I have two words for you(well, one word and a symbol): 'Google +'

Google has some very sharp people, this much is undeniable; but enough hubris will fuck up the best of us; and they've succumbed to that from time to time.

It's also probable(though not assured) that a fair chunk of games are carefully designed to avoid IOPS-heavy demands because they are supposed to run from an optical disk in a console, a situation that makes an unremarkable HDD look positively random access. The PC version will still have more trouble with other processes butting in, but anyone whose game or game engine imposes load that craters an HDD is not going to have a pleasant time in the console market.

The PCIe devices are faster; but (since they also tend to be either substantially similar to SATA devices; but packaged for the convenience of OEMs who want to go all M.2 on certain designs and clean up the mini-PCIe/SATA-using-mini-PCIe's-pinout-for-some-horrible-reason/mini-SATA/SATA mess that crops up in laptops and very small form factor systems; or tend to be markedly more expensive enterprise oriented devices that focus on IOPS) it isn't clear why you'd expect much improvement on application loading workloads.

SSDs are at their best, and the difference between good and merely adequate SSDs most noticeable, under brutal random I/O loads, the heavier the better. Those are what make mechanical disks entirely obsolete, cheap SSD controllers start to drop the ball, and more expensive ones really shine. Since application makers generally still have to assume that many of their customers are running HDDs(plus the console ports that may only be able to assume an optical disk and a tiny amount of RAM, and the mobile apps that need to work with cheap and mediocre eMMC flash), they would do well to avoid that sort of load.

HDD vs. SSD was a pretty dramatic jump because even the best HDDs absolutely crater if forced to seek(whether by fragmentation or by two or more programs both trying to access the same disk); but there aren't a whole lot of desktop workloads where 'excellent at obnoxiously seeky workloads' vs. 'damned heroic at obnoxiously seeky workloads' makes a terribly noticeable difference. Plus, a lot of desktop workloads still involve fairly small amounts of data, so a decent chunk of RAM is both helpful and economically viable. Part of the appeal of crazy-fast SSDs is that the cost rather less per GB than RAM does, while not being too much worse, which allows you to attack problems large enough that the RAM you really want is either heroically expensive or just not for sale. On the desktop, a fair few programs in common use are still 32 bit, and much less demanding.

You might be able to solve the problem(at the expense of a great deal of additional workload) by larding the caseload with samples specifically constructed to be non-matches; but then blinded and packaged the same as any other sample, to identify people who just lean positive; but that would probably require a lot of additional work to do in enough quantity to counteract the obvious pressure.

Why on parity?

In their capacity as (ostensibly) trustworthy, neutral, expert testimony, they both victimize the defendant and betray the public's trust in the criminal justice system and the duties of their office.

Punishment-on-parity seems like the absolute bare minimum, with no acknowledgement of the aggravating circumstances of abuse of authority, the corrosive effects on rule of law and public trust in the existence of rule of law, and so on. I am sympathetic to arguments that mounting their heads on spikes outside the courthouse might constitute a public nuisance, because of the smell of decay; but that would bring the requisite gravity to the situation.

Is there any reason, aside from the reflexive deference to allegedly legitimate authority figures, why they use the phrases 'gave flawed testimony' and 'overstated forensic matches in ways that favored prosecutors' rather than the more honest 'committed a fuckton of perjury'?

I don't know why anyone bothers, given that DJ spew is one of the most insufferable aspects of radio, without even the crass-but-compelling monetary justification of ads; but odds are good, on many channels, that there isn't even necessarily a DJ specific to that station. Once you can their obnoxious chatter, you can programmatically sprinkle it into the playlists of multiple stations in different markets. You only really need to be more specialized if the chatter is supposed to have some 'local' flavor, in which case you do need recordings matched to the appropriate market.

If you want to be pedantic about acceptable variations choosing something with such a long history and such wide use in various disciplines is a terrible plan.

"Percent" is probably the most common flavor currently; but 'per cent', 'per cent.', 'pct', 'pc', and likely others are still within the realm of accepted use. Hell, the '%' sign isn't even entirely settled, unicode has something like four defined variants. And that doesn't count the archaic, but historically used and still recognizable, specimens that cropped up between Latin and the present day.

I take it that you were exposed to basic literacy and only basic literacy, none of that messy intermediate stuff.

The trouble is usually that the broadcasters just hear 'extra channels!' and zone out. You can have higher quality and redundancy; but using those bits to squeeze in a bunch of extra channels and then pretend that the results are acceptable has a tendency to win out.

It also doesn't help that digital transitions are when broadcasters usually give in to the temptation to squeeze in a bunch of extra channels. When they get really greedy, the results are so bandwidth starved that they sound like horribly compressed crap(because they are) even under ideal circumstances. Even if they don't push it that hard, they haven't typically been very conservative about building in a lot of margin for degradation.

This probably has something to do with the fact that 'HD Radio' is a proprietary non-standard that is whatever iBiquity Digital Corporation say it is, and costs whatever they say it does. They obviously want it to be adopted, because they get nothing if it dies; but that's pretty much the only incentive encouraging them to cooperate on licensing or keep prices reasonable.

There is a pitiful veneer of 'standardization', courtesy of the NRSC; but 'NRSC-5c' is more or less a very lightly de-branded generic descriptions fleshed out by the incorporation-by-reference of the iBiquity documentation.

It makes the various MPEG standards and dealing with the MPEG-LA look like some kind of FOSS hippie commune by comparison.

While this sometimes pays off, when circumstances line up correctly, it is vital to keep the limitations in mind:

Lower cost has made it much more likely that random bystanders have some level of video recording, rather than none; but entities with ample resources also take advantage of reduced costs, which is why, say, nontrivial areas of the developed world are effectively saturated with automated LPR systems. There is a win for those cases where it previously would have been the word of someone who counts vs. the word of some nobody; but elsewhere reduced costs and improve capabilities make having a big budget and legal power even more useful.

Improved surveillance only changes the game at the 'evidence' stage. If legal, public, or both, standards aren't sufficiently in your favor, improved evidence is anywhere from irrelevant to actively harmful. You can have all the evidence you want; but if the DA refuses to indict, or the 'viral' pile-on targets the victim rather than the aggressor, it doesn't help you much. Had McHenry's tirade been a bit cleverer, or her target a shade more unsympathetic, odds are good that the attendant in question would be being hounded as we speak.

I'm pretty sure that letting the contractors bilk NASA is the point of the exercise at this stage. The SLS isn't referred to as the Senate Launch System for nothing.

The article does not mention where the cost of this error is going to fall. This seems like an important detail. On a sufficently complex project, one of the bevy of subcontractors fucking something up isn't a huge surprise; but I would be very, very, disappointed if NASA wasn't able to contract sufficiently vigorously to make the vendor eat the cost of delivering the goods as specified, rather than paying them for their effort no matter how well or badly they do.

It's like the beloved classic '42.zip'; but can be delivered directly over the minecraft server protocol and will be naively parsed by the server, no social engineering required... Never trust the client.