Add support for Gulp

[Pentiado]

I've been working recently on awesome project and I was tried of waiting until my code gets compiled and browser refreshed. It is not perfect, on my old macbook pro I have to wait like twice longer than when editing in Chrome console but I understand that compiling coffee or stylus and page refresh might take a while, about blink of an eye.

What it do?

• most of the stuff that grunt but FASTER and in awesome way ; )
• sass
• first question asks if you want to choice experimental Gulp instead of Grunt so there will be no harm for Grunt lovers

What it can't do yet?

• compass

[eddiemonge]

Can you strip out the stylus stuff for now? One change per commit is better. Also, your commit message needs to be on one line

[Pentiado]

Done. I had issue with yeoman, very similar to this one yeoman/yo#69, I thought that problem was in my code but reinstall of yo helped.

[addyosmani]

Some thoughts:

I would love to see us break down generator-angular into generator-angular-grunt and generator-angular-gulp where both of these generators are offered as options and utilize our composability feature to call out to them.

@eddiemonge would this be a good opportunity to go down this path or did you want to land this option directly in this generator and then consider breaking things out at a later point?

[Pentiado]

Breaking this down to separate generators is for sure better option than assuming that we will be able to keep everything in sync forever but this option leads to separating community and it's as good as building new JavaScript framework each day.
Grunt and Gulp are just tools that are making work easier and faster. I think for most of the programmers it really doesn't matter which tool is used, they just care if it do everything what they want, how fast it is and how easy they can add more things that they need.
My suggestion is to give Gulp a shot as experimental part of generator-angular and observe how it will evolve. Community might love instant result after hitting save and the fact that they are writing JavaScript instead of some strange configuration objects so much that there will be no need to support Grunt anymore.
When I heard about Gulp I was really angry that someone created next a'la Grunt tool but those two podcasts convinced me completely that Gulp should become number 1 between task runners.
http://javascriptjabber.com/097-jsj-gulp-js-with-eric-schoffstall/
http://javascriptjabber.com/074-jsj-grunt-js-with-ben-alman/

Anyway, if you decide to split it into two generators I volunteer to be a maintainer of the Gulp one

[eddiemonge]

I do like having them in the same project.

Couple of things:

• Needs tests added
• Needs testing (not sure who as I might not have time to get to it)

[sindresorhus]

Also need to fix merge conflict.

[ianks]

drools in anticipation

[bsiddiqui]

+1

[sindresorhus]

ping merge conflict :)

[stereokai]

I tried cloning this repo on my Windows machine to resolve the merge conflicts, but alas. Github for Windows keeps failing.... It seems there are very long filepaths in this repository and Windows can't handle them. I guess I'll have to manually duck-tape the generated app locally with this Gulp setup.

[hppycoder]

Until the community has complete census across which build system it wants to implement we will always have to maintain parity between gulp and grunt. As a company we used grunt and are moving away from it.

I am agreeing with @addyosmani on this issue. I believe that it should be broken into a generator-angular-gulp project. Keep the current one as grunt as the default, but does have a fork for gulp.

Also found this project: https://github.com/henyojess/generator-gulp-ng

[color2life]

+1

[fdxuyq]

Can support Gulp now? Waiting for the feature

[arthurvr]

@OneSeven There's still an open PR, #837.