Ask HN: Those making $500/month on side projects in 2022 – Show and tell

It's definitely something I am having in the backlog, but I cannot promise if and when it will be implemented.

I was curious... from what I can find online the wholesale price of an e-ink display is not that much cheaper (if any) than buying an equivalently sized Kindle. What is the viability of a business model that involves rooting a Kindle, loading whatever calendar display software you need, and shipping it inside a pretty wooden frame?

There seem to be 8-inch e-ink displays for sale on AliExpress for $20-$40, actually much cheaper than I expected. No idea about the quality though.

https://forms.gle/tNcCcYrNBu5nWKgJ9

Multiplex is actually nice since it's cross laminated and thus retains its shape. I experimented with solid wood and it started arching after a few weeks.

Any idea what could be causing this? I am at a loss.

I’ve done a lot of research on CE/FCC and while it seems possible to do CE yourself (since you can self-certify) you are on the hook if you miss something, like certain required tests (and doing some tests can be very expensive).

CE mark isn't actually required in the US, but you'd need it for Europe and other locations. It's more involved, but all-in testing can be done for <$5K for EU if you're careful.

> Did you have some investor or paid from your own pocket?

> Asking as somebody who thought making some embedded / HW projects, but the initial cost seems to much to be paid by myself.

Crowdfunding is how it's done for HW products. Investors aren't going to be interested in anything small time (less than $10-100mm potential revenue + recurring subscriptions) unless they're friends and family or something like that.

It's a lot of money, but it's not out of reach for someone with a tech job who uses crowdfunding for the major production push.

Especially worth noting how things can snowball after several months of consistent shipping.

You're not just starting from scratch every two hours, you've already got work to build on top of, to keep you growing.

I'd recommend the book Traction by Gabriel Weinberg for ideas.

It's Next.js, so React + Node.js, written in TypeScript.

I have different experiences to others on the market, thus I resonate with some more than others.

If you check out https://nopixel.hasroot.com/, they maintain a list of all Twitch streamers currently streaming nopixel.

nb: I have no affiliation with nopixel.

Roleplaying games are really great for exercising your social skills and creative expression though, that's for sure.

- Participating in the web dev communities

- "Powered by Linkz.ai" footer in the Link preview popups

- Campaigns on Product Hunt, Reddit & similar

- Soon: Affiliate, LTD, targeted ads for Webflow/Squarespace ecosystems

Technically rich link previews save visitors from the tab-overload.

I've got so much more that I'm afraid to publish. Might have to reconsider.

"Make decisions that will give you more choices, not fewer."

I was the founding engineer and Head of Eng at Reforge the past 4+ years while I was building Closing Credits. I left in August 2021. So, it was a side project for nearly 5 years.

If I have 3 side projects and no full time job at this exact moment, where do I stand? I'll delete my post if I'm violating the side project rule.

Tech choices: I never reinvent the wheel. I just take working pieces from other work that I've done and glue it together. Anything custom, I'll read how others do it.

Lessons: I probably should've chosen a different market. If I had targeted companies and taught their employees professional education rather than poor amateur voice acting hobbyist, I'd probably be making $20M ARR. But I don't mind, this is still fun.

Twitter is a perfect example: removing "some" of your tweets is purposefully made cumbersome for users so that they get discouraged to manage their tweets. Searching through your tweets by date range or by keyword, and a button to delete all results? "Nah, too complicated for us silicon valleyers" :)

One thing to remember with GDPR is that it is not a law that protects customers. It is a legal framework that specifies a set of requirements, which companies must abide to in order to do whatever they need/want with your data. Once you understand this, GDPR becomes much clearer :)

Good problem to have I suppose ^^

The monetization model is just referral links to Namecheap, where I get a 10% commission. I want to make that a bit more elegant (especially for people with uBlock Origin, which it doesn't track), and also add a few other referrals (logo makers and maybe hosting).

Couldn't think of other ways to monetize this without making it obnoxious (I hate ads, and making it pay-to-use also seems restrictive to me). If you have any ideas, I'd be open to hear them!

Revenue breakdown is roughly equal between android, ios and web, somewhat surprisingly. Android converts worse but has higher user numbers. Web converts much worse, but converts at a higher price (justified by the fact that hosting/maintaining the web stuff take a lot more time and money)

When I started the site, I mainly used CC0 licensed sounds others had recorded.

