I was, on November 14th 2014, at the first edition of dotCSS, a conference by dotConferences, also responsible for dotJS (which was 3 days later), dotScale, dotGo, the future dotSwift and the one-shot discontinued dotRB.
Long story short: dotCSS was a good conference, which clearly found its audience, since the venue was quite full, but did not find its sponsors, which led to the reduction of its duration (from a day to an afternoon) and a dramatic lack of dotCSS T-shirts.
The absence of big sponsors is also probably the reason why no 2015 edition was announced (when dotJS 2015 already is): basically, only Google helped a bit. Since there were plans for combos tickets this year, I can only assume that dotCSS's future is uncertain.
I won't detail each and every talk, but here's what I'll remember from dotCSS:
Daniel Glazman, co-chairman of the W3C CSS Working Group, gave a talk explaining CSS' various problems, and how hard it seems to be for it to move forward while trying to reach good compromises on browser vendors needs and priorities. For example, he mentionned the infamous difficulty to vertically center elements of CSS, and talked about how hard it was to even name properties.
Kaelig Deloumeau-Prigent talked about his work at The Guardian to bridge the gap between developers and designers. Sass, and especially clearly named variables, allow better understanding for both developers and designers of how front-end elements are thought and standardized throughout the project. His idea was to put the design at the center of the project.
Harry Roberts (@csswizardry) presented his 10 principles for effective front-end development:
- The simplest option is usually the best
- Reduce the amount of moving parts
- Understand the business
- Care less, care appropriately
- Pragmatism trumps perfection
- Think at product level
- Do not design systems around edge-cases
- Do not make decisions based on anecdotal evidence
- Don't build it until you've been asked for it
- Expect and accommodate change
Maxime Thirouin summed up what was basically the feeling of the whole conference : "We love CSS, but it's frustrating". That it still has (without preprocessors) no variables, math, customization... which is why he is building cssnext, to get the next standardized features now.
Victor Brito told all web devs to put more efforts to make the web accessible
Gregor Adams talked about functionalities in CSS to dynamically style elements (http://slides.pixelass.com/dotcss2014/assets/player/KeynoteDHTMLPlayer.html)
Guido Boman talked about CSS parallax scrolling done right, and an implementation of it by Keith Clark full css implementation of parallax scrolling. Full details at http://keithclark.co.uk/articles/pure-css-parallax-websites/
Finally, Tim Pietrusky talked about one-element rules.
Hugo Giraudel, in his talk Keep calm and write Sass, talked about how preprocessors are used to bridge the gap between the simple beginnings of CSS to style webpages and the complex web applications it must now give style to.
He presented principles to work with Sass:
Estelle Weyl talked about less-known things CSS can do, like counting with counter, then about selectors specificities in her CSS? WTF! talk, which ended with notes about how to (not) use !important and how to override it with CSS animation quirks.
Nicolas Gallagher talked about scalable CSS, and how, in a large company like Twitter, the whole UI has to be designed in order to avoid having developers stepping on each other's toes.
Bert Bos, co-creator of CSS, talked about typography, giving clear examples of the challenges linked to it, which may or may not be fixed with CSS. For example, quotes are different in French, English, Dutch ; punctuation and style can be different (spaces before ! and ? in French, not English), so is it possible to markup in a consistent way text?
Ana Tudor (@thebabydino), the last speaker, showed a series of neat demos about distribution, creating fun shapes with CSS/Sass. Her neat demos are on Codepen.
And that's it for dotCSS!
It'd be a shame if the conference does not return. Perhaps if it was part of dotJS (or a dotHTML or something like that) as a 2-days event, part of it could survive?