USB Inventor Regrets Making Them So Difficult To Plug in Correctly (mashable.com) 289
An anonymous reader shares a report: While plugging plug a mouse, a phone, or a thumb drive into your computer, you try to stick the USB into its slot, only to find it stopping prematurely. You flip it around, but it still won't go in. So you flip it back to the original position and it slides in without a hitch. We've all been there, and the inventor of the USB sees our pain. Ajay Bhatt, the leader behind the IBM team that gave us the USB in the mid-'90s, revealed in an interview with NPR Friday that he is well aware of the annoyances the public has with USB, or Universal Serial Bus, but there's a reason it's designed the way it is.
"The biggest annoyance is reversibility," Bhatt told NPR. For outsiders, it seems like designing the USB so it can be reversible would be an easy fix to everyone's problems, so no matter which way you stick it in it's a success. Bhatt told NPR that would have doubled the cost of the technology, requiring double the wires and circuits. Another option that the Intel team floated was a round design, but that would have been even more difficult to plug in correctly. Although the rectangle design we all know was ultimately chosen and adopted by pretty much every hardware manufacturer since Apple first put USB ports into its computers in 1998, Bhatt acknowledges that there may have been a better way. "In hindsight, based on all the experiences that we all had, of course it was not as easy as it should be," Bhatt said.
"The biggest annoyance is reversibility," Bhatt told NPR. For outsiders, it seems like designing the USB so it can be reversible would be an easy fix to everyone's problems, so no matter which way you stick it in it's a success. Bhatt told NPR that would have doubled the cost of the technology, requiring double the wires and circuits. Another option that the Intel team floated was a round design, but that would have been even more difficult to plug in correctly. Although the rectangle design we all know was ultimately chosen and adopted by pretty much every hardware manufacturer since Apple first put USB ports into its computers in 1998, Bhatt acknowledges that there may have been a better way. "In hindsight, based on all the experiences that we all had, of course it was not as easy as it should be," Bhatt said.
So IBM screwed this up? Why am I not surprised.... (Score:2)
Sure, it is a difficult problem, but given the number of instances this was expected to be deployed in, you would expect a bit more care.
Re: (Score:3)
Re:So IBM screwed this up? Why am I not surprised. (Score:5, Insightful)
“A committee is a life form with six or more legs and no brain.”
Robert A. Heinlein
Re: (Score:3, Interesting)
Re: (Score:2)
The "problem" have been solved before by using round connectors, like the 3.5mm stereo connector. How many times have you had a problem plugging in a headphone jack unless you were totally wasted or blindfolded?
Of course then someone would have plugged in their headset directly into the USB port. But that would have been resolved with a different size of the jack and plug.
Re: (Score:2)
Which Apple would have promptly abandoned in favor of its own connector or doing away with the connection like they've done on their phones.
Re: (Score:2)
USB has 4 wires including 5V and ground. You could make a connector and plug with four contacts, but you run the risk of shorting the 5V or ground to one of the other connections when you insert or remove the plug. I think.
Re: (Score:3)
Put the ground on the tip and the +5v on the outer ring so it makes contact last. No accidental power on a data line nor can bus-powered devices be in a "running but unable to communicate" state due to a loose cord.
Re: (Score:3)
Put the ground on the tip and the +5v on the outer ring so it makes contact last.
You have it backwards. The tip will briefly contact every connection going into the jack, so you do not want power on any of them. The thing it contacts last should have the +5. The normal ground connection should be ground. Always.
But the solution is much easier than that. It doesn't take "double the wires and circuits". You make the jack with one set of contacts on one side and the same on the other -- two "boards" with a space in between. The plug is a single board with one set of contacts that can con
Re: (Score:2)
There's this shape called an isosceles triangle...
Re: (Score:2)
Re:So IBM screwed this up? Why am I not surprised. (Score:4, Insightful)
They didn't screw up. They chose the most cost-effective solution.
Making it non-reversible wouldn’t have been such a problem had they made the proper polarity obvious by design of both the plugs and sockets. The best design of this kind would be one that would be obvious to a finger when the socket is out of sight in the back of a device.
Re:So IBM screwed this up? Why am I not surprised. (Score:5, Insightful)
Indeed. And when you look at, for example, Display Port or HDMI (full size), you can see that this can be done right for a new design.