Then I started recording my own sounds. How much work it is is very situational - if you regularly find yourself in an environment which has the sound you want to record, and not many other sounds around, then it's pretty trivial. For example, you want to record rain in the forest, and you regularly walk in a forest where it rains and there aren't many other noise sources (e.g. other people, planes overhead, singing birds, etc). The actual recording itself doesn't take much work, because I shoot for a level of sound quality that will satisfy 80%-90% of people, rather than a real "audiophile" quality level.

On the other hand, if you want to record something that only happens occasionally and with lots of other noise sources nearby, it can be a ton of work. For example, you want to record the sound of thunder, but you only get occasional storms, you live in a city with lots of other background noise, and it usually rains when it storms and you want rain on the recording. In that scenario, you might have to travel far and burn a ton of time trying to get the right conditions for recording.

Today, HR teams usually post jobs on LinkedIn and perhaps one or two more platforms. A world of fragmented job boards would be difficult to navigate for non-specialists.

https://cushychicken.github.io/grow-your-mailing-list-by-bei...

We're seeing a steady and growing trickle of organic traffic, too.

I see that name pop up on all sorts of queries and I'm shocked it's not in the db. Digging into that presently. Thanks for the suggestion.

Similar story with csv imports, it's half supported but not in the UI yet.

The defunct social network Orkut is a good example of this. It's Órkut, not Urkut.

Of course, none of this matters :)

Q: I am a developer, Do I need this API?

A: No, if you can build your own rendering system with all these features and able to make it run this fast. You don't need this API.

It's a very early draft, but hopefully provides a glimpse into the techniques I use.

(tl;dr: it's way more low-tech than ML generated art.)

This is strange; I don't really have any special password requirements. What's the exact error message you were getting? The only requirements are that it's at least 6 characters long and different than your username, and in each case it should tell you exactly what's wrong.

There's nothing extra running on the server; no reverse proxy (the app itself automatically fetches/renews the HTTPS cert), no database, nothing. Just the app, the SSH server and the default system services.

Sorry, I'm a little confused? The pronunciation is listed for every word; that's the hiragana next/on top of the words. (:

> Do you know of something similar for Chinese by any chance?

Alas, I do not. Maybe I'll make something like that in, like, 20 years if I'll ever be able to make a living off of this. (:

Eventually I do want to make it a one-stop-shop which will teach you everything and take you from a complete beginner to someone who can immerse in native media as soon as possible. We're not there yet, and the site works best if you're at least an advanced beginner. The bare minimum requirement is that you know hiragana and katakana already.

Your best bet would be to start with a textbook of some kind and/or some actual lessons with real teachers to learn the basics. The more of a beginner you are the more the human touch helps; the more advanced you are the more you can depend on apps.

Yeah, it's still pretty much a very niche resource that many people do not know about. (:

Indeed, Heisig's keywords can be janky. Mine are not perfect, but in general they should be better than Heisig's. Well, at least for most of the really common kanji; I still need to change/improve the keywords for some of the more rare kanji and tweak a few more common ones. (As you can imagine doing that manually for a few thousand characters is a lot of work, so it has been slow going.)

It's Electronjs with vue3/typescript, threejs, tailwind2 and konvajs for timeliner(because canvas render much faster then html update, so it's keep high FPS on edit mode). And vitejs for tooling.

Now all I get in support requests is the occasional account management email, and those are pretty streamlined at this point. (I could eliminate nearly all of them by creating a proper account management page, rather than directing people to email me, but given the number of requests I get, the cost of doing that is way higher than just handling it by email.)

It helps that the screencast is aimed at more advanced developers. But that's also why it didn't end up being viable in the long term, I think. That and the "long play" format, I suspect.

The “Lessons Learned” channel was more elaborate. The video at objectplayground.com [1] is my most popular example. For those, I would make slides and animations in Keynote, then record while playing the slides. I had an outline, but not a script, as I felt that made the narration less stilted. Sometimes that resulted in minor mistakes going into the video. For more serious ones, I would record a new line and edit it into the video.

