Posts by Michael Wojcik

Re: Top notch investigation

Sigh.

The "everyone knows this already" argument is the lowest sort of comment someone can make for a story on some sort of research, yet there's always someone around to make it. It's an anti-scientific, anti-intellectual, knee-jerk reaction.

The things that "everyone knows" are high on the list of things that we need actual empirical evidence and methodologically-sound analysis to back up.

Re: As pointless as going through paychecks -

that's a deliberate delve and rightly deserves serious action

Email sent and received through the corporate MTA belongs to the corporation. If someone is authorized by the corporation to read that email, then they have the right to do so. Tough luck for the sender and recipient.

Historically, particularly when email started to become more common among non-tech businesses, some organizations routinely scanned email and had employees review flagged messages.

(Rainbow Rowell's novel Attachments explores some of the social ramifications of such a policy, though Rowell rather sweetens the well by applying tremendous pressure to sympathize with her characters. It's a pretty tale, but perhaps not her most realistic. Of some formal interest as well; it's a hybrid epistolary that alternates between intimate third- and first-person narration.)

Re: It’s all about the Greed

But he used BLOCK CAPITALS! How can you disagree?

Fine. Then by your own examples, we had delivery services long before the "gig economy" created with aggregators and apps. And we will have them after the gig delivery services are undermined by Evil Government Regulation (scary!), should that happen (unlikely), if there's demand.

Re: What goes around come around.

those exceptional cases where you're travelling more than the car's full range in a day

For some of us those "exceptional cases" are the norm. If I'm driving at all, there's a good chance I'll be exceeding the car's range. (And since I can get about 700km on 5 minutes' refueling, no, 400km in 10 minutes is not "on par". It might be good enough. Maybe.)

Personally, I'd much rather have fewer batteries and an on-board ICE generator to charge them. The combustion engine could keep running at peak efficiency, there'd be no heavy transmission (and that engine would likely be air-cooled, so no fluid cooling system either), range would be better, and refueling would be quick. You want to get rid of petrol? Fine, run it on propane; we have a mature propane infrastructure already.

Re: They look stylish...

What, a compromise between reduced drag and practical cargo capacity? Those bastards!

As the relative success of the new Ford Maverick truck shows, configurability is itself a selling point for a significant chunk of the market.

Even I might someday consider an EV (though there'd have to be much better charging options on the routes I most often drive) if I could get it with adequate cargo capacity, a wagon-style body, and good-quality seats — because after ten or twelve hours that does start to matter — and without idiotic "tech" features like a touchscreen, smartphone integration, etc. And there's no reason why a standardized platform, or even a proprietary but customizable one like what Kia are pitching, couldn't offer that.

Re: Nothing more annoying

Well, there's differential privacy. But computing information leakage is complicated and expensive, and de-anonymization is indeed easier than many people think.

Re: Nothing more annoying

Agreed, though my approach has been to stop using sites that have annoying ads. Also, I use a Javascript blocker, and only enable ad-related scripts for sites I really, really want to support, and where the risk of malicious scripts is small.

I'm happy to participate in advertising-based revenue streams for content I want, as long as the advertising is approximately no more intrusive than it is in print, or was back in the days of broadcast commercial television (which I dimly remember, though I haven't watched any in quite a few years now).

Re: Insert Elon Musk snark ..

Getting tedious ..

He certainly is. Thank goodness for the snark.

Re: Hahahahahaha.... no!

Dang, that pair of downvoting Musk fans are busy today.

Re: Easy way to avoid it

support for Carplay and the Android equivalent

Personally, I wouldn't trust either of those for a moment. But YMMV.

Re: Inkwells

That'd make a lovely scene in a period film. (Yes, I know, a mere 65 years ago. Ancient history for today's audiences. The Angel of History faces backward.)

I agree I could have done without the hours in primary school learning to write cursive. On the other hand, back in the day being able to read cursive was indeed very useful, sometimes indispensable.

And I do occasionally wish I'd been forced to learn some sort of impressive script, like copperplate, in my formative years. Had an acquaintance in grad school who had, and his handwriting — just the everyday stuff he did taking notes in class or drafting papers — was a thing of beauty. I eventually developed passable calligraphic skills in Japanese (not that I remember a single kanji these days), but my English handwriting is a scrawl illegible to all others.

Re: Wrong

A ban has no chance of "mak[ing] ransomware go away entirely". None. Zero. It is remarkable that so many are naive enough to believe otherwise.

"Just reinstall from backup" is the way commentators on ransomware stories spell "I can't be arsed to pay attention".

Re: Brutality

But since they're often state-sponsored, they're probably being given a flash car and a promotion.

Hell, in North Korea, "my family eats tonight!" is probably sufficient incentive.

Re: Escalation?

Oh, they have been attacking each other. There's plenty of turbulence in the IT crime sector.

The problem here is that there are at least two major sources of innovation in extortionate IT crimes (ransomware, etc). The private-sector gangs mostly operate on an affiliate model; since the gang leaders aren't performing the attacks themselves, they need to introduce innovation in order to justify their own existence and distinguish themselves from other gangs to their affiliates.

