I've had a rather uncreative year, and I'd love some inspiration going into 2020. Please share your side projects/businesses/hobbies, ideally with links and traction numbers. :-) Have a lovely new year!
Adopt Animals (https://www.adoptanimals.io/) - a charitable passion project for free, independent, and ad/tracking-free animal rehoming listings in the UK.
We're partnering with one shelter in Edinburgh (Scotland) to start, and built the website and a pair of apps to showcase animal listings. We've had a few success stories of people finding pets already, which is really motivating!
If you know of a shelter in the UK who might want their listings on there (ideally they'll have a means of exporting them and we'll build an importer), let them know to get in touch with us :)
For interest, we also just submitted our annual report to our regulator, if you'd like to read about the first year of our parent charity https://www.kale.charity/reports
There are currently 3 trustees of the charity, including myself, and no employees. It’s illegal for trustees to be compensated for work relating to the charity. We’re careful to pick problems (and solutions) that we can make and maintain.
I focus on the engineering, my partner focuses on anything design related, and our third trustee manages finances.
We do it because we’re passionate about it, and the legal structure gives us a lot of weight and ability to negotiate reduced rates with services we depend on. It also sets us up to pay other developers, if/when we decide there’s too much for us to do as a team. Folks might want to donate their time as well, but we will need to be mindful of properly compensating people for their time.
I am a classical musician, programming for fun. I found guile scheme about 2 years ago, and I have been writing all my software in it since. Such a nice language to be working in!
I wrote a SRFI (scheme request for implementation) for transducers, which are efficient composable algorithmic transformations. They allow you to eagerly transform collections, say like using map and filter, but without building intermediate collections. The SRFI document is here: https://srfi.schemers.org/srfi-171/srfi-171.html
They are zero cost (apart from negligible macro expansion) and provide a homogenous way to iterate through various collections. I am now in the process of implementing foldr for it, which will allow for a general way of writing lazy iterations.
SIXTEENmm (https://sixteenmm.org), a streaming site for old, hard to find and indie films.
I've spent most of the year working on colourisation and recovery techniques. Colourisation is probably 70% of the way, but restoration is as hard as ever.
The project isn't meant to be a unicorn. The tech stack is boring, and probably wouldn't scale.
There's about 50 people on board, and about two or three videos streamed a day - so it isn't remotely "successful" yet, but I set a four year timeline for it.
I don't have any specific stats to offer - because I don't collect it. I collect as little data as I can get away with. Both for personal belief, and economic reasons.
Hard to get massive fines for the inevitable breach when there's no PII to steal.
This is great, my wife was looking for a site for old movies the other day. At the risk of offending could I suggest working on the design a bit to increase trust and look more established. If I stumbled upon your site I don't know I'd be willing to put my credit card details in with the current look and feel of the site.
No risk of offense. Styles are not my best skill, though I have found people trust more when you throw things like pointless animation at them, which is something I'm specifically avoiding. (Part of the site design guidelines is that my highly autistic niece can use it without freaking out).
But, you don't have to put your CC in. Not until you've established a relationship.
The biggest thing I have built this year is Hexadecimal (https://tryhexadecimal.com). It is my first SaaS business, so it is a pretty rough endeavor, both on the development and the business side. Built on the vanilla Rails stack. As boring as it could possibly get. I have described in some detail the tech behind it: https://runninginproduction.com/interviews/9-running-a-websi...
Lessons learned:
* If you'd like to start making money on the Internets, don't start with a SaaS
* Making your first $currency will give you a (much needed) morale boost
* If you're just starting out and you're in for the long-term, optimize for learning and building relationships
* Worthwile Things take time
* You probably won't get it right from the first time (whatever it is). It is far more important to keep iterating rather than getting the right answers from the very beginning.
* Most minor decisions won't matter in a few months', let alone in a few years' time. Don't overthink it. Make a fast decision and if necessary, re-evaluate it down the road
* Don't rush to automate tasks
* Build it, and they will do absolutely nothing
* Businesses live and die by their distribution channels
* Running a lean operation (i.e. low-cost) is a competitive advantage
* Having an audience is an unfair advantage
* Writing is a gift that keeps giving. Write more!
