Posts by TechnicalVault 33 publicly visible posts • joined 8 Apr 2019
Unsurprising, as it is instability that is not in the Chinese economy's interest. They already have a problem with a property bubble that they've had to address. If this Ukrainian war then shifts both Russia and the Western world into recession then that means suddenly China's customers aren't buying. It didn't help that this was all started during China's Olympics so kinda insulting to them deliberately stealing their moment on the world stage.
Tbh that comment is so ill researched you should be ashamed you call yourself Moist, he'd be smarter about it.
NHS waiting times are long for some procedures yes, however like the US money works, if I want to jump the queue I do the same as any moderately wealthy Yank would do, I go private. Most people who want private medical care neither want nor need to fly to the US for this, it's overpriced and the standard isn't much better than UK private. Most of the people who do go to the US for treatment cancer sufferers reaching out for some treatment where the manufacturer is massively overcharging and US insurance wouldn't cover it unless you get a Platinum plan. If it's something that the UK private hospitals don't provide then you can easily fly an hour or so. Somewhere in Europe will have it, eg you can get advanced cancer treatments like anti proton beam therapy privately in Spain.
Instead I just phone up my insurance company and go to my local private hospital. Insurance company you say? Yes you can get health insurance in the UK but because UK insurance companies aren't on the hook for ER and indigent care thus it is massively cheaper. If I really want it cheaper than the UK I fly to Eastern Europe where if I am careful about my choice of hospital I can get medical procedures like dentistry or cosmetic surgery for a few thousand quid.
The difference is that if I can't afford medical treatment and get hit by a bus in the UK, I don't wake up in the hospital with a few hundred thousands of dollars of medical bills because the ambulance took me somewhere out of network. The result of medical bankruptcies and indigent treatment costs in the US is socialised medicine, you're paying for someone else's treatment, you're just in denial about it.
The proper punishment for defiance of a court order like this one, is not just a fine but jail time for executives as a sanction for contempt of court. It might seem harsh but it's not punishment for the action itself but for defiance of the authority of the court.
Problem is those AGRs have to go. Nuclear power stations have a design lifespan and whilst it can be extended there are limits. Hinkley Point B for example could have been extended a bit but ultimately the corrosion and graphite bricks warping in the core caused by a lifetime of ionising radiation was always going to be limiting. These are 2nd gen reactors, we learnt from them, it's time to move on.
France's Flamanville 3 was the prototype for the 3rd gen EPR design, expensive but informative. Hinkley Point C has gone better, especially the 2nd reactor but it's still basically a 2nd gen prototype. Sizewell C will be even easier to build but it's the last one for which the build is near certain. Whether they go on to build more depends whether politicians hold their nerve. These are long term projects that take longer than the lifetime of a parliament to reap the benefits from... Basically if we want this to happen there needs to be one hell of a grassroots pro-nuke lobby going on.
One alternative is to keep burning gas, which given peak oil in the North Sea was 2001 means either relying on unreliable partners, imports or a move back to "town gas". The UK lacks the storage the Germans do to stabilise this market, unless that is rebuilt I suspect it'll be a while until that market stabilises.
Another is to put a metric butt ton of investment into pumped storage. UK doesn't quite have the geography for this that Norway does so it'll cost a lot and take time to build. It's also energy agnostic, it can smooth out demand from anything.
Lastly you can insulate more. British buildings are still pretty useless in terms of dumping energy from the inside to out, especially all that dodgy 80s and 90s double glazing which past their service life and pretty useless to boot. Again, energy agnostic and saves money whatever you choose for generation. Previous attempts at this have flopped because it's not as fancy as a new infrastructure project, but it damn well works.
Frankly all the good options right now require mega project style investment, the do nothing option ain't gonna fly.
The UK's Copyright, Designs and Patents Act 1988 section 50B [legislation.gov.uk] which allows reverse engineering for interoperability and explicitly says: "Where an act is permitted under this section, it is irrelevant whether or not there exists any term or condition in an agreement which purports to prohibit or restrict the act (such terms being, by virtue of section 296A, void)", is going to likely make this an interestingly expensive case.
