A complaint about Tescos

Here in Ireland we effectively have just a handful of supermarkets to choose from -- Tescos, Dunnes, SuperValu, Lidl, Aldi and maybe a couple of other over-size corner shops like Centra and Spar.  Tescos have gone absolutely all-in with their clubcard over the last few years, but the price with a clubcard is the normal equivalent price in other supermarkets - and if you don't have a clubcard it will be much more expensive.  For all other supermarket cards, you effectively have to just download the app and start using it.  With Tescos, you need to provide a whole bunch of personalised information for them to be able to track you and sell your data to other people as well (I assume linked either via phone number or email).  I also have other gripes with Tescos, namely

• Their weighing scales at the automated checkouts are so strict they seem badly misconfigured.  On several occasions I have placed a single can of diet coke on there only it to tell me it's the wrong weight.
• The local Tescos in town is downstairs, out of range of mobile phone networks, and as such, you can't use the clubcard app.   The nearest Tescos near my house closes 8pm most evenings and doesn't really have much that my local SuperValu doesn't have.  And the SuperValu is nearer and similarly priced.  Also, why is the clubcard scan buried like 4 levels deep in the app - the design is awful.  Compared to the very simple SuperValu €5 off €35 which seems great for a loyalty card.
• The meal deal is fine if you're looking for calories.  If you're on a diet, the meal deal is not great, and you can get better food elsewhere for less than the meal deal. 
• Tescos has very little unique stuff that's worth a special visit and generally the quality is higher in either Dunnes or SuperValu, and either similar price (with clubcard) or more expensive in Tescos.  For me, the only things that I go into Tescos for are:  Tinned new potatoes (they only seem to have these in stock roughly 20% of the time anyway), Toffee cheddar (only available christmas time and wasn't available at all last year), Flat gammon steaks (which I have like once a year).  Puff pastry (for some reason my local SuperValu never seems to have any)

 The thing is, if Tescos is not your nearest supermarket and you don't currently have a clubcard for whatever reason, I can't see why people would start using Tescos instead.  Without a clubcard, everything is outrageously priced, like M&S prices for slightly better than Lidl quality.  So why sign up in the first place - just go to a different supermarket.

 

 

 

 

 The Rest is History Drinking Game

  

Dominic mentions Wolverhampton

Tom mentions cricket

Dominic says "Blimey"!

Tom says "sacral"

Dominic says "poor form" or "disappointing behaviour" 

Tom describes someone as a "lad" (a massive lad, etc.)

 

 

 The Simplest Diet

 Assign a value of -1, 0 or 1 to every meal.  Add these up for the day.  Add one for alcohol in the evening and one for a snack in the evening.  Subtract one for significant exercise.  Aim to get zero every day.

 

On Intelligent Life

There's constant search for extra-terrestrial intelligence but could intelligent life have developed on earth in the past?

First let's define intelligent life - dolphins are intelligent, as are ravens, as are octopuses, as are wolves.  There are two things that make homo sapiens seemingly unique amongst intelligent species of today - tool usage and communication.  Whilst other species have tool-using abilities (ravens, octopuses, the great apes) and rich communication (dolphins and other cetaceans), no other current species combines these two characteristics, and it's allowed humans to dominate the planet.  It's unlikely that these two characteristics haven't occurred together in the same species at some point in the earth's evolutionary past.

So, could the kind of intelligence that we're looking for in the SETI programme have existed in the past history of earth?

We need to look at the artifacts left by our civilisation - if humans disappeared tomorrow, what artifacts of our culture would remain for future species to discover?   The world without us covers this in significant depth.  The plastics we've created should remain as a thin but detectable layer of strata though it's probable that a microbe will evolve to eat most of the synthetic polymers we produce.  The other significant relic of human civilisation is that from the nuclear industry - nothing in nature concentrates isotopes the way we do and these will leave a significant signature that should be relatively easy to detect - indeed the discovery of the natural nuclear reactor at Oklo was initially thought perhaps to be the product of an ancient culture.  The satellites placed in high earth orbit will probably be there until the sun expands to a red giant, but are extremely hard to detect.

