I am not understanding the value of the open source builder, if I am not able to download the code and host it on my own.
What does the opensource gives me additionally apart from integrating the studio on my own website.
The hobby/free plan already has unlimited projects, pages and team members.
Looking into the builder, I cannot see the code but just the css. Would have been good if I could create one of components and be able to download the code using the open source builder.
it’s open source you can add the feature to download the code yourself if you want it. being open source is also just a good thing in itself because it’s contributing to the total body of code available for others to extend and use in whatever way they want.
Syncing and publishing what? What's so difficult about `git push` which pushes to GitHub and automatically deploys to Vercel? This doesn't look cool, but those two features are not going to be the ones I'll seek after.
Syncing the site data from builder to local fs and then building everything and configuring the app for publishing to vercel.
Just pushing to github only helps you if you only have static html/css if you wanted to publish a site as gh-page ... I guess you need to clarify your use case
Funny enough a WYSIWYG editor is the reason Radix components were originally created. I worked on some of the original Radix components while working at Modulz[1].
The main offering is their primitives (https://www.radix-ui.com/primitives): essentially skeletons of common components that plain HTML makes difficult to style.
They take care of cross-browser compatibility and accessibility and you apply your CSS to their wrapper element. (generally the term for this is headless UI components)
Most recently they released a styled version of those primitives
I feel like there aren't a lot of use cases for custom elements (W3C web components), but a generalised component library like this is probably one of those use cases.
Particularly given the ability to style shadow pseudo-elements like drop downs in a non-hacky, browser supported way.
Personally I refuse to signup just to check out something I'm only passively interested in, and I will never understand why startups do this. I imagine they're immediately turning away >80% of potential customers vs. just letting people try out the damn product
Any product that allows you to consume lots of resources including API calls, interact with other users, or share anything basically needs to be authenticated to avoid misuse. This covers virtually everything.
Sign in with google/apple can be virtually painless and only lets them know your email address, name, and profile pic if oauth is configured reasonably and its easy to tell on your side if its not because it asks for the additional permissions. You can also sign up with email without making the person go through a lot of malarky. Simply send them a sign in link instead of doing the whole normal dance. Click it and continue on.
If need be they can be prompted to fill out additional data if features require it.
Yea but if I haven't even seen the product yet, I'm not going to bother and give away my email + personal info.
I understand sometimes you need auth, but the app could just show the live app and then pop up a signup modal when the user tries to do something that requires auth (that's what I do on my apps)
It’s not the only way to prevent misuse for example: captcha’s, rate-limiting etc…
I bet they implemented the login anticipating misuse… and like all other startups, there’s only a tiny chance the product would be misused, but a large chance that many people won’t use it due to the login wall.
Effective captchas a much worse than logon with google or send a login link to email. Rate limiting might prevent the system falling over or spending all your money but it does near nothing for vandalism.
Rate limiting is like handing your football players packs of condoms instead of cups. It might be necessary but it sure as hell isn't sufficient.
I would suggest that if your app requires interaction with others you provide them with a test experience where they can read live data but not effect others. Gate functionality that might be misused with a request for an oauth2 login or email which you can send a login link to. Near zero commitment don't have to share anything beyond your email and name don't even have to make up yet another password. A few clicks and you are done.
You can provide almost as good an experience as you want without people getting their lols or their dollars off your other users.
i doubt 80% figure. Email is constant source of abuse and requires additional infrastructure so more and more indie products are sticking to social logins only.
I prefer the split approach. I use SSO on the inside of my home network and for hyperscalers and things like Tailscale. At least partly so I could enforce hardware auth when not supported or sane to implement natively.
Good for you. Keep your SSO login. Nobody's asking for it to be removed. But, your SSO login isn't going to be affected if email login is added, would it?
It is a privacy leak to tell Github about what other websites or products you are using, that you otherwise wouldn't if you just used normal email/password signup.
GDPR only has a tangental relation to privacy. It might be "GDPR compliant" for me to publish my nudes to twitter, but I've still (voluntarily) lost some privacy for doing so.
Standard compliance, especially in something like a website builder should be a top priority. If the builder doesn't work in major browsers, why there be any confidence that the sites it produces will?
So, you are saying that there is no plan for it to work outside of Chrom(ium) browsers? It doesn't solve the problem of forcing someone to use a specific browser to use the platform.
It does not inspire confidence in an end product will work everywhere and be of high quality when the product to create it is buggy and limited.
Clicked pricing. It's all per month - no way to own it / buy - it's a lock-in-lease, where "your monthly bill only increases when you're making more money!"
A little off-topic: I am seeing the same design in 1/3 of links that get to to HN front page. The design is clearly a copy of the Vercel's design (and maybe they stole it from somewhere else). The same typography, gradients, border radius, and definitely the same blur effect.
I am happy that 'Corporate Memphis' is dying, but I am not happy it is getting replaced with another fad.
Makes me wonder, does someone keep track of web design trends trough history?
I don’t understand why downloading code is such a difficult feature for a website builder — I get the competitive aspect. BUT! It would bring so many users to your system and really would be great marketing — why not compete by giving people what they want and just practice love in not lock in? Especially when you’re already there with the OSS license?
I haven’t heard this phrase before, it’s such a great way to describe how I like to do things, which I would have written as “compete on quality rather than treachery”
Lots of companies will behave this way... when times are good. They can be fairweather friends at best. Eventually it's time to batten down the hatches, and that's when licenses change, interop gets removed, etc. like we've seen again and again.
