We’ve been doing behind-the-scenes work on some fundamentals for performance, and now we’re moving to a phase where we want to engage more with specific examples. We’re looking for qualitative data around how our users are perceiving performance issues. We are also forming a “focus group” of apps with slower performance that we can try to target with improvements.
Your help and feedback here is greatly appreciated! Please participate here - it’s a relatively quick survey!
Hi @allenyang ,
Bubble app performance from a data perspective is incredibly slow in editor, capacity usage is inconsistently reported, scheduled api workflows are unreliable, and batched transactions without the ability to set the order of operations given the way capacity management is handled results in bad and/or missing data.
Hi @8it2 - thanks for that feedback, but could you also fill out the form in the first post please? Thanks!
@allenyang Done. But I sure hope this time is different, since I have and others have been raising these issues to the Bubble team for months.
Yes, we are committed to improving performance. You may not have seen an explicit “This project = Improving performance” feature launch from us, but trust us that work is happening behind the scenes with this in mind and to lay a strong foundation for performance improvements.
The gathering of this data and forming a focus group are hopefully two more signs that
@allenyang But that’s been said before! And, on the roadmap! And, customers have reported right here in the forums that performance has gotten worse! Granted, I’ve seen some improvements, but the real issue here in my opinion is what’s going on with capacity management, units, reliability, and transparency. Capacity should be clear as day and consistent, as opposed to some mystical unmanageable, unreliable thing.
Furthermore, when I contact support with these issues, it would be great to get some real support. Help me out, don’t ask for another bug ticket, don’t ask me to join a focus group. Look at my logs and ask some good questions! Or, just let me know there’s a problem and you don’t know how to address it right now, but I can take steps so that I don’t keep facing the issues, instead of buy more “units” or capacity.
Call me, maybe?
I’m really confused why you didn’t build your own survey with bubble… That’s pretty lame. I’m not going to fill it out on principle until it’s built on your system.
You should show off your system, not use someone else’s.
I’ll take the opposite perspective on this one. I’d rather the team spend time where it’s most impactful. Taking time to build a survey tool to get feedback feels like a poor use of time if there are already ready-made tools for the job.
One thing I have been noticing with Bubble apps compared to other frameworks is the time it takes to retrieve data. That’s keeping in mind that yes, the way you format your search does impact speeds a lot.
Say you have 1000+ users in your social app, you want to display a user’s data on their profile. You make a search for users with a unique username as a constraint. It takes seconds to find that user and load in their data. Compared to basically instantly with any other social network.
@allenyang I’m wondering in detail why that takes longer in Bubble?
I have my app on a personal plan and it’s blazing fast. I’m surprised about that since I’ve experienced Bubble to be very slow for the past 2/3 years that I’ve been a member.
That said, I will fill in the survey now
It’s hard to know what’s at play here without looking more closely at the app. If you’re so inclined, please submit this on the form!
Good to see the bubble team investigating on performance issues!
What does “Do you want to be considered for the performance focus group?” actually mean?
I did the survey, but I want to point out a major performance difference I just found. Check out text elements pulling from an API source versus inputs. When running multiple tests, I’m seeing load times cut in half when using an input instead of a text element.
@mirko.heimburger - As part of this push, we’re gathering a group of apps with noticeable performance issues, which we can ‘beta test’ our performance optimizations on. If your app is chosen for this focus group, we’ll be asking you to provide a test account in the live version of your app that has enough (fake) data associated with it to show the performance issues.
@josh10 That’s a bit odd - do you have a repro case? If so, would you mind filing a bug with that information?
Thanks to all those who have completed the survey thus far! Friendly bump on this thread in case anybody else wants to submit.
@allenyang I will submit my app in the future if the form is still up, if not I will send it to the Bubble team for analysis if possible.
I was thinking more in general but let us take a look at an example, the Bubble app itself. I am loading up Bubble | Build Powerful Full-Stack Apps Without Code, the page itself loads fairly quickly but loading in my applications takes like 5-6 times longer than first draw.
I have a fast fiber connection but still takes a good 2-3 seconds just for the data to load after the static page has been loaded.
As a suggestion, a spinning loading circle might make it feel faster, although it will not fix the underlying issue of course.
That is what I am wondering about, I am guessing maybe the connection has to go through multiple servers to work with the Bubble database?
Or maybe me being in Europe makes a difference if the server is in America?
Try using this site, or the network tab in Chrome. It will help you identify the slowdown.
Hi Stone, Can I ask why you are not using ‘Current User’s x’ instead of running a search for a unique user?
I think if the tabbing issue is solved properly and workflow data not saving on some occasions you will make a lot of users happy about the performance as it’s really driving me nuts.
Overall I would definitely consider myself as a promoter
They are not using servers in Europe. Switching to Cloudflare has reduced loading time but I’m waiting eagerly for European servers to be accessible in the options menu.