Project Jacquard
Back to punched cards.
42407 publicly visible posts • joined 16 Jun 2014
These questions have been posed in the comments of just about every GDPR article.
As regards backup, there is wording about technical feasibility (this is an off the top of the head rely so you'll have to go and check it if you want the exact words). The upshot of this would appear to be, no, if it;s not feasible you don't have to go and edit your backups. OTOH, if you were to restore one of those backups you'd need to have a means of re-doing the deletes that happened between the taking and restoring of the backup.
There's also wording about not having to remove data relevant to performance of transactions. So you'd be able to justify holding onto warranty data until the expiry of the warranty.
But as with any legislation, the ultimate way to decide what to do is to read it to find out how it affects yourself.
"consulting companies are not always practicing what they're preaching."
OTOH in such a field the you should expect to be judged by the way you run your own business. If that isn't very good why should you expect anyone else to buy your services. In fact, you're no better than all those would-be SEO specialists who write from gmail addresses and don't seem to have a domain name that should logically appear on first page in Google if one were to search for "first page in Google".
"How would you feel if low income workers who were fired for incompetence or misconduct had to give back several years of historic pay?"
They're not going to be the ones making company policy or overseeing company policy being carried out. The reason CxOs are paid a lot of money is that they carry responsibility. If they fail to discharge that properly that over a long period why shouldn't some of that money be clawed back?
"having monitoring of the conditions would be useful"
Every freezer I've had monitors itself. If the temperature rises above a limit it sounds an alarm. If a fridge or freezer isn't located where the alarm wouldn't be heard it's not beyond the wit of man to run a bit of twin core to sound the alarm somewhere where it would be. It doesn't need to be connected to the internet; that's just needless - and dangerous - shiny for the sake of shiny.
"Things have to be done RIGHT NOW because hospitals are vulnerable"
Let's look at that a little more carefully.
What has to be done right now?
I'd say the first thing that has to be done is to find out what has to be done (yes I did spend time living in Ireland ;).
The medics have a term for this, it was used in the article: triage.
Some things can be done quickly but everything can't be done at once because you always have a finite number of people to do it and things have to be done in order: if you think the network needs to be rearranged so as to isolate the more vulnerable equipment then that has to be planned, otherwise you may accidentally fail to do so by missing out on some bridge or you may separate a piece of equipment from other systems it needs to work with. Then you may need to buy more kit which has a lead time.
You can start doing things right now (Starting telling TPTB you're going to need to budget time and money is one). You will actually be able to finish doing very little right now.
Above all you need to avoid the politician's syllogism: something needs to be done, this is something therefore it must be done.
"The rest makes sense but that's just a bit far fetched."
There was an internet connected dishwasher mentioned here a few months ago: https://www.theregister.co.uk/2017/03/26/miele_joins_internetofst_hall_of_shame/ That turned out to have been intended for use by medical services so it doesn't surprise me that they would have actually found internet connected fridges. This sort of exercise should lead to questioning the wisdom of such devices.
"wonder what the lessons learnt are, that a hospital could be put immediately into practice."
Why would the lessons learned have to be put into practice immediately? Some, maybe. But the fact that some or even all may take longer doesn't invalidate the exercise.
Have you ever carried out a DR practice? If you have I'm pretty sure that at least the first time you would have learned a great deal about how to prepare for a real DR event. Given that you don't see the point of this exercise I'd guess you haven't.
"Yes, you can. I can too. But $600 so I don't have to fix it on my mum's computer when it breaks would be the best $600 I have ever spent, BAR NONE."
The point is that the workable solution you knock up on a Pi doesn't cost $600 to produce at scale. You have to ask where the mark-up's going.
The ongoing need for a means of mediating some connections is a real one. We don't need yet another slurping operation.
Perhaps the solution is to accept that such an ongoing service needs to be paid for. There could be a multi-tiered service - say a way of advertising one's current IP address (to provide for ISPs allocating addresses dynamically) for the peer-to-peer. The next tier could be a store and forward facility for passing messages to someone currently off-line and an additional facility for advertising this more permanent address. Add on a means of advertising a public key so that stored and forward data can be encrypted at rest whilst being stored. Another tier could add more encrypted storage, calendar functions etc. Is there anything in this list that seems even faintly novel except for the notion that such services are actually the product in their own right?
"Granny probably doesn't have internet at all."
How long do we have to put up with this juvenile stupidity?
Of course those of us in the70+ age group have internet. My 90-yo cousin has an iPad (and a robot lawnmower which is an actual thing these days). His wife has a Linux PC (the answer to having been successfully phished with ransomware) which also, via Calibre, acts as a server for her Kindle. Another cousin and her husband have 2 Linux PCs & a laptop on their internet connection; one of those PCs was bought from Time which, for those who know, indicates how long they've been online.
