The things I *would* like to fabricate would be plastic or metal parts that is part of a larger assmebly, but has broken. In that case, it's much harder to measure every dimension, put it into a design package, print off a sample, see where it doesn't fit, modify the design and repeat the whole process until I get one solitary example that fits, performs and doesn't contain any manufacturing flaws that weaken it.
You measure the dimensions, measure them again as a check, put them into a design package, make the part, measure it as a check, and it fits the first time. A machinist's tool box is mostly full of measuring tools.
I've spent a lot of time with Autodesk Inventor open on a computer, and a part and digital calipers in hand. It's very common at TechShop to see someone measuring and modeling an existing part in a computer. For many standard parts, you can download 3D models for SolidWorks or Inventor.
The minimum fee set by default on the client the Bitcoin Foundation maintains (Now called "Bitcoin-core") was changed.
Any other clients or anyone who feels like doing their own compiling can set the minimum fee to anything they like, including 0, but there's no guarantee their transaction will ever get included in a block if they set it very small.
Right The big mining pools get to decide what the transaction fee should be. Only miners can validate transactions. The default value in the wallet is only meaningful if the big mining pools are willing to accept it.
Also, remember that the block chain limits Bitcoin transactions to about 7 per second. That limit was hit a few times back in the day (last fall, when Bitcoin was booming) and no and low-fee transactions didn't get through for days, if at all.
This clerical shop processes once in a lifetime events. Once the retirement data for an employee has been calculated, it goes into a pension payout system that automatically generates the checks every month. So it's not bad that it's mostly manual.
Some years ago, I got a look at the USAF Satellite Control Facility, which until the mid-1990s controlled all USAF satellites from a big blue building in Sunnyvale, CA. They "drove the bus" - handled orbital insertion and adjustment, stabilized the satellite orientation, monitored solar panels and batteries, and handled operational problems. (Payloads, such as cameras, radars, and such were controlled elsewhere by the owning agency over separate data links. Very USAF.) The systems used were so antiquated that one was a custom-built emulator for a tube computer. For each satellite pass, physical patchcords had to be set up to interconnect three computers (one to buffer data, one to decode it, and one to compute orbital mechanics) to process the data for the pass. The consoles looked and worked exactly like the 1960s ones from the Apollo program. The operation took about 600 people to run.
Yet they never lost a satellite through an error made at that faciilty. The USSR has lost satellites through such errors. NASA has. COMSAT has. But not all those old guys in Sunnyvale.
There were two attempts to modernize the facility; one using mainframes, and one using VAX computers. Both failed. It was finally replaced, cautiously, with a new facility at Falcon AFB. I have no idea what they're using. I wouldn't be surprised if some of the old software for some of the older satellites is still running in emulation.
It's a Zeppelin NT. One was based in Silicon Valley for several years, but didn't make money after the price of helium doubled. It cost $400 for a sightseeing tour of the Bay Area.
I've heard a talk by the company CEO, who'd piloted the thing. It handles much better than the classic Goodyear blimp, which he'd also flown. With three steerable props and computer coordination, it's much more controllable during landing. It doesn't require a large ground crew hanging onto ropes to get the thing tied down. That's why Goodyear is going with the NT, even though it's more expensive than their classic blimp. There are videos on Youtube of both types landing.
If you want to see what it's like to fly one, the open source FlightGear simulator has a good model of the NT, including the mobile docking truck.
Google is on the right track with their "certificate transparency" scheme, with a public log of all certificate generations, but, like most Google schemes, it involves Google as a central party. The public log needs to be decentralized.
We know how to do this. The Bitcoin block chain is just such a decentralized public log. The Bitcoin block chain could be used to secure the cert log, by putting the Merkle tree into a Bitcoin transaction every 10 minutes or so. Then there can be multiple copies of the public log, and anyone can check them for consistency and validity.
["Frozen"] took the No. 13 spot on the all-time worldwide box office list this week, passing "Star Wars: Episode I - The Phantom Menace" and "Jurassic Park."
The point being that ditching gender stereotypes in mass media can have a very big financial payoff.
Strong female heroines have done well in Disney animation for decades. It's been a long time since "Someday my prince will come". "Brave" did about $0.5bn against $185m production cost. "Mulan" did very well. So did "Pocahontas". "Frozen" is a classic Disney musical, with music, merchandise, and othe tie-ins. (And, inevitably, sequels. "Frozen 2" is in development. Disney used to have a whole division, Disneytoons, devoted to crap direct-to-DVD sequels, such as "Mulan 2" and "Pocohontas 2". It's not clear whether the Pixar tradition of good sequels or the Disney tradition of crap sequels will win out.)
