Posts by vtcodger 2029 publicly visible posts • joined 13 Sep 2017
Pascal: I don't really understand blockchain well enough to evaluate it. But based on the long history of complaints of processing delays and the high transaction costs we're told are associated with cryptocurrency transactions, I think it likely there really are problems.
I think for a digital currency to work, it has to have VERY low transaction costs in addition to being convenient and secure. I don't think India or anyone else knows how to do any of those things, much less all of them simultaneously.
As for blockchain, crypto, NFT. The whole thing reminds me a lot of the Albanian pyramid schemes of three decades ago that managed to pretty much demolish the economy of what was (and remains) one of the most primitive and undeveloped countries in Europe. https://www.imf.org/external/pubs/ft/fandd/2000/03/jarvis.htm
A bad idea dubiously well implemented.
There is (or used to be) something called COTS (Commercial Off The Shelf) procurement that allowed the military to buy some stuff at the local stores sometimes. But the rules were and probably still are complex. And how do you keep Sgt Jones from buying his batallion's paperclips and toilet paper from his brother-in-law who marks supplies for the government up 500%? It's difficult.
Another issue is things like durability and temperature range. Back near the dawn of time when the US required most able bodied young men to do some time in the military I played Company Clerk for a US Army Reserve unit. The Company Clerk then and probably still today was a sort of junior assistant office manager. Did filing. Prepared the Morning Report of significant events. Ran around and found people. Got folks forms for requesting vacation and such. Made sure the completed forms got to where they needed to go. Handled the mail, etc,etc,etc. There is nothing the clerk did back then that couldn't be handled by a Z80 CPU, a few hundred K of memory, Word Perfect and maybe a spreadsheet like Lotus-123. I doubt the job has changed much. BUT if the company deployed to the field their computer might have to boot and run in desert heat (probably COTS would do that) or in a subfreezing tent. (Maybe today's low end stuff will. 20th century OTS hardware mostly wasn't fond of low temps).
Bottom line -- it's complicated.
"I have sneaking suspicion that what costs an ordinary person $100, the american military pays 10 - 100 times more for."
Decades ago, when the world was young(er), I chanced to be a fly on the wall listening to a conversation between a group of mid level managers at a large defense contractor discussing whether they could manufacture a single -- quantity one -- aircraft qualified toilet seat for some ridiculous price. ($3000?) Their conclusion. They couldn't. Not with all the proper paperwork and required testing and reviews.
On the other hand, you don't want some complex hardware unit failing in use because an overzealous assembler put it together with hardware store Loktite when the stockroom was out of the Mil-Spec qualified stuff. And yes, I actually saw that happen. Took them all morning to get the lid off that box.
So yes, the military pays a lot for stuff. But it's not ALL due to stupidity or laziness.
In the late 1990s and early 2000s before I switched to Linux I used Windows 95 plus a few service packs ... well, maybe more than a few ... 25 or so as I recall. It ran fine. Really. Better than Windows 98 which crashed a lot, was prone to trash the Registry, and often refused to shut down. I had to reboot WIn95 every few weeks. And it didn't support USB -- which wasn't important as USB back then rarely worked on any OS. I'd still be using it if (largely dubiously necessary) changes to application software hadn't made keeping Win95 impractical. In point of fact, from a user viewpoint, Win 95 was just fine. Office workers could use it. Scientists could use it. Six year olds who were still a bit hazy on that alphabet thing could use it. What more do you need?
Of course, Win95 wasn't really an appropriate OS for a server in most cases. But why the heck anyone would want a Windows server when Unix is so much less obtuse has always eluded me.
Sounds utopian to me. but maybe it can be made to work. At least to some extent. I can't think why my doctor's office, for example, wouldn't make the effort to play nicely with my Pod.
But I don't see how this solves the problem of Google et.al spying on every living creature on the planet. It's what they do. Why would they stop?
And I don't think Pods are going to make financial activity on the Internet any less risky. It isn't safe now. It probably won't be safe if done with Pods -- unless we make unauthorized access to Pods a felony and convince 195 countries to jail anyone who attempts to access a Pod (other than their own) without permission.
