Jump to content
House Price Crash Forum


  • Posts

  • Joined

  • Last visited

About ChumpusRex

  • Rank

Recent Profile Visitors

3,100 profile views
  1. There is no restriction to 1st party processing; the processing can be carried out by a 3rd party (e.g. where necessary, and where the same protection of the data can be assured). Recital 48 of the GDPR makes this explicit, by stating that sharing of data among multiple data controllers may be a legitimate interest. It is easy to see how this applied to financial institutions and CRAs. In their guidance, the ICO use this specific example of lenders sharing data with CRAs, and CRAs then sharing that data with 3rd party lenders as an example of acceptable data processing.
  2. You don't need consent to hold and process personal data. However, you must have a reason (for each individual type of processing). Consent is just one possible reason you could give for processing data. As posted earlier in the thread, other valid reasons are: contract, legal obligation, vital interests, public tasks and legitimate interests. So, as long as a credit reference agency has a valid reason in the above list, they do not need consent. Legitimate interests is a rather broad category, but basically means that you have a strong argument that the benefits (to the public or to the customer) of processing outweigh the privacy risks. In effect, consent is required only in those cases where another reason for processing does not exist. For example, if you order a product from a retailer, then they can process your personal data for the purposes of fulfilling your order, and also completing their tax returns, under the contract and legal obligations reasons. However, if they wish to send you marketing bumf, then that does not fall under any other reason, hence explicit consent would be required for processing for the purpose of marketing. In the case of credit reference, there are strong arguments that legal obligations and legitimate interests exist: Banks and other financial institutions lending money are legally obliged by the FCA to verify that any credit they offer is affordable and avoids over-indebtedness. Part of this includes checking an individual's other credit commitments, hence there is a strong argument that credit reference agencies are processing personal data as part of this legal obligation. CRA data is also potentially valuable for detecting and preventing financial fraud, as a result there is a strong argument that there is a legitimate interest in processing the data, for the purposes of crime detection and prevention. Similarly, CRA data is potentially useful for the tracing and recovery of bad debt, so there is a further argument of legitimate interest.
  3. I don't even recall ever releasing a version for download. Anyway, it's well dead. I gave up on it after I realised I had no idea what I was doing.
  4. However, not everyone gets a good experience. Friends who have developed serious conditions like MS or a stroke have found their private health insurance to be of very limited value, because their insurance would not pay for the ongoing care needed, and instead directed them to the NHS. One of the things to remember is that typical private medical insurance in the UK is that the maximum amount that they will pay for treatments is very limited. For example, if in hospital for illness or surgery, and recovery is slower than exepected, then it's common for them to only pay the hospital bills for the normal convalescence period, so that you have to pay the hospital bill, or more usually, the hospital will ship you off to the nearest NHS hospital.
  5. There are many reasons behind this, but some is made a lot more difficult due to changes in the way of working. For example, junior doctors who do a lot of medical leg work and documentation increasingly work complex rotas where every day they see a different random selection of in-patients, so they have no knowledge of individual cases. For example, this means that if there is a dangerous test result, the contact details on the request form often turn out to be someone who's doing a totally different job on that particular day, and not the person that needs contacting. Equally, it can also mean incorrect documentation as the reason for test. Then there is pressure of time. It would be nice to have time to double check things where possible. 10 years ago when I was training, if a doctor sent a request for a complicated test, like an MRI, there would be some time to check the notes before doing the scan, and double checking what questions need to be asked before scanning to ensure that the correct body part is scanned with appropriate scanner settings, etc. to give the best chance of answering the relevant question. This was fine if a department was doing 30 complex scans per day. Nowadays, it might be 200 scans with only 2x the medical staff. The first thing to go has been the double checking. The problem is that the NHS has been notoriously bad at communication. Most hospitals are still using numeric pagers as the main mode of communication, which makes it impossible to tell if a call is urgent, requires access to a telephone, and assumes that no one else will try to call the paging party and block the line. When I was a junior in one job I used to get about 100 pages per shift, about 20% of those failed because by the time I could get to a phone which was not engaged, the other phone was engaged, or the calling party had been paged to something more urgent, so someone random picks up the phone and knows nothing. There are ways to mitigate this of varying complexity and technology, but getting buy-in is difficult. It also doesn't really help accurate, time-consuming and mentally intensive work when someone is getting interrupted on average every 5 minutes.
