You're right about that. The good sustainers are usually the last ones left in the building, if only because products are usually sold with an agreement to maintain support for X number of years. Even if the product is ultimately a failure in the marketplace, the few customers who bought and deployed it have a contract that says they'll get maintenance releases for X years.
Unfortunately the higher-ups see that sustaining effort as purely a cost, so it's now one of the first functions they attempt to outsource.
The other thing about comments is that they can be funny. I remember a couple of bugs we tracked down where a failed hardware access resulted in an infinite loop. Sure enough, when we dug in to the code, that while() loop had a comment before it along the lines of
/* i should probably put a guard counter here */
Heh. And then I think of the pain and expense those bugs caused.
It's fun to hack away at home on personal projects, but professional efforts - especially big ones - really do require decent documentation and comments in the code. Unfortunately my experience is that the requirements documents, design documents, and even the comments in the code are usually out of step with where the code actually is at a given time. I've seen a lot of code written by the local genius that only that same local genius can maintain. That's not doing anyone any good. How many times have you looked at a chunk of code and thought it'd be easier just to re-write it than spend the time understanding what's there in order to modify it? Long story short, my vote is for well documented, well commented code. And keep it simple too. Unless you're under contraints that require you to get creative, just write it in a way that others will understand.
I agree that these guys have a right to some privacy. Most interesting to me is that the comments here on/. are generally supportive so far. What a different thread it would be if this had been Bill Gates and Microsoft instead of Steve Jobs and Apple.
Another excellent book that touches on what quants do is When Genius Failed, by Roger Lowenstein. This book charts the rise & fall of Long-Term Capital Management, a hedge fund that relied heavily on mathematical models to guide their trading activity. It's a cautionary tale about placing too much faith in mathematical models of markets that are not always rational.
Technology like this is what forces American criminal organizations to outsource their counterfeting and ransom operations overseas. You're putting American criminals out of work!
The Britannica article does not present the theory as a scientific fact. It uses the qualifier "probably," which means it is likely true but there is no conclusive proof.
Yes, Google is a business. They're successful because we - the owners of the eyeballs they attempt to attract - believe they provide the most relevant search results, queried from as much of the web's content as possible. If they're going to censor results, they should be honest about it and let their customers know that they are doing it. Their customers can then decide whether or not to continue using their service.
Google's been such a great service/product/company that most of us have let them happily in to our lives. We trust their search results. We trust their e-mail service. The worrisome part of stories like this is that with Google's penetration of the market, they're powerful enough to dictate what sites are seen by the majority of people. We know about Abu Ghraib, so if we search for images and don't find them, we know something is up. But what if we search for something we're not as aware of and no results are returned. We may just assume there was nothing there to find.
Let's hope Google's "don't be evil" mission statement wasn't just a trojan horse disguise to sneak in to our lives and slowly become "evil."
Back in high school local companies would donate their old - read: scrap - mainframes to our electronics lab. The computers weren't functional, but we salvaged as many components as we could. Rather than painstakingly de-soldering chips from the PCBs, we'd take a blow torch to the back of boards, shake them and let the chips fall to the floor. The TTL logic chips always survived this brute force approach. With CMOS chips, it was 50/50. Our brain cells likely faired much worse, given that we ended up breathing toxic fumes from the burning boards. Of course, some would say we didn't have that many to begin with considering our choice of chip removal technique.
OK, so there's an apples and oranges thing going on with the specifications. For equivalent time sampling, the signal needs to be periodic (ie. no audio waveforms, well... maybe techno, hehe).
I looked at the Bitscope web site, and they claim a bandwidth of 100MHz but they sample at a maximum of 40MS/s. This doesn't even make sense by the lower limit of Nyquist, where you need to sample at twice the frequency. In this case, the bandwidth would be only 20MHz.
Am I missing something?
Slashdot Mirror
You're right about that. The good sustainers are usually the last ones left in the building, if only because products are usually sold with an agreement to maintain support for X number of years. Even if the product is ultimately a failure in the marketplace, the few customers who bought and deployed it have a contract that says they'll get maintenance releases for X years.