* The true validation is people paying you money
After many months (or years?) of procrastinating, I finally published my personal website (https://jmstfv.com). I have been meaning to do this for a long time but kept putting it off for various (artificial) reasons. So, I hand wrote the HTML, copy pasted the CSS from my other projects, and called it a day.
Lesson learned: start with the least painful solution.
This is cool. I'm curious about what your main marketing channels are. There is another project this reminds me of another project called Simple Analytics.
* Making money with affiliate models in this space can be hard if you don't already have a big existing audience (ie. from a Youtube channel) - wrote a bit about it here
It shows which players gigged together, and colours the players based on the country they live in. It's a wiki, so anyone can add themselves or other players, as well as add a link to a YouTube video to promote their music.
A bunch of people (Gypsy jazz a pretty small world!) added themselves (there are over 200 players on it now) then it died down.
Totally! Building a community has helped me to know what customers wanted.
How I did it:
- Started a small Facebook group, added friends
- Started connecting on FB with productized service entrepreneurs, added them to the group
- Mentioned the group in various places (Twitter/Indie Hackers/...)
Interesting, thanks for the info! Did you create a separate account for your product, or did you just use your personal account? What sort of information do you post to the group? Thanks again!
This tool was built as part of a larger suite of tools that highlight and resolve the cost performance gaps businesses face when entering the Chinese market.
2. Sentiment-tagging (positive/negative) for financial news.
Personally, I don't believe there's a lot of alpha to extract from this, because news usually lags market information. But A LOT of people believe differently, and this article shot up on relevant Google searches, way ahead of academic papers or other sources of authority.
https://towardsdatascience.com/a-new-way-to-sentiment-tag-fi...
In this initial version, theres photo sharing. But yes, planning to integrate messaging system as well.
I take photos and record videos on my phone. All my friends and family too. I used to share them on different media sharing platforms. Used to have a meaningful discussion/conversation on the shared content with the people I know.
Things changed and got worse. I hated infinite scrolling and public feed which I still do. I never really liked the idea of "liking/hearting" - rarely "liked" stuff. Also, these days it's all about influencers and things get lost in the feed.
Keeping all this in mind, inspired by good aspects of the existing platforms and really based on what I needed, I decided to build Storry.
I built a prototype of an ingredient list scanner for Keto people. It's still not launched and I'm sure there's bugs, but I'm hoping that posting this gives me some motivation. :-)
Blook (https://blook.io): Helping US and Foreign entrepreneurs register their company in the US as an LLC or C-Corp. Go-live will be on the first :) Particularly looking to work with the latin america market.
Looks great. Wish you all the best for the go live. We are in the same business. I do this for Germany with firma.de. Trying my best to increase the amount of entrepreneurs here.
If you ever want to talk about bundling other service offerings into the company formation offering let me know. We had our fair share of learnings there :)
I built a new web application front-end to an old warehouse inventory tracking and catalog management system for a very large corporation everyone is familiar with. But thanks to the contracting arrangement, I can't talk about it. Makes it tough to attract new clients.
I just released my first product (https://www.keepmeon.top). It fetches and filters content from Reddit, Hacker News, Lobsters, etc. and generates a RSS feed or sends a periodic email summary.
There are a lot of things to improve, but it's a start...
The Grassmann.jl package provides tools for doing computations based on multi-linear algebra, differential geometry, and spin groups using the extended tensor algebra known as Leibniz-Grassmann-Clifford-Hestenes geometric algebra. Combinatorial products included are ∧, ∨, ⋅, *, ⋆, ', ~, d, ∂ (which are the exterior, regressive, inner, and geometric products; along with the Hodge star, adjoint, reversal, differential and boundary operators). The kernelized operations are built up from composite sparse tensor products and Hodge duality, with high dimensional support for up to 62 indices using staged caching and precompilation. Code generation enables concise yet highly extensible definitions. The DirectSum.jl multivector parametric type polymorphism is based on tangent bundle vector spaces and conformal projective geometry to make the dispatch highly extensible for many applications. Additionally, the universal interoperability between different sub-algebras is enabled by AbstractTensors.jl, on which the type system is built.
