As I continue to question my decision to continue building in Bubble, I am continuing to come closer and closer to becoming so frustrated to the point that I actually just leave Bubble…after having to yell multiple times due to the frustrations that exist currently in the past one hour, I thought maybe others are not experiencing the same levels of frustration as I am and so maybe a ‘me thing’ and not a ‘bubble thing’…
Question: How many times a day do you think “I hate Bubble now”?
0
1
2
3
everytime I try to do anything in the editor
only when using new features not previously QA tested by Bubble
I work full time on bubble apps and I am beyond frustrated. Without a doubt every hour I work in the bubble editor I am yelling at my screen and exclaiming WTF to another bubble bug I’ve found.
Or worse - something that used to be working fine (header menu) has now been “redesigned better” and the new design is worse than the original - and I’m stuck with a seriously downgraded core function of the editor.
I’ve given up reporting bugs because:
there are so many
they never get resolved
the bug report “smart” chatbot just makes me angrier
I can’t be the only dev who finds every bug in bubble editor… surely dozens of others have already filed the same bug report
A few weeks ago I seriously looked at moving to weweb but after a half day of using it I decided it was not quite ready. It’s definitely on it’s way to being a solid replacement to bubble though - hopefully by years end. Seriously can’t wait!
This is like being okay with the plane going down because ‘everybody on here is going down too’ but still, the ‘going down’ part is not okay. I’m glad I’m not alone in my frustration, but it is not okay that we are having to deal with these things.
I’d love Bubble to just take a 3 month hiatus from ‘improving’ the platform with redesigns or new features, and take the time to fix all of the outstanding issues. Would likely go a long way with user retention.
Honestly its quite reassuring that someone with as much experience and time in the game as yourself is voicing similar sentiments to myself. The editor is a mess, I’m having to bounce between different browsers because there are different bugs in each one (copy with workflows seems to be 50/50 as to whether it will paste with workflows now). Need multiple tabs open in case I accidentally click “backend workflows” and completey lose where I was in the editor (baffling how this has been completely ignored despite many many complaints about it). Which would be fine if the editor didn’t constantly crash/run out of memory.
Its extremely hard to get in the zone and concentrate on actually developing when most of your time is spent trying to maintain an operational environment in the first place, or your focus is persistently intertupted by these things.
And now with the looming new workflow editor being forced upon us its set to get even worse. I’d avoided it up until now as I didn’t see any benefit (if workflows branched like Flutterflow does under If/else conditions then we’d be getting somewhere), but have been using it on one project for the last few days since its going to be inevitable. Needless to say, browser crashes and bugs aside, my productivity is massively reduced. Switched into another project where this isn’t enabled and it was a relief to know what the hell was going on again.
Honestly, I’m looking to wrap up one of my projects in the next few weeks which will free up some time and give me a chance to commit some serious time to where to egress to from here, I’ve kind of had enough.
I think it helps to be aware of a cognitive bias called “Prevalence-Induced Concept Change.” Basically, as Bubble solves more problems, the remaining ones, no matter how trivial, become more apparent. Like, I definitely take for granted some of the frustrations of code that I don’t have to deal with because Bubble has removed them. Just a bit of nuance to my previous post.
The biggest one, and this affects 100% of people, is how when you combine two groups into a single group, 90% of the time it just breaks all the groups inside. I don’t experience a lot of other bugs that people talk about, probably because I don’t use plugins or multiple tabs.
There’s also a brand new bug where if you click Deploy the first time, nothing happens. And you need to click it again.
This is why its so frustrating to be honest, I doubt anyone would deny that Bubble is a marvel in and of itself. I don’t want to move away from it by choice, but these issues highlighted by myself and others…people have limits.
And yes, I’ve also noticed a need to double click deploy. And when switching branches it has a habit of throwing up nonsense errors, usually relating to established API endpoints that have nothing wrong with them, though the error checker begs to differ. Or in the case of last night, it added a bunch of conditionals onto a completely unrelated repeating group from another repeating group elsewhere in the app. Once these were deleted (after carefully checking I wasn’t about to completely bork my app) it insisted that two errors remained in the text values despite the errors not relating to anything relevant. “Resolved” by completely closing down the browser and reloading a few times. This is not an isolated incident.
Yeah branches are very spooky to me, had one of those errors you’re talking about. Also very cumbersome to switch between live and main. Takes so many mouse movements and clicks.
I really like the direction WeWeb is going personally, after playing with it for quite a while (~6 months) in an attempt to replicate a page out of my main app. Like you though, I just couldn’t get the functionality I needed in it. That was quite some time back though, probably due for a revisit…
I don’t often encounter bugs in my use of the editor, but I am probably a “bad example” - as I only use bubble for the UI now, and have shifted the entire backend to Xano. I just don’t have time to redevelop an interface from ground up in another platform (and have yet to find one with equivalent functionality, but again - I haven’t had a lot of time to look either!).
