Of lot of the problems in the computer industry stem from lack of proper well thought out standards. As well as the lack of licensing individuals and tool implementations. Diversity of implementations is good if the products adhere to standards. Everybody and their brother creating their own tool chains and proprietary (but open!) widgets that all solve the same basic problem has become the problem itself. We would still be in the preindustrial age if we were unable to standardize even basic things like thread patterns on bolts.

The software industry is the equivalent of recent HS grad noticing that his neighbor built a house using haybales (http://en.wikipedia.org/wiki/Straw-bale_construction) showing up at the city council and convincing them he can build a bridge over the local creek with haybales faster and cheaper than the local engineering firm. They proceed to hire, him and he in fact manages to build a bridge over the creek in an afternoon. Gets a lot of money, fame and further jobs. Its only 6 months later when the creek floods and washes the bridge away are the design tradeoffs apparent. By then, the kid has spent the money, moved out of state and is building cars out of cow manure for a large company.

The problem is, designing a building, pulling a tooth, or fixing a toilet is relatively easy. There are not so many different ways to do it.

There are actually a lot of different ways to do what you describe, you only think its straightforward because the industry has standardized on methods/tools/etc. Pumbing would be a lot harder if pipe sizes weren't standardized and every plumber around designed a toilet from scratch using tools designed by the guy down the street.

The problem with the computer industry is that lots of times people without a lot of formal training/experience are allowed to create a computer languages, application framework/etc. Those kinds of things are limited to the higher levels of licensing and standardization bodies in most industries. Its this very limit that results in the standardization, not the other way around. You can't even get a Journeyman plumbers license in TX without 4 years work experience as an apprentice.

Frankly, the basics of computer operations tend to be standardized by natural selection, the problems are the fact that there are a million different toolkits (often designed by people without a clue) with a million different bugs/edge cases to bite developers. Plus, a person certified with some level of "web development" wouldn't necessarily be tied to an industry like banking or medical software, especially if there was a higher level industry specific "master" license or some such that would sign off on the work. Its similar to the "architecture" positions in these organizations that are responsible for the design of the system, while the lower level coders do a lot of the grunt work.

In fact the PE licenses for civil/etc work much the same way. Getting a civil PE allows you to work in lots of different industries, but usually there is a senior person with a specialty that signs off on a given project. Freshly minted PE's don't get jobs signing off on large bridges, buildings, etc. Plus, a Civil PE usually will enlist the services of a Electrical PE or Mechanical PE for parts of any given project larger than a tool shed.

So, the problem you describe is a symptom of the lack of licensing not the other way around. Frankly, its only the ease of creation of software that allows this to propagate, any other industry would be hamstrung if it couldn't even depend on something as simple as bolt thread patterns and heads being somewhat standard. Sure programming might not be as exciting if everyone had to learn and prove some level of mastery of Ada but I'm betting there would be fewer stupid mistakes being made by bright people writing an application in languages they have never used before because its the "cool" language du jour.

Heck, I am not even allowed to pull a simple tooth without a proper license. If that is not a real scandal....

Maybe the real scandal is that any idiot can convince themselves they are the next Dijkstra, blow some smoke up a midlevel MBA's backside and get hired to write the front-end for the bank you use, or the local software contractor building some part of the control system for the car you drive.

As a society we have decided that we actually want people to have some level of licensing before they design a building, pull a tooth, or even fix your toilet. Yet we allow people without any formal licensing, or competency to design/write critical software.

If you want to have the same level of usefulness now as there was in the 80's, learn Mathematica, MatLab, or Learn Perl,Python, or Ruby

If you want a career in programming maybe, but most of the people writing spreadsheet macros and dbase forms in the 80/90 weren't professional programmers either. Excel and Access are still around, and can solve the same basic problems that they solved back in the 90's without the huge learning curve of ruby, rails, html, javascript, css, datastructures and algorithms, etc just to perform a couple fairly simple computations or graph some data, or create a couple basic input forms.

Sure the solutions created may not be professional programmer quality, or scale to millions of users, but they are often simple solutions created in a matter of hours for real problems being faced by real people.

Frankly, the web is the absolute worst platform for many of these kinds of applications because its takes a problem solvable by a non programmer in a few hours and turns it into a problem that can take a team months. I suspect that a lot of web developers would really have their eyes opened if they spent a few dozen hours slapping together a couple little applications in Access VBA.

