My App Finally Released | Image Edits

Coffee Lounge
1

Hope you are all well!
i just wrote this up quickly, so alot probs wont make sense, just me getting things off my mind. Just me rambling, not important to read.

So ive finally migrated my app away from bubble.io.
My app being: dash.imageedits.com ,if anyone wants to check it out.

However this isn't built with Wappler.
Im just simply making this post to say a huge thank you to people in this community , i was given great advice and general information regarding software, not just Wappler related information. I am forever grateful for that, with this help and direction, i am now quite confident in my skills as a self taught web dev :slight_smile:

And i personally love Wappler, even though i didn't use it in the end for my web app, it would be my first pick for any no code/low code tool. (however i wouldn't pick Wappler or any other no code tool over coding directly)

I had the web app developed to about 70% in wappler (with help of slashhash wappler dev team), before i started again in pure code.

My experience is coming from being a photographer, starting a web based business with bubble.io, attempting to migrate bubble.io to Wappler, then from bubble.io to full code (with a bunch of other tools tried in mean time but didnt like enough to really try).

Ill be giving my opinions on Wappler and whom should use it below. But first a quick over view of my new app.

  • Its is a specific niche marketplace design basically.
  • There are alot of roles involved.
  • --- has roles for clients and their team members, also roles for editors/service providers and their teams, and then the web admin and web team roles.
  • Next.js (pages router), on vercel, all basic api via lambda, with long living functions/api on separate express backend (and for websockets).
  • Stripe for checkout, but also using stripe credit balance option for client. Also monthly invoicing. Website can handle multiple currencies.
  • Postgresql for database, with multiple layers of backup, from wal replica, and then a bunch of dump backups, and replicas just at ready in different clouds.
  • Redis for queue management and also rate limiting.
  • wasabi for s3 storage, uploader that can handle nearly any amount of files.

Currently the code base is at 190k lines of code (ignore my migration code which was about 40k). lines of code dont mean anything, but just to give idea that i like to think this is fairly medium to large size project.

(i haven't used Wappler in a bit over a year, my code journey was long, so some points below may not be relevant any more)

why didnt i go with Wappler? 95% of my mistakes were spelling mistakes, which was just eating me alive. Idk if i was doing it wrong, but Wappler couldn't provide me recommendations needed for nesting/looping over arrays and such.

That was my main reason, other reason:

  • Typescript, related to spelling mistakes (i cant live without a typed language)
  • Components, i couldnt figure out reusability with wappler.
  • Talent, sadly Wappler doesn't have much hireable talent , and idk if that will ever change.
  • Specific knowledge base, you are basically locked to this community, and that just isnt enough for amount of issues you can run into with code. Wappler can do basically everything you could need, however its just different, and isnt popular enough to have vast knowledge on it.
  • ---- example with next.js if im struggling implement solid chat system, there is multiple popular open source repos i can look at to look at their implementation.
  • AI, ai to me is mostly useless when comes to business logic and large context understanding, however it is great at predicting what i am trying to write, and give me recommendations same as how i would write it. Ai is great for getting specific functions that do one thing without much need of context. And can also break down nearly anything in a simple way to understand. This to me, has basically removed any benefit i previously used to get from no code. I dont know if ai will be useful for Wappler as idk if there is enough information about Wappler for it.
  • my opinion, Wappler team is talented but doing way too much, i cant imagine what the code base is like, but id have to imagine it would be a night mare to try manage everything, and all possible options of Wappler, (php, .net, and i remmber someone recommending GO, and i just think, they are human not robots, its just too much to main and have all work together)(i could be wrong tho)
  • bootstrap, for a while i thought was bootstrap was okay, but after using tailwind, yeah no, tailwind is a huge thing for me now, i can make my website quite unique without much effort (currently its pretty standard but not hard to change)

who should use Wappler, imo, people migrating from a different career/starting a web based business (with very little web dev expirence), but not sure if web dev is really going to work out in long term.

