* Posts by Michael Wojcik

12299 publicly visible posts • joined 21 Dec 2007

Brit watchdog shows some teeth over McAfee antivirus auto-renewals

Michael Wojcik Silver badge

Re: Norton

What about Norton Antivirus is worth even £20 / year?

Michael Wojcik Silver badge

Re: McAfee?

Hell, even on Windows, Microsoft Windows Defender is at least as good as most of the commercial offerings for most users. The value proposition for McAfee appears to be corporate central control (which, I can say from experience, is primarily used by IT departments to make work impossible). I assume the value proposition for Norton is "hey, we use the name of this guy who used to have a good reputation for PC stuff about a million years ago".

Michael Wojcik Silver badge

Re: It's not just McAfee...

But anti-malware software is extremely useful and well worth the money, so there's no incentive to use dirty tricks to extract money from customers, right?

Or, perhaps, it's mostly useless snake oil? Nah.

Michael Wojcik Silver badge

Re: An easy way to quit.

Virtual cards from a service such as privacy.com (which I use) also work well for this. One click and the card is closed. Privacy cards are also locked to a single merchant, so card-information thefts have much smaller risk, and it's easy to set and change spending limits.

Michael Wojcik Silver badge

Re: Good. Now for the rest.

Ugh, those Prime Student pushes are so damn annoying. I try to avoid Amazon, but Every. Time. I. Order. Something. they bug me to sign up for Prime Student (as well as bugging me to sign up for regular Prime, which ain't happening, Amazon).

If only other online bookstores (B&N, etc) could sort out their ordering and fulfillment systems. I had a minor holiday crisis with B&N some years ago where they held up an entire order of gifts for several people because they'd cheerfully let me order one book which was not in stock, and indeed has never been in stock since. If I hadn't called, discovered the problem, and browbeaten them into sending the rest of the order at the last minute I'd have been out of luck.

(I shop in local stores when I can, but often people want things which aren't available locally.)

Hard cheese: Stilton snap shared via EncroChat leads to drug dealer's downfall

Michael Wojcik Silver badge

Yes, there are no privacy implications whatsoever from our infallible police forces, ever-fair guardians of justice, combing through photographs to extract palm- and fingerprints. I have the utmost confidence they will use this technique only to find those they know are Bad People, or are pretty sure about, or figure might be more or less Bad People, or would maybe do some Bad Stuff at some point in the future, or might annoy them at some time, or might annoy someone in a position of influence, or ...

But, I mean, an alleged drug dealer was caught,1 so that excuses anything. Mustn't have drugs. Except alcohol, obviously. What good are my civil rights if someone else somewhere is getting high?

1We're told he confessed, and that's indisputable proof, right?

Michael Wojcik Silver badge

Re: Hard cheese

A clear queso unintended consequences.

(Thought I'd wedge one more in there. These aren't hard.)

Michael Wojcik Silver badge

AFAIK the police infiltrated EncroChat simply by joining

According to reports, by getting a mole hired as an employee and compromising the software. End-to-end doesn't help in that case, as long as the client (on either end) is compromised.

ObReference to "Reflections on Trusting Trust", etc, etc.

Michael Wojcik Silver badge

Yes, but in the EncroChat case they did it without "cracking the encryption" per se. Police forces (according to reports) got a mole hired by the company, who introduced a back door into the system.

Is that any better? Debatable.

Microsoft: Behold, at some later date, the next generation of Windows

Michael Wojcik Silver badge

there will no longer be such a thing as the tech sector

Yes, everyone will be running their own back-office insurance-adjustment applications under their personal copy of IMS. No tech sector at all.

Honestly, why do we pay C-level execs who insist on spouting such obvious rubbish?

Michael Wojcik Silver badge

Re: and I’m incredibly excited

Let's be fair. Wouldn't any amount of excitement about an upcoming Windows release be incredible? It needn't be a lot of excitement.

Michael Wojcik Silver badge

Re: dual boot.

