Posts by Dunstan Vavasour 433 publicly visible posts • joined 14 May 2007
If you look at nVidia's forecasts, and the new GPU production facilities being built by TSMC it all points in one direction: our energy efficiency savings being cancelled out by endless GPU barns so that the Powerpoint jockeys can bore on about how Aye Eye is, like, shifting the paradigm of everything.
We should follow Ireland's example and start turning applications for these latter-day gas guzzlers. IMHO.
Autocomplete on steroids is a pretty accurate summary.
Parent commentard's observations are also mine - LLMs confidently produce absolute shite when asked about something specific, will sometimes do a reverse ferret when challenged or, as Lee says, double down producing a pyramid of piffle.
When asked to produce a piece of code, what comes back is typically broken: the main benefit over an incompetent human developer is that it screws it up faster and more cheaply and doesn't argue about what you said in the first place. Try asking ChatGPT for something in OpenSCAD - it won't say it's beyond its capabilities it will just produce poorly parameterised code that (at best) is a broken object. As described, it doesn't care about solving the problem just about producing a plausible looking reply.
And this isn't a harmless bit of nonsense, it is:
a) draining investment out of productive economic activity
b) crowding substantive issues out of political and management discourse, and worst of all
c) an energy guzzling catastrophe, reversing years of savings from energy efficient computing
Always my beef with the new-improved...
I'm not concerned with how it works when new, I'm concerned with all of the failure modes over time. BUT, if the concrete coming out is chemically the same, or similar, to the stuff we get from big carbon producing kilns, I'm all for it.
And yet, I wonder if humanity has leapt into overconsumption of concrete, as it has with plastics, and the better solution is to stop using the stuff.
GEC was quite the outlier with its own cocktail of disfunction.
Credit control was hugely damaging - the rule was that suppliers weren't paid until they put you on stop and you needed something. So projects would grind to a halt because components hadn't turned up. The opening position with any new supplier was "we'll require six months credit".
Acquisitions would be bled dry and then shut down. I remember an early 1980s robotics company that Weinstock bought, moved from Watford to Rugby, then starved of development cash until its products were old-hat, at which point it was shut down.
The end result was the infamous cash mountain which meant that GEC became a takeover target. The good thing was that after one junk-bond financed takeover scare, Weinstock panicked and merged swathes of the company with Alstom, at which point some of the financial fuckwittery lifted.
This was all before the Harvard MBA/McKinsey types got into their stride and started hollowing out business after business. And so now we have long-view Chinese companies that make long term business decision and, when they pay off, the MBAs try and put up trade barriers rather than relearning how to run engineering businesses.
Not a new sentiment - Harry Truman used to refer to Chaing Kai Shek as "Cash my cheque". But that was back when Taiwan had China's seat in the UN Security Council and it wasn't clear Mao Ze Dong's regime would last.
Now, if all the Harvard MBAs and McKinsey consultant's hadn't recommended the offshoring of chip manufacture, the world would be different. But then they wouldn't have been able to charge those juicy fees. It really grinds my gears when bean counters sacrifice supply chain security for cheapness then whinge that their supply chains aren't secure.
I remember in the 1990s going down to a Novell office to see their take on a SGML based word processor. Their internal markup approach lent itself very well to morphing into generalised markup (I'd always used WP with reveal codes on up to that point) and when pointed at the right DTD and stylesheet it gave you WYSIWIG SGML.
It never saw the light of day, or if it did then it wasn't properly promoted. The world wasn't interested in proper document management when they could just produce great big hairballs of MS-Word.
Quite simply, anything developed for the public sector should be public code. Councils do about the same thing as each other. Hospitals have the same IT requirements.
Basic public service software: general ledger for a council; appointments booking system for hospitals --- should be developed by central government and made freely available, and UK public sector bodies should have *&^% good reasons for not using it. Councils are not that complicated, the legislation is the same for everyone, and to fuck up a basic ledger system is so inadequate.
https://publiccode.eu/en/
Once again, the problem is stated as being data going to a Chinese company, when the real problem is the data going to *any* company.