Id use Wappler as a quick starting point, while teaching fundmentals of coding. From there you can determine if you enjoy it or your business is going to work, and decide if you want to jump face first into pure code, happy with Wappler.
if you choose to move away from Wappler, youll be able to migrate alot easier than any other no code platform.

Bubble and other no code plateforms will be easier to start, but they teach you concepts completely wrong on a fundamental level, my bubble.io app was calling db directly from client side and i think 99% of bubble.io apps are doing that. bubble didnt save me any time, it slowed me down overall over the years by an insane amount, where i reckon 40% of my effort was focusing on migrating away and unlearning concepts from bubble.io and retraining myself correctly, with Wappler, the fundamentals from the start atleast are basically the same as coding.

I cant really think of anyone else id recommend it too, over coding, however that applies to any no code tool, but that could now just be because of my understanding of code and forgetting how difficult it is when starting out. It took me years to understand.

So ima stop my rambling here but i have alot more to say, but generally stay away from bubble and any other similar tool that seems super easy, they make the easier stuff even easier but any medium task difficult, and diffcult tasks near impossible. Wappler doesn't have those down sides of bubble.io, just the main downside of less knowledge out there.

again, please for love of god, dont use bubble.io or any other no code tool, if you want a visual builder use Wappler. (there are others out there now based in react but i ahvent used them)

ill be making a big rant about bubble.io on forum later because this no code has caused me so many damn sleepless nights.

3 Likes
2

Me after reading the article

image

3

Nice job! Good read!

I came from the other end of the spectrum, from full-code to Wappler, and I'm happy I did so despite the bugs I came across. I haven't left full-code entirely, for the reasons you mentioned regarding types, autocompletion and such.

4

honestly haven't see/interacted with many people going from code to no code/low code tools.

Can i ask do you feel more productive/quicker with Wappler over code?

5

We did the same. Decades in development writing bespoke applications.

Definitely. We can deploy quite complex applications to mobile and desktop within weeks. The codebase is standardised. We have a huge repository of blocks/flows we have created to select from. Our Clients are happy! We don't just develop our own applications. And we select not to undertake Projects for other Wappler users (despite being offered Projects multiple times a month and for good money). We have enough on our plate.

It is swings and roundabouts. What works for you may not work for others and vice versa. If you achieve your objectives that is what counts. Productivity, longevity, stability, performance, security (not necessarily in that order). Yes some things do require a little extra attention here and there but for the most part it is all the same despite writing the codebase by hand or using a WYSIWYG editor. The end result is equal (good input = good output). Tools are just that, they are tools. Each of us use the tools we are comfortable with and that get the job done. Having the majority of 'tools' in one location can be incredibly beneficial. Being able to work on the front end and back end with a single 'tool' is advantageous in most regards...

Good luck with your endeavors @Hinky

4 Likes
6

Hi
Just curious in terms of how you manage your code blocks and flows across projects. Do you just have a single project as reference or are you using git.
Thanks, Paul

7

Congrats on the release @haydn! Very happy for you.

Claude to the rescue! I prompted it to kept it as close as possible to the original! :smiley:

Hey Wappler community!

Just wanted to share some thoughts on my journey from no-code to full-code. This is me rambling, so don't feel obligated to read it all.

The Big Move

So, I've finally migrated my app away from Bubble.io. For those curious, it's dash.imageedits.com. Feel free to check it out!

Now, here's the kicker - it's not built with Wappler. But stick with me, because this post is actually a massive thank you to this community. The advice and info I got here, not just about Wappler but software in general, was invaluable. Thanks to you folks, I'm now pretty confident in my skills as a self-taught web dev. :blush:

Why Wappler Still Rocks

I love Wappler, even though I didn't use it for my final product. If I had to pick a no-code/low-code tool, Wappler would be my first choice, hands down. (Though these days, I'd pick coding directly over any no-code tool.)

