So Skynet's main weapon will be hairballs of slightly stinky plastic thread and the occasional penis?
Suddenly, I feel less worried.
97 posts • joined 26 Oct 2011
Interesting, what have you all got against eLoran?
From my reading of it, it seems like a less jammable, nearly as accurate (with modern receivers), lower running cost navigation system that can be rolled out across the globe as needed / wanted / as our empire inevitably restores itself post-brexit.
Useful antennas currently exist down to 2cm square, and I imagine that'll gently shrink with time - so not as tiny as GPS, but GPS isn't going to stop existing, LORAN would just ride alongside as a less fragile backup for when you really care.
Educate me, downvoters.
+1 for an eLoran style ground-based option, rather than yet another sat system that'll be just as prone to jamming as the rest. Possibly not as good for guiding individual rounds in (is it?) but better at letting squaddies and assets know where they are,
Edit: Also, third parties are much more likely to chip in to help build out & run an eLoran. Nobody will give a single fuck about yet another GNSS. I'd imagine that any receivers will be for Are Boys (and emergency services) only, hence huge, power hungry and crappy, to the point where they all just use a phone instead.
That's right in the sweet spot of annoyance - Too few for any kind of volume (real plastics, cheap parts and PCBs), and too many to casually knock out in an afternoon.
SLA plastics, then? Would help explain the grotty buttons - but the clear display cover looks like a real moulded part?
While it's no great loss if the Vega+ dies on its arse, it would be a shame for all the tooling to have been made, and then just a run of fifty cycles, then in to the bin. Especially for the people whose crowdfunding paid for the mould making.
£500K doesn't get you many goes round at design, tooling, testing. Does it look like they nearly got there, or is this a product that's 90% there, and just needs a second, and possibly third 90% to be adequate?
I'm slightly surprised that the pressure dropped. The outside pressure is about 1/4 sea level (isn't it?) , and there was a sodding great hole, right at the front, with the plane presumably going at 600ish mph?
Anyone do the maths / physics for this?
Does the data recorder log cabin pressure?
(I'm not in any way saying that it wasn't appallingly windy, cold and tricky to fly...)
I've got a facebook account (under a pseudonym, natch.)
How in hell is that a good UX? Can't find things, can't dig back, can't organise. It's just a dumping ground. Photos get resampled, cropped and generally befouled. Videos likewise. Coments don't thread. Ads are poorly targeted bollocks if I disable blockers. Recommendations for 'you might like' are nonsensical babble.
It just lived off investor cash until it got big enough to sell the emperor's new ads, and now it's reached critical mass.
Horrible mess. But clearly successful. Does beating FB mean playing it at its own crazy game? Interoperating with it?
Anyone who's tried to laser cut a digestive can tell you that they're surprisingly resistant, so I reckon a huge one would be worth a try for a heat shield. Getting SpaceX to replace their boat-borne catcher's mitt with a giant cup of tea would be a fine thing, too.
Hang on - I'm a rural business, and bandwidth means I can Skype (or Zoom, or hangout, or whateerthehellcustomerwants) in to meetings without having to drag my arse across rural roads for an hour or more. 10Mbps will do, but crappy asymmetric and flaky *DSL, no chance.
It's not (all) just form-filling farmers around here. We'll even pay for service (but not necessarily what BT/OR want to charge). Fat wireless links from chrches, as long as the backhaul's ok, would seem to fit the bill.
It does sound like it's just a way to get masts without the planing hassles, though.
>It will be interesting to see a map showing 4G coverage and absence of FTTP/FTTC.
Out here in the fens - EE's coverage map is good, and BT's service is shit.
Obviously anecdotes are not data, but for me 4G at £60 per 200G trumps sub-megabit flaky ADSL. Sure, unmetered would be nice, but I'll take what I can get.
 I should stop saying this stuff, or people will sign up and my tower / backhaul will get contended to hell. I certainly don't miss the home-from-school and early evening VM slowdowns at my last place, on a notionally much faster service.
It's easy to mock, but some of us are stuck on glorified dial-up. The line here delivered a flaky 300kbps when I killed it. Fortunately I'm in a decent 4G area, so as long as I'm prepared to pay £1/G, I can have a usable service - which I need for both work, and reading ElReg...
How are BT going to deliver this?
As few FTTC boxes as possible, and hope that advances in *DSL will allow them to keep up with any increase in obligated speed?
Bite the bullet and overprovision, with some FTTP as well?
Just ignore it, say they're working as hard as they possibly can, and swallow any fines that arrive?
There are people on ebay selling Farnell's stuff, using photos and text scraped straight from the web site, with a 400% markup and 99p delivery fee, despite the fact they get free drop shipping straight from Farnell to the customer. One of them, "Top Quality Tools", has 126,000 feedback - there's clearly a (perceived) need for it. Of course, people will probably whine if they ever see the actual markup, which the ebay sellers manage to conceal effectively.
The design rules are pretty tight, though - 0.8 thou track and gap (20 micron) , I'm seeing from manufacturers.
