IMHO Everyone should write at least one framework of some kind. Bonus if it's open source.
It's a huge learning experience. Gives you a deeper understanding of how other other frameworks work. Helps you understand user and customer support. How to write documentation. Great for your resume. Etc.
In other words, if you only ever consume frameworks it makes it difficult to know what to do or how to approach solving some of the lower level issues a framework may present.
Do you have any personal experience in doing so? They seem like tempting projects from the outset, but getting started with some clear goals in mind, alongside philosophical/idealistic differences from what's already out there so you're not simply rewriting something for the sake of it has prevented me from getting too far down a particular rabbit-hole.
My first php framework was something called php4gl back before ruby on rails was created. It used the command line to bake controllers and such. Just a private project never released. Lost the code now!
I also wrote another couple more php frameworks. I used one for all my personal side projects. And I wrote a couple for day jobs website over the years.
Actually it wasn't until Laravel came out that I was convinced NOT to write or use my own php framework ;)
So, now I've stopped writing php frameworks. But, I still write a CSS component framework from scratch for every project.
I wrote a really quirky javascript framework called Double Dollar that worked using a kernel pattern. Anyone I showed it to thought it was nuts. I abandoned it and now just use jquery for projects I work on like https://nugget.one
Probably the most sucessfull open source project I created was ezSQL. Wordpress ended up using it as their original database class and then as the API interface for their current wpdb database class. I think a lot of people probably curse me for that ;)
I like fast, small web pages because not everyone has a broadband connection at all times. Can you tell me what other mature, tested, well-supported CSS frameworks have a feature set that large in less than 25% of Bootstrap's size?
[EDIT] - BTW, I mean this with less snark and more did-you-know-all-these-existed? I knew of purecss & skeleton (and I've forgotten a few, because it used to be very common to see these), and I've seen bulma before but some others were new to me (and I favorited them to give a try on some random project)
PureCSS was one of my favorites for a very long time. These days I use component frameworks like ElementUI[0] more than anything else to maintain consistent styling (since most things on the page happen to be components).
For me, the default answer is "yes" for many reasons:
* I'm not smart enough to handle most browsers well in a short amount of time
* I'm not primarily a designer and these frameworks are often gorgeous by default
* I don't want clients to pay for me to reinvent the wheel, and I don't want them left with my personal framework when there are many, better documented ones available
* These designers thought more than I did about accessbility, and I place a high priority on it
* These frameworks can sometimes deliver a lot of bang per byte, and I'd rather be writing stuff only I can do than wasting time trying to get mine as small and efficient as theirs
Perhaps the issue is with the constant use of the adjective "modern" in many new libraries and frameworks. It implies that there are no other notable benefits than the implementation being based on newer technologies and development practices.
Is it faster (than what's current out there), more stable or easier to use, would be the answer that most prospective users look for...
Curious what people are using css frameworks for these days. I used to use minimal ones to get some kind of grid support. But now I just use css grid. Are people still using frameworks due to needing to support old browsers?
We don't need another "lightweight, responsive, modern" CSS framework.
It's a solved problem.