In all cases, I would edit out pauses, and if I misspoke or stumbled over my words, I’d just repeat myself and edit out the mistake. I got pretty good at editing—because there was no video of my face, I could make cuts that were basically invisible. My livecoding videos each had hundreds of cuts, typically just half a second or so. It would take a 20 minute recording down to a 15 minute final video. It took about 4 minutes of editing per minute of final video.

[1] The video on objectplayground.com isn’t showing up on my phone for some reason. You can find it here instead:

http://www.letscodejavascript.com/v3/comments/lessons_learne...

Not to say that this is your issue, but I find it helps if I dress up a lot.

I have nice, embroidered pearl snaps, custom cowboy boots, fancy hats, and a couple nice vests.

That is to say, it's pretty easy for folks to confuse me with panhandlers. I have no problem with folks panhandling, but my tips are way better if I look like a performer.

Anyway, accordion OP you should get a QR code for tips :)

I should probably look into that... it seems like a reasonable idea.

I will say, the one time that someone asked me directly about it, when I told them I didn't have venmo they gave me a $100 bill. Not to say that will ever be repeated.

1. These are not HFT models. They trade once every 2-4 weeks on average. They scale to billions in capital, so selling them does not inhibit my own returns. 2. There's a lot of demand. Part of the reason I decided to launch is I had some friends IRL begging me to let them use my models as well. Seeking Alpha has over 15M MAU and hundreds of thousands of premium subscribers, and in my opinion this service provides more direct value and wastes less of your time with noise. 3. All the money I make from this goes back into my own bot and compounds. Sure the $1400/month doesn't add up to much right now, but 1000+ premium subscribers one day would make a huge difference to me financially. 4. I believe we are at a precipice with passive investing, and the next bear market which could be right around the corner will dishearten a lot of folks. The more money investing smartly in the market at a reasonable fee makes the markets and overall economy less fragile. More automated smart investing saves society the drag of lots of suits on wall street mostly playing a big marketing game without producing much alpha or liquidity or any other measurable benefit in contrast.

Also, on a lighter note, running a business is just a lot of fun. I love seeing people use my product and get value from it.

By contrast, our models only send out signals on average of once per 2-4 weeks depending on the model. Our signals are always macro signals--they have an opinion only on where they see the overall market (S&P 500) heading next. The signals are irregular. During extremely calm times like in 2013 or 2017, there can be months without a signal. On the other hand, during highly volatile times like the COVID crash in 2020, there can be multiple signals per day, but there will never be more than 1 per hour.

Naturally, their drawdowns will be worse because individual stocks are inherently more risky than the index. As far as which will be the best absolute performer over the long term? I guess we'll have to wait and see :)

One last note about ours that could make a difference depending on your account type and country of residence. Since our models are macro models, if you live in the USA and implement them exactly by trading the S&P 500 Futures, you'll save money in taxes as all gains are taxed at 60% long term and 40% short term regardless of holding period. Alphahub gains will be taxed always at 100% short term rates unless held for more than a year which is unlikely with their systems.

I'm glad to hear you are interested in setting up your own automatic trading system! I'm in the process of writing some docs on exactly how it works, but in essence, we'll give you an access token for the API and then you can call it from your bot locally or in the cloud to get the live signals for all the models you have access to.

Lastly, the 75% is a bit of an anomaly. As you can see, the top model is only expected to return about 60% pre-tax. It's unusual that in live trading we have exceded the back test--2020 was a bizarre year and we also started after a big bear market so we benefited for a huge bull that year + nailed the June and September crashes (check the trade history). 2021 was a below average year by contrast, and over the long term, any algotrader should expect the live results to be a bit under the backtest as despite our best efforts (and I consider my anti-overfitting engine to be the biggest accomplishment here), all systems are to some degree overfit to past data.

Lastly, as I mentioned in another comment, I make no guarantees of always publishing this site and signals. If I decide in the future (perhaps once my account is to that 50-100M range) that the site is more of a liability for me than an asset, I may close it, with or without grandfathering. However, obtaining 1% of Seeking Alpha's userbase would be 5M+ per year in revenue from the site which is a few multiples of my current net worth. Even with a hypothetical 50M net worth 10 years from now from 40% post-tax, post-living expense returns, 5M is still meaningful. If we fail to gain that much traction, or the bot does so well that I hit that kind of figure much earlier than expected, I may halt the publishing.

