What I don't understand is why you are not allowed to air mail a battery by itself in a sealed container, while you are allowed to air mail the same battery inside a device. I am not that familiar with battery technology, but I would expect that a battery connected to a circuit to have additional ways of catching fire compared to a battery by itself. I mean if a fault happens inside the battery you are screwed whether it is in a device or by itself, but AFAIK there are cases where the problems were caused by the electronics connected to the battery, so you get an even higher chance of something going wrong. Maybe they are afraid the density, i.e. shipments with just batteries which would make more batteries per volume than say a shipment of laptops? But still, there would be rules about density then.
What am I missing?

Problem is battery density. If you're sending batteries, the amount is far more dangerous per unit volume than if the battery was in a device.

Shipping lithium batteries in bulk is what caused the downing of a UPS cargo plane a while back which is why they're no longer allowed - one battery caught fire, which then caused other batteries in the same container to catch as well.

Whereas if it was in a laptop, it may destroy the device and the pack, but the density of cells is lower and its less likely to catch more packs on fire.

It's why hoverboards are particularly dangerous - their packs of 10 or 20 cells wrapped tightly together - when one goes off, it will more than likely cause the rest to off as well.

Also, raw lithium and aluminum don't mix - which means the sprayed lithium can damage structural aircraft components as well.

Google's compatibility list shows it's only for recent (under 2 year old) Chromebooks. It's not that the older Chromebooks are too slow or anything, they just are older than 2 years old. After all, the list excludes many x86 based Chromebooks (including Google's original Pixel).

You can easily side load a lot of stuff yourself using the free personal developer accounts. The apps expire after 30 days though so you have to keep re-adding it every month. I've got a couple apps on my phone that apple would never approve on the store, no jailbreaking.

Even better, Apple generally wants you to do this with apps with source code - the developers of f.lux tried it, but they released it as binary only and Apple called them out over it.

It's one of those things you really wish you could ask RMS about - a commercial closed-source OS that allows open-source to be loaded on, with enforcement of the "source" part - no releasing of binaries that may or may not match the source, but an OS that requires you to build the app from source code.

Lock him away and take all this data and hardware and when he submits the bugs to Apple, make Apple pay him the bounty and let him go with a nice clap on the back.

Well, or trust him not to sell the exploit to someone else or have it stolen. This must be worth a lot of money, much more when it is not submitted. People have been stolen from, killed or tortured for less.

Exploits are the new plutonium. You can prepare for war with stockpiling and weaponizing them.

Well, given there are three parties who would pay for it. First is Apple, as part of their bug bounty. He'd probably get a cool quarter million out of it.

The second party is pirate app stores for iOS - they often sell access to their pirated apps and do have some money to spend. The Pengu jailbreaks were basically this.

The third party is state-sponsored agencies. If you were in it for the money, you WOULD do this because they really pay - a cool million dollars or more for something like this.

It's traditionally why Apple doesn't pay for bug bounties - Microsoft, Google, their vulnerabilities sell for around the same price as the bug bounty - typically a 10-20K. But an iOS bug is big-time, easily $1M+.

Have you seen the "<iframe allowfullscreen>" tag? Seems to be a marketeers wet dream. And every web users nightmare.

They probably implement that like full screen video - it has to be a user-requested action for that to actually happen and browsers have the ability to block a site from going full screen (as well as doing the overlay that says "site is now full screen" with "allow" and "cancel". And which is NOT overridable - the browser throws it up and the user can cancel it.

What's our take away on this supposed to be?

(A) These evil scoundrels are cheating on the government tests

(B) The people who are designing the government tests epically suck at their jobs, should be fired, and have competent people hired in their places

I'm going to have to vote "B" here, folks.

Or maybe the test is designed so comparisons can be made between years, models and history?.

I mean yes, you can design the test to be different and updated every single year, but then you lose the ability to compare models from this year versus models from last year - and while you can run it under both tests, no one wants to pay for that. Perhaps the TVs take 40W this year. Last year's test had those TVs taking 30W. Was it the test that changed that caused all the new TVs to take 10W more? Or are this year's models less efficient?

That's the reason why tests are standardized - they're designed to emulate real-world use, but be scientifically repeatable so there's a consistent basis to which comparisons can be made.

It, just like the VW scandal, can result in cheating because of this - the test is standardized so if you detect the test you can game it. No way to avoid it

This is absolutely the case. Autoplay is a static attribute to the HTML5 video tag and can be set to be ignored by browsers, but there is also a Javascript API for HTML5 video playback and it is trivial to start playback from there. Technically that is not "autoplay", it is just, "play" that happens to be triggered on page load via Javascript.

And what happens is that the play command gets lumped in with the popup blocker commands that get ignored on page loads, which lets the play command work AFTER the page loads, but it doesn't start if it's part of the page load.

