Re: Unfair dismissal
He probably will have got something to keep quiet: Apple will want to keep this out of the headlines and, therefore, out of the courts and he could easily launch a discrimination suit.
12180 publicly visible posts • joined 16 Apr 2007
Account databases do get compromised, and any username and password so exposed can be easily fed to a bot that will try the combination out at popular websites, a technique known as credential stuffing.
I hope that passwords are no longer being stored unencrypted as that should now count as negligence: hashing with a salt makes reverse engineering much, much harder.
then there aren't really many useful alternatives to C.
Depending on the domain, you may find plenty of Fortran and Lisp fans out there… ;-)
PyPy does a fantastic job of optimising Python byte code and has over time achieved quite astonishing results. Unfortunately it doesn't play well with Cython, which is a great alternative for those occasional hotspots in libraries.
But the general drive to shoe-horning types in Python so that compilers can optimise should be resisted: code becomes bloated and unreadable and thus difficult to maintain. And then it turns out that YAGNI but it ticks a load of boxes, which is great for the project managers.
Where Python does need to improve is on multiple CPUs and, due to the way Python works, this is hard™ but there have been significant improvements and investments, due to the ecosystem built around Pandas and Numpy.
No, they don't care because they've grown up being trained to be "good" customers, ie. ready to buy the next load of crap.
In general, most people seem happy with the "nothing to hide, nothing to fear" argument. But, in particular, they tend to find it creepy. It's easy to demonstrate this by reading their public streams and then asking them about things that they may have thought were only being read by a few people.
Python has a history of changes, some of which are backwards incompatible: dropping support for backticks, demoting filter and reduce, etc.
While the move to Python 3 was very poorly handled, the ideas behind it were reasonable and generally approved. Code that runs on Python 2.7 will run on Python 3 with a minimum of changes, this was the reason for supporting 2.7 for so long.
But that battle is largely over now and new ones are starting, particuarly the one to force type annotations onto everything.
This is a major problem for Android and iOS.
IOS, yes, Android less so. A friend of mine with a collection of fruity devices routinely gets annoyed about having to buy newer version of apps to support newer versions of IOS or be forced to get newer versions of IOS or new devices in order to use newer versions of software. Android is, for better or for worse, far more pluralistic so it's only recently that support for Android < 6 has been dropped from apps.
I liked the S5 so much I had two of them! They ran fine with LineageOS apart from a slight problem setting the storage location for photos in more recent versions and a slight flakiness in Bluetooth when using it for navigation.
I passed them on to friends when I bought myself an s10e, which while most defintely a better phone and a worthwhile update, also just highlighted how good phones in general have become.
I think the S8 was released on the cusp of Project Treble which means, that while Samsung might be dropping support for it, it should still be receiving updates directly from the Play Store. Samsung has since then clarified and AFAIK extended support for its "flagship" products.
Yes, a lot of the ideas are half-baked. But this is why proper technical discussion with others, such as Apple, would be welcome. It goes two ways with Apple's proprietary web-kit extensions foisted on an unsuspecting world: viewports handled by meta tags, CSS support for notches, etc. This goes back to the day Apple decided it was job done with web-kit, removed developes from it and essentially withdrew from WHAT-WG.
To their credit, Google has consistently fought for the web as a platform. Yes, this is because the web is where they sell ads, but that shouldn't detract from things like webP/webM, WebRTC, etc, which have made the web a more open environment, including for companies like Zoom. Even after they essentially won the browser war they have continued to pour resources into web development. They should be taken to task (or even court) for their privacy and monopolistic practices and criticised for some of their more hare-brained suggestions, but applauded for their continued commitment to development.
Thanks for the links and additional information. Good road signs are a great advert for the importance of design. The story of Frutiger and Roissy Airport is also interesting. And Microsoft did learn some lessons from that. Unfortunately, and this happens only too often, it decided that it wouldn't limit the number of tiles. And it added animations.
Limits are important. I remember driving around the Scottish Highlands a few years ago and being not just overwhelmed but downright confused with the signs along Loch Ness: too many places in two languages.
The Microsoft Blog has some pretty, and pretty useless pictures of the new typefaces but apparently no playground for users to try them out: surely no problem with WOFF but whatever.
I've always hated German road signs and the typefac they use. They are not at all averse to mixing different sizes, but basically they're decorative and explain why, before the era of sat navs, many Germans tried to memorise the routes before setting off, or even doing trial runs.
Of the typefaces and road signs that I know I do like both the UK ones – it helps being able to say "The North", "The South" but this isn't always possible – and the Dutch ones which apparently use the typeface from the US highways, though points can be deducted for "andere richtingen" being used too often and more than a few people wondering why the never get there.
