Sass 3.3 is released | Hacker News Reader

Sass 3.3 is released

blog.sass-lang.com

75 pointsFWeinb31 comments

> You can now write &-suffix (or &_suffix, or even &suffix if you really want) and Sass will make it work.

Now you can finally use BEM in Sass without the need of verbose mixins. Good stuff!

AlwaysTwisted has a really great guide on this for more info and copy-and-pastability. http://alwaystwisted.com/post.php?s=2014-02-27-even-easier-b...

So excited about this, absolutely brilliant. I did notice some frameworks that I was using (I think it was Bourbon) breaking with SASS 3.3 though.

I think they've been waiting on Sass 3.3 to release the next Bourbon update. There have been pull requests that were accepted long ago (hyphens mixin comes to mind) that still aren't in the released version of Bourbon.

Agreed this would be great but I would be curious to see how Bourbon would deal with their extensive and growing LibSass user base which has the feature set of Sass 3.1.

It's been a real bummer using sass the last 6-12mo with no sourcemaps support in chrome. It used to work, then Chrome changed something and it all went to hell.

I'm using it with rails, anyone able to get things working with sprockets, etc?

Still can't import pure CSS files without having to rename them to .scss. It's been in talks for over two years now, very disappointing.

I don't remember if I've tried it but I thought this was fixed in 3.0.

http://sass-lang.com/documentation/file.SASS_CHANGELOG.html#...

That adds a CSS import rule, doesn't embed the content.

What's so bad about having to rename a file?

If you're using dependency managers like bower you either have to import the stylesheet in your <head> or use the @include in a sass stylesheet, both will be a new HTTP request. Whereas if you compile it in one file, you save an extra request (now imagine you use multiple dependencies each with their own css file). Renaming a file would defeat the whole point of dependency management.

Just use the link to the CSS file in development an have your build tools concatenate everything for production. Certainly not a show stopper.

Whatever line-wrap/justification setting they're using is horribly, illegibly broken on mobile Safari. Some irony.

It doesn't look good on mobile Chrome, either, and in horizontal mode the menu overlay dominates almost half the screen.

Looking very good here at latest chrome on latest Android.

http://imgur.com/kuSeXfn

I wouldn't call random, non-hyphenated word breaks like that "looking very good"!

For some reason, at width <= 480, it sets

    word-break: break-all

which sets the character-splitting to CJK (Chinese/Japanese/Korean), where splitting between characters is OK. This should generally not be used on English text.

Perhaps the author meant to use

    word-wrap: break-word

Which allows it to split a word that is too long to fit on a single line, instead of stretching/extending outside of the text area.

The blog theme was made by a Chinese developer, which is probably why that is set the way it is.

While I don't dabble often in markup, I'm curious.

Is Sass still mostly entrenched in the Ruby world (any alternative implementations?) or has it picked up adoption with other devs compared to Less?

There's lib-sass[1] a C implementation, but it's not at feature parity yet.

[1] http://libsass.org

Thanks for posting that! I started looking for an implementation that I could embed/distribute easier just the other day. I didn't stumble across this one and it looks promising. This is one area tools built on Ruby, Node, and similar platforms often fall short.

There is nothing that requires Sass and other Ruby programs to be endemic to Ruby world. The programming language it is implemented in is an implementation detail after all.

This is technically, and therefore the best kind of, correct, but in practical usage there's Ruby and JRuby and, uh...does it run in IronRuby?

I mean, libsass is a thing, but it's not nearly feature-complete. You end up bolting a Ruby thing into your pipeline somewhere, which is a bit of a drag depending on how you're used to doing things. (For me, I just let Play handle it, but if somebody hadn't made play-sass I probably wouldn't be using Sass at all.)

I am not a designer, and the only place I use sass is a bunch of static blogs, including my primary one. The command I run to compile my sass files is as follows:

  find . -name '*.sass' | onchange -s make css

where onchange is a simple command I've written that watches filesystem for events on files from stdin and runs command when they occur, and make css is simply

  sass --update _sass:dynamic/styles

So long as sass is in the $PATH, I'm good to go. I don't care how my sass files are translated to css files, I only care whether they get translated or not. In the last place I worked, they used guard for compiling less (sth. like sass) files and concatenating them, which was a frightening program with a huge configuration file, written in JSON. For concatenation, I'd do

  find dynamic/styles -name '*.css' | onchange -s cat dynamic/styles/*.css > dynamic/all.css

See, when your abstractions are processes, everything is easy. It's like git, one simply does not need to know if git commit is a shell script or a C program. It is a process, you need to call it, and that's it.

edit: put in absent redirection into the last command line.

It’s seen wider acceptance. Some argue even more so than LESS. Sass is very popular in the Drupal community, for instance.

Personally, Sass was the first tool to really push my toolset towards its current polyglot of whatever works best.

I'm not a Sass expert here, but does anyone know if anything in this release helps with the problem of not being able to @extend items within media queries?

Yes, the new [@at-root](https://github.com/nex3/sass/blob/master/doc-src/SASS_CHANGE...) directive. It is verbose but will allow you to extend root level extends. However this creates selector spaghetti in your CSS and should be avoided if possible.

Sass 3.2.15 was also released today. I'm staying with it for a while to avoid breaking anything I have right now.

It's just some kind of script. It is not CSS.

No one claimed it was CSS.

Thank you for weighing in, professor. Yeah, I really can't imagine why would anybody ever care about this tool that tens or hundreds of thousands of people use every day to build web interfaces. Especially here, on Hacker News, of all places. What is the world coming to? It's not even CSS!

Reply on news.ycombinator.com