Become a fan of Slashdot on Facebook

 



Forgot your password?
typodupeerror

Comment Re: Port Design (Score 1) 134

Now that explanation makes sense. We're still relying on external equipment to function per the spec, though, and actually have overcurrent protection; we're also hoping that said protection activates before our equipment is destroyed. And it's still the much less likely scenario, compared to overvoltage, which is more difficult to protect against; if that's not being handled, it still doesn't make sense to handle reverse voltage.

Comment Re: Not Harassment (Score 0) 428

I'll add to this by suggesting, for a moment, that we pretend that Cortana is actually a real woman and not simply a computer program. If the only way she can be taken seriously is through social codification, saying that she should be taken seriously because she is a woman and it is sexist not to take her seriously, what kind of a win is that for women? Hoe is it not sexist to say a woman must be taken seriously, rather than allowing the same measure of seriousness that is applied to me to be applied to women? Provided that some measure other than penis length is used, there shouldn't be anything wrong with that; and, then, we are free to ask Cortana, the computer program, whatever we want.

Comment Re: All I know is that this: (Score 1) 260

I'm not sure if you are trying to argue, or agree and expand. I'll give the benefit of the doubt and assume you're simply expanding on my points. I'll also clarify, for the benefit of those reading along.

It is absolutely false that GitHug benefits in any way from encouraging people to use them as a central repository; in fact, as evident by the discussion happening here, people using their service (incorrectly) as a central repository has a direct negative impact on GitHub every time the service goes down and those users start bitching that they can't get any work done due to the outage. That said, I do agree that they don't emphasize the correct way to use their service, as an additional home for your repo, which brings a number of additional features. This is something they can, and should, certainly work on.

I also never said, or even implied, that GitHub should get a pass on the stability of their service, let alone on account of features Git natively brings to the table. I merely pointed out that they built a fair bit upon those features, adding many of their own; if you only view GitHub as a place to host Git repos, you're missing the bigger picture by a mile. Additionally, it is worth noting that GitHub Releases, which some projects use to host builds, are being used incorrectly by those projects. All a Release is supposed to be is a tag, pointing to a specific revision of the source, that GitHub lists on the Releases page.

The problem is that people are using GitHub incorrectly, the complaining when the service goes down, not because they lose access to the additional features GitHub offers, but because they lose access to things that should live elsewhere in the first place. If I was hearing complaints about the temporary loss of GitHub's additional features, I'd consider those complaints valid. But complaining about bad things happening when you misuse a service? Come on.

Give them valid criticisms where they are deserved, there's plenty to talk about there; likewise, when it is pointed out that you are misusing your tools, accept that criticism yourself and become a better developer for it. There is no need, not benefit to anyone (yourself included), to blame GitHub for your misuse and misunderstanding of their service.

Comment Re: All I know is that this: (Score 1) 260

No, as long as it will com back up, with all data in-tact, it has plenty of value. Or are you saying there is no value in a remote warehouse full of your backup tapes because you can't access them immediately and sometimes the facility closes down and you can't access them at all? Not that anyone should be using GitHub as a backup solution, but it's the same principle: a datastore.

If your data, and ability to access it at all times, is important, you plan for that. In this case, that means hosting your own remote alongside GitHub and configuring Git to push to both; that way, when Git is up, you get all of the added value it brings (and there is much) and, when it is down, you can still clone your repo from a known location, without having to collaborate with another developer, who may be unavailable, to clone from his repo.

You can do this for $5/mo if your repo is <15GB or so. Or, if you want something just ever so slightly more reliable, you can do the same for $10/mo and let me get a little commission on the deal (and get an extra 4GB of storage). Hell, if you're willing to trust me with temporary access, I'll even set it up for you (one time, maintenance is on you) on Linode if you've used my referral link.

Comment Re: Port Design (Score 1) 134

I lack an EE degree as I'm a software developer and not a hardware guy (some day, perhaps...) but I do generally get the basics. It still comes down to planning for the least likely scenario when you can't do anything about the more likely ones, but that's product design, not port design; even if that level of protection was in the spec (and I haven't looked at 3.1 in detail) there's nothing preventing manufacturers from leaving it out and just not using the logo. Beyond that, while it would have been preferable (for the end user) to have reverse polarity protection, you know as well as I do that accounting and marketing run engineering anymore; if it's not marketable, accounting insists it gets cut. Input protection is only marketable in deveopment boards and high-end testing gear, so it gets left out of consumer kit. Welcome to modern life...

I'm not saying I agree with it, just that this is the state of consumer product engineering and your choices (especially as an EE) are to either put yourself in a position to change it, or get used to it. We lowly consumers are expected to have out hands around out ankles at all times, didn't you know?

Comment Re: Port Design (Score 1) 134

Wow, where'd you get your EE degree? That's impressive!

Your complete lack of understanding of how electricity works, that is. And USB, for that matter.

For starters, the way USB handles overcurrent conditions is by limiting current to slightly more than requested (or slightly more than 500mA, 900mA for USB 3.1, if no negotiation has occurred) and monitoring current draw. In that way, even a dead short can not damage the device and the moment an overcurrent condition is detected, the port is shut down. This is all handled in the chipset and is a requirement of USB certification; the chipsets used in the Chromebook Pixel and the test gear mentioned in the articles I've read (I didn't read this article so I don't know if this particular article mentions which equipment was used, but I've read several others on this incident) are USB certified.