Only makers of proprietary modules, and sycophants and nay-sayers complain about the Linux kernel not having a static API and ABI for kernel modules.

Ok, so this is going to sound like flaming but its not.. Because _ANYONE_ who has used linux for any extended period of time in a non trivial manner and, who is being honest with themselves has had problems due to the kernel and associated drivers being tied together.

Thats because without fail, the drivers for every single piece of hardware and the kernel/userspace API's don't all tend to move at the same speed. This means that when you have a problem with a driver for a particular piece of hardware in your system, being forced to take newer versions of drivers for every other piece of hardware is a recipe for something breaking. I have a pretty functional laptop running a recent distribution, I was pretty estatic when about 90% of my use cases actually worked. Sound, 3d graphics, suspend/wakup, wifi, EFI boot, on and on. But, in the end, i'm not running the latest GL SL, because its not supported, I have to hand modify the bluetooth driver to get my wireless mouse working, and I had to hack KDE and the OS to get the back-light dimmer to work properly. Then every couple weeks/months a new kernel gets rolled out to fix some problem or another and I have to reapply my changes.

I could write a 1000 page book about driver/kernel problems I've personally seen, and you need only spend a little time on LKML to notice that driver developers on linux don't really appear to be any better than any other OS. They make forward progress most of the time, but they also regularly regress some feature or another resulting in plenty of pain. Expecting the few hundred developers involved in the kernel/drivers for any particular machine to GET IT RIGHT for in any given kernel release is an exercise in fantasy land. Expecting the ten thousand or so involved in every driver in the kernel to GET IT RIGHT for every single machine running linux is even more fantastic.

So, a problem like this has a solution, ignoring it for religious reasons doesn't.

While purists might believe that analog without the mess of breadboards, wire-wraps and soldering isn't analog, it fills a real-world need.

Its not possible to do modern analog design with breadboards and wire-wraps. AFAIU, this is the kind of analog work being discussed, and from my perspective (comfortable with digital and low frequency analog) modern analog design is pure magic. I'm not sure how any of this can be learned with a "hands on method" today. Board design, maybe, the rules being used to determine microstrip/etc transmission lines are understandable by a normal EE. But analog design on a sub-micron IC? HA!

How many schools can even fab a modern IC without shipping it to MOSIS? Where I went to school there was a "fab" as part of the ECE department but even then the IC's being fabbed were far from cutting edge.

Heck even MOSIS's education processes are looking pretty old.

There's no magic wand they can wave to fix their excessive oversubscription.

Magic wand? or just putting some effort into maintaining the network?

I actually don't see any over-subscription (in Austin) of the last mile. I regularly see my rated (30/5) bandwidth to nearby services.

In my case i'm pretty sure they could wave their wand in front of the provisioning console and upgrade my service to 200Mb/20 (from 30/5). That is because my modem reports a 8x4 DOCSIS lock, with (at least) 16 available channels from the head-end. For the lazy, i'm provisioned less than 5% of the available capacity for my neighborhood. Austin was a backwater until about two years before google announced here, when pretty much overnight TW finally decided to roll out DOCSIS 3 (6 years or so after everyone else). Since then they have been allocating more channels and apparently splitting the nodes.

Maybe they will have an over-subscription problem (if they bump everyone, which is what the rumor mill says), but really i've never seen that here. Instead we just got shitty slow provisioning. Between ~1999 when I got road runner to about ~2010 the best speed that TW would sell a customer account didn't even double.

A high energy electromagnetic field will do just fine. Works on earth... it will work in space.

You just need a fusion reactor.

I don't think electromagnetic shielding is that far fetched anymore. http://physicsworld.com/cws/ar...

Seat of the pants calculation says, its probably smaller than an MRI machine and could be powered with with a similarly sized fission reactor.

Not small by any standard, but completely doable with today's technology.

You mean like ~$180k? Because that is how much the fuel costs are for a falcon 9. If space-x can make the whole thing reusable and get the launch rate up to a couple times a week, the raw cost to put a human in orbit could be just a few times the cost of a first class intercontinental flight.

If the fuel costs is 30% of the total launch costs (about the same as the airline industry) then the expected 6 people per launch would be ~$100k per person, which matches the roughly $500/lb numbers musk has been quoting.

