
Re:And more revisionist history (Score:5, Funny)
You probably didn't have a SCSI card,
Jesus, SCSI cards...now I'm having flashbacks like a shell-shocked Vietnam vet.
Once you got them all sorted out they worked great, but sorting them out was often a nightmare right out of Dante's Inferno.
Gonna go pop a Xanax, be right back.
Re: (Score:2)
How did you plug in an external drive into your PC before USB? You probably didn't have a SCSI card, and I don't recall there being any other external buses for that.
I did have a SCSI card, you insensitive clod! They also had LPT-connected HDDs, CDROMs, etc. They were slow AF, but they did work. And my very first PC, an IBM PC-1, had an external HDD... on ST-506. It had a 30MB full height Quantum MFM drive on a Xebec full-length 8 bit ISA card, and the upgraded ROM BIOS to be able to boot from HDD. I had MS-DOS 3.0, IIRC. The interface cables left the rear of the machine wrapped in a copper foil shield, which was wrapped in a rubber layer for protection, and passed thro
Re: (Score:2)
My mistake. Been a long time since I had my 3000, I traded it for a BeBox, and traded that for an Indigo. I last had a 1200, which obviously does have IDE. I am now Amigaless. I wish I'd kept the 3000, which IMO is the best of them to have today even if it doesn't have IDE :)
Re: (Score:2)
I had an SCSI card.
Re: (Score:2)
You do know that you could probably a new Bentley for less than a 5MB SCSI drive in 1983.
Re: (Score:3)
Yes, actually it did.
Just not overnight, but nothing changes overnight.
Circa, 1995 - 200, if you wanted to plug in a peripheral, you had to use a smattering of printer-parallel, serial, mouse, keyboard, OS/2 or gameport, telephone, and vga. If you were tech savvy, you added an eide board to support scsi and ethernet. Now, you plug in usb, period. You can use modern video ports and ethernet of course, but what else do you need other than usb in any of its progressive specifications?
Like anything else new,
Apple desktop bus (Score:2)
The first time I got a mac-- my job provided me with one-- I was mystified and amazed that all the things (keyboard, etc) just had one common phone connector. Where were the giant fist size parallel ports and completely incompatible keyboard plugs, and rs232 bare wires? how can one plug do everything?
Re:Apple desktop bus (Score:5, Informative)
The first time I got a mac-- my job provided me with one-- I was mystified and amazed that all the things (keyboard, etc) just had one common phone connector.
I'd by mystified and amazed by that too, because it never happened. You can see from this image [wikipedia.org] that the mouse port, for example, was a DB9. (The RJ connector for the keyboard is on the front of the machine.) They also put DB9 serial ports on the original mac.
The macintosh never had just one kind of port for peripherals and input devices. Never, ever. In the early days it had the keyboard on its own bus, the mouse on its own bus, and peripherals on RS-232. By the time the Macintosh Plus came along they had created LocalTalk, a protocol-managed (AppleTalk protocol) RS-422 current loop serial interface. So the machine had 4-pin DIN ports which were identical to the connectors used with S-Video for the LocalTalk, and it had 8-pin DIN ports for serial. These ports were garbage, because you need 9 pins for full hardware handshaking. They persisted right up until Apple adopted USB.
Where were the giant fist size parallel ports and completely incompatible keyboard plugs, and rs232 bare wires?
On the original Macintosh.
In short, you have no idea what you are on about. The only macs with a "phone connector" were the original 128k, 512k, and Plus machines, they only had one, and all of them use the same DB9 mouse port.
Re: (Score:2)
I'd by mystified and amazed by that too, because it never happened.
Your memory seems to be incomplete, and OP made a mistake. OP is referring to the Apple Desktop Bus, which used a mini-DIN-4 plug, not a phone connector.
ADB was standard on all Macs from the Macintosh II to ~ the G3, and was one port for not quite everything, but it did a lot: mice and keyboards (with daisy chaining), and pretty much all other input devices. It was brilliant (zero configuration, just plug your device in and IT WORKED), and far superior to the mess that existed on the PC.