Just look at all the jobs in Task Scheduler that are triggered by boot or initial logon. Windows does an insane amount of stuff when it first comes up. Whether most or any of it is desirable is another question. (I disable most of what Microsoft crams into Task Scheduler whenever I get a new work machine.)

Then, of course, there are all those oh-so-necessary services that start up at boot. Why, you wouldn't want your attack surface to be missing UPnP, would you?

Fluid components and custom scenes: Microsoft pitches developers with new tools for Teams

Michael Wojcik Silver badge

Re: Have they fixed the bug where it rings but there's no notification to answer incoming calls?

Or sorting the Teams list alphabetically. Currently you have to sort teams manually by dragging and dropping, which is onerous if you're on more than a handful of teams (I'm on 38 at the moment) and works poorly (as drag-and-drop often does, particularly in crap Electron apps). And if you don't sort them, finding the team you want is an enormous pain in the ass.

This would be trivial to implement. It's something an intern could do.

But yes, Microsoft, continue to give us eye candy no one asked for.

Teams is awful. The underlying technology is awful (SharePoint – the worst way to make data available since clay tablets were invented), the UI is awful, performance is miserable.We used to use RocketChat for the chat / message-board functions; it was far from perfect, but it was much better.

I'll admit that for internal videoconferencing Teams is actually no worse than the many other products we've used (PVX, Bridgit, Skype, Lync/SfB, etc), and better than some (such as Go To Meeting or WebEx), in my experience. But its other functions are pretty much rubbish.

Apple is happy to diss the desktop – it knows who's got the most to lose

Michael Wojcik Silver badge

Re: "Legacy of single user on a disconnected PC"

It's not compatibility with the Win 3.x line which harms the security architecture of the NT line; it's user habits. That's why Vista introduced the split token and UAC – because Microsoft had given up on trying to wean people off doing everything with administrative access, and figured they'd just try throwing a half-height security barrier in the way.

Users, even technical ones, are notoriously resistant to security measures that affect their workflows.

That said, there are some problems with NT's security architecture, like the excess authority required to monitor processes (owned by other tokens) for termination. And the biggest issue with Windows security remains its enormous and crufty attack surface. Even without excess permissions, if you can drop a keylogger through an RCE you can eventually capture the current user's credentials, and pivoting and escalation will almost certainly be possible.

Michael Wojcik Silver badge

Re: That Harvard Guy's bio ....

James Mickens is hilarious. Watch some of his recorded presentations sometime – he has a great one on Byzantine consensus protocols, for example. You're not supposed to take that stuff seriously.

Honestly, the ability of some people in IT to utterly fail to understand jokes remains impressive even after decades of observing it.

Virgin Galactic goes where it's gone twice before, for the first time in two years

Michael Wojcik Silver badge

Re: When does the final crash test dummy go up on a test flight?

To be fair, I didn't know that was his plan either. But then I also don't particularly care whether he goes or not.

It's nice that we have multiple private firms innovating in this area, and I think it's fine that Branson is trying to use tourism to subsidize and advertise a bit. Not a personal priority for me – research is better done by machines, the long-term survival of H. sapiens is not something I'm invested in, and space travel will never be practical for all but the tiniest fraction of the population – but improving the technology is good.

Michael Wojcik Silver badge

Re: I thought space was 100km?

They could have gotten a bit closer. White Sands is at an elevation of about 1900m above sea level. Wheeler Peak, the highest point in New Mexico (which I can see from the Mountain Fastness), is at 4000m. Just launch from there and you've made up nearly 20% of the difference!

(No, I'm not seriously suggesting this.)

Michael Wojcik Silver badge

Re: I thought space was 100km?

Our space is the best! It's yuge! Other countries have to go further just to get to space, but in America it's right there.

Tesla owners win legal fight after software update crippled older Model S batteries

Michael Wojcik Silver badge

Re: Carbon neutral

EVs are the future, even if they come with a fuel cell, or a thorium reactor, on board

Or one of those fancy internal-combustion engines I keep hearing about.

