Minecraft's my Nirvana. I found it hard, it's hard to find. Oh well, whatever... Never Mined The future often arrives looking like an expensive toy. From the first microcomputer to the latest self-piloting drone, these "toys" hide a larger truth: they're the canvas upon which our imagination plays, as we dream up braver, bolder visions. We think physically, with our bodies, and our toys help us get our hands around what …
Wasnt expecting to have a thought provoker before my coffee kicked in :)
Couldnt agree more on the premise of "The future often arrives looking like an expensive toy", dovetails with a theory/prediction of mine that VR/AR will gift the world a cure for blindness as a side effect of implants for direct optic nerve manipulation developed to overcome the problem of the headset...
So I am to take it that the author was not around during the early micro days. Say 1975 to 1980. If he had been he would have know that the first "toy" microcomputer that had much success was the ZX80. In 1980. All the other successful 1'st and 2'nd generation micros being the price of a small car you pretty much only found them doing real work in business, science and education as a direct replacement for mini computers. Which cost ten times as much. Pretty much all S100 systems were used in business and science. And it was the same for even Apple II's. The first time I ran into one being used purely a home computer was in 1983. So not many toys back in the very early days. Far too expensive.
The other stuff. Mapping.. All that information has been in GIS data-sets for the last 20 plus years. The problem has never been the resolution or density of the data. Its just that apart from traditional mapping related applications there is no other viable use for the data sets. Even games. This has always been a verical market. Always will be.
The AR problem, no viable long term market, has been around since the early VR hardware days 25 years ago. Just because some AR headset for $400 reproduces about 80% of the quality of the $100K plus SGI set up back then is neither here nor there. It is still demo ware. A few minor hit games does not make a completely new platform. Now there is a huge market for this hardware. But as its the VAR industrial and business segments it is way too unsexy for any of the players to invest any money in building out the market. So a multi billion dollar market still goes unclaimed.
> The other stuff. Mapping.. All that information has been in GIS data-sets for the last 20 plus years. The problem has never been the resolution or density of the data. Its just that apart from traditional mapping related applications there is no other viable use for the data sets
That's a little strident. Take traffic data, for example, as gathered by Waze and Google maps. It's data that has been gathered by users' phones, not by satellite imagery.
Well the realtime data sample density from apps like Waze is not that much different from what we had access to in areas like the Bay Area back in the late 1990's. At the time I was using real time sigalert data for my daily commute to what was at the time my first serious GIS R&D project. The only time the 280 commute was real bad. And the realtime traffic data density from WSDOT's system in Seattle around that time was even better than what was available in California. I could check my school run traffic conditions realtime through a web interface. That was in 2000.
Those data sets twenty years ago were from real time road sensors and local police / highway patrol and Catrans / WSDOT sources. Waze adds little to the mix. For usual traffic conditions data sets from apps like Waze and Google Maps are useful but in unusual traffic condition the human with very large data sets from experience going back decades will always beat the routing of those apps.
Waze/Google etc. bring one important thing to the mix - the traffic flow sensor moves with the driver, allowing data to be gathered from a much broader range of roads than would ever be realistically possible using fixed position sensors, and also allowing data to be gathered at much finer spatial resolutions than even the most heavily instrumented roads (based on my UK-centric experience at least). Here, the traditional methods of traffic sensing might be able to tell you there's slow traffic along a half mile section of road, whereas Waze/Google/etc can, given sufficient users in the area, give you traffic flow data down the a few metres of resolution.
There are a growing number of traffic authorities and other official bodies with an interest in traffic flow data (e.g. emergency services) who are partnering with the likes of Waze specifically because of their ability to provide this finer level of detail which complements the data they already get from the traditional sources, and in some cases investing no small amount of time and resources to fully integrate with these new sources of data. Given how cash-strapped some of these authorities/services are, I doubt they'd be so eager to do all of this work if there weren't genuine benefits for them compared with just continuing to use their existing systems...