Re:Apple desktop bus (Score:5, Informative)
The first time I got a mac-- my job provided me with one-- I was mystified and amazed that all the things (keyboard, etc) just had one common phone connector.
I'd by mystified and amazed by that too, because it never happened. You can see from this image that the mouse port, for example, was a DB9.
Your memory seems to be incomplete, and OP made a mistake. OP is referring to the Apple Desktop Bus, which used a mini-DIN-4 plug, not a phone connector.
I am well-acquainted with ADB. It uses the same connector as S-Video, which is convenient when cable-shopping since you don't have to pay the Apple tax on an S-Video connector. My whole point was that OP didn't know what they were talking about.
Besides keyboards and mice, there were also modems that would plug into ADB, from Global Village. I've got a 28.8k one to go with my accelerated Mac SE. (I actually sold this machine, and then tried to make numerous attempts to drop it off at the guy's house, and he flaked on every one. Weird, but then, Mac fans are weird.) ISTR there being ADB hand scanners as well. I'm sure there were also lots of other things with which I'm not familiar.
ADB was standard on all Macs from the Macintosh II to ~ the G3, and was one port for not quite everything, but it did a lot: mice and keyboards (with daisy chaining), and pretty much all other input devices. It was brilliant (zero configuration, just plug your device in and IT WORKED), and far superior to the mess that existed on the PC.
ADB ports existed on every Macintosh until they became translucent. Then, if you were committed to ADB devices, you had to get a ADB to USB gateway. Actually, the original blue and white G3 was the last Macintosh with ADB, the original iMac [in]famously lacking it. It also had a data corruption bug because of Apple's implementation of the cheapass CMD IDE chip they put in there. They were hardly the only ones — it also appeared in some early UltraSparc machines, as well as an assortment of other hardware. Apple never replaced the affected hardware, though, and they eventually deleted their acknowledgement that a problem existed from their public documentation (When they rolled ye olde TechInfo Library into their new Knowledge Base.) Both older and newer TIL articles made it in, but that one mysteriously disappeared. I recycled that machine.
As you can see, I'm more than a little familiar with Apple history, especially the Macintosh kind. The first computers I ever used were an assortment of Apple 2s, the first computer I had in my home with storage was a ][+, a friend had a 512k, and my Junior High had a Mac Plus. My mother had a IIci, which I inherited, and on which I ran netbsd for a while. Besides the SE (with Radius 16 MHz '020 accelerator) I also have the very last model of dome-shaped iMac G4, with the 3dfx GPU. I'd like to get that one to someone who would appreciate it.
TL;DR: Feel free to try to educate me on Macs, but GLWT. I've got the particular opinions I have on Apple not because of ignorance, but because of lengthy and extensive familiarity.
Re: (Score:2)
I am well-acquainted with ADB. It uses the same connector as S-Video, which is convenient when cable-shopping since you don't have to pay the Apple tax on an S-Video connector. My whole point was that OP didn't know what they were talking about.
Just because someone made one minor mistake about the shape of the connector doesn't mean they don't know what they're talking about. The rest of his post was accurate.
Re: (Score:2)
Just because someone made one minor mistake about the shape of the connector doesn't mean they don't know what they're talking about.
Because he got a DIN confused with an RJ, I know that he got original Macs confused with much later ones.
Re: (Score:2)
The CMD IDE chip worked OK in the Ultra 5/10 from SUN, so the hardware bugs could be worked around in software. At least SUN managed it, don't know about Apple.
Apple's suggestion was that one either purchase a PCI card and put their disks on that, or use FWB toolkit to switch the device to PIO mode, because data loss only occurred in UDMA. I ran the data corruption test tool and verified that I had the problem, failed to find a cheap ATA card, and junked the machine. I hadn't paid much for it, and It was taking up space, and I've got the best iMac G4 model anyway. In storage :P
USB was a return to ADB (Score:2)
After the days of the Apple desktop bus the macs slowly accumulated a proliferation of ports like scsi for disks, then firewire. The USB was thus just their plan to get back to the apple desktop bus with s single daisichainable system. Specifically not just a port but a bus. If you think about it this was what distinguished every mac perifrial from the PC world. Most PC plugs were single use. You had a single purpose adpater card on the IMB card slots then a cable to the special device. On a mac, scsi
It significantly changed the way I use my computer (Score:2)
Re: (Score:2)
Re: (Score:2)
Re: (Score:3)
There is a trick that a keyboard engineer can use to get around the boot-or-NKRO problem: present two keyboard interfaces, one flagged as boot-capable (for BIOS) and one not (which the BIOS ignores).
