Re: Gooey Coating
And binoculars, not even much used. Time degrades the rubber, not my hand grease, I'm sure.... :-(
Definitely not used for viewing any "speciality" content, before you start....
382 publicly visible posts • joined 20 Apr 2007
Nike and Nike Hercules were primarily anti-aircraft missiles, radar-guided by humans looking at screens and twiddling knobs to guide the missile in, each battery of 4 launchers could control only one missile at once; the intent was to detonate 30-50 miles off the coast of San Francisco for example, when the bombers were 3 or 4 minutes out.
Sure, they were nuclear tipped, to take out multiple high altitude bombers - same characteristics, >70,000ft, >0.9mach as our V-force - without having to be too accurate. But they weren't designed to stop ICMBs.
I can believe that in extremis launching some to a predetermined place and detonating would have been considered as a tactic if ICBMs were incoming but it wasn't their design goal, that was definitely centred on guiding to the radar return from relatively slow planes.
Yeah well it's not the chipset covertly doing something via a sekrit backdoor, without the CPU et al knowing, like Intel's secret onboard management chip with its own MINIX OS et al, it's the software that gets the assistance data for Qualcomm's GPS chipset. All GPS systems do stuff along these lines, sending lists of audible WiFi and cell towers to a server to receive in return relevant ephemeris and almanac data (describing exactly where the satellites are), instead of downloading/decoding all that data at 50 bits per second (sic) from the satellite signals like a standalone old-skool GPS would have to.
The conflation of "phones with chipset X do this" and "chipset X does this" ain't quite valid; it's got to be software.
Methinks the real issue, if any, is that the comms is not encrypted, rather than that it goes to a Qualcomm server /per se/.
Hah, once at a show I was poking about at a new thrilling NeXT machine, with its MACH microkernel and all that interesting stuff. There was a normal non-privileged shell open for having a look around.
There were many processes with PID -1 and interesting names, obviously the privileged microdaemons that did all the kernel work. I couldn't help myself: I typed "kill -9 -1" and nothing ever happened ever again.
Um, I think you have me conflated with someone else re: Archimedes demos; I didn't do a lot there.
I was working with the ARC folks on ARX when Arthur/RISCOS was arriving, and on RISCiX (BSD4.3) when later RISCOS/Archimedes things were happening.
But I did write Beeb 6502 ADFS and took DFS from 0.9 to 1.2 which more-or-less worked correctly. And the ARM ISA test suite, used to verify the silicon, which a couple of decades on, ARM was foolishly distributing binaries to ISA licencees with all the swears on failure left in. No-one had ever made an ARM chip whose shifter was fsckd until Intel made their own StrongARM/XScale machines from scratch! They were, as I understand it, offended, and ARM mgt sent a rocket about unprofessional code blah blah... I was long gone, never joined ARM.
And Atom Snapper and Atom Invaders... :-)
Heh. IMHO one of the reasons ARX failed to be anything like ready in time, was Modula-2.
"Extra-linguistical objects" - the M2 world and the compiler folk rejected the idea that a program was going to run on actual hardware.
I was writing the hard disc file system, so I needed to lay out the on-disk structures in memory, then issue a read or write to actual hardware.
There was real dogma about finding out, programatically, the address in memory of an object, or the offset into it. New compilers would change the size of my nicely-organised disk-block object from 256 bytes to 257. They even said things like "you don't need to know the units of sizeof() because you only ever feed it back into other routines that know the units". Eventually we got the academic idiots to understand that this was actual physical hardware, and the compiler doesn't get to change the block size from 256 to 257 bytes (or even 2048 to 2056 if sized in bits!) and they came up with the label "extra-linguistical objects" and a separate set of built-in functions that used units of bytes. Like every sane environment has. Sigh.
I mean, if a program cannot change in any way external objects, then you can optimise it away to "main() {return;}" can't you.
Separately, opaque types. This was because there was no initialised static data at all, IIRC. I mean you could not even set a static int to zero, nor assume anything about its value at all. So every module that held state had a "opaque FooData Foo::init()" member and you pass the opaque back to every single function call. When it came to making an efficient OS executable they were talking about ideas of running it up to a certain point where all the init()s had been called, then snapshotting the memory state and building that into the image, into the BSS section, outside of anything the language could do. FFS.
Such bodges required to make a sensible executable image.
Anyway, you'll have gathered that we hated it.
I used to work somewhere that had one of the first HP/PA HPUX systems (I think, mid 1990s) outside the USA. They were very pleased with their automatic voltage sensing auto-switching PSU that just worked with 110V or 240V or whatever, supposedly. We were in Barrington Hall, in a village. At least the place had 3-phase supply.
We had brownouts.... we also had power cuts and the supply was delicate enough that procedure was to turn off everything we could when it went dark; sysadmins would then come round and turn it on one by one so that the startup load of all the discs didn't happen all at once and trip it out again.