This is really well done. FYI, in Firefox on macOS, the search box doesn't seem to do anything for me. I don't see any outputs or requests in the JS console or network tab.
I didn't get rich by writing this book, but it's very useful for networking in blockchain community. It's like an expensive name card. It also boasts my reputation.
For those who are familiar with blockchain programming, Mamba is like Truffle but instead of based on Solidity+JavaScript, it is based on Vyper+Python.
For those who are unfamiliar, Mamba is like Ruby on Rails but it's for writing decentralized / programmable money applications.
My plan for 2020 is to write more tutorial articles in Mamba website and develop more features in Mamba framework.
For business: I am helping a company build their programming bootcamp (not related with blockchain).
So, I bought the app, and am in the process of creating a hub with 27000 videos, about 5 tb worth of videos, using 4 drives symlinked into one folder :) Let's see how it works!
Initial thoughts:
Would love to see a 'display title of folder' upon hovering on a thumbnail in compact view. I like compact view more than normal, but without displaying filenames I can't see what things are.
Would like a way to change the font display/size/formatting of the filename display somehow, but at least changing the preview sizes increases the font size of it (it was too small a font size by default for my monitor setup). Although now I'm getting bigger thumbnails by increasing preview size even if I only want the font size increased, but better than nothing.
Tabbed hubs with a top bar instead of using recent history on left?
I like how there's such a good variety of search tools.
Love the auto tagger. Would be cool to fiddle with how it works/sensitivity settings in terms of the threshhold for word frequency/whatever else is triggering tagging conditions. This could be incredibly useful for ppl with large libraries.
Since I have so many videos in one hub I'll probably end up wanting additional sorting/categorizing options. Thinking I'm going to have to do that with file system folder reorganization and then potentially doing 'show folders' enabled within Video Hub currently eh? (Or separate hubs for each category I suppose, maybe I'll end up doing that if this symlink stuff is too unwieldy).
On a related note, it would be cool to add in more autocategorization functionality like the autotagger, and bundle in sort of 'library organizer' functionality. A lot of people have tons and tons of videos that are really difficult to sort/categorize by hand. There's probably some sort of library management software out there but I haven't looked into it. If you know about the software Calibre it has all these ebook databases it crosschecks your files with and then auto-tags/auto-genres/ISBN classifies everything. Not sure if something like that exists for video files, but there's probably ways of detecting genre/patterns in the metadata and filename themselves I'd think.
Potentially but also maybe not, a minor bug/unintended functionality is clicking on a video to 'show similar videos to' plays the video in my default player. But maybe there's a toggle for single click > double click video playing that I'd ideally set it to double click because I don't always want to play a video if I just want to see what's related to it. Can't check the options panel right now cause I'm at video 9000 in the loading bar of the 27k total ;p. Another solution for this instead of single/double click would be if one can just click the text filename it'll show similar videos to without playing the video whereas clicking the thumbnail just once triggers the play video. Also that 'similar videos to' has really interesting potential, can see that also being very useful depending on what triggers a similar condition for large libraries as well.
Anyways, great software! Will recommend to others. Pleased to find it because it really does come in handy. Switching over from default Windows File Explorer to One Commander was also another great thing I did in the past few weeks that's going to improve my workflow like Video Hub will -- anybody who's on Windows should really checked out One Commander, I prefer it now to even directory opus.
I didn't implement a font-size feature because I thought the 'zoom' would be good enough -- it resizes the entire UI, but that's a benefit for some people (my parents for example, or me -- with a 55" 4K monitor I use as my desktop - but I sit far away).
I _just_ implemented a threshold feature (will be released in 2.1.0 -- probably in January). You can see the PR: https://github.com/whyboris/Video-Hub-App/pull/322 -- I hope that's what you had in mind.
