Posts by John Smith 19

The economics shoudl be *interesting* to say the least

Nuclear companies do not make money *building* nuclear reactors.

They make it selling the proprietary non standard "fuel elements"

As readers may have noticed this thing does not have *any* of them.

Without an *existing* nuclear construction company "sponsoring" the design it's YANR design. Might be brilliant, might be c**p, but no one is going to put the serious amount of cash on the table to find out.

It's got some of the features of the molten salt thorium concepts and some of their issues.

Time will tell if they have deep enough pockets to get it to do anything.

Subtle & sneaky

It's kind of weird, you'd have thought Intel, with it''s *huge* investment in chip fabs, would be looking for exactly this type of approach (clever tweaks of the *existing* tools and processes it uses) to give those improvements.

Instead the start-up using what appears to be a *deeper* understanding of what is going on has achieved this leverage.

I'd guess someone has been caning their HPC hardware to run the sims that worked out the theory in detail.

Thumbs up for this, with the proviso that it has to be translated into *products* first.

So CO2 falls, -> temperature falls but ice starts at *much* higher ppm level of CO to begin with

First point supports the warming hypothesis.

Second is rather interesting.

And what *started* this cooling in the first place?

Then again does CO2 track temperature or vice versa?

In signal processing it's important to know what is signal and what is noise

I'm not sure this paper does.

On a side note how does this compare with *previous* predictions of the same authors?

About what they always said it was? Vastly bigger? Vastly smaller?

Although they do seem to have included what their assumptions were.

Something the academics of the CRU seemed to have been a bit vague on.

Kind of surprised they didn't start with the dodo.

Small(ish)

Recently extinct.

Easy to add "McNugget" as a suffix.

Actually *all* of Ian Fergusion's list sounds delicious (suitably prepared).

People who've had cat tell me it's quite gamey, like rabbit (rumored to be sold as rabbit in France).

600LY, but what is that

In Congressional election cycles.

About the *only* measure of time that the US gov is concerned with.

21500 Jeffries does not seem much.

That said it should get a fair sized HPC box.

But how much does the software cost to run on it?

It's that surprise h in the middle.

Read the headline and thought "Didn't know Craig was a big DS9 fan"

@Colin Millar

"BB would liek to remind you that "all your crats are belong to us""

Not so.

All those 'crats belong to Rupert.

How about releasing the results of drug tests, *including* the deaths.

Only the UK has a *very* big pharma friendly testing regime.

You don't have to report *all* deaths of patients on a trial. Scouts honor that you will if they are *significant*. Handy if your testing anti depressants and your side effects turn out to cause suicidally inclined people to top themselves, after all they were going to do that *anyway*.

Medical devices (like heart valves) are "grand fathered." After your 10-years-in-R&D-unobtainium-and-fairy-dust design is *just* like all those other heart valves. No need for any sort of *extensive* testing (let alone reporting).

As for "annonimised" health records, Google might have some advice to give them on that.

The other 99%

Will make various uses of the *official* allowed interfaces and produce results ranging from average to astonishing.

Some will brick it (possibly on several occasions)

Still a very cleverly engineered idea.

It seems to attack Sinclairs problem (good price but *inconsistent* quality) while retaining the keen pricing Sinclair was known for.

fat, plastic and used to high temperatures

Sounds like a humvee?

China offers to upgrade *all* UK broadband

While offering free VoIP and CCTV back services.

And will even lay it's own cable back to it's Beijing base routers

What can possibly go wrong?

People with access to private information tend to abuse that access

Only politicians *ever* seem surprised at this.

It's an abuse of *trust* which (no matter how much they hate to admit) the public have to have in the police for them to do their job.

Now suppose one of those police persons was actually more of a rapist and looking for people with a history of (say) drink offenses and were viewed as an unreliable witness, or as their attacker would view them "perfect" victims. Note I've made no comment on the gender of the attacker or the potential victim.

BTW North Wales has a fairly extensive history of sex offenses against boys in council care homes.

