Speeds up compass compiling.
Please forgive me if this is not the place to put this issue.
That is the right place for this.
I feel like +1, but only if it does not create to much overhead in the generators Gruntfile. If watch is actively used for other tasks, I don't think we should switch compass over.
what about using it in conjunction with a paralleling task such as grunt-concurrent?
@wiebl That's what I consider as "much overhead". We've had lots of criticism about the complexity of the Gruntfile provided in our generators and we've been trying to improve that for the last couple weeks.
If compass is a speed bottleneck for your app, then feel free to customize it yourself.
But I'm pretty sure using the compass watch option + grunt concurrent is not a good option for most official generators. I'm not saying compass watch option alone isn't one, just the setup including grunt concurrent.
@SBoudrias fair enough :) I used to have my sass files compiled automagically using compass.app and that only took .5s or something while compiling with grunt can take up to 4s
I respect your first-things-first approach 👍
Even when I don't setup a generator to work with compass, grunt takes around 3s/4s to compile my .scss files.
This makes live reload a bit useless for me.
👍 the slowdown in the compass compiling for livereload is a downer.
Anyone would like to do a PR on the webapp generator?
What about use Libsass to compile sass files?
Using compass watch in a concurrent task hides output from Compass to the terminal. IMO debugging trumps speed for the serve / development task, so one vote against.
@bernardogfilho i think that, speaking about changing the compiler, grunt-sass and/or node-sass would be easier than a C library. I'm trying to open that discussion in #1233.
Yes, you're right @alejandroiglesias.
I just saw the discussion and it seems we need wait for a more stable future for libsass and node-sass.
Closing for lack of activity. We're better off just using node-sass when it stabilizes.