Posts by Tom 7 8318 publicly visible posts • joined 11 Jun 2009
Yeh, I chuck my watch away when the battery runs flat.
I'm not convinced fast charging it the answer to all electric cars usability. The London Electric Tram Company used to be able to turn around a tram in 3 minutes around 1910. I'm sure with a little thought and a few standards you could have thousands of places around the UK where you could swap in a 150 miles or so while having a lash.
The broken promises are familiar. Its far too easy in this country to avoid your responsibilities - want to dig an open cast mine, sure just make sure you re-wild it afterwards. 30 years later the re-wilding responsibility is owned by a shell company and no-one can remember who allowed this shit to happen.
Water is devastating stuff. I lived near a village called Wray which suffered a devastating flood in 1967 purely from an angry rainstorm. Where I lived we had a river ran through the bottom of our garden that shared the hilltops that fed the Wray flood and swelled from a 1' deep 10' wide stream to a raging 60' wide and 20' deep torrent of full grown trees and massive rocks in a matter of a couple of hours and re-wrote the river landscape for miles. No-one died here though, many sheep met their doom and a cat and kittens survived in car engine bay that was swept 5 miles down stream. If you've never seen anything like this its almost impossible to envisage the power and destruction water can wreak. All I can say is those people at the cinema that night were lucky a couple of rocks didnt jam in the wrong place and re-direct the flow to their detriment. Just down stream from our house the 60' wide river valley was moved a hundred yards to the left due to the bank collapsing (we think - it wasn't there in the morning) and at other points thousands of tons of rock had simply been brushed away.
I find it interesting that it was the people short-sightedly trying to save money that caused the virus to become endemic and will have cost most of them far more in the long run. Its actually still possible to eradicate the disease - indeed it would be the cheapest option in the long run - but it would take worldwide co-operation. NZ, Taiwan, Vietnam and Thailand showed they could reduce the infection rate by a factor of over 1000 in a couple of months by a proper lockdown with financial support for everyone. That would bring levels down to less than 1000 a day worldwide which should be controllable tending to 0 in the 21st C. As it is we'll need to control new variant breakouts every few months as the thing mutates and or crossbreeds until we get a vaccine that can protect against anything resembling a coronavirus.
I'd say it depends on the problem. Humans are pretty damn good at high knowledge low component optimisation. I can easily see AI being able to optimise vast networks that humans simply wont have the time to understand well enough to actually improve. Some of the best layout engineers in the world may be as an MBA - simply fucking with stuff to make themselves relevant. But this stuff is hard - 30 odd years ago I was reading a couple of PhDs a month to try and suck ideas from all aspects of circuit and chip design somewhere that knew I'd probably get results as a result. I have a feeling it would take 20 years in the industry now to get to anywhere near the sort of global view of the whole engineering surface you need to be able to see doing something here would ripple through to a little space there without having to squeeze that lot there making this track there a lot longer so needing a larger driver which means the power in this area is a little higher than wanted so the leakage will probably mean the ram there will chuck up one more error in 10**15 which will mean the software error corrector will be too slow ...
I do wonder about AI complainers some time - people saying it doesnt explain how it got to where it did. TBH people do things they cant explain to others because others dont live the way you do,
I used to investigate and try and use every algorithm and trick in the book I could code 30 or more years ago - biggest device I made was 10,000 transistors of ECL which was nice and regular. But the computational times then for some stuff was heat death level. I could get code to get within a few percent of experienced human layout for small groups but a guy I know worked on paper and if you really needed something as small as possible he was your man and he'd bring out something 10% smaller than I could do on a good day and 20% smaller than some seriously computational try every last thing and annealing several annealing algorithms. As chips get more complicated people adopt sub optimal area wise solutions because the time taken for humans and computers to do much better grows exponentially.
I'd bet certain things were sidelined simply because they couldn't provide the results fast or cheaply enough. I was designing stuff on things with megs of ram and MIPS of core - machines run by other people. I dare say some of the code could be run on GPUs with 10s of GIgs of ram in times where they are useful now. We've just got to a price and performance point where old tricks become useful again against the ever increasing volume of data and rules to crunch. As all the different layers of technology progress different things are going to pop in and out of usefulness. Humans may well pop out of it with AI and chip layout.
