Posts by Graham Cobb 1535 publicly visible posts • joined 13 May 2009
Sure. But I'm with MMcF: that is an important lesson and this experience needs to be highlighted in the training of future flight controllers. With hindsight, it is obvious that that was the right process. But nothing wrong with reminding people of the obvious on occasion!
I mostly use SearXNG nowadays (I used to use SearX).
I run my own instance in a Docker at home but often just use https://priv.au/ directly. I have other instances set up as browser search engine options as well, in case some are down.
If the "community" had asked for some kind of comments that could have led to machine-generated documentation, it woudl have been far, far better.
Unfortunately I can't agree with this. My real-world experience with comment-based machine-generated documentation was that the only thing you could assume about the documentation was that it would be completely unrelated to the code. Even the best programmers often forgot to update the documentation comments so the resulting documentation was the worst of both worlds: it was neither up to date, nor did it correctly document any earlier version of the system!
That was supposed to be one of the main reasons for the OO approach in the first place: the code would be "self-documenting" because many relationships would be explicitly defined in the object hierarchies and declarations. Of course, it didn't work as hoped but it was probably still an advance on documentation based on text comments completely unverified by the compiler.
My local co-op now has almost no fresh vegetables on the shelves (although still a reasonable amount of meat - I presume because that is expensive and moves more slowly) and certainly only one sort of anything (one sort of mushrooms, etc). Milk looks quite low as well, unless they are keeping some in the back to reduce panic buying.
And signs saying it is due to the reported problems. Looks like its logistics systems are badly affected.
(The page below is 'protected'- had to 'view source' to copy this. That says it all.)
I suspect that (in this case) it isn't deliberate. Certainly, in my browser, I could do Select All and I could also start a selection drag by clicking in the fixed portion of the text (near the top) and dragging further down. Once selected, I could Copy with no problem.
It looks like careless capturing of clicks for the opening of sections. So FU rather than conspiracy. Still piss-poor programming.
The real Signal app cannot be monitored by any security services as far as anyone knows.
However, using the real app means that when he loses his phone (or has it taken away from him by his boss/the authorities) he can't retrieve the messages he might weally need to get him out of trouble with said boss/the public/the authorities of whatever other country he wants to impress.
He doesn't know that you can have security, or you can have archives. Not both.
The fentanyl "problem" will solve itself just as soon as Trump's friends in Big Pharma work out a way to make it legal so it turns from being banned into "the very best way to live a long healthy life - me and all my gerontocratic friends use it every day!"
If consumerism has gone crazy then just increase sales taxes. A simple and fine-tunable weapon, fully under your own control - and it could even bring down government debt.
Actually, syncing is really hard. Many, many years ago I worked on syncing for PIM data (structured data, for calendars, addressses, etc). In general, it is impossible to do correctly -- all you can promise you can do is make the two datasets the same. In general, if there can be changes on both sides, you can create conflicts which the software will not know how to resolve "correctly". Even if you timestamp changes.
To see that this is obviously the case, imagine you and a colleague are both editing a document (say, a proposal to a customer). You notice that the price and the delivery date are inconsistent (as a price change is coming up). You make a change to your copy which changes the price, and your colleague makes a change which changes the delivery date. The software cannot reconcile those changes - in this particular case either would be valid, but changing both the price and the date would not be valid!
That is a carefully chosen example, but it turns out that in almost all structured data there are similar cases of conflicting changes which do not look like they affect the same field but which cannot be resolved to consistency.
However, that doesn't excuse OneDrive! It doesn't attempt to make semantic promises - it really only needs to make sure it saves one of the versions of the file and lets the user know of conflicts.
If a commercial airplane can be brought down by any use of a perfectly ordinary piece of equipment carried by almost all of the passengers then I don't want to travel on that plane.
I am amenable to the argument that avoidable risks should be avoided, but I reject entirely the idea that a plane that can be interfered with by legally carried devices in the hands of untrained personnel can be approved as "safe to fly with passengers". JFO.
