glimcat

I wish I had more than two choices at 4 mbit. You can pick the phone company or the cable company. Neither is very good, and neither cares - they don't need to.

Heck, I wish I had one choice where I could call tech support, tell them that I've been getting dropped packets and 3000 ms pings for over a week, and not have them blow me off because the bandwidth tests fine.

The hierarchy of slides goes something like:

    Relevant images which support the speaker
    Relevant graphs, rendered for projection
    Relevant microcopy which gives context or focus
    Relevant videos which are kept as short as possible
    Bullet slides which double as effective microcopy

    Bullet slides which double as speaker's notes
    Most slides which animate in point-by-point
    Irrelevant or unnecessarily long videos
    Unreadable diagrams
    Unreadable walls of text

The zeroth rule of presentations is don't waste your audience's time.

"Asserting that AP classes are not equivalent to classes where he went to school is unhelpful. Mr. Tierney fails to mention that he went to Yale, making this anecdote disingenuous at best."

I did my first two at a community college. While I don't doubt that you can find some high school courses equivalent to or superior to their college analogues, this is not the prevailing reality.

But such positive outliers have little to nothing to do with the AP brand, and much more to do with the course's teacher. They certainly have nothing to do with AP testing for equivalency credit, which occurs after the fact. AP testing is a shoddy placement tool which frequently and provably tends to overstate students' preparedness for advanced studies.

In STEM fields, anyway. Freshmen distribution survey courses don't do much to prepare you for anything particular, even at Yale.

Maybe it's the engineer in me, maybe it's the astronomy geek who watched two shuttles explode.

But I am way more impressed with "engine failed, still got to orbit safely" than I was with the already titanic feat of making it to orbit in the first place.

For any problem with optimal solution k, a superset of that problem formed by relaxing the rule set will have an optimal solution at least as efficient as k.

It comes up in AI a few times.

The fastest way to grok memory issues is to do some work in Assembly. Few programmers will use Assembly for regular work, but the intuition it fosters will serve you everywhere.

This is all very good advice which will save you much stress and many wasted hours.

I would add:

Document any reference material you use, including the source and why you're including it. Cache any digital content, either in the project path or using a management tool like Zotero.

Keep a research log. Minimally, annotate your trials. Coming back even a week later and trying to figure out which trial was done on what hunch with what results is extremely time consuming without this information.

Users make my head explode sometimes, even when I'm expecting it.

I've caught users recommending a product of mine (feels awesome), then they turned around and insulted it in the same conversation. Upon a little anonymous interaction, they were upset because the CSS didn't match Google. Users, WTF???

Thank god for A/B testing.

I'm thinking that some monkey software buried in the human psyche is interpreting giving someone money as an act of submission, while another bit is assuming that free is a promise of all-you-can-eat and thereafter interprets limits to that as cheating.

Someone with actual professional psychology experience could probably shoot that full of holes.

I refer to it as persona-based pricing

I hereby declare this to be the correct technical term for it. The guys in marketing may be calling it something else, but I don't think you'll find a more succinct description of the strategy.

I've been known to follow up a couple days later with a concise report on some issue they mentioned having in-house.

It seems to be better received than null-content pings.

Images are up now on the raw feed. They're getting F5ed hard, so actually seeing anything may take a bit.

http://mars.jpl.nasa.gov/msl/multimedia/raw/

Images are grouped by Martian solar days (sol 0 == day 0, length 24h 39.6m).

One major thesis of MSL/Curiosity is that sample return missions are hard. Taking a jaunt over to Mars, collecting samples from various locations, then shipping them all back to Earth is not something we can accomplish right now - so Curiosity has been designed as the next best thing.

Take a look at the instruments section of that Wikipedia article. Curiosity is crammed with every sensor and analysis tool they could possibly fit.

http://en.wikipedia.org/wiki/Mars_Science_Laboratory#Instrum...

I started, well, pre-kindergarten. Or in my teens, for anything major and self-guided. Or at about twenty for producing professional results. Or in my late twenties, as someone who could do research-level algorithms work. Or today.

If I stop using a language for a while, I have to relearn it. The term language is very appropriate - learning and retaining a programming language acts very much like learning and retaining a human language.

My take-homes about "when to learn programming" are:

* If you have kids, expose them to "Hello World" programming early. By which I mean, once they start reading. The cognitive benefits of multilingualism are greatest at early ages when neuroplasticity is high. There are numerous languages and projects which are suited to this. Lego robots should probably feature prominently once they're of an age to not eat the pieces.

* You're never too old to learn programming. The biggest jump in capability by far is between non-programmer and someone who has been at it seriously for a few weeks or months.