Diesel-electric works for locomotives. Why aren't the EV fans pushing for a hybrid pure-electric powertrain with ICE onboard generation? (Hybrid drivetrains are idiotic.) Solves the range and refueling problems, and if you want to charge it from the grid when it's parked for long periods, you're free to do so.

The only passenger cars of this sort that I've seen are exotics. It would make more sense than an electric-battery design for a pickup, or for anyone who needs to drive long distances.

Amazon continues its ban on allowing police to use its facial-recognition software

Michael Wojcik Silver badge

Indeed. Police departments in the US have already demonstrated that they're more than happy to evade warrant requirements and other regulations by purchasing data from aggregators and resellers. This is a token PR move by Amazon; they and the police know it's essentially meaningless.

American insurance giant CNA reportedly pays $40m to ransomware crooks

Michael Wojcik Silver badge

Re: "Phoenix is not on any prohibited party list and is not a sanctioned entity."

Indeed. It's very difficult for a ransomware crew to change its name and ... oh, wait.

Michael Wojcik Silver badge

Re: Governments need to intervene here

That will not work.

There's already a strong incentive not to pay: it costs money, it's risky, it's bad PR, it looks bad to investors. Yet companies pay anyway, because the alternative is worse for them.

Executives can always find a proxy and construct plausible deniability for making payments. Prosecution would be very difficult, and prosecutors hate difficult prosecutions. (See Eisinger, The Chickenshit Club.)

And (as I keep pointing out) even reducing payments by orders of magnitude won't eliminate ransomware attacks, because the cost of mounting those attacks is extremely low.

Governments already promulgate all sorts of IT-security requirements. The Biden White House just issued a new batch. They haven't helped much yet, and there's no reason to believe they will in the foreseeable future.

Michael Wojcik Silver badge

That's not generally the way it works. The attacking organization has a botnet probing for known vulnerabilities it can exploit to drop a ransomware package, which will then encrypt files and notify a C&C server. The humans only find out about it after a victim has been compromised. There aren't a bunch of pasty-faced yoots in hoodies hunched over keyboards manually encrypting a file at a time.

Some ransomware includes exfiltration of data; some doesn't. A given crew might, at some point, upgrade their botnet to deliver a package that includes exfiltration capability, but while the money's still rolling in there's no great incentive to do so quickly.

There are probably ransomware operators who still work manually, but the smart ones will be automating the process as much as possible. And aside from developing packages with novel capabilities, it can all be automated.

That's one reason why outlawing payments won't stop ransomware attacks.

New IETF draft reveals Egyptians invented pyramids to sharpen razor blades

Michael Wojcik Silver badge

Re: RFC is "rough draft"?

I-Ds (Internet Drafts) are drafts of some sort, rough or otherwise, and that's what we have in this case.

Frankly, while Kumari's draft might be amusing (I haven't read it, and the excerpts quoted in the article didn't inspire me to do so), I don't have much sympathy for his complaint. Some people will cite I-Ds as authoritive. So what; people will cite all sorts of things. Those who understand the IETF know that I-Ds are not normative and neither are many RFCs, only some of which are even on the Standards Track.

The archive that the IETF maintains of I-Ds is nonetheless useful, because some I-Ds never make it further but nonetheless become de facto standards, or at least a guideline for implementation, where no other standard exists. draft-ietf-tn3270e-extensions-04.txt (which I don't think made it to an RFC) is one example.

There are others of historical or theoretical interest, such as draft-ietf-usefor-useage-00.txt.

Blessed are the cryptographers, labelling them criminal enablers is just foolish

Michael Wojcik Silver badge

Re: Peer review

The suggested venues might get some attention from hobbyists, but real cryptographers and cryptanalysts typically have more important things to do with their time. Poking holes in ciphers with no compelling advantages from unknown authors gets tiresome quickly.

Michael Wojcik Silver badge

Re: Peer review

Frankly, if you're serious about cryptography, you almost certainly shouldn't be devising new ciphers, except for your own amusement and practice. Anyone serious about cryptography should understand the state of the art, and that state is "we don't need new generic symmetric ciphers". Barring a historic event in cryptanalysis, no one who knows what they're doing is going to go through the huge cost of rolling out a new symmetric cipher that isn't PQ1.