Double-click toggle is something others requested -- I'll see if it's easy-enough to implement. Perhaps single click would then just show similar as you suggest, while double-click would open.
Just FYI -- all the generated screenshots & thumbnails are based off the filesize, so when you decide to create a new hub that has a lot of the same videos (perhaps when you're creating smaller hubs from individual folders from the symlink one), you could just copy/paste the screenshots and the screen generation process would go by quicker (the app will just check that the screenshot exists and move on). After that, when you click the "Any changes" resscan option - it will delete any screenshots for videos not in the hub.
I got tired of fighting SSR and implementing the same optimizations over and over again so decided to launch http://oya.to/ an optimizing cloud proxy - kinda like Cloudflare. The main feature right now is prerendering as an alternative to SSR, but there are a lot of things planned for 2020.
In a similar vein I found myself doing a lot of work that I think should be unnecessary just because I wanted to buy some fonts and not depend on yet another subscription service so decided to build https://woff.cc/ aka "Web of Fonts". It's currently in the planning stage but when launched, it will allow you to upload your licensed or (SIL) free fronts (including all Google fonts already built in), optimize them (subsets, css embedding, etc.) and either host it there for free or download and host it yourself.
I’ve been improving my production process for the wordclocks[0] I build. I’ve also just barely started to try some basic marketing of them. I sold 2 this year.
Part of my marketing effort is improving visibility on Etsy and so I’ve started selling much less expensive items on Etsy too[1]. They are not so much for the revenue as for the traffic, ratings, and learning how to best use Etsy. These secondary products are all based on things I’ve designed and made for/with my young kids.
Two observations:
- I’ve really enjoyed the path of developing better processes of how to fabricate things
- I’ve really enjoyed getting my kids involved with making. It’s fun to observe them thinking about how to make stuff and working with them to figure things out (they are 3 and 6).
I built and launched Trunk[1] which is a SaaS that syncs stock levels in real-time and does other inventory management features for sellers that sell on multiple platforms (e.g. Amazon, eBay, Shopify, Etsy, Squarespace, Square, Faire, etc).
Being a Disney fan, I needed to browse Disney plus movies with IMDB rating. Also, my friends wanted to view Disney plus catalog without registering. This site serves both of these purpose.
Looks like a cool idea! If this worked like Pingdom, where I could start for free and then upgrade based on frequency of checks or number of sites / pages, I’d sign up :)
2019 was not a great year for me, too. However, as a side project I launched a micro-social networking app for introverts. It's available for both iOS and Android.
website: getpixelpal.com
To be honest, though, I've fallen out of love with it and I'm considering shutting it down or making a pivot to something else.
At the beginning half of this year I finished up an mvp of an idea that I had, took me a bit of time to build, where I could make it a bit easier for product managers to get analytics, and saas app feedback (https://cruisedirector.io). I was hoping to diversify my income a bit, but I've been pretty busy this year at my day job and haven't done any marketing or improvements. I have also redirected any potential customers, that I chatted with while building the app, to my competition. The competition is absolutely killing it in this space. Hopefully I can find some time this year to chat with some customers, maybe diversify a bit, and be able to do a bit of marketing.
While it's not mature yet, I started something I had in my mind during a long time, I called it atbswp (https://github.com/rmpr/atbswp), it basically allows you to record your mouse and keyboard moves and reproduce them at will. There's something similar called tinytask unfortunately it's neither multiplatform nor opensource. Back in time I used it to play automatically some games (asphalt, plant vs zombies, ...) but I think the usage can go far beyond that, for example you can use it to test software with a GUI, to present a demo at a conference, ...
It's kinda usable but I need to polish things up (settings , language, ...)
I built Metadigest (http://metadigest.uzpg.me). It's a weekly newsletter that sends the most popular tech content from around the web.