But Microsoft has rarely got it right with typefaces. Even the almighty Apple backtracked slightly from force feeding everyone the designer's favourite Neue Helvetica because, even on hi-res screens, it still takes sophisticated hinting to render nicely.
But for my own stuff I use the Le Monde Journal typeface. I remember when it was introduced for the paper after extensive testing and concurred that it was a distinct and distinctive improvement without being OTT: great at different sizes and has an extremely legible italic. YMMV
Without more details it really is just a number. Being sold to private equity could mean that there's going to be more for investment but it's more likely to be some kind of debt and equity deal: Dell is keen to pay down some of that massive debt it piled on. Boomi will then be stripped prior to being sold to a competitor in a couple of years, if the gearing is right: that will almost certainly be a cash transaction.
and probably trained on a rig consuming Megawatts
In which case you should also consider the energy used to train the meatware…
But the comparison is irrelevant. What is interesting, and why the market is so hyped, is that these kind of puzzles were often the target of the first AI systems in the 1960s, which soon proved too complicated. Some business process map quite well to these kind of challenges, which is why they're starting to be automated.
Of course, some AI, a bit like nuclear fusion will always remain ten years away…
It's just more: it's not what you know, but who you know. The polling/electioneering data slant is a great door opener as most politicians would sell their soul (if they still had one) for better polling data.
Now that we have computers powerful enough modelling based on statistical extrapolation has gone from being merely fashionable to seemingly essential and the provisos that should accompany such things ignored. The models are now so detailed that they increase the illusion of control. Even though we know forecasts don't give us control of the weather it's difficult to shake off the feeling they do.
And tell me, the UK has a serious problem with stabbing assaults and deaths,
It may have, yes, but less of a one than the US has.
An armed society is a SAFE society!
Not if you go on the data from the US it isn't. But other countries have similar rates of gun ownership and seem to get by with fewer mass shootings so I'd say the point is moot. But maybe it isn't about gun ownership than about what kind of guns and who has access to them. For people in countries where gun ownership is restricted, travelling to countries where it isn't certainly doesn't feel safe.
As far as I am aware the Covid-19 vaccines do not suppress or protect against influenza.
It certainly doesn't. My point was that AFAIK more people opted for the flu vaccine last year than usual. But, more importantly, the flu as another airborne virus was also denied many of its usual means of distribution through reduced mobility of the population. This was observable in the Southern hemisphere earlier in 2020.
Flu vaccines are known to be far less effective than COVID – I wonder whether this may change – not least due to mutability of the virus over the course of time, which is part of its modus operandus. Hence, sometimes we have mild seasons and sometimes (eg. 2017-2018) we have more severe ones, which we just tend to accept. The counter-factual test is difficult to demonstrate, but it could be conjectured that, with an ageing population, mortality might be expected to rise (until this becomes self-limiting). Whatever the reason: the flu season 2020-2021 hasn't really happened yet.
Of course, as you point out, if we don't die of the flu (this year), we will eventually die of something else. However, the main argument for vaccinations is that, like dental hygiene, they're dirt cheap compared with medical care and the potential long term effects. And then there is quality of life of the survivors.
The irony is, again of course, that because they're cheap the phama industry hasn't really been bothered with them: it was the university of Oxford that worked on the Adeno-based vaccine and the work on mRNA was initially done with an eye on developing treatments for conditions like MS.
In many countries over the summer case rates and, hence, mortality rates fell, as they are doing again. This is largely down to people being outside more where the rate of infection is much lower. But when looking at the global mortality rate, ie. of the population as a whole and not of those infected, you have to take vulnerability into account and many of those who died in the first half of 2020 would most likely have died at some point in the year – most likely in the cold and flu season in the autumn and winter. The deaths were, in a sense, brought forward by a couple of months. If this sounds callous it's not meant to be: this is just an observation of well-established data. We all die at some point and respiratory diseases are one of the main causes, especially of those with pre-existing conditions where it doesn't take much to overwhelm the body. My sympathy is with everyone who has had a bereavement as a result of the pandemic.
What we're now seeing is possibly the inverse, with the precautions (restrictions but also vaccinations) taken to reduce the spread of COVID-19 also suppressing the flu. We should at some point expect a return to the mean, though for the UK at least that might not be this year. And, unfortunately, I expect all politicians to go back on their word regarding funding for public health just as soon as it becomes politically expedient to do so: nurses and care workers will continue to be paid below average, funding for research will be cut, etc.
But it's not a total elimination of the risk.