And someone who's serious about (machine, production) cryptography ought to know that. That's a basic fact of the market. You don't even need to understand things like linear cryptanalysis and the Random Oracle Model to understand that replacing the AES infrastructure would be enormously expensive, and doing it with a cipher that hasn't received many years of scrutiny would be enormously risky.

1That is, resistant to algorithms in BQP.

Michael Wojcik Silver badge

Re: Peer review

We don't need new generic symmetric-encryption algorithms.

No new block cipher is going to have compelling advantages against AES unless a practical novel attack is found against AES; after decades of cryptanalysis by a wide range of experts, that seems very unlikely. A new algorithm, on the other hand, isn't going to have decades of cryptanalysis by a wide range of experts.

Meanwhile, we have widespread hardware acceleration of AES. No new cipher is going to have that advantage.

No new stream cipher is going to out-compete AES in a streaming mode for the same reason.

Simplicity isn't automatically a virtue. RC4 is extremely simple. It turns out to have high-order correlations which are not at all obvious and make it too dangerous to use in the modern world.

Development these days is focused on other areas. Post-quantum cryptography, for one (though it's too late to get into that game unless you have a major breakthrough). Homeomorphic encryption for another. Partial-information-preserving encryption. Integrating encryption with differential privacy.

But new generic symmetric ciphers? They're a dime a dozen, frankly, and they're all risk with no return.

Michael Wojcik Silver badge

Re: security by obscurity

I had a friend who had a soft-top convertible in Boston. He kept signs in the windows that read "DOORS UNLOCKED" to discourage people from cutting the roof to get at the interior.

Open-source JavaScript project Babel 'running out of money' after employing paid maintainers, sponsors pull out

Michael Wojcik Silver badge

Re: Holy shit

In many parts of the US, $11K / month pre-tax is not a lot of money to live on. Even in the cheaper areas, if you're supporting a family, paying off college debt, etc, it's comfortable at best.

SolarWinds CEO describes overhauled Orion build system after that 'very small, unique' security breach

Michael Wojcik Silver badge

Re: a “very small” number

Known to have been compromised. And I would call that a "very small" consolation.

Considering that in the same sentence he managed to jam in the patent untruths that the attack was "unique" and "very novel", I think it's safe to dismiss the entire speech as utter bombast and bullshit.

Making excuses to avoid making amends.

Uncle Sam wants 'ethical hackers' to crack its planetary defenses, but don't expect a pay-day from this bug bounty

Michael Wojcik Silver badge

Re: "The US DoD has opened up all of its publicly facing systems and apps to investigation"

They don't need to.

These systems are publicly-facing, so they're already "open" to state-sponsored actors and other professionals. "Opening" these systems in a case like this just means "we won't hassle you if you look for vulnerabilities".

Since publicly-facing systems are already under attack (all of them, constantly), there's nothing new here as far as the professionals are concerned. And, of course, by logging attacks and feeding those logs into SIEM / UEBA systems, you learn some information about your attackers.

Michael Wojcik Silver badge

Re: The prawn of doubt

Sure, everyone comes down on the Strategic Lobster Initiative, but it's the major justification for the National Melted Butter Reserve program, and if Congress cancels that we're all in trouble.

‘Staggering’ cost of vintage Sun workstations sees OpenSolaris-fork Illumos drop SPARC support

Michael Wojcik Silver badge

I'd enjoy having a Sun-3 for the same reason. I think that was my first UNIX workstation. Used it at university for C, LISP, and Scheme coursework.

Or an IBM RT PC, which was the first workstation I used at work, and consequently wrote considerable non-trivial software for. Mine ran AOS (IBM's BSD port), not AIX. Unlike the SGI machines there was nothing sexy about the RT PC, but I have a perverse fondness for the ol' boat anchor.

Telcos crammed 8.5m fake comments against net neutrality into FCC's inbox

Michael Wojcik Silver badge

Re: Confirm email address?