As of now, and for the forseeable future, the site makes financial sense to me, and it's been a lot of fun, so it also makes pleasure sense for me. Life is about more than money of course.

My models do not use any leverage, and they perform better in volatile markets than stable ones. In fact, the only way they can gain alpha is from avoiding drawdowns because during long signals they get exactly the same returns as SPX.

And yes, you are right, it is incredibly difficult to make a profitable bot. 2020 isn't the full story, that's when I decided to give it my full time effort of 80+ hour weeks.

As long as the bots continue to perform near the backtest, yes eventually I would become wealthy with them without other monetization, and as mentioned in other posts, a net worth in the high 8 figures, probably dwarfs what I can make with this website and service so I make no guarantees that it will always be around. That said, I'm enjoying running it and having real people use my product, and it does not hinder my own returns, at least not at our current scale.

Side note: FWIW I think it's probably harder to make a successful crypto bot than equity bot because crypto is far more speculative and doesn't have as much fundamental and macro backing it up--would have to be mostly TA, sentiment and there's bound to be more inherent randomness to the movements since it is so new and speculative.

Ultimately somebody needs to provide these images and if nobody else does it I guess you can make okay money from it!

It's probably automatable with some effort!

Bark Like a Dog -- https://www.amazon.com/Iguana-ASD-Bark-Like-Dog/dp/B07C1BKWK...

Truth or Dare -- https://www.amazon.com/Mitchell-Harris-Truth-or-Dare/dp/B073...

The Song that Never Ends -- https://www.amazon.com/The-Song-That-Never-Ends

and my personal favorite, but doesn't make much is Rap Battle -- https://www.amazon.com/Mitchell-Harris-Rap-Battle/dp/B0742JQ...

No advertising except for $20/month for Apple Search Ads… I guess I should advertise more.

My advice if you want to use our service but in a region besides Chicago: use our service and email us the region you'd prefer. In the mean time, the extra 20ms round trip latency probably won't bother you and we'll let you know when we expand in the future (and your voice will help decide where!).

It is more like we provide you a network interface that, instead of being plugged into your modem for Comcast/AT&T/etc, it's plugged into a datacenter.

You get a public static IPv4/IPv6 connection with Reverse DNS - just located somewhere else.

Making your machine publicly routable is the primary use case (e.g., for email, web, whatever you want if you have trouble with NAT/CGNAT or simply don't want your home IP associated with the service), but it also encrypts your traffic between your PC and our datacenter such that your ISP can't snoop - which is not to say we're a good privacy solution - that is not what our product is designed to do.

About the logo – I'll take it as "imitation is the sincerest form of flattery" ;-)

https://dheera.net/projects/4x5/

My original intention for this is gigapixel images, but I ran into some issues cancelling the CRA correction that the Pi HQ cam does, and also, my newly bought Sony A7R4 can shoot 240 MP shots with pixel shift so it's suddenly less interesting again since it's already within an order of magnitude of a gigapixel, and if I got a shift adapter mount and shot with a medium format lens I could probably easily exceed a gigapixel with the A7R4. I might try to shoot 4x5 film some time though.

You should know your way around Linux (if you already know Ubuntu, you can use it on the server as well).

You need some software: - For receiving and routing mails: I use postfix - For authenticating: Dovecot - For serving mails via IMAP: Dovecot is a popular choice

There are some good guides on the net how to set up these on Ubuntu or Debian (which is pretty similar), just google for e.g. "postfix ubuntu".

It's critical that the mail you route arrives at the other end, so you must prevent spammers from abusing your systems. Otherwise you get blacklisted. As a start you need to implement state-of-the-art things like: SPF, DKIM, DMARC. MTA-STS, TLS-RPT are also useful. None of these are hard and there are also good tutorials out there, it's just a bit of work.

If you were looking for news then sUASnews is great for some catching up on drones. For robotics I often find articles on IEEE Spectrum and The Robot Report interesting.

So far the best way I’ve found for growing subscribers is trading shoutouts with other newsletters but I didn’t experiment much with paid ads.

Do you have a source to support your perspective?