This whole thing seems a bit surreal. Seems to me like I'm being asked to believe that the FCC doesn't know how to allocate frequencies and/or that cell phone companies don't know how to test their transmitters for spurious emissions even when they are told what band(s) to stay out of and/or that RA designers can't/don't know how to design bandpass filters and don't have some way to distinguish their return signals from interference as, for example, from the RAs in nearby aircraft.
I don't think that I believe any of those things.
What am I missing?
In the US there are laws that try to discourage dealing in "securities" backed only by "The wide blue sky of Kansas" Those blue sky laws actually are enforced at times as are laws requiring some amount of truth in public statements. It'll be recalled that Elon Musk ran afoul of those. Heck, the US has even been known to put popular celebrities that do particularly outrageous things in jail for a few years -- e.g. Martha Stewart. So yes, Well Street may be a casino. But it's a regulated casino that tries to keep the actual odds in favor of the house close to the apparent odds.
It's hard to see how cryptocurrency fits into that picture.
I'd be more impressed with block chain if its supporters could demonstrate even one beneficial use. You'd think, for example, that block chain could somehow be used to conveniently authenticate Covid vaccination or for website login. But somehow that hasn't happened. So I think probably crypto's critics are correct. It probably really is a hive of ill considered, resource wasting, scams.
"It is going to take a long time to get back to anything approaching normal wrt travel."
Why would one want to return to the pre-Covid normal? At various times in my life, I've commuted by foot, bike, motor scooter, car, public transit in US (lousy), public transit in Japan (better). I've also worked from home. IMO, home is by far the best with walking second and traveling by car a distant third.
I'm pretty sure I'm not the only one. The small company my daughter works for is giving up their New York City office space -- permanently. She's moving back to Vermont. Their latest hire is in Texas and may not meet her coworkers face to face for years ... if ever. I wouldn't be surprised that's the future for a lot of companies.
Of course, not every job can be done remotely. And there are huge problems with remote working. But I think overall, that cities as we knew them in the 20th century may be headed for long term decline.
Overall -- not all that great a loss I think.
There are lots of accidents on UK roundabouts ...
Are there really? I ask because when I stop to think about it, I've seen very few accidents of any sort in recent decades. Roundabouts or no. And the few I've seen while driving or on TV seem mostly to result from slippery roads, booze (apparently no tree is safe in the wee hours of the morning), and/or poor judgement about following distance or passing clearances on highways.
When was the last time I actually saw or heard about an accident at an intersection? I really can't remember. Many, many years ago for sure.
Downvoted because in my opinion safety of SMALL roundabouts is wildly oversold. The places where they work well are largely places where signals or a multiple way stop would work equally well. The big problem is that there is no way to tell if a car approaching around the roundabout on the driver's side is going to turn onto the street the driver is coming from or will continue into the space the driver will be turning into. Turn signals? They'd likely help a lot. If drivers used them properly. But proper usage would need to be 100%. It's nowhere near that. And in some more complicated geometries, it's far from obvious what proper usage would be
Caveat. My objection doesn't apply so much to large roundabouts (although the signage on the one large roundabout in my area is so poor that traffic there doesn't flow smoothly if anybody from out of town is trying to navigate it). Roundabouts will presumably be fine in about forty years when (almost) all vehicles will probably be able to talk to each other, ascertain intentions, and negotiate passage. (But what about vehicles with no or failing communicaiton gear, or heavy fog or rain or smoke or snow that impedes V2V communication?)
"... quite often the show would end with a horse drawn buggy (or cart) accident where it was carelessly driven off a cliff."
Hmmm. Is isn't driving off a cliff exactly how the owner of Segway Personal Transport Systens died in 2010? https://en.wikipedia.org/wiki/Jimi_Heselden. According to Wikipedia
On the morning of 26 September 2010, Heselden was killed when he fell from a cliff footpath into the River Wharfe, at the village of Thorp Arch near Boston Spa; a Segway vehicle was found near him. The West Yorkshire Coroner concluded that Heselden had died of "multiple blunt force injuries of the chest and spine consistent with a fall whilst riding a gyrobike".
Well, the longest direct (great circle) distance possible on Earth is -- pretty much by definition -- 20,000km. Assuming various delays double that, the maximum latency would be around 40000/300000 sec = 0.133 sec = 133 msec. Annoying, but not intolerable for most usage I think. And, of course, typical delays possibly would be less. How often do most of us talk to our antipodes?
