We just released a major update of how your app’s CSS is rendered in the browser (no more inline style), making your apps faster to load both:
• in the WeWeb editor while you’re building, and
• for users visiting your published app.
You’ll notice the biggest improvements on larger projects and slower connections
The new architecture also significantly reduces memory usage: in our internal tests, RAM consumption was roughly cut in half.
But there’s more behind this update than page loading speed…
It also lays the groundwork for more upcoming improvements to how we serve public pages, making your content easier for search engines and AI assistants like ChatGPT, Claude, and Perplexity to discover and understand.
This is a welcome improvement. The email that went out the other week about checking custom components didn’t really allude to the performance impact. Thankfully I didn’t have any to fix. I do appreciate the pre-made ai prompt that was provided. That made the check much easier.
Interested to learn more about the ai discoverability improvement. Sadly, I just spent the last two weeks creating a new marketing/landing page site for my app using Astro soley for improved ai discoverability in search queries. That meant moving the WeWeb app to a subdomain. A little scary considering it’s a live production app, but it all worked out.
As a side comment… is the downturn in the community engagement a sign of slowing user adoption of WeWeb, or just things working better? Not a whole lot of discussions like there was a year ago.
That’s huge! I do feel the that the editor is way faster too
And as a side effect, it also unlocked style inheritance! If I set a property to a parent, the children will also inherit them. For instance, having the text props available at the div level can simplify some use cases
I notice the speed increase, which is great. BUT how do I perform a smooth transition to the new verion?
I re-published the app.
Every user should refresh its browser - that should normally do the trick.
However, today I had a couple of users that reported that the webapp was unresponsive and a refresh did not work. For example: a button that should navigate them to a certain url was unresponsive, but when the url was typed in the browser the webapp did work/navigate. What can be the cause of this? And what can/should I ask the users to do next time an update is rolled out? Is there a quick way to do a full reset of the webapp on a browser?
Note that I asked this before: updates are announced as if there’s nothing to do on my part. After some questions on my behalf earlier, it turned out that after each update I should test the entire webapp to be sure. But now I find out (of course in worst moment when an user has a meeting with its own clients) that things don’t work - that probably could have been prevented. I’m just not sure how.
The nice (even though unprobable) hypothesis would be that they’re routing (and handling) the issues to their support ticketing system and so there are less people asking in the community, the more realistic one would be this.
Another hypothesis would be that AI handles more of the questions, witch decreases the use of a forum. I did not do any research to confirm or deny, but it could be plausible.
Also google trends - I do not know if they manage to see if the interest is lowering vs just moved to AI outside of googles AI.
While people that are familiar with coding today might be much more likely to build directly using claude code or similar, the number of people that are not familiar with code are probalby 1000x in quantity, and are much more likely to choose a nocode solution assisted by AI even today.
Only Weweb knows if their customers trends the same way. But seeing the huge improvements that weweb has done the last 6 months, with MCP that actually are helpful (compared to weweb’s own AI), CSS changes that hugely helped speed, prerendered pages helping SEO for static pages, etc. etc., I would think they are on the right path to survive in todays fast evolving market, even as a nocode solution.
