5 ms·

> I realized that my work was technically obsolete as hell, and not really of any value. For all of the benefits of the internet, I see this as a tragic side e

by tspike 12y ago

> I realized that my work was technically obsolete as hell, and not really of any value. For all of the benefits of the internet, I see this as a tragic side effect. Rather than just trying something and seeing if it works, the first instinct is to Google it to see if anyone else has done something similar. Even if you manage to start on something, eventually you'll be exposed to someone who's done it so much better and more thoroughly than you and get discouraged and give up. Of course, plenty of innovation still happens and people can find inspiration in superior implementations, but the loss feels similar to the loss of biodiversity, or language diversity resulting from globalization -- that is, isolated populations create vastly different solutions to the same set of problems in ways that aren't possible without those barriers.

LnxPrgr3 12y ago
Experience is worth something, even if it comes from toy projects. The downside of "lazy learning" is you can't know what tools you'll need in the future. I was able to win at one of my favorite work projects because I'd played with signal processing and machine learning for fun. Had I not, I would've concluded I couldn't beat the terrible solution we had—I wouldn't have had the tools to know better. And of course, your reinvented wheel might turn out to be an actual improvement!
- anon4 12y ago
  I would say that lazy learning is a pretty big boon overall for both you as a person and society at large. On one hand, you have niche skills that can be applied professionally, and on the other, you can find people with knowledge in some niche subject that they can develop on the job pretty quickly. You will never not find an application for some bit of knowledge.
  - LnxPrgr3 12y ago
    Apologies for the late reply, but that's precisely my point: if I only did lazy learning, I would've missed awesome opportunities. Of course, I can never know everything. I do both: I learn things as needed to get something done, and I learn things just because they fascinate me and might prove useful later. That latter part—gaining knowledge for the sake of knowledge—has been an amazing benefit, specifically by giving me a better map of what's possible, and how the seemingly impossible might be done. IOW, I consider lazy learning necessary, because no one can know everything, but not something I'd encourage for its own sake, because sometimes you have to know something before you can know when it'll be useful. I absolutely encourage playing around just to learn more about your craft—you'll learn things that will prove useful later.
- vegedor 12y ago
  To reinvent basically means repair, but reinvention of the wheel also implies that nothing was broken in the first place.
pm90 12y ago
I actually see that as an advantage: instead of reinventing the wheel, you're using others work as a basis of something that has probably got more functionality and is more useful. I do agree that this does mess with tinkering with the fundamentals and stuff. But here's my thesis: when you work with a technology for a long time, after a while you will naturally be curios about how the lower level stuff is working. I think this is why a lot of application developers still find systems programming exciting. And also why the hacker news audience is also keenly interested in stuff besides the very narrow focus of cs/ee.
- AnthonyMouse 12y ago
  > I actually see that as an advantage: instead of reinventing the wheel, you're using others work as a basis of something that has probably got more functionality and is more useful. It's really both. The trouble is you want e.g. a protocol to secure your communications, and designing one yourself while making it secure is a huge ordeal, whereas you could just use TLS for free even though it's full of warts. So everybody uses TLS. Which then snowballs into everyone using OpenSSL for TLS even though it's ancient and full of cruft. The problem is you do want everyone using the well-reviewed long-tested thing, but then that thing becomes a decaying designed-by-committee monstrosity which nobody can fix because it would break compatibility. See also HTML, HTTP, DNSSEC, javascript, etc.

by tspike 12y ago

LnxPrgr3 12y ago

Experience is worth something, even if it comes from toy projects. The downside of "lazy learning" is you can't know what tools you'll need in the future. I was able to win at one of my favorite work projects because I'd played with signal processing and machine learning for fun. Had I not, I would've concluded I couldn't beat the terrible solution we had—I wouldn't have had the tools to know better. And of course, your reinvented wheel might turn out to be an actual improvement!

anon4 12y ago
I would say that lazy learning is a pretty big boon overall for both you as a person and society at large. On one hand, you have niche skills that can be applied professionally, and on the other, you can find people with knowledge in some niche subject that they can develop on the job pretty quickly. You will never not find an application for some bit of knowledge.
- LnxPrgr3 12y ago
  Apologies for the late reply, but that's precisely my point: if I only did lazy learning, I would've missed awesome opportunities. Of course, I can never know everything. I do both: I learn things as needed to get something done, and I learn things just because they fascinate me and might prove useful later. That latter part—gaining knowledge for the sake of knowledge—has been an amazing benefit, specifically by giving me a better map of what's possible, and how the seemingly impossible might be done. IOW, I consider lazy learning necessary, because no one can know everything, but not something I'd encourage for its own sake, because sometimes you have to know something before you can know when it'll be useful. I absolutely encourage playing around just to learn more about your craft—you'll learn things that will prove useful later.
vegedor 12y ago
To reinvent basically means repair, but reinvention of the wheel also implies that nothing was broken in the first place.

pm90 12y ago

I actually see that as an advantage: instead of reinventing the wheel, you're using others work as a basis of something that has probably got more functionality and is more useful. I do agree that this does mess with tinkering with the fundamentals and stuff. But here's my thesis: when you work with a technology for a long time, after a while you will naturally be curios about how the lower level stuff is working. I think this is why a lot of application developers still find systems programming exciting. And also why the hacker news audience is also keenly interested in stuff besides the very narrow focus of cs/ee.

AnthonyMouse 12y ago
> I actually see that as an advantage: instead of reinventing the wheel, you're using others work as a basis of something that has probably got more functionality and is more useful. It's really both. The trouble is you want e.g. a protocol to secure your communications, and designing one yourself while making it secure is a huge ordeal, whereas you could just use TLS for free even though it's full of warts. So everybody uses TLS. Which then snowballs into everyone using OpenSSL for TLS even though it's ancient and full of cruft. The problem is you do want everyone using the well-reviewed long-tested thing, but then that thing becomes a decaying designed-by-committee monstrosity which nobody can fix because it would break compatibility. See also HTML, HTTP, DNSSEC, javascript, etc.