None of them got investigate until *very* late in the day.

21st century lithography, 21st century materials 1970s instruction set

BTW Atomic diameters are roughly 1/10 of nm. So *roughly* 10 years is foreseeable in terms of density.

The business about the fin shaped transistors having better *off* characteristics is interesting as in principle that should lower the off current by (potentially) quite a bit.

Handy when you've got an effect that's wasting power at *any* clock frequency even when the processor is doing nothing.

And on the subject of Mr (Dr?) Faggin's pessimism on cognitive computing I'd remind people of AC Clarkes observations about leaned men saying what can and cannot be done.

On the whole depressing.

Nice little earner for Intel.

Think how many companies they licensed it to.

Still not to worry. They will promise "Thinks will be better with HDCP II" and work another round of security-by-obscurity.

Until the next PhD student cracks that of course.

Note the crypto used by Sky Digital *seems* to remain invulnerable.

But as it was licensed from a company set up by one of the developers of the RSA algorithm you'd expect that.

Note. *None* of them manages better than 70%.

So a bit less than 1/3 of *all* the subscribers of the *best* ISP think they are s**t.

Let me suggest this will *never* improve *unless* people make it clear *why* they are leaving and what terms they would come back under.

If *enough* people walk a business either starts taking notice or it's out of business.

I personally would blacklist *any* ISP involved with Phorm or anything like it (TalkTalk's friendly deal with the Chinese for example).

US still has a fair number of working nuclear reactors.

I'll take a wild stab that *some* of them could be loaded with the odd fuel element with Neptunium 237, the elements removed and returned to a central processing site.for low volume re-processing to get the Pu240.

Security *should* be manageable and the extraction process is understood. It's hard to believe that given the scale of US gov nuclear research and production facilities there are *no* unused processing cells where the necessary hardware and chemistry could not be set up with the necessary containment and security.

The question is how *much* Pu240 does the US gov need. These are thermocouple generators (*possibly* moving to Stirling cycle systems), not reactors.

Just a thought.

Science should explain the *past* as well as predict the future.

And this study seems to do *exactly* that, except the *dire* consequences that *should* have happened (with the *current* models) don't.

But note that quote by the study lead author. A 2c *drop* in temperature caused a 120m drop in *ocean* levels.

Which raises 2 questions.

Is there enough ice in the *remaining* Earth's ice sheets to raise the worlds oceans by 120m if it goes the other way?

How much of the world is <120m above current sea levels?

Because if there is and a lot of land is below that level things can still get pretty bad.

A *proper* scientific theory seeks to explain *all* the observations in its field of study, and make *measurable* predictions about what future observations *should* be (if your theory has trouble doing one or other of these things, it's *not* science).

I'd like *all* sides to do more of *this* kind of science.

"kidnapping a gimp-wearing BDSM fetishist "

You make that sound like it's a *bad* thing.

It would seem rock star might consider ploughing some of their cash into upgrading the physics engine for their games framework.

Just a thought.

Extraction, Tranlate and Load

3 Little words that are often forgotten in migration projects (which *essentially* this was. I'd never heard of this IBM thingy till I read the article. It *sounds* quite impressive. Sort of mult-media and hypertext before those ideas really existed).

The answer the manager seems to have come up with in *each* of these cases was "We'll think of something."

*Heroic* effort but icon for the plan being well, stupid..

Brilliant, elegant and simple.

And as others would have said, any honey harvested is pure gravy.

I'll get my anti-bee suit on the way out.

WTF was BT Engage?

Dittoo for Basilica and Lynx.

Seriously I'm guessing they are some kind of con-sultancy con-tracting operation for Global Megacorps.

In which case they should have some high profile customers they brag about.

Such as....

"parked in handicapped spaces"

Hmm.

Did "handicapped people make handicapped faces"

"I'm an a**hole" "he's an a**hole" "what an a**hole"

That's mine with the Dennis Leary DVD in the pocket.

A tablet with a 2 week life expectancy

That's pretty good to me.