Its not at the level of buying a bus ticket, but its less than many corps are spending for their private aircraft to fly C level execs around.

See GOG.com (where I purchased witcher 2, which I have yet to play). Same basic idea, lots of discounts, nice organization, etc.. Only no DRM! None at all!

The selection though is a little more limited, For me though I tend to enjoy old games just as much as new ones. The combination of GOG, humble indy bundles, and a couple other similar sources keeps me well stocked with DRM free games.

I have a plumber buddy.. He does new construction... That solves the crawling around in other peoples shit problem. Of course it doesn't solve the hauling couple hundred lbs water heaters into attics in the 100+ degree texas heat problem. Then sitting up there running a torch in 140 degree temps..

A lot of these trades are really nice part of the year, and suck the rest of the time due to weather. Plus, the ones that pay well actually require someone intelligent enough to pass the licensing tests. Those tests are often fairly difficult and filled with more math than your average liberal arts student sees in college. Algebra and trig are basically a prereq which is why a high school degree/GED is needed.

The same commercial in different shows may share the same blocks even more than the shows.

Having worked on some deduplication systems, I would be _REALLY_ shocked if anyone's deduplication system can find similar video segments from two different shows and deduplicate them.

Unless the raw footage is being decoded and stored there are two problems. First even if the encoding of the video data results in identical P/B frames/slices it will probably exist at different offsets. Many of the block deduplication systems won't re-align block data. Aka the start of a P frame in one stream may be at offset X and in another stream its at offset Y, so even though the data may be identical for long periods of time the deduplication system won't detect it. The smarter systems might be able to detect small runs inside the frame sequence but even small changes in the transport metadata will probably screw this up. So, it would basically require a full blown video transport decoder to make this work properly. I'm not aware of any deduplication systems with video decoders, the ones with stream decoders like that tend to decode things like oracle databases.

Secondly, the commercials tend to be inserted in real time, with a real-time mpeg encoder. So even with a new I frame at the beginning of the commercial (doubtful), the state of the huffman coders/DCT heuristics/etc will cross over from the content into the commercials. Meaning that the compressed data is going to be completely different even though the video stream may look identical. Even if the data were decoded back to a raw uncompressed format its quite likely the frame data still won't match.

That said, it is possible with some frame matching heuristics to find identical sequences in different video streams. Hmmm, do I smell an automatic commercial removal tool?

OTA can send the full, uncompressed digital signal

I think your a little confused. The OTA (ATSC) standard is still sending compressed video (mostly MPEG2) , just that the bit rate tends to be higher than what most (some?) cable companies provide for their digital content.

See also netflix, which tends to have even lower "HD" quality than the cable companies. With the advent of lossy compression the quality of a show just as much to do with bitrate and compression algorithm than resolution/color depth/framerate/etc.

(I assume that if 1000 of their customers record the latest episode of Big Bang, they keep 1000 copies of it - if not, their argument would be a bit weaker).

And I wonder what that means.. In an age of deduplicated storage your storage vendor's stack is going to detect that there are 1000 copies of the same shit on the storage device and only actually store one copy.

On top of that, property taxes are the highest in the country. Expect to pay over $10k/year on a $400k house (which is a pretty modest house, maybe 1500sf)

That sounds about the same as Austin, maybe a little more. Houses selling here for ~500k, with an appraised value of ~400k are ~$9k a year in taxes, more if you can't claim the homestead exemption.

Although, you _CAN_ get pretty nice houses on the outskirts of town in the areas with bad schools for ~200k. If you want one of the better schools $400k is sort of the starting point for anything over 1200 square feet. Moving farther out doesn't help much. The house prices in Leander and Georgetown are getting expensive too and can often mean commutes > hour depending on where your commuting to/from.

A big part of the commute problem isn't the distance but rather that the roads all seem to move at 10 miles an hour during rush hour due to the fact that Austin hasn't really upgraded any of the core transportation infrastructure since the 1970's. Both mopac and 35 (the two north south roads) haven't had any major work since the 1970's. Recently they started adding an additional toll road to mopac, but that probably won't alleviate the extra traffic added by the toll road extension just a couple years ago. Then there is the problem that there isn't actually an E-W corridor between 183 in the north and SH 71 in the south leaving all E-W traffic on the surface roads.