Thanks for the kind words. I agree about the tests. To elaborate on your point about going beyond C++: not a whole lot of truly mathematical code is written in anything other than C or Fortran. I see the reasons for that as threefold.
First, a library written in a "simple" language is easier for a domain expert to enhance, especially since most domain experts will not be particularly expert at OOP.
Second, libraries created by Fortran and C are relatively simple to link, and enjoy great platform independence.
Third, most mathematical code gains very little from OOP. Wolfram and Mathematica aficionados may make a case for LISP, very understandably, but inheritance is but infrequently helpful.
I see a lot of posts from non-mathematicians here. Most are correct and useful, but not necessarily fitted to your situation. As a fellow mathematician who has dabbled in this stuff over the years, I can think of a few projects where your particular skills may be helpful:
(1) Work on the GSL. This is the highest-quality numerics library out there. I'm not aware of particular weaknesses - it's really great - but I'm sure there's something. (2) Work on the open-source spreadsheets. These are basically OpenOffice, Gnumeric and KSpread. All three lack some numerical routines that could be useful. (3) Work on Scipy/Numpy, R, Octave or Maxima. The numerics routines in these libraries are not always GSL-quality, particularly special functions. The built-in optimization routines tend to be basic and sometimes light on features. Even translating unit tests between packages would be a useful, enlightening and somewhat interesting project that I believe no one has ever undertaken. (4) Work on an open-source optimization library, such as OpenOpt. (5) Performance/accuracy comparisons of existing spreadsheets, libraries, and programs.
Actually, Japan's military spending is the sixth most in the world, and not that far behind China's. Those constitutional clauses are observed mainly in the breach.
Everything on this earth heavier than lead (atomic number 82) comes from supernovae. And most of the other heavy stuff (heavier than iron) comes from them as well.
You're right that 1080p is a red herring. That is why if you read my post carefully you'll understand that I really care at the 1080i level. The AppleTV does not support that either, which is a significant limitation for those of us receiving uncompressed 1080i over the ether.
I had been hoping for (but not expecting) an AppleTV capable of working with 1080p (or at least 1080i) video. Instead, they went the other direction, making it tiny, less power-hungry, and cheaper.
Many folks will probably be happy with this one, but I would have liked something more DVR-like to go with my OTA antenna and MythTV/Plex setup.
For those keeping score at home, Plex just went to v9 and is supposedly even more awesome.
The global effect of a single country rejecting software patents is much less than, say, the effect of allowing decryption and reverse engineering tools a la DVDCSS.
New Zealand authors and users will be safe from lawsuits, but users who download and use code in countries where the patents apply remain vulnerable to lawsuits, despite not being the authors.
I also use both, quite heavily. Rpy2 is a real improvement over RPy, but I share your preference for allocating tasks to the two languages according to their strengths, and the rpy/rpy2 data structure representations are much trickier than I like.
I find myself gravitating to R when the builtins are useful (robust estimation, smoothing kernels), where I need to split and analyze data subsets (using by() and its cousins) or where I want to plot things. Python is the ticket for interfacing with C, talking to the outside world, parsing, and expression of simple mathematical models.
The function of the industry, from a broad perspective, is "efficient capital allocation". That is to say, making sure that resources such as personnel and materials end up going where they will be used best, such as making iPhones instead of Palms.
This is worth some fraction of society's output. My gut tells me 5% or so. However, the finance industry takes probably about 20% of the corporate profits in the USA. Hence it is about 4 times too large, and three quarters of us are clearly here inefficiently. You can make similar arguments about attorneys and government employees.
Why are we all here? Well, like nearly everyone else in the world, our career choice wasn't particularly altruistic. It's morally about the same as piling on as yet another web developer during the dotcom boom. Those web developers might have been better schoolteachers, and we quants might have been better algebraic geometers, but such career choices are truly altruistic. Society does not reward them lucratively, 'knowing' that the good feeling arising from being, say, a schoolteacher is enough to attract some people.
I know that I would probably be happier in a research setting. Cthulhu knows I tried. But when you're staring at 8 years of tiny postdoc salaries in undesirable locations, all while running like mad on the publishing wheel and competing against other brilliant people for rare tenure track positions, it looks like a pretty lousy life. Even sysadmins and web developers have it better.