The Orion capsule has done a full high apogee re-entry test (https://en.wikipedia.org/wiki/Exploration_Flight_Test-1) where it was sent up to a high apogee in space then went through all the forces that you'd experience coming back from the moon. The SRBs and RS=25s are hardly new tech, we've got reams of data on them both in flight and from ground tests. The bit where we lack data is this all up integrated test, that's what this mission is.
We have Starship in-atmosphere tests but nothing reaching space yet and so we don't have much data on how the craft will perform in space. The all up test for this one (https://en.wikipedia.org/wiki/SpaceX_Starship_orbital_test_flight) is scheduled later on this year. I'm optimistic but space is a challenging environment so it's all about gathering that data.
Musk is not unhinged, rather his greatest strength has become a weakness. Like Jobs before him, he has a tremendous amount of ability and self belief to go with it. Like Jobs he is a bit of a micromanager, trying to achieve perfection.
Unfortunately in this modern corporate world these same traits will be reinforced ad infinitum by the people around him like Shotwell, covering for him and "protecting him" from getting real feedback. The more unchallengeable he becomes the worse it'll get, until it really bites him in the ass. The best thing he could do is to do what similarly the unchallengable kings of history did and get himself a jester, someone with permission to tell him he's being an ass without any fear of being fired.
I've heard of people trying to do the same to people who voluntarily resign. Just a quick reminder that doing this is pointless and stupid. If someone is quitting they will obviously know before you do and if you have a regular habit of escorting quitters out the same way you kick out fired people then you can guarantee they already know how you will behave. Save yourself some organisational memory and don't burn that bridge. If someone tells you they are quitting then lock them out on their last day not before.
"The report also mentions that in the last three years, NASA users have made over 12,000 requests for elevated privileges – just the sort of thing that could lead to more information reaching the wrong eyes."
Or, it could be because they've locked everything down so inappropriately tight that people cannot do their jobs without requesting elevated privs. If these kind of broad brush metrics are high, it doesn't mean users shouldn't be asking for these privs it means you designed your system badly. A good security system is inobtrusive to normal users fulfilling their normal job role, and should only become a regular feature of people's jobs when they start to deviate from that or to an attacker.
Also most elevated privs requests shouldn't be decided by IT, usually it's managers of the area or of the data that should be deciding (and if you've done it right actioning) whether or not an exception to policy should be made.
It's the same as publicly announcing to a small charity you were going to give them £2000 a year and then turning around and only transferring £1000 one year without telling them you'd changed your mind. The charity is going to make their budget based on having £1000 more than they had and the deficit is going to come as a surprise. This means you've caused them to then make promises they now can't keep. Likely that deficit is going to come out of grants they cannot now make, and which if you had told the charity in advance, they would not have promised.
Yeah it's your money, but there's right and a wrong way of going about things once you've made a commitment. The right thing to do is to be as public about changes to your donation as you were about the initial donation. All it will cost you is kudos but at least you can still have a good conscience.
We faced similar challenges at the Sanger Institute same data, but much larger scale. We had to implement a few policies to make people behave:
- No legacy BAM files for archiving. Same data 100GB BAM or 30GB CRAM, lossless compression, easy choice.
- All sequence data is archived to ENA/EGA public archiving at the point of creation, policy is added later (it's not visible until that is added). Funder requirement that happens to ensure there's a separate copy with another organisation that is unlikely to be deleted.
- No going off on your own with a cloud service. Clouds usually give you an unlimited spend allowance, and make it hard to watch how you spend, we have a limited budget. We have a cloud budget, you put in a light weight budget proposal before you go dancing off. You blow the budget? You're going to be explaining it to the rest of the scientific faculty not us.
- Models built to show the costs over time of various storage options before we make big decisions. It's amazing how many people miss the lil gotchas, like "big fee for taking more than 10% of the data out of storage".
Yep, a fair number of the people with these domains are decision makers myself included. For me this is a final nail in the coffin, personally I am pretty much likely to banish any chance of GSuite ever being a supported thing in my workplace. They make good software but they seem to have the attention span of a kitten. Time and again they have taken a perfectly good stable service and depreciated it because seemingly it's not exciting enough for them. I need a nice boring service provider who is not going to force me to do a massive migration for thousands of users because they've decided they wanted to change tack again.
