Code output right now is not meant to be used by humans.
Readable output will come when we create no code components.
The only thing current build is meant to do is to publish it as-is
Import is currently also not implemented
Remix is basically a router and data loader
Is that something that you guys are wanting to implement in the future, is that on the roadmap?
This is realistically the only thing that is making me a bit uneasy about Webstudio.
One of the biggest issues of WYSIWYG builders is messy, bloated code output. Webflow is the best when it comes to this, so I thought Webstudio prioritized similarly in that regard.
I think you guys have so much potential here for more technical users. The fact that code runs on the canvas is already amazing, if the code output was clean it would really be the perfect solution.
Anyways, thank you for releasing such a great tool, I can definitely see myself using it for smaller projects.
Def. Our code is clean af, just not meant to be human readable
You can switch in project settings to get a bit different output, but atomic is super optimized for production.
When you're comparing what we have done with old school dream viewer, sort of builders which simply output shitty code, you are completely misled
You are probably right. I am not as technological as you guys when it comes to development, so don't take my messages as bashing what you have built, I am just trying to understand the cons and pros a bit more deeply.
I just wonder, if you guys stop supporting Webstudio at some point, and Webstudio.is goes down - how can I make the code usable and scalable outside of Webstudio? Is there any other services that would be able to assist with this? What if a client decides to switch development to a more traditional agency? Will they be able to work with my site?
These are all questions I have been asked by agencies in my city when pitching Webflow, and the response has always been "Just export the code and integrate with another CMS".
- Webstudio itself is open source
- Yes in the future code we generate will be meant for humans as well
And that response just export the code and integrated with other CMS is the total b*** because the code you can export from webflow needs to be heavily modified to be used for a proper site and then there is actually no good way to use it with a CMS because you need to build dynamic server not just a static website
In case of webstudio you don't actually need to export at all what you do is you are using the CLI to generate the code and everything is already there The entire CMS integration that you have inside of builder is completely downloaded and it's a functional app that you can host now anywhere else you don't need to do anything with it which is why it's not really important how readable the code is because it's not for you it's for the server
The code itself will be readable once we have components because then you can actually generate code specifically as a design system or a set of components that you want to be using in a custom code base, which is a completely different use case
Agreed. Webflow's exported code does need to be heavily modified. I only really use Webflow as a template builder for the HTML / CSS, no interactions. (Which is why I was so excited to hear you guys embrace GSAP instead of a built in interactions system) So in my case transitioning into another CMS is much simpler than an average Webflow site. I use Statamic for my CMS of choice, it's been doing wonders for me in the process.
I am really looking forward to the new updates of Webstudio though. It is so close to being the perfect solution for me, I just gotta be patient for the updates. Thank you once again for creating such a great tool!
Good explainations, Oleg.
Great explanation Oleg and very interesting discussion. You should explain it on your blog on a more technical blog post! For sure it would be interesting and reassuring for clients and community!
That's a request for @John Siciliano
I have a feeling no matter how much we explain nobody reads it anyways
I am just honestly waiting for a better AI assistant so people can ask these quetions and get an answer from AI. We are almost there.
Problem is nobody wants to read an in-depth article, people want an easy to read explanation for their particular questions, presented in a digested way, which is only possible when you know the exact question.
I guess we should write articles for AI to feed it with informtion anyways.
You've right lot of people want to have direct answer (the same for my business nobody is reading the FAQ lol). But I read all your blog and it's very instructive and that's why I really a fan of your vision and product! So I think it’s 's useful but at some point its not prio. A discourse for SEO would be maybe a good idea!
I'm perhaps one of the rare exceptions. I LIKE reading and understanding the products I'm using. I've been through 90% of the YT channel videos and I THINK 100% of the help documentation. And the full Essentials course by @John Siciliano .
I somehow missed that you have a blog. Will check that out.
AI for questions will be great. But I hope you won't stop creating articles, documentation, and tutorials -- esp for the advanced and edge cases. I thrive on that. I need that because I always seem to be trying to do something just outside the norm. And usually get stuck.
I have read all the blogs, watch all the youtube videos, follow all the tweets and read the documentation. 🤓 I don't do this with any other project, tech, platform though.
Oh yes, I belong to the same class of individuals that tend to read up on all the docs and watch the tutorials before asking a question. Besides, in order to use a tool better, you need to read up on how it functions, any tips & tricks to optimize not just your own productivity using that tool, but for your overall workflow as well. I can't imagine not having access to Webstudio's blog and various modes of information dispersement. Please keep 'em coming! 🙏🏼
its a separate class indeed
