Posts by Andy 73

Re: councils do not like central gov

This doesn't have to be a centralised solution - just a common set of tools that authorities have enough information and understanding to apply to their specific needs. We absolutely do not want one central monolith "to rule them all" - we want to eliminate the mistakes that are consistently made when specifying, implementing and integrating what are fairly common organisational tools in government.

And that is where a centralised 'centre of excellence' can provide education, skills and advice (and some common tools) that authorities could utilise before spending billions on failing projects.

Sigh...

There is a very broken concept of independence and competition, where this is one of the areas that has the strongest case for shared, common solutions.

Of course our government (let's be clear, all UK political parties and the devolved governments) are heavily lobbied and 'supported' by the big players in such a way that they do not believe there is a credible alternative - they get offered a "huge discount" on the headline price (and no concept of just how badly over inflated it is), and get sold inflexible monoliths that are deeply tied to consultancies that are the only choice when it comes to (absolutely obligatory) customisation.

This is further encouraged by local council officers who need to protect their budget, staff and the perception of the value they provide. They are heavily incentivised to believe that the problems they are solving are unique and complex, and the big providers are going to do absolutely nothing to persuade them otherwise.

If the civil service was worth a fraction of what they claim, they would be able to provide an analysis that shows the commonality across the many local authorities, the areas that can be "solved once", ways to break apart the monolith with well defined boundaries, and patterns for customisation at cost. This does not necessarily require getting rid of the (deeply embedded) big suppliers, but education for local authorities on how to engage and specify work. There is absolutely no excuse for the civil service not to make this priority - both as a facility for authorities and as a skill set within the service. Yet... here we are..

Re: If our chips are that good...

I'm not sure what you're trying to say here. If the UK is "25% poorer" due to the decline in Sterling, then it's 25% cheaper for the global market to invest in Britain - which do you want it to be?

As it is, FDI in the UK appears to be broadly the same, though we've had massive volatility over the last decade for some very obvious reasons that are global, not domestic in origin. It's also worth reiterating that contrary to popular opinion, manufacturing has not gone away in the UK, but in common with most Western countries, services have massively grown over the last few decades.

Of course poor countries are developing - that's where the cheap labour is. It's what gave China decades of growth (that's now slowing as their labour costs and quality of life increases), and what is driving industries to India and South America. The West will not see growth like that in our lifetimes - which is why investors keep pumping money into meme stocks and technology bubbles. Expecting otherwise is ridiculous.

Nor should it be surprising that the UK is struggling, when most of our political classes have a poor grasp of economics and just want to argue over minor readjustments in the budget. If you want to see serious growth, you need to encourage an environment where many options can be tried without investors and companies having to risk their existence each time. That 'support for risk' is what has allowed America to dominate the tech industry. Conversely, the tendency for political groups in Europe and the UK to want to tightly control the direction of investment and innovation is one strong reason for our relatively poor performance. It turns out politicians are both slow and ineffective when it comes to picking up tech trends.

Unfortunately, as usual the discussion is framed in terms of how many handouts can be directed to the right people rather than how the wider economic environment can support increased activity. In that sense, Brexit Britain is a myth - in or out of the EU, our environment for innovation has not changed due to the usual institutional feet of clay. Arguing over tax breaks and handouts is missing the point that we need more change, not less.

Re: Moon landing

I'm not sure if you've fully grasped the physics here.

3 tonnes of mass (random figure for illustration) around the nose of Starship (50 meters up) has the same rotational inertia as it does on Earth. 20 tonnes of fuel at the bottom has one sixth the weight (and therefore exerts 1/6th the stabilising force). In essence, Starship acts like a rocket six times taller when it comes to trying to balance it on the moon. That's exactly the problem the two recent landers have had, and they didn't go for the Buck Rogers 'tall thin tube' design.

IMs design and learnings have nothing to do with SpaceX's own design - other than to point out that this is a non trivial task, where basic design can make things significantly easier or harder. Pointing out that the Starship geometry is intrinsically less stable (no matter what the distribution of fuel is) is not to say they cannot do it, just that they have chosen a design that makes the task more difficult. How many times has SpaceX successfully landed Starship on Earth so far?

Re: Moon landing

Great appeal to authority there. It would carry more weight if two groups of exactly rocket scientists hadn't both independently failed to land rockets upright on the moon.

Turns out some jobs are difficult - who knew?

Re: Musk? Who trusts this guy?

That may be the case, but Musk has sufficient control of the general direction of SpaceX (committing to contracts, general process of development, business direction), that he still appears quite capable of introducing uncertainty, if not complete rabbit holes.

Certainly the solution they have committed to for Artemis appears to be extremely poorly architected and planned, and the payments for development are... interesting. Regardless of how clever the folks are at SpaceX, they've been put in a position where they may be regarded as the single point of failure for a hugely (and possibly unnecessarily) complex mission, required to solve too many novel problems to meet the deadlines and without sufficient budget to complete the development process without huge losses.

Musk is not just a figurehead at SpaceX - he controls the money, and with it their obligations and overall direction. I'm pretty sure Starship would not have the current architecture (or even exist) without his direct influence.

Re: "focus the team on the launch of the next-generation vehicle"

Damn autocorrect.

Re: "focus the team on the launch of the next-generation vehicle"

Sinclair was late, usually because he couldn't get manufacturing going fast enough. He rarely announced products that didn't exist.

Musk on the other hand fakes product videos, claims features that don't exist and viscously attacks anyone who calls him out.

I think I preferred Sinclair's approach. He also didn't claim to be saving the planet.