It's the stacking of the boards through that via-laden interposer board that I find interesting - although the amount of waste from all those 'middles' seems sad.
(That's just the good photo from ifixit of the stacked interposer)
Also interesting that the BGAs are backfilled, but they haven't gone the next step and filled the whole assembly for thermal and rigidity reasons. Maybe it makes things worse with heat coming from different chips at different times?
Yeah, but you'd have to point your lens in the right direction, which is hard.
Picking the AC signal out of the mush is 'just' analogue and processing, no movement. (and blocking the DC from the sun is trivially easy).
I'm impressed that they got this much bandwidth, though. Good stuff.
Yeah, I know. It's only BT directly with whom I have a gripe - any dealing with them has, over the last 30 years, been somewhere between unsatisfactory and a shambles. I've learned my lesson.
Dealing with the weasel's tentacles, though, doesn't seem to be doomed (yet). Cellnet - tolerable. OpenReach - tolerable. EE, tolerable. Wish me luck.
"Sorry, your property is in a RED area which means no properties are planned to have access to fast broadband. Scroll down to see the options available to you."
Ooh, subsidised satellite, maybe that's better than the <250kbps over the landline.
Nope, it's subsidised installation, but with a long contract and still over £1/Gbyte, but with added latency. I'll stick with my 3G & 4G dongles. I guess that's broadband - and at least I'm not giving the BT weasels anything.
I've got to admit that I'm reasonably happy with EE's 4G->Ethernet widget. Shame it's not unlimited, but it gets the job done. Reluctant to squander precious bytes on a speedtest, but it never feels much slower than the VM 50Meg I moved away from. If you're getting crap service but have 3G or 4G coverage, it's worth a look. Port your phone number to VoIP and ditch the wire.
Oddly, I don't grumble at all when 35 (or lower) is an appropriate speed. It's the oblivious '35 is the right speed in any circumstance' that I (and others, seemingly) find vexing. I just don't understand how one can drive like that? Is the act of steering so overwhelming that there's no capacity left for altering the speed? Clearly it's not a safety thing, or they'd back off through the lower limit sections. I'm honestly baffled. Anyone who tends to do this want to pop on an anonymous mask and give me a hand?
Just on case it hasn't already been patented...
Can I have bluetooth, wifi or other short-range wireless homing between car and customer's phone, so it can actually manage the final stage of finding the customer. I don't think the mapping is yet good enough to reliably locate which house / flat / hovel is which, and it'll be annoying to have to walk half a mile to get to a car that's arrived by postcode only... Likewise, finding the autonomous car that's come to pick me up from the station, rather than the hundred other people. This doesn't work well, even with experienced taxi drivers, and is going to need a solution - and I haven't seen it discussed yet.
Having driven far too much last Sunday - is there going to be a speed control button on these things, so the elderly can set it to FUCKING DAWDLE at 35 on country roads, and then FUCKING SPEED at 40 through 20-limited residential areas?
If not, I can't see them getting a lot of use.
Automatic transmissions piss me off - 'like giving your drunk friend in the passenger seat the clutch pedal and gear lever'. I look forward to handing my drunk friend full control, oh yes.
Maybe Keysight will ditch the daft name and logo, and start making HP gear again. You know, the (mostly) good stuff they used to. Among test gear people at least, the HP name carries a lot of fondness.
Agilent / Genital was a crazy rebranding, but maybe it made sense to try to decouple from the systems borg HP was turning into?
“The Holmes-Ginsbook Device” , Isaac Asimov, seemingly only released in the collection Opus 100, is a fine short story about two devoted scientists and their invention of an improved document reading device, as the current state of the art made it too difficult to look at girls and smoke.
I can't remember many of the other stories in that book, but this one has stuck with me.
Considered, and put it on the 'Edge case, hard, solve later, got a product / demo to ship ASAP' pile.
Self-driving cars strike me as the hardest, most edge-case-ridden, timescale-pressured project I can think of. If I was younger, I'd be trying to land a job, I think. Sounds like fun.
The article also mentioned training up your vision system against LiDAR data - that makes some sense. If that's a planned route to ditching the LiDAR, then nobody really cares how expensive they are in low volume, as long as they work and can be bought. 6 months delay will concentrate the mind, though. Maybe people will design the systems while they wait, rather than just knocking out some code and hoping it compiles?
Reflections add all sorts of fun. Things appear to be travelling at twice the speed and angle. I'm guessing that LiDAR tends to stop at first return signal to reduce this, but that'll also make it more susceptible to other interferences.
Dammit, this stuff is hard to get right. Still, good enough will be fine, eh?
However, first one to reduce the BoM by 90% will be happy - able to shift volume crippling the competitors rather than themselves. No matter how deep your pockets, it's nice to have a gap between your costs and your price, to pay off the development / fund more of the same / spend on beer.
I'm not sure that all of these LiDAR manufacturers will still be in the game in 5 years.