Then when an OS starts up, it tells the boot-interface "I'm not a BIOS", at which time the keyboard switches from talking on the boot-interface to talking on the other.
You very rarely see mass-produced keyboards do this, but it is fairly common in open source firmware for DIY keyboards.
BTW, unrelated to this pro
There were better ports available - like FireWire (Score:2, Insightful)
Comment removed (Score:5, Insightful)
Re: (Score:2)
I don't believe any of this was about license fees, more the fact that if you were to build a computer with Firewire, you still needed to include USB or PS/2 or ADB, whereas if you built it with USB, you didn't need anything else, and didn't need to make the hardware more complex.
Apple did originally charge a licensing fee for IEEE1394, but it was only $1/device, and later lowered to half that. It was not excessive though, in that due to the added complexity which you mentioned, it cost more to make firewire devices than to make usb ones.
USB had a cheaper connector than 1394, had cheaper cables, and was cheaper to implement. All of that was more relevant than the licensing fee. However, 1394 ports remained common until USB2 became prevalent, and even FW400 was faster than USB2 in t
Re: (Score:3)
The margins for commodity electronics are razor-thin, like 1%-5%. If you buy a $100 motherboard, the manufacturer is only making a few dollars profit. $1/port (not per device [cnet.com]) was massive. The licensing fee for other similar technologies was jus
Re: (Score:2)
The margins for commodity electronics are razor-thin, like 1%-5%. If you buy a $100 motherboard, the manufacturer is only making a few dollars profit.
And yet the majority of my motherboards have been around $100... and have had 1394. The one I've got now is the first one in ages that doesn't, and it was slightly more expensive than most.
I think Sony was the only big PC company which licensed it, and they eventually exited the PC business because they couldn't maintain a profit from it.
Sony was the only PC manufacturer to go all-in on it, giving it their own name. They wanted to use it for other devices; it's a pretty snazzy bus. They had a roadmap for it up to 3.2 Gbps using fiber, but obviously that would be silly. If anyone would have done it, though, it would have been Sony. Instead, we got USB3, wh
Re: (Score:2)
USB became more complex (Score:2)
The REAL reason Firewire didn't get far was because Intel and AMD had to make controller chips with support for it otherwise it cost extra to build in a special controller chip for it. USB added more complex chips as it evolved to be more like firewire; it's backward support is why cheapo cables and devices still exist.
Today USB and thunderbolt being totally mismanaged! The two NEED separate plugs instead of the same plug with different tiny icons. The newer USB plugs were not as durable either... I thi
Re: (Score:2)
Or to summarize: Firewire was much more expensive (special chips, more expensive connectors, more expensive cable) and offered nothing, in terms of performance, for the vast majority of USB-connected devices (mice, keyboards, printers, scanners, etc).
The Lightning plug is backwards, though. Apple put the flexible fingers on the phone, and the flat pad on the cable. The parts that wear out over time are the flexible fingers. So you've built-in the wear and tear on the expensive point (note that most Light
Unfortunately it took awhile to really work (Score:2)
Re: (Score:2)
I think you could argue that it wasn't until USB 2 high speed that it was all that useful other than for keyboards and mice. USB 2 high speed made it pretty useful for storage.
But then you could probably argue that it wasn't *truly* useful until USB 3, where you could get SATA-like disk performance on it.
I think with some of the latest speed variations, it could theoretically replace a lot of SAS cabling, too.
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
I'm not aware of this problem, and I used XP from the beginning. Now, if you plugged your keyboard back into the wrong port, it wouldn't work, because the driver wouldn't install if you weren't logged in. That was annoying all right, but you could boot up in safe mode and fix it.