In contrast, "Tangled", which has a female lead who's not that heroic, wasn't as profitable. ("Tangled" cost $260 million to produce. Six years of software development on the hair. Really.) So we can expect to see more heroic princesses come out of the factory in beautiful downtown Burbank.
Most Japanese anime seems to be about female heroines. It all started with "Dirty Pair" in the 1980s, and never stopped. It's such a cliche now that there are anime which parody it.
Probably... they were 'magically discovered' when it became clear to Mark that the authorities might be contemplating criminal charges against him.
Probably the best explaination. But it's not enough to save Mt. Gox. 650,000 Bitcoins are still missing.
A lot of people are looking very hard at Mt. Gox. Mt. Gox now has a court-appointed bankruptcy supervisor in their offices, authorized to look at all records and talk to all employees. The Tokyo Metropolitan Police Department is investigating. Maybe they didn't know much about Bitcoin a few weeks ago, but if they need outside experts, they can bring them in.
Fluke did their job well. Now it's up to Sparkfun.
Whoever gets those will have the experience of using a good multimeter. I have a Fluke 21 on my desk right now. It's over 20 years old. Autoranges over inputs from 1mV to 1KV. Auto power off. Runs for years on a 9V battery. Test leads have good strain reliefs and don't wear out in normal use. Finger guards on the probes so you don't slip into a live circuit. Ohms measurement still calibrated properly; goes to 0.00 if you hold the probes together tightly.
It's out of production, but its replacement, the Fluke 77-IV, sells for about $260. If you want an original Fluke 21, they're for sale for $100 on eBay. These things last for decades.
That's the largest genome that's been fully sequenced, not the largest genome known. See Comparison of different genome sizes. Genome sizes for plants vary over a huge range, and aren't closely related to organism complexity. The largest genome known is for an amoeboid.
Right. After reading through all this, it's not clear why Homeland Security bothers. A customized news feed from Google would be about as useful and much cheaper.
These "media" guys apparently just send out little bulletins. There's no indication that they're tied into the operational side in any useful way. Homeland Security presumably has an operations center that handles active incidents. But the media guys don't seem to use any info from there, even a list of active incidents. If the operational center got a report of an explosion somewhere, one would expect that the "media" team would immediately get that info, then start looking for info about the incident appearing on social networks and tie that to the incident, both for immediate and later use.
The base Bitcoin technology is surprisingly good. Nobody has been able to double-spend yet. The "mallability" bug has to do with programs which incorrectly decide a transaction didn't go through and redo it.
Most of Bitcoin's problems aren't with the software. Bitcoin's irrevocable money sends to anonymous remote parties are the con man's dream. At last, you can rip people off without ever giving them enough info to find you. That's why Bitcoin is such a scumbag magnet.
Mt. Gox's problems stem from a combination of incompetence and criminal activity. They're not technical. Karpeles was running a business that handled a billion dollars a year without an accountant, a controller, an inside auditor, an outside auditor, or a compliance officer. You can't do that and succeed. You have to have enough separation of functions that no employee can steal without detection. Mt. Gox didn't have that. Probably so that Karpeles could steal.
There's a lot to be said for this. Formal analysis of analog systems is possible.The F-16 flight control system is an elegant analog system.
Full authority digital flight control systems made a lot of people nervous. The Airbus has them, and not only do they have redundant computers, they have a second system cross-checking them which is running on a different kind of CPU, with code written in a different language, written by different people working at a different location. You need that kind of paranoia in life-critical systems.
We're now seeing web-grade programmers writing hardware control systems. That's not good. Hacks have been demonstrated where car "infotainment" systems have been penetrated and used to take over the ABS braking system. Read the papers from the latest Defcon.
If you have to do this stuff, learn how it's done for avionics, railroad signalling, and traffic lights. In good systems, there are special purpose devices checking what the general purpose ones are doing. For example, most traffic light controllers have a hard-wired hardware conflict checker. If it detects two green signals enabled on conflicting routes, the whole controller is forcibly shut down and a dumb "blinking red" device takes over. The conflict checker is programmed by putting jumpers onto a removable PC board. (See p. 14 of that document.) It cannot be altered remotely.
That's the kind of logic needed in life-critical systems.
The listing terms that the HKEx finds objectionable are centered around the proposed structure of the company, which would allow their 28 partners to control a majority of the board - even though they only own around 13 percent of the company.