Of course, there's no guarantee that Starlink when fully loaded will have latencies as low as 133 msec. 133msec is just a number that might, maybe, be achievable.
In the 70 years since Miller-Urey, dozens -- maybe hundreds -- of reactions have been recognized that create complex organics from simple components at rational temperatures and pressures. The modern Earth has a wide variety of environments with differing acidities, temperatures, with and without various reactive chemicals, etc. Why would the early Earth be any less diverse? With no microbes to tame the mess, I'm guessing that an organic stew with hundreds or thousands of distinct complex molecules mixed might have been present once liquid water was present. And liquid water seems likely to have been present early on even though our "knowledge" of stellar evolution seems to tell us that the planet couldn't have been warm enough to have liquid water oceans.
On the other hand, we don't actually have any route(s) identified that lead from "organic stew" to "life as we know it" Not even in concept. Much less in detail. And we aren't likely to any time soon.
Could life arrise spontaneously in that situation? Yes, possibly. Could some space traveling spore or critter, survive and thrive in the "stew"? Yes, possibly. (Although I'm ignoring the possibility that some elements of the "stew" are likely to be toxic to any given lifeform). So, is spontaneous organization of life on Earth possible? Maybe .. or Maybe not. How about panspermia? Maybe. ... Or Maybe not.
AYL has a point. I'm in the US. If I hack a few computers in Russia from the US and use the data to trade on the Moscow Stock Exchange from the US, is that even a crime in the US? What crime? A felony? Or a misdemeanor? If Whatshisname is turned over to the Russians and is tried for anything, is he likely to receive any punishment other than a stern "I don't want to hear about you doing this again" lecture?
("I don't want you to hear about it either your honor. I'll try harder not to get caught next time.")
You neglected to mention that the emacs aircraft -- while armed with truly impressive weaponry -- is controlled by a rotary telephone dial using 6118 unique and totally non-intuitive dial in codes.
(I actually use emacs because of the configurability and the broad capabilities of org-mode. But, I'd never, ever, recommend it to anyone else.)
"They are the slightly more modern equivalent of the telephone menu system"
In recent years, I find I have become quite fond of telephone menus. They are usually fast, efficient and do not waste my time displaying gaudy, but useless imagery, and annoying ads. Neither do they leave me stranded when some step in the process I'm engaged in fails to render or won't accept my input without explaining why.
Unfortunately, telephone menus have increasingly been replaced in recent years by computer driven garbage that wants to TALK to me. Those rarely work worth a damn. My policy has become NEVER, EVER make the mistake of talking to a computer. At least not one one that purports to be speech capable. If you stubbornly refuse to say anything, they'll usually fail you back to the old telephone menu or route you to an actual person (Although with some companies -- Comcast for example -- the latter is not much of an improvement.)
Old News -- 10Ks usually are not very informative. An awful lot of publicly traded companies seem not to be very anxious to share actual information about their finances with that selfsame public. I doubt that will change much even if the reporting requirements change. They have people to make sure that transparency doesn't happen unless it just can't be avoided.
Nonetheless sometimes things can be seen. In 2000 or so, on a whim I compared the 10Ks for Worldcom and Sprint which seemed to be in much the same business and concluded that Worldcom looked to be the world's largest freestanding collection of IOUs. It turns out that not only were they deeply in debt, they were claiming imaginary assets. Lots of them. They filed for bankruptcy in 2002. Similarly in 2006 I compared the 10Ks for Ford and GM. Ford was making a little money. GM was losing a little. That was in good times. What would likely happen when times turned bad? Yep, Ford survived the 2008-2009 financial crisis. GM didn't.
So., 10Ks probably aren't entirely useless. But I wouldn't expect to learn much from most of them. With or without better requirements.
What does the emergency calling system do if you don't pay up? Probably it calls in to the local authorities to report each and every speed limit violation, rolling stop, or other legal transgression you commit. All for your own good of course.
"Guess where the electrically driven pumps for the fuel were ?"
Serious question: I'm not an engineer as this question will probably demonstrate.
I think that a mechanical pump can PUSH fluids a long ways. 20 stories? Sure, why not? But I don't think it can SUCK fluids all that high. 10m (four stories?) max for water? A bit further for diesel which is less dense? But not 20 stories?
OK, maybe the pumps can't be on the roof with the generators. And the bottom where the water to be pumped is flowing in possibly is not a good idea. What's plan B?
"yep... but a Chinese company drops 300 billion and it's just chickenfeed..."
Not to worry man. The Evergrande default situation is "contained". The Chinese have told us so.
Of course the cascading defaults of dubious US loans and weird financial instruments based thereon in 2007-2008 were "contained" as well. Federal Reserve chairman Ben Bernake told us so.
I think Python in your browser is probably a bit more likely to work as intended than Javascript which seems often not to do quite what the website author intended. I mean, why would one intentionally put an ad with a non-functioning Close button on top of the order form I am trying to fill out? But I agree that working better not necessarily a good thing.
Don't get me wrong. Python in a browser seems an RBI (Really Bad Idea) from a security viewpoint if nothing else. But also it sounds like the real problem is the WASM capability that enables Python to run.
Does WASM have anything much to offer users other than a probably small rendering speed improvement and potential bundle of major grief?
Yes, the technical folks should determine what equipment is needed.
But the decision whether to buy or lease usually depends on things like tax and cash flow considerations. And the people who understand those are the accountants.
Yes, a lot of companies make stupid decisions. In my experience, that usually isn't really the fault of the technical folks or of the accountants. It's the fault of management whose job is to talk to their experts and come up with the most viable solution for their particular situation.
As with most things Musky, it's not all that easy to find technical details on Starlink. But I think it likely that the satellites in higher orbits mostly are there to act as relays -- which won't do wonders for latency, but would decrease the number of ground stations needed.
As for needing a license. India either is, or will shortly be, the world's most populous country. Of course ground stations there will be needed. If not immediately, eventually. And licenses. And, one suspects, substantial bribes to India's notoriously corrupt bureaucracy. All in good time?
"It nice to know the Chinese have bought in to the quantum snake oil, jut like everyone else."
You've got a point there. A bit of skepticism is probably appropriate. On the other hand, China (and US and EU and ...) probably can't afford not to assume that quantum or other advanced decryption techniques might become available at some future time.
"The only data that should be shared is data concerning Interpol and global police operations"
That's a bit too narrow I think. You don't really want to shut down all international commerce I expect.
And laws making the activities of the NSA,CIA and their numerous counterparts in other countries illegal are kind of pointless. Intelligence Agencies don't care if they are breaking other nation's laws. Breaking laws is one of the things they do. (Although a lot of their activity like reading and analyzing foreign newspapers is quite legal and seems harmless and appropriate).
But laws constraining Google, Facebook, et. al. from spying on people seem entirely reasonable and proper. Even more effective, might be to tax their data stores at rates high enough to discourage retention of much more than basic login identification and the details of active transactions. Unfortunately, I can't begin to figure out how that can be done. I suspect it's impractical.
"Not sure how robo drivers are going to make that any better"
My understanding is that Waymo only works within thoroughly mapped areas. Good technical reasons for that. I imagine these are the same. You give it an address (How? If you don't speak a Chinese language or have a basic understanding of Chinese characters, that may be a problem). It goes there. You pay with a credit card or phone app or something. Maybe cash even, No tip presumably. If the address system is like Tokyo (buildings numbered in the order they were built), the last 100 meters or so of your journey may be a problem. But that's not the taxi's fault. And a human driver might well not be any better unless you share some common language.
Nothing wrong with digital currency I think. Really, for significant amounts, in much of the world, those of us not in the drug trade rarely see much physical cash anymore. And we haven't for decades. We use credit cards, checks (in North America at least), and such.
But why blockchain? Blockchain wastes more energy than Las Vegas lightshows and seems necessary only if you don't trust your "bank". But Japan is one of the world's more honest countries. World Population Review (https://worldpopulationreview.com/country-rankings/least-corrupt-countries) ranks it number 19 between Ireland and Iceland. That's lower than the UK,Canada and Australia (tied for 11th place), but better than the US (25th place). Why would one not trust a consortium of Japanese banks? For that matter, why would one trust blockchain? It's extraordinarily complex and seems a good deal less transparent in practice than in theory. At least it appears that a significant amount of cryptocurrency has somehow found its way out of the pockets of those who thought they owned it and into the pockets of ... well, somehow that seems a bit unclear.