My main bugs are generally with changing data types on groups or in the API connector, which sometimes results in me having to rewrite numerous conditionals - often (and randomly) resulting in me having to take a screenshot to “rewrite” the conditional, since fixing the issue wipes out the rest of the conditional.
If you asked me ~2 years ago, I would have been in the 5+ category of your poll - I was venting most the day! But I have since found workarounds for most the issues I was encountering, and by writing my own plugin and using a non-bubble backend (both of which I was forced to do due to the WU changes so my company could even release a product and fund it!) - I brought that down to <1 per day.
I tend to think of the platform these days as being so incredibly filled with legacy and messy code (try reading the source when debugging - I’ve done it!) - it’s an absolute mind-bending thing. That’s the downside of lots of functionality - a ton of legacy that is difficult to change in a non-breaking way.
I doubt bubble is sitting on their hands - I just think they underestimated the amount of effort it would take to modernise the core, whilst still satisfying their VC investors. Credit where it’s due - the editor at least is beginning to look a little more modern - but that workflow editor… shudders
I fully believe Bubble needs to keep building future-facing features. But It is also true that without making our core experience fundamentally robust and great we will leak users faster than we can get them.
I’ll bring these points (and others that have been made in other topics) to the team.
Almost always on list operations or reusable operations. After 2nd creative solution pushing for final 3rd most creative solution I spit out ‘‘I hate Bubble’’. Latest one is now I need to look up some css in order to have some horizontal
scrolling function on a simple group
On the rewrite, my workaround is to either work from my live version in a split screen (lots of bits you’ll be able to copy out) or create a branch before changing the API call.
When getting stuck into really large chunks of logic I now quite often break them up using arbitrary text. It’s a bit of a nonsense but definitely nice to be able to edit halfway up the block and/or copy and paste otherwise laborious bits of code, since you can paste full expressions into the arbitrary text. This of course comes with it’s limitations since output conversion is pretty restricted.
I think a really great solve for a lot of these issues would be something akin to abritrary text, but more along the lines of an “arbitrary block” that you could set an output type on, allowing you to chunk code instead of writing small essays. If we had a similar level of output control on format as text and grouping operations it would provide a pretty phenomenal boost to bubble’s power level.
One of the things I do fairly often - but not really practical if I’m dealing with a data type in any form. In my case, due to the lack of native loops and if/then functionality - I end up with VERY messy and convoluted conditionals. Trying to get the brackets into the right position is often problematic too, so I completely hear you on that.
On branches… oh how I wish the personal plan had this!!! I have never encountered a platform that gates branch management to only higher tiers, especially when the functionality of bubble tends to lend itself to “feature creep” being inadvertently introduced along with numerous bugs (both caused by the editor or the developer), which are then difficult to remove. I’m perpetually disappointed at this decision on the branches.
The WUs was the start of the decline. This fallacy that vibe coding only gets you 80% of the way there and then streaming in Q1 2025 with no updates, a forever beta for mobile apps and lack of focus. I’m out.
Every AI tool is the worst version it will ever be right now, they are only going to get better, and faster, with more funding and a more motivated team. I guess I owe Bubble some respect for forcing me to separate my backend from the frontend with the WU saga so atleast my apps are more flexible to new homes as a result. I just feel we have outgrown bubble at this point which doesn’t point to the future they are trying to sell of going from MVP to scale with Bubble.
Ah yes, we have one app on a max plan so I forget those are gated. For what it’s worth… you’re not missing much haha. I am terrified of versioning in bubble, as I’ve seen it just wipe changes on multiple occasions.
I once had a datatype duplicated, which broke 90% of my workflows relating to it because they didn’t know which one to point to. Bubble support informed me that I had clearly just created another data type, as they could see that the new one had a different id.
The fun part… the new “id” was just the name of my first field in the datatype. Unresolved, support did their usual amble, blame, ignore and moved on, so I had to spend a week fixing my app with no help. Versioning is pretty broken, so I try to avoid making too many branches for fear of it destroying my app.
My biggest fear with bubble is that one day they suddenly do what some other no-code platforms did: setting upper limits on the number of APIs you can configure. Bubble can’t (and shouldn’t!) be charging us for API requests to a 3rd party service (e.g. Xano) for requests from a client browser, but if they did either of these things - I would be shutting down my app immediately and rebuilding elsewhere. Both these scenarios cost bubble nothing - so any attempt to charge WU or require a higher tier plan for their use is just pure greed if they were to do so.
So if they do improve the API connector - I just hope it doesn’t go backwards from where it is. Honestly - the only things I want to see from an “improved” API connector are:
Request templates (e.g. something I can copy from for each API).
Ability to move an API to another group.
The ability to rename an API that carries through to the editor!!! (this is my #1 bugbear with the API connector - the number of times I have needed to recreate an API from scratch since renaming it doesn’t work - even when duplicating).