Apparently, the HKEx regulators still cling to the quaint notion that small investors are important. I guess those HK guys have a thing or two left to learn about how real capitalism works.
There was a time when the New York Stock Exchange didn't allow that, either. They caved about a decade ago.
Now, both Google and Facebook have two class "president for life" stock issues.
What might solve the Google problem is to increase the penalties for attaching advertising to someone else's work. People who make content available for downloading but don't put ads on it are doing it as a hobby, and they're not going to become too big, because serving that data costs money. It's the ones who pirate material and add ads that are the problem. They're doing it as a business.
I've always thought of San Francisco as the new Sodom and Gomorrah.
That's a decade or two out of date. Maybe more. Here's this weekend's list of sex and fetish events in SF. There's a nostalgia tour for tourists of SF's sex history, and a screening of porn films from a Berlin festival. Yawn.
I'm not seeing this. It's a dull period for San Francisco. The first dot-com boom was more fun. Connecting up everybody and everything was important. This boom is all from ad-based companies, and most of what they're doing is rather banal. So are many of the people doing it.
Almost all the artists who need more than a desk and a laptop moved out years ago.
SF used to have lots of big empty warehouse and factory spaces that were used for art projects and wild parties. That's what SOMA was. Those are gone, replaced with "live/work lofts" or giant bullpen workspaces.
I do not get why tech people want to live in the Mission. I've had friends there for years, and it's tolerable, but not a place to live in by choice. Wednesday I went to a stand up comedy improv thing in the Mission where people tried to put together presentions from random PowerPoint slides. Heavy bouncer presence outside because it was right next to a service center for homeless people. The comedy sucked, too. That's what the tech crowd is bringing into the area.
Here's a typical Mission location, one which also happens to be a Google bus stop. "Cafe la Boheme" has crappy food, and it's had crappy food for years. The place with the graffiti is an upstairs dance studio which is hanging on. "Chinese Food and Donuts" isn't very good at either. That corner has looked the same for many years. There are some decent restaurants a few blocks over on Valencia, but not at this corner. There are cool places to live in SF, but this isn't one of them.
This is what the European Union really does - they set standards so stuff works all over Europe, across borders and across vendors. Like GSM phones. In the past, over 20 years they moved the 220V and 240V countries to 230V. That was completed in 2003. Trying to get the whole EU to use the same AC power plug, though, was not successful.
Half a billion dollars has been stolen. Where's the Tokyo Metropolitan Police Department? This is their job. It's embarassing that they haven't made any arrests.
It's not "youth" that's the problem. It's banality. "The best minds of my generation are thinking about how to make people click ads. That sucks." - Jeff Hammerbacher, Facebook. Most of the "app" companies are not "tech" companies. They're fad publishers. The technology for doing routine web apps and phone apps is pretty much standardized now.
The engineering that goes into phone hardware is just awe-inspiring. Electronic design today is brutal. You barely get to use any power, the budget for each function is tiny, the size has to be very small, you have to operate multiple radios without interference right next to each other, and there's a new product to get out every six months.
Most of that engineering is not done in the US. That's a big concern. The US probably doesn't have the technology to build a cell phone any more.
It's not as bad as the first dot-com boom. This time, there's usually revenue. Income, even. Even Twitter claims to be profitable (although they're not, really. Look at the Generally Accepted Accounting Principles results, not the ones excluding "one-time expenses".)
The distributed, eventually-consistent blockchain anchored by mining works and is quite robust against attack. Nobody has yet successfully attacked the basic Bitcoin system and stolen money. So the low level technology appears to be secure.
Irrevocable, remote, anonymous transactions are the con man's dream. Especially when they're assocated with a whole community of suckers who think anonymous anarchy is a good idea. The scam level in the Bitcoin world is huge. Over half the exchanges have gone under, and that was before Mt. Gox. Bitcoin-oriented "stocks" and "Ponzis" have an even worse record.
Personal computers are not secure enough to store money. "Bitcoin wallet stealers" are a major problem. Many "online wallet" services turned out to be scams. Storing Bitcoins safely while still being able to use them is quite hard.
Volatility is far too high for Bitcoin to be a useful currency. Since last October, Bitcoin has gone from $100/BTC to $1100/BTC to $600/BTC. Daily variation often exceeds 10%. The companies that accept Bitcoin for real products have to reprice every few minutes. Bitcoin behaves like a pink sheet stock. Too many speculators, not enough real customers.
