Half-Man Half-Biscuit
P, P, B McG, C, D, G.
Is it Dusty Miller? Dr Mopp? Pugh?
Mine's the 2017 "Trumpton Riots" jacket, the one with the mop of orange hair on it…
12180 publicly visible posts • joined 16 Apr 2007
I suspect they're just hitting problems with carrier NAT. At some point even that doesn't help any more. If it was only 700 million devices they have to worry about then NAT might suffice but man + dog usually has more than one IP device and then they're all the embedded devices to worry about. China could reasonably claim to want to use all the possible addresses in IPv4. As could India.
In open source development half a million is full time company budget. In MS environment it's consulting fee for how to do things.
Are open source developers that much cheaper than other ones? That would be news to me.
Governments of all types have an unenviable record when it comes to commissioning software. I don't see why this would have been any different.
I think this might be the case in France where IT is centrally coordinated. But won't work in Germany where it's each council for itself.
There was work done back in 2003 to justify the initial switch and there was the hope that "if you install Linux the developers will come". Didn't happen so time to pull the plug. Now, if they'd started by simply moving to OpenOffice and sponsoring that and other projects things might have worked out differently. MS doesn't make a lot out of selling the OS, it's Office that makes the money.
Treesaregreen and her merry little band of fascists (ruling by the stick and intimidation) will probably have to watch this get referred to the ECJ whose judgment they will then have to respect and possibly even act into law before they can think about repealing it… Fuckwits.
Excel is good for spreadsheets and the 2016 version has some big performance improvements.
However, it's increasingly being used for reporting only with analysts switching to Jupyter + Pandas + Matplotlib for analysis and putting the results in an Excel file for further consumption. As the maintainer of an Excel library I'm proud to be contributing to this development.
Schrems original motivation was as I explained - permissionless tracking.
Maybe, but I don't see the two things being as legally equivalent. But IANAL. This close to GDPR day they'd pretty stupid if they were collecting "personally identifiable data" without permission.
I think Google's collecting the data for other reasons like footfall analysis and possibly even network planning. I noticed recently that it is starting to say how long people spend in certain shops.
The Quartz article makes it clear that Since the beginning of 2017, Android phones have been collecting the addresses of nearby cellular towers—even when location services are disabled—and sending that data back to Google.
The comparison with Schrems' case against Facebook over data protection isn't valid. Doesn't mean the procedure isn't suspicious just that it's not about data protection.
I'm not sure the UK has particularly more CCTV cameras than anywhere else.
Compared with other European countries it certainly does. CCTV schemes are always sold as crime prevention schemes because this is popular. They are at best aids to crime solving. But usually they serve as excuses for lower head counts in police forces and invasions of privacy.
Proving the poster above you's point exactly. You understand Python and think it's excusable, you don't understand JavaScript so dismiss it.
What do I think is excusable? Using a variable that is not in scope will raise a NameError
and not a SyntaxError
. If you find yourself wanting to use global
in Python you're most probably doing something wrong but the language treats you like an adult and doesn't try and guess what you might mean.
Python's scoping combined with it's mutable types can cause problems in class variables.
I completely agree. I quite like Python, but don't understand why it doesn't insist that variables are explicitly declared.
Because it has strong and dynamic typing. This was a design decision related to Python's target audience of non-programmers. As a consequence there is a strong preference for descriptive variable names. Static typing is still considered relevant only for compiler optimisations.
And I read "Markup Language" every time I see ML. It's usually best to avoid these abbreviations wherever there is a chance for ambiguity. But journalism 101 says in any case that you should define the abbreviation at the first use in a text. Sub-ed should have picked this up.
Automatically creating a local variable when I have forgotten to specify that as 'global' or have typed a variable name wrongly has driven me nuts using Python.
Sounds like your problems are more than with Python. global
does exist but you will never need it. Like exec
and eval
it's there for a reason but the overwhelming majority of Python code will never need it.
Whitespace or braces is personal preference but in team work whitespace is much easier to enforce.
At a programming level Python can be criticised for dynamic typing, memory use, the global interpreter lock, speed but for any of the random inconsistencies that Javascript throws up.
There's nothing terrible in JavaScript, nothing at all.
There have always been problems with JavaScript. Hardly surprising for something that was cooked up as quickly as it was. More surprising perhaps, is how far it has managed to get despite these limitations.
But the writing for ECMAScript is on the wall: Web Assembly. Most core JS developers, those that work on the language, seem to agree that they've come about as far as they can with it without a major rewrite and if you're going to rewrite a programming language…
But where does that leave JavaScript? Like Python, it been at the forefront of web development for about 20 years
For most of that time they've been doing very different things. And, while I'm a Python fan, I don't think it would be correct to claim Python has been at the forefront of web development at all.
