# Have you considered HTML character entities?

### Yihui Xie / 2018-07-11

As a software developer, I prefer lightweight stuff and avoid dependencies when possible. Over the years, I have been leaning more and more towards “frugality”. I used to be big fan of libraries like jQuery and FontAwesome, but now I rarely use them any more. I started to write vanilla JavaScript after I suddently discovered document.querySelector() one day (before that, I thought \$() was the only way to select a node in the world).

Similarly, I stopped using special libraries such as FontAwesome for icons and symbols after I looked through the full list of HTML character entities. Yes, this list of symbols is much shorter than that in FontAwesome, but I often refrain from using fancy symbols, so this list is rich enough for me.

Most of the time, I only need a couple of symbols, so I really don’t want to import a full external JS/CSS library. For example, yesterday I wanted to add icons after those external links on the rmarkdown website, to indicate that they will be opened in new windows. Eventually I found and added the entity &boxbox; (double boxes ⧉), which was a perfect symbol for my purpose.

Last year, when I was designing the hugo-xmag theme for Hugo, I also got an inspiration from the list of HTML entities: I realized I could easy turn a letter into a Blackletter, e.g., &Afr; is 𝔄, and &Bfr; is 𝔅, etc. So the default site name is generated in Blackletter in the hugo-xmag theme. It looked cool to me, and didn’t need any third-party dependencies.

Certain HTML entities may not work in all web browsers, but I don’t really care. For my personal websites, I have no intention to support legacy browsers.