Something this close to the atom wide sharp bit of the pointy end should tell you pretty much exactly WHAT went wrong a small fraction of a second after it did. All power to the engineers being able to read the log files through the Niagra falls of sweat this would induce in most people. Once you've done that the WHY should be pretty clear thought the HTF do we fix it might take a couple of minutes going over the pre-written disaster recovery plan, which should include a big 'make sure this cant get in again' post mortem procedure which should explicitly exclude bean counters.
Could cause more problems than it solves. If all three cores are close to each other on the die and the error is one of the 'field type' (where lots of certain activity in a certain area of the chip causes the problem) then all three cores could fall for the same problem and provide identical incorrect results thus giving the illusion all is ok,
A small typo in the header of a chip mask set data I generated once meant the 5mm wide chip would have been 500m across.Fortunately only laughter was produced. I have been very keen on writing sanity checks for ever last thing I can think of - especially where I've fucked up before!
I had the good fortune to get my first job out of Uni working as a chip designer at BT Research Centre as it was then. I used to work on high end CAD systems, one of which cost 50 times my annual salary!!!! Large colour screen at eye level (how I miss that) keyboard and digitiser tablet with a 4 button crosshair puck. Something I fell in love with was the ability to use scribbled as commands - Z to zoom in on an area , writing it backwards to zoom out etc. You could also make your own command strokes so long as you moved over the imaginary grid in a unique way while holding down a button so each 'shape' could have 4 different results. In my line of work at the time i was an incredibly powerful tool and I've reproduced it in Javascript to some resounding 'mehs'. Even when I'm far more productive than anyone else because of the gestures no-one seems to want to use them!
When I was there >40 years ago I was told the solid core diffuses into the guide making a lens that tends to encourage a more tight 'beam which results in less spreading of the pulse. The hollow core allows for longer paths and so spreads the pulses. Like I say this was over 40 years ago and we were working at 9.6Ghz so the physics may be a lot different using modern glass pulling and better lasers.
<Sgt Major Voice>Present chips to groin.
Micturate.</Sgt Major Voice>
My local lug is probably going to jump ship - I'm hoping we're going to look into setting up a redundant array of servers on our home DMZs and DDNS and then publish the shit out of our findings.
And then extend it to Jitsi etc....
I think one of the problems with trying to work out how our solar system formed is the tendency to go for simple and assume it was a fairly even disk of dust. Given our solar system almost certainly formed from two or more supernova shock-waves crashing into each other I think a nice homogeneous disk where the sun formed before the planets is probably one of the lowest probability scenarios there is,
You may be onto something there. I know of several people who were paid by MS to not work for other companies. They dont really work for MS on anything useful - they are given a sandpit and lots of resources and a good living to keep them amuses but MS has no real use for them. I guess many other large companies do the same. I wonder if they keep them in their sandpits or let them do 'something useful' when they realise they are merely being kept unproductive on purpose.
In almost all forms of development co-operative development structures - ones that assume no malice - are orders of magnitude faster than those that assume malice. In many ways kernel development has been lucky that no-one has tried this before. The effort involved in back filling the development process with mechanisms to ensure this doesnt happen again would be considerable.
Interestingly the people who are going to suffer most are the new chip/ concept developers.
SystemD will find it much much harder to get new 'features' added. Bug fixes like the dreaded buffer overflow will probably be easy to check and approve by normal methods.
There are a group building networks based on real networks from wildlife. They've done flight control and visual from the Honey Bee and produced networks that are far far better than trained versions and 50 times faster. I am looking forward to the bit where they find the bit that does the graph optimisation things we need to know to manage all this old shit. I bet someone will try and patent it but I dare say they wont be able to cover all species with rounded corners.
There are (were?) a special breed of plugs/connectors that look and feel as if they are connected when they are not. Often the only way of telling they are connected is by the device working. I've found that some dead disk drives can often be brought back to life by putting a couple of loops of string around the connector and drive and tightening it gently by twisting a pencil round and round and sometimes the connector moves visibly and other times the drive just starts up with little if any noticeable movement. As for RF stuff I've found merely straightening the way of piece of coax approaches the connector can make it work, even though it worked the other way for years before failing. That could be because the core is fractured but I've never spent the time checking.
But in buying ICL there may very well be information that means they have to be treated as British. Fujitsu and ICL were playing together in the early 80s so there may be 40 years of communications between governments and bosses to accidentally be released into the public domain should the PR dept need some squirrels and dead cats.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