I'd love to see this article rewritten without the "China" angle, stating the real concern: user data is being scraped from UK government websites and shared with commercial third parties.
I've written to my MP about this several times. At its core, the Horizon scandal wasn't about mistakes, it was about wrongdoing. It was about Fujitsu management being told by their development teams "Yes, this could happen" and then swearing on oath that it couldn't.
Miscarriages of justice work both ways: getting away with flagrant wrongdoing, whether at an individual or corporate level, is as obnoxious as banging up an innocent person.
"I actually like the ribbon UI. It's a much more efficient use of space that allows significantly more features to be exposed without the need for multiple levels of dialog boxes."
Most screens are much wider than they are high. So taking away a whole load of screen height for the ribbon, which cannot be moved to the side, is perverse.
System reliability matters when you have monolithic applications. Once you build the resilience into the application layer, system reliability doesn't need to have as many nines after the point.
Now, that's not to say that distributed apps with distributed data is a silver bullet - you still need to do grown up failure mode analysis - but mainframes aren't your only path to all those nines.
I'd be amazed if they intend to power them with coal. They've got the manufacturing costs of both PV panels and batteries down to the point where any high energy business is now built with renewables nearby to avoid the cost of mining and transporting any sort of fossil fuels.
The only reason China is building coal generation now is because they can't build out renewables fast enough. For this sort of planned migration the renewable generation will be a pre-req.
A long time ago Sun Microsystems' datacentre-in-a-container had "fail in place" designed in. You didn't replace failed components, you just blacklisted them and carried on working around them.
If I were running AWS I'd expect something similar - when components fail you take them out of service forever and operate the rig round them. If at 5 years you have 10% of failed components then it might well be worth operating them another year.
I think it goes far beyond the auto makers, but you're exactly right. The principle of the Method of Constraints (usually refered to as Just in Time) is that, to avoid carrying a lot of work-in-progress, you have components and sub-assemblies become ready just at the point where they will be used.
Decades of "value engineering" where bean counters have:
- minimised buffer stocks
- screwed down suppliers on price
- assumed that supplies can be assured by having a contract in place
have got us to where we are now. When you outsource and de-skill, that skill is effectively gone forever. And where you choose to be dependant on a supplier, your interest has to go beyond the contract you have with them.
Trouble is that with the long supply chains in almost every industry, no country feels as if it does science and engineering. But a fair proportion of the worlds airliners have engines made in Derby, most of the FIA motor racing teams are based here, and we devised a cheap and cheerful COVID vaccine for the world.
Which all begs the question, why would you use this rather than just writing Terraform?
And I suspect the answer is that the PHB has heard of Microsoft and hasn't heard of Hashicorp. We all know the type, claim they're "risk averse" as an excuse for running hardware and software that's near or past its end of support. Believe that nobody got fired for buying Microsoft (or IBM). When the head of development comes to him for budget, he'll go over the proposal and believe he's adding value by insisting they use Bicep rather than Terraform.
At the same time there will be analyst charts showing a tick in a box for AWS with CloudFormation and a tentative tick in the same box for Bicep. The fact that most developers don't use either doesn't stop it being a "feature gap" for Azure.
"Both 737 and 747 were 1960s engineering that formed the platform for continued update and improvement for decades and outlasted their Lockheed and McDonnell-Douglas rivals. In the 1960s engineers at Boeing were able to come up with such robust designs that were solidly businesslike and pragmatic and out and out market leaders."
I think it's far more that, despite there being not a single common component with the original aircraft, they still have grandfather rights for certification. From top to bottom, these aircraft types have design choices that wouldn't be permissible were they being certified from scratch: for example, the passengers in the nose of a 747 have only one escape route, backwards. Other types such as the A320, have grandfather rights too, but they don't go back to the 1960s.
^
|
This.
