Re: OK, let's catch that lie: Facebook is NOT free.
"You ARE paying for Facebook"
An alternative view is that you're not paying for Facebook. You are what Facebook is selling - to anyone who wants a slice of you.
42400 publicly visible posts • joined 16 Jun 2014
"the only data they are selling is my anonymous Facebook account data" and that of "my extended family who are spread out all over the country, and there isn't a day goes by that I don't get to see and enjoy photos and videos of my young nephews, neices and cousins as they learn to walk and talk, play in their gardens, start at their new schools, attend their proms, graduate from their Universities, and eventually post their own videos of themselves getting wasted in fancy dress."
"And If you end up unwittingly serving up a nasty trojan-bearing ad the reputational damage is all yours! Using an ad broker at least spreads the blame around a bit."
If you serve up a nasty from your own servers then you're open to actual damages so have good reason to be careful after the first time.* Irrespective of using an ad-broker you're likely to get the reputational damage anyway - the users don't see the broker.
*Or maybe a whole lot of times if your CEO happens to be a baroness and you're a bit slow moving.
"Maybe people aren't switching because, on the whole, UK banking is very good? I've only ever moved to take advantage of deals not because of bad service."
Really? More likely people aren't switching because they're uniformly crap having dead-heated in their race to the bottom. I've only ever moved because of bad service and am running out of places to go.
Given that Microsoft have more or less given up on the RT as a source of profit wouldn't it be a good idea to unlock them? It wouldn't cost them anything and would be good PR. The only problem appears to be that, AFAICS from the article, the same keys are used everywhere so they couldn't restrict this to the RT.
"I would venture that very few hackers are actually interested in taking over control of your steering or brakes."
There are few idiots who drag obstacles onto railway lines. Nevertheless they exist and are a nuisance at best and very dangerous at worst. I don't see any reason to suppose that those who'd do that wouldn't also be interested in crashing your car if it came in range.
"If we get to the stage where oiks can stand on motorway bridges, press a button on their Pi-powered gizmo, can crash a stream of cars into each other, can do it time after time at any time of their choosing, we will indeed have a problem. But we are a long way from that."
So you're saying that the time to solve a problem is after it's gone into production? Surely conventional wisdom is that you build security into the design, not try to add it on afterwards. Because by the time we have that problem it will be far too late to fix it.
"To me an on-premise system is one that sits ... in secure data centre to which your office has suitably fast connectivity"
I'd have thought that a better minimum definition of on-premise is that if the connectivity goes down you can see the digger that did it. Or the cleaner who pulled out the plug.
'they nabbed the whole "company" and have records that could cause the victims to claim restitution.'
Claiming is one thing, getting is another. I doubt any company who has lost money during the monitoring period would have consider the eventual outcome to have been worth the cost of their losses.
There's always a trade-off between gathering information and allowing harm to continue. It will always seem easier to err on the former side when it's someone else's harm; allowance should be made for that.
They monitored the operation for several months. According to another report they did attempt to contact at least some of the victims to warn them (and were themselves suspected of being scammers) but essentially the let the scam run rather than blowing the whistle PDQ. So why are they not being treated as accessories?
" Regular people are simply scared of an upgrade breaking something..."
And when those upgrades have gained a reputation of doing that, quite rightly so.
"... and having to learn new things."
Experience with smartphones and Chromebooks indicates that that's not the case.
"When they buy a new PC they will not have a choice."
That depends on the H/W vendors. If they decide that Windows has become a toxic brand they'll offer a choice.
" Windows XP = 10%...The Windows XP segment isn't going to shrink very quickly. It clearly consists of users (or applications) that are happy as they are."
As this is online usage the real XP share might be higher than that as it will include boxes that are tied to specialist H/W that never even provided W7 drivers but never go online.
"and those who are not can easily get the files back with a linux live CD."
After which they can just install from the Linux CD.
In the midst of reading these comments I had an alert pop up about 4 updates (to fontconfig stuff). They were downloaded in less than a second and released 105 K of disk.
"Why single out the UK government? Did they think it would be an easy target?
Why not try the same thing against the US government"
Because it's UK legislation that's being challenged.
Because the UK is a signatory of the ECHR where C = Convention.
Because the UK is subject to the jurisdiction of the ECHR where C = Court.
Because none of the above apply to the US.
Is it really so difficult to understand?
"it is inevitable there is a case where guilt or innocence on the available evidence comes down to a simple mathematical probability at which point a threshold has to be determined."
Not really. If you take a typical trace material investigation you're effectively saying there's a 1 in x chance of this material being found on someone being hauled in at random and asking the jury to decide if x is a big enough number. But the situation is that the defendant shouldn't have been hauled in at random, there should be other reasons to have arrested him. That should also be part of the evidence. You should have multiple lines of evidence coinciding.
What I'm not happy with (as per Bazza's fingerprint case) is someone with a match in the records being charged and that match being the only evidence submitted.
Basic scientific method is that you form a hypothesis (for some good reason, not at random) and then try test it by trying to disprove it. You look for tests which make it difficult for the confirmatory result to be obtained by chance.
If fingerprint or DNA is the most discriminatory technique at your disposal then if that's your original reason for suspicion your confirmatory evidence, by definition, isn't going to be as good. If you allow for the fact that there could be an innocent basis for the fingerprint or DNA being found (including contamination in the case of DNA) then the case is actually resting on less good evidence.
'For the life of me, I cannot understand why M$ thinks it's OK to remove control over the configuration of production machines. I guess they just assume that only Enterprise Edition customers use their computers for "real work," and the rest of us just play Candy Crush Soda Saga.'
Haven't you worked that out? Even if you only use one computer for real work you're supposed to go out and entangle yourself in enterprise-scale licensing and costs. There may be an option coming a little later for a remote desktop terminal install which just signs you in to do all your work on Azure.