Unfortunately the higher-ups see that sustaining effort as purely a cost, so it's now one of the first functions they attempt to outsource.
The other thing about comments is that they can be funny. I remember a couple of bugs we tracked down where a failed hardware access resulted in an infinite loop. Sure enough, when we dug in to the code, that while() loop had a comment before it along the lines of
/* i should probably put a guard counter here */
Heh. And then I think of the pain and expense those bugs caused.
It's fun to hack away at home on personal projects, but professional efforts - especially big ones - really do require decent documentation and comments in the code.
Unfortunately my experience is that the requirements documents, design documents, and even the comments in the code are usually out of step with where the code actually is at a given time.
I've seen a lot of code written by the local genius that only that same local genius can maintain. That's not doing anyone any good.
How many times have you looked at a chunk of code and thought it'd be easier just to re-write it than spend the time understanding what's there in order to modify it?
Long story short, my vote is for well documented, well commented code. And keep it simple too. Unless you're under contraints that require you to get creative, just write it in a way that others will understand.
I agree that these guys have a right to some privacy. Most interesting to me is that the comments here on /. are generally supportive so far. What a different thread it would be if this had been Bill Gates and Microsoft instead of Steve Jobs and Apple.
Another excellent book that touches on what quants do is When Genius Failed, by Roger Lowenstein. This book charts the rise & fall of Long-Term Capital Management, a hedge fund that relied heavily on mathematical models to guide their trading activity. It's a cautionary tale about placing too much faith in mathematical models of markets that are not always rational.
Rasmus must have dual-citizenship. I'm pretty sure he's Canadian as well.
Technology like this is what forces American criminal organizations to outsource their counterfeting and ransom operations overseas. You're putting American criminals out of work!
The Britannica article does not present the theory as a scientific fact. It uses the qualifier "probably," which means it is likely true but there is no conclusive proof.
Yes, Google is a business. They're successful because we - the owners of the eyeballs they attempt to attract - believe they provide the most relevant search results, queried from as much of the web's content as possible. If they're going to censor results, they should be honest about it and let their customers know that they are doing it. Their customers can then decide whether or not to continue using their service.
Google's been such a great service/product/company that most of us have let them happily in to our lives. We trust their search results. We trust their e-mail service. The worrisome part of stories like this is that with Google's penetration of the market, they're powerful enough to dictate what sites are seen by the majority of people. We know about Abu Ghraib, so if we search for images and don't find them, we know something is up. But what if we search for something we're not as aware of and no results are returned. We may just assume there was nothing there to find.
Let's hope Google's "don't be evil" mission statement wasn't just a trojan horse disguise to sneak in to our lives and slowly become "evil."
There's a link to images on the announcement. Follow that, and look for a pic of the da Vinci team. The caption describes them as "racketeers". Ok. :)
Back in high school local companies would donate their old - read: scrap - mainframes to our electronics lab. The computers weren't functional, but we salvaged as many components as we could. Rather than painstakingly de-soldering chips from the PCBs, we'd take a blow torch to the back of boards, shake them and let the chips fall to the floor. The TTL logic chips always survived this brute force approach. With CMOS chips, it was 50/50. Our brain cells likely faired much worse, given that we ended up breathing toxic fumes from the burning boards. Of course, some would say we didn't have that many to begin with considering our choice of chip removal technique.
TIP: If you buy these through Parallax there's a volume discount that cuts the price down to about $170.
OK, so there's an apples and oranges thing going on with the specifications. For equivalent time sampling, the signal needs to be periodic (ie. no audio waveforms, well... maybe techno, hehe).
I looked at the Bitscope web site, and they claim a bandwidth of 100MHz but they sample at a maximum of 40MS/s. This doesn't even make sense by the lower limit of Nyquist, where you need to sample at twice the frequency. In this case, the bandwidth would be only 20MHz. Am I missing something?