Agreed, except the quality of data is significantly worse. GPS inaccuracy is quite high - which leads to interesting situations where carpool lanes and on/off ramp pileups damage reliability of data accuracy for people traveling through.
There's also the update problem: in theory, a new traffic signal or a temporary obstruction (construction, etc) should be noticeable, but the density of data inputs is a lot less than most people imagine.
Take Waze: it shows other Waze drivers, and the densities are very low in reality.
This is vastly different than implanted road or bridge sensors. The sensors are far more accurate, far more detailed and much less reliant on sparse users.
Waze doesn't show accurate user densities, the user icons you see on the map are just a random selection chosen, with a few minutes delay to provide a degree of anti-stalking/privacy, from all of the users in the area. And things like car pool lanes are also taken into account in the map data, allowing the back end algorithms to aasociate variations in traffic speeds along the same segment with different lanes rather than having it aggregated into a single incorrect average for all lanes.
For sure the GPS inaccuracies can cause issues, but there are ways to mitigate these, and the benefits of having data (even if it's only geolocated to within 20-30m accuracy, which would be a crap GPS chipset on a bad day - my current phone regularly reports positional accuracies in the low single digits) provided en masse on higher usage roads, and provided *at all* on roads that would otherwise never have any sort of traffic flow data associated with them, cannot be understated.
The problem with GPS inaccuracies is that the mapping algorithms know they exist and try to compensate for them. I have many times found the GPS flavor of the day telling me that I'm on a road that I'm not. Sometimes it's a simple transposition from the access road to the major highway, but the really fascinating ones are from the passenger train to the parallel highway. When they diverge, the GPS tries desperately to keep me on that highway and usually falls over dead when the distance gets too great.
It was TrafficMaster in the 1990 in the UK that gave access to roadside sensors traffic information in real time. It was a subscription thing but everybody shared the logins. I used to check my route home.
Now it’s realtime in a display in my car which looks ahead and offers diversions and the actual end of a queue of traffic agrees with the in car display within about 3 metres.
"Those data sets twenty years ago were from real time road sensors and local police / highway patrol and Catrans / WSDOT sources. Waze adds little to the mix."
Possibly a fair comment in the Bay Area, but Waze adds a fuckload to the mix where that infrastructure doesn't already exist and there is no viable prospect of having a government capable of creating that infrastructure eg in developing countries.
One other poster mentioned the problem, sample density. Even in the Bay Area and LA the Waze/etc real time sample density is really not that high. And suffers from very high temporal and locale clumping. It is a very useful extra piece of information but of limited use outside very built up areas. Which have high density high quality senor networks anyway. Out in the boonies you are not going to have very many real time sample points per road mile. If any.
A friend who depends on Waze/etc in urban areas has been let down an several occasions when trying to get useful info on long distance interstate drives when conditions are marginal. I once got her across Minnesota in a blizzard by using the state highway real time info plus the webcams on the front of the snowplows keeping the interstate open. Using the state road sensors and the realtime feeds from the snowplows I was able to tell them which interstates were still passable. While the apps said they were all closed. Now sure how much info is extracted realtime from from the traffic cams but I have always found them a very useful extra data point. Especially when driving mountains/ passes / and in the great white north in the wintertime.
Eh, no! I had two microcomputers before 1983 (KIM 1 in 1978 and Osborne 1 in 1982), and there were very active computer clubs and bulletin boards at the time. These microcomputers certainly didn't cost as much as a small car, and they weren't just used for business. Also, I bought the Osborne 1 at Micro Center, and they had a decent business from hobbyists.
A quick look at old issues of Byte on archive.org from from around 1977or 1978 will show you lots of $2K and $3K machines and a lot of software to run on it. Starting around $200 and quickly going to $500. I am talking about who actually bought the machines. In large numbers (for the time). I am talking 1975 to 1980. Mass market cheap micros started with the ZX80. The Cosmac Elf and other such interesting items did not sell in large numbers. I remember the Kim. With a 6502. Yeah, sure there were lots of clubs but most of the microcompuers sold, S100's, Apple II's etc were very much used for serious work. Even the next generation, the first IBM PC's, were still very expensive and almost never bough as home computers.