No, that's why herd immunity is based upon three factors: the infectiousness of the virus; the efficacy of the vaccine; and the percentage of population immunised. I should add that herd immunity is, like full employment, somewhat of a misnomer: full immunity is never possible but as long as outbreaks are self-limiting without additional interventions you've effectively there. You can refine this in the case of COVID-19 where the morbidity and mortality rates are so heavily skewed to the over 70s, which is why this group has generally been prioritised for vaccination (a significant departure from the norm). As soon as they have been vaccinated, the risk posed to the population as a whole is significantly less – this is observable in the inverse correlation between vaccination and mortality rates since mass vaccination started. And, as this is the group most at risk, there is little to be gained from placing restrictions on it.
Vaccine passports are standard in Germany and there is a certain logic to them, especially in a country where most people don't have GPs with a list of their vaccinations. Pet owners are used to them for when they cross borders with their pets so I can't really see why it should be that different for people.
That said, as soon as the documents can be used to gain an edge you should expect fakes and sure enough the going rate in Germany for a fake passport with the relevant vaccination stickers is around € 150. These documents were never designed with anti-fraud measures… but give me them any time over fucking solutionist app nirvana!
Permission is required for non-essential cookies. As state cannot be preserved in a browser without cookies, the cookie that needs to be set representing your decision, must be reset every session, hence the banner for you in every session. That's not the issue: the issue is the way websites attempt to gain consent for everything with misleading "accept all" buttons.
On the continent the general rule for traffic is vehicles coming from the right have right of way. This makes sense in terms of traffic safety because it means you generally have to approach a junction and expect to have to give way. It's less good for traffic flow but it does tend to mean that everyone has a chance, which is not always the case on a roundabout where there is a main road with traffice generally going straight on. Then there are the exceptions: Bundesstraßen (A roads) have priority but you're often unlikely to notice them in cities and this can lead to endless fun! In Brussels I was in a bus that was driven into in the side by a car and the bus driver was considered to be at fault: that the person driving the car obviously wasn't looking where she was going at the time (dealing with a small child) did not matter.
The reasoning behind roundabouts is fine but it failed to take the meatware into consideration and the accident rate starts to go up once the traffic density reaches a certain point (something like 20,000 vehicles an hour or day, but don't take my word for it). Unfortunately this took several expensive and tragic beta tests: in Manchester the junction of Princess Parkway and Barlow Moor Road was remodelled many times in the 1980s and 1990s as smashes on the old roundabout increased: ditto for the roundabout at the end of the Mancunian way; this is still a roundabout but it now has traffic lights and an underpass. I assume it's similar in other places.
Microsoft's decision to stick with the name Windows 10 when pushing out "feature updates" indicates that it is moving towards Apple's policy. It's probably only due to the large enterprise customer base that it will continue to provide support for older versions.
While I can understand Apple's general approach, I have occasionally been extremely annoyed by their insistence on fixing some bugs only in new versions: a bug in Bluetooh in Lion was particularly annoying and AFAIK never fixed. There are always good reasons for not upgrading immediately to the latest MacOS as evinced by the frequency of subsequent patches.
Because compared to SpaceX's Falcon 9 rocket, the Rocket Labs Electron is a toy.
When it comes to rocket science, size doesn't really matter. In its own way Rocket Labs is every way as impressive as Space X, which has also benefitted from some juicy US government contracts.
Musk understood the marketing advantage of reusability: all those videos of rockets returning to base remind me of the "smart" bomb footage from the first Gulf war. But other innovations, such as the printed engine and other fuel sources are probably at least as important.
Musk has more money and is more interesting to the media. He has an army of "surrogates" who are more than happy to defend all his actions to all and sundry. He has good ideas and bad ideas but above all knows the importance of keeping the stock price high to guarantee dirt cheap access to capital: if that is cheap enough you can can afford to do whatever you want.
Well, as developments over the last decade have shown: in order to have an industry, you need to be involved in developing the technology. Remember Nortel, Lucent, Motorola? Huawei, ZTE and others have been producing great equipment for years – Huawei particularly should be singled out for developing low power base stations – that we nearly all use.
17 years olds drinking with their teachers has much to commend it. Less likely to get bladdered and the kids see that drinking sensibly can be fun.
I certainly remember that growing up in a culture of prohibition inculcalted a desire in my generation to get drunk. Hence, a fondness for cheap and disgusting Thunderbird, White Lightning, etc.
I also remember trying to take my younger brother into a pub with me for a breather during Christmas shopping and be refused entry – he'd have only got and wanted a soft drink. Because I'd already been to the continent a few times where in many countries it's perfectly normal to see whole families in bars, this really jarred and still echoes discrimination againts families, though I think things have improved a bit since then.
