Posts by breakfast 1557 publicly visible posts • joined 24 May 2007
I bet one could make a pretty great procedurally generated Astrology service that worked if anything better than most regular astrologers. It could also draw information from the best sources of superstitious sky-nonsense to derive internally consistent forecasts.
That would be a fun and interesting challenge, albeit one that may have been completed numerous times, judging by the weird one-liner astrology columns that often turn up in newspapers.
The thing with AI is that whenever you start getting deeper into questions around it you turn out to be doing philosophy. An entirely necessary and worthwhile activity, but not one that has the strongest record for coming up with practical answers over brief timescales. Whether philosophy will answer the AI questions or AI will answer some philosophy questions will be an interesting path to follow over the next few years.
As others have said, the problem is not that there are no ways to create working distributed architectures, the problem is that distributed programming is hard and whatever architecture you use will require you to solve some very tough problems. Splitting a service into decoupled microservices is not a terrible idea if they are truly decoupled, but most applications don't work that way in real life, so we tend to end up having semi-coupled complex systems that don't do what you expect when you expect as well as opportunities for the creation of endless race conditions that you never thought of during development.
If it was easy, anyone could do it.
The accepted wisdom is that security is hard, and just once in a while wisdom is accepted because it is true rather than simply because it is easy. The problem is that in general you want to offer people the simplest thing that can possibly work, but that either relies on human factors ( such as users being able to choose a good password, which is tricky because users are humans ) or mechanical factors ( such as locking an account to a single machine with a specific certificate on it which become problematic when the user loses access to the exact mechanical configuration they were using or needs to access the service from another system or location.)
There are good ideas around but it doesn't seem as though anybody has really got to the heart of this problem yet and a lot of very smart people on the usability and security sides have been working on it for a long time.
That may well be because you don't use it, which does make it easy to discount. It's a very good way of having very low-friction interactions with interesting people whose opinions you enjoy hearing. There is a lot of interesting stuff I would never have heard about if not for Twitter and it has given me the chance to talk with some of my favourite authors, comedians and scientists as well as keeping me aware of a lot of what is going on in my field.
It's also super-great for puns, though not everybody sees that as an advantage.
As someone who likes Twitter best of the Social networks and uses it a lot, I have no idea how they imagine they will make money from it and I don't think they really know either.
They've got a lot of users, but I doubt they'd keep them if they bumped the service around too much to get pointless adverts in for everyone to ignore.
At some point someone on the investment side is going to realise that having users doesn't mean automatic money and a lot of companies are going to evaporate in a puff of debt.
Agile works pretty well in some circumstances on some types of project - the problem is that it is a good match for a situation where you don't know everything you want in the finished product and you don't have a definite deadline for when it is finished. This is actually alright in a start-up type situation where you can start from core functionality and iterate, but it's no use at all for a major project with a hard deadline at the end of it.
Fundamentally the problem is everything in this thread. Anybody involved in government IT should be required to read these comments.
Seems like Corbyn has a lot more in common with Bernie Sanders, who is the only candidate on either side that appears to have even a slightly positive voter approval rating in the wider US electorate. Whether that will result in anything interesting happening is a moot point.
This makes perfect sense with the rest of the comment - nobody actually read the Britannica, so nobody actually knows whether Wikipedia is a drop in quality. It's certainly cheaper though, which is the important thing where the knowledge of the world is concerned, right?
"Mixing wet-ware and silicon-driven cars on the same roads, however, is a recipe for disaster as far as I am concerned. Ideally there should be seperate roadways for each mode of transport, but that is unfortunately completely out of the question."
Your suggestion answers itself in the previous few paragraphs. People are bad enough on the ground, but automated systems can manage flight quite adequately.
Self-driving flying cars it is. Problem solved.
"Them" and "their"?
Not seeing your own $segment_of_society as the default for everything is not the same as your $segment_of_society being discriminated against. But until your $segment_of_society has been discriminated against, people often don't understand that. It's called privilege and most of us who work in this field are lucky enough to be born into at least some of it - nothing wrong with that but if we assume that our $segment_of_society is the default, we exclude people who don't belong to it and we probably do that without even meaning to.
It's not political correctness, it's just good manners.
In a lot of places they use non-robotic mules for that rough-terrain work and they are amazing at it.
It's easy to get excited about what we can do with clever and powerful machines to the point we lose touch with things that have worked for us consistently over thousands of years. I'm not saying we shouldn't be trying the machines, but evolution and selective breeding have also put some impressive tools at our fingertips if we choose to use them.
Jedi Knight was one of the best games of the period, and probably the only FPS I have played through more than once - I wanted to see the dark side powers but at the moment when I got my light sabre I knew that this was a tool I could only use for good and ended up playing through trying to use only the light sabre from that point forward. A truly great game.
I will also never forget the great disappointment that was X-Wing Vs Tie Fighter. It was just not enough of a game for the money and at a time when network-only games were a tricky proposition but I suppose that it was a valuable preparation for the disappointment we were soon to experience when the sequels arrived.
I do wonder how much speed increase one might achieve by simply designing a completely new architecture based on modern design techniques. As far as I know, most computers we use these days are highly refined versions of the Von Neumann architecture - Quantum computers are a different beast so they have exactly that opportunity leading me to wonder whether this is part of why they show some impressive jumps in pace.
A lot of people complain about how government is inefficient, particularly where IT is involved, but one reason this comes up is that if you are a private company and someone can't use your service, you lost a customer. That happens, and businesses can deal with it. If you are the government, your services must be available to everyone. That introduces a vast number of complexities and edge cases that people who have never worked in this area are unlikely to think of.
I'm sure many cost overruns are a combination of the extra effort involved to mitigate those problems combined with the fact that companies willing to pitch in low to get the business are either unaware of the complexities involved or working with a plan to increase the amount they bill later. Many "overruns" are most likely to be close to the cost of the project that would be projected by any rational person, but because the job goes to the lowest bidder companies have to make up fictitious cheap prices that they are well aware they will never achieve.
I'm not defending the process at all, but a lot of the perceived waste is more a question of people trying to do a very difficult job. The government could probably save a lot of money by having some decent in-house development teams and some expert contract negotiators, but invest-to-save doesn't bring extra cash to companies where ministers have directorships, does it?
I did that, but I didn't think it would be very believable if I completed the whole thing so I only switched around one face.
Of course, the downside of this was I made it impossible to solve the cube after that. Wasn't until my cubemaster cousin tried and failed for some time that anybody realised my deceit...
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