Re: (Score:3)
Does that really help though? Whenever I need to plug a DP cable into the back of a monitor (without looking), I end up doing the same USB switcheroo thing because the differences aren't pronounced enough to tell which side is which immediately. It does help a bit when you can both the plug and socket though, since with USB you can't even easily tell which way the plug you're holding is oriented.
Re: (Score:2)
Re: (Score:3)
Re:So IBM screwed this up? Why am I not surprised. (Score:5, Insightful)
You are right; further, as someone who remembers PS/2, serial, parallel, ADB, SCSI, etc.. I have to say "thank you" to Bhatt and colleagues who created USB. Its been a long time coming, but current iterations are close to "one connector to rule them all". Non trivial problem, and the original USB served us well on the path to panacea.. I don't see a screw up here.
USB is screwed up though. There does not appear to be a retained-connector design for the system. USB is fine for devices that are hot-plugged often, but sucks rocks in commercial settings where cables benefit from mechanical latching to retain them in their sockets.
DB9/15/25 achieved this with screws. Centronics did it with hinged clips on the sides of the socket connector. Micro-Centronics and higher pin-count SCSI did it with push-clips on the sides of the plug connector. The Registered Jack standard does it with a tab, albeit a bad design that breaks fairly easily.
USB, Firewire, and many other standards suck because one cannot latch the device into its socket where it doesn't pull out. This leaves such devices ripe for casual tampering or accidental disconnection of important devices. Had they chosen something that works like those micro-Centronics connectors, or like fiber LC, or even like the retainers used for SATA then it might have worked well, but instead they went for no real locking system, and problems have nagged us for a couple of decades now.
Re: (Score:2)
On all of my USB cords the USB symbol faces up when plugged in correctly. The only issue I see is if the port is installed incorrectly or if you don't have an orientation for "up" like if you installed them and turned them sideways.
Re: (Score:2)
Indeed. But the worst option that still works was chosen.
Re: (Score:2)
Good point. Thing that everybody is forgetting is that PC's were still a niche thing, the big "everyone has a PC" surge really didn't start catching on in a lot of countries until around 1998 or so, it missed some countries too. Even then, most people looked at the entire thing like it was some gigantic esoteric mess that only a really smart person would want to stick their fingers into once it was all in working condition. So now everyone gets to look backwards, and forget that a lot of design decisions
Re: (Score:2)
I'm so very sorry for being off by a year and change princess. I'll try harder next time.
Re: (Score:2)
The PC was released in 1983 and USB 1.0 was released 16 years later in 1996.
1983+16=1999. Something doesn't add up in your argument.
Bad design all around. (Score:2)
Re: (Score:2)
My son actually fried my computer thanks to a broken USB slot. The plastic tongue had come off leaving just the metal prongs. My son tried to plug in my computer (without looking at what he was doing) and the power cord went into the USB slot by accident. A metal prong in the USB slot entered the power cord's hollow center and made a connection. Instantly, my computer shut off and wouldn't boot back up. Basically, it connected the circuit and fried the motherboard. (Thankfully, the hard drive was fine. I wa
Re: (Score:2)
He probably didn't fry the computer.
My daughter had a similar event with her computer. I realized that the problem was that the 5V connectors were shorted out, so the motherboards's 5V supply was shorted out. I separated the connectors and filled the now useless socket with epoxy to stop it happening again and the computer was used for a long lime afterward.
Re: (Score:2)
My son learned a lesson about looking where you plug things in ...
That will come in handy when it's time to have The Talk.
Mains plugged cable (Score:2)
My son actually fried my computer thanks to a broken USB slot.
Let me get this straight, so your son was holding a power cord that at one end was plugged into the mains, and then tries to connect the other while the cord was powered ?
There are so many ways this could have gone wrong.
(Well unsless it's a car charging cable that needs proper handshake between the charger and vehicle before letting the power flow).
My son learned a lesson about looking where you plug things in
He should also have learned to avoid manipulation mains-powered cables. The proper sequence would be plug the device-end first and then and only then plug the o
Re: (Score:2)
Re: (Score:2)
relatively easily broken
I've never broken one in my life. Define "relative"?