I too (I especially!) would like to live in a society where the quants were researchers instead. And while we're at it, a society where the best researchers are rewarded a tenth as well as the median pro ballplayer or actor. But powerful forces work against it.
Your claim that the finance industry is corrupt is wrong, for usual meanings of the word involving sentient intent. Yes, the industry is bigger and more influential than it 'ought' to be, but that is an emergent property of this modern economy.
In case anyone's interested in what I think of the financial crisis, here's a synopsis. We had a classic asset bubble, as has happened throughout recorded history and probably before. This particular one was in real estate, which allowed more people to participate than one in oil (1973), stocks (1999), or gold (1979), etc. At the same time, this large and incredibly efficient finance industry greased the wheels, allowing the bubble to extend further in size, penetration, and time.
In the aftermath, much more money was lost in people's house valuations than was lost by the finance industry, GM, and the like. In the case of houses many of the losses were basically just erasures of paper profits. But because industrial profits are all paid out as dividends, salaries and taxes, the losses in the corporate world were more real. So despite being far smaller, these latter losses attracted all the news and attention, with the exception of occasional newspaper articles on foreclosures.
I use Plex and MythTV (+HDHomeRun tuner) on a Mac Mini. It's been very reliable, and I'm happy with everything I can do, including videoconferencing. My only wish is that I could play MythTV recordings from within Plex, but really the latest version of Myth is pretty nice to use. Oh, and that Myth on OSX would do AC3 sound passthrough.
It's worth noting that I almost never browse the web using this setup any more, because most of the websites with interesting content (Hulu, YouTube, Comedy Central) have already been integrated with Plex.
instead of one "grand ending" there are now more smaller ones. While reaching a big ending may be quite satisfying, not all players are able to achieve it with 20-40 hours of gaming time. So several small endings may help them to enjoy games more.
So, you're basically saying the new thing is tantric gaming?
MythTV, especially v0.22, is really an awesome DVR but I agree it doesn't do a great job of providing the rest of the media center experience. I've been happy, though, using Plex media center and Myth on a Mac Mini, which of course has no problem with DVDs. (Blu-ray is a thornier issue).
You can read a little more about my setup and experiences here.
I think the real modern tragedy is the encrypted channels on cable and satellite: no decent HD PVR solution is possible any longer for these media. My solution was to fire the satellite company and go broadcast only.
I have a simple suggestion that eliminates all the security risks you are worrying about: write an expiring power of attorney for your mom (or other trusted friend or relative). It will be cheaper and more reliable, and mom might even like to get the occasional phone call while you're backpacking across the continent.
My experience with configuring Myth was that it was very difficult to do on my Mac Pro (due to a bug in Myth that I think is now fixed). But setting up Mythbuntu (both frontend and backend) on my laptop literally took me only about 15 minutes! If you go with KnoppMyth or Mythbuntu, you can expect a pretty painless configuration process, just by accepting defaults for all but a few obvious screens.
Granted, this was using "easy" hardware in the form of an HDHomeRun, so I had no video capture card to configure. And my experience on the Mac Pro demonstrated to me that if you don't use a dedicated Linux distro, you will have to spend some serious time and effort.
Eventually, I put everything on a Mac Mini (running OSX Leopard) which has been working very nicely.
Perfect on a Mac Mini (+link to howto)
on
MythTV 0.22 Released
·
· Score: 2, Interesting
I set up an SVN snapshot of Myth on a Mac Mini about six months ago. I wanted to save power, so the Mini runs both the backend and the frontend. If you like, you can see a full description of how I did it. (The guide is out of date in the sense that I resolved jumpy playback issues by reducing the priority of commercial-flagging jobs.)
It's been wonderful. I get full HD video and convenient scheduling. I've had exactly zero crashes, and the automatic commercial skipping has been very reliable (maybe one mistake every 5 or 10 shows). I also really enjoy the ability to watch TV on any computer in the house.
Right now, I'm working here and there on integration with Plex because I'd like to have all media in just one interface.
I'd like to add one thing to the many great points above:
The United States has a higher-than-average GDP and wage structure. If you assume that science skills and language/culture skills are evenly distributed, then you would expect to find science fields disproportionately populated by immigrants.