A few years later my first car, used, was half the price of a IBM PC AT at the time. You could get a new pickup or low end compact for about the same price as a PC AT fully kitted out.. Or for a fully kitted out S100 with a few 32K cards a dot matrix printer and a 5Meg hardrive back in 1979.
Looking at old issues of Byte is a reminder of just how expensive early microcomputer were. Or even in 1986 or 1987 for that matter. And a quick look at the software and articles for the late 1970's will give a pretty good idea of what most were bought for at the time. Business, science, or teaching. By 1981 this was all starting to change. When good microcomputers with real power started appearing. Like the Atari XL's and the Commodore 64. But they were still around the $500 plus range. Still not exactly a impulse purchase.
That fits well with the "toy" hypothesis. Until the likes of the Atari, Commodore and of course Sinclair, even cheap computers - and "cheap" is deceptive because £400 computers were about the same as the average monthly wage in the late 1970s - were hardly toys. Acorn's System I at £60? and Sinclair's own Cambridge Mk 14 at £40 (perhaps as much as a family of four would spend on the weekly shop back in the late 1970s?) were cheap enough, but they were absolutely not mass-market because they required a good degree of knowledge and skill to operate, and the willingness to tinker with something which had no discernable use for most people. The more expensive computers such as the RML-380z (the first computer I really "got my hands on") were little better.
That changed with the Vic-20, the ZX80/81 and similar "ready to run" computers which could perform a variety of useful (for certain values of "useful") tasks when connected to the family TV. Once the utility of a computer had been proven, the later C64, Spectrum, BBC Micro, TRS80, Dragon, Oric, etc. etc. were seen as valuable purchases and no longer just toys.
I paid around £1,200 for an Archimedes system in 1987 or 1988. It *was* a lot of money, but the only "proper" computers which were cheaper were the low-end models of the ST and Amiga ranges and the very low-end of the "PC" market, such as Amstrad's 1512 and 1640 units. The Archimedes (slightly biassed opinion) blew most of those out of the water.
I used to have issue 2 of the UK Computer Shopper. It was sobering to look at the adverts. It seems to have disappeared somewhere along the line of the several house moves I've had in the intervening years :-(
M.
But all those "toys" you seem to dismiss, if I am not mistaken, were all doing real work in my experience. Quite often replacing minicomputers. One conversation I remember from around 1978 was the price/[performance trade-off between getting a S100 backplane and a Z80A/B board / memory for a couple of $K. Or a PDP11/03 for about 4 times the price. The option of building the Heathkit H11 was not availability. And anyway, I think the Z80A was actually faster than the H11. In the end the lab got a great deal on a DG Nova. And bough another Apple II with what was left over from the budget.
For me the great learning experience in the first few years after seeing the first IMSAI ads in 1975 was when I was in one of the first computer stores in early 1979. I was chatting to the guy behind the counter when in comes this very obvious businessman who just asks "Which of these computers runs Viscalc?" The shop owner points to the Apple II sitting in the window. The business guy says "How much? Then immediately writes a check for best part of four K and then walks out with computer and Viscal box under his arms. Until 1980 I dont think I saw a single "toy" microcomputer. Even the single board units were when not used to teach programming were being used as some form of hardware controller. Again replacing some other hardware which cost far more. Thats why RS-232 interfaces were so popular. They just did not talk to teletypes.
Anyway by the time 1983 rolled around there were plenty of "toy" mass market microcomputers but the real microcomputers of the time, the Victor 9000's, the IBM XT's, or the Apricot PC's, were not to be sneezed at. And by the end of the following year we had the first Macs which had the horse power to do very serious work. The 512K. The following year I was writing software that enable the Mac to replace $40K workstations and 120K special purpose computers. Progress is when you can deliver 1/3rd of the performance for less that 1/10'th the price.