Why are usb cable plugs all female (Score:5, Interesting)
A huge part of why the whole situation os so screwed up is that all of the USB cable plugs are female.
So have these vary small connectors (except for USB-A), and having them all be hollow means they are extremely fragile and if you force it even a little bit the cable end is loose for all eternity. I have that situation with a PS4 cable I bought, somehow I pulled on it or pushed on it wrong and now it falls off the controller at a glance.
Compare that to all of the connectors people love - various audio jacks are mostly male on the cables, they plug in with a satisfying feel and are durable as heck. Apple's Lightning connector, which works extremely well and is also extremely durable, is just as nice and robust to use.
USB-C is better for sure because at least you can use it in either orientation so there is not as much of a forcing issue - but again the cable ands are somewhat delicate and you have a thin slot inside the device you are plugging into which I can see getting damaged in ham-fisted attempts at pushing in a USB-C connector.
I am glad everything is going to USB-C at it really is better to have everything on one standard, but I am a little sad Apple will have to move away soon from Lightning for the phones, even though it means fewer cables when I travel.
Maybe the USB group could invent an "inverted USB-C" standard that would essentially be "Lightning for Everyone".
Re:Why are usb cable plugs all female (Score:5, Interesting)
Making the cables use female plugs is the best choice. If the cable is male and it breaks then chances are part of it will end up stuck inside the computer, i.e. it damages the expensive bit.
Cables are cheap, and you want them to break first. USB is actually decently robust, having a metal shield that helps guide the cable in straight before it comes into contact with the plastic part and electrical contacts.
Apple's Lightning connector is inside out. If you shear it off the tip can get stuck inside your phone and is very difficult to remove. Also, because the contacts are against the outer edges, it's got less electrical shielding. With USB you have the metal shield of the connector and the metal shield of the plug, connected to the internal shield of the cable, and as such you get extremely high data rates and native support for stuff like HDMI. Lightning can't do that, Apple had to make an HDMI dongle with VNC to HDMI processor in it.
Re: (Score:3)
Yeah, but that's only an issue if you use a Lightning connector in an situation which might see a lot of stress and abuse, like for headphones. Surely nobody would do that?
Except that doesn't happen (Score:2)
If you shear it off the tip can get stuck inside your phone and is very difficult to remove.
In practice that doesn't happen, and if it did you could use needle-nosed pliers to pry it out. I have never had that happen to any Lightning cables I have owned, nor have I had any friends that ever happened to. Basically the way cables are the material beyond the plug would sheer off long before the tounge inside the port did, it's very easy to make that part strong enough you cannot easily break it off.
How many
Re: (Score:2)
If a male end gets stuck, it's easy to remove it from the expensive bit. Just use a pin bent into a small hook to pull it out.
OTOH, with the current design, the male end on the computer can actually break off, damaging the port forever. Making the cables female was a stupid design choice.
Re: (Score:2)
To break in inner part of the USB-C port you would have to apply enough pressure to deform the metal housing, in which case you have probably ripped it off the PCB anyway.
With Lightning you only need to apply enough force to break the thin plastic that the end of the cable is made from.
Re: (Score:2)
Re: (Score:3)
Yeah, unfortunately USB-A is large enough that you can shove things in there to damage it, including micro USB and USB-C plugs.
Re: (Score:2)
A huge part of why the whole situation os so screwed up is that all of the USB cable plugs are female.
We could switch to Boy George [wikipedia.org] genderless connectors. If that's what gets you off.
Re: (Score:2)
Except the data says not (Score:2)
The idea behind that was that when people inevitably did something stupid, they were more likely to end up with a broken replaceable cable instead of a un-fixable soldiered port
I agree that was the theory, but the reality is this - with USB both the computer AND cable connectors are more delicate, and either side may be likely to develop issues.
We know from the data on hundreds of millions of iPhones now, that a male connector on the cable is simply more durable all around. I've never had an issue with a L
Re: (Score:2)
Still fair (Score:2)
I agree about the original USB design, there was for sure cost consideration there so fair enough, that gets a bit of a pass (though the agony it has caused over a decade or so is no less real, and I think there's an argument to be made it was a bad tradeoff even then).