I am now working on Devolio (https://www.devol.io), a welcoming community for developers to share and discuss. I am really excited about letting users custom-design their profile and integrating with Github. I wrote about what I learned this year and how I got into programming at http://uzpg.me/2019/12/28/projects-and-learning-in-2019.html
I rewrote my social, score-tracking app for Pinball in react-native. Previously it was in Angularjs (ionic/cordova). Has ~6k registered users, ~300 daily active users, and earns ~$200 a month on Patreon.
The idea came from being in the trades in my previous career. Companies in the trades spend a ton of money on advertising but fail to capitalize on a lot of their incoming leads. We use humans to answer incoming text, fb messenger, and webchat questions for people like electricians, hvac techs, plumbers, etc.
Long term idea is that as we get more data on how these conversations typically occur, we can build some ML systems to lower the cost of answering the questions.
It's still MVP stage right now. Got a few customers and planning on ramping up marketing in January.
I released ModularPro, a tool for the Unity game engine to massively improve the process of assembling modular assets during level design. It started as a side project and I ended up releasing it in the Unity store. Would be great to get some feedback from anyone who actually does level design for a living to see if they think it would be useful for them.
I recoded my HTML5 canvas Javascript library from scratch. Partly to add fun stuff like modules, web workers, promises etc - but mainly because I'm determined to make the canvas element much more accessible (and easier to add analytics to, etc) - progress report here: http://scrawl-v8-progress-0919.rikworks.co.uk/
Question: traction numbers? If this is "how many people are using my side project", I'm fairly sure the answer is "nobody" - which has the bonus that I don't need to worry about supporting backwards compatibility.
I built promnesia, a browser extension to connect together different data sources and enhance browser history.
E.g. if you visit some blog post or youtube video you had in your bookmarks, you would be able to tell if it was coming from your reddit saved posts, or if your friend sent it to you on telegram.
Another feature is displaying annotations from any sources (e.g. pocket/instapaper);as an overlay on any page
I wrote with a friend his memoir about climbing - The Crux: A Climber's Search For Meaning In Sport, Death, and Change. https://www.amazon.com/dp/B0834P85QG/
It was a fun project that took a year to write. The programming bit went into developing a page for pre-orders and fulfillment using stripe and other tools. But suffice it to say, I liked writing a lot more.
I used the JUCE framework, which is free for personal/GPL usage (or you can pay for a closed source license) and gives you cross platform. The examples are pretty good for getting the basics, there is also The Audio Programmer YouTube channel that has a lot of content (the only bad thing I would say about it is that some of the early videos sometimes use bad concepts as it seems this was a sort of a documentary series for learning plugin programming for the channel owner). The JUCE community forum is really friendly and helpful (and there is also a discord channel for TAP). Otherwise search engines were my main source to learn (mostly to refresh my C++ knowledge), and I managed to land myself a day job programming a plugin that helped a lot just by looking at the existing code, but that obviously a very rare occurrence!
Slowly putting everything on kvr and have been very slack with the YouTube videos. I think mainly have been blessed by having our record label with same brand be the main driver for sales, and I've been asking well known artists in the psytrance scene to help with promotion (being my friends I give them freebies and ask without obligation to post something about them on their social media).
Probably don't have many sales outside that scene to be honest, so I'm hoping with time that will change.
Construpdate (http://construpdate.com/): Project management synchronization for construction teams.
Project managers import a project schedule from Primavera P6 or Microsoft Project via Excel. Each contractor gets a personalized link where they can submit daily updates on the status of their tasks. All changes are tracked for auditing purposes.
It's a work in progress, comments and suggestions are welcome!
It's an interactive database of the top 100 Shopify stores, with 19+ e-commerce data points available for each one.
I'm really happy that the average time spent on the website is over 4 minutes and 30 seconds - I never expected people to be this interested, so now I'm thinking of ways to expand this idea.
I started building this in January and am at the point of sending out for prototype PCBs: https://github.com/roscopeco/rosco_m68k - MC68010-based single-board computer
I can really recommend a project like this as the opportunities for learning are vast!
Tagmap (https://tagmap.io/) - allows communities to have a map of their members and easily message people nearby or similar to you to make new friends. Currently have a few thousand MAU, hoping to pick this up significantly in 2020.
