* Posts by Paul Kinsler

423 posts • joined 9 Aug 2007

Page:

PoisonTap fools your PC into thinking the whole internet lives in an rPi

Paul Kinsler

Hmm...

Presumably there's some way of configuring dhclient so that it only tries known/pre-specified interfaces?

I can see from the man page how to set options for specified interfaces, but not how to ignore others which might appear (with potentially unexpected names or numberings).

0
0

Boffins find Galaxy making killer radiation, rule out Samsung phone as source

Paul Kinsler

[mathematics] "frightens even physicists"

There is also an interesting article/comment/reply sequence discussing this here (all in NJP hence open access)

http://dx.doi.org/10.1088/1367-2630/17/1/013036

http://dx.doi.org/10.1088/1367-2630/18/11/118003

http://dx.doi.org/10.1088/1367-2630/18/11/118004

1
0

CERN boffins see strange ... oh, wait, that's just New Zealand moving 2m north

Paul Kinsler
Happy

Re: That's one hell of an expensive seismometer

But now I'm just left wondering "where is the data from the proper (LIGO) seismometer?"

0
1

Twitter trolls are destroying democracy, warn eggheads

Paul Kinsler

Re: following it up with a comment from the Librarian!

We are always interested to hear more about the works of Dmitry Alexandrovich Gromov.

0
0
Paul Kinsler

Re: Pre web? Did you never see/use Usenet?

Early internet "social media" were (at least in my experience) just that, /social/ media: e.g. usenet was based on groups, irc had channels. You might have thrown your ego about in such forums, and some were admins or moderators, but essentially the forum was about the group or social interest of those who turned up and posted/talked there.

In contrast, contemporary social media tend to be explicitly about ego: e.g. /my/ twitter feed, /my/ facebook page. I think it changes how we see and use the medium, and what we expect of it. Likewise, the up/down-vote scoring sometimes seen in this kind of "ego-media" changes the default experience of the internet glasshouse.

2
0

What should the Red Arrows' new aircraft be?

Paul Kinsler

Re: Lightnings! For about thirty seconds until they were out of fuel.

Perhaps Canberras then?

1
0

How do you make a qubit 10 times as stable? Dress it up for work

Paul Kinsler

Re: And the bullshit keeps rolling on

Science almost always /isn't/ the stuff which make it into science press releases - do not confuse media reports of science with the real thing.

Feel free to peruse some open-access real-science at e.g.

http://iopscience.iop.org/journal/1367-2630

https://www.osapublishing.org/oe/home.cfm

https://journals.aps.org/prx/

1
1

NHS patients must be taught to share their data, says EU lobby group

Paul Kinsler

Re: When you can guarantee me that my data won't get pilfered by hackers ...

I think that even with the best will in the world that guarantee will never happen.

More practically, there should be defined compensation to be paid by each entity using the data, to each person whose data it is, in the event of a breach and or/leak and/or misuse.

Then we will be able to be sure there will be real incentive for the data to be handled correctly, and so have a reasonable hope that it /will/ be handled correctly... but if not, compensation will be paid. Perhaps 10k per record over tens of millions of patients might concentrate the mind a bit.

2
0

The answer to Internet of Things madness? Open source, of course!

Paul Kinsler

Re: "Look, I can do stuff with it over the Internet" as if that's a good thing.

Well, to be fair, I might well have bought several sorts of mostly pointless iot-like objects by now, if only for the entertainment value of mucking about with it over the net ... ... ... except for the fact that I manifestly couldn't trust them, what with the various "communicate with manufacturer's servers", security problems, no proper patching, not open source, obscure control protocols, and other fun features they have.

3
0

AI software should be able to register its own patents, law prof argues

Paul Kinsler

Re: if you can't comprehend enforcing a patent then you don't get one.

It seems to me you could also use this as a test to determine whether humans (or indeed anything else) should be able to patent things. Is this the intent of your criterion, or are you just trying to discriminate against things that might be called AI's? :-)

It seemed clear from the article that the prof was just thinking ahead, since from the third-last paragraph: "The idea of computers achieving legal personhood status is a long way away, he acknowledges". So it doesn't seem likely that he was thinking about contemporary so-called "AI" systems, which is the ones you seem to be objecting to in this context.