I'm not sure I buy the it's too hard/expensive argument. Not when Cambridge University has been covering large portions of Cambridge town centre with Eduroam WiFi for years. Incidentally, the Wellcome Genome Campus not far down the road from Duxford is has some excellent demonstrations of how easy it is to do outdoor Wifi where it's needed. It's just a matter of having the right antenna and base station.
Yes this all requires moderately sophisticated Wifi kit but these days that is off the shelf enterprise stuff and most other government departments already have procurement contracts in place for this. There is even GovRoam https://www.jisc.ac.uk/govroam to help you provide your unclassified wifi to staff. If uni's can do it for the network mess that is students then squadies should be easy.
The legacy content industry is too tied up in petty IP conflicts and repetitive formulas to thrive and innovate. Look what has happened to the History Channel, now all about how aliens built the pyramids and then sank Atlantis.
This is where YouTube (and TikTok) has thrived, where else can you find:
- blokes repairing Apollo AGCs
- teaching you how to build log cabins
- numerous makeup tutorials
- guitar and magic lessons
- wilderness shows that aren’t OH NOES a beer every 5 minutes
They filled a hole they aren’t going away:
One advantage of bigger debris is it's a lot easier to observe and thus predict (plus there tends to be a lot less of it). Given the laws of physics haven't changed lately, those estimates should be pretty good.
The one detail that might complicate estimates is that larger debris in LEO has more atmospheric drag and this means it's orbit will change (decay) a lot quicker than something smaller. I suspect that's probably only significant over a timescale of months rather than hours though.
Unfortunately you are wrong, someone might use them for marketing in aggregate (how much demand is there for this drug) but they really are useful to researchers.
I have already worked with EHR in scientific research and they are a treasure trove because they allow us to reduce the amount of confounding variables people don't know to tell us about when we're researching. You can recruit patients with rare diseases without knowing who they are by sending a message to their GP asking them to pass on the invite. You can perform longitudinal studies of huge groups of patients without the recruitment ascertainment bias you get from regular studies. Basically don't knock it till you've tried it.
What those doctors were doing in Italy was standard triage, when resources are limited, prioritise the patient most likely to survive. Healthy 30 year old gets the ventilator over the 20 year old with CF lung scarring because one is much more likely to survive than the other. Given the limited time to make the decision and lack of resources to carry out deep health assessments the best option is to use existing data. It's brutal but given you implicitly condemned it, I would like you to present a fairer alternative?
A big problem here is the software foolishly exposing the 97% number in the UI. Your average layperson does not understand that does not necessarily mean that it is a match. In this case it probably means your training data is woefully deficient in non-matching black people, so your algorithm has learnt the wrong thing. Honestly they should not be selling this kind of thing with this naïve a UI.
What software aimed at minimum wage staff should be doing is telling the staff what they should do next (customisable per company). It should be saying something like: "I think this is this person, please manually compare the images and check the person's ID to confirm the match".
Looking at the facts of this case, it looks like if they had done it on a true project basis it seems it would be fine. A real project based contract would specify the project, a delivery time and a fee scheme based on the job not the hours worked. The building industry does this all the time.
That way if you get it done in less hours than estimated it is none of their business, they contracted you to provide a service, you provided it. Instead they are specifying how many hours a week he should work, that's employee territory. If you want to work that way it is fine, it is not that hard to employ someone on a temporary basis, retail does it all the time.
When you write an IT services contract you write it for what you are doing now and what you can foresee. Once the contract is signed your contractor is a fool if they do anything other than what is broadly stated on the contract, otherwise they go broke.
This works okay if your business is likely to carry on doing the same thing. Unfortunately for the administrators of universities their means of research production is academics, whose requirements change with the field they are studying. I am not convinced they can write a flexible enough contract to deal with that.
There is likely to pretty much no time when that building doesn't have at least a skeleton crew running the place. Tube runs 24 hours on 5 underground lines on Friday and Saturday and even when the tube is shut down there will be a team in coordinating engineering work out of there. It's not like they're going to leave the place unguarded either, there's always going to be security bods hanging around.
The problem is executive directors are also employees with management responsibilities and thus cannot be as easily dismissed as a non-executive director. Employment law applies and that means any dismissal will take time and negotiation. I suspect that long term their positions will be untenable as it is clear from the vote that they have lost the confidence of members, but it will take a while to unwind matters.