The reason for this is that scientific skill translate across the cultural barrier much more easily than other skills. The theory of comparative advantage (the only important result in all of economics, some think), then points out that native citizens will tend to populate fields where their advantage over immigrants is largest. That's law, business, etc., and not science and engineering.
Another way to say this is that, even if Americans are better at science than the average immigrant, the immigrants still enjoy a comparative advantage in science, due to the smaller absolute difference in scientific versus cultural skills.
One result of this is that wages in technical disciplines will tend to revert to the world average far more quickly than wages in nontechnical disciplines.
Can I just contradict all the pessimists here for a sec? The only people I know who currently use Twitter are techies who use it to exchange systems support tidbits with each other. Things like "How do I make Cyrus IMAP listen on an alternate port?"
As long as the questions come with subsequent answers, Google will probably do a good job with this stuff. And don't forget the information value in any links provided in response tweets.
(BTW I don't use Twitter myself, and have never even tried it)
Having owned a Kindle 2, I am not a fan of the e-Ink screens. They break way too easily, especially for a device one hopes to travel with. I also found the contrast to be merely borderline acceptable for my 41-year-old eyes (fine in bright light, bad in dim light).
After my screen broke in a 2-foot fall from a coffee table, I was completely unmotivated to replace the Kindle. If you get one of these things, treat it carefully!
I would estimate the half-life of a Kindle subjected to the usual slings and arrows of class-to-class migration at 4 weeks. They don't talk about durability in the article, but I know from (tragic) first-hand experience that those e-Ink displays make 1990s LCDs look tough.
A Kindle is good as a travel reader of linear texts. For anything else, the contrast, fragility and slow speed make it highly inferior.
Slashdot Mirror
Thanks for the kind words. I agree about the tests. To elaborate on your point about going beyond C++: not a whole lot of truly mathematical code is written in anything other than C or Fortran. I see the reasons for that as threefold.
First, a library written in a "simple" language is easier for a domain expert to enhance, especially since most domain experts will not be particularly expert at OOP.
Second, libraries created by Fortran and C are relatively simple to link, and enjoy great platform independence.
Third, most mathematical code gains very little from OOP. Wolfram and Mathematica aficionados may make a case for LISP, very understandably, but inheritance is but infrequently helpful.
I see a lot of posts from non-mathematicians here. Most are correct and useful, but not necessarily fitted to your situation. As a fellow mathematician who has dabbled in this stuff over the years, I can think of a few projects where your particular skills may be helpful:
(1) Work on the GSL. This is the highest-quality numerics library out there. I'm not aware of particular weaknesses - it's really great - but I'm sure there's something.
(2) Work on the open-source spreadsheets. These are basically OpenOffice, Gnumeric and KSpread. All three lack some numerical routines that could be useful.
(3) Work on Scipy/Numpy, R, Octave or Maxima. The numerics routines in these libraries are not always GSL-quality, particularly special functions. The built-in optimization routines tend to be basic and sometimes light on features. Even translating unit tests between packages would be a useful, enlightening and somewhat interesting project that I believe no one has ever undertaken.
(4) Work on an open-source optimization library, such as OpenOpt.
(5) Performance/accuracy comparisons of existing spreadsheets, libraries, and programs.
some of the best commercial fishing in East Asia .... So while the talk about a fisherman is noteworthy, he's just a pawn
A pawn for prawn, as it were.
Japan doesn't have much of a military
Actually, Japan's military spending is the sixth most in the world, and not that far behind China's. Those constitutional clauses are observed mainly in the breach.
Everything on this earth heavier than lead (atomic number 82) comes from supernovae. And most of the other heavy stuff (heavier than iron) comes from them as well.
So we live among a lot of supernova remnants.
You're right that 1080p is a red herring. That is why if you read my post carefully you'll understand that I really care at the 1080i level. The AppleTV does not support that either, which is a significant limitation for those of us receiving uncompressed 1080i over the ether.
I had been hoping for (but not expecting) an AppleTV capable of working with 1080p (or at least 1080i) video. Instead, they went the other direction, making it tiny, less power-hungry, and cheaper.