0
0

Simpsons creator Matt Groening once drew Mac heaven for Apple

Paul Kinsler

Alpha

There were a couple of Alpha's in the physics department at UQ before 1994; they may have been blindingly fast, but seemed less so when you didn't have (i.e. couldn't afford) enough memory and had to do large matrix calculations by swapping virtual memory off disk. I recall my computations were so i/o bound that I got maybe 4% of the cpu speed.

Not the fault of the machine, though, which was quite nice.

0
0

Super Cali: Be realistic, 'autopilot' is bogus – even though the sound of it is something quite precocious

Paul Kinsler

Re: Autopilot on an aircraft is seen as an aid to the pilot, ...

I think "autopilot" is likely to be taken to mean "pilots itself in most ordinary circumstances".

What might count as a perfectly adequate set-and-(partly)-forget autopilot for an aeroplane, allowing a pilot to take care of other tasks whilst being prepared to take back control is hopelessly inadequate for coping with the complexities of ground/road traffic: aeroplanes have a much reduced traffic density to care about, and aren't confined to bendy roads with intersections, multifarious obstructions to vision, and so on. A road-traffic autopilot needs to be vastly more sophisticated than one for air-traffic to achieve the same "pilots itself in most ordinary circumstances" ability.

4
0

Ofcom punts network-sniffing Android app

Paul Kinsler

Re: So why download an app whose it's main purpose is to gather data?

Just because they are ok with the collection with one sort of data (signal strength/ network performance or whatever) does not mean they are necessarily happy with the collection of other data - e.g. their use of games or social media apps.

That said, I've no idea what the data collection was or is in this case - just making a distinction which may be helpful here.

4
0

Infected Android phones could flood America's 911 with DDoS attacks

Paul Kinsler

Re: I can't see what the Android angle is

"The hackers used a discrete event simulator (DES) and a handful of Samsung phones to test their work" ... presumably android ones, since that's what the majority of Samsung smartphones are?

0
2
Paul Kinsler

rules would have to be tweaked to allow dropping emergency calls from unknown subscribers/devices

A more minimal tweak might be to allow calls from "known" & "unknown" s/devices to be queued separately, and in the event of capacity problems, to answer each queue alternately. Then, in an anonymized DDos, the two queues would be very different lengths so that "known" calls would be more likely to be answered, but you wouldn't ignore all "unknown" calls.

2
0

Typo made Air Asia X flight land at Melbourne instead of Malaysia

Paul Kinsler

Re: The crew cancelled a ground proximity warning,

They might consider themselves lucky they got to cancel it.

I recall that the ground proximity warning was the last thing that the pilots of Air NZ flight 901 that crashed into Mt Erebus heard. That was the result of a coordinate input error also, if I recall correctly (albeit a somewhat more complicated one).

2
0

It's time for humanity to embrace SEX ROBOTS. For, uh, science, of course

Paul Kinsler

actual evidence for anything being a "gateway"?

Just because there are many dubious and/or overblown "gateway" claims, typically being somewhat doom-laden, does not mean there are no "gateway" like scenarios which actually occur.

Notably, I might recast a common, and largely uncontentious narrative of the form "inspirational teacher in year X led me to do Y" as a "gateway experience in year X led me to do Y" - the actual events haven't altered, just the language used.

Further, there could easily be competing analyses of a situation, with A claiming that there is no significant gateway effect since very few in its presence went further, while B might be able to point out that the posited "gateway" was crucial in the progression of those susceptible individuals.

It seems to me that the existence of a "gateway" does not demand that everyone walks on through it; and evidence for or against a claimed gateway effect needs to look at the specific situation. But I don't think that it's an unreasonable thing to hypothesize that one /might/ occur in a novel situation, as is done here.

7
2

Dark scientists' LUX-ZEPLIN doubles down on WIMP hunt

Paul Kinsler

Re: 'But theoretical physics is based on Mathematics, not observation. ' could not be more wrong.

Well, sort of.

Theoretical physics is based on taking tried and tested mathematical models based on existing phenomena, and applying them to new situations. The resulting/new predictions can then be tested by experiments (at least in principle).

