Add Sass Support

[gesa]

Because some people don’t like LESS I suppose.

[gvn]

:+1:

[k88hudson]

@gesa Personally I don't think this is worth maintaining since none of our apps use Sass

[gesa]

Yeah I am going to make this a low priority, but it is worth discussing whether or not we wish to add Sass (and ruby) to our stack if it is the right/a better tool for the job.

[jbuck]

If we decide to switch Makerstrap and our tools and libraries to SASS, we need to have really good reasons to do so. I am firmly against switching anything over to SASS because our entire technology stack already heavily uses LESS.

Additionally, I think that adding any non-node dependency (ruby or compiling libsass via node-sass) is something we should avoid. Having a single thing (node) that you need to install to contribute to any Webmaker app is a win for both paid contributors and volunteer contributors.

[humphd]

Yeah, I would echo @jbuck here. I think this is less about LESS vs. Sass, node vs. ruby, and more about the fact that we've got substantial investment, as a company, project, and community, in a particular stack already. There was a time that switching the stack made sense (i.e., when we moved away from python/django and toward node), because we were at the zero point of what we needed to do, and anything we did was new. Now, however, we're maintaining and creating things on top of an existing and well established base.

I'm going to close this. I think the spirit of this bug is right, though (I read it as "let's try to also include this other awesome tech, and the people who like it."), and as such it was good to flag it. We can and should evolve our stuff where the benefit of doing so brings us discernible value--the introduction of Angular was one such example, I think. But I'd encourage us to be pragmatic about debt we'll incur due to switching technologies otherwise.

[gvn]

Hey all. I think there's some confusion here. @gesa isn't suggesting we deprecate LESS in Makerstrap, simply that we support both LESS and Sass in Makerstrap. This doesn't mean that we have to switch over existing projects from LESS to Sass. That can be decided on a project-by-project basis. We are not inherently incurring technical debt with this approach. For example: I personally would NOT switch webmaker.org over to Sass (I already tried to mixed results). However, for some of our new apps: Profile and Events in particular, the migration effort would be quite minimal.

Personally, I think we should try to support Sass in Makerstrap. It's got much more velocity and usage in the front-end community. Better tooling also exists for Sass. Makerstrap is a mostly non-domain specific library. We should be able to attract more users and contributors by supporting both precompilers.

I think this bears further discussion and understanding before we decide to close this ticket. I'm going to re-open it so that can occur.

[gvn]

I also wanted to add that AFAIK node-sass has no Ruby dependency.

[toolness]

Since Makerstrap seems to be something that we're trying to use to make the Webmaker community ("big tent") bigger, I think supporting both LESS and SASS potentially lowers the barrier to entry, since any developer can use whichever one they're already familiar with--or whichever one is best for their task at hand.

That said, if adding SASS support is going to be something that quickly falls into disrepair, such a strategy might end up backfiring. It's cool that Bootstrap offers support for both, for instance, but it seems like their community can continually support such a thing.

[gvn]

Yeah, I think that's probably the real question here: what is the LOE for maintaining Sass and LESS in parallel for MS, and is it worth the time? Will integrity be maintained? Honestly, I'm not sure. It's certainly difficult to quantitatively measure the benefit.