But in some shapes of brownout, the HP/PA machine's clever PSU auto-switched to 110V. Great. But it was very slow at switching back to 240V when the mains came back up, and failed with smoke. Every time the project was delayed a week whilst they shipped another PSU; after some months they sent a better PSU without the stupid in it, and all was well.
It's not just dates; I had a customer saying "your data is useless" when they had opened the CSV in Excel, just to
check it wasn't corrupted, AND SAVED IT BACK. Thus corrupting it....
All the 64-bit integers were now double floats. Thus removing the significance.
These were 32 bits of latitude, and 32 of longitude, in 1/2^32 units of a full circle.
Handily that resolution is about a centimetre, we were very pleased with that encoding scheme for mapping.....
I learnt long ago that trying to do two things in parallel in tabs is *dangerous*.
Bit of a windfall, not sure what to do, so of course I wanted to open one fixed-rate fixed-term saving account, and one variable-rate to spread the risk of inflation.
Filling in many tedious details in two windows next to each other. Disaster; neither opened successfully, but at least neither debited my money either. (TBH it could have been so long ago that "apply online" meant "fill in a form online and we send you forms, you sign and return a cheque" - I can't remember)
At least Amazon seems to get that right these days.... I'd still be scared to try it with real finance.
"Delete" only deletes an entry in a directory (folder) and decrements the reference count on the actual bag-o-bits file. When the file has no references and no open handles on it, *then* the file is deleted.
Normally, in the simple case, of course "rm foo" just removes folder entry foo, and also the bag-o-bits underlying it at the same time....
If only the bosses had said they were going to abstract from the database to make training examples, then the workers would have known why the guidelines on cleanliness of made-up names. That's the root cause of the problem right there.
Reminds me of test software I wrote for the very first ARM CPUs at (troubled Cambridge micro-maker) Acorn in the 1980s. Testing that the silicon actually did the ISA, as it were. It was for use only within our small friendly team. So one of the fail messages was "Shifter fucked". But later ARM span out, became a monster success, and was shipping that test suite as binaries. You can guess the rest.
I have JS enabled and a blacklist of sites where it's disabled (in $Chrome). The disabled sites are the more or less the newspapers that I read daily. I experimented with globally-disabling JS and having a whitelist instead; it was about the same size so I didn't take that plan further. But it was soooo close, very much on the cusp.
I wonder how long it'll be before disabled JS is popular enough - or the default on enough browsers - that we'll see nagware like the "you appear to have an Ad Blocker..." saying "Please enable JaveScript to continue..." :-(
Civilian GPS really is 1023 chips/cycle so the PRN shift-and-feedback 10-bit register never has to go through zero, 'cos if it did it would stay on zero. Repeating every millisecond (nominally, subject to doppler and clock errors).
About the mil signal, the really long cycle time is for security only - makes it really hard to find unless you have pre-loaded tracking data somehow, from that encrypted almanac as you say. But that aspect doesn't add at all to accuracy: only a higher chipping rate can do that, which as you say is 10x the civvy version. But only 10x, so each chip is ~30m. You might get 30cm correlation data out of that.
The ionospheric correction - the naive algorithmic one (ramps up linearly from 0600 local to noon, ramps down from 1800 local to midnight, times cosine something I think) in civilian GPS standards docs anyway - is under 10m, ISTR it being 5m to 7m max usually. So the naive model is good for a system whose overall accuracy is 5-10m anyway. Interesting that of course the mil system needs better Iono corrections from the ephemerides (I guess) 'cos its native accuracy is higher.
But it can never be reliably that accurate. I had a GPS antenna on the windowsill of an office about 15m up. It saw half the sky OK. I generally got accurate positions, being of course the position of the antenna not my computer connected to it. But when it rained, a sloping roof across the road turned into a radio reflector and I could "see" much of the other half of the sky too. Great! you might think, but it saw the new sky with extra path length, since the new signals went across the street and back extra. So my positions were solved for about 10m away, in the middle of the road. Ain't nuthin' you can do to prevent that sort of situation short of rejecting some signals in the solver - but which ones? - whatever your chipping rate and correlation accuracy, it can still be fooled by reflections.
You're conflating turning off selective availability (SA), which they did 'cos of various gulf wars, and making access to the military signals available. They didn't do the latter.
SA dithered the civilian signals to cause a wandering inaccuracy of several hundred meters. It is defeated by using a fixed ground station to work out the current offset, and delivering the difference data to receivers - this is Differential GPS. So people were working around it anyway.