2
0

Microsoft to rip up P2P Skype, killing native Mac, Linux apps

Paul Kinsler

Re: Is anyone using Linphone?

I've had that on my system for years, but never made anything but a few test calls. Currently I'm trying ekiga instead, which seems to have fewer dependencies, which makes getting to the point of making a test call easier. :-)

I'm no fan of skype, but since that's all anyone else wants to use, I seem to be stuck with it.

0
0

If we can't find a working SCSI cable, the company will close tomorrow

Paul Kinsler

Re: Always start with the physical layer,

Absolutely. First check that your Higgs boson has the correct energy, ... :-)

33
0

Graphene is actually self-folding origami, proclaim physicists

Paul Kinsler

Under the microscope, graphene looks like chicken wire.

Perhaps it's a scanning tunneling microscope...

4
0

IoT puts assembly language back on the charts

Paul Kinsler

Apollo assembler code on github

I can only hope there's some 80 year old guy out there who has a quick nostalgic browse, spots something that needs improvement, and submits a patch.

2
0

The Great Brain Scan Scandal: It isn’t just boffins who should be ashamed

Paul Kinsler

Sound to me that "Deep Learning" in Google's Go machine is mostly human curated database.

Hmm. Human chess players not infrequently discuss and analyze famous games, replay them with variations, read books about openings and endgames. It seems reasonable to me that Go players might well do the same. If Human players can make use of a "curated database", surely the computer players can also?

Although it'd be interesting to see the computers make their own curated database by playing themselves. Perhaps they could then write a book about "how to play" for the humans who need help :-)

1
0

Man killed in gruesome Tesla autopilot crash was saved by his car's software weeks earlier

Paul Kinsler

Re: Those are still rare in the U.S.

Really? I recall a conversation I had with an American guy on irc who told me he was inputting some gruesome traffic statistics along the lines of car-under-truck/decapitated-adults lines ... some twenty years ago. On the plus side, he did tell me that small children often survived because their heads were lower down.

0
0

Gun-jumping French pols demand rapid end to English in EU

Paul Kinsler

In the long run, my money is on Mandarin.

I think you may have misspelt "Marain" :-)

6
0

Physicists build simulator, hope to stand up beautiful Standard Model

Paul Kinsler

how they are going to validate their generated models

As an experiment, it /is/ an embodiment of the process of validating scientific models. If the experimental analogy works, we all cheer; if it fails, we try to work out why.

1
0

Hey cloud lawyer: Can I take my client list with me?

Paul Kinsler

Re: try to hide it by deleting texts

I don't doubt that in many cases what you say is true.

However, if one was to believe the well-known quotation: "Give me six lines written by the most honest man in the world, and I will find enough in them to hang him.", then replace "me" by the employer in question, you might also consider that such "delete" actions might just show an element of caution.

1
0

Telegram crammed: Hackers find way to send massive messages

Paul Kinsler

Re: Telegram?!

As a long time slackware user, headlines referring to some newfangled "slack" thing/app are also irritating, since they are both IT-ish. It's not that I don't quickly get the new context, it's just that "slack" generates a sort of brain interrupt that I could do without unless the article really was slackware related (like, when 14.2 is finally announced, or something).

4
0

Bendy see-thru mobes? Materials scientists make nanofilm advance

Paul Kinsler

Re: what is going on with that hand?

" I'm going to guess that they applied the film to someone's hand and that the red color you're seeing is the actual sheet itself. "

Your guess is consistent with the relevant figure caption in the article "Images of CuEW-TCEs (duration of the first electrospinning stage t1st = 30 s) transferred to a human hand and ..."

0
0

'We accidentally hit wrong button on Dell buyout – here's $194m for the cockup'

Paul Kinsler

Re: History revised...

I read that some smaller shareholders were forced to sell once a sufficient majority agreed; they might well be aggrieved at being short changed even if voluntary sellers shouldn't be.

This was covered either in an earlier (semi-recent) Reg article or its comments thread.