Sounds more like the bespoke plastic optics and complex mixed-signal ASIC(s) that they're relying on to bring the cost down don't exist yet, so that little module is rammed with general purpose optics, analogue and FPGAs, at eyewatering BoM and manufacturing cost.
Once the design's stable and they know what they want to make, then the lead time for all of the above begins - and it's not short, and no amount of wishing will make it much shorter.
(Veteran of a sorta-similar project)
I also wonder how jammable / spoofable they are.?
LIDAR's spiffy and all, but I'd really hope there are other sensors involved, and then all you have to do is decide what's lying. In real time, with the risk of expensivel consequences. Easy peasy.
(Old cartoons with two guys carrying huge sheets of plate glass also spring to mind)
Return of the MS Home server? Maintained, headless, running as backupp, streamer, IoT hub, firewall, AP?
Most of the parts are in place, and maybe customers would be more amenable to paying a subscruiption to keep that fresh, in a way that they won't for OSes?
I'd rather have a FreeNAS or whatever box under the stairs, but I can see an MS offering 'to keep you, your kids and your IoT crap safe from the big bad internet'...
Would people be more likely to buy stuff that's had a bit of third-party pen testing and promises updates for <n> years?
Enforcing those updates is a different kettle of fish - but if it was possible to buy non-shit cameras and widgets, I might be tempted. At the moment, it's all shit, and maybe it's time to change before the net just turns into a swamp of DDOS and takeovers?
(Ah, AC above covered much of this obvious stuff - but some way to find premium products that actually work - surely that's worth something?)
I was wondering how susceptible TBytes of NAND SSD would be, given how much effort it is to make them remember stuff at ground level. This interesting bunch of slides form ESA
if I followed it right from slide to slide, suggests that, in LEO, and with a 5mm aluminium shield, NAND should be good to go. Maybe LEO, shielding, a short life and a swarm of the damn things for redundancy means that failing hardware won't be the limiting factor here. Sanity, on the other hand...
Sure, I could buy a cheap Android for $100 each year, or I could buy a featurephone for £20, replace it once or twice a decade. Also, with a featurephone, I don't have to wrestle with a fuckawful touchscreen, risible battery life, and I can have something water, dust and drop proof. Horses, courses, etc.
These numbers are really quite high for a single die. High capacity nand chips tend to have a bunch of (back-ground, so they're waffer-thin) die stacked on top of each other, and a bit of cunning electronics to make it look like one honkin' great device, all in a bit of black plastic with some balls on the bottom. Tablets will then often have a (small) number of these packages to reach the numbers they want.
reckons a stack of 32 layers is 4um high, and you can stack stacks more or less forever.
I look forward to a PLCC-thickness package, with the middle 2mm made of 16000 layers of storage.
Sure, it'll take a while to fab, format and test, but that's a lot of data.
Anything to stop us growing a few dozen layers of NAND over a microcontroller, to save dicking about with external storage? Maybe a layer or ten of DRAM, too? Surely that's much the same process...
These aren't stacked die, they're grown layers on the same die, if I'm understanding right.
Thin, thin layers. What I'm wondering is - what's the limit? Do earlier layers diffuse a bit, each time through the oven? It can't be that the layers interfere - 48 layers is close to 'n'...
How long does it take to test these once they're fabbed? How many times round do you have to go, erasing, writing, checking and building the bad block list?
Even if you're rocking along flat out, 32GBytes is a lot of space to work through.
(I miss the bad sector lists that used to be printed on HD labels. Maybe if they wrote really, really small.)
Is it possible to see any of this structure in a die shot, or does it just look like a regular NAND device, and you can't see under the top layer?
Ah, some die shots back here. Mildly interesting, want more!
also, 4.3Gb/mm^2 . Lordy.
I'd sort of assumed a decent speed - say, 100mph straight-line. Nowhere near easyjet speeds, but with a 5 minute to strap-in, rather than 3 hours.
That would be enough for me to want to / be able to live outside the megacity, if it wasn't a daily commute.
How fast do we think these embiggened quadcopters can go? (and, come to think of it, how come they're so much more compelling than regular helicopters? If they are, why do helicopters still exist?)
Can he bribe Apple to start building stuff locally, by allowing them to repatriate their massive cash stash with a temptingly low tax rate?
Or, of course, threaten them with the opposite and see what happens? Do Apple actually want the cash in the USA?
I'm so glad I'm insignificant in all this. Head down, keep making stuff, see what happens.
Ah, I'll put it back since it's got a reply, looks weird otherwise. Wasn't convinced it added much apart from my general uneasiness with arbitrary companies having endpoints in my premises.
As I understand it, IPV6's addressing will allow a direct link between the IoT widget and the mothership. That feels more or less the opposite from what I want. Guaranteeing that the widget is landfill as soon as either the mothership or my net feed hiccups, and unfettered comms between the widget and the mothership. I can't see how to firewall this, the widget is either going to be connected or it isn't going to work. Do I just hope it's benign?
Nope. Still not keen on this interpretation of IoT.
Biting the hand that feeds IT © 1998–2019