Ask HN: What are some hard coding projects I could create? | Hacker News Reader

Ask HN: What are some hard coding projects I could create?

19 pointsdoubleponk20 comments

So I want to push my coding skills. I've spent more than enough time on coding challenge sites now and implemented enough datastructures/algorithms. I've developed enough simple Saas sites. I want to start doing something useful and challenging.

I've thought about writing a database, but the world really has enough databases already. I've messed with compilers, but the world has enough languages out there already.

I have no itches out there I feel I need to scratch, I just enjoy coding.

What are some other challenging projects out there that aren't already over-saturated with implementations?

View on HN

You are welcome to hack on postgresql! From making queries faster (better plans, computing on graphic cards, column-stores,compiling queries) to making better the storage engine (column store, tokudb-like stores etc) to new type of indexes to new type of queries (ex increment jsonb fields even if they dont exist) to sharding it (and even different types of sharding (pg_shard, pg-xl)) to lowering the overhead (each row has 22bytes of overhead!) to so many different things that are listed here: https://wiki.postgresql.org/wiki/Todo.

Must the hard part be about coding?

There are plenty of projects with hard challenges. I'm working with live videostreaming using WebRTC and I've come across plenty of problems I don't want to touch with a ten-foot pole - my approach is exactly opposite yours, I want to go along the path of least resistance and leverage existing infrastructure as much as I can.

The Mozilla team welcomes contributions, maybe you can ask to help on the WebRTC team?

https://www.mozilla.org/en-US/contribute/

Also Chromium or the WebRTC team:

http://www.webrtc.org/contributing

Yes. Coding is what I enjoy and I am getting bored of writing stuff that no longer challenges me.

Thanks for the suggestion, that is a great idea! I will look into them.

If you want a challenge get into computer vision projects like image processing, image enhancement or video processing work etc. There are some hard problems to solve. You could start by picking an existing algorithm that people like to use but it takes too long to process and see how you can make it better.

Or if you want a change, write firmware, drivers or other low level code that is tough to do right and most require significant design, thought and testing. IoT projects all need this type of work.

Another idea, work on distributed computing problems, they are still ripe for innovation and problem solving. If you want a real challenge, combine all three of my points.

I personally moved away from web development overall as while I get it can be fun, I am just tired of it. I like doing the backend and solving scalability issues, distributed issues, but designing, writing and handling web forms etc is just not anything I care to do on a regular basis anymore.

At the same time on a non-technical level, I am personally expanding on my marketing work as I have always done really well there so it is fun to help others with it. But you mentioned coding specifically, so I know you are focused there, but you might keep an open mind to what you can do outside of it too for a challenge.

You should learn a new language, a weird one that isn't like others you have used, e.g. if you have only ever done object oriented stuff learn a functional language. I spend some of my free time just learning new, mostly impractical languagess, and I have found that new ideas come through from the text at hand, or from thinking about problems in a different fashion.

I suggest J, Erlang, and Idris. The first two have been around a while and have a good set of larning utilities. Idris on the other hand has a little more stuff going on, but learning about dependent type systems is fun. For me at least.

J: http://www.jsoftware.com/help/learning/contents.htm

Erlang: http://learnyousomeerlang.com/

Idris: http://docs.idris-lang.org/en/latest/index.html

I'm not so sure - I'm already languages junky. There comes a time when you have to stop learning new languages and start using what you know to create stuff.

Idris is on my list already, but J actually looks really interesting. I've only really seen it used in Code Golf before. Might have a little play.

Image segmentation under occlusion. Email me if you'd like to discuss: alexander.s.farley at gmail.com.

Thanks for responding. Yes, a lot of interesting stuff there. Looks like a great rabbit hole to fall down!

So, you want to push your coding skills...to what end? Become more hire-able? Make (a little) more money? Make something to show HN? Once you dig deep and find your reason why you want to push your coding skills, optimize toward that.

Fortune and glory - if I'm being completely honest.

But do you value fortune more, or glory more? King or rich?

https://gigaom.com/2007/12/10/do-you-want-to-be-rich-or-be-t...

http://hbswk.hbs.edu/item/5543.html

Are Kings poor? They seem pretty well off.

But I digress, that's just a naive interpretation. It seems like the articles are more about issues of boundaries and control of the company.

From the original poster's comments, I thinks he values being recognized and technical archievements over founding a VC-funded crappy SaaS and becoming rich :) , and he might be looking for a well-paid technical position where he can work on hard problems. I might be wrong :)

A system that auto-generates requests for hard coding projects perhaps?

Cocos2d-x game engine always needs help: https://github.com/cocos2d/cocos2d-x

do you want to make money, help the world or just code something hard? I think the first two (with emphasis on the second) are(should be) way more important to a coder.

All 3!

I want to challenge myself and become a better coder so I can create stuff that is useful for people, which in turn would land me a better paid job.

maybe pick an open source project to start fixing bugs?. the challenge* is how many you can fix and get accepted by the maintainers.

*I realize this is not a purely coding oriented challenge. but it seems a worthy one if you're bored and are looking for something challenging do. also, bugs fixed + accepted counts toward resume fodder

how about trying to make a project that provides some service and figure out how to automate getting new users for that project. That is quite a challenge

strong AI.

Reply on news.ycombinator.com