Databases. Oracle, Microsoft, etc. have figured out a lot about how to make high-performance query execution engines and transactional storage systems and written about very little of it.
Research has caught up some, but it definitely lags.
6,154 karma · joined August 7, 2008
hn@danieljackoway.com
I'm happy to receive emails from HNers.
Research has caught up some, but it definitely lags.
I have some experience encouraging groups who tend to be less sure of their abilities in an academic subject to study that subject and am friends with many people who are much more involved in these efforts, and I know that being confrontational is counterproductive and that you have to be very careful about how you suggest to them that they have a "problem", even if you are telling them that they can overcome that problem. You want to stay positive and show empathy toward their struggles, emphasizing that the struggles are not any failing or inherent shortcoming on their part and are very common.
patio11 talks about this a lot. Here's a podcast where he talked about it (more in the context of B2B software, but many principles still apply): http://www.kalzumeus.com/2012/09/21/ramit-sethi-and-patrick-...
If you scroll through his comments, you'll see many instances of him telling people to charge more for things and explaining some of this stuff news.ycombinator.com/threads?id=patio11
I like the end, where you emphasize that math is a great exercise in abstract thinking and modeling. But I would start with something more neutral saying something like, "Many people had suboptimal experiences with math in school because it was often presented well, and some teaching styles don't work for some people. And some may have just always found other things more interesting. However, my goal with this book was to make math more accessible and interesting to many people by presenting material in more a more intuitive manner and emphasizing the core skills of abstract thinking and mathematical modeling of the real world."
That's definitely not perfect, but it's something to start with. Try to imagine that you're someone who never found math all that interesting and always thought that literature or history or art was far more interesting and meaningful, and that possibly you had some bad experiences with math and got the idea that you're not good at it. Now try to imagine what might trigger those bad feelings about math and turn you off to a book, versus what might make you decide to give it a chance. If you don't trust your ability to imagine this way, try to find a close friend who is in this situation and will be honest with you.
(Not trying to sound confrontational--it's just not a space that I thought had much more innovation left in it, so I'm curious to hear what you've come up with.)
If this were implemented, for probably 90% of votes that he casts, we would see very low turnouts with voters coming from two groups: special interests--the small group of citizens who are deeply affected by the outcome of the vote and thus find it worthwhile to vote; and political junkies who get satisfaction out of voting on every issue, but who probably don't actually have time to really be informed about every single bill, especially as they evolve.
Implementation would also be extremely challenging. How do you ensure that Oregon residents (and which residents?--those who are eligible to vote? Those who are registered to vote? All, even children, felons, etc?) can vote once, and all others cannot vote? How do you handle people whose credentials for voting (whatever they may be) are compromised? What happens when a bill is amended--do people who have already voted on it have their votes stand by default? Are they able to change them? If they're able to change them, is that at the expense of anonymity? Do you actually have anonymity anyway? How does the Senator vote if someone DoS's the voting system? And obviously if there are any security vulnerabilities in the system, lots of problems could arise. These are very hard problems; there's been quite a bit of research into evoting, yet it is still very much an unsolved problem. If someone wins on a direct democracy platform, these problems will become very apparently very fast.
That said, this adds complexity to their systems with the only benefit being cost savings. Given that we can assume that no code is perfect, it's likely that at some point the auto-downscaling will cause an outage or period of slow responses, which could easily lead to lost usage and trust that costs them as much as they're saving on ops.
All the failover problems I can think of since they moved to this architecture 4 years ago have been coordination problems where something undesired happened when transitioning from one member of a pair to another. In this case, network problems lead them to a state where both members of a pair thought they were the master.
1% downtime is over three days. They've had some big outages, but I think this may be their longest of the year and it was a < 6 hours. They could have one of those every month and still only have 72 hours of downtime, which is 99.18%.
The tricky thing is that with hacking, the line between piecing together the vulnerability and exploiting it doesn't involve as much clearcut intent and harm.
There's no way that in a post that explores the limits of technology, he's actually trying to make some claim that he missed a huge red flag in an online dating algorithm's output, nor that we should interpret the situation in that way.
Cell-phone carriers have formed an oligopoly due to the extremely high cost of building nationwide service and the limited amount of mobile spectrum. This will naturally lead to a situation where the industry can stagnate and overcharge with there being very little pressure from new entrants to the market that try to exploit the market failure. Carrier-lockin makes this problem even worse because they make it even more expensive to use their services unless you lock yourself in longterm in exchange for a phone subsidy that amounts to a discount.
Sure, there are lots of concerns about Google having lots of information about you. But at least you have choices, and at least you don't have to pay Google $100/month and pay them more if you want to leave.