An alternative is to ignore the play command if the tab isn't visible. And just in case, we can have a right-click menu option to send the play command as well in case someone tries to be tricky and assume the video start playing immediately.

Around 2006/7 I had a used Tmobile Dash with a voice/text plan and no data. At the time, my university ran a 28kbps dailup internet service that students could use for free. It was a relic of another time, but it was still there.

On the Dash/Excalibur (and presumably other Windows Mobile devices) you could dial into these services with the built-in modem, and since I rode the bus a lot (at least two/three hours a day), I used that service.

It was hilariously slow, but it worked. I could visit websites, read articles, and chat. By turning off the images, it could be done with reasonable comfort.

Don't underestimate the value of being able to log on at any speed.

Actually, you could do it on practically any 2G device that supported data with a facilitating carrier. It's called Circuit Switched Data (your usual GPRS/EDGE/3G/4G/LTE is packet switched data) and it establishes a traditional modem connection, using a voice channel.

It's slow because the voice channel is 9600kbps or there abouts. When you establish a CSD connection, the cell modem tells the network to use one of its modems to dial the number and establish a modem connection. The data is then sent over the voice channel as raw data (the modulation itself wouldn't survive the GSM coding, so it's just data). You phone then opens the virtual serial port and accesses the data, which is probably just a PPP session. In those days, the connection was identical whether it was CSD or GPRS/EDGE - it's just the endpoint is different - in CSD, the PPP data goes over the air to the number you dialed, while in packet, the PPP session goes to the modem to interface with the packet data hardware.

Incidentally, GPRS/EDGE use unused timeslots and frequencies and combines them to get data connectivity so the less busy the cell, the faster the data connection.

The problem with tinfoil is it grows all spiky really quickly. You brush it off and all those spikes go everywhere, killing everything electronic you touch, and making it hard to plug stuff in and out without getting shocked.

Al foils, though, doesn't have that problem. Hell, even tin cans are steel these days.

All of our phones and digital cameras have a maximum SD card limit, most 64Gb.

There are two known limits to SD cards.

First is the standard old SD card - FAT16 formatted, up to 2GB. Then there's SDHC, FAT32 formatted, up to 32GB.

For larger cards, there's SDXC, which uses exFAT and has a 2TB limit. 128GB cards are common today, and if you can take 64GB, you can take this card as your device is SDXC compatible.

Some SD cards were 4GB using the SD method, which was a very creative way of interpreting the standard - as such, they may work in some devices but not all. If you're curious, the SD storage medium used a signed 32 bit integer for byte-level storage access. SDHC turned that integer into block-level access and that's why it was pretty trivial to add SDHC support.

The issue here isn't the materials cost of the epipen. You don't even need an epipen to deliver the medicine, just a syringe and an epinephrine vial. Any school nurse worth her salt will know how to use a needle. If school districts wanted to give a fat middle finger to the pharma industry on this they could go and purchase them.

The issue here is that Mylan (the makers of the pen) lobbied the FDA and government to require its purchase be done by school districts and then jacking the price up to gouge the taxpayer (ie you and me). Now school districts have to purchase the pen instead of going the route I outlined above.

The only way to hack the regulatory process is to donate a retarded sum of money to "charity" of a specific presidential candidate and various other lobbying groups. Though that isn't really hacking. Just, "business as usual" in the fairy tale land known around that stretch of highway known as the beltway.

The cost of it was around $50, then it got jacked to $150 and now $600+.

Epinepherine is cheap - an EpiPen contains about $1 worth of the drug.

Mylan basically created an auto-injector that made it possible for anyone to administer it - remove a cap and jab, which is why it was popular - because anyone nearby can go grab it. Just like an AED, it can be put in a lot of places and be ready when needed. So there was a lot of demand to have it everywhere, especially in schools where an anaphalactic shock can mean you have only a minute. And because it's easy to use, anyone can administer it.

The problem is after getting everyone to demand they be made readily available, Mylan jacks the price up. They've been on the market for years the cheap price. In fact, that's why they're sold as a two-pack - the FDA found one injection wasn't necessarily enough of a dose, so instead of redesigning it to accommodate twice as much drug (at the price of going through the whole approval cycle again), they just sell them as a two-pack. And this was a decade ago. (The patent's actually expired) The only reason they did this was simple - the competition generic brands were actually having difficulty with their products and thus are down for the count for a few years.

Vacuum tubes create a noisy signal, but in a weird coincidence, they do it in a way that is pleasing to the ear. The clipping and distortion sounds "warm," and there's an added depth in the sound (harmonics) that you don't get via transistors -- unless you create circuits that mimic the behavior of a vacuum tube.

Metallica? Tube amps. John Mayer? Tube Amps. Clapton? Tube amps. BB King? Tube amps. Eric Johnson? Steve Vai? Garth Brooks? All tube amps.