There are scaling problems. Currently, every user has to have a complete copy of the entire transaction journal back to the first Bitcoin, and has to keep up with all the transactions as they happen. The confirmation process has a 7 transaction per second limit. Confirmations take about half an hour before they can be trusted; longer during busy periods.
"Mining" is more centralized than expected. The original idea was that "mining" would be a spare-time activity of each user's computer. In practice, "mining" is done in large data centers with custom water-cooled ASIC chips. Two mining pools control more than half of Bitcoin's mining capacity, and they have the power to set fees and change the rules.
The algorithm used requires that the length of the password be
within configurable length limits, and that the password not
have triplet statistics similar to those associated with words
in the English language. This is an inversion of a technique
used to find spelling errors without a full dictionary. No
word in the UNIX spelling dictionary will pass this algorithm.
Users should be advised to pick a password composed of random
letters and numbers. Eight randomly chosen letters will
pass the algorithm over 95% of the time. A word prefaced
by a digit will not pass the algorithm, although a word
with a digit in the middle usually will. Two words run
together will often pass.
(The code linked is the original version in pre-ANSI C. Yes, kiddies, that's what C code once looked like.)
The article lists only one Wikipedia article, and it's for a silly game. The article isn't particularly bad, although it could be trimmed a bit. It looks more like fancruft than promotion. A better (worse) example is needed.
Microsoft bought SoftImage, as a part of the effort to displace high-end Unix workstations with PC's running NT. It was all over, but the shouting. Alias transformed Wavefront into Maya in roughly this timeframe, while MS starved out "dot release" life support on SoftImage...
I wrote Falling Bodies, the ragdoll physics plug-in for Softimage, back in 1996-1997, so I got to see this happen. Back then, Softimage was #1 in Hollywood. Microsoft bought them, and when I went up to Redmond, the Microsoft guys were talking about making Softimage mass-market software. But that never happened. It was too hard to use, and required more graphics hardware than most users had back then. (I had a $2000 Dynamic Graphics card in an NT workstation back then. Every low-end GPU today has far more power.)
So Microsoft sold Softimage to Avid. Avid made overpriced film and video editing systems, sold with semi-customer hardware and built into cool-looking furniture. Softimage had a good video editor in addition to the 3D line, and that's what Avid really wanted. They had no clue what to do with the 3D product. They did convert from Softimage to "Softimage XSI", which broke all existing plug-ins and didn't have a plug-in API that worked. That's when I dropped Softimage.
As video editing went mainstream and Avid's sales of overpriced furniture declined, Avid sold off the 3D product to Autodesk. Autodesk had sort of become the default acquirer of 3D animation products. Most of them came from small companies with tiny product lines. Maya came from the merger of Alias and Wavefront and the mess at SGI. Autodesk picked up Lightwave and some other stuff, and of course they already had lots of 3D engineering tools.
This worked out well at Autodesk. The architectural design programs were integrated with the good renderers from the animation world, and images of what new buildings were going to look like got really good. (Adding a radioisity renderer with very realistic lighting models allowed architects to get all the right light fixtures in the right places.) Autodesk's real business is tools for making real physical stuff (their internal slogan is "If God didn't design it, one of our customers did"), but there's a lot of crossover between 3D design of real-world stuff and 3D design of animated stuff.
Softimage has pretty much been a has-been product for years now. After 20 years, it's probably time to phase it out.
The main problem with Mt. Gox was not that the code was a mess. It was a lack of basic financial controls. Mt. Gox lacked a chief financial officer, a controller, inside auditors, outside auditors, a board of directors, an audit committee, and a compliance officer. Yet they were doing a billion dollars of transactions a year. It's not even clear that they have a general ledger listing all transactions. Lack of financial controls is usually considered an indicator of fraud. I've been making this point on bitcointalk for the last year. None of the "Bitcoin exchanges" have proper financial controls. None have an outside auditor and published audits. Yet they're handling far too much money to operate that way.
As for "The National Police Agency seems to lack the ability to analyze the bitcoin trading history of Mt. Gox", that seems to be correct. One would think that the Japanese National Police Agency would have a cyber-crime division, but they don't. In 2013, they were trying to beef up their capabilities in the computer area. This is embarassing for a developed country. Today, any sizable financial mess involves computers, and Tokyo is a major financial center. Untangling any business collapse requires computer forensics and forensic accountants.
The Tokyo police have a backup option - putting Mark Karpeles through one of their standard 23-day interrogation sessions. That's probably going to happen at some point.
Mt. Gox didn't have that high a transaction rate. They only did two or three money transactions a minute on average. They had a lot of traffic from people querying their site for market info, but that's all read-only traffic, and they had nginx and Amazon AWS to help with that.
