Posts by Charlie Clark 13422 publicly visible posts • joined 16 Apr 2007
Any form of equal opportunities legislation makes it illegal by definition: if you prefer a candidate for a reason not related to the job description, you are discriminating against others who can, therefore, take you to court.
As John notes, there are schemes that encourage and reward employers in some situations. For example, many workplaces may not be fully accessible for disabled people and it can be expensive to make them so, grants can help. Other schemes may involve financial support for placements for disadvantaged school leavers. Some of these schemes have turned out to be successful, others, of course, little more than gravy trains,
Positive discrimination is illegal in most European countries. However, it should be noted, that where there is anti-discriminatory legislation, violating it is either way a criminal offence. This has kept Europe largely free of the civil suits asserting discrimination and seeking compensation: they do occur, but normally after a violation has been deemed to have occurred by a court and rewards are much lower than are possible in the more litigious US.
I've never been a fan of flag-waving "social justice" employment practices, which are often little more than window-dressing. But I think many good employers have learned to think beyond their bias (we all have one) and maybe give some candidates a second look. But they're also likely to engage in active outreach campaigns to the more socio-economically disadvantaged parts of the society, something that it often conspicuous by its absence the US. Are you poor and from a racial minority? Then why not join the army? McDonalds and the like should be given credit for sometimes actively giving chances to such people, though normally after working them, ahem, like slaves for a while, but at least they've learned to promote from within.
Have to agree with you and the real bug is somewhat obscured by some fairly hair-raising deployment practices: it should never really get that far:
$(echo "SELECT COUNT(1) FROM gw_sessions WHERE session_key = $quoted AND session_type = 'sdcust' AND (expiration IS NULL OR expiration>NOW())" | $db)
Where $db = /path/to/psql. I mean, really?
Not only do we have the underlying bug in Postgres, but also quoting to the shell and a requirement for psql locally and apparently without further authentication.
But I'm impressed by the skills of the exploit.
No, I don't mean that. I mean it much more generally. Currently, we don't have a culture and legal framework that will accept that, while bugs may occur, these are rarely malicious, deliberate, or even the result of poor practices, mitigation is more important. Threatening sanctions could be useless — think of the many at least partly abandoned open source projects — or even counterproductive. But neither is the current shoulder-shrugging promise to do better next time from the likes of Microsoft and Apple.
But the responsibility isn't solely with the developers because — and I don't mean you should get out the jossticks and whalesong — in a sense we're all involved in this. Can we come up with ways to set and follow standards and, where necessary pay for mitigation? Difficult to see how current approaches avoid creating a gravy train for those probably least suited to improving things! And limited sanctions for commercial offerings may be part of the solution, though I suspect stricter disclosure and resolution timeframes may be both more acceptable and effective. We are going to have to do something because at some point, some awful disaster is going to be directly attributable to a software error.
…those that from afar look like flies…
The biggest problem is that you often can't know that something is a bug until someone works out how to exploit it. This is the reasoning behind the general exemption from liability that software has enjoyed for the last 40 odd years.
In other trades, unlimited liability does apply in the US but has occasionally led to curious outcomes: once it became possible to identify the sex of a foetus early in pregnancy, it didn't take long for the first class actions and then for obstetricians to refuse to tell patients or even take them on.
What we could probably do with is a better culture of accepting responsibility and mitigating the issues in a way that can be applied both to commercial and open source code.
A dictionary is just the Python term for a hash table, so there's every reason to suppose that they could benefit. However, I'm not sure we'll notice much in everyday life because they're generally not that big but we might well see differences in databases, which are designed to be big and which use hash tables all the time.
But advances in mathematics do often lead to noticeable performance improvements, even in a language often dissed for "poor performance" as Python! ;-) And there have indeed been noticeable improvements in some of the more recent releases. Tim Peters recently posted about an unexpected boost in Pythion 3.13 and, though I can't remember the details, something involving number theory and some other goodies demonstrating that code, which was once state-of-the-art, or at least good enough can often be improved upon.