Python is more suited to machine learning than Javascript because of the eco-system and libraries that have grown up in the scientific community over the years. Python isn't great for machine learning because there are TensorFlow bindings but because there is NumPy, SciPy, Pandas, etc. Python's syntax is also very popular with the non-developer types, who've suddenly found themselves being classed as "data scientists". This combined with the extensive libraries of a general purpose language makes Python eminently suitable. Node.js simply can't compete in this area and why should it? It's great for some stuff though its existence is now threatened by the agreement of the major browser developers to support web assembly. An article on this would have been far more interesting than the collection of straw men in this article.
Real incomes have declined significantly over the last 10 years. Experience shows that people will spend (sometimes money don't have) on things that give them immediate pleasure, if they're required. A new phone doesn't do that anything like as much as it did, hence the money gets spent on something else: music and tv subscriptions spring to mind as worthy replacements.
Google have nothing apart from your location...
Sure, but I was referring solely to the address of the cell masts. Anything else is GDPR relevant and Google is smart enough to want to avoid any potential turnover-based sanctions.
But mobile phone companies might start to worry about what Google might do with all that telemetry: knowing where the masts are and how many people are in any particular cell at any time. Google's already shown that selling advertising is not the only business it wants to be in.
The complaint is that if they gain ownership of HBO, Warner Brothers (movies/TV), and the Turner channels then they can then hold other TV competitors to ransom by demanding higher channel fees that get passed on to consumers
Disney has already announced it's going to do this with its catalogue and privilege Hulu over Netflix. Don't remember hearing the DoJ making any noises about that. Maybe it's because Disney lets the Trump fondle and fuck as many hostesses as he wants whenever he visits Disneyworld.
That none of the states have joined the DoJ on the suit speaks volumes. A court case will simply delay things.
America has a problem with cartels: ARPU on networks is much higher than in other comparable markets. Cheap debt and lax regulation are encouraging horizontal merges in lots of industries.
Analysts estimate Apple makes more money from iOS than Google does from Android. Which is really embarrassing
Not really. It's not as if Google doesn't make money from Android. Any comparison should take into consideration the capital cost (Google doesn't really make many phones so it's much lower) and also the market. Google might mot make money in China but it still makes money in places like India, Africa and South-East Asia where Apple's devices are too expensive for many people. It's also an almost pure services play which means Google continues to make money whether or not people update or buy new phones. And it provides Google with enormous amounts of data for future services.
The GPL, unlike the MIT license, protects against people adding a small thing to an open-source program that might become necessary to use it - thus taking it out of being usable in its open-source form. So the GPL is a good thing.
This is completely untrue. Anything that's MIT/BSD/Apache licensed keeps stays that way when code is contributed to it. What those licences don't try and enforce is whether somebody tries to embed the code in a proprietary product. This means more coding, less lawyering.
I suspect you are aware of the grid scale experiment that informed my comment (Falkenhagen): I used to work for the company concerned. Technically it works a treat, as long as you don't mind the losses.
Actually, no so thanks for the tip.
Lossy closed-loop systems make more sense where you can take producer/consumers off the grid. This would have the advantage of helping to reduce gross base load and insulating the wholesale market from the the problems associated with excess renewables.
Who knows what continued research will come up with? Got to be better than pouring more billions into things like carbon sequestration.
The cost of fossil fuels will need to treble at least before that happens, or some process to be invented or discovered that makes the production much more efficient.
In some situations it's possible to imagine closed loop systems that produce CH4 from excess power and use it as storage instead of batteries. In this case it's competing with the inefficiency of the battery charge / discharge cycle. But it has to be closed loop to prevent arbitrage or downright illegal profiteering.
Petrol or LPG are great ways of storing energy. If we ever get a way of making them from renewable power, water and air (CO2) that is cheaper than digging them out of the ground then we solve a lot of the problems associated with batteries.
Electric vehicles are currently popular because they are effectively subsidised by not having fuel duty.
The obvious solution is to make batteries easily swappable.
Not really. Electricity itself is far more fungible than battery packs. Not only do you have to devise simple, safe and quick ways of doing the swap, you're going to have non-trivial infrastructure for charging and storing them – protected from theft – and probably the need to ship them around. I think the company that wanted to do this for cars folded, but I could be mistaken. Much easier to build charging points even with dedicated MW lines.
Sounds like some kind of Nervana to me...
From Wikipedaia: the Sunway TaihuLight uses a total of 40,960 Chinese-designed SW26010 manycore 64-bit RISC processors based on the Sunway architecture. ARM and RISC give customers options they didn't have a few years ago. Even Intel has started making noises about custom silicon and FPGAs…
PS. I think you mean Nirvana…
But I - nor most people I dare say- don't use all browsers.
APIs are for developers and there care a lot about this sort of thing.
Most people won't give a damn about a theoretical 30% speed increase
People care a lot about perceived speed which for websites can be as little as 10% (0.1 second speed up for a 1 second page is notcieable). More important in Firefox are the changes in the threading and memory use. These changes are far more important than notional rendering speed. Quantum isn't perfect but after a couple of years chasing UX unicorns the Firefox team turned their focus back on the browser and I think they got most things right and will probably move back to Firefox from Vivaldi which now seems to be chasing unicorns…
You don't like it? Then fork the old code base.