Of course, much of the "tone" we guitarists revere comes from overdriven and abused guitar tubes -- cranking up the volume on the tube so that there's massive distortion and noise -- which again sounds pleasing to some people.

Now, it's one thing to overdrive a tube or change the bias on it to get a particular sound from your guitar, what about building a tube amp to just listen to music?

Well, I suspect this is essentially "remixing" songs. Adding a bit more depth, dirt, or warmth (from the noisy tube) might sound better but that's subjective, and it's all about personal preference.

So, sir, you might argue that you dislike what a noisy tube does to your signal, but you can't say some people won't perceive it as improved, as it's about personal taste.

In other words, you use tubes to process your signal. Any amplification on their part is a just a coincidence.

I get it - its an effects processor. You just keep it in the analog realm, given you can do the same thing digitally afterwards (just without over driving the input).

There are general audio blocks - you have ADCs, DACs and amps, which just have ONE job. Anyone claiming any of those parts are producing a "warmer" or "bright" or "sounds better" is lying because that part is malfunctioning and thus not transparent. (The jobs of DACs and amps is so well defined that we've pretty much achieved perfection).

Then you have a processor, whose job IS to alter the sound, and processors all sound "different" because that's their intent - to alter sound.

And everything that alters sound (EQ, etc) should be kept in processors. A lot of the "tube sound" is caused by the distortion (i.e., imperfect amplification) introduced by tubes, as well as the impedance issues caused (tubes are a high-impedance output, hence the output transformer is required to lower it) which causes EQ because even with a transformer, its still fairly high output impedance compared with the speaker or headphones. (Speakers/headphones have a frequency-dependent impedance, which causes all sorts of EQ to take place - when the impedance is high, those frequencies are muted, etc).

Yes, we solved that, we went to transistors which have a stunningly low output impedance that they can give a flat response to dramatic impedance changes (if you use the 1/8th rule, where output impedance of the amp is 1/8th the nominal impedance of the speaker or headphones, then the variance is around 1dB or so as it goes from nominal down to 1 ohm or lower).

NAND have limited write cycles... doing a re-write NAND bruteforce attack is not going to last very long!

A 4 digit PIN code has around... 10,000 combinations. You get 10 tries before it wipes, but 6 tries before the delay gets long (1 hour). So if you guess 6 tries per flash, you only need to do 1,667 reflashes. Given the nature of the effacable storage flash, this is well under their use limit.

Apple had a reason for not making an iPod Classic with a flash drive. If they had sold it at a comparable price to the version with the miniature hard drive (and they could have given the current price of flash memory; 128GB is around $30 retail now which means it costs Apple less than that), it would have exposed the lie that is their current pricing of the versions of the iPhone and iPad with more than the minimum amount of memory. Those big and important parts of their customer base would have rebelled and demanded lower prices, which would have cost Apple a LOT of profit.

Well, when the iPod classic was discontinued, the iPod Touch was popular, which DID have 128GB of flash storage.

And the Classic really was the only one in the lineup with spinning rust, which was going away because flash was more efficient, cheaper and smaller - the 160GB hard drive in the last Classic could be replaced with flash memory with a volume that was much smaller. At the same time, SoCs had native flash interfaces and didn't require CE-ATA or ATA interfaces (which was getting rarer).

Apple could've gone with a SSD manufacturer, but PATA SSDs were and still are pricey. Which would've pushed the Classic price up into Touch territory, further diminishing sales - for most people, the Touch would be far more useful than the need for a USB hard drive at the same price.

I'm sure Apple considered expanding the Nano line of iPods to have 128GB of flash (which is what a Classic would be in the end) but decided the added cost wouldn't make economic sense, especially at a time when iPod sales were tumbling into nothingness. People just weren't buying iPods, especially large ones anymore with their smartphones. The small 4/8/16GB iPods still had some sales purely because of their low cost and practical disposability which meant they were popular at gyms and running and all that.

It's a wonder Apple hasn't killed off the iPod altogether - even the iPod Touch isn't selling.

Why VIM doing rendering at all? and better yet, platform-specific rendering?

In text mode, no, there's no rendering going on. But in graphical mode, you have to translate the characters to pixels, and that's not a trivial task anymore. Add in scrolling and other tasks you do with blocks of text, and you're pushing serious pixels. Especially if you operate in a high resolution mode (4K monitors aren't hard to find these days).

DirectWrite is a GPU-optimized font renderer for Windows - it loads the fonts into the GPU memory, and then the font rendering programs, plus text effects like ClearType. You then send it the characters and the GPU renders it down, like it did in text mode, but without the blocky pixellated look. Given how many pixels it has to push these days, it's a task a GPU is much better suited for, and it makes scrolling and displaying large windows text all that much faster and with lowered CPU load.

GDI allows a lot of nifty tricks, but 99% of the time, no one uses all those tricks - they just want to spit text onto the screen. So Microsoft created a GPU optimized way to accelerate the common task.