Their use of PHP wasn't the real problem. From the leaked code, a big part of the problem seems to have been that the front-end system that talked to web users also handled the money. Banks have a separation between the front-end web system and the money system, with standard-format transaction items flowing between them. All those transaction items are logged, often by a third system that just does logging. This allows auditing. It's separation of function that's important, not the language. As far as anyone can tell, Mt. Gox had nobody on staff who understood this.
This all screams "inside job". If you're running a business that handles a lot of money and you lack financial controls, you're scared that someone will rip you off. Unless you're the one doing the ripping off.
Slashdot Mirror
The things I *would* like to fabricate would be plastic or metal parts that is part of a larger assmebly, but has broken. In that case, it's much harder to measure every dimension, put it into a design package, print off a sample, see where it doesn't fit, modify the design and repeat the whole process until I get one solitary example that fits, performs and doesn't contain any manufacturing flaws that weaken it.
You measure the dimensions, measure them again as a check, put them into a design package, make the part, measure it as a check, and it fits the first time. A machinist's tool box is mostly full of measuring tools.
I've spent a lot of time with Autodesk Inventor open on a computer, and a part and digital calipers in hand. It's very common at TechShop to see someone measuring and modeling an existing part in a computer. For many standard parts, you can download 3D models for SolidWorks or Inventor.
The minimum fee set by default on the client the Bitcoin Foundation maintains (Now called "Bitcoin-core") was changed.
Any other clients or anyone who feels like doing their own compiling can set the minimum fee to anything they like, including 0, but there's no guarantee their transaction will ever get included in a block if they set it very small.
Right The big mining pools get to decide what the transaction fee should be. Only miners can validate transactions. The default value in the wallet is only meaningful if the big mining pools are willing to accept it.
Also, remember that the block chain limits Bitcoin transactions to about 7 per second. That limit was hit a few times back in the day (last fall, when Bitcoin was booming) and no and low-fee transactions didn't get through for days, if at all.
This clerical shop processes once in a lifetime events. Once the retirement data for an employee has been calculated, it goes into a pension payout system that automatically generates the checks every month. So it's not bad that it's mostly manual.
Some years ago, I got a look at the USAF Satellite Control Facility, which until the mid-1990s controlled all USAF satellites from a big blue building in Sunnyvale, CA. They "drove the bus" - handled orbital insertion and adjustment, stabilized the satellite orientation, monitored solar panels and batteries, and handled operational problems. (Payloads, such as cameras, radars, and such were controlled elsewhere by the owning agency over separate data links. Very USAF.) The systems used were so antiquated that one was a custom-built emulator for a tube computer. For each satellite pass, physical patchcords had to be set up to interconnect three computers (one to buffer data, one to decode it, and one to compute orbital mechanics) to process the data for the pass. The consoles looked and worked exactly like the 1960s ones from the Apollo program. The operation took about 600 people to run.
Yet they never lost a satellite through an error made at that faciilty. The USSR has lost satellites through such errors. NASA has. COMSAT has. But not all those old guys in Sunnyvale.
There were two attempts to modernize the facility; one using mainframes, and one using VAX computers. Both failed. It was finally replaced, cautiously, with a new facility at Falcon AFB. I have no idea what they're using. I wouldn't be surprised if some of the old software for some of the older satellites is still running in emulation.
It's a Zeppelin NT. One was based in Silicon Valley for several years, but didn't make money after the price of helium doubled. It cost $400 for a sightseeing tour of the Bay Area.
I've heard a talk by the company CEO, who'd piloted the thing. It handles much better than the classic Goodyear blimp, which he'd also flown. With three steerable props and computer coordination, it's much more controllable during landing. It doesn't require a large ground crew hanging onto ropes to get the thing tied down. That's why Goodyear is going with the NT, even though it's more expensive than their classic blimp. There are videos on Youtube of both types landing.
If you want to see what it's like to fly one, the open source FlightGear simulator has a good model of the NT, including the mobile docking truck.
Google is on the right track with their "certificate transparency" scheme, with a public log of all certificate generations, but, like most Google schemes, it involves Google as a central party. The public log needs to be decentralized.
We know how to do this. The Bitcoin block chain is just such a decentralized public log. The Bitcoin block chain could be used to secure the cert log, by putting the Merkle tree into a Bitcoin transaction every 10 minutes or so. Then there can be multiple copies of the public log, and anyone can check them for consistency and validity.