Besides, whom does it really help for the news of layoffs to get lots of play? If there's a salient case for this move being sleazy, it only holds up if he's actually wronging some party in the process.
The only argument I can really think of is that the more other employers that hear about it, the easier time the laid-off have finding jobs, but that's a very tenuous argument to make
This makes total sense; we don't really know what direction Coursera and Udacity are headed as businesses, nor even what online education is going to look like or who the students will be. For Stanford, having the option to cut ties with Coursera is critical to Stanford's goal of being a leader in online university-level courses. And the platform is open source, meaning that depending on what happens, it could become a standard that more and more schools adopt.
And I, like you, think this is good for everyone. The more different the players are this early in the game, the more likely that at least one of them ends up doing it right.
So you're probably still looking at an hour and a half, but maybe not quite as bad as you describe.
Of course, this isn't a problem if you just work in SF too, and there's more and more great companies in the city. As awful as Muni is, it should offer you a <= half hour commute in most cases.
To make Unity and the DisplayPort port (as opposed to just VGA) work, I had to go into the BIOS and change a graphics setting to make it not automatically switch between the integrated and better graphics card. This made unity work, but it also made something terrible happen when I actually plugged in a monitor with DisplayPort. I believe that both the laptop screen and monitor went to 1024x768 and couldn't be coerced out.
The next step, to get full resolution on our 30" monitors, was to install some different video drivers. That was going to be harder to reverse, so I just gave up and dealt with nonoptimal resolution via VGA.
I also had a terrible time disconnecting monitors. Half the time when I unplugged the VGA it would just give me a blank screen.
Also, a few times when listening to music, the entire machine froze and it played the last ~2 seconds of audio repeatedly, with no sign of escape besides restarting.
We've come a long way from when several hours of work to make wifi work was a given, but I still don't think I've ever owned a Linux laptop where all hardware worked perfectly.
Care to give me compelling reasons to change?
I'm probably taking Networking and Machine Learning for credit. I'm excited that McKeown is teaching networking for the first time since I've been around (not sure if he ever did). He's done some really interesting networking research and usually teaches the more advanced networking class, which I've heard great things about.
So I think it will be an expensive second-class citizen for the forseeable future. It's not in their interest to change that.
From Amazon's perspective, a great Kindle is something that you use extensively, carry with you everywhere, and buy many books on. There are very few books that Amazon sells that are seriously helped by the DX's larger screen, save maybe for someone who needs to crank the font way up.
The DX, I assume, is great for PDF's with formatting/figures that won't convert well (which the small Kindles are pretty bad at). But Amazon makes no money on you reading PDF's on your DX.
A DX with a great touch screen and a stylus could also make an outstanding replacement for paper that you write on. But again, Amazon won't make money for each page you scrawl.
Add to that the fact that the market for the DX is presumably significantly smaller, and you get where we are now. Amazon charges a lot for the DX because they need to actually profit from it, instead of relying on content sales. Amazon doesn't have a big incentive to improve the DX because it doesn't make them much money.
Maybe in the long run, they'll get around to it, especially if technology does get to the point where you can make really good lines on e-ink. Then maybe they'll push for the Kindle to be the digital piece of paper that everyone owns, and oh by the way, now that you can buy content from your paper, maybe that will be some very profitable paper for Amazon.
But right now, the incentives don't align.
I don't think that 50k is competitive for any developer that they would be willing to hire, regardless of where they're located or how much they promise the equity will be worth in a few years.
Edit: Didn't even realize that it said you had to relocate to SF. For SF, one of the most expensive cities in the US and the one with by far the most competition for developer talent, 50k is not competitive under any circumstances (It's barely livable.), and even 100k isn't very competitive for anyone with even a little good experience.
Just one example: I work for Cloudera. As I understand it, we raised a fairly large round when we had almost no employees or revenue, and weren't exactly sure what we were doing. And we absolutely needed to. Hadoop wasn't mature enough for most businesses yet, most businesses weren't really looking to do the things that Hadoop is letting them do yet, and the company was nowhere near mature enough to know how to support enough big enough customers to pay the bills anyway. We needed time, so we needed funding.
3+ years later, everything has matured and there will certainly be no talent acquisition.
More generally, your claim doesn't seem to make much sense. Why does raising money make a nascent company more likely to sell out for a small talent acquisition? The real reason that companies that raise early sell more than companies that raise late is that the set of companies that raise late is a biased sample. It is, by definition, a set of companies that got too big and successful to be talent-acquired.
In the current landscape, early companies that haven't yet started printing money may well find that their best option is to take a fairly large check and work for someone else. I can't see how investment would make them less likely to get past that stage; investment gives them more time to figure out their business, and less incentive to sell because investors will get a cut and not really want them to.