You do realise zepto works in IE10+ only, and even then you need an additional module for support. I wouldn't say it's worthwhile. It is not a viable alternative.
I'm working on a widget that other people will embed on their page. I cannot make assumptions about jQuery being available and I cannot afford to add jQuery as a dependency. Assuming the widget was valuable to you, would you want to add my widget to your page if it included 94kb (jQuery 1.10.2 minified) of JS before my code was even added on top? If you cared about performance, it's highly unlikely you would.
So for that reason I've been working with plain JS for months now. We have browser support back to IE9 (maybe even 8), and our entire codebase is around 25kb minified (gzipped < 6kb). We have one or two collections of utilities, but most of it is application code that would still be there even if we had jQuery as a dependency. All I can say is it is not that big a problem. If you're not sure how to get something to work across browsers, look at the source of jQuery (this website [0] is fantastic for this purpose) or any other library that is well regarded.
But you know what? Most of the time it isn't even a problem. And with modern tools [1][2] and a good test suite, it's not difficult testing across multiple browsers to quickly find issues.
Before even working on this project I decided to go on a self-administered "jQuery diet". I haven't used jQuery in any personal projects [3][4] for at least a year and it's great. It took a little adjusting, but not much. If anything it was a little shocking. I thought I was a good JS dev, but it really opened my eyes to how little I knew about the DOM and other native browser APIs and, honestly, I felt a little ashamed.
Conclusion: sometimes going it alone is the right decision. jQuery is absolutely the right choice in some environments, but it doesn't make it the right choice absolutely
I'm working on a widget that other people will embed on their page. I cannot make assumptions about jQuery being available and I cannot afford to add jQuery as a dependency. Assuming the widget was valuable to you, would you want to add my widget to your page if it included 94kb (jQuery 1.10.2 minified) of JS before my code was even added on top? If you cared about performance, it's highly unlikely you would.
So for that reason I've been working with plain JS for months now. We have browser support back to IE9 (maybe even 8), and our entire codebase is around 25kb minified (gzipped < 6kb). We have one or two collections of utilities, but most of it is application code that would still be there even if we had jQuery as a dependency. All I can say is it is not that big a problem. If you're not sure how to get something to work across browsers, look at the source of jQuery (this website [0] is fantastic for this purpose) or any other library that is well regarded.
But you know what? Most of the time it isn't even a problem. And with modern tools [1][2] and a good test suite, it's not difficult testing across multiple browsers to quickly find issues.
Before even working on this project I decided to go on a self-administered "jQuery diet". I haven't used jQuery in any personal projects [3][4] for at least a year and it's great. It took a little adjusting, but not much. If anything it was a little shocking. I thought I was a good JS dev, but it really opened my eyes to how little I knew about the DOM and other native browser APIs and, honestly, I felt a little ashamed.
Conclusion: sometimes going it alone is the right decision. jQuery is absolutely the right choice in some environments, but it doesn't make it the right choice absolutely
[0] http://james.padolsey.com/jquery/
[1] http://karma-runner.github.io
[2] http://vanamco.com/ghostlab/
[3] https://github.com/WickyNilliams/headroom.js
[4] https://github.com/WickyNilliams/enquire.js