["Frozen"] took the No. 13 spot on the all-time worldwide box office list this week, passing "Star Wars: Episode I - The Phantom Menace" and "Jurassic Park."
The point being that ditching gender stereotypes in mass media can have a very big financial payoff.
Strong female heroines have done well in Disney animation for decades. It's been a long time since "Someday my prince will come". "Brave" did about $0.5bn against $185m production cost. "Mulan" did very well. So did "Pocahontas". "Frozen" is a classic Disney musical, with music, merchandise, and othe tie-ins. (And, inevitably, sequels. "Frozen 2" is in development. Disney used to have a whole division, Disneytoons, devoted to crap direct-to-DVD sequels, such as "Mulan 2" and "Pocohontas 2". It's not clear whether the Pixar tradition of good sequels or the Disney tradition of crap sequels will win out.)
In contrast, "Tangled", which has a female lead who's not that heroic, wasn't as profitable. ("Tangled" cost $260 million to produce. Six years of software development on the hair. Really.) So we can expect to see more heroic princesses come out of the factory in beautiful downtown Burbank.
Most Japanese anime seems to be about female heroines. It all started with "Dirty Pair" in the 1980s, and never stopped. It's such a cliche now that there are anime which parody it.
Probably... they were 'magically discovered' when it became clear to Mark that the authorities might be contemplating criminal charges against him.
Probably the best explaination. But it's not enough to save Mt. Gox. 650,000 Bitcoins are still missing.
A lot of people are looking very hard at Mt. Gox. Mt. Gox now has a court-appointed bankruptcy supervisor in their offices, authorized to look at all records and talk to all employees. The Tokyo Metropolitan Police Department is investigating. Maybe they didn't know much about Bitcoin a few weeks ago, but if they need outside experts, they can bring them in.
Fluke did their job well. Now it's up to Sparkfun.
Whoever gets those will have the experience of using a good multimeter. I have a Fluke 21 on my desk right now. It's over 20 years old. Autoranges over inputs from 1mV to 1KV. Auto power off. Runs for years on a 9V battery. Test leads have good strain reliefs and don't wear out in normal use. Finger guards on the probes so you don't slip into a live circuit. Ohms measurement still calibrated properly; goes to 0.00 if you hold the probes together tightly.
It's out of production, but its replacement, the Fluke 77-IV, sells for about $260. If you want an original Fluke 21, they're for sale for $100 on eBay. These things last for decades.
That's the largest genome that's been fully sequenced, not the largest genome known. See Comparison of different genome sizes. Genome sizes for plants vary over a huge range, and aren't closely related to organism complexity. The largest genome known is for an amoeboid.
Right. After reading through all this, it's not clear why Homeland Security bothers. A customized news feed from Google would be about as useful and much cheaper.
These "media" guys apparently just send out little bulletins. There's no indication that they're tied into the operational side in any useful way. Homeland Security presumably has an operations center that handles active incidents. But the media guys don't seem to use any info from there, even a list of active incidents. If the operational center got a report of an explosion somewhere, one would expect that the "media" team would immediately get that info, then start looking for info about the incident appearing on social networks and tie that to the incident, both for immediate and later use.
The base Bitcoin technology is surprisingly good. Nobody has been able to double-spend yet. The "mallability" bug has to do with programs which incorrectly decide a transaction didn't go through and redo it.
Most of Bitcoin's problems aren't with the software. Bitcoin's irrevocable money sends to anonymous remote parties are the con man's dream. At last, you can rip people off without ever giving them enough info to find you. That's why Bitcoin is such a scumbag magnet.
Mt. Gox's problems stem from a combination of incompetence and criminal activity. They're not technical. Karpeles was running a business that handled a billion dollars a year without an accountant, a controller, an inside auditor, an outside auditor, or a compliance officer. You can't do that and succeed. You have to have enough separation of functions that no employee can steal without detection. Mt. Gox didn't have that. Probably so that Karpeles could steal.
There's a lot to be said for this. Formal analysis of analog systems is possible.The F-16 flight control system is an elegant analog system.
Full authority digital flight control systems made a lot of people nervous. The Airbus has them, and not only do they have redundant computers, they have a second system cross-checking them which is running on a different kind of CPU, with code written in a different language, written by different people working at a different location. You need that kind of paranoia in life-critical systems.
We're now seeing web-grade programmers writing hardware control systems. That's not good. Hacks have been demonstrated where car "infotainment" systems have been penetrated and used to take over the ABS braking system. Read the papers from the latest Defcon.
