I build software, systems & weird things with code.
Joined July 2023
- Tweets1.6K
- Following154
- Followers176
- Likes3.6K
Aruodore retweeted
printed 200 stickers for a candle brand, then searched the name. a tire shop had it.
check instagram, tiktok, the .com and the business register before you spend a penny on a business name. two minutes.
Aruodore retweeted
Onboarding
Subscription screens are a lot more than presenting prices.
You're influencing how people perceive value and make purchasing decisions.
The subscription page is easily the most important page on any app at all, but it doesn't seem to get enough love in many cases.
It's a small interaction, but I think these are the kinds of decisions that make interfaces feel more intentional.
When you set a default offer, you are literally making a suggestion on which of the offers you want your Users to go for. Thats what you are doing.
Some offers present more value to the User, some other offers present more value for the business. How do you pick your defaults?
Just like the design of literally anything else, designing pricing interfaces is mostly a game of moderation and deliberate emphasis where and when needed.
Sure, you wanna give your users options so they can have some kinda anchor price to make a choice, you do not want to overload them with options.
You you have to show them the list of benefits, but do not let that become a blocker. Users should not have to scroll to find the checkout button.
People buy in different ways and not everyone wants to have a logical soliloquy before making a purchase.
Some just want to click click click, and making them have to scroll by all means is def going to make you lose quite some sales.
The less effort required to move from making a decision to acting on it, the better.
One thing I've come to appreciate about pricing interfaces is that presenting options is only half the job. The other half is helping people make a decision.
Aruodore retweeted
Search is not the same feature everywhere.
You could search for the same thing on Google, Netflix, Amazon, and Spotify, and get four completely different experiences.
And that's intentional.
Google would show you what other people are searching for, even before you've finished typing.
Netflix would show you a title it doesn't have and recommend something similar instead.
Amazon can suggest products you didn't explicitly search for, but they do have.
Spotify can turn a single search into an entirely new listening experience, while ranking songs you've recently listened to.
...
People search for different reasons:
to find something they already know,
discover something new,
understand what's possible within a system,
acquire information,
compare alternatives,
make decisions,
complete tasks,
revisit previous discoveries,
find inspiration,
solve problems,
monitor changes,
or investigate subjects in greater depth,
and each of these intentions should influence how a product designs its search experience.
...
Another thing to consider is the minimum parameters a user needs to make a successful search.
Searching for an item on a jumia store needs just one parameter (item name), while booking a ride needs at least 2 (pickup and destination).
...
In summary;
Search should be designed around what users are trying to achieve and what the product exists to help them accomplish, rather than simply matching keywords to results.
Ultimately, good search design isn't just about returning accurate results; it's about helping users reach meaningful outcomes, even when those outcomes differ from what they initially searched for.
Attached is a fragment of a search experienced i designed for Prysm - A game community app.
Aruodore retweeted
Road to 6k 😴. @Serticode you think this 2k this morning will keep your top spot
@wolkahq
Aruodore retweeted
Everything your team needs to work together, in one place.
Chebu brings business email, transactional email APIs, cloud storage, and collaborative Docs, Sheets, and Presentations into one workspace.
Write together. Edit together. Share files. Manage emails.
One workspace for your entire team.
Get started free at chebu.io
Aruodore retweeted
Chebu is officially LIVE 🚀
We’ve been building something bigger than just another email service.
Starting today, you can sign up at chebu.io and get a free email address, 5GB of cloud storage, and access to our online Docs, Spreadsheets, Presentation Editor, and PDF signing tools.
And if you’re running a team, Chebu brings everyone into one workspace with shared email inboxes, individual mailboxes, and tools to work together without juggling multiple services.
One place for your email, files, documents, and team.
We’re just getting started.
Aruodore retweeted
The JPEG-LS decoder we built from scratch in Rust 🦀 is impressive and lightweight enough to handle decoding on the frontend in bulk, saving server compute so we can serve more users.
Even then, memory and system resource use stays significantly lower than before. We plan to offer a free desktop app anyone can install to store, forward, and securely transfer images to Xenrad (if you use it), or simply use as a viewer.
Aruodore retweeted
Hamiltonian Monte Carlo uses gradients and momentum to move through a probability distribution without the usual random-walk crawl.
This interactive makes the leapfrog trajectory, energy error, and acceptance step visible:
aruodore.com/pieces/hamilton…
Aruodore retweeted
The prototype is currently in the red 💀. Let’s see what a little refinement can do…
Aruodore retweeted
The pipeline is open source.
Rare-Studio
GitHub: github.com/Joshualeexy/rare-…
Aruodore retweeted
Start your very own faceless channel with Rare-studio
The pipeline is open source.
Rare-Studio
GitHub: github.com/Joshualeexy/rare-…
Aruodore retweeted
Added dynamic watermark to my Rare -studio video production pipeline using openCode and Ollama
Aruodore retweeted
Gibbs sampling moves one variable at a time.
When two variables are closely linked, that simple approach can become surprisingly slow. Try it yourself:
aruodore.com/pieces/gibbs-sa…
Aruodore retweeted
A scatter plot is a chart that uses dots to show how two things relate to each other.
A scatter plot turns individual items into dots and positions those dots using two measurements, an x-axis and a y-axis.
In this case the y-axis is the length of the pitch from post to post, and the x-axis, the width of the pitch, from throwing-line to throwing-line.
The individual dots give you data, but the arrangement, and movement of all the dots together gives you information.
everything visual was made entirely in
@figma
the only exception was the sound which was added using @capcutapp