I'd agree that Postgres isn't really replacing SQL Server in the corporate space, even though companies would generally be far better off running Postgres containers than the ridiculous overhead of Windows VMs needed for SQL Server. This is more down to legcay, inertia and change resistance than anything else: I'd be surprised to see many real world use cases where SQL Server is the better choice. But migration costs and some less experienced DBAs probably appreciate the GUI management tools.
However, with Microsoft's shift to getting everyone on Azure, you can understand their renewed appreciated of open source as a way to improve their margins.
For any new project, Postgres is probably the best place to start and be prepared to learn how backup, replicate and move your data to avoid expensive lock-ins. A couple of weeks training will quickly pay dividends.
NoSQL originally meant "Not only SQL" and referred to ACID + various combinations of key/value, document and graph databases for particular domains and since then we've seen a revived interest in column storage. The fundamentals of relational calculus that underpin relational databases are what guarantee its superiority for general use and make it essential for reliable transactions or analysis and other systems suffer a significant penalty when they reivent the wheel in the query engine rather that data definition. Denormalising for specific purposes – whether buckets for ephemeral and irregular data or materialised views for analaysis – have always been an option.
But what I find most interesting is MS contributing openly to a liberally licensed open source project which Postgres being BSD. This gives the lie to the notion that only GPL software will receive contributions from commercial users.
I think there's also a paradox in there about how you'd prove it…
I think that, one day, we might find effects that look like FTL communication might be possible but it'll turn out to be a "dimensional issue" or never applicable so getting the results of the 3:30 at Chepchester before the start of the race is still a no-goer.
But that's the whole point of digital computing using binary devices where you can easily detect on/off states. The first ones were indeed more or less cobbled together but this allowed for continual improvements, and occasional step changes, along the whole chain. But the principle – that any equation that can be expressed using binary logic can be automated – meant that calculations that had previously taken days, could be done in hours, if not minutes. This was required for a military increasingly keen on understanding fluid dynamics and, er, other things, but also potential commercial advantage for businesses.
The innate "fragility" of the equipment, of couse, was the direct genesis of the term "bug" that we still use. But it also encouraged the development both of more robust – solid state – components but also of resilient strategies so that less would be lost the next time a butterfly got in the wrong place of a valve failed.
Steam is just as hot as the water that made it
This is wrong, steam is hotter than the water it has just boiled out of: the transition usually isn't at the surface but at the base as the superheated steam bubbles it's way up and out. As you only want around 90°C, you don't really want to produce any steam at all.
The table is incomplete and needs to include the times for poached and coddled eggs at a minimum and we need to know more about the toast: how thick is it, how much butter (and what sort) is on it and how long it's in the toaster for.
Honestly, they seem to handing out academic titles to anybody nowadays.
Yours sincerely
Prof H. Helibecnof
My understanding is that turbines have greater efficiency than ICEs – but it doesn't really matter, something that relies on oxidising hydrocarbons to generate power – the important thing is separating power generation from the drive train. This allows for modular systems and, therefore, potentially greater improvements of their various components.
We should give credit where it's due: to the engineers for their hard work, but also to Musk for his nous, conviction and even his ability to ensure cheap finance for the company. But, as others have been saying for years: where's the next version? The Chinese have been working on all aspects of the chain for years and now lead in most of the technology. And now they have by far the largest domestic market which will help work on fit-and-finish, routing and automatic driving.
That said, I think battery-driven cars are a dead-end due to the limits both of the chemistry and the means of delivery: charging networks. Electric drive trains are almost certainly the future for most of us, I suspect they're likely to be powered by hydrocarbons either as turbines or fuel cells for a while yet.
There are several statues which will limit both Musk's access to data and ability to act and Trump cannot remove these by decree. Elevated security clearance and access will require Senate confirmation and the Emoluments Act may well come into force for many of the "cabinet", even though Trump himself is now probably immune. It will all be a big distraction from the business of government, though that may be just what some of those involved want.
