Andrew Marcuse's Blog
Live Last read · last published · next check
I originally wrote [RSS.Style](https://www.rss.style/) using XSLT, but Google, as the malevolent friend of the open web, is deprecating XSLT ([flamewar on Hacker News](https://news.ycombinator.com/item?id=45954560)). So I switched to JavaScript, which (as far as I can tell) works everywhere. It is pretty easy to add: just a single line to the XML of your RSS/Atom feed: you don't need to host the…
I used the [Tabulator](https://tabulator.info/) JavaScript library for a recent project at work, and was blown away by how flexible and easy to use it is. I ended up cranking out a whole bunch of sites using it: * [UnicodeSearch.org](https://www.unicodesearch.org/) - the complete Unicode character database (over 150,000 characters) with instant filtering and searching. *…
I recently started work at a company([Fiserv](https://www.fiserv.com/)) that has quarterly Hack Weeks where you can work on anything you want. I wanted to brush up on my front-end skills, so I made a [Luhn Calculator](https://luhn.fiserv.com/) to calculate numbers which pass a [Luhn algorithm](https://en.wikipedia.org/wiki/Luhn_algorithm) check. Very useful if you need a random credit card number…
While working on [RSS Style](https://www.rss.style/), I discovered that a lot of RSS/Atom feeds in the wild are (cough) sub-optimal. The W3C has a [Feed Validation Service](https://validator.w3.org/feed/), but it only looks at the feed itself, not at the linked pages. And a lot of the problems that it finds are about things in the content which can mask more serious errors. Anyhow, give it a try…
I am thinking of using [Material UI](https://mui.com/material-ui/) for a future project and wanted to try it out in an actual project, so I built a sitemap viewer that reads a `sitemap.xml` file (and any linked files) and displays them in an outline. It works pretty well (IMHO) as long as you have paid attention to your URLs when you built your website. Links: * [Sitemap.xml…
As part of [RegexPlanet](https://www.regexplanet.com/), I have implemented the same basic functionality (a straightforward [JSON API](https://github.com/regexplanet/regexplanet-next/blob/main/CONTRIBUTING.md#backend-api)) with multiple technologies. I originally did them where-ever was easiest, but have been migrating them all to run in Docker, and it is interesting to compare the size of the…
This is a review of available logo APIs. The focus is on getting an icon (aka logomark) for displaying next to the text company name. It does not include sites that just have a lot of logos to look at, such as all the [websites indexed by LogoSear.ch](https://logosear.ch/sources/index.html). I also made a [logo API comparison page](https://resolve.rs/info/logo-api-comparison.html) for the ones…
One of the things that (IMHO) is hindering RSS/Atom is poor discoverability. The major browsers no longer support it, so clicking on a RSS/Atom link brings up either a "wall of text" or even worse, a "Save As" dialog, both of which are useless and especially hostile to newcomers. This has lead to authors avoiding putting visible links to their feeds in the HTML body. So now newcomers never even…
Are you losing track of all the repos you have? I am. Have you used Github Actions to automate everything in existance? I have. So I built a tool to generate a page full of Github Action badges. And, of course, it runs as a Github Action! (It also works from the CLI) It works really well for Github organization profiles: public or private. Note: badges for private Github repos are only visible if…
I saw a snippet in the Bun docs that [benchmarked a minimal Node.js webserver](https://bun.sh/docs/api/http#benchmarks), and thought it would be interesting to see how it would work with something a little bigger than "Hello World". It was surprising easy to do, with a lot of advantages: * No dependencies * No build step * Not even a `package.json` * Docker image is a single step straight from…
I know it is a little late to be joining the party, but I am starting a blog. I want a place to post random things that are not necessarily part of a specific [project](/side-projects.html) or a [things-i-learned](https://til.marcuse.info/) note.