Linux never had this problem, instead it had the problem that you'd move a device to a new port and it would get assigned a different device because it was detected in a different order.
No way, man (Score:2)
If serial ports are hard, turn in your nerd card.
Keyboard (Score:5, Funny)
I'm still using a 25 year old split keyboard that has a DIN connector. The DIN connector is connected to a DIN to PS2 converter. The PS2 converter is connected to a PS2 to USB converter. The USB converter connects to a USB cable, which connects to my PC. And the best thing is, it just works!
Too bad... (Score:2)
... gamers must use ps/2 mice if they want high polling rate. USB still has issues with latency.
Moore's Law and standardization. (Score:5, Insightful)
RS-232 serial was a standard from 1960 and lasted up until the late 1990s. Not a bad run.
Moore's Law made USB possible. The parallel printer bus wasn't much more than a few buffer transistors on the system bus for most computers. And RS-232 UART chips were pretty basic too. USB chips probably had more transistors on the die than an 8008 chip.
I used to own a Digital PDP-8e, built in 1971. It was based on the DEC Omnibus. There were two serial options, an RS-232 card and a current loop card. These cards were about 8X10 inches and covered with transistors.
https://en.wikipedia.org/wiki/... [wikipedia.org]
Re: (Score:2)
Interfaces similar to RS-232 are everywhere, still. They are downright ubiquitous because they are cheap to implement. TTL serial is still commonly used on devices for debugging or configuration, and you still find serial ports on networking equipment and the like for the same reason. Even though every microcontroller of note does I2C and SPI, we still see serial interfaces on all kinds of equipment (like GPS modules.)
Re: (Score:2)
USB needed a number of things to come together to exist.
It uses differential signalling, which existed for a long time before USB but wasn't cheap enough to be mass market. The connectors were fairly small for the time and low cost manufacturing for that kind of thing was relatively new.
USB also required a lot of built in intelligence in the device, which meant microcontrollers or ASICs which were relatively complex compared to earlier protocols like RS232 and PS2. Again, USB arrived just as the got cheap e
Re: (Score:3)
You will still find a IDC-10 header for RS-232 on even the latest PC motherboards in standard ATX or mATX form factor.
You'd just have to get an expansion port bracket with the port on, which connects to the header via a ribbon cable and you will be up and running as if it was 1995.
Be mindful though that some older DTK/Intel brackets use a different pin numbering on the ribbon cable connector than the modern norm.
Remember IRQ conflicts? (Score:5, Informative)
Remember IRQ conflicts? Trying to find an interrupt that wouldn't clash with something else? It sucked.
If you came into computing after the Great IRQ Wars, of the 1980s, consider yourself lucky. There were a strictly limited number of IRQ lines available and deconflicting them could be a nightmare.
I went so far as to buy something called "The Discovery Card" from "Allmicro", which was a card that fit into an ISA slot. It would detect which IRQs were in use, and it boasted on the packaging that it "never reported an IRQ or DMA being used when, in fact, it is not".
It worked like a charm. I still have it even though ISA slots are ancient history. Good times, lol.
The USB port is possibly one of the greatest advances in peripheral connectivity since wire was invented. Although why they couldn't take 5 extra seconds during the design phase and make it a either a keyed port or an "either-way-fits" like USB C is beyond me.
Here's an article about it from September 1993: The Discovery Card [google.com]
Re: (Score:3)
Remember IRQ conflicts? Trying to find an interrupt that wouldn't clash with something else? It sucked.
Thank-you! People complaining about USB need to consider the factors that made it attractive in the first place.
Re: (Score:2)
I remember them. Thanks for reminding me.
Re: (Score:2)
Remember IRQ conflicts? Trying to find an interrupt that wouldn't clash with something else? It sucked.
Yes, but the transition with early, buggy implementations of automatic assignment that didn't mix well with fixed configuration and sometimes even itself led to plug & pray instead of plug & play, because who needs manual overrides. Magic when it worked, an even greater PITA when it didn't. Not that I want to go back, but usually it wasn't that much more advanced than making a list and going through your extension cards to see what address they all used.
