Planning to fork jekyll-s3 and create a gem for deploying static websites to S3 – what do you think?

[laurilehmijoki]

The only thing jekyll-s3 has to do with jekyll is the assumption that the website resides in the _site directory. Otherwise jekyll-s3 has no dependencies whatsoever to jekyll.

With small changes it is possible to turn jekyll-s3 into a more generic S3 website deployment tool. I've started the work and created a new gem called s3_website. It is based on the jekyll-s3 codebase, and you can use it like you use jekyll-s3 if you rename your config file from _jekyll_s3.yml to s3_website.yml.

I've chosen the name to be something as boring and concrete as s3_website, because I find the naming of Unix tools like rsync, cp and find clear and descriptive. A potential problem here is that it's difficult to google a gem called s3_website, because the web is full of resources that address topics like S3 + website.

What do you think? I welcome any feedback on this matter.

Edit:
Unlike jekyll-s3, s3_website uses thor. It is now possible to say s3_website push – and the gem will synchronise your local files with the S3 bucket. New users can create a placeholder config file with s3_website cfg create. I'm also thinking of adding the command create, which would behind the scenes call configure-s3-website --config-file s3_website.yml.

2nd edit:
You can now sync any directory with the S3 website bucket by specifying the option --site. E.g., s3_website push --site my/website will push files my/website.