I actually got about 70% through developing my app in Wappler (with help from the awesome slashhash Wappler dev team) before I started over in pure code.

My Background

For context, I started as a photographer, then began a web-based business with Bubble.io. I tried migrating from Bubble to Wappler, then eventually went full-code. Tried a bunch of other tools in between, but none really stuck.

The New App

Quick overview of what I built:

  • It's basically a niche marketplace
  • Lots of roles: clients, their team members, editors/service providers and their teams, web admin, and web team roles
  • Tech stack: Next.js (pages router) on Vercel, Lambda for basic API, separate Express backend for long-living functions/API and websockets
  • Stripe for checkout and credit balance, with monthly invoicing and multi-currency support
  • PostgreSQL database with multiple backup layers
  • Redis for queue management and rate limiting
  • Wasabi for S3 storage, with a beefy uploader
  • About 190k lines of code (not counting migration code)

Why I Didn't Stick with Wappler

  1. Spelling mistakes were killing me. Maybe I was doing it wrong, but Wappler couldn't give me the recommendations I needed for nesting/looping over arrays.
  2. No TypeScript (related to the spelling issue - I can't live without a typed language now)
  3. Couldn't figure out component reusability
  4. Limited pool of hireable Wappler talent
  5. Smaller knowledge base compared to more popular frameworks
  6. AI tools work better with mainstream tech stacks

Who Should Use Wappler?

In my opinion, Wappler is great for:

  1. People switching careers to web dev
  2. Those starting a web-based business with little dev experience
  3. Anyone unsure if web dev is their long-term gig

I'd use Wappler as a quick starting point while learning coding fundamentals. It's way better than Bubble or other no-code platforms that teach you the wrong concepts from the get-go.

The Bubble.io Rant

Bubble and similar tools might seem easier at first, but they'll slow you down in the long run. Trust me, 40% of my effort was spent unlearning bad habits from Bubble and retraining myself correctly. Wappler, at least, teaches you the right fundamentals from the start.

Final Thoughts

I can't stress this enough: stay away from Bubble and similar tools that make easy stuff easier but medium and hard tasks nearly impossible. If you want a visual builder, use Wappler. It doesn't have the downsides of Bubble, just a smaller knowledge base.

I'll be posting a bigger rant about Bubble on the forum later. That no-code nightmare cost me way too many sleepless nights.

Thanks for reading my ramble, and huge thanks again to this community for all the help along the way!

4 Likes
8

We used Wappler to create an application to store all references to our blocks so we can search them quickly. Our ethos is to use a unified source for all our Projects, all based off our own styles we have created over the years. All of our applications share the same database schema (with only minor changes required in most circumstances). Our Flows are archived in the same way. As they are within the pages themselves (not specific to schema or Actions) they can be easily imported in to new Projects. 90% of the time we found ourselves doing the same thing repeatedly so we decided to just create a standard codebase from which we now select from. Essentially a library of things... Nothing special.

We now have dozens of Projects indexed and every style, element, layout (everything we have done in the past several years), is all stored in the database. We use tags to add in Client names or specific keywords for easy identification. No we do not use Git. No need for it.

3 Likes
9

haha even thou i use claude alot, idk why i didn't do this haha

10

can i ask what sort of projects you are building?

if its sharing a db schema, i assume your building same sort of web app, or more of general marketing site with cms. (unless your talking foundation schema like auth and then expanding on that per project basis)

that last part with everything stored in db, is quite unique and cant wrap my head around it but sounds very interesting

11

Overall yes, but it's not entirely straight-forward.

Front-end-wise yes, it's wonderful to build pages visually instead of just using a code editor. I'm not a front-end person.

Back-end-wise it's a bit of mixed feelings. Wappler lacks modern web development practices that are present in frameworks such as Laravel. Wappler's database manager is relatively fun to use though.