The internet has been available as a commercial proposition since the 1990s, PCs since the 1980s. Many of us have been using computers since the 60s or maybe even earlier. Do you really think the internet, computers & all sprung into existence when you got your first mobile phone (BTW I was working in the mobile industry in the mid '80s)?
The only thing such would-be witty comments show is the depth of the commentards' ignorance.
"for about the same money you can roll your own with a Pi, camera & a touchscreen monitor, but it won't look as nice"
My reaction exactly. And for the price of an HDMI cable you could hook it up to a bigger display. The knob on the side (as opposed to the knob who's selling your data) is presumably presented to the software as a mouse.
Given the number of people selling boxes to house your Pi even the appearance could be a quickly solved problem.
" their privacy policy is the first I have seen, ever, that is so one-sided it contains not a single word about protecting YOUR details and data"
OTOH "We take customers' privacy extremely seriously" is so frequently trotted out in the wake of blatant failure to do just that that it has become a de facto admission of guilt so maybe that's why they avoided it.
"Instead of thinking of them as self-driving cars, try thinking of autonomous vehicles as trains that go where you need them to go."
But not necessarily when you want one for the simple reason that you'll want one at more or less the same time as everyone else and most of the time someone else will have got them before you.
"Bike for short journeys, train for longer ones."
That works for you so it must work for everyone else. Everywhere is reasonably flat. Everywhere has good local train services ("good" because I've had experience with not good) and everyone has a good enough sense of balance and fitness to ride a bike. And, as per Charles 9's comment, everywhere has an agreeable climate.
"I don't get how car sharing reduces congestion"
The predictions we're usually given are not so much private individuals' cars being shared but commercial autonomous fleets, something Uber-like but without the drivers. According the Beeb's report (but not included in el Reg's coverage) one of TfL's reasons for pulling Uber's licence was that the Uber cars increased congestion.
"As UK government is run from London and dominated by English politicians, from the Scottish perspective it's not so much the United Kingdom as 'England et al.'"
From an English perspective Scotland is seen as making a lot of decisions in its own Assembly with its own govt. but decisions involving England are made in the UK Parliament.
I'm not convinced that, had the Scottish independence referendum been held on a UK-wide franchise,* it wouldn't have had the same outcome; the English might have voted to have rid themselves of the Scots. However being effectively thrown out wouldn't have sat well with the SNP.
Not setting up an English Assembly along with the Welsh and Scottish ones was a mistake. Arguably we might have had the Brexit nonsense although I suppose we might have had the numpty wing of the Conservative party dominating the English Assembly.
Home rule for Yorkshire, I say - and within its pre-1974 boundaries.
*Quite justifiable on the basis that it affected the UK as a unit.
"Yes, the Epidiascope. They had one at my school in the early 1970s."
The epidiascope also had a slide projection option (the dia bit).
Although most speakers had 35mm slides someone turned up with over-sized slides. No problem, we have the biiig slide projector fitting on the epidiascope. Except his slides were too small for that one. They ended up being suspended in the gate of the epidiascope with sticky tape.
He was a local speaker. The lecture room was the one his own department shared.
"I am not sure that would be fair on the prior art front. It would be impossible for the USPO to have perfect knowledge of everything ever invented."
Raise the bar on originality. Recognise that everyday programming is simply applying logical thought to everyday problems. The applicant should provide evidence that the problem has been around for a good while and acknowledged to have no existing solution - as documented in peer-reviewed journals, for instance.
Or simply ban software patents in their entirety.
A client for whom I used to do occasional work got so pissed off with the support from a package vendor* that he cracked the licensing file in order to not have to keep paying their "maintenance" charges. On the basis that I really wanted to keep my distance from that I didn't pay too much attention to how he did it - I think it was simply a matter of resetting some text every few months or updating some date related number.
*He had my sympathy. One gem was that I ended up having to edit the Informix sysindexes table to bring it into line with the actual indexes their C-ISAM S/W had created.
"I bumped into an old boss who'd bumped me."
Sometimes it works the other way round. In this case the boss was actually a client of the company I'd been working up to about 9 months previously. The conversation more or less went "Would you like to come and work for us?". The subsequent interview was more or less "Do you still want the job?". They knew how many plates I'd been spinning.
It's not that computers will become indistinguishable from humans. In too many support centres the scripted humans are already indistinguishable from computers. Turning the job over to the computer is just the next logical step and will be undetectable to the hapless customer.
I hope the reviewers ( https://www.theregister.co.uk/2017/09/18/first_annual_privacy_shield_review/ ) of the Privacy Figleaf are taking note of this decision. It indicates that the US legal system does not regard disclosure of PII as being a harm in itself but only as a possible facilitator of harm.