# Chapter 2. CSS **URL:** \ **Category:** 2019 Web Almanac\ **Created:** [October 30, 2019, 11:57pm UTC](https://discuss.httparchive.org/t/chapter-2-css/1757 "2019-10-30T23:57:17Z")\ **Posts on this page:** 9\ **Page:** 1
**Author:** ![rviscomi](https://yyz1.discourse-cdn.com/flex035/user_avatar/discuss.httparchive.org/rviscomi/32/835_2.png) [@rviscomi](https://discuss.httparchive.org/u/rviscomi)\ **Post date:** [October 30, 2019, 11:57pm UTC](https://discuss.httparchive.org/t/chapter-2-css/1757/1 "2019-10-30T23:57:17Z")
Share your questions, comments, and ideas about the CSS chapter of the Web Almanac… ---
**Author:** ![ldevernay](https://yyz1.discourse-cdn.com/flex035/user_avatar/discuss.httparchive.org/ldevernay/32/1963_2.png) [@ldevernay](https://discuss.httparchive.org/u/ldevernay)\ **Post date:** [November 14, 2019, 2:53pm UTC](https://discuss.httparchive.org/t/chapter-2-css/1757/4 "2019-11-14T14:53:23Z")
It could be great to gather some data on unused CSS (which drags performance and sustainability down). ---
**Author:** ![bodo](https://avatars.discourse-cdn.com/v4/letter/b/f0a364/32.png) [@bodo](https://discuss.httparchive.org/u/bodo)\ **Post date:** [November 18, 2019, 9:52am UTC](https://discuss.httparchive.org/t/chapter-2-css/1757/5 "2019-11-18T09:52:52Z")
Small correction > (e.q., `mm` , `in` , `cm` ), should be a `g` not a `q` 😃 And yes, I wasn’t aware the q-unit existed. My guess for the usage percentage would be wordpress. ---
**Author:** ![rviscomi](https://yyz1.discourse-cdn.com/flex035/user_avatar/discuss.httparchive.org/rviscomi/32/835_2.png) [@rviscomi](https://discuss.httparchive.org/u/rviscomi)\ **Post date:** [November 18, 2019, 8:27pm UTC](https://discuss.httparchive.org/t/chapter-2-css/1757/6 "2019-11-18T20:27:06Z")
@zcorpan noted that something weird was going on with the Q unit: > I filed [this issue](https://github.com/HTTPArchive/almanac.httparchive.org/issues/513#issue-524579367) to investigate and I think there was a bug in the query that falsely counted base64 gibberish. I’ve rerun the query with a fix and we’ll be updating the chapter accordingly. @bodo the typo you found is also in this section so we’ll get that fixed as well. Thank you! ---
**Author:** ![rzegnam](https://avatars.discourse-cdn.com/v4/letter/r/d2c977/32.png) [@rzegnam](https://discuss.httparchive.org/u/rzegnam)\ **Post date:** [December 19, 2019, 8:12am UTC](https://discuss.httparchive.org/t/chapter-2-css/1757/7 "2019-12-19T08:12:06Z")
Guys, thank you for the great overview! I am just beginning my story with CSS and thanks to your article already have an idea what to learn. ---
**Author:** ![tpowell](https://yyz1.discourse-cdn.com/flex035/user_avatar/discuss.httparchive.org/tpowell/32/1712_2.png) [@tpowell](https://discuss.httparchive.org/u/tpowell)\ **Post date:** [January 9, 2020, 5:19pm UTC](https://discuss.httparchive.org/t/chapter-2-css/1757/9 "2020-01-09T17:19:03Z")
The HTML survey acknowledged the sample bias problem of tags in the survey because of the uniqueness of homepages and their typical tags. In the CSS case I wonder given the frameworks being a 25%+ how much of the data is a function of the weight of the way the competency of the framework authors is being recounted versus what vanilla CSS writers tend to do. I think you would see a very different answer if you subsetted to just vanilla CSS and compared. TL;DR - web platform knowledge is very unevenly distributed and the repeat count of authored libraries and their tactics likely significantly skews the data suggesting more awareness than there likely is. On a more !important note 🤣 how could use miss sampling the usage of everyone’s favorite/hated kludge? Seriously though !important might be quite interesting esp compared to the library thing mentioned above (with and without the frameworks) Look forward to more data analysis coming soon. ---
**Author:** ![Nooshu](https://yyz1.discourse-cdn.com/flex035/user_avatar/discuss.httparchive.org/nooshu/32/1476_2.png) [@Nooshu](https://discuss.httparchive.org/u/Nooshu)\ **Post date:** [February 18, 2020, 12:50pm UTC](https://discuss.httparchive.org/t/chapter-2-css/1757/11 "2020-02-18T12:50:45Z")
I’d love to see some information about the `font-display` property in terms of adoption. Specifically around what values are being used. My guess would be `swap` would be the most popular (if used), since it seems to be the most talked about. I’m interested in seeing when and where `optional` is used, as for low-speed / unstable connections I can see a real advantage in the browser being able to abort the font request completely. ---
**Author:** ![tunetheweb](https://yyz1.discourse-cdn.com/flex035/user_avatar/discuss.httparchive.org/tunetheweb/32/1671_2.png) [@tunetheweb](https://discuss.httparchive.org/u/tunetheweb)\ **Post date:** [February 18, 2020, 1:04pm UTC](https://discuss.httparchive.org/t/chapter-2-css/1757/12 "2020-02-18T13:04:04Z")
Your wish is our command - you’re just looking in the wrong chapter 😀: [https://almanac.httparchive.org/en/2019/fonts#fighting-against-invisible-text](https://almanac.httparchive.org/en/2019/fonts#fighting-against-invisible-text) ---
**Author:** ![Nooshu](https://yyz1.discourse-cdn.com/flex035/user_avatar/discuss.httparchive.org/nooshu/32/1476_2.png) [@Nooshu](https://discuss.httparchive.org/u/Nooshu)\ **Post date:** [February 18, 2020, 1:04pm UTC](https://discuss.httparchive.org/t/chapter-2-css/1757/13 "2020-02-18T13:04:51Z")
Doh! Many thanks @tunetheweb 🙂