Except that, in this case, Trump really isn't that charismatic.
In the past, it has needed the leader to be charismatic, but the US experiment - with a two-party system that can't be broken, an electoral college system that means only a tiny minority of states count for anything, and a head-of-state with much too much power - has failed to cope with the destruction of serious journalism and the dominance of social media populists.
I hope someone is taking notes for the period of reconstruction after the global structure from the last 100 years collapses and is eventually rebuilt (after 1000 years of a new Reich?).
What happens when fewer babies are born than the number of people who die in a country?
The impact of the human beings from that country on the Earth is reduced. Seems like a good outcome to me, in moderation. So, governments can, and should, tweak that control knob in line with a careful analysis.
who'll be around to pay the taxes to fund care for the elderly?
As long as the previous item is done with care, analysis and forecasting - the declining population.
All perfectly manageable as long as there are grown-up politicians, being advised by experts doing research and analysis. Not one-line, quotable, slogans.
Uh... why do you need an iOS version of Thunderbird? There are many iOS IMAP email clients - all of them can access the same IMAP email server that Thunderbird accesses.
Touch GUIs are very different from mouse GUIs, so I don't understand why you want to try to replicate Thunderbird on a phone/tablet? Why don't you use a client designed for that GUI? That is what I do. I use Thunderbird on my PC, and I use other clients on my phone and tablets - all access the same folders remotely.
Sorry, it is your post which is "simply untrue".
It is true that the interview system allows them to consider many factors - not just exam grades. As part of that process, they may take into account whether a candidate has been deprived of the opportunity to fully develop or demonstrate their qualifications.
But this is nothing new. In my day (about 50 years ago) almost everyone applying to do maths had top grades so the interview process was pretty much the only way they selected people. At the time, it was already the case that grammar school kids (like me) had an easier interview process because they had recognized that the public school kids had had a better opportunity to learn how to impress in interviews (and also had specialist teachers to cram for A-levels and for the entrance exam).
So, their attempts to assess who can best benefit from the Oxbridge opportunity have never been based on "pass marks for entrance exams". It's never perfect, of course, as demand massively exceeds supply. But it isn't some modern "diversity" policy, or quotas - just an attempt to assess who really are the best qualified, combined with a lot of luck.
I am no expert on cyber-security at scale but I can see a few principles, which seem to be completely different from the approach of the Government...
1. Fix the bloody personal data problem!!! The biggest risk to people is the problem of personal data theft. There is one, and only one, real answer to that: prevent companies from requiring (or acquiring) any more personal data than the minimum required for their service to operate! At the customer's option, they can ask the company to store more data to provide a more personalised service but that can be withdrawn at any time and must be unrelated to the price charged. I might allow my TV provider to keep information like how far I am through a particular series, or what sorts of films I like to watch, but that should be unrelated to how much they charge me and I must be able to delete some or all of my data at any time I wish.
This single item would dramatically reduce the amount of personal information stored and the attractiveness of many of the cyber attacks.
2. Critical national infrastructure (power, water, communications, transport, etc) funding must be strictly controlled and the companies operating it must have strict responsibilities (especially for security, safety and reliability), which can be enforced against some entity which cares (not limited liability shareholders).
3. Private companies providing services to government (particularly in areas of national importance) must have some sort of strict liability to their customers so the company invests in the necessary cyber-security.
Sure, these are easy to say and hard to do - but this needs to be the debate, not fines which will never get paid.
I won't disagree, except to point out that they spend a lot of money which the datacenter guys want some of.
To be honest, I'm not aware of another application which will really be speed-of-light limited in the UK - although we did actually worry about it in applications like Prepaid Mobile Charging in large countries like Brazil and the USA. And, of course, anywhere which tried to do anything much over satellite connections. For a while some of the attempts to build large ring-based network architectures threatened to be a problem in some places, especially under failure modes where traffic got sent "the long way" round a ring (more from switching delays than true speed-of-light considerations).