It's not that they're buying IT services from other people, they've had that outsourced for years. It's that they are doing it badly. Whilst IT might not be a core competency of the NHS, Information is and must remain so.
There is a reason why practically every hospital has a different patient records database and the links to GPs are poor or in some cases non-existent. They need the people who understand the IT + the information flows to be in the position to make the decisions rather than the bean counters if they want to achieve "transformation". Until they do they are going to keep failing and blowing money.
If I had to bet I'd say it's mostly engineers (and in previous years sales staff) who were travelling to customer sites. In the end it will mostly be just a credit shift, the employees will be effectively giving Fujitsu a short term interest free loan for the fuel until they claim the expenses. Most of those staff do regular expense claims already so I suspect they are quite likely to find that savings are nowhere near what was predicted.
The problem is the bean counting way does not accurately measure costs and risks. Regardless of whether the risk materialises if you are exposing the business to a risk that will cost £100m at 20% likelihood to save £10m then you are an idiot, especially if you do not recognise that the risk is not really 20% but a rising curve. Often you get lucky, but eventually in the words of Susan Ivanova: "No boom today. Boom tomorrow. There's always a boom tomorrow."
Business needs to get better at weeding managers that are gamblers out.
At my place we are running 2 OpenStack installations and they are quite happy little private clouds. If I had one thing to say about it, it is that we were early adopters but Openstack is a lot more mature than when we first started. There are still a few more features we would like but it has finally reached the point where it can run happily without choking everytime RabbitMQ dies.
You can go it alone and try and put something together directly from Openstack but you are probably better off working with someone who knows where all the bumps in the road are. We have just switched Openstack distributions from Redhat to StackHPC on our primary and so far so good.
If you are spending £900m on it and it is pretty much your "means of production", IT is a core competency. You cannot afford to palm it off to some third party as if it is workaday stuff like desktop support or cleaning. It's not like you're going to be using that much off the shelf software because having a monopoly on tax collection is pretty much the definition of a national tax agency. HMRC is a software company, they should embrace that.
If they are burning employees out like this then they need to pay for the damage done, it is the only way to achieve desired corporate behaviour. A nice little tax on net profits (gross is too easy to do Hollywood accounting on) proportional to the number of current employees below SNAP level (prorated by the number of hours they do, no cliff edges) + long term unemployed after working there +disability claims in the 1st 6 months after leaving employment there.
Actually it's a breach of the data protection act/GDPR for them to knowingly hold incorrect information too.
"(d) accurate and, where necessary, kept up to date; every reasonable step must be taken to ensure that personal data that are inaccurate, having regard to the purposes for which they are processed, are erased or rectified without delay (‘accuracy’);"
Also given that it is your personal data (the mobile number) that they are processing without either consent or in fulfilment of a contract it is illegal for them to process that data too.
Your best bet to get this things resolved is to write an email reminding them they are processing your data illegally to dataprotection@morrisonsplc.co.uk
The brutal truth of outsourcing is that you will get exactly what you specified in the contract and nothing more. Everything else carries a fee and unless you have robust discipline and have planned really well (which few organisations do, especially in the first year) you will frequently find yourself dipping into contingency money just to carry on normal operations.
The reason is that same as why when requiring people to keep timesheets you often experience productivity drops; there were all those little extras people were doing. Someone who was staying on that extra half hour suddenly doesn't when watching the clock. Sadly this is a lesson which is mostly taught by experience.
Firstly you've let it out for a commercial purpose and you're in breach of your contract with AirBnB, and that's just some of your civil law problems.
For criminal law south of the Border it's a bit more murky but there are some new laws on Voyeurism, as well as the data protection act which actually comes with some really fierce fines since the passage of GDPR. Up until recently it was hard to even prosecute someone for putting a hiding cam in a public toilet.
If you pull the same stunt north of the border however it's much clearer, the offense is voyeurism. If you film anyone in a state of undress without their knowledge in a place where they would have a reasonable expectation of privacy (it doesn't matter who owns it) you commit an offense. If you think about it, the law being indifferent to ownership makes sense otherwise for example a stepdad could secretly film his stepdaughter in the bathroom without repercussions.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2023