Second, this was not an overcurrent condition, but a reverse-voltage condition, so the above would be irrelevant if it weren't in response to your own irrelevant mention of current and overcurrent conditions. This had literally nothing to do with current and literally everything to do with polarity.

Third, that's not how diodes work. Diodes restrict the flow of electrons (not entirely block, mind you) in one direction, the don't short circuit anything and they certainly don't affect current. One could argue that putting too small of a diode in a circuit would add resistance, which would affect current, but then you're runnign a component out of spec and it is acting as a resistor, not a diode, because that's now what diodes do; and even in that instance, it would be limiting current, not drawing it and creating an overcurrent situation. That means, while a diode may have saved the device, it would not have involved a polyfuse in any way, shape, or form; it would have done so by reducing the reverse voltage, not by "short circuit[ing] the reverse polarity situation causing an overcurrent".

But, even then, diodes aren't magic. Remember, they don't magically stop current flow in one direction, they simply massively restrict it. To a point. Once that point is reached, the floodgates open and *bam*, current flows through. That that means is, if the diode is rated with a breakdown voltage of 5v (and why wouldn't it be in a 5v circuit?) and you reverse 5.1v through it, well... there you have it. The net result is either -5.1v through the circuit, if the circuit was not already energized, or -0.1v if it was. Negative voltage, either way. And let's not forget that it is common (and even specified) for slightly higher voltages to be used when charging which, I'll remind you, is also specified for USB 3.

That's not to say there aren't solutions, or that one or more of those solutions doesn't involve diodes; there certainly are solutions and they should have been employed here. A diode with a breakdown voltage higher than what one might expect to see sent down the wire by mistake (say 12v, the highest rail voltage in a PC, or 19v, the common laptop battery pack voltage; a 24v breakdown voltage should suffice) should adequately protect the device from and reverse voltage condition it might encounter. And before you suggest that it wouldn't protect against mains voltage, I'll remind you that mains is AC; there's no point protecting against that voltage in one direction when it's going to be allowed through in the other. You're toast by then anyway; and your polyfuse blew long after the overvoltage killed your USB controller.

As for why there wasn't a diode in the power path with a 24v breakdown voltage? If I had to guess, it's because the common expected failure mode is excessive forward voltage, which is much more difficult to deal with. Engineering is hard; it's even harder when you don't know what you're talking about, which makes it easier for you to think you've got it right (as you seem to believe) while getting it massively wrong. There's not much point in protecting against the thing that has 0.00001% of killing your device (reverse voltage, under the assumption of using properly made cables) if you're leaving the thing that has a 99% chance of killing it wide open (overvoltage, under the same assumption); that's engineering.

Comment Re:Port Design (Score 1) 134

What if the perfectly approved Apple cable has been chafed and is now shorted to ground? Fail gracefully.

That's what you might think, but I'm gonna go out on a limb and say a cable that would fail gracefully when shorted to ground would not pass mains voltage mains to ground (e.g. the phone case). Read the articles before you blame shoddy aftermarket chargers, as the 2nd one involved a genuine Apple charger; but what you're saying is that the cable, connector, and/or device should somehow prevent this.

Comment Re: Context On the Issue (Score 1) 401

Hint: fix your broken update system first.

The Nexus line, being vanilla Android without any vendor- or carrier-specific modifications, has a very well-defined update system that works quite well. Google can't update system images that have been modified by 3rd parties (even on a Nexus; if you root or otherwise modify the system, you must flash a fresh factory image to update) lest they break things by replacing modified binaries with new, incompatible ones. This is why Google only directly updates their own devices.

No buying a nexus doesn't fucking count.

So you'd rather blame Google for the actions of Samsung, HTC, LG, AT&T, Verizon, Sprint, and T-Mobile?

Not everyone wants a shitty spy on me phone.

Which is why they don't buy from Apple, Microsoft, Samsung, or HTC, or use any phone sold by AT&T, Sprint, or Verizon (CarrierIQ). That leaves LG on T-Mobile, cheap Korean and Chinese knock-offs that likely have their own spyware and backdoors baked in, and the Nexus line. Oh, and Blackberry, but really?

Or were you trying to insinuate that Nexus phones are covert spy devices? And, if so, what makes you think Google would allow any other Android device to be different if they were in charge of updating them all?

The amount of outright ignorance in your post would be astounding if you weren't AC.

Comment Re:Don't have a problem with it (Score 1) 401

Could have bought a Nexus maybe, but that would be the only equivalent in the Android world.

All the other Android devices are alternatives; you're right, though, they're not equivalent. So, where are all the alternatives in the iPhone world?

I see this argument a lot and, really... REALLY? Of course the Android equivalent to the Apple model of developing the OS and hardware is the only line of phones Google develops. Duh much? If someone doesn't like what Apple or Google are offering on the hardware side, where are they to go? Windows? Hah! Blackberry? Hah! So they end up with one of the hundreds of other Android models available. Where is that selection for iOS?

Comment Re:So what? (Score 1) 401

Where can I read this EULA before I buy the phone? (nowhere) If I disagree with the EULA after buying the phone so I can read it, am I able to return it for a full refund, including any shipping charges, with no restocking fee? (no)

When corporations stop using EULAs as consumer-rights WMDs, you'll have an argument.

Slashdot Top Deals

Never buy what you do not want because it is cheap; it will be dear to you. -- Thomas Jefferson

Working...