With SA turned off as it has now been for decades, the civilian signals are still only accurate to a few metres 'cos the chipping rate of the PRN code (orthogonal CDMA code) is 1023 chips per millisecond, so each chip is about 300 metres long. Curve fitting to multiple correlation results gives timing to fractional chips but it's inherently noisy - so you might get actual results accurate to 10s of meters (30ths of microseconds). Solving with many satellites reduces the combined error, sure, but it's stlll meters not centimeters.
AIUI the military signal has a faster chipping rate with longer PRN sequences and so it's inherently more accurate because the basic unit of "knowing when the signal is" which you have if you're managing to track the signal at all, can be not about a microsecond = 300m, but a fraction of that. It's a 10th of that - 30m - in the old versions, and it's not in wikipedia for the latest actually secret stuff. I only ever worked on civilian receivers, so had no need to look at the mil stuff.
The Shuttle SRBs nearly had a "become safe after lighting" feature which involved explosive detachment of the top cap of both SRBs, so they both became like one of those new wanky double-ended lightsabers. With no net thrust, was the idea.
This was in case only one of them ignited, to prevent making a catherine wheel out of the whole affair... it wasn't implemented; possibly the top-end flame thrower would have killed everyone anyway, and they got SRB ignition reliable enough that they always lit both, or neither.
I read in a public, academic book, that when Skylab launched, NASA was keen to see what the wrongly/party deployed solar panels, and associated damage, looked like. The NRO had secret spy satellites, of the kind which drop a film cartridge when it's been filled, which NASA were not allowed to know about, nor anyone else really. Apparently NASA had a visit from sunglasses and dark suits, who said "Your top 4 engineers on this, in here, now; no-one else. Here are some photographs. You may look at them for 15 minutes. They do not leave this room. We were never here, you never saw this." And it helped them decide that fixing it on-orbit was feasible, and what they needed to take and do. According to one of those NASA engineers.
Problem might be there are few extensive texts in suitable form for input, which were written by Asians or Africans about the white men, nor by woman of any kind, which reflect their attitudes to the cruel imperious selfish arrogant white European male, no? Just choosing English for the language used puts that bias on the whole thing.
Can't see an easy solution, but maybe it could be improved by pre-biasing the corpus the other way.
It's legal according to US law for the US to use StingRay in non-US places; doesn't mean it's legal according to the law of those places. What it means is that the US government won't act to prevent itself using such tech outside the US.
Remember, according to US law, US law applies over the entire universe. Really.
As there are nerds here, I'll ask: I've never managed to work out why "we" consistently used V12s the "right" way up, and the Luftwaffe consistently - so it seems - used inverted V12s. Are there advantages to inverted beyond the view from the cockpit - but then you need longer undercarriage; I'd have thought having a single sump is better, but OTOH do you get more reliable valve lubrication if it's all inverted? But what prevents oil pooling in the underside of the pistons besides their violent motion?
The leak may not even be data, it could be as small as a timing change from the spec-ex path taking longer to fault, or not, allowing the attacker to probe the kernel space for not/valid pages - so defeating to an extent the kernel memory-map randomisation? Worse might be a spec-ex branch on spec-read secret data which affects timing similarly, without directly exposing the data itself.
The sentence in the article is a bit mangled, do you mean the muscle cross-sectional area decreased to about 80% of what it was originally, the range being 86% in the best case and 72% in the worst case?
Or if it decreased by 86%-72% does that mean only 14% to 28% remained (which seems terrifying)?
In Cambridge it's a Court not a Quad; Quads are from The Other Place ;-)
But yes it's a great photo, so of its time.
Re the pic on the steps of 4a Market Hill, L-R David Johnson-Davies ("DJD"), Hermann, Chris Curry, Roger (now Sophie) Wilson, Nick Toop, holding an Acorn Atom. 4a MH was never an ARM office; only Acorn.
The Turing Room in King's is the College Computer Room as I understand it; named after the great man. But I don't think there's any implication that he lived in that basement room, it's a cellar. As a Fellow he might well have lived in the rooms above, though.
I thought the licence those big players have ("foundry") lets them re-implement the hardware as they like, but absolutely not fork the instruction set and software interface generally (ISA, eg. cache behaviour) to make sure code portability can never be sacrificed. Intel IIRC learnt this when they acquired DEC and its Xscale designs, and tried to fiddle with the ISA, and found they couldn't because the ISA is copyrighted in some way separate from the IP licence.
WTF because this is a complete surprise. I could believe it's as simple as SB having a huge mound of cash and looking for the best possible investment, as ARM is starting to work in servers too, and the IoT bubble.
This does seem separate from the protection afforded to common-carrier-like websites in that law, in that this is abusive readers/consumers of the site, who cause no change to its state -- not abusive speech published via/though/on the site. Tricky.
It does suggest that reader accounts should strongly be vetted, in such cases, and contacts should be only through the site so that they are logged in an evidence-worthy fashion. Otherwise it's just like telling Facebook everything with entirely public access.