Our culture also produces objects of high durability yet are likely difficult to detect over geologic time - glass and ceramics will eventually be ground to a powder or eroded; stainless steel cutlery and bronze sculpture lasts an incredible amount of time yet will eventually become an unusual deposit in rock strata; Gigantic engineering achievements such as the Suez canal or the Hoover dam would silt up and be gone within a few hundred years.

It's interesting that if we were looking for a civilisation equal to our own in the distant geologic past, we would not be looking for evidence of cities (eroded to unrecognisability in a few millennia), agriculture (reclaimed by native species within decades) or even the spike in atmospheric carbon dioxide (could be due to extremely large volcanic eruptions instead).   Instead, we'd be looking for buried bronze sculpture, rubbish dumps, and radioactive waste.

On balance, it's unlikely that a society as advanced as our own has existed on earth before - though the discovery of inhuman bronzes, or unpleasant waters would certainly change that.

Dark matter and an electron-poor universe

This started out as a fantastic idea but quickly I realised that it just didn't stack up!  But as a physics idiot, I'll try to demonstrate here how to first set up and then destroy your own theories instead of going down the pathological science route!

Hypothesis

The universe as a whole contains far more nuclei than there are electrons.

What it explains

Dark matter - without electrons the only interactions between nuclei is via the repulsive electromagnetic field between the postitively charged nuclei and should they be travelling fast enough, the strong force.  The lack of electrons means no spectral absorption lines, no bonding means little to no clumping together of matter, and nuclei are massive enough to avoid being "hot" dark matter like neutrinos.

A bit more detail/Research needed

When Big Bang recombination occurred, the universe went from opaque - where a free electron would capture any passing photon - to transparent - where the universe had cooled enough to bind electrons to nuclei.  If there were a large excess of electrons in comparison to the number of nuclei the universe would not become transparent (it would be "foggy" whereas we see the opposite).  If the number of nuclei was the same as, or was in excess of the number of electrons around then we get transparency.  Existing theory suggests that the universe is electrically neutral [need cites] - i.e. the number of nuclei matches the number of required electrons.  [need to check if there's evidence]

The problems

Why does the local area have an excess of electrons?  (possible answer: It's related to the fact that the local area is a supernova remnant).  (Maths/reading required).

Ramifications

Early stars composed of primordial BBN materials will be electron deficient.  Whilst this doesn't affect fusion, it will affect electron degeneracy pressure at the core of stars - thus no helium flashes/carbon detonation due to the lack of free electrons.

Testable Predictions

In areas where BBN primordial abundances are observed, the amount of lithium and beryllium should be lower than the theory predicts.  This isn't because less lithium is actually produced by BBN, it's because one of lithium's electrons sits in the second shell requiring a lower temperature for the electron to bind, making it more likely that H or He would capture that electron beforehand leaving a bare lithium atom that doesn't show it's absorption lines in the CMB.

I don't think we can test the overall positive electrical charge of a galaxy (this theory suggests that anything with a lot of electron-less nuclei around then there would be a large net-positive charge) - there's not enough for it to interact with?

Decay Channels

I started looking in detail at lots of decay channels - but it's actually very simple when you consider conservation of charge.  Essentially, every proton converted to a neutron (e.g. in diproton He2 decay to deuterium H2 but other decay channels as well) would emit a positron to conserve charge - this positron would then annhilate with an electron producing gamma ray energies.  Very generally, for every neutron you get around here (i.e. earth) then a positron was created in the process that then annhilated with an electron.

Follow Up (Hypothesis goes bang)

Somebody has actually come up with exactly the same theory too but I can't see them addressing the "why is it not like that round here" problem.


This paper points to the universe being electrically neutral during BBN, and in addition to this, we would see different results in the very early universe regarding expansion - it'd be influenced more by electromagnetism rather than gravity.  We'd also probably see large scale electromagnetic effects from such a postively charged proton cloud that would wash-out any gravitational effects - and these would probably be observable.

And finally, the most important question "why is it not like that round here" has been utterly trounced - in large supernovae that would produce the elements that we see on earth, we would actually expect vast numbers of electrons to be destroyed in the process - one electron destroyed for every neutron created.  Instead, we have an excess of electrons around here.  This does lead to a further question - given that large stars and particularly type II supernovae will destroy a vast number of electrons by annhilating them with positrons - and we live in an area where a type II supernova occurred (because we have lots of elements > He around here!) - why do we still have charge balance?

