etrain

Adaptor combinators. Because they're functional.

And, by the way, the two commands are something that could be done in O(n), rather than O(n*log(n)) - but this little procedure is so damn easy to write, that on relatively thin inputs that are less than 20m lines long, I usually just do this.

sort | uniq -c | sort -n

is something I use all the time to get sorted frequency tables.

Basically. And that's a good thing.

Controversial views are a necessary but not sufficient condition for visionary status.

To put it another way - if a visionary isn't controversial, she's probably not a visionary.

He's also an ACM Fellow, from it's first class - along with people like Knuth, Cerf, Rivest, Codd, etc.

Emptiness.

In this situation, I did not receive compensation because of my employment agreement with my then-current employer. That was my choice, and I'm sure we could have figured something else out.

I have seen candidates handle it differently, though - one potential hire we paid for a 3-week engagement as a contractor, for example.

Ah, didn't realize that - throwback to my xargs days! Thanks for the tip.

I realize it's a one-off, but you could potentially mix lines of output with a script like that.

If you gnu parallel (xargs++) installed - turn that second line into a script (and replace $line with $1).

  cat domains.txt | parallel -n 1 -P 50 script.sh

I will tell you my favorite hiring trick, by far. Before I do, though, I'll set the scene. I was looking for a new gig a few years back. I was mostly happy at my job, doing interesting stuff, not killing myself, and making good money, but itching to do something new. Wanted to get in the startup scene, etc. So I see a new company's job posting, and it looks like a perfect fit (location, mission, everything). I meet with the CEO, have a few conversations with investors and employees, and things are looking good. Then they throw me a curve ball - "We want you to do something interesting with our product." - I spent the next two weeks working my ass off nights and weekends. Mind you, not compromising my full time position in any way. In the end, I did something interesting, and it turned out to be a great thing for the company. More importantly, it let me get to know their product deeply, and I understood the kind of environment I'd be entering.

The moral of that story - when hiring, challenge people to do something for you. Give them a week to work with you. Pay them even. Just see how they actually work. It'll work well for you, and them - even if your rate of hiring is lower than you'd like.

As we move toward languages with big standard libraries, productivity becomes more and more about knowing what you don't have to do yourself.

Case in point - even as a Python veteran, I've never seen/used collections.Counter before. It'll save me from writing the same stupid three lines over and over, and is probably implemented faster than I naively do. I'd say that this will improve my productivity.

The tails.

The premise that "information gathering" is a manager's most important task is a pretty interesting assumption underlying this post.

I recommend reading Andy's book "High Output Management" for more insight into this assumption.

Good to see some constructive advice on how to configure mongo, instead of just bashing it.

Even if it's not your favorite technology, sometimes you end up in a position where the rest of the company is using something, and you need to work within those constraints. It's important to understand the technologies you're building on, their configuration options, and to understand the best practices way of working with them.

This, by the way, is not restricted to mongo.

The author points out something about these Machine Learning contests and Machine Learning in general that I've noticed for a while - feature selection tends to dominate learning algorithm selection. It's good to see that there are modern academic methods for feature discovery that seem to be on par with (or better than) a domain expert manually selecting features.

I'm pretty stoked on Go. I freaking love C. Love it. It's like they took C and went to the best disciples of C's creators and said "look - we've got these problems now nobody envisioned 40 years ago. Can you make C for us again, but better?" And they said "yes!" and it was good.

We get some nice concurrency primitives, garbage collection, cleaner syntax, something between structs and objects that fits the right feeling, automatic bounds checking, cute array syntax, and a big-ass, well defined standard library. Oh, and this concept of interfaces that is so well executed it's not even funny.

Except. I feel like they are forcing the fanboy mindset. At one point in this slide deck, there is the following bullet: "The designs are nothing like hierarchical, subtype-inherited methods. Much looser, organic, decoupled, independent."

I didn't see the talk. But that is the most vapid, meaningless description I've ever seen of a feature of a programming language. Rob might as well have said, "it's hipster better," which would have conveyed exactly as much meaning.

So here's my question - and I hope there are real answers - can someone point me to >3 real, big systems that are built using Go? I'll accept Google internal systems on faith.

We subscribed to Newsweek throughout my childhood, even living overseas. I recall excitedly going through it every time it arrived. As far as the news digest format went, in those days they mixed topic areas fluidly, while keeping their readers informed.

I canceled my personal subscription years ago. To me, the format changes had made it less comprehensive, the quality of writing had declined, and it had become too political. Or, maybe I'd just outgrown it.

Sure, they'll have a digital version, but they have far smaller staff (and impact) than they once did. They're a marginally iconic publication, who were unable to adapt to the new world, and are thus a victim of the changing landscape of journalism.

Another classic is his article on solving sudoku puzzles.

http://norvig.com/sudoku.html

Though languages with constraint systems built in have even nicer solutions (e.g. Oz)

I had a very similar experience about a year ago. Except instead of running out of RAM on a 32-bit instance, I was running out of disk on a 64-bit instance. That's right, my database ran out of disk, and the driver didn't throw an exception.

Yes, I realize that there's a "safe=True" option to my python driver. But I'm writing to a database. As others have said here and elsewhere, the default behavior of a database and its drivers should be to complain loudly when a write fails. It is ridiculous that safe!=True by default. If I want to turn off this feature to improve performance, I will.

The signup/confirm/login metaphor is really tired and annoying. Don't give me two screens of meaningless text, show me what your product does and let me interact with it before making me sign up. For example, open the strategy builder window up to anyone, but make me sign up before I can save my strategies.

On the product itself - security specific rules are fine but you're missing the point of backtesting and automated trading. The ability to spot patterns across any security (or perhaps in an industry) is the key here. I wish there were some way to specify this. I realize it's v1, but your rules are too simple for most traders.

Finally - you should call out that your market data is EOD. You should also offer VWAP (I know Xignite offers it) as a reasonable alternative for prices you "could have gotten" as a retail investor.

Also, the IP thing scares me, and I wouldn't use the service until you change it.

Want to comment just to stress that this feature is EXTREMELY useful and has saved me from all sorts of tricks with file intermediate output with filenames that match process ids, etc.

Serious question for phk: Despite your rant, why do you continue to use autotools in Varnish? Is it because libraries that Varnish depends on require autotools? Or did you simply follow convention?

The startup game has tons of formally educated people, and their failure rate is just as spectacularly high, if not higher. In my experience, the formally educated tend to build "science projects" at a higher rate than businesses, when compared to those building "twitter for teens"

Love the data and the site. Would be fantastic if you'd open up an API to this data to make it easy to ingest and analyze in external tools (R, Excel, whatever).

It would be easy enough to scrape it, but formal support for data export is such a huge win for sites like this.

Really strong support for code reviews.

I don't mind the rtomakyo dotfile style. (https://github.com/rtomayko/dotfiles)

I haven't implemented it myself - mostly relying on standard tools as the author does. Vanilla bash, vim, etc.

The other trick I like to employ (if I'm in control of the network) is have every machine on my network mount a central (NFS) home directory, with all my dotfiles in it. As long as that's backed up, and configured on every machine, it works very nicely.

"Certainly by the time you read this." Code isn't live yet, at least for me, ericd - is this conversation giving you pause about deploying the new feature, or is it standard ops issues?

Also, this couldn't come at a better time for me. I love padmapper and am moving to the bay area in 3 weeks, I need this, man!

Yeah, Wrangler will turn your "manipulation" steps into scripts in Python/MapReduce/Javascript.