1
0

Wi-Fi hack disables Mitsubishi Outlander's theft alarm – white hats

Paul Kinsler

Re: the worlds infrastructure is so easily compromised

I saw the other day a poster for the sequel to Independence Day ... and remembered how ludicrous it seemed that some random ape could hack into a vast starship built by aggressive and intelligent Alien Things, and sabotage it so easily.

But nowadays it seems much more believable :-)

3
0

The Fog of Cyberwar: Now theft and sabotage instead of just spying

Paul Kinsler

Re: while a cyberattack is much, much cheaper to build and sustain.

Only because it's parasitic on existing infrastructure. One could hardly say the internet didn't/doesn't require money and sophisticated manufacturing and maintenance capabilities... just that the cyberattackers don't have to pay for it.

0
0

Systemd kills Deb processes

Paul Kinsler

Re: you'd 'nohup' them to keep them running when you logged out

Indeed. Only if systemd is killing nohup'd processes is it really misbehaving IMO (although I accept that modern expectations may differ, and notwithstanding systemd's other misbehaviness).

0
0

Twitter expands beyond 140 characters

Paul Kinsler

Re: [twitter] keeping in touch with others in my field of science

For myself, I can't see it as useful in any sense beyond the superficial. OK, so I suppose some of the shorter emails that I send might be twitterable. But 140 chars is way to short for me to have any kind of technical conversation, even given that a technical conversation is likely to include mathematics of some kind.

Also:

I can see how you might do PR for science via twitter, that would work well. But science PR isn't the science itself - even a one page press release on a result isn't science, so however apt the analogy you might use in your tweet (or pr) still isn't a fair representation. Which is not to say the twitterpr is worthless, just that it's a very distinct thing to the necessarily detailed science itself.

Science publishing might be broken to some extent, but I fail to see how any valid scientific article can be usefully replicated on twitter or social media - it seems to me that whatever the failings of journal-style articles, whether in traditional journals or not, twitter is not a replacement. It seems to me you have conflated the actual scientific content with the pr/publicity you would like to achieve for that content.

The conference thing is interesting. For myself, I just read/skim the timetable from the conference site and look up the authors or work independently. Twitter/hashtags might give a sense of 'buzz' around some particularly noteworthy result, which I guess might be helpful. But my projects tend to run over timescales that incline me to prefer to avoid obvious bandwagons and seek out still interesting but less trodden paths. I skim a much wider range of article/conference output than is ever likely to even microtrend on twitter.

0
0

Being an IT trainer is like performing the bullet-catching trick

Paul Kinsler

"Captain Haddock cosplay contest at a Geography teachers’ convention"

oh, very good.

41
0

Boffins achieve 'breakthrough' in random number generation

Paul Kinsler

Re: Sure this may have been "common wisdom"

... and not only that, it is very likely possible to combine them in the wrong way so as to enhance any correlations (or make new ones) and so make the randomness worse.

2
0

Would we want to regenerate brains of patients who are clinically dead?

Paul Kinsler

Re: how it feel to to be wrenched back from an even more definite terminal event.

... or, as the article posits, possibly wrenched back from someone else's terminal event...

2
0

Ooh missus, get a grip on my notifications

Paul Kinsler

Re: published article 'explaining' away illicit messages ;-)

That gave me a most brilliant idea for a research paper on Quantum Steganography, only I now see that someone's already done it. But I suppose that technically the topic here is anti-steganography, so maybe there's still an angle...

4
0

Archaeologists find oldest ever ground-edge stone axe

Paul Kinsler

Re: journalist: "they found an axe (which took hundreds of hours to grind)"

Perhaps the extra hundred+ hours were for grinding it down from a normal sized axe to a tiny micro-axe.

6
0
Paul Kinsler

Re: Hmm.

A quick bit of searching confirms my recollection that the Neanderthals had a range in europe/central asia, see e.g.

http://news.nationalgeographic.com/news/2003/03/photogalleries/neanderthal/

Not sure what Homo thingius lot might have predated the current native-Australian population, if any. But afaict, not Neanderthals.

6
0

29 years of data shows no mobile phone brain cancer link

Paul Kinsler

The El Reg article quotes from an article in 'The Conversation' ...