Looked around and clicked the tag that had a few thousand users (because everything else was a hundred or less) and the few near me were bizarre NSFW shock value profiles. Is that normal for this?
There's been a few malicious users here and there and some interesting communities, we usually remove them pretty quickly, especially when they're reported. I'll do a check right now and try to remove whatever it was you stumbled across, thanks!
We're also working on a more robust system to filter content that some users might not want to see, although anything extreme should be removed either way.
It could also be the community. For example, I don't know what "r9k" means but it's far and away the one with the most users. Considering the number of 4chan related tags I saw maybe it's something unsavory.
Kind of a project-feature / workspace-switcher for macOS.
Facilitates human context switching by allowing you to save and load all open applications and their state (open windows, tabs, etc.)
Numberer - https://pdfpagenumber.com , a native Mac app to add page number to PDF files, I did this in a few days (main function is just one for loop that loop through each page).
Built Remote Leaf (https://remoteleaf.com) - We hand-pick thousands of remote jobs and send you a personalized remote jobs list based on your country/timezone and skills.
reply
Tomorrow I’ll be launching https://typehut.com, a super simple publishing platform for blogs, changelogs, newsletters, announcements, etc that I’ve been building during this last month.
I built out my idea for a VR urban planning tool earlier this year & more recently I built a network based music recommendation system as part of a hackathon project. I’m still in school but hackathons really help keep my creativity flowing!
I decided to build out my “automated” SEO agency, myseosucks.com. It’s been quite the rollercoaster ride but super exciting to see when clients / agencies on-board themselves and can self-manage the SEO process.
I released my first iOS app! Shirabe - a Japanese/English dictionary that combines spaced repetition, periodically reminding me of new words I leaned. Https://Shirabe.app for anyone interested :)
While my knowledge of IPFS is pretty basic, this strikes me as a great use case for it.
IPFS (https://ipfs.io) is a P2P internet protocol that's made to enable a decentralized internet...I believe its goal is to supplant (or at least supplement) HTTP(S). As I said, my knowledge of IPFS is still pretty basic, but I find it fascinating and I thought it might be relevant for this application.
I am probably wrong but my understanding is that IPFS is an addressing scheme. While it can be private or peer to peer it can also be public. As an addressing scheme it continues to follow the web’s client-server model. A client-server model on the web is inherently public and anonymous.
A dedicated peer to peer system focused on privacy has no server and no third party. Instead the model is client-client without third party routing. Privacy is then achieved because connections are direct, in an application layer sense, between nodes and each node is a client running the same application in the same way. This is always confidential and never anonymous. This is achieved because there is no application layer addressing used at all, not even DNS.
Think about it more in terms of a shared disk drive on a private network mapped to a local file system address instead of a web location.
* Currently it uses IP addresses for connection, which is less than ideal.
* Currently every device is treated as a separate user.
Road map:
* I am working on a web service that will issue named accounts and digital certificates. The named accounts will identify their current IPv6 address to the account service so users need only to connect to an account which can be any IPv6 address for portability.
* Once accounts are a thing I will need to change several things on the client application. My goal in this step is to separate users from devices so that a single user can connect and manage multiple devices. That extra layer of abstraction will allow options for management, security, and convenience that are not currently available.
* I also want to enable end to end encryption for all users/devices. That is just a simple matter of public key exchange built into the already designed invitation model.
* Then I want to create a text client that supports markdown.
I built a web browser you can deliver through a web browser. Technically a view layer for a browser you connect to via DevTools. > 500 stars on GitHub. https://github.com/dosyago/skateboard
I also build a way to archive anything you browse online so you can read it again offline as if you were still online. > 500 stars on GitHub. https://github.com/dosyago/22120
We're partnering with one shelter in Edinburgh (Scotland) to start, and built the website and a pair of apps to showcase animal listings. We've had a few success stories of people finding pets already, which is really motivating!
If you know of a shelter in the UK who might want their listings on there (ideally they'll have a means of exporting them and we'll build an importer), let them know to get in touch with us :)