However where I get off the acceptance train for sure is USB-C. Here they had a chance to make the connector RIGHT, to make the cable end male and the whole system more reliable. But they chose to instead just take on the reversibility aspec
Re: (Score:2)
Re: (Score:2)
I think a bigger problem is the small size of the socket relative to the long length of the plug body. A little lateral force on the plug will have a lot of leverage to break the socket off the circuit board. That was a problem with micro USB, too.
Not actually a big deal (Score:4, Informative)
Too much is made of reversibility. USB devices, initially, were not plugged in / unplugged as often as today. The connector was also tons better than many of the alternatives at the time.
USB spec mandates the USB logo on the top of the plug. Not all cable manufacturers follow this and when they do it isn't always very visible. But it makes this problem quite simple. It is crazy how many people don't know about the logo on the cable, but it would be quite easy to solve by having it more visible and having a note on the package.
What is nuts is plugging in random devices to your computer port or plugging in your phone into a random USB port. People may have gotten more used to USB charging than USB data transfer and don't realize you are plugging in unknown harder into a computer bus that can do anything.
Re: (Score:3)
Top becomes left or right if the slots are 'portrait' oriented and there's more left-right confusion if you're reaching behind the device!
The 18 USB sockets across my desks PC,monitor & external drive are all rotated/portrait as are the several NAS boxes USB expansions. Only my XBox1 actually has a top to match with - but only after I sat it flat after an office tidy.
Also add this (Score:2)
It was so difficult and unforseeable . . . (Score:5, Insightful)
. .. that people didn't come up with the idea of keying the shape until 1952. [wikipedia.org]
Seriously, the lack of reversibility is excusable, but the failure to visibly key the external shape was not.
Re: (Score:3)
Seriously, the lack of reversibility is excusable, but the failure to visibly key the external shape was not.
My personal favorite was the original IDE cables... notch? No notch. [computerhope.com] Pin 1 indicator? All gray. [wpclipart.com] Though I suppose you weren't supposed to plug and unplug them very often, the design was almost an insult saying you must be a computer technician and track the twists and turns of the cable to do it right.
Re: (Score:2)
I also look forward to lecturing upcoming whippersnappers on setting master/slave/cable select jumpers.
Re: (Score:3)
Back in the bad old days of IDE there were a lot of unkeyed cables (without even a stripe, let alone a physical key) and unshielded headers without even a key mark silkscreened on the board. What's more, it wasn't all that unusual to not even have the connectors numbered! You'd have to find the square solder mask on the back side of the board! Most of my hardware was cheap crap, especially at the time (what with ATA drives being new and all) so that surely led me to encounter more of that kind of thing than
Designer (Score:2)
Re: (Score:2)
Bie (Score:2)
Also design the next one so it aligns and glides in in the dark, kthxbie.
Re: (Score:2)
Yeah, this. The thing that sucks worst about USB is that it's square, there's no taper. Even D-Sub has a taper. If it had a taper, then the problem of thinking you have it flipped over the wrong way when you in fact have it right but the plug is just hanging up on the edge of the socket would be at least reduced, if not eliminated. They managed to get this wrong even on Type C.
Simple fix (Score:2)
Re: (Score:2)
The USB standard specifies that there should be some marking on the top side of the cable, usually the USB logo. Unfortunately with cheap cables it's either omitted or it's just moulded into the plastic and almost invisible.
At the time it wasn't bad compared to other connectors of the day. PS/2 was round and difficult to correctly align. D-types were nearly as bad a USB, slightly easier to see but not by much. The software side which a much bigger screw-up.
Re: (Score:2)
Not really that big of a problem (Score:2)
Yeah, there are non-compliant devices that for some reason have the logo on the wrong side for no apparent reason.
But can't blame usb consortium for that.
Re: (Score:2)
I thought I had read it somewhere, but can't find it anymore.
Most vendors do follow this convention though.
Re: (Score:2)
compliant devices have the USB logo on the top of the plug, so just by looking at that you know how to plug it in.
Except if you're: plugging something in the back of your computer it's hard to see; if the USB port is sideways, it's not so
clear which is the "top"; manufacturers often put their logo on the other side, which makes it harder to tell which is the top from a
very quick glance (i.e. you can't just look to see which side has markings on it)...