SSL is broken

There's a magic padlock icon that appears in your browser indicating that you're secure - and that nobody in the middle can read the traffic - and it's probably broken.

Certificate Authorities

The problem is not that the encryption scheme is broken - the public/private keys structure is fine and has been demonstrated to be secure, it's that there are way too many certificate authorities and a single mistake or deliberate outside party interference (for example, from governments) can allow a man in the middle to decrypt all traffic, read what's being sent to and fro.

There's a lot of certificate authorities - most are telecoms companies or related to a national government in some way, and all of these can issue certificates for any website.  In addition, it's possible for a certificate authority to issue a wildcard intermediate certificate to an organisation that does exactly the same.

The way an SSL certificate is validated is that when your browser contacts a secure site, the site returns their certificate and a chain up to a valid root certificate.  As soon as the browser finds a root certificate that it already knows about then it assumes it's all fine and the connection is secure.
The problem is that if one of these certificates is compromised, or abused in some way by the company owning it, a man in the middle can read all of the traffic in between.  This is not a hypothetical situation - it's happened already - with Turktrust, with Nokia, and with DigiNotar.  What's worrying is there are lots of certificate authorities and it only takes one of them to be incompetent to render traffic insecure.
 
There's also other instances of https traffic being decrypted with varying levels of validity - company firewalls occasionally do it by using their own CA authority (which requires modification of each client computer), and anti-virus software with parental controls can also do this.  Needless to say I believe these should simply not be allowed.  By installing parental controls (on some anti-virus systems) you are effectively giving permission to your anti-virus company to view your bank details, and I don't think most people would be happy with that.

Just as concerning are government security agencies.  Whilst the examples above are the result of incompetence, security agencies could go to the certificate authorities directly and request a wildcard certificate - which would probably be granted.  But this means that the security agency could happily decrypt all traffic and nobody would be able to detect that they are doing so.

When a root CA is found to be compromised in some way, revoking it is a deeply painful process that can take months whilst each browsers list of root certificates is replaced. Even worse are embedded systems which may never have their CA list refreshed.

Whilst most companies applications won't change, I would recommend that all banking and financial transaction apps use some man-in-the-middle prevention - namely, EKE encryption to detect that this is taking place and prevent data being transferred.  Whilst some banks do this already, Natwest does not and it really should!

Quantum Computing

A little further ahead, we have quantum computing, especially with the new we-think-it's-quantum-but-we're-not-really-sure D-wave systems.  Using Shor's algorithm and a sufficiently powerful quantum computer, all root certificates could be compromised (again, probably by national security agencies) and again, it would be extremely difficult to detect that a man in the middle attack was being perpetrated.  There's something the certificate authorities could do now to combat this, and that's use an algorithm that can't be solved using Shor's algorithm - i.e. not prime number factorization or the discrete logarithm problem - there's other ways but these don't seem to have any take up right now by certificate authorities.

On Minecraft

Sound in games is very difficult to get right - it's a bit like make-up - a little makes everything a lot better, a lot makes things worse. And having none at all is better than too much.

Minecraft gets it right - the sound is functional and pretty spooky at times. When there's a zombie hanging around outside your door or down a passageway that you can't see, the sound of it is incredibly atmospheric, as well as being informative. What you don't get (and would be added in a big budget software house) is the sound of birds, crickets and various paraphenalia that you simply don't need. In some ways the sounds of minecraft reminds me of that from a very old game, Dungeon Master - functional, informative, and downright terrifying at times.

Minecraft also gets music right. Instead of being constantly on, the music fades in and out - there will be long periods with no music at all, but when it does arrive, it's incidental and adds to the experience rather than distracts from it. A bigger budget game would have swamped the music channel with crap that 95% of people turn off. (Seriously, game designers: If I want to listen to music whilst playing a game, *I* will choose what it is, not you). The one proper commercial game that I can think of that does music right is the GTA series - again, it's not always on - it's incidental and doesn't distract.