  6. That is the standard NHS charge - the dentist doesn't get that, they collect it on behalf of the NHS. The £20.60 includes an examination, scale and polish, X-rays, sealants, photographs, minor complexity assessments and similar minor procedures as required. (i.e. the fee is paid once for as many minor assessments as are performed on that attendance). The dentist bills for that session at a pre-agreed rate to the local clinical commissioning group. The prices are locally negotiated. The dentist then puts in an invoice for what work has been performed (e.g. examination, X-rays and scale/polish), according to the terms and conditions of the contract. One of the ways that CCGs have been putting pressure on the front line is to only fund a certain number of procedures. For example, they are putting caps on the total number of procedures in each billing cycle. For example, if a dentist's CCG will only agree 4000 check-ups per year, then that's all they will pay for. For a dentist with who runs a tight ship and can see 25 cases in a day, that basically means a 3 day week, after which point they aren't getting paid.
  7. https://www.theguardian.com/society/2018/mar/10/housing-budget-817-million-unspent-astonished-mps Councils have returned £871 million pounds ear-marked to build new affordable housing. While a large underspend is always nice, in this case, lack of good quality affordable housing is a critical problem in many parts of the country, and one which many councils are very concerned about. So it seems a bit baffling why councils are cutting back on commissioning new housing when central government are showering them with cash to do so.
  8. I've been watching the bitcoin community tear itself apart for too long, and doing nothing about it. I lectured someone at work who happened to ask about bitcoins, about the fractured community, the BCH/BTC angst, the fee crisis, etc. and then completely failed to do anything about it, except for rotate into BCH at the top. I'm also now seeing bubbles everywhere - even my mother, who can barely manage e-mail is now getting adverts for ICOs, bitcoin trading, and all manner of other scams; they are even popping up at work where I restricted my web use to only directly work-related stuff. The mad volatility of the last few days has made me decide to rebalance, and only really keep play money in this market.
  9. Difficult to know. However, the marginal cost is more easily calculated. Based on the most efficient mining technology in widespread use, the energy consumption to mine 1 block (currently generating 12.5 BTC) is at minimum about 500 MWh. So, about 40 MWh per BTC at present. That's around 12 tonnes per BTC assuming 300 g/kWh, but obviously this is highly dependent on the source of electricity. Due to the historically low margins on mining, most BTC mining tends to be in areas where electricity is cheapest which tends to be regions with large amount of hydro.
  10. Everything that goes on in the channel is subject to change. However, it is possible to structure transactions such that there is no need for trust. Consider the following: Alice goes to a coffee shop, and opens a channel, containing $100. The $100 is transferred to a multisig address, requiring co-signature from both Alice and Barista. Effectively, she has put the money behind the bar as a tab. She orders a coffee. Alice signs a transaction for $5 to the coffee shop and a $95 refund, and then presents that transaction to the barista. The Barista could sign the transaction and broadcast it, closing the channel. Or keep the partially signed transaction on one side and wait and see what happens. Alice orders another coffee. She signs another transaction for $10 to the coffee shop and a $90 refund, and presents it to the barista. The barista could sign the transaction .... After the 3rd coffee and pecan slice, she signs a transaction for $20 and an $80 refund. You get the idea. The point is that at any point, the most recent transaction signed by Alice is proof of available funds. The barista merely needs to sign the transaction to claim the funds and close the channel. Additionally, only one of these staged transactions can ever be valid on the blockchain. The barista cannot sign the $5 transaction, present it to the bitcoin network, and then do the same with the $10 transaction. In this case, the barista waits until such time as he wishes to close the tab, and broadcasts only the final transaction presented by Alice. In the event that Alice makes no purchases, and the barista refuses to close the channel, then the channel has a time out, at which point, the multisig requirement expires, and the channel's funds can be recovered by a single signature from Alice.
  11. Even thought the Segwit 2x fork had been officially called off, some users decided to continue running the client and mining software, and initiate the fork anyway. The result was catastrophic failure, as the software froze before the scheduled fork block. Multiple bugs have subsequently been found in the fork activation code, as well as problems with the default configuration of the software. Considering that the fork was only cancelled at the 11th hour, and this fork was supposed to "become bitcoin", the incompetence (the main bugs were an "off by one" error in the fork activation logic, and a "variable used before assignment" flaw in the mining code) and lack of testing is telling. Meanwhile, the hard fork of BCash has resulted in a number of people failling to upgrade and running/mining the old fork, which seems to have had some life injected into it by the emergency difficulty adjustment system.