Not that complicated if you've ever had to look at radar return waveforms. A bit puzzling if not.
If the moon were a glassy smooth sphere, you'd only get one return signal -- from the center of the moon. Any other reflections would go off to your sides. But the moon is bumpy. Craters and such. So in addition to the return from the center point, you'll get some signal bounced back from other lunar features. Those are further away from you than the center of the moon. It takes a tiny bit longer for your signal to get to them and a bit longer for the signal reflected from them to get back. Result: instead of one tidy set of pulses that is mirror image of your outgoing signal, you'll get a reflected signal that's smeared out over (I'd guess) 20 or so milliseconds.
I may have some of the details wrong. It's not like I spend my spare time bouncing radio signals off the moon. But I think the above is the general idea.
Reasonable question -- deserves an answer.
LoRa is a proprietary spread spectrum technology that is used in license free bands. It presumably allows multiple user pairs to communicate over the same band simultaneously without requiring them to find a free frequency slot. The Wikipedia article is at https://en.wikipedia.org/wiki/LoRa
There's also a Wikipedia article on EME (Earth Moon Earth) communications in general at https://en.wikipedia.org/wiki/Earth%E2%80%93Moon%E2%80%93Earth_communication
I wouldn't be surprised that the motivation really isn't to save money. It's quite possibly a not entirely unjustified fear that at some future time their operations could be hostage to the whims of MS management. Potentially, MS could increase their subscription fees by 35% a year every year for 20 years, and those users who are locked in would have no choice other than to pay up.
But MS wouldn't do that? Haven't dealt with many MBA equipped beancounters have we?
I'm guessing, but it appears to me that local and regional governments do an awful lot of stuff that may be difficult in some cases to do well with the generalized tools in office packages. Birth and Death records and producing the corresponding certificates on demand. Land titles and zoning restrictions and zoning variances. Police records. Probably medical data subject to complex access rules. Library catalogs perhaps. Maybe lists of library cards and expiration dates. Dog licenses. Canine rabies shot certificates. Tax information. Property ownership data. I actually looked at the latter once for some project of my esteemed spouse's. It's public record information here in Vermont. The property locations and valuations were orderly. But man-oh-man was there peculiar stuff in a few of the owner and owner address fields. The property is jointly owned by several trusts with multiple addresses in multiple countries?
My guess is that sort of stuff they want to do in OS independent environments -- browsers apparently.
My take is that becoming OS independent is possibly feasible. And it's probably a good idea. But it will not be easy. And they will be lucky if they can do it at current budget levels.
This is an interesting problem. Here in North America small farmers have pretty much been decimated because it seems that every economic shock of any sort makes their life more difficult. A century or so of that and it's no wonder that most farmers have sold the family plot and taken up web site design or retired to someplace warm. Currently, the percent of US population engaged in agriculture is slightly over 1 percent. And US/Canadian farms mostly weren't/aren't all that small by Asian standards.
China seems to be doing much better agriculture wise than most developing countries, but they seem to have developed a weird system with some family members intensively farming their leased plot while living in a house they own built on land they lease. And other family members work in sometimes distant cities. It's difficult to imagine such a system working at all much less being transplantable to India. Maybe Japan or Europe offers some answers that will work for India.
So Modi is probably correct that Indian agriculture needs massive structural changes. His problem is that any such revolution is likely to involve politicians hanging from lamp posts. I doubt he favors that part.
And door locks themselves can freeze up in cold climates -- especially on older vehicles where the weatherstripping has stiffened and cracked with age allowing water to seep in. Even if the lock -- mechanical or electronic -- works, the latch may not free up. Solutions involve a hair drier (if you care about the paint job) or a propane torch (if you don't) or entering through a rear door (if there is one, the weatherstripping on those will likely be less worn) and climbing over the seats in order to start the car which can then be run for 20 minutes or so with the heat on full. But won't the latch freeze up again when you turn the engine off? Very likely, Yes.
Waiting for Spring, or moving to a warmer climate will also work.
