First stable version

[jr-cologne]

Hi everyone,

I am aiming to release the first stable version of this Gulp Starter Kit by July/August this year at the latest.

In order to achieve a version which is very comfortable to use and also reflects the needs of your projects as good as possible, I am now asking for your suggestions and feature requests.

I really appreciate any ideas for potential improvements and would be happy to discuss them with you. So, feel free to comment on this issue!

In case you don't really have own ideas, here's a list of features and improvements I am already considering to implement:

• support of HTML templating languages like Pug (first implementation: #19)
• support of more CSS preprocessors (e.g. Less, Stylus) (first implementation: #21)
• better directory structure for Pug: (implemented: #20)

  src/pug/
  partials/_header.pug, _footer.pug
  helpers/_mixins.pug
  templates/_default.pug
  index.pug

• option to convert images to webp (not before beta)
• use incremental builds for better build performance (implemented: #25)
• CLI improvements:
  • configuration options for installation process (e.g. choosing the CSS preprocessor)
  • improvements regarding the UX and UI of the CLI (e.g. better visualization with chalk and ora)

Please let me know what you think about the current list as well.

Thank you all in advance!

Best regards, @jr-cologne

[ivan-vilches]

I am using it, works nice thanks you very much, its possible if you have template engines likepug it could be optional? thanks a lot

[jr-cologne]

Hi @ivan-vilches,

thanks for your feedback. I am really happy to see people using it.

Templating languages and such things will definitely be optional. So, you won't have to worry about this once version 1 is released.

Best regards, @jr-cologne

[revelt]

hi @jr-cologne Thanks for sharing this kit, I think it's good. Often devs are not courageous enough to share their own boilerplates or don't bother sharing at all.

It's very likely that this kit will be used only as a starting point or only some pieces will be "borrowed" so I wouldn't worry much about extending too much or covering all possible stacks. If somebody needs Nunjucks instead of Pug (like me), they already know what they are doing and they'll wire it up in no time.

I think what really matters is performance. Specifically, putting a good foundation, to prepare for later, when user's Build System that originated from this kit will grow very big.

Speaking about features, I haven't noticed incremental builds here, maybe it's worth having a look.

In general, it's easy to wire up a Gulp starter system. It's hard to wire it up efficiently and do it early in dev lifecycle. What happens is, we often wait until Gulp build system is too sluggish and then we implement perf features (like incremental builds). But by that time it's too late, the Gulp Build task is ginormous and a bit messy from all that patch-editing over months, so it's often easier to start from scratch, to seek some starter kits (like this)...

[jr-cologne]

Hi @revelt, thank you so much for your feedback! To be honest, I didn't know that something like incremental builds even exists. I am definitely going to implement this very soon. Best regards, @jr-cologne