I don't handle the fulfilment myself. I wrote a software integration that connects NewBusinessMonitor with a direct mail company in the UK. They have special machines that can print, envelope, and send all the letters automatically, at large scale. The letters are printed in colour (on rather nice quality paper), trimmed by guillotine (for full-bleed printing), folded in half and inserted into a windowed envelope, before being dispatched 2nd class, usually the following business day.

I typically don't need to manually intervene in any part of the process; I just try and do hand-holding with people when I onboard them, because I think good business is about building relationships with people, and many (most?) of my users specialise in areas outside of computing.

What happened when you git Sign up?

Thanks a bunch for the feedback!

Everyone’s path is different, but generally speaking, if you publish enough blog posts (especially on LinkedIn), someone from a content shop will notice and ask you to write a book or create a course. Both flows are similar, but I’ve only done courses and can speak to that in detail.

You propose a course. The content manager you work with decides if it fits their learning plan or whether there is demand for that kind of material. If they think people will view the course, you’ll be asked to come up with a course summary. DevOps and Kubernetes are super hot right now, so these were easy for me to think up.

Here, you’ll pitch the course. What is it? Who’s meant to view it? What will you cover? This gets reviewed. If it looks good, you’ll be given a producer, and they will ask you to write scripts.

Sometimes the content manager will ask for a demo of how you’d present a course. Here, they want to see whether you have enough “presence” to instruct. Sometimes they’ll provide feedback for you to fix, but this didn’t happen in my case.

You’ll also get an advance, with the first half being paid out around this time. These advances aren’t like the music industry where you use the money to pay for production and marketing. It’s literally money in your pocket.

Script writing can take FOREVER. especially the first time. It took me three months to write enough scripts for a two hour course. There’s a lot of back and forth. Since my courses are very hands-on, I also write lots of code and test everything while putting the scripts together. Very difficult to do all of this with a full-time job, doubly so when that job is demanding!

Once the scripts are cleared, you’ll be asked to record. There are lots of different ways this can go. Mine are audio-only with a terminal walkthrough, so I just recorded and screen capped everything. Some courses have live-action and acting. These require studios. I haven’t done one of these yet, so I can’t speak to it.

Beta testing happens after recording. People internally go through your course and make sure that everything gels. Sometimes you’ll need to fix stuff after this.

Once beta tests complete, your course goes live and starts racking views! You also get the other half of your advance.

You get paid royalties for every learner who views your course. I don’t know the specifics of this, but it’s kind of like YouTube’s ad platform. However, you don’t earn anything until the grand sum of your royalties exceeds the advance you were given. Depending on how your course performs, this can either never happen or it can happen very quickly.

I recently added a "call for sponsors" blurb in the intro of one of my newsletters once I felt I hit a level of subscribers where it made sense. I immediately had a few readers reply with interest that ultimately led to my first few sponsors.

From there, I did some cold outreach via Twitter DMs and email. My newsletter is very workspace item/tools centered so I put thought into what sort of sponsors made sense there and it led to the next batch. Think: companies who would organically be featured by a guest anyway... show them the value.

I'm pretty transparent with the growth/numbers on Twitter as well so I think that helps when doing the Twitter outreach for sponsors. They are easily able to look back at the Twitter engagement if they don't already know what Workspaces is.

I wrote it with Microsoft Word. I initially tried a few different apps but all of the organizational tools they provided got in the way. I broke most conventions and just wrote the book straight from start to finish, and re-read it countless times until I felt satisfied. Then I sent it to a ton of friends who helped me edit. A professional editor would've been a better idea, but I didn't really have the money for that at the time.

This year they lowered that $20,000 to $600. So anyone who receives more than $600 in one calendar year will trigger Venmo/Paypal or whoever to send the IRS the proper tax forms. If it is for personal use of course you are still not required to pay any tax.

As far as I know none of the loopholes for taxing the uber rich have been changed so this is really just going along with the IRS' current policy of increasing revenue from small time outfits vs going after the bigger fish (who can often afford to hire expensive accountants and lawyers to fight the IRS)

I am not against paying taxes but having to wrangle the info from a bunch of different sources and then also deal with manually adjusting for other factors just makes this whole thing a huge pain.