"Dreadful design."
Think of it as a safety feature. If you can't get into your Tesla you and the rest of us are safe from Musk and Company's sometimes dubious design decisions.
Key fobs ARE convenient. I've been known to install an Avital unit into an older car without remote entry in order to get that convenience. But twice in the past three decades, I have had to deal with key fobs with weak batteries. Once -- on a subfreezing night in rural Vermont -- warming the fob by vigorous rubbing got the door open and engine running. The other time I jury rigged something with a battery from another device and some scraps of wire. Neither car was mine incidentally. I think cars without physical locks are a problem waiting to happen and won't buy one.
As for a vehicle that requires an internet connection for entry and use. That'd be really odd. Maybe for an armored car or some other special purpose vehicle there's some rationale. But for just another set of wheels... Really, that's bizarre. I have to believe there must be some alternative method of access even though it may not be obvious.
Incidentally, having a physical key in the fob doesn't help if there is no physical lock to insert it into. Tesla is not the only car maker whose designers possibly are better suited to jobs in the food service industry.
I'm sure the Chinese among others would be happy to submit a very attractive bid and might very well deliver a superior product at very low cost. With adequate backing from their intelligence people, they would not be burdened by unpleasant constraint like the need to show a profit.
Of course the US requires a US contractor for a DOD data management system. They'd require that even if the country didn't have much in the way of qualified candidates. Fortunately the US does have a number of companies that look to be as qualified as any in the world. Now whether there is a meaningful job there. That's a different issue.
You got me curious, so I spent 15 minutes on the internet and am therefore now an EXPERT. It looks to me like "sandboxes" maybe aren't exactly built into the Javascript language or its interpreter as many of us assume. Perhaps we are expected to bring our own sandbox to the party which we do through our choice of browser(s). I certainly hope I'm wrong about that.
Perhaps someone who actually understands Javascript sandboxes would care to explain to you and I how they work and how they can possibly be anything other than a thick application of cosmetics over a massive collection of security issues.
BTW, I'm not sure one actually needs admin privileges to overwrite the MBR even without the help of Javascript. As least in older (pre UEFI) disk setups, the MBR (and all of track 0 for that matter) are not part of any file system and aren't protected by the usual OS mechanisms. You may have to bring your own "device driver" to overwrite them, but maybe that's not all that hard to do.
Would Google sell my fingerprint to an advertiser? Of course they would. They'd sell my saliva if they could find a customer and had a way to collect the spit. (Log on by tongue-print anyone?) Would any advertiser actually pay for it. Ad people are pretty weird. But not THAT weird. ... I think ...
My GUESS would be that the sensor and associated logic works just fine for some fingerprints including the one you use, those of the Google developers and those of the Google test team. But that it fails for some. I wouldn't be at all surprised that if you spent a rainy afternoon testing the fingerprint sensor with all ten of your upper extremity digits, you might find that some work reliably and some don't. They're all different y'know.
As for "enhanced security". If logging in doesn't work, IT'S SECURE. Maybe not the optimal solution to providing security ... But, to a spin artist I can see "works poorly if at all" morphing to "is much more secure".
"It's in a low earth orbit, at a lower altitude than StarLink satellites. As I recall, there is enough atmosphere at that altitude that StarLink satellites would deorbit themselves in just a few years"
Not that simple. Nothing ever is it seems. Last I looked, there were to be three batches of Starlink satellites. One batch (most of the satellites) to have perigee around 330km, a second batch around 550km (comparable to Hubble) and a third batch around 1000km. It's the low orbit birds that give (some?) users lower latency, but burn up quickly. But that really is (if they are approved) the majority of them.
Hubble is expected to reenter in the mid 2030s. Presumably it could be moved to a higher orbit before then if it is still working. That probably doesn't require a manned mission.
Depending on what's wrong with HST, I wouldn't ignore the possibility of one or more unmanned repair mission(s). I think that we're a long way from robots that can do everything that humans can. OTOH, we can make robots that can do many straightforward mechanical jobs and whose operation can be tweaked a bit remotely. And if (when) things don't go according to plan, you can probably simply stop and think things over without having to worry about problems like running out of Oxygen. And you don't have to bring robots home safely after the repair.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