Of course [and I didn't downvote you]. But the reality is that power is not time sensitive in the same way!
Meeting the latency requirements of many of the data centre customers (which is what drives the price they are willing to pay) is what is hard - or even impossible. Power for the data centre costs a lot of money - but is relatively easy to solve by siting the datacentre near a major grid point. Putting a data centre in Dinorwig (to continue my silly example) introduces an absolute minimum of 2ms latency (just from speed of light delays - more with switching).
And that was so weird! I was there (for the Olympics) and I, along with everyone else, assumed it was a terrorist attack that would be escalated - but then there was nothing.
Do we know what really happened? Was it really just a protection money threat that the French government quietly paid off?
It's probably easier to move power than data. For network capacity, round-trip delays and fault tolerance, it would probably be better to keep it within South/Central England as much as possible. You don't want to try to build the network capacity to shift a significant amount of London's data processing demands to, say, Dinorwig.
First year at uni was all on their cluster of VMS VAX 11-750's that we accessed via an X.25 network.
You may, or may not, be interested to know that the DEC X.25 software (for both PDP-11 and VAX systems) was entirely developed in Reading. Some hardware (and firmware) came out of Ireland and France, but the US was almost entirely uninvolved.
Please, please stop just believing the mainstream media, they are as deceiving as any Internet source.
No, that is a false statement.
The "mainstream media" (for whatever definition of "mainstream" that you personally choose) is not "as deceiving as any Internet source". It is often wrong, of course - and some definitions of "mainstream" are wrong more often than others - but there are "Internet sources" much worse than even the worst "mainstream media".
Of course. All that is true. But that doesn't mean that AI summaries are not useful.
Many people, on many occasions, have no interest in a particular point , and an AI summary can mean that they still learn something or help them decide whether the article is something worth reading (now or later).
I'm no fan of the rubbish that is called "AI" today. But I realise that even today's crap AI can have some uses, on some occasions, for some people.
Can't believe I'm backing Rupert Murdoch over Brave
I think it tells you something about yourself.
Rupert Murdoch is a very old experienced business person. I do not lose any sleep worrying about his business.
I do worry about people who believe that the way to save journalism is to artificially restrict the tools people have for consuming it. If the newspaper thinks it is good business to make (some of) their content available for free, they can't complain when someone reads it and summarises it for others. All newspapers do that all the time, after all.
The technology that can do that without a person involved is the Jacquard Loom of publishing. Yes, it will destroy some things, and it will enable others. There will be big changes for the people who's livelihoods depend on the old way of doing things. But that is inevitable.
Yep. Same thing with my Brother DCP-L8410.
I have many times recommended Brother printers as this one has been an excellent performer and works well in my (Linux-based) environment.
As it happens, I do use Brother toner (my usage is low and so the cost is small - and my local IT shop provides it on 24 hours notice, so the convenience is worth the price). However, if they really start doing these tricks I will certainly not be recommending them again!
I was referring to the GR effects. I think I read that the difference in clocks between the satellites and earth-based receivers is enough to be noticeable and has to be allowed for.
My question was whether the difference between the satellites and the moon causes an effect that is larger or smaller than that (although of the opposite sign, of course).
I run all my own routers and networks, including a few different VPNs for different purposes. I do log all outbound connections but I admit that looking at those logs gets very tedious.
At the moment, my compromise is that I run 3 separate main WANs.: one is for visitors - it can't access anything locally but also doesn't block much stuff from the outside world; the second is for gear (like printers) - it can't access the outside world but can communicate with other devices on its own network and on the third network; the third is for our own devices - they can access the second and third networks and the outside world, with some filtering set up. All the rules are enforced in my routers (running OpenWRT with my own configuration).
It's not perfect - and I have had to open a few tunnels I would prefer not to - but it generally works well and it blocks a LOT of access. No devices (printers for example) send any data to their manufacturers, for a start. I also watch the logs to see what sites things like PCs are accessing and block many of them.