By the late 1980's the real toy microcomputer were the 3'rd game consoles running 6502 and Z80's. Which have kept that market niche ever since.
Maybe to those on the outside looking in the only so the "toys". But to those of us lucky enough to be on the inside pretty much from the very beginning we tended to see mostly the serious use of these incredible machines. Which is what they were mostly used for in my experience. Anything but toys.
I think these new things that look like toys at first only look like toys to people who weren't intimately familiar with the products and their users. Take small single-user drones. Perhaps, like me, you didn't really think of them very much until lots of people were buying them for no real purpose except to fly them around and take pictures of their houses and make annoying noises. However, certain people had been buying them before they got to that level, and most of them wanted specific information, such as aerial surveys of farmland or photos of something best viewed from a great height. We didn't see them, because they were not flying them near us, but they existed. I think the same probably applied to microcomputers; while some people clearly bought an early and expensive one because they wanted to play around with it, a lot of other people bought one to use it for computing. By the time everyone was getting one, the future had already arrived. People were just noticing it.
those "toys" you seem to dismiss
I'm not dismissing "toy" computers. My reading of the hypothesis was that until the price of computers came down to a level where purchase didn't require a separate column in the capital budget, but could be funded out of the tools budget (or whatever) they remained very much a niche item. You had to have a reason to have a computer because you had to persuade the holder of the purse-strings to loosen them.
Microcomputers began to take off in the late 1970s when prices came down to the level that departments could afford them, even perhaps justifying them as "we need to see what one of these can do", but when they came down to "toy" prices - the sort of cost you could put on a department credit card without too many questions being asked, and particularly once the sub £100 computers came along in the early 1980s that were not only cheap but didn't require a background knowledge in single-stepping assembler hex codes into the things - that's when things really started to explode.
Users of the S100 systems and their predecessors looked at the little ZX80 and its siblings and said "toy", but the fact was that it was just serious enough that you could (with patience) do real work on it with almost zero additional outlay (plug it into your TV and the cassette recorder you already have lying about) and you could do all that for less than the cost of - ooh, let's say a Tarbell S-100 cassette interface for the S-100- based computer you had already spent hundreds, if not thousands of pounds on.
You could argue that a similar thing has happened with the Raspberry Pi. Whatever its failings as a general purpose computer (and you have to admit that the ZX80 had a lot of failings of its own), the Raspberry Pi is "good enough" for a vast variety of uses in exactly the same way as the ZX80 was and - even better - it costs less than taking the family to see the latest Marvel movie at a big chain cinema. It's a pocket money machine in a very real sense*.
The fact that they have now sold over 25 million of the things (scroll down to "Leaps and Bounds") let alone the impact on third-party suppliers, the copycats, clones, compatibles and (arguably) the Arduino-alike market too must tell you something about "toy" computers :-)
M.
*A pre-built ZX80 was originally just under £100 in 1980. According to the Bank of England that equates to over £420 today, which would buy you the Pi, a memory card, a case, a power supply, a nice keyboard, a mouse, a monitor, several electronics kits and tools to muck about with, a desk on which to place it all and still have money left over for a night out at the cinema followed up with a takeaway.
Sorry, every computer, no matter how serious its intended purpose, has had that toy aspect. The first US government computer I "repurposed" in 1972 was either a Teledyne-Ryan, SDS Sigma 7, or HP 200C at the Department of Defense Computer Institute (DODCI). I went into the computer lab with my father on a Saturday during his training course about computers and played a game of "Star Trek" (Mike Mayfield's 1971 version) on the teletype. When I got bored with that, I discovered the programming language behind it (somewhat BASIC-like) and reprogrammed it to calculate Fletcher Pratt's Naval Wargame ship values (long before World of WarShips). In my case, the game was used to familiarize Silent Generation Officers with these new-fangled computers. My reprogramming showed the future flexibility and general-purpose of computers (Teledyne-Ryans were used in the Field Artillery Digital Automation Computer system).
