1. Spend less time on the internet - I spend 2 days of the week without internet access so I can finish my projects in peace
2. Use as little technology as possible to finish them. I find having a limited tool-set actually lets me think about the creative part instead of working with too many tools and being distracted learning more tools.
3. Don't become attached to projects. If it doesn't feel natural, discard it and work on a new one. I regularly finish projects, but I also have a large pile of discarded ones.
2. prototyping is often about allowing oneself to stitch things in the ugliest possible way.
Right now my current "project" is building a stock of photos so that I can post a continuous stream of photos on social
https://mastodon.social/@UP8/tagged/photography
This is pretty consuming because my developing process is intensive. I took about 3500 photos of two basketball games on Jan 31 and picked out about 400 to develop in DxO, I chose one color grade for all of them because a lot of the girls had bad acne that I wanted to minimize, now I am cropping them and applying local microcontrast to the parts in focus (but not faces of athletes with acne) I expect this take a few evenings, then there are many more batches of photos to develop including other games as well as tranches of flowers and landscapes. I have probably 15k undeveloped photos where those came from.
Another stream of that work is building an autoposter which will make it easier to keep these coming, for a while I was posting to Mastodon + Bluesky but lately I am posting to Facebook and LinkedIn, want to add Instagram (hate the UI for uploading) and Threads and anything else I can. That's blocked by my work to finish my Python API for using Postgres as an arangodb-like document database which blocks a whole bunch of projects that both involve moving forward, like a general purpose trainer for text classification and getting YOShInOn (my RSS reader) off arangodb. Feels overwhelming at time but the train is moving.
https://en.wikipedia.org/wiki/Kanban
My RSS reader is ‘active” in the sense that I use it every day but I haven’t touched the code in months. I’ve been using Kanban control for side projects since maybe 3-4 years ago when I started doing side projects again. At the beginning of the year I had a lot of projects that were more programming heavy like I wrote a Python chess program that could beat my son that didn’t have time control, I did some work towards a Java chess program which would have had time control so we could take it to the chess club but my son quit going to chess club so I mothballed it.
The photo thing is converging towards transitioning from semi-pro to pro, figuring building my rep as a ‘social media photographer’ will get me work. I have other imaging projects in mind such as something to view stereograms with WebXR and something that can compose individual flowers to make ‘infinite flower fields’ that could be printed on fabrics but these are at the ‘pile of notes” phase waiting their turns.
For me, I commit myself to work on it every day except vacations. I use a kanban board for tracking work and just do things that feel natural. When I'm grinding out a big story, sometimes I switch and just focus on small bugfixes or random ideas to rebuild the motivation.
My other side project is much less worked on. But when I have a need in the software, I just focus on it for a week and document everything in Github issues. Usually the needs are quite complex and the fun of solving and getting a nice solution is fun enough to help drive me through.
I only work on things that I have a need for and I think that helps a lot with the mindset of finishing it off.
Too many of my early projects sort of became a jumble of different aspirations, most of them unrelated to the core product idea behind the side project. I want to learn Elixir, I want to try Next Auth, I want to try Remix, I want to learn Haskell, I want to build my own Shopify, I want to -- and the list goes on. Being clear and honest with myself about what those things are makes the scope of a project more clear to me and I can make judgment calls this way on what stays or goes.
I've started to do things like just email my list and explain that I'm considering making something. what questions or concerns would they have about it. And I consistently ask my email list if there's something I can improve about something I already launched.
The Book Launch by Jeff Walker helped put into plain terms the actual questions and how I go about pre-launching something I haven't made yet. It's helped me build about three or five new courses very confidently because I had been asked a bunch of questions from people who would go on to take that course.
Obviously you're not going to complete entire projects in two hours, so you cut scope, while ensuring some version of "the thing you want to do" is possible.
So for insane forms with dozens of fields, you start with one field.
Then you iterate.
The idea is that at any given point outside of those two hours, your project can be used. It doesn't do ALL the things, but just enough to be useful.
Did this approach work consistently, or were possibly other factors at play?
has been 4.5 years straight for this one project, doing two hours on workday mornings
So I just won’t start building as many. If I find the engineering part interesting enough I might just build the part that’s interesting to me. That typically means I skip the front end and only focus on the core business logic.
In the past i used a simple 'build-release.sh' whereas lately i tend to use Github actions or other CI/CD workflows.
Another habit is to focus on stuff that you need or want for yourself and if you are in the flow finish it, then switch to other requests. As soon as I Start with something that feels boring but was often requested, I lose interest very quickly and get distracted.
I also got better at validating ideas. Again, I take longer before getting excited and promising anything.
Not long before every project was a way to learn new technology and I quickly lost interest once "the hard part" was done.
People generally envision something that is more akin to a product than a prototype, and one thing money is good at is that if you have enough, you can get far closer to that vision faster by dumping it.
Don't focus on what is still missing. Instead, focus on one or two features that can be made more useful.
I'm able to take my ideas to a good stage and actually do the justice now. I'm not limited by my own skill/time/patience.
Just the usual HN nonsense. Shruggie. I appreciate calling it out!
Here are things I tried to do:
1. Having a sounding "board" person.
2. Taking only 5sec to help you decide.
3. Post-it the thought/idea/notion all over the house to make focus.
4. Talking it through to yourself. Using yourself in a 2/3rd person.
I need all of this...