  12. Indeed. Close to 60% of hash power left mining BTC to move to BCH during the pump, leaving the BTC starved for capacity. The proximate problem with the BTC difficulty adjustment, is that it is very slow to respond (2016 blocks, 2 weeks under normal conditions, much longer under hash rate starved conditions). This is made worse because the BTC chain is running very close to capacity. Over the last 6 weeks, blockchain capacity has averaged about 91%. Due to the vagaries of mining and transaction/block propagation, absolute peak blockchain utilisation averages at about 95%. With hash power, and therefore blockchain capacity, down to about 40% of normal due to the abrupt shift in mining, the transaction backlog has grown extremely rapidly. BCH uses the same algorithm, but modified to provide a dramatic and rapid difficulty reduction under severe hash rate starved conditions. This has been widely gamed over the last couple of months: as soon as BCH becomes unprofitable to mine, everyone stops, the rapid difficulty reduction kicks in, the mining becomes super easy, everyone jumps on, and mines 2 weeks worth of blocks in 24 hours, at which point the difficulty increases makes it unprofitable, everyone leaves, and the cycle repeats. However, the most recent shift has not been because of this modification, but due to the dramatic rise in price of BCH. This phenomenon of miners switching rapidly was discovered pretty quickly among altcoin developers as auto-switching mining pools were developed. These algorithms varied, but in general, they responded rapidly with smooth rises and falls in difficulty as hash rate appeared or disappeared. BCH is switching to a moving average type system tomorrow with a new hard fork, this bases the difficulty of the next block on the average hash rate of the preceding 144 blocks (24 hours). While not a perfect solution, this is likely to be significantly less disruptive for both BCH and BTC.
  13. Tethers are intended to be a pegged and collateralised cryptocurrency. Rather than being produced by mining, they can be produced or destroyed by an issuing authority, in this case by being collateralised by USD. You pay $100 USD to the issuer, and they issue 100 tethers; and vice versa. The tethers can then be transacted in the same way as any other cryptocurrency. The intention was to make trading of cryptocurrencies easier, so a small exchange does not need to deal with cash, and instead just deals in tethers, which can be redeemed or deposited by customers. This moves a ton of AML/KYC overhead out to the issuer of the tethers. As tethers are a reasonably convenient method of entering the crypto market, it is not that surprising that the quantity outstanding has increased with the value of cryptos in general. The concern is that tethers are the worst of both fiat and crypto: central control, unlimited supply, regulatory uncertainty, no real consumer protection against loss/fraud/theft, and more. Additionally, there is a counter party risk, as you trust that the controlling entity has properly collateralised the tethers and correctly accounted for them, and will be willing to exchange them for cash. On top of that, there have been some allegations of dubious behaviour with tethers, including fractional reserve lending or somewhat dodgy accounting by some institutions. It is hard to know how much truth there is in these allegations. I'd view them as a marker of small and relatively unsophisticated consumer entry into the crypto market. However, they are anathema to crypto purists, and even for pragmatists, while they may have utility, like custodial crypto services/wallets, they bring additional risks that may not be desirable.
  14. I don't think it's that good in reality. Realistically, with 100% segwit transactions capacity would be increased by about 70%. At present, segwit transactions have approximately 12% penetration and current block size is hovering around 1.07-1.08 MB during the current major transaction backlog. Extrapolation to 100% segwit transactions would suggest a maximum block size of approx 1.5-1.7 MB. At the moment, block capacity is also wasted by miners either mining only partially full blocks (there are lots of exactly 750 kB blocks even during periods of severe backlog), or not releasing the extra block capacity provided by segwit (i.e. delivering 1.00 MB blocks which, due to segwit might be only 90% full). Interestingly, the miners most aggressively pushing for BCash or segwit 2x are those who are mining statistically the smallest blocks. I'm not quite sure of the motivation behind this - possibly it is incompetence, possibly they are pushing 2x by deliberately exacerbating blockchain congestion, but with transaction fees, now at around 3 BTC per block - some miners are throwing away close to 0.3 BTC in fees (about $2.4k) for every block mined. I think core, at the moment, are more concerned with the fallout from a hard fork. Around 3-4% of nodes are still running code which is not segwit compatible and which would have to operate in compatibility mode. This is despite the segwit code being released over a year ago, and lying dormant waiting for the triggering conditions to activate it, and the intervening code releases containing a number of other compelling upgrades. Similarly, many large users (exchanges, etc.) may run customised versions of the core code, which may require redevelopment to support a hard fork, hence the need for a sufficiently large notice period. I think an increase in the block size, beyond segwit is badly needed, and indeed, the solution used by BCash (where the maximum block size is decided by consensus) seems a reasonable enough approach, as it avoids the need for future hard forks, should even more capacity be needed. The problem is that there was vehement opposition even to activation of segwit, so I don't know how any consensus approach taking over BCash's approach would work. At any rate, the segwit 2x approach is the worst possible option - a hard fork, with no built in upgradeability, should capacity be inadequate.
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue.