HN Explorer

Ask HN: Any “useful” side project ideas?

25 points by tonmoy ↗ HN
At my work I get a couple of weeks of low intensity work for every 4-6 months when I have about 20 hours/week I can spend time on doing side projects or maybe learn something new. In the past I have taken on multiple projects that couldn’t be finished in 2 weeks and when I come back to them 4 months later I would have to basically start from scratch due to lose of flow hence it would be difficult to get motivated. Is there anything useful I can do that may be helpful for the world in these two weeks. I’m not too concerned about putting stuff on my resume. My primary goal is to be useful and secondary is to have fun. While I feel I do have a pretty good knowledge of algos and data structures, being a hardware engineer (RTL, HDL etc) my software skills, especially knowing the latest frameworks are not the strongest for sure.

16 comments

[ 1.6 ms ] story [ 33.8 ms ] thread
jacquesm
That depends, really on whether or not you will be around to support them after the launch, throwing something out in two weeks usually works well if it does not see adoption, then it's just a way to train your brain. But if it does see adoption then you have a responsibility towards your users and in that gray area where it requires serious time but not enough to live off you will end up harming your reputation and your users. I usually decide against side projects unless I'm 100% sure they will not become a burden and my users interests are safeguarded.
tonmoy
While I don’t know the future for sure, if I see even some users are getting benefit from my work, I will be disciplined enough to spend at least 4-5 hours a week to support my creation. In fact I am not opposed to the idea of starting to support an open source project
jacquesm
Neat! Much good luck, I'm curious what you will end up doing. I launched pianojacq.com in very much the same spirit, and a lot of the design was influenced by those considerations. So far it has worked out very well.
akg_67
What are your interests outside the tech domain? Try applying your tech domain knowledge into something in your area of non-tech related interest. It can be as simple as monitoring energy usage in your home or as complex as automated smell removal and seepage from garbage can. Do a side project for yourself.

I like to study new stuff, mostly mathematics/quantitative oriented. When I am reading a book on a topic, I try to figure out how I can implement the things explained and written in the book.

tonmoy
I think that is another problem I have, I find everything interesting when I first start. However when I start doing it I end up not keeping interest once I start digging too deep
johnebgd
I just started soldering again. It’s been 20 years.

I hope to use these skills to save some retro game systems otherwise destined for a landfill.

thinkingemote
Depends on what you mean by useful for the world.

One response could be volunteering for a Wikimedia tech project, another for a charity etc. Another may be to offer tech support to the old people in your town.

Figure out what kind of thing you'd like to help first and then look for projects. Or figure out the type of work you want to do and then look for openings.

samsquire
Check out my profile links. If you want data structures ideas check out ideas4.
mod
I spent about 20 or 25 hours building a website to republish some county data that people frequently check. It's been a big hit with my local friends.

Another one I will probably do soon is a script to check weather & water conditions in my area to tell me if it's a good day to go fishing. Namely, I'm going to check the water level of the local river via the USGS, and see if it's within a certain range that are ideal conditions.

I think a good value add for me would be some home-automation style voice commands, but with [whatever I dream up] types of outcomes. Like an actually-useful set of commands added to Alexa, let's say.

everforward
> I think a good value add for me would be some home-automation style voice commands, but with [whatever I dream up] types of outcomes. Like an actually-useful set of commands added to Alexa, let's say.

There are some projects in that space already. Mycroft is one, but it seems slow going and didn't support the 8GB Pi4 when I tried. There was another that just did voice to text and fired webhooks with some of the data; I can't recall the name but it seemed perfect other than needing at least 2 services running (the voice part and your backend).

If you mean that you don't want to write the code, I think HomeAssistant has some voice integrations. Haven't tried it though, so I don't know how good they are or what features are available.

gameman144
I'm curious what the county data in question is? I'm a big fan of local-government data, so this is a super interesting niche to me.
mod
I'd rather not mention my county for personal privacy reasons, but I will just say the data in question is arrest/bookings/mugshots. The county uses some really unusable software to manage a "roster" of inmates, and to see arrests you have to catch it before someone is released. Even then, it's just highly difficult to use.

Here's a county that isn't mine, using the same software, so you can see what I mean:

https://www.dallascountyarsheriff.com/inmate-roster

People are dropped from the roster immediately upon release.

One of the main users of my site is a friend of mine who is the crime editor in the local paper. It helps him to not have to sit and refresh the page all day.

kebsup
One formula you could use is: take public ugly data, display it in nicer way. For example in my country, there is a law about data that government has to publish, but going through monthly excel sheets is not very practical.
jawmes8
I recently took up Golang and had a great time building an IRC client. I was never around when IRC was bigger, but I enjoy hanging out on it now. Kinda reminded me of how HAM radio guys must feel about building their own rig.

This book helped a lot: https://quii.gitbook.io/learn-go-with-tests/