Many folks will probably be happy with this one, but I would have liked something more DVR-like to go with my OTA antenna and MythTV/Plex setup.
For those keeping score at home, Plex just went to v9 and is supposedly even more awesome.
Form 1040 EEZ
== == == ==
Instructions: Fill in the form. Send payment in the amount listed in the final cell to
Department of the Treasury
Internal Revenue Service
Fresno, CA 93888-0102
== == == ==
What was your total income last year? $__________.___
The global effect of a single country rejecting software patents is much less than, say, the effect of allowing decryption and reverse engineering tools a la DVDCSS.
New Zealand authors and users will be safe from lawsuits, but users who download and use code in countries where the patents apply remain vulnerable to lawsuits, despite not being the authors.
Mod parent up!
I also use both, quite heavily. Rpy2 is a real improvement over RPy, but I share your preference for allocating tasks to the two languages according to their strengths, and the rpy/rpy2 data structure representations are much trickier than I like.
I find myself gravitating to R when the builtins are useful (robust estimation, smoothing kernels), where I need to split and analyze data subsets (using by() and its cousins) or where I want to plot things. Python is the ticket for interfacing with C, talking to the outside world, parsing, and expression of simple mathematical models.
Armstrong and the other astronauts got to walk on the moon. What do I get for billions of dollars thrown at more human spacetravel? Nothing.
I'll take the robots and the science instead, please.
I'm a quant in the financial industry.
The function of the industry, from a broad perspective, is "efficient capital allocation". That is to say, making sure that resources such as personnel and materials end up going where they will be used best, such as making iPhones instead of Palms.
This is worth some fraction of society's output. My gut tells me 5% or so. However, the finance industry takes probably about 20% of the corporate profits in the USA. Hence it is about 4 times too large, and three quarters of us are clearly here inefficiently. You can make similar arguments about attorneys and government employees.
Why are we all here? Well, like nearly everyone else in the world, our career choice wasn't particularly altruistic. It's morally about the same as piling on as yet another web developer during the dotcom boom. Those web developers might have been better schoolteachers, and we quants might have been better algebraic geometers, but such career choices are truly altruistic. Society does not reward them lucratively, 'knowing' that the good feeling arising from being, say, a schoolteacher is enough to attract some people.
I know that I would probably be happier in a research setting. Cthulhu knows I tried. But when you're staring at 8 years of tiny postdoc salaries in undesirable locations, all while running like mad on the publishing wheel and competing against other brilliant people for rare tenure track positions, it looks like a pretty lousy life. Even sysadmins and web developers have it better.
I too (I especially!) would like to live in a society where the quants were researchers instead. And while we're at it, a society where the best researchers are rewarded a tenth as well as the median pro ballplayer or actor. But powerful forces work against it.
Your claim that the finance industry is corrupt is wrong, for usual meanings of the word involving sentient intent. Yes, the industry is bigger and more influential than it 'ought' to be, but that is an emergent property of this modern economy.
In case anyone's interested in what I think of the financial crisis, here's a synopsis. We had a classic asset bubble, as has happened throughout recorded history and probably before. This particular one was in real estate, which allowed more people to participate than one in oil (1973), stocks (1999), or gold (1979), etc. At the same time, this large and incredibly efficient finance industry greased the wheels, allowing the bubble to extend further in size, penetration, and time.
In the aftermath, much more money was lost in people's house valuations than was lost by the finance industry, GM, and the like. In the case of houses many of the losses were basically just erasures of paper profits. But because industrial profits are all paid out as dividends, salaries and taxes, the losses in the corporate world were more real. So despite being far smaller, these latter losses attracted all the news and attention, with the exception of occasional newspaper articles on foreclosures.
I use Plex and MythTV (+HDHomeRun tuner) on a Mac Mini. It's been very reliable, and I'm happy with everything I can do, including videoconferencing. My only wish is that I could play MythTV recordings from within Plex, but really the latest version of Myth is pretty nice to use. Oh, and that Myth on OSX would do AC3 sound passthrough.
It's worth noting that I almost never browse the web using this setup any more, because most of the websites with interesting content (Hulu, YouTube, Comedy Central) have already been integrated with Plex.
