Nearly unusable. Changing the sizes of elements lags a few seconds. Changing font sizes. Moving stuff around. Everything lags a second or two.
Tried different browsers, same deal
Using a Mac M1
Internet connection on fiber, low latency
Anyone has found a solution here? Why would it be slow during editing? It’s not like I’m running any complex recurring workflows.
My app is very large, however. Dozens of pages, about 100k entries in the DB from user-generated content, etc. But again, I don’t understand why would this affect the editor.
Yea, I found the same using safari. But it seems to start to tire out after a while. Perhaps the Mighty Browser will be the solution. Or get a new Mac MAX with 64GB ram… …or hope that the new bubble editor comes packed with performance enhancements.
I upgraded to an M1 specifically for the HTML5 on chip assistance the M1 gives. I’m getting great performance.
I use Chrome, mainly because Safari wont copy/paste elements in bubble.io between two windows of the browser.
I strongly suggest getting the arm64 M1 native version of Chrome. Chrome install won’t auto detect your chip and install this version. Go to Google Chrome - Download the Fast, Secure Browser from Google Click download and select the “Mac with Apple chip” version. Trashcan your currrent Chrome version (don’t worry, it’ll keep your chrome settings) and install the new one. It’s reported to be 80% faster than Intel native.
When a page is full of elements, it becomes unusable, slow and clunky. You have to transform your main groups into reusables and work the logic inside everyone of them independently.
States and url parameters are your friends to communicate between reusables).
I’d recommend starting with reusables from the start.
Same experience here, using Bubble with the Canvas template on a large page is very slow. Had out_of_memory errors multiple times on such pages. Granted I’m working on a lower-end PC with just 8GB RAM, I still think there’s a lot of scope for optimization using webassembly or something like that behind the scenes, like Figma.
For now, slicing large pages into multiple smaller reusables works well.
There seems to be no mention of the possibility that the Bubble server may be the root cause of these slowdowns.
Do any of you run realtime monitoring of your machine? What is happening with RAM, CPU, network traffic, disk, etc. while you are waiting for the Bubble app to update?
The editor is certainly more capable while working on individual reusables. However, it would be grand if the root cause of the performance issues were addressed. For new users hopping into our beloved Bubble, working with reusables isn’t intuitive.
Bubble is undoubtedly a resource hog. This is what needs to be addressed.
My sense is that they (Bubble) can’t really fix this. It has been an issue for years and performance hasn’t really been improved much.
I liken Bubble to a Swiss Army knife. Because it does so many things, it has to be bulky by nature. There is no way around it. There was a suggestion years ago to allow Bubblers to turn off features that they don’t need (such as loading so many fonts, and such). Or even a stripped down version of Bubble (Lean-Bubble) for people with larger apps…ie check the box to run Lean-Bubble. Nothing came of it.
A couple of years ago, there was a guy who got some traction with his financial app built with Bubble, but he ran into scaling limitations, and eventually left Bubble and rebuilt his app using a traditional stack. His advice was to do what he did. Prototype in Bubble, then if you get traction, raise money and rebuild it with traditional code.
To have raised $100M from investors, presumably, they must have convinced them that they can solve the performance/scaling issue at some point. Perhaps they can, but they haven’t let on how they intend to do this or provided a timeline. If they can, I suspect it might entail an entire rebuild…Bubble 2.0 or something…hence the need for $100M.
Agreed. Been on here for 4 1/2 years and there really hasn’t been any major performance improvements except for perhaps Cloudflare (but even that has seen marginal performance increase). Repeating groups are still slow to load more than 25 items, editor still chokes under many elements, page loads still in 4-6 seconds for even small pages… As Nigel had indicated as well, when you get to “big data” with thousands of rows and database calls, there’s a definite limit on what Bubble can do.
Performance is the only reason I’ve used Bubble for prototyping only and not expanded into making full blown web applications (which, for me would seem to be the holy grail of money making on Bubble vs. the casual Bubble hobbyist)
Very interesting, thanks for sharing this. A Bubble rebuild sounds like a very expensive, but perhaps necessary idea… lest some lean startup builds “bubble 2.0” themselves and eats Bubble’s lunch.
Guess we’ll be stuck with editing our apps on a per-reusable-element basis. This might hurt Bubble’s urgent growth requirements (Investors get hungry for returns after a $100M investment).