It's about power, about weilding that power and about subjugating those who don't comply. Get the big lie over in the title: "Inclusive Language" is about excluding those who won't play. We have devised Big Brother outside the state: we have Newspeak, we have thought police and we see groups engaged in the two minutes' hate, directed at the bête noire (oops) of the day.
Sometimes it seems like a parlour game ("Am I still allowed to say <insert word or phrase>") until the mob turns nasty, and you are deemed an enemy of tolerance. And unless it is constantly pushed back against, they advance through their own constant micro-aggressions.
I just hope the brand doesn't get sold off and slapped on tat. Thinking of "Polariod", which was slapped on all sorts of cheapo electronics before it was bought by The Impossible Project and put on their worthy range of instant cameras.
But then as long as Olympus are making medical optical instruments they'll probably keep the name. My appreciation of the brand was further enhanced when one of their instruments was used to remove a pre-cancerous polyp from my bowel the year before last.
Exactly this. If there was a discrepancy, the franchisee *had* to accept it before they could start trading the following day. The only alternative was to stop trading. And because they'd clicked on "I accept this reconciliation", their case for having the discrepancy investigated was lost.
We all remember the "cold fusion" story from a couple of decades ago - some physicist couldn't explain where the energy in an experiment had come from, so they decided there must have been a nuclear fusion reaction at room temperature. This was similar: there had been no daily reconciliation before Horizon, so when the Post Office introduced it and found discrepancies, they deemed that the only cause could be fraud.
Most of the settlement from the Post Office has been consumed in legal costs. Costs which were incurred because of their and Fujitsu's intransigence and alleged perjury. Without doubt this is the worst miscarriage of justice this century.
Azure is where they see their future income stream.
The competition between cloud providers is healthy enough that Azure needs to move towards the customers and developers, they can't strongarm customers towards running their cloud apps on Windows (however much MS Classic would have liked to do so).
So the better they can make the GNU/Linux on Azure experience, the more customers will choose to run their clouds on Azure.
This is all very well, but unless the handoff between the ISP and BT-OR is fixed nothing will change.
My village was blighted for two months by a dodgy linecard in the cabinet. More than half the properties had an outage at some point, and they all had a visit from BT-OR to check their cabling. I don't know how much these cost the ISP, but it was clear that OR were filling their boots instead of fixing an obvious common pattern (the faults from this cabinet had spiked by 1000s of percent). Many of us were unable to work from home, and had to travel to offices, even the vicar was offline from parish emails. But this was secondary to BT-OR being able to send out newly trained repair men and charge ISPs.
It's clear from BT Retail's advertising that this isn't going to change: if your broadband is offline they know it won't get promptly fixed, so they'll send out a 4G hotspot.
Well I've been following the Ad Contrarian blog for several years, and he has been calling out the fraud in online advertising for all that time.
The root cause is that the only source of metrics available is from the ad brokers themselves, so nobody is independently auditing the page impressions and click-throughs. And nobody dare say that the emperor's naked, and that the funding model for half the internet is based on fraudulent counting.
http://adcontrarian.blogspot.com/
How often is it the copper final mile that's the limiting factor on domestic BB?
As long as we have ISPs racing to the bottom of the "unlimited 80Mb/s for just £XX per month" we'll have congestion, packet loss, latency, not to mention crap service when faults develop from ISPs who get the runaround from BTOR and just pass it on to their customers.
FTTP may be a nice totemic aspiration, but is a waste of time if the ISPs can't up their game.
They're diversifying, but the core product can be thought of as grep on steroids.
There are shiny product pitches, but basically it collects logfiles from, or runs scripts on anything, chucks it into an index cluster of brobdingnagian proportions and then greps your amalgamated data really really fast. Main use cases are IT Ops and security analytics where the swiss army knife approach to schema-on-the-fly means you can correlate anything with anything.
It's proprietary, licensed in gigabytes per day of indexed data. Lots of places index terrabytes per day: they pay lots of money in licensing and get lots of value out of it by being able to fix stuff before it breaks and costs money.
[Disclaimer, I've got a load of Splunk certs and think it's actually pretty cool]
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
