"Is that a real program or is that something somebody wrote?" | Hacker News Reader

"Is that a real program or is that something somebody wrote?"

lists.canonical.org

187 pointsasciilifeform31 comments

A memory walk. Recognizable in many ways. Some lessons are applicable to more than one stage in career and life in general ("I didn't need teachers to teach me things; I needed teachers to force me to get out of my comfort zone, and partners to get me unstuck"). A good read.

I agree, that particular line struck me as the most interesting lesson in the whole thing.

I think what people do is help with inspiration (in the broad sense). Other than that, we can read books, play with source code etc.

Wow! Thanks for that!

Not only was that a great read that I can empathize with, but I got a ton of great book suggestions out of it.

I'm assuming that was written by the same kragen that frequents here... Kragen, if you read this, well done!

Yes. Thanks!

Really great read. I started programming BASIC and Logo at a young age as well, but I don't recall the details nearly so clearly as the author, and my path was quite different taking a massive detour through HyperCard (and HyperTalk!) then web design and finally back into programming by way of a CS degree, PHP, Ruby, and the desire to actually write decent Javascript.

Although I think working through an MIT-based CS curriculum allowed me to short-circuit a lot of the early challenges he mentioned, the "learning to read" part might be my next step. Pair programming with a team of better programmers sounds like the perfect career development move.

A fine, honest article. The best part was "learning to read", and that happened recently in the author's cronology. One of my friends tells me that "if you can read, you can cook" to which I retort "then, if you can read, you can program". The answer to that is "no, not so". And I know that is not exactly what the article meant. But good reading habits are important.

In the sense I meant, I think that if you can read, then you can debug. Cooking is more like executing a program than like writing it.

I've never read "Algorithms in C" before. I'm trying to track it down in Amazon, but it seems like there is a paperback called "Algorithms in C" by Sedgewick, as well as a 5-piece bundle also called "Algorithms in C, parts 1 - 5" by Sedgewick. Does anyone know if the former is just part 1 of the latter?

I believe the latter is just a later, expanded edition of the former. I do know that they are all part of a series that he has been revising for two decades, starting with "Algorithms" which was in Pascal, then branching out to have editions in C and Modula-3, then revisions in several volumes/parts. The first edition of Algorithms in C was just a single volume. The paperback you mention is probably just a copy of that edition - if I'm looking at the same one as you, it looks like a paperback of the hardcover I remember from fifteen years or so ago.

[redacted]

In stores, I've usually seen parts 1-4 in a single volume.

(Note that these are the Indian LPEs (Low Price Editions) by Prentice Hall India.)

Parts 1-4 are in one volume, and then part 5 (graph algorithms) is big enough that it gets its own volume. I own the parts 1-4 book, and it's very pleasant and readable. The pictures of the algorithms in action really help make the concepts clearer. (They're also a cool hack; the author wrote implementations of the algorithms with a bunch of printf calls to spit out PostScript code, which used a high-level graphics API that he wrote, in PostScript, for the purpose of making diagrams of running algorithms.)

Algorithms in C is a classic, Algorithms in Foo is, sadly, not. Sedgewick's text, much like Appel's compiler text, is a one-language book with a bad gettext translation for other programming languages.

Read Sedgewick in C and Appel in ML. Skip the C++, Java, and C# versions of both books, they usually pop up in bookstores at the end of summer.

The bit I've skimmed of Algorithms in C++ seemed like it did a pretty reasonable job of using C++ templates effectively; I don't think it was a bad gettext translation.

An excellent read, especially for people considering going into software engineering. It gives a good overview of the life cycle of a "typical" software engineer.

I hope my life cycle didn't end in 2007!

Heh, ok, life cycle up to maturity perhaps?

I don't think I was mature before I had written http://canonical.org/~kragen/urscheme/.

and it again rises the question where one stands in this path.

I can for see a future when this question will actually be competent, perhaps when AI can write programs (sort of) in a commercial setting.

Interestingly, kragen is on here too: http://news.ycombinator.com/user?id=kragen

I wonder if he's seen this yet, and what he's done since this post in '07.

Just saw it. The last couple of years have been pretty busy; some of the product is on kragen-tol and kragen-hacks and on github, but by no means all of it.

What's the next phase? What else does he have to learn?

Embedded systems. Distributed computing. Real-time systems programming. Software reverse engineering. Compiler construction. Digital signal processing. Machine learning. Natural language processing. Operations research. Numeric computing.

Give each a decade and you will be in hacking bliss for the rest of your life.

Yes. Eg. operations research can be quite rewarding from a math point of view. Esp. combinatorial optimization (as a subfield) has a lof of interesting algorithms and concepts.

I can't claim to be an expert in any of these, but I have dabbled in many:

Embedded systems: http://canonical.org/~kragen/light_sensing/

Software reverse engineering: http://canonical.org/~kragen/demo/klappquadrat.html

Compiler construction: http://canonical.org/~kragen/sw/urscheme/ http://github.com/kragen/peg-bootstrap http://github.com/kragen/stoneknifeforth

Machine learning: http://angstro.com/ (this isn't solely my project; I'm only doing a minority of the work)

Numerical algorithms, digital signal processing: a bit at ERIM

I can't say I've done anything very interesting with distributed computing, real-time systems programming, natural language processing, or operations research.

I have dabbled in all, mastered probably none.

An older friend of mine made millions from an RTOS he wrote in the 80s. He also wrote his own tcp/ip stack, cross assemblers, built his own ICE debuggers and other stuff. However, he knows very little else :-P He is almost always surprised by "what computers can do". Specializing did something strange to this man; he does the impossible and remains ignorant of the trivial. He only programs in C and Assembly languages but was surprised to discover there were syntax-highliting editors for assembly. He knows most algorithms that we toss around in conversation in chip form; to him MPEG, AVI, MD5, and SHA1 are all silicon things, never "written". I taught him what a "packer" was for executable binaries.

OT:

kragen, Opera de-hearts your website: says you destribute malice :-(

http://i27.tinypic.com/2s6svap.png

wtf?

I heard about that a few weeks ago. I think I need to talk to Opera about it. Just in case, I grepped for the obvious things, without any success.

Uh, I mean I searched to see if there were any obvious browser exploits being propagated from the pages. I can't rule out the possibility of something subtle but I didn't notice anything.

Reply on news.ycombinator.com