I think you mistake the point of the exercise.

Pai already knew what he was going to do – he had his instructions from his bosses at Verizon. The whole public-comment process was just a show to make it appear some sort of deliberation had taken place. As such, it was useful to have a large volume of comments, but not at all important that they be genuine.

JET engine flaws can crash Microsoft's IIS, SQL Server, say Palo Alto researchers

Michael Wojcik Silver badge

I can see their point

The team also said Microsoft dismissed some of their findings as not worthy of a fix.

"Look, we could fix this, but it would still be JET."

I admit I wouldn't be in a hurry to crack open that can of despair either.

Microsoft has gone to great lengths to push its tech, but survey suggests many devs slipped through the .NET

Michael Wojcik Silver badge

It's possible to put windows side-by-side, so you can change the focus by moving your mouse. Or to run msbuild from within vim.

But, yeah, there's really no good debugger for .NET / CLR programs on Windows. WinDbg with SOS (or whatever they're calling the managed-code debugging extensions these days) really isn't viable; I use WinDbg for native-code debugging on Windows but it's largely useless for managed code. I don't think mdb is even supported anymore and it was always only marginally usable at best.

When I have to debug, I use Venomous Studio, much as I loathe it. It's uniformly terrible, but it's the only thing I've found that works, at least for the sorts of things I have to debug.

Of course, a lot of the managed code I'm debugging is written in managed modern-syntax OO COBOL, which source-language-sensitive debuggers will have trouble figuring out. (Visual Studio can because of our extensions, obviously.)

Belgian parliament halts China Uyghur 'genocide' debate after DDoS smashes ISP offline

Michael Wojcik Silver badge

Re: Give us a clue

There are a lot of people in China, and it's wildly improbable that none of them have access to a botnet. This attack could have originated in China without having been requested or sanctioned by the Chinese government.

It's foolish to simply declare "it was China" with no evidence, because as you say there's little incentive for this sort of thing as a matter of foreign policy. It's equally foolish to eliminate all of China – government and private citizens – from consideration.

The position the Reg took in the article is the sensible one. We don't know. And, really, it doesn't much matter.

American schools' phone apps send children's info to ad networks, analytics firms

Michael Wojcik Silver badge

Re: Really ?

You'd have to be brainless not to figure that a healthy, literate (and armed) population will not tolerate much abuse.

Or have a competent understanding of psychology or history.

Healthy, literate people not only tolerate abuse, they actively participate in it. Being healthy and literate is better than the alternative, but it is by no means a magical prescription for freedom – and the capability (which is dubious however many personal weapons might be spread among the populace) to subvert the government's monopoly on violence doesn't change that, because the participation in abuse is almost entirely an effect of ideology, not repression.

Octopath Traveler: Love letter to JRPG golden age has great combat but retro graphics highlight the genre's tedium

Michael Wojcik Silver badge

I have to admit it sounds interesting to me, in much the same way that SaGa Frontier's "Free Scenario System" was. (SaGa Frontier is actually in the same family as the Final Fantasy games – the "SaGa" brand was used for the latter in Japan.)

In SaGa Frontier, you have your choice of seven protagonists, and with each one you can eventually assemble a party containing the other six. Then when you finish a protagonist's story, you go back and pick another one. So the entire game consists of taking the same seven characters through seven interrelated stories in the same setting, from seven points of view.

Based on the article I understand Octopath's approach is different, but I still like the idea of assembling a narrative from pieces rather than having it just supplied by the game, as with most of the Final Fantasy titles. Not that I haven't enjoyed several of the latter. (I haven't played them all because I'm a very late adopter and I generally only play video games while running on my treadmill, so it takes me years to get through a typical RPG.)

Lambda School, a coding bootcamp that takes a cut of your next tech salary, now takes a 30% cut in staff

Michael Wojcik Silver badge

Re: Identured Servitude Agreement

I can't wait to see what Slavery will be modernized euphemistically into by these clowns.

"Rehabilitation." Already done.