Actually I find a tablet useful when travelling (I'm retired now and don't always want to take my laptop with me). I have gone through several tablets with /e/OS (I tend to leave them on planes - the downside of not being a laptop!).
Currently I am using a Samsung tablet, bought inexpensively off eBay - it says SM-P615 on the back. I guess it is convenient to have a Murena supported device but I prefer to choose one that fits my use and then put /e/OS on it.
OK, I don't use either OBS or Fedora, but this problem has been around ever since distros were invented. I don't think users find it hard to understand the choice between a safe, stable, supported but rather old release and a new, sexy, but buggy release.
All this issue means is that (i) distros should make it easy to switch between using the latest distro version of an app or the (main) non-distro version, and (ii) the distro should commit to tracking serious upstream bug-fixes and incorporating them into their version of the package (or, if that is going to be too hard, dropping the package from the distro).
All I ask from my distro is that the packages it ships are supported, with critical (particularly security-related) bugs fixed in a timely fashion. If I want the latest sexy upstream features I can uninstall the distro version and install the flatpack version.
Of course, it would be nice if the distro made it easy to select, on a per-package basis, whether to use the distro version or the upstream version. Preferably in a standard way for all packages that choose to participate.
Unfortunately, you have been cruelly misinformed. Postscript (and, hence, PDF) has always been a complete language, and creating documents on the fly has always been a fun sideline.
After all, you can have a short PDF which contains all prime numbers - all infinitely many of them - and prints them all out if you want - just keep supplying paper and ink.
I would make do with just being able to tell my cheap chinese Christmas lights strings (for my Xmas tree, outdoor trees, draped over the garage, etc) to come on without bloody flashing!!
They all seem to be similar - they have a controller built into the power plug and (mostly) have a remote control. Either switch can be used to cycle through a variety of annoying patterns. But there is no way to set the startup to be my desired pattern (actually just on - no flashing, fixed colour).
I would be happy to have something that could either replace the power supply/controller or the remote control. Including a timer (so I could replace the mains timer I use for each string) would be even better.
Anyone know of any such gear?
I've not been running vendor-supplied software on any router pretty much since my employer pulled out of the router business in the 90's. Now that the LEDE spat is resolved, OpenWRT works well and TP-Link hardware seems to be a good choice for it nowadays.
Of course, I am only 98% certain that TP-Link firmware doesn't operate a clandestine capability for China to snoop or interfere with the operation of OpenWRT. So, how about, instead of taking potshots at banning TP-Link, how about the US government puts the effort into detailed security testing to convince itself (and us) that TP-Link is not interfering when the router is running OpenWRT.
At the same time, various other governments (including the Chinese) could do similar testing to make sure that Cisco isn't doing the same thing to allow the US to spy on us.
I, similarly, use a very tightly locked down Firefox instance for my main browsing (as I type this, for example) - with every tab in its own container, javascript disabled by default, and uBlock Origin, Privacy Badger, various Fingerprint Defenders and a few other tools.
However, in the last few months, I have found that more and more sites just won't run in that environment. Mostly they do nothing. In many cases, I end up using a disposable Brave instance - better than nothing, even though it is Chromium-based.
Yeah, I did that last summer with a non-governmental entity in Paris who wanted obnoxious ID requirements. It wanted to scan my ID, and then use my camera to see a picture of me. Then it would let me make the reservation I wanted to make.
I didn't lie to it about who I was but I decided to use the experience to see how hard it would have been to lie to it. The answer: not hard. I scanned my UK driving licence (which seems to have a small and poor photo of me) - on a fairly low resolution on my scanner. Then I told v4l to send a still photo of my face as if it was a videocam. I had no trouble getting in.
In this case, I was actually identifying as me. But I could probably have done those two things with a photo of <insert name of favourite actor> and a photoshopped driving licence scan.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