One thing Ive never understood about portables is why they don't shift the clock frequency.

Full speed for rendering and caching every possible page from *here*, then shut down the processor to a few Mhz.

How much crunching does it take to detect a) someone has touched the screen b) time to do something about it.

Always assuming you haven't put it on slow scroll so it just has to render chunks, DMA them to the screen and shut down.

coming to the UK soon

Do you *really* think the UK will buy a specific "Designed for UK" smart meter?

Of course not.

Yes they are normally supplied by outside companies (Landis and Gyre IIRC are quite prominent).

*Some* attempts have been made to check their security (reported by El Reg) and showed them pretty vulnerable (no authentication, ID #s sniffed etc. Shades of the remote adjustable insulin pump). but then the meter mfg stopped making samples available to the testers.

Here's a thought. Do the penetration testing *before* you roll out.

A few notes on renewables.

UK energy generation is roughly 56-60GW and UK power stations have historically been in the 1GW range. 20-255 of it is Nuclear.

The British Hydropower Association estimates the UK has 2GW of viable hydro power in terms of micro hydro systems (so probably eligible for renewable support). Some might freeze in the winter (but likely to be fast flowing enough not to) and likely to last as long as Earth has a viable weather system.

In the late 70s Reading U looked at the problem of tapping heat from North Sea oil wells for on board power for the oil rig. They looked at the idea of using *single* wells (so even a dry well counts) with a down well heat exchanger driving an inert fluid driving a turbine(no noxious chemical being released or needing to be re-injected). Indications were each well could drive 500-1000 Kw.

On shore a single well could probably supply 50 houses. No *big* Uk energy company will bother. However a business *could* be made which offered a package of drill well/install (and replace) generating hardware/system maintainence /billing, which might involve taking their costs off the top and any power not being used by the cluster of houses being "exported" to the national grid.

It would mean *some* home owners would loose some of their property for the well head/generating package but this could be rented from them or offset against their bills.

Anaerobic digestion is believed to be capable of supplying 1GW of UK power needs

*But* if electric vehicles take off things get a *whole* lot worse.

Molten salt reactors using thorium seem to be the only ones that can be designed *not* to depend on precision made fuel elements and to burn nuclear wast as part of their input mix, while tapping off the Xenon isotope which is the *biggest* poison whose accumulation is the main reason for fuel element re-processing. .

As others have pointed out it's virtually impossible that *one* replacement (like wind) can meet *all* power needs (and wind is a *very* poor choice if you're going for just one) so a mix of renewable/sustainable (not sure if geothermal is renewable but with a life time in the *millions* of years I'd say it's sustainable).

BTW I don't think PV's are that bad. The US with JPL as lead agency spent *lots* of money in the 70's to radically lower PV (Especially Silicon but various others as well) and I *strongly* doubt the payback period is now 14 years.

I'd like them to find out how long one of these really is good for.

Soyuz is rated at 206 days (IIRC) but *why* exactly? It's better than 6 months but why not go for the whole year?

The even in 1975 was the Apollo Soyuz Test Program (ASTP) although no other tests were done.

Note that bacteria or plants or whatever is being tweaked to make oil

are *not* evolved to make it in the first place.

Evolution selects for *survival* over the conditions that the source species have encountered over their evolution.

It's *highly* unlikely those conditions would have driven their body chemistry to make them *perfect* oil substitute producers, or anywhere *near* that level of efficiency.

Which suggests there is *lots* to play for.

However there is the issue that whatever is produced will need raw materials (including sunlight) to operate.

The state of the art

In exam crib sheets.

*amazing* materials science *if* it works

Shrunk a large *rigid* glass fluorescent tube into a flat *flexible* plate.

Eliminated Mercury from the process.

In principle then can leverage the *decades* of work on phosphors and activator chemistry and tailor the output precisely.

Actually as an *idea* it's surprising this has not been tried *decades* ago.

Of course the driver hardware needs work but that is the same with all current CF technology.

*Cautious* thumbs up on this.