Bitcoin is ‘disgusting and contrary to the interests of civilization’ says famed investor Charlie Munger

Michael Wojcik Silver badge

Re: Insert meme here

In order to get power over Bitcoin you need a huge amount of computing power.

Wrong. Network-partitioning attacks against Bitcoin are not only possible, they're quite common. There are multiple papers on the subject.

Ex Netflix IT ops boss pocketed $500k+ in bribes before awarding millions in tech contracts

Michael Wojcik Silver badge

I think you're giving him too much credit, and he just thought he was being clever. Criminals often overestimate their own skills.

Michael Wojcik Silver badge

Re: This is Not Embezzlement, This is Capitalist Entrepreneurial Spirit

Looks like a Poe factor of 0.8 on that comment.

Of course, meta-Poe says we can't tell whether you're trying for a high Poe factor (a deliberately ambiguous post intended to elicit a maximally-mixed reaction), or for a low one and you just aren't making your tone clear.

If it's the latter, name checks out.

Michael Wojcik Silver badge

Re: Don't trust those with purse strings!

The reason white collar crime is so prevalent is because its mostly unpunished. Fix that and you fix white collar crime.

I find your abundance of faith disturbing.

Pretty much everything we know about human beings, particularly from psychology and behavioral economics, tells us that people are not rational economic agents. The credible threat of punishment may deter some crime; it does not, and never will, eliminate it.

Yahoo! and! AOL! sold! for! $5bn! as! Verizon! abandons! media! empire! dreams!

Michael Wojcik Silver badge

Re: Those CDs were not SPAM

I knew people who hung them in windows to warn birds away. (There was a popular notion this would discourage birds from flying into the glass; I have no idea if this has been tested in any methodologically-sound fashion.)

For a while it was popular in some circles1 to hang one from the rear-view mirror of one's car, which would occasionally reflect the sun right into the driver's eyes.

And, of course, they were commonly used as coasters.

1Heh.

Michael Wojcik Silver badge

Re: how come yahoo is still a thing?

The original Yahoo information model was quite clever. Unfortunately they abandoned it and the search architecture built on top of it, losing their only distinctive feature.

Michael Wojcik Silver badge

Re: Brilliant Business Strategy

Don't forget the previous incarnation: AOL + Time-Warner in 2000. $182B. But, of course, it made TW what it is today (part of AT&T and worth less than half of that 2000 price).

Good: Water vapor signal detected for first time on distant planet. Bad: Er, we'll let one of the boffins explain

Michael Wojcik Silver badge

Re: So it's not a dry heat then?

Apparently many people find the word "moist" cacophonous or have unpleasant associations for it. It shows up regularly in lists of most-disliked words. ("Panties" is another frequent flyer.) It doesn't bother me, personally, and alternatives like "slightly damp" are usually awkward.

Appeals court nixes online blueprint sharing ban on 3D-printed 'ghost guns'

Michael Wojcik Silver badge

Re: Obviously more guns needed

Ah, some fine kookery from one of our resident kooks to liven up Monday.

"As soon as I find a HOLLOWED-OUT VOLCANO for lease in BRITISH COLUMBIA you will all be forced to RECOGNIZE MY GENIUS."

On the other hand, it's hard to pick between SSg7's 20 bazilliawatt laser and the coming flood of 3D-printed assault rifles being mass-manufactured by terr'ists and gangbangers for Fake Threat to Ignore o' the Day.

I'm not a fan of guns, any more than I'm a fan of chainsaws; they're both tools, and dangerous ones,1 and it would be good to keep them away from idiots and assholes but that's largely infeasible. But 3D printing does not seem likely to greatly aggravate the gun problem in the US, and forbidding it will almost certainly not help anything in any significant way. Regulations controlling the sale of such guns might help somewhat, at least in making them even less economically attractive than they already are.

1JFTR, at this time I own one chainsaw and zero guns. That could change; there are critters about the Mountain Fastness, including some rabies and plague vectors, among other possible reasons for wanting a firearm.

Michael Wojcik Silver badge

Re: Why bother with 3D printing

Name checks out.