Re: Musk vs. the investor class

Yes, just wait till you find out where your pension was invested...

Re: Backwards

"Eventually they'll want to do everything a fully specc'd computer is able to do" (I paraphrase).

That's an argument for giving primary school kids a Unix computer and telling them that eventually they'll be coding in Python, so they'd better learn to download endless libraries. I don't think that's how education should work.

You say that "It can't do that" is off-putting - I'd suggest that so is "Yes it can do that task, but first you have to install these five programs, set up the correct directory structure and install exactly the right versions of these six libraries".

A genuinely educational environment abstracts away as much as possible so the pupil can (a) learn some key concepts without distraction and ideally (b) be rewarded by a positive result. Yes, sure they eventually want to be able to write a full networked game in Unity, but there are a lot of steps along that path that benefit from not exposing them to the full horrors of a modern computer.

Re: Backwards

That, unfortunately is exactly why it isn't in every secondary school in the UK - very few schools want to manage a collection of SD-Cards, SD-card writers and other paraphernalia to deal with "Miss, it's not working". Fewer still want to waste time on a constant upgrade cycle of slightly better releases, security updates and fixes. Even fewer actually understand anything about selecting a release version.

That's ignoring that there is no official "for schools" release (which would ideally drop kids into a robust environment with simple tools ready to teach, rather than a desktop and command prompt).

Again, a cheap Android tablet or even an Arduino gives a more reliable, focussed education platform than a 'easily fixed' Linux computer.

This is a source of some frustration, as I see again and again techies decide something is "educational" based on arbitrary convenience for an experienced developer rather than a practical understanding of the challenges of teaching in a school. Yes, I understand that these are used in code clubs, but this is not an experience that is available without experienced and enthusiastic evangelists. That shouldn't be a prerequisite for an "educational" device.

Re: The tube

Yes, it certainly would have. Many, many hardware designs were tested out by cobbling together interfaces to 8-bit computers of the time, since most of them exposed pretty much the entire system bus to peripherals. As an example, the Apple II had a whole range of expansion cards that allowed you to try out alternative and sometimes exotic processors and operating systems.

Re: Backwards

I've argued elsewhere that the Pi is not an educational machine - at least in the sense that it has a meaningful presence in the UK curriculum or schools (outside of computer clubs). Our kids have sailed through primary and secondary schools, all the way through A-level computer studies without having once turned on a Pi. That's despite us living in Cambridge.

Whilst Upton has talked about it being "like the BBC", I've always understood that to mean a standalone, easily configured, hardware accessible 'building block' - that turns out to be very useful in industry, hobbyist areas and higher educational establishments where electronic design is relevant. The Pi itself is not actually particularly designed to be a 'computer for schools' - it neither fits the curriculum, nor is it actually a very practical tool for schools that really don't want the hassle of cobbling together a working system from a bunch of parts. Sure, it's cheap - but by the time you've bought monitor, keyboard, cables, mouse and all the rest, it's not very competitive with a boggo Android tablet and hugely less convenient for busy teachers.

It's low cost... but that's about the only side of it that might be considered "for education".

Re: The glory days of UK IT

"Pissed away by successive governments"..

Not really, most European countries read the tea-leaves wrong and stuck with oddball personal computers when America was moving wholesale onto the PC standard. Acorn was still pushing RISCOS when it was finally bought out, and many users were still on Amigas, Atari ST's and other long-forgotten machines right into the 90's. None of that was the government's fault, so much as the obsession of UK (and European - hello Nokia) companies to chase "smart ideas" and fragmentation.

We kept trying to out smart people, when what was really needed was a consistent, affordable and compatible experience.

Re: The tube

I'm not sure it was the presence of the Tube that made the Beeb sell - just that it was the flagship educational machine. The Electron came later when Acorn realised they were missing out on the much larger market Sinclair had captured, but not only was it too late, it also managed to be incompatible and slightly underpowered when directly compared with the top selling Spectrum and C64. The absence of the Tube was not as important as the relatively limited and slower graphics capabilities.

Re: Building software is hard...

It occurs to me that building software per-se is not hard - we really do know how to do this stuff, and there are companies that excel at it.

What is hard is building anything (train tracks, nuclear power stations, ERP systems) where the project spans multiple years and has multiple, changing owners and stakeholders.

The longer a project goes on, and the more it has to accommodate tweaks and changes and budget shifts and new ideas, the more likely it is to fail. Perhaps one reason large infrastructure projects in China are finished at lower cost and significantly faster is that a project mandated by the CCP *does not change* once it has been decided upon.

No feature creep. No changed specifications. No late discovery. No new management with a "better way to do it". No revised budget.

And in particular, no client who believes that because they are a subject expert in their activity, they are also a subject expert in implementing a software system to manage it.

Re: Hahahahahaha

Note I never claimed closed source is more secure - just pointed out that opening source code up does not magically make it secure, but without question does make it easier for people to look for vulnerabilities. There is an important difference here between the idea of 'security' and 'vulnerable to attack'. I can grep "software that uses library X with newly discovered vulnerability Y" in moments on GitHub.

Equally there is nothing about software being closed source that prevents it from having sane data availability. Clinicians and insurance companies should probably make this a base requirement for any device they approve for their patients. The fact that they don't probably says more about how they think about such devices than the implementation choices of the manufacturers.

The same applies to updates - and I note that a family member with a pacemaker had regular updates on the 'high end' unit she had fitted, and almost none on the 'standard' one - this is entirely under the control of the medics (and, sigh insurance and government) that make selection decisions between devices. Whether or not it's open source is utterly irrelevant.