If you have to do this stuff, learn how it's done for avionics, railroad signalling, and traffic lights. In good systems, there are special purpose devices checking what the general purpose ones are doing. For example, most traffic light controllers have a hard-wired hardware conflict checker. If it detects two green signals enabled on conflicting routes, the whole controller is forcibly shut down and a dumb "blinking red" device takes over. The conflict checker is programmed by putting jumpers onto a removable PC board. (See p. 14 of that document.) It cannot be altered remotely.
That's the kind of logic needed in life-critical systems.
The listing terms that the HKEx finds objectionable are centered around the proposed structure of the company, which would allow their 28 partners to control a majority of the board - even though they only own around 13 percent of the company.
Apparently, the HKEx regulators still cling to the quaint notion that small investors are important. I guess those HK guys have a thing or two left to learn about how real capitalism works.
There was a time when the New York Stock Exchange didn't allow that, either. They caved about a decade ago. Now, both Google and Facebook have two class "president for life" stock issues.
What might solve the Google problem is to increase the penalties for attaching advertising to someone else's work. People who make content available for downloading but don't put ads on it are doing it as a hobby, and they're not going to become too big, because serving that data costs money. It's the ones who pirate material and add ads that are the problem. They're doing it as a business.
Like YouTube.
I've always thought of San Francisco as the new Sodom and Gomorrah.
That's a decade or two out of date. Maybe more. Here's this weekend's list of sex and fetish events in SF. There's a nostalgia tour for tourists of SF's sex history, and a screening of porn films from a Berlin festival. Yawn.
I'm not seeing this. It's a dull period for San Francisco. The first dot-com boom was more fun. Connecting up everybody and everything was important. This boom is all from ad-based companies, and most of what they're doing is rather banal. So are many of the people doing it.
Almost all the artists who need more than a desk and a laptop moved out years ago. SF used to have lots of big empty warehouse and factory spaces that were used for art projects and wild parties. That's what SOMA was. Those are gone, replaced with "live/work lofts" or giant bullpen workspaces.
I do not get why tech people want to live in the Mission. I've had friends there for years, and it's tolerable, but not a place to live in by choice. Wednesday I went to a stand up comedy improv thing in the Mission where people tried to put together presentions from random PowerPoint slides. Heavy bouncer presence outside because it was right next to a service center for homeless people. The comedy sucked, too. That's what the tech crowd is bringing into the area.
Here's a typical Mission location, one which also happens to be a Google bus stop. "Cafe la Boheme" has crappy food, and it's had crappy food for years. The place with the graffiti is an upstairs dance studio which is hanging on. "Chinese Food and Donuts" isn't very good at either. That corner has looked the same for many years. There are some decent restaurants a few blocks over on Valencia, but not at this corner. There are cool places to live in SF, but this isn't one of them.
This is what the European Union really does - they set standards so stuff works all over Europe, across borders and across vendors. Like GSM phones. In the past, over 20 years they moved the 220V and 240V countries to 230V. That was completed in 2003. Trying to get the whole EU to use the same AC power plug, though, was not successful.
Half a billion dollars has been stolen. Where's the Tokyo Metropolitan Police Department? This is their job. It's embarassing that they haven't made any arrests.
Read the whole article. It's quite good.
It's not "youth" that's the problem. It's banality. "The best minds of my generation are thinking about how to make people click ads. That sucks." - Jeff Hammerbacher, Facebook. Most of the "app" companies are not "tech" companies. They're fad publishers. The technology for doing routine web apps and phone apps is pretty much standardized now.
The engineering that goes into phone hardware is just awe-inspiring. Electronic design today is brutal. You barely get to use any power, the budget for each function is tiny, the size has to be very small, you have to operate multiple radios without interference right next to each other, and there's a new product to get out every six months. Most of that engineering is not done in the US. That's a big concern. The US probably doesn't have the technology to build a cell phone any more.
It's not as bad as the first dot-com boom. This time, there's usually revenue. Income, even. Even Twitter claims to be profitable (although they're not, really. Look at the Generally Accepted Accounting Principles results, not the ones excluding "one-time expenses".)
What we've learned so far from Bitcoin:
How do we know that the next update on linux is safe?
That's a very good question. Linus's position on the Intel random number generator not needing additional enthropy indicates he can no longer be trusted.
Sigh. My obvious password detector, published in 1984:
The algorithm used requires that the length of the password be within configurable length limits, and that the password not have triplet statistics similar to those associated with words in the English language. This is an inversion of a technique used to find spelling errors without a full dictionary. No word in the UNIX spelling dictionary will pass this algorithm.