No, you're conflating his personal immunity from prosecution with the legality of his decrees. Courts can, and are, already ruling against some of the degrees. For example, Trump could pass a law requiring the death of left-handed people and start the ball rolling by cooly shooting one. SCOTUS brainfade probably means that nothing would happen to him but the decree would be ruled unconstitutional and the protection would not extend to any lackeys that tried to enforce it; though Trump could, of course, pardon them and I think we may see an awful lot of pardons over the next few years.
Sales are one thing, profits another an Tesla made most of its profits from emissions certificates. As it can no longer do that and is facing competition from cheaper and better vehicles
"Hey, Elon have some more of our money!" said a guy in a red baseball cap…
The economics and laws are different: getting funding for the MVP beloved of Silicon Valley is almost impossible in Europe and the idea of being able to scale up quickly to dominate the market doesn't make sense. And a safety-first culture tends to lead to overengineering before any product is launched, whereas US companies will often happily pay to see things crash and burn and equally as happily go bankrupt if claims against them stack up too much. They can walk away and start a new company immediately.
I also like the opportunities for arbitrage in data centres. Tariffs like those proposed would make them immediately uncompetitive in America but you could be build sham ones and run the calculations elsewhere, even in Europe. Difficult to prove otherwise.
As others have mentioned: Motorola still made phones in the US until recently but these were ones with the most problems… Apple could lauch the A-Phone along the same lines…
It worked out pretty well for Germany: free trade tends to favour specialisation. This could be agriculture in countries with lots of sunshine and water; it could be manufacturing with good logistics, energy and skilled labour force; specialisation solely over cost of labour has rarely lasted for long. When one country focuses on one or two areas to specialise in, it invariable opens other opportunities for others in other areas and net exports usually run into problems as to what to do with their surplus: Germany has for years being investing its profits in stupid American schemes – infamously writing an already bankrupt Lehman Brothers a big fat cheque.
"innocent" owners? New to the shipping business are you? All these ships are being registered with the help of some think brown envelopes. Owners will try as far as possible to aim for plausible deniability and, since ship registration fees became an issue, registrars will turn a blind eye to the ships and the crews. That is until someone mentions sanctions, ie. problems for all ships registered. The only exception thus far was the Chinese ship with the Russian crew. The Chinese refused to cooperate but we haven't seen many of their ships doing the same route. For all its bluster about the South China Sea, China is extremely keen on maintaining a general freedom of navigation because it has so much to lose.
Most of the ships in this shadow fleet are pretty much scrap, so it wouldn't be a great financial loss. Bringing it into port and finding out as much as you can from the ship and crew and putting the "Maltese" owners under pressure might bring greater reward.
I don't see how Broadcom can expect to be able to segment the market so neatly.
The charges will force some customers out and prevent new ones from joining. Difficult to see how this isn't going to introduce a new generation that doesn't know VMWare fairly quickly and the know-how for migration will continue to spread amongst those who must or would like to.
Also, as others have said, this might give the impulse to some to ditch at least part of their farms of VMs in favour of containers, where this makes sense – I know containers aren't that much different and memory, and particularly IO, will limit what you can do, but in some situations you can drastically limit your footprint this way.
the reference to core *nix distributions is simply erroneous here. Multimedia systems are much more susceptible because they invariably interact, at some point, directly with hardware.
If you're talking about the many eyes make shallow bugs ideas, this is a fallacy. While I'm a big fan of open source, making the source open does not avoid bugs, where open source does come into its own once a bug has been identified.
Apple's developers have made a common mistake that probably could and should have been caught by code review, testing and static analysis. But Apple steadfastly continues to refuse to improve its release and update procedures, so that, instead of frequently releasing minor patches, it prefers to release monolithic sets updates and new features once a quarter, which often and unsursprisingly introduce new bugs because of their scope scale. But they can point to millions of satisfied customers – I've got a Mac myself but no I-Thingy and I prefer not to live on the bleeding edge, usually keeping at least one major version behind – no matter what we say.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