Meanwhile, the government-sponsored and -allied groups are under more direct performance pressure from their task masters. If you were running a North Korean ransomware team, would you want to show Kim a graph that's not trending in the right direction?

So some of the innovation may be due to supply constraints, probably as much because of competition as because of improving security posture by victims. But a lot of it is due to increasing demand.

Re: A far better

True. But security is a matter of economics. Improving defenses and removing the vulnerabilities you can identify raises costs for attackers, and that in turn reduces successful attacks.

Improving software security is justifiable even if it can't achieve perfection.

Re: A far better

There are alternatives to using passwords for authentication. While there are problems with all of them, there are problems with passwords, too.

Re: A far better

That's why either liability or regulation is needed. Those are the ways we turn externalities into direct costs, and make them apply equally (well, in an ideal world) to all vendors so as to remove the market incentive to violate them.

Re: Would it not be possible to give a patient list to the police...

I don't think those consequences were unintended.

Of course the situation is always more complex than any synopsis can convey. I have friends and relatives in the police, and I've known other good police officers who are deeply concerned with avoiding unnecessary violence and injury, applying the law fairly, and so forth. The post-2001 militarization of police forces is indeed a big problem. So is the ever-broadening mandate of police forces to deal with all the social problems that states and the public have given up on: domestic strife, child welfare (aggravated by popular dangerism), homelessness (often coupled with mental illness), alcohol and drug abuse, and various other ills. The war on social services that began during Nixon's presidency and has continued since has severely overburdened many police forces, and put officers in situations they're not adequately trained for, nor given resources to address properly.

Then we have the problem of similarly under-trained officers being asked to respond en masse to (supposed) hostage situations and the like, as in the Finch case, where they're inadequately informed and under-supervised.

And we have various Federal agencies trying to use local law enforcement as proxies (e.g. the TSA) or in conjunction with their own forces (e.g. the DEA, ATF, ICE, etc).

Meanwhile, abuse of qualified immunity, endorsed and encouraged by SCOTUS (particularly Antonin "Drop a Tree" Scalia, of blessed memory, and Clarence "Nothing's Cruel" Thomas), has emboldened those who'd like to see more police thuggery. And politicians know that on the balance "tough on crime" often gets them votes, particularly since a majority of Americans are convinced that violent crime is getting worse even though the opposite has been true for decades.

Re: Would it not be possible to give a patient list to the police...

Isn't any level of suspicion noted for calls originating outside the geographic area of the alleged SWAT worthy situation?

Yes, but that's not a reliable indicator. Barriss used VoIP (via a public library's guest WiFi) to call Wichita City Hall, where he convinced the person who answered the call to transfer him to Wichita PD. For the police in that case, the call was local, because it was routed through City Hall.

Now, they should have noted the call was not coming from the claimed address; and the person who transferred the call should have flagged that it was being transferred. There were a lot of mistakes made in the Finch case from the moment it started — and no one paid for any of them. (The officer who shot Finch was promoted to detective, for example. None of his superiors were disciplined for their mishandling of the situation either.) But it wasn't quite as simple as the police falling for an out-of-state caller.

Re: Would it not be possible to give a patient list to the police...

The police are pretty good at doing that already

Citation needed. What fraction of police responses are triggered by swatting? In what fraction of those have police identified the culprit? Show your work.

Re: Would it not be possible to give a patient list to the police...

if a swatting should be attempted (given that it has already been threatened), the police can visit the premises in a slightly less gung ho method

Yeah, good luck with that.

In smaller communities — towns and smaller cities — you're probably going to get a more careful and proportionate response regardless. In larger ones, there's a good chance the police force will go all Wichita without bothering to check any "list" that was provided to them. In the Wichita Finch case, the swatting call didn't even come through emergency response — Barriss called Wichita City Hall and asked to be transferred to Wichita PD. If that didn't give them pause, do you think they'd stop to look at a list provided by some random medical firm?

And, of course, since there are no consequences for the police, either for supervisors or for officers who actually pull the trigger, in these incidents, they have no institutional motivation to try to correct the problem. Certainly there are many good police officers who are sincerely troubled by it, but there's no structural pressure.

Yes, the only reason anyone would object to torture is because they support the potential victims. For our foes, it's just fine.

Or maybe you and the OP and all the upvoters should get a fucking grip and try to behave like civilized people?

Re: "is Perplexity's AI the answer?"

Kind of surprised I had to go halfway down the comments page to find a mention of it.

Re: AI not needed in search

Web search (I typically use DDG) works fine for me, for the cases where I've used it in the past. For real research, I use real tools.

Certainly web search could be better. What tool couldn't be?

But also certainly "AI" won't make it so. We already have a technology for doing conversational research; it's called "research librarians". SotA LLMs are far, far behind their capabilities.

Re: Where To Even Begin

Where do Trump and Xitler find these fools?

There are strategies under which this is a good tactic — particularly at this historical moment. Even if it had no chance of succeeding (and that's by no means certain in 2024), it's a way to exhaust the NLRB and the employees it's representing, and push for arbitration or a settlement.

Re: Dear Elon (you ignorant slut):

Since ~half the US population appears to have decided that an authoritarian con-man bully is the better alternative, I don't think that "rising up" will do us any favors.