Users should be advised to pick a password composed of random letters and numbers. Eight randomly chosen letters will pass the algorithm over 95% of the time. A word prefaced by a digit will not pass the algorithm, although a word with a digit in the middle usually will. Two words run together will often pass.
(The code linked is the original version in pre-ANSI C. Yes, kiddies, that's what C code once looked like.)
The article lists only one Wikipedia article, and it's for a silly game. The article isn't particularly bad, although it could be trimmed a bit. It looks more like fancruft than promotion. A better (worse) example is needed.
Microsoft bought SoftImage, as a part of the effort to displace high-end Unix workstations with PC's running NT. It was all over, but the shouting. Alias transformed Wavefront into Maya in roughly this timeframe, while MS starved out "dot release" life support on SoftImage...
I wrote Falling Bodies, the ragdoll physics plug-in for Softimage, back in 1996-1997, so I got to see this happen. Back then, Softimage was #1 in Hollywood. Microsoft bought them, and when I went up to Redmond, the Microsoft guys were talking about making Softimage mass-market software. But that never happened. It was too hard to use, and required more graphics hardware than most users had back then. (I had a $2000 Dynamic Graphics card in an NT workstation back then. Every low-end GPU today has far more power.)
So Microsoft sold Softimage to Avid. Avid made overpriced film and video editing systems, sold with semi-customer hardware and built into cool-looking furniture. Softimage had a good video editor in addition to the 3D line, and that's what Avid really wanted. They had no clue what to do with the 3D product. They did convert from Softimage to "Softimage XSI", which broke all existing plug-ins and didn't have a plug-in API that worked. That's when I dropped Softimage.
As video editing went mainstream and Avid's sales of overpriced furniture declined, Avid sold off the 3D product to Autodesk. Autodesk had sort of become the default acquirer of 3D animation products. Most of them came from small companies with tiny product lines. Maya came from the merger of Alias and Wavefront and the mess at SGI. Autodesk picked up Lightwave and some other stuff, and of course they already had lots of 3D engineering tools.
This worked out well at Autodesk. The architectural design programs were integrated with the good renderers from the animation world, and images of what new buildings were going to look like got really good. (Adding a radioisity renderer with very realistic lighting models allowed architects to get all the right light fixtures in the right places.) Autodesk's real business is tools for making real physical stuff (their internal slogan is "If God didn't design it, one of our customers did"), but there's a lot of crossover between 3D design of real-world stuff and 3D design of animated stuff.
Softimage has pretty much been a has-been product for years now. After 20 years, it's probably time to phase it out.
The main problem with Mt. Gox was not that the code was a mess. It was a lack of basic financial controls. Mt. Gox lacked a chief financial officer, a controller, inside auditors, outside auditors, a board of directors, an audit committee, and a compliance officer. Yet they were doing a billion dollars of transactions a year. It's not even clear that they have a general ledger listing all transactions. Lack of financial controls is usually considered an indicator of fraud. I've been making this point on bitcointalk for the last year. None of the "Bitcoin exchanges" have proper financial controls. None have an outside auditor and published audits. Yet they're handling far too much money to operate that way.
As for "The National Police Agency seems to lack the ability to analyze the bitcoin trading history of Mt. Gox", that seems to be correct. One would think that the Japanese National Police Agency would have a cyber-crime division, but they don't. In 2013, they were trying to beef up their capabilities in the computer area. This is embarassing for a developed country. Today, any sizable financial mess involves computers, and Tokyo is a major financial center. Untangling any business collapse requires computer forensics and forensic accountants.
The Tokyo police have a backup option - putting Mark Karpeles through one of their standard 23-day interrogation sessions. That's probably going to happen at some point.
Mt. Gox didn't have that high a transaction rate. They only did two or three money transactions a minute on average. They had a lot of traffic from people querying their site for market info, but that's all read-only traffic, and they had nginx and Amazon AWS to help with that.
Their use of PHP wasn't the real problem. From the leaked code, a big part of the problem seems to have been that the front-end system that talked to web users also handled the money. Banks have a separation between the front-end web system and the money system, with standard-format transaction items flowing between them. All those transaction items are logged, often by a third system that just does logging. This allows auditing. It's separation of function that's important, not the language. As far as anyone can tell, Mt. Gox had nobody on staff who understood this.
This all screams "inside job". If you're running a business that handles a lot of money and you lack financial controls, you're scared that someone will rip you off. Unless you're the one doing the ripping off.