That said, I think recent research indicates that young adults are particularly sensitive to alcohol so moderation really should be observed.
Depends a lot on their background and their relationship with the teacher. IT could be pretty "cool" and this was probably their first trip in the big wide world so the heady call of the "exotic" too strong to withstand, especially after a couple of pints. We all have to learn, at some point, how shabby and disppointing these places are.
I know of quite a few companies that happily bought, issued but didn't log a lot of equipment. Note this doesn't include essential IT supplies such as the battery chiller (beer fridge); dashboard panopticon (120" borderless TV); network-phase testers (cattle prods); flame retardant sheeting (carpet rolls); flame retarder (quick lime).
At some point the beancounters are going to want to know where those keyboards, monitors, webcams, headsets, etc. went…
Not at this performace. You're kidding yourself if you think that x86-64 emulation was "expected" to match a core i5.
Which particular core i5? Intel has been struggling for years to improve performance of its chips. Ten years ago, x86 was significantly faster than ARM for single core work but since then ARM has improved a lot. Hence my 2020 MacBook Pro is not much faster than its 2016 predecessor. Over the same period Intel has only managed incremental improvements.
I'm not denigrating Apple's achievements, just trying to give them more perspective than your fanboi comments. Intel has been losing market share across the board because of its failure to produce high performance chips with low power for years. This is why Google and Amazon have designed their own chips and why nVidia has been pushing GPUs for machine learning. And even AMD has got back into the game, both in the data centre and increasingly on laptops.
One of the main reasons why Apple is a pioneer is because Google decided to cripple Android as a desktop OS in favour of ChromeOS. With Office for Android and IOS, Microsoft has demonstrated that ARM is good enough, as you'd know if you'd ever tried Samsung's DeX or similar. So well done Apple for seizing the opportunity.
As for your comments about the semi-conductor industry: Intel's CEO had to go because the process engineers have failed to keep up with TSMC and Samsung. ARM was put up for sale by a cash-strapped Softbanka and nVidia saw an opportunity to increase its market share in ML environments. Qualcomm has been wanting to get into the data centre market for years. Samsung struggled a bit with the Exynos but continues to improve in chip design and manufacturing.
It cannot be overstated how huge this might be.
It can, and you've just done it. Impressive as the work is, it was largely to be expected. Apple has had both the time and the resources to tune the M1 for emulation, something that wasn't possible with the PowerPC to Intel transition which was done entirely in software. It helps that there are more similarities between x86_64 and ARM64 than there were between PowerPC and x86, and that modern CPUs generally have cycles and RAM to spare, which wasn't the case 15 years ago. It also helps that a lot of processing, such as codecs, has already been offloaded to the GPUs, though full hardware acceleration isn't available to all: read up on restrictions for ffmpeg (used by Handbrake) on MacOS on x86 for more information. But this means that Apple can more easily intercept the relevant calls and route them through hardware acceleration. At the same time it profits not only from its own optimisations but the work already done by ARM,
It has been covered elsewhere that the M1's memory has been optimised for certain operating conditions and where this works, the results are spectacular. Where this isn't possible, performance won't seem so great.
But the work done on the compiler and the toolchain have also made it much, much easier for developers to target the new architecture and release universal binaries. This has certainly been helped by owning the toolchain making incremental changes over the years. And maybe developers have learned from the last transition that asking users to pay for universal binaries is not the best way to get on their side.
But not all that glistens is gold: look at the number of updates since release and it's obvious that there has been a lot to fix. This, and Apple's determination to build up the walls around the garden, is why I haven't installed Big Sur on any of my machines. Nevertheless, what Apple has managed to do is impressive.
From the description it sounds like it took around 30 minutes. Presumably the current limiting factors are energy input and size. Ideally, you'd also want to be able to source locally some kind of reduction agent in the hope of being able to reduce the temperature required. However, seeing as CO2 to CO is considered an essential part of our own renewables strategies, it's to be hoped that improvements here can be expected .
I assume you're talking about the web version? I always found that to be a bit of a hog. Until I came across Teams as an Electron app: has the fan whirring with any kind of video feed.
Meet, previously known as Hangouts, has always had excellent networking performance. I've also use the call out function a couple of times: useful when you're away from home and making international calls. While this has been phased out, Google has at least refunded the balance, which can't be said for all such providers.
The new version offers support for up to 250 participants. If that includes video then that means some serious network performance. As I normally disable video I'm not worried about a screen full of faces, but that's a lot of multiplexing! The inclusion of realtime subtitling and international dial in numbers for all calls will also be useful for some.