* At the same time, start learning programming as early as possible. Learning to be a great programmer takes a lifetime. But very little of that is learning syntax! Great programming takes marketing, art, math, psychology, and whatever else you can cram in. Pretty much any field of human endeavor has something to contribute. "What to program" is far harder to learn than "how to program" - and people who start learning programming later in life will have a lot more to draw on here than some 12 year-old.

I'm still very nervous about the landing system. It's ambitious, to say the least.

But if it works - oh, if it works! This really is the most exciting project in exogeology since men brought rocks back from the moon.

(Table of localized landing times: http://timeanddate.com/worldclock/fixedtime.html?msg=NASA+Cu...)

I think they appeal to different audiences.

He's also effectively selling preferred access to classes, which is something of an ethical issue.

And many universities handle waitlisting on a department or class level so they have leeway to deal with various factors as appropriate. Ever tried to implement university, departmental, program, and class policies simultaneously, while keeping them up-to-date, while handling who can override the computer under what amalgam of policies? No? Well, that's what you'd need to do in order to get automated waitlisting working at most universities.

The fact that there's not a university-level waitlisting feature isn't an excuse to hack around policy, especially not while violating ToS and misappropriating resources for commercial resale.

I had to can a project recently after spending over a month fruitlessly trying to get a response from Google's support resources.

On the up side, I made $70 and learned some interesting stuff.

But here's yet another way for Google to run false positives which send you into their black hole of non-support. So one of the things I'm still learning is "don't let Google be a mission-critical part of a project."

After being at this for a while and being successful enough to get a label account on iTunes.

I'm sold, although I'd still start by sitting them down with a copy of Feynman's QED. When doing math about a thing, it helps tremendously if you already have some sort of intuition about what the math is describing.

Also, the historical stuff is fun and interesting. Physics students should probably be taught this stuff regardless, but maybe not as part of a class on theory.

MRI is mathematical voodoo even when it's working - it's easy to fool yourself at any of several stages in the process. "New scanner gives 10x improvement in detail, shows amazing and unexpectedly simple result" is just screaming for verification.

But it needs it obviously enough and the claim is interesting enough that further verification (or falsification) should happen soonish, as research timescales go.

> You have to work on a problem you know exists.

At the same time, experience can often mislead you into thinking that your problems are everybody's problems. Experience gives you a head start, but you should still do customer research to confirm the "traction potential" of your proposed business model.

It depends on what you want for your career.

Either on a resume is just saying, "Hey look, I have a bit more education than the next guy in the pile!" - and little else. It can be a marginal advantage there, but the effect is often negligible compared to networking.

What's more important is acquiring skills and experience. Do you want to learn more advanced CS theory, or do you want to start learning business theory? Keep in mind that you can choose one and still take classes from the other.

Also keep in mind that an MBA is really intended for corporate middle management. If that's the job you want, go for it. If you're more interested in startupville like much of the population of HN, then pursuing practical experience with product launches and building a business may be a more efficient use of your time.

Rule 34 likely applies.

Good as a functional prototype. Sensor placement would be tricky when moving towards production due to different hands and grips.

I have a hammer, but I have no clue how to get from driving nails into boards to building houses. I don't know what the road map should even look like, or what skills I would end up needing, or how to evaluate them along the way.

Thus why software engineers often get a degree, buy books, pay for classes, and so forth.

The self-education threshold tends to be even higher at the beginning. Most people have no clue how to get from owning a computer to Hello World, to FizzBuzz, to simple scripts, to a well-designed and documented program.

The real failure is when people turn to university (etc.) to learn a skill, get a "QC Okay" stamp at the end - and then still can't FizzBuzz.

You need shorter feedback cycles which focus on finding and correcting errors, not monolithic course grades and monolithic degree pass/fail stickers. You wouldn't expect it to end well if you tried code a major project that way. Why would it work any better as an education plan?

The time cost of signalling and unit computation is fundamentally physical. Drift and error in inertial and position sensing is fundamentally physical.

You can make very good goggles if the money and time per unit isn't an issue. Registration to the point that there are no detectable flaws is very hard. You have a hard timing constraint from the human visual system on the order of 5 to 50 milliseconds. In that time, you need to sense the visual environment, determine position and its derivative vectors, figure out how your simulation needs to be updated, render the new data, and display it to the user.

Not something you can necessarily solve by throwing money and engineering hours at it until it works. In fact, there may be a few Nobel prizes in physics between here and there, or possibly biological integration well beyond our current horizon.

I think we're still waiting for the first big thing.

Right now, you don't need to get rich. You need to get some of the pressure off and get stable. If you can reasonably land a six-figure job which you can stand doing for a few years, DO IT.

Desperate and looking for get-rich-quick schemes is not conducive to successful entrepreneurship. Most successful businesses take years to build, and most business ventures are not successful. You should consider whether it can be a part of your ongoing career plans, but it's not a likely solution to your immediate problems.

Student loans tend to be ineligible.