Hindsight (Score:2)
"640 plug-in retries ought to be enough for anybody"
Token Ring Connector (Score:2)
Should we expect any less of the company that designed the beautiful token ring connector?
https://en.wikipedia.org/wiki/... [wikipedia.org]
Non-fitting end (Score:2)
When he dies, put him in his coffin upside down.
Note quite first (Score:2)
Although the rectangle design we all know was ultimately chosen and adopted by pretty much every hardware manufacturer since Apple first put USB ports into its computers in 1998, Bhatt acknowledges that there may have been a better way.
My company had a Compaq Presario 3020 in 1996 that had USB, but it wasn't a problem because nobody made USB peripherals until the iMac popularized the connector two years later. :D
Meh... fancy BS (Score:3)
Even as a UI enthusiast, I've never been a fan of reversible designs as it indeed makes them more expensive and adds extra wires that makes the plugs larger. Hermaphroditic cables are just wasteful and over-engineered. This is especially true with tiny, fine-pitch designs, as it just increases the likeliness that things will wear, cross-talk, and even short out. Even mobile cables need to be robust, not fancy.
Face it, we've never had trouble plugging in DB9, RJ45, or HDMI. The real problem is that USB and SATA are perfect rectangles. Even worse, it IS actually possible to plug in a USB cable upside-down with enough force. I've seen people do it without even breaking the connectors, so it's easy to think it's plugged in correctly when it isn't. Try doing that with HDMI.
Yep, BS for hyping up USB-C (Score:2)
Agreed, non-reversible isn't the problem. They're just trying to do the slight-hand trick to make the public think so ... and that it's all solved now, so if they just pay for that upgrade ... for every device.
But since USB-C is so fragile, then there'll be another round of "upgrades" to USB-D in a few years.
Re: (Score:2)
"adds extra wires that makes the plugs larger."
This is not inherent to reversible designs. Even USB-C only requires two wires left unconnected due to reversing (the second USB 2.0 pair).
"we've never had trouble plugging in DB9, RJ45, or HDMI."
This claim indicates a lack experience plugging things into the back side of equipment. (BTW, the 8P8C socket for that RJ45 will accept the 6P4C or 6P2C plug for an RJ11.)
Re: (Score:2)
IANAUE (Score:2)
I am not a USB engineer, but ...seems pretty stupid?
""The biggest annoyance is reversibility," Bhatt told NPR. For outsiders, it seems like designing the USB so it can be reversible would be an easy fix to everyone's problems....Another option that the Intel team floated was a round design"
Reversibility IS slightly hard, so that means the solutions are slightly more expensive.
Round is even HARDER to plug in correctly, as you have 360+ "initially appearing correct" orientations, instead of the 2 USB has toda
Re: (Score:2)
"like you have in the RJ11/RJ45"
You can accidentally put an RJ11 plug into an RJ45 socket.
double standard (Score:2)
Got the other end right. (Score:2)
Reversable USB double the cost of the technology? (Score:2)
Not really, some steering diodes on the power lines and a bi-directional databus to transfer data.
Quote is BS (Score:2)
Bhatt told NPR that would have doubled the cost of the technology, requiring double the wires and circuits.
This is BS. Adding a second set of contacts would not require redundant circuits. The added contacts could simply be connected in parallel with the first set of contacts.
In a similar vein, I'm often annoyed by magnetic stripe readers such as the ones on gas pumps, with their sometimes hard-to-interpret drawings of card orientation. For a small additional cost these could be made to work in either orientation by adding another magnetic head on the opposite side.
First World Problem (Score:2)
Seriously? Is this something that needs to be "fixed" at all? It's a First World Problem. Move on.
Re: (Score:2)
Why do they Call it Serial? (Score:2)
While we are on this topic, I have always wondered why it is called Universal SERIAL Bus?
Not only are their more than 1 pin/wire, but there is more than 1 data line.
Am I completely crazy, or does that not make it parallel? What am I missing? But then the old serial cables back in the day still had like 8 pins, so I have no idea what is going on.
Re: (Score:3)
If the data is sent one bit at a time, it is a SERIAL connection. If mutiple bits are sent at one time it is a PARALLEL connection. It has nothing to do with the number of wires physically required to implement the connection. Both USB and the old serial cables are indeed serial.
