Current look of the website and paper-prototype for next version

The good thing about having a website is people can criticise it. Or they don’t, which is also valuable feedback, because it can mean a number of things. Instead of wondering what those things could mean, it’s usually quicker to try a different way of communicating, and get critique in person. That way I learned that showing a printed out screenshot of metaexplorer was better than a video on my phone. Now with COVID, meeting up personally is not a good idea any more, but instead a video recording of myself using the software does an equally good job. And here’s where the learnings of the recent months come into play: I’ve been in the enterprise software industry for long, now working for a Startup I can see that I’ve built metaexplorer with an enterprise in mind. Startups need copy/paste plus some glue and scissors, whereas the museum software I did before had to run for years (technical speak: avoid memory leaks). Then I had a project with a bank, where the software was supposed to survive ReactJS (the technical framework) ideally, but we scrapped that idea at some point. At a minimum, the software had to survive an audit.


No need for audits in Startups, and even less so in bootstrapped businesses, yay! So that leaves a little issue: From my personal experience, I only know people who don’t need metaexplorer in its current form. I built it to create a platform where Chinese could get a care package from someone packing it in their home. So that meant I’d have to be able to iterate on a Chinese and English/German App, but it shouldn’t be packer=German, recipient=Chinese. People send goods to friends and family all over the world, so ideally that should be language-independet. The Chinese User Experience (UX) is very different from German and English because their language needs much less space on the screen and fits in squares. That means you can write it up and down and in a block of four characters.

Example of English, Chinese and German text on a Material Design card

If you search for Arabian UX, which is written right-to-left (RTL), you can find that they’d turn the whole layout around, putting the image on the right.

Each packer or recipient should have had their own app, each person in their own language. So text and visuals should have been very adjustable, with the core functionality staying the same. When I started in 2016 there was no technical solution for that, so I started to build my own.

Now this type of package-business is already not possible with China any more, mostly because packages get rejected at random. This happened earlier than I anticipated, but at least I had some anticipation and built metaexplorer as an independent tool. It increases the iteration speed for UX implementation and it adjusts itself to the data it’s running with.

Platform businesses are nothing special any more, but software to build platforms still is. If we take it one step further, we end up at the dream of the blockchain projects: To create something that solves a problem on a big AND a small scale, and the way in which they do that is by establishing a protocol. That’s not my dream, but the protocol part is interesting.

The technical part

github copy repository URL button

As the most popular Application Programming Interface (API), REST has moved on to OpenAPI, which has then seen the rise of GraphQL. So it seems there’s a need for a standard for data that gets sent over the internet. Why not try to build a standard for sending things to a user – so the user interface? Actually here a lot of well-adopted design systems and libraries add different flavours to the internet, so the users don’t have to relearn the UI for every website. What if one could access the data of the website in a familiar way instead? What if there was a tool on one website, and one could use it on another, like different flavours of “copy to clipboard” – buttons? 

google sheets copy link to sheet

So this is part of what metaexplorer is aiming at. I’m not inventing anything, the concepts exist, but business use is rare. Linked Data and Semantic Web concepts have only partly found its way out of research. But Wikipedia is entirely readable through the SPARQL-language, and Google “understands” what you type through a knowledge graph. That means “tell me the birth places of all Oscar winners who have a .org-website” is something you can already query Wikipedia with, not Google. But those open data sets and APIs are rare, and my assumption that their number would increase a lot hasn’t become reality. To surf the Web of data in your own ship, explore and interact with information as you go, that’s where metaexplorer has its name from.