You can see my setup log here: http://public.boonstra.org/MacMiniHTPCSetup.html
As you probably already know, there do exist certain costly counterexamples.
So, you're basically saying the new thing is tantric gaming?
MythTV, especially v0.22, is really an awesome DVR but I agree it doesn't do a great job of providing the rest of the media center experience. I've been happy, though, using Plex media center and Myth on a Mac Mini, which of course has no problem with DVDs. (Blu-ray is a thornier issue).
You can read a little more about my setup and experiences here.
I think the real modern tragedy is the encrypted channels on cable and satellite: no decent HD PVR solution is possible any longer for these media. My solution was to fire the satellite company and go broadcast only.
Madman curie sounds like the name of a stunt motorcyclist
or a stunt physicist (whatever that is)
I have a simple suggestion that eliminates all the security risks you are worrying about: write an expiring power of attorney for your mom (or other trusted friend or relative). It will be cheaper and more reliable, and mom might even like to get the occasional phone call while you're backpacking across the continent.
My experience with configuring Myth was that it was very difficult to do on my Mac Pro (due to a bug in Myth that I think is now fixed). But setting up Mythbuntu (both frontend and backend) on my laptop literally took me only about 15 minutes! If you go with KnoppMyth or Mythbuntu, you can expect a pretty painless configuration process, just by accepting defaults for all but a few obvious screens.
Granted, this was using "easy" hardware in the form of an HDHomeRun, so I had no video capture card to configure. And my experience on the Mac Pro demonstrated to me that if you don't use a dedicated Linux distro, you will have to spend some serious time and effort.
Eventually, I put everything on a Mac Mini (running OSX Leopard) which has been working very nicely.
I set up an SVN snapshot of Myth on a Mac Mini about six months ago. I wanted to save power, so the Mini runs both the backend and the frontend. If you like, you can see a full description of how I did it. (The guide is out of date in the sense that I resolved jumpy playback issues by reducing the priority of commercial-flagging jobs.)
It's been wonderful. I get full HD video and convenient scheduling. I've had exactly zero crashes, and the automatic commercial skipping has been very reliable (maybe one mistake every 5 or 10 shows). I also really enjoy the ability to watch TV on any computer in the house.
Right now, I'm working here and there on integration with Plex because I'd like to have all media in just one interface.
I'd like to add one thing to the many great points above:
The United States has a higher-than-average GDP and wage structure. If you assume that science skills and language/culture skills are evenly distributed, then you would expect to find science fields disproportionately populated by immigrants.
The reason for this is that scientific skill translate across the cultural barrier much more easily than other skills. The theory of comparative advantage (the only important result in all of economics, some think), then points out that native citizens will tend to populate fields where their advantage over immigrants is largest. That's law, business, etc., and not science and engineering.
Another way to say this is that, even if Americans are better at science than the average immigrant, the immigrants still enjoy a comparative advantage in science, due to the smaller absolute difference in scientific versus cultural skills.
One result of this is that wages in technical disciplines will tend to revert to the world average far more quickly than wages in nontechnical disciplines.
Can I just contradict all the pessimists here for a sec? The only people I know who currently use Twitter are techies who use it to exchange systems support tidbits with each other. Things like "How do I make Cyrus IMAP listen on an alternate port?"
As long as the questions come with subsequent answers, Google will probably do a good job with this stuff. And don't forget the information value in any links provided in response tweets.
(BTW I don't use Twitter myself, and have never even tried it)
Having owned a Kindle 2, I am not a fan of the e-Ink screens. They break way too easily, especially for a device one hopes to travel with. I also found the contrast to be merely borderline acceptable for my 41-year-old eyes (fine in bright light, bad in dim light).
After my screen broke in a 2-foot fall from a coffee table, I was completely unmotivated to replace the Kindle. If you get one of these things, treat it carefully!
I would estimate the half-life of a Kindle subjected to the usual slings and arrows of class-to-class migration at 4 weeks. They don't talk about durability in the article, but I know from (tragic) first-hand experience that those e-Ink displays make 1990s LCDs look tough.
A Kindle is good as a travel reader of linear texts. For anything else, the contrast, fragility and slow speed make it highly inferior.
It's a solar sail, right? It can just luff up.
(kidding)