...will take too kindly to people starting to dig up the pavement to try and find diamonds.
The question prevails that if you can't dig to get resources... then how are you going to get those resources in order to build in the overworld? .. and the portal to the Nether will probably found in your local religious establishment.
Does this mean I will suddenly be dodging idiot minecrafters suddenly jumping out onto the bike lane, just as I had to do continually at the height of the pokemon go craze? There are already enough idiots whose attention is completely absorbed by their smartphone that step happily onto from the pavement into oncoming traffic without some silly game encouraging them to do so.
</rant>
@DwarfPants. Agree with you whole heartedly the fewer millenials that trek out to my bit of country the better. Most of them have no idea what the country side is for other than to build motorways, hell most of them don't know that eggs don't get made in Sainsburys but comes out of a chicken's arse.
I heard one mother (in my last town -still rural but becoming suburbia and generally attracting social climbers who think they are minted but really aren't, along with a growing population of drug addicts, yet the local authority denies they import them from other areas despite their accents not in anyway being local) utter to her son when he suggested buying free range eggs "No such thing, just a way to get you to pay more, I've never seen a 'free range chicken in my life'"
My dad then muttered "she'e clearly never left town then...."
At one time I had over 80 birds, including Buff Orpingtons and Wyandottes, they produced fantastic eggs but most people seemed to prefer to buy anaemic eggs from the supermarket instead of my fresh free-range eggs.
In the end I did a deal with the shop.in the village run by a Sikh family, they had a key, collected the eggs and sold them then gave me credit against my groceries and beer.
Townies are weird.
I conclude you're saying with the exception of replacing the quill and ink with an electronic doo-dad, not much about the human element of cartography has changed in the past 500 years: Still only the brave and adventurous travel into areas of the not so easily known. I'll have a cold one ready to greet them.
...as these toys become tools and the map of the world becomes the world itself.
. French philosophers notwithstandnig, maps will never become the world. If you walk across a crenelated black line* on a map, life goes on as normal. In the real world... not so much. Until you respawn, anyway.
* https://media.geograph.org.uk/files/1ff1de774005f8da13f42943881c655f/NAT_black1.jpg
A few years ago, everyone played. Now all the servers are empty... they spend $2 billion just to kill the most popular game in the world.
But it is probably worth it for Microsoft because back then so many kids were learning Java as their first programming environment... and it actually worked on Linux. This is the Skype model... destroy anything that works properly on Linux.
"Yes. It's Minecraft all the way down."
When a senior Microsoft executive saw the amazing virtual computers that have been assembled within the Minecraft environment, he instantly realized that, in principle, Windows itself could run be on a virtual block-based computer built within Minecraft.
At first, he naturally assumed that Windows would run very slowly on a virtual Minecraft CPU, but such concerns about speed and performance are never really considered to be a showstopper at Microsoft. So he allowed himself to continue his thoughts.
Then, suddenly, it dawned on him that he could employ infinite recursion so that the Minecraft executable itself could also be run on its very own self-same Minecraft virtual machine; the block-based CPU merely needed to be multithreading compatible.
Clearly this brilliant concept is deep into Fields Medal territory. With one fell swoop of genius, the physical hardware is no longer required. Everything runs on the new multithreading Minecraft virtual machine, including the Minecraft virtual machine itself.
Yes. It's Minecraft all the way down.
Then, amazingly, it's trivial to type in whatever clock speed you want. There are no longer any physical limits. Performance is unlimited. And Intel hardware is no longer required.
So Microsoft will own the world. See? The upsides are unlimited.
But apparently they're still working on it. Rumours say that it should be working and released shortly.
You may find this explanation to be completely unbelievable, but I've found no better explanation for the question:
Why on Earth would Microsoft spend billions of dollars to buy Minecraft?
Therefore this must be precisely what has happened.
Q.E.D.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