Add a bump! (Score:2)
Easy? You have to push like a gorilla! (Score:2, Insightful)
How the hell did you manage that? ... And in NO case ever attempt to turn on the brain!".
Like "It doesn't go in, let's push even *harder*!
I bet you also plugged in VGA connectors with the pins crushed and bent against the outside metal.
Re:It's bad but not as bad as micro usb (Score:5, Interesting)
It's easy to muscle past the slot shape and plug it in upside down and destroy the cable and socket at the same time.
It's easy to do, but not easy to do accidentally. You have to slip and fall on it while inserting it for that. Do crush every egg you try to crack, too? Every time you open a jar, do you shatter it?
Re: (Score:2)
Maybe they are one of those people who just rams it in harder when it doesn't seem to fit. They were probably deprived of those "match the peg to the hole" toys when they were young.
Re: (Score:2)
There have been cables with reversible micro-USB plugs on the market for a few years now.
I'm not sure which one was first. There were several crowdfunding campaigns for them starting in mid 2015 (MicFlip, MicroFlip, Lightors, ...), but it did not take long before multiple cable brands were available in retail stores.
With the exception of my one micro-USB 3.1 device, these type of plugs are the only micro-USB plugs I use.
Re: (Score:2)
Hulk smash!
Re: (Score:2)
Didn't your father ever teach you "Never force anything. If you have to force something, then something's wrong."???
At least, your first girlfriend would teach you that.
Re: (Score:3)
A 4 wire Audio type jack would have solved all issues.
Audio jacks have horrible contacts. That's why they frequently crackle when you wiggle them. High speed digital signals and power delivery would not be a good application for that type of connector.
Re: (Score:2)
It's because the contact surfaces aren't flexing, if you have spring load on all contact surfaces that's a non-problem. A bit more expensive, but not impossible.
Re:I call BS on this (Score:5, Funny)
rectangular square
Still trying to figure this one out.
That's the trouble with topologists. They don't know the difference between a cup of coffee and a donut.
Re: (Score:2)
It's actually a squangular rectare, but let's not mince words.
Re: (Score:3)
Re: (Score:2)
Re: (Score:2)
Except cost. USB-A can be printed onto a standard thickness circuit board, basically for free - that was the point.
Re: (Score:2)
I don't think USB has an controlled impedance lines, and the connectors are definitely not controlled impedances. The cables are just two twisted pairs inside a foil shield (sometimes).
Re: (Score:2)
Back in the Apple G4 days I discovered Apple made USB keyboard and extension cable that had a notch in the metal shell. It did not affect plugging anything into the computer, but only the keyboard could be used with the extension cable.
Even when using it as designed, the notch [reddit.com] made it more finicky over how you plugged it in.
Re: (Score:2)
Ever see someone shave down a 9-pin serial port or 25-pin printer/midi port connector then tell you the peripheral didn't work? Yeah...all kinds of special people out in the world. And if you think that's bad, just remember most of the drive cars too.
Re: (Score:2)
I don't see this obligatory gif here, so here you go [tumblr.com]. I don't entirely disagree that it was a major improvement, but some UAT by the masses would have demonstrated some of the major failings and we might have started closer to USB-C.
Re: (Score:2)
Yeah, you turned the mini-DIN plug in the socket until you could feel that you had got it in the right orientation. Then you pressed it in and it made a connection.
The big problem that I think people overlook when discussing USB is that the older signalling standards did not hot-plug.
If you plugged an ADB or PS/2 (or any of the other [deskthority.net] keyboard/mouse connectors), you would risk frying electronics in the host and/or the peripheral.
Re: (Score:2)
Headphone connectors work exceptionally well!
They really, really don't (unless you're talking about the big-ass 1/4" plugs, on equipment that never gets moved with a headset plugged in anyway). They collect pocket fluff, they get unreliable after a few years of plugging and unplugging, and they provide way too much easy leverage to put bending force on circuit boards.
It's amazing how Apple taking the damn things out of their phones can make people forget decades of walking around with a finger on their walkman/discman headphone plug or setting heavy
Re: (Score:2)