Ask HN: Constant impulse to create something, yet never act? | Hacker News Reader

Ask HN: Constant impulse to create something, yet never act?

3 pointsendorphine12 comments

Quite often when I see other projects on GitHub or other Show HN posts, I feel an impulse to "create something" (as in, software). Then I think to myself: "I should sit down and write a toy DNS server in language X, as a means to learn about DNS while also learn the language".

I especially feel this when I see small toy/pet projects or projects that are built throughout a long period of time (i.e. curl). For the latter case, I admire the dedication some people put in their passion projects, building/evolving/maintaining it over a long period of time (again, curl is a prime example).

However, despite these thoughts, while I do sit and try to learn a new language, I never start a project like this. But on the other hand I constantly feel this impulse to "create something", even if it's a toy project.

Does anyone else falls into this endless loop of unsatisfied creativity impulse? If so, have you found any way to tackle it somehow?

Well outside of work and more as spare time working on side projects or simply as a hobby, I think it depends on one's energy and if other things in life are getting in the way. When we're young, and a full time job not the most pressing thing on one's mind, such exploratory activities are generally spontaneous, we were a bundle of energy and enthusiasm and sheer determination to get the less than best coding or idea in general to work ... it wasn't uncommon for the young enthusiastic programmer to find their day fades into night and back into dawn and it's hardly been noticed ... days soon becoming weeks. When we get older, we've often got other life issues, commitments, bills to pay and as such many of us unconsciously self impose a structured order to the day, spare time being more to get over the day, to relax and not to tax the brain too much - and often we're so stuck in our thinking, the rigid structure remains even if one has found themselves out of work. However (not that I had much luck back in the day) I've seen over the years a couple of people push themselves to practice taking the time to pick something and spend the day working on it until it was just something they'd do, and did so without concerning themselves what else they had to do or were meant to be doing, they just ploughed right in.

This resonates so much with me. Most of the time I find that I'm just in this automatic mode of operation of working->family_time->sleep->back to work for 5 days a week. By weekend, I'm so tired that I just want to relax by listening to music or walking.

However, there are some rare moments when I have much more energy and willpower, and I push through and really am onto something, until sleep deprivation catches up to me again.

I guess some people have more natural energy and determination than others, and probably I'm on the lower half.

I'd love to try and make it so that I don't subconsciously fall into this mode of operation when I really need to be creative.

I think one reason is that the internet makes it seem like things are easier than they are. You can browse the internet and in one afternoon see a hundred projects that each happen to be someone's life's work, and you are seeing them as a result of survivorship bias that has filtered out every project that was bad or unlucky or made by a normal person. The lizard part of your brain makes you think you can do a project like one of those ones in an afternoon even if you know better.

That's so true and a great reminder to be mindful of survivorship bias. Thanks!

It’s probably because the projects you have in mind are too big and so you don’t know where to start.

Don’t try to make a “DNS server” right off the bat. Make a program that is hardcoded to respond to a query for the A records for a single, hardcoded domain. Do this first before you give any further thought to “DNS server” functionality. Use the “dig” or “drill” command line utilities to make the query, and use Wireshark to inspect the the resulting request and response from an existing DNS server such as 1.1.1.1 or 8.8.8.8. Then make it so that your program can respond similarly to requests from dig or drill, for some single domain of your choosing. Test it a few times to make sure it works. And only after that should you continue with thinking about what to do next for the DNS server.

Constant impulse to create something on which I always act.

The result is a myriad of half finished things, some now wildly out of date, and most difficult to pick up again (ie undocumented). Also not a very productive place to be.

I used to be like you. Still have dozens of hardware and software "prototypes" languishing in drawers and random directories.

I'm addressing the problem by making sketches on paper (yes, very old school) and writing rambling "design" documents before and during my work. That way, I am better able to focus and if I do abandon a project, I have some scraps from which to recollect my original thoughts should I attempt to restart a project after I have gained some applicable experience or knowledge by working on some other idea.

Yes, I was only today thinking I might reuse an old esp8266 thing I made and I thought - how did I even program that? And I would have been on Windows then, so how do I do it on Linux? I hope it will come back!

I was accutely annoyed by the planned obsolescence of some critical (for my computer usage) software components. I started to write heavily derived/alternatives directly in assembly (I realised that even lean and simple C was not enough in those cases). This de-facto cancels planned obsolescence based on compilers and some language syntaxes but the work is beyond reasonable.

I don't know why, I am so mad about it, I am doing it anyway.

Sounds fascinating, what are some examples of the things you are building?

I figured out only recently that lean and simple C was not enough, namely, no real assembly stuff published yet.

In lean and simple C I have some (easy to port to not gcc and not clang, namely tinycc or cproc/qbe etc).

Since it is a big component (for the wrong reasons), I will show it on HN when it is actually working and published, coze it is so much toxic to write (it is not the assembly which is toxic, but the standard on which it is based is completely obsolete/broken/toxic) I may send everything to hell and drop it. I guess I have still months of coding, a lot of them.

My solution is: Just do it!.

By starting something small and incrementally adding to it, learning as I go along.

For example, I was curious about multi-colored LED lights. So I bought a strip of WS2812 LEDs, an Arduino. Downloaded Arduino IDE and some examples. Bit by bit, I learnt C++, some electronics, need for external power, etc. Now have several "works" that respond to sound, rejuvenated antique lamp, etc.

Reply on news.ycombinator.com