Re: (Score:2)
If you check the USB 1.0 spec for the connectors they seem to have though everyone would agree on which way around they would go and the cables would be clearly marked with the USB logo on the top, so there would be no real problem inserting them. I think they very quickly realized that mistake.
Re:Remember IRQ conflicts? (Score:5, Informative)
hah, my first computer came from the shop with the mouse (the port it was connected to rather) and modem both using IRQ 3 -- the net effect was that data would transmit unless the mouse was moving.
I was ~11, and my parents knew nothing about computers. Took quite a while to pin down what was going on, not even 14.4kbps dial-up should have been that slow.
Good times.
Certainly changed security. (Score:2)
One thing that changed was the security of systems. Before you knew what was being plugged in to which port and it had limited capabilities. The capabilities of USB devices are now software defined which makes each and every one a threat to system security. To make things worse, most USB devices have reprogrammable memory that only requires sending it the correct byte string and the new unsigned firmware. This means, malware isn't just in the memory off storage devices but rather it controls the operati
Re: (Score:2)
Then all the attacker has to do is send they keyboard labeled as "free sample keyboard, from your friends at Microsoft!" to some high value target and wait for them to plug the thing in.
But in practice, BadPS/2 is harder to pull off because in order to fool the user, the attack device has to be physically shaped like a keyboard, not (say) a storage device.
Re: (Score:2)
The PS/2 port is also far more limited, you can send a stream of keyboard presses but you can't do anything else. You can't introduce a storage device via PS/2, nor any other kind of random device. There have been some attacks against USB where rogue devices specifically attack bugs in the driver code for certain types of device etc.
A PS/2 port only has a single purpose, it's not a general purpose connector.
USB solved two big problems (Score:2)
2) It gave PC manufacturers a standard to adopt for peripherals.
Yes it wasn't necessarily the best or fastest standard out there. In fact one could say the adoption was given an anticompetitive nudge from Intel. Sadly these thing happen in tech. Working with hardware would still be a pain in the ass for awhile due to driver quirks in Windows, but it was still a step forward compared to dealin
Re: (Score:2)
1) It streamlined management of hardware to the point you didn't have to work within the IRQ framework anymore.
IRQs were only ever a problem on the ISA bus, and even there they weren't a big problem once ISAPnP was invented. You did have to be scrupulous about locking out any IRQs and ports used by legacy hardware in your BIOS, but that was hardly a serious concern. External peripherals never had IRQs, and though the ports they were connected to did, the user didn't have to think about that stuff any time except when adding expansion cards.
2) It gave PC manufacturers a standard to adopt for peripherals.
This was the part that mattered. They even had standard interfaces that didn'
Re: (Score:2)
IRQs were only ever a problem on the ISA bus, and even there they weren't a big problem once ISAPnP was invented. You did have to be scrupulous about locking out any IRQs and ports used by legacy hardware in your BIOS, but that was hardly a serious concern. External peripherals never had IRQs, and though the ports they were connected to did, the user didn't have to think about that stuff any time except when adding expansion cards.
1) Your Delorean is not going far enough back in time. ISAPnP wasn't standard at bios level until 1996ish. By then ISA was on the way out anyway. You're just waiving your hand and saying "Oh this was invented and it's all good" but there's a whole decade of PC use in there (mid 80's to mid 90's) that did not have the benefit of ISAPnP. This is the era that spawned the search for something better than IRQs.
2) External peripherals did indeed use IRQs. They used whichever was assigned to the slow-ass ser
Bad-day rant (Score:2)
No they didn't. Only the PS/2 ports weren't hot-pluggable, so they did require a power cycle, but after that they just worked. Something that I can't say of USB, still today.
Re: (Score:2)
No they didn't. Only the PS/2 ports weren't hot-pluggable, so they did require a power cycle, but after that they just worked. Something that I can't say of USB, still today.
PS/2 mouse ports are hot-pluggable. PS/2 keyboard ports aren't. You can fry your kbd interface by plugging in a keyboard while the system is hot. I've done it myself. Not sure if you could get away with hot-plugging both on Intel, though. Intel boards used to and may still let you plug in your keyboard and mouse backwards (into one another's ports) and have them still work.
the original USB interfaces were very slow and CPU-taxing, so USB went to replace only the slow connections.