The Reg article says "Lead author Simon Chapman, emeritus professor in public health at the University of Sydney, explains at The Conversation..." - i.e. the article in the Conversation was written one of the researchers who worked on the study and wrote the scientific article under discussion. Unless anyone on the Reg is a cancer specialist of some kind it's debatable whether reading the original research article would have helped.

It might be said, however, that the Reg could have contacted some other researcher in the same general field to get an (alternate) opinion.

Is there something about the original research you want to know in particular?

1
0

Revealed: HMS Endeavour's ignominious fate

Paul Kinsler

Re: the 'RRS Decision by naming committee pending'

Actually, I think I prefer that even to Boaty McBoatface.

Can we get the voting reopened?

3
0

Do you know where your trade secrets are?

Paul Kinsler

There should be no over-ruling this ...

So, we need someone to develop a "write-once" filesystem? Although I suppose even that fs could be subverted, given suitable system privileges, by installing a hacked rw version of the same and mounting (and then altering) the backups filesystem with it.

OK, so all (suitably paranoid) backups should only be to write-once media, then?

0
0

Engineer uses binary on voting bumpf to flag up Cali election flaws

Paul Kinsler

Re: There are 10 types of people. Those the understand binary and those who don't

I agree there are 10 types of people, but I would like to suggest a small correction. The list is: those who understand binary, those who dont, ... and those who understand Gray code.

3
0

Astroboffin discovers exoplanet by accident ... in 1917

Paul Kinsler

Always keep your records in order

Here's an interesting discussion of old astronomical records, and how to preserve them... and their annotations.

The Scientific and Historical Value of Annotations on Astronomical Photographic Plates

Sara J. Schechner, David H. Sliski

http://arxiv.org/abs/1602.03475

The application of photography to astronomy was a critical step in the development of astrophysics at the end of the nineteenth century. Using custom-built photographic telescopes and objective prisms, astronomers took images of the sky on glass plates during a 100-year period from many observing stations around the globe. After each plate was developed, astronomers and their assistants studied and annotated the plates as they made astrometric, photometric and spectroscopic measurements, counted galaxies, observed stellar variability, tracked meteors, and calculated the ephemerides of asteroids and comets. In this paper, the authors assess the importance of the plate annotations for future scientific, historical, and educational programs. Unfortunately, many of these interesting annotations are now being erased when grime is removed from the plates before they are digitized to make the photometric data available for time-domain astrophysics. To see what professional astronomers and historians think about this situation, the authors conducted a survey. This paper captures the lively discussion on the pros and cons of the removal of plate markings, how to best to document them if they must be cleaned off, and what to do with plates whose annotations are deemed too valuable to be erased. Three appendices to the paper offer professional guidance on the best practices for handling and cleaning the plates, photographing any annotations, and rehousing them.

7
0

Reusable autonomous swimming microbots soak up 95% of spilt lead

Paul Kinsler

Re: I'm confused.

quote: "the inner platinum layer decomposes hydrogen peroxide for self-propulsion"

So, with 2 H2O2 -> 2H2O + O2, presumably the bubbles of oxygen are produced in a way/position so as to push the thing one way (or at least move it around a bit more than otherwise).

3
0

Watch: SpaceX finally lands Falcon rocket on robo-barge in one piece

Paul Kinsler

Re: "Standard" brand fireworks. Made proudly here in the UK.

I much preferred those old-fashioned Congreve ones.

1
0
Paul Kinsler

Re: Dat link:

More to the point, is there some way to see the rocket better, using portrait? These widescreen formats are just terrible for rocket take-off/landing videos :)

3
2

Vivaldi Jon: Mobile – yes. Feeds and an ad blocker… probably not

Paul Kinsler

Re: Vastly more powerful?

Presumably, pocket computers are "vastly more powerful than pocket computers were before", not " vastly more powerful than current desktops".

3
0

Intel tock blocked for good: Tick-tock now an oom-pah-pah waltz

Paul Kinsler

Re: Bluntly the greatest requirement for CPU anybody really needs is for gaming,

So, my 3D electromagnetic simulations /aren't/ maxing out the cpu on my workstation after all? And I don't have to wait for them to finish?

perhaps: s/anybody/most users/

2
0

Page:

Forums