12

Sure. We create asset management solutions for Clients, their Agents, their Support teams, and third parties. Multiple roles, rules, and permissions. Each asset is assigned on an individual basis, can be shared across Groups/Teams, and within those assigned Groups can then be assigned to sub-Groups based on the asset Managers preferences and their respective Client/Team base. Schedules can be synchronized and shared, time management, designation of tasks, feedback and tracking, historical data can be assessed and comparisons made, and predictions calculated. Integrity and accountability are essential so every step of each process is monitored. That is the short and curly of our business model.

We have learned over time that the requirements are quite standard. It is User, access, permission, task, progress, support, accounting and finance, file/media storage, and instruction. We commence with a shared schema and if anything outside of this remit is required we simply add it in based upon the specifics of the request, more often than not we already have a generic solution available. We separate our databases based upon specific sectors so we have multiple schemas available to select from (we pick what we need from those). These have been created over many years and account for most scenarios.

It is far more simple than it sounds. Essentially when we create a new style for an element we store it in the database, if we create a new table design with filters and paging we store it, new layouts are stored, containers, rows, columns, headings, paragraphs, buttons, calendars, etc. All treated the same way. Generic Flows which we make use of are stored (everything adopts our in-house naming convention so it is second nature to us upon recycling any element or functionality that we know and understand it intimately, we have used it many times before). Likewise for things like Data Stores, Cookies, Sessions, Server Actions, the whole nine yards, all follow this same principle. This gives us a repository we can then filter against and pull what we need, we then track its usage so we have an overall picture of when and where it has been implemented. It is all a little bit OCD if I'm honest. But on the flip-side if we need to re-purpose something in some way we already have a template to work from, and if it is new it is itself added to this repository. We also document pretty much everything we do, just for our own sanity if nothing else.

1 Like
13

Hopefully not Claude Pro. During these last weeks the quality has gone down the drain. I had to cancel due to how horrible info it was providing.

At least API is fine. Cursor is a beast with Sonnet (API)

14

Offtopic
@JonL you owе me 18 euros for Claude Pro subscription. Before you advertised this AI-model so much so I started to think it's much better working with text comparing to ChatGPT... But it's not.

It's a joke.

2 Likes
15

Hah, good luck collecting that debt man!

The underlying model itself is great. No joke, it's better than what OpenAI is peddling right now. Anthropic's Opus 3 is straight up magnificent for anything involving text, reasoning, logic, all that jazz. Sonnet 3.5 - that thing is an absolute beast when it comes to coding. It'll code circles around you.

But here's the rub - the "Pro" version, the one they use for their slick web UI and fancy mobile apps? They're constantly messing with that, changing things around and sometimes straight up gutting functionality. Why? Oh, the usual bullshit - "capacity constraints" or "performance tuning" or "alignment adjustments" or "content filtering". And don't forget the classic "we're just running a lil' A/B test, nothing to worry your pretty little head about."

But hey, it ain't just Anthropic pulling this crap. OpenAI does the same damn thing to their ChatGPT Plus model, the one powering their web UI and app. They'll screw the pooch on the regular too.

So word to the wise - don't go getting too attached to any particular web or mobile version of these services. Keep your options open, be ready to rage-quit and take your business elsewhere on the drop of a dime. Anthropic acting up? Peace out and hop on over to OpenAI. OpenAI starts trippin'? Bye Felicia, Anthropic here I come!

But if you want some goddamn stability and consistency in your life (and who doesn't), your best bet is to go straight to the API. See, with the API, you can freeze that shit to a specific model and release date. Find a version that works for you and lock it down tighter than a drum. Then you're golden - no more surprise "improvements" jackin' up your flow.

So yeah, good luck with that debt collection my guy! And remember, when it comes to these AI services - keep your head on a swivel and always have a backup plan. Stay sharp out there!

Based on my notes and written by my intern who drinks too many energy drinks: Opus 3

4 Likes