Even USB2 would punch your CPU in the nuts back in the early days, when you really loaded up the bus. Then CPUs got faster. AFAIK USB3 d
Re: (Score:2)
This has nothing to do with the narration of a smart, lone engineer at Intel vs the dumb rest of the world that I was objecting to.
No you couldn't. And by the way, 10 megabit ethernet was long obsolete when USB came o
Re: (Score:2)
A firewire keyboard would have been pointless, there were no devices which were firewire only so there was always some other way to connect a keyboard. Firewire was much faster than usb and far less processor intensive when operating so it had a niche for devices where high performance was required.
Because it was aimed at the high end, many low end devices didn't offer it at all and firewire peripherals also tended to target the high end of the market and thus cost more.
I remember those days (Score:2)
One of the most common and widely-sold USB devices were speakers. You could waste your limited USB 1.1 bandwidth by plugging in a speaker instead of just using the audio-out jack of your sound card. Almost every device would claim it wouldn't work right if you used a hub, so unless you were lucky and had ones that did, you used up the two (or four, lucky you!) ports of your system quickly. Windows had support from vendors but Linux was hit-or-miss. Things may have turned out well enough, but it was a far ro
Re: (Score:2)
One of the most common and widely-sold USB devices were speakers.
Many manufacturers do this today with speaker bars. For four years we purchased Lenovos at work and every sound bar had a USB connection. This meant loads of fun when people would plug headphones into the audio-jack and not understand why nothing came out.
You have to go into the speaker system and manually change your output to headphones then, when done, go back in again and manually change sound back to the sound bar.
Not to mention instead
Firewire wasn't really an option before USB (Score:2)
Shape of USB came from Compaq (Score:2, Interesting)
I was in an early USB design meeting with Origin and Compaq engineers in 1994.
Compaq was barely mentioned in this article, but it was Compaq's engineers that tackled the important problems of manufacturability, cost and tolerances of the connector itself. They designed something that was cheaper to make than the existing keyboard plugs and PS/2 connectors despite the fact that the USB connector had a chip in it and the other connectors did not. USB would have been a dud if it had not also been cheaper th
USB good? (Score:2)
Or "the port that almost did nothing" (Score:2)
I had gotten into IT in '94 and remember the first USB ports showing up on new machines. For the first few years USB ports generally were not used for anything. Keyboards were all PS/2 or DIN, mice were all PS/2 or serial. Printers were parallel or serial connected. External modems, serial of course. It really took a while for USB to catch on. Once it did however, yes connecting everything got much, much simpler.
What did USB fix? (Score:2)
It might have reduced the need for specialized cables but that hasn't borne out in practice.I have more crap plugged into USB hubs--as opposed to the motherboard or something plugged into a motherboard I/O slot--with those cables permanently plugged into hubs and with each cable labeled for the particular device that it's meant to be used for. So much for a universal connection to peripherals. Failing to leave cables plugged into means that it's much easier to misplace or even lose a cable. It's an effin' m
Universal? (Score:5, Interesting)
At His His Funeral (Score:3)
When Ajay Bhatt passes away (hopefully many years in the future), during his funeral they'll lower his coffin into the grave, and then have to pull it back up and rotate it 180 degrees before it will fit into the grave.
Re: (Score:3)
Actually, it won't fit on the second try. They'll rotate it 180 again back to original orientation, and *then* it will go in the hole.
Re: (Score:2)
Apple was first to drop all the other ports, though. The original iMac only had USB, ethernet, phone, and mic/headphone ports. 400+MHz versions of the iMac then sprouted a Firewire port, and Apple never tried to do USB-only again. Later machines have USB+Thunderbolt.
Re: (Score:2)
And the PC adapter cards that went with each of those cables.
Re: (Score:2)
nope, protocol not universal either. USB 1.0, 1.1, 2.0, 3.1, 3.2....various power, charging, "on the go", and max length differences... what a mess