I'm pretty sure a good portion of Canva is vibe-coded too. You could argue Canva has tests, QA, etc. But somehow I feel we just tend to associate quality to commercial products.
Interestingly, Claude has generated a LICENSE file which demands commits & changes are GPG signed. But the repo's license file has no GPG metadata (nor does the rest of the app), so it's impossible for anyone to commit to, extend, or meaningfully change this code without violating the repo's license.
In short, nobody can commit to, change, or meaningfully adjust this repository while adhering to its license.
Sorry to break it to you, its from our internal project (only the editor part) which had way less feature. And you should watch commit history of open source projects more often every single one usages Claude/codex, people no longer code by hand its over.
Great but imho Canva alternatives that stand a chance should be made API and MCP-first, not just an afterthought.
That's the thing I miss with Canva right now. Yes there are API & MCP options (4 different ways even I think, it's pretty messy), but it's just not sufficient. Ideally you can fully vibe code a poster or a video from the context of your repo, which has all the context and assets, then edit the result online in their point-and-click editor.
But if it was just made with one prompt if you considered that feature important why would you use this as a base and not just start your own project for the clout.
If something was easy for someone to spin up there's little reason to build on their work rather than just make your own thing you fully control.
I think the poster is saying that adding an MCP feature to it is trivial, which doesn't say anything about the complexity in building the tool in the first instance.
Just making an observation that people vibe coding these projects don’t want to contribute to shared efforts, they want to be the guy who did the thing.
> Ideally you can fully vibe code a poster or a video from the context of your repo
Ideally for who?
Maybe I'm reading too much into it, but this seems to be a tool for people who at least have a slight intention of producing something coherent, consistent and generally nice looking. What you're suggesting is to turn this "tool for a human" into a slop producer, so it's able to produce as much content as fast as possible, with as little human input as possible. Couldn't you fork this and implement this "slop factory feature" yourself instead of asking this tool to be re-built from the ground up to cater to this?
I do find it funny how many people on HN consider a product needing a bot layer a necessity as if the producer of the product derives some kind of benefit from allowing anonymous bot traffic to their site or their app; as if bots have some added value as opposed to be nerd parasites on the internet's store of value.
I do find it funny how many people on HN think products are needed.
They're not, that's what AI is killing. The true successor of Canva isn't some vibed clone, it's just a bag of tools, without branding, and this will have to come from outside because obviously the company behind Canva isn't willing to reduce Canva to what it truly is, once you cut out the branding, which is... just a bag of tools.
I don't understand why someome capable of self hosting would use a "Canva alternative" instead of something normal like Inkscape/Figma/...
Canva is so bad. It's impossible to align stuff properly. Text rendering is awful. Official builtin templates have horrible kerning. It's genuinely one of the worst graphic editors I've used, and I'm not even remotely a professional designer. I don't want a "Canva alternative", I want a normal editor.
AI doesn't change this. Would I rather use some hobby project or the real tool? Depends on my budget and how much I care about functionality/quality – sometimes the hobby project, often the real tool. How long each took to develop is not a factor in my purchasing decision.
I'm not expecting the people who take the vibecode shortcut to have the discipline and commitment and talent to steward their projects in the manner of the successful ones I've come to trust in the past.
Yeah, I would and do, but many people like shiny and just jump around for some reason. And the ones that have a reason and make products popular overnight, have as reason that they need likes and upvotes as they are influencers.
People who have jobs, lives, other stuff to do in other words, keep sending me 'the latest new bla' many times per day, every day. All products that will be gone tomorrow.
I mean, AI made my own canva because my wife uses it and hates their support, bizarre amounts of long standing critical showstopper bugs and attitude to long standing members. So I hope they do get vibecoded to death, however, I would not release such a project until after a year of testing and fixing, which is where our version is about. Canva in the basis, like Miro etc is a one shot with Opus 5.5: it works fine-ish. But you will need months of grind to get it so people will want to use it.
Edit: in the vibecode age, 2000 commits would be the minimum to consider a project maintained. And then I do have a script to analyse the 'effort' of an average commit because some people fake commits with crap to inflate the number.
A lot of what makes Canva stand out comes from its IP coupled to their ecosystem/product (templates, assets and stock photography), its otherwise a pretty simple tool with basic functionality (but with good UX) compared with more professional software like photoshop or illustrator, or Figma. I'm sure you can built an open source equivalent in terms of software, but I don't think it's really possible for open source to produce a true replacement or alternative.
THAT ASIDE, it might be worth focusing on UX. This looks vibe coded, maybe it isn't.
- The UI is broken in a few places (the tools text in the shapes menu)
- Consider prioritising more important tools based on available space in the ui, the header is a bit cluttered.
- Some of the tool menus are awkward to use, like in Canva those are in a side bar and when they open the have panels that don't disappear when their parent looses focus. Probably worth making sure you're not using their exact icons as well as these look pretty similar.
- Your add page button is creating an awkward scroll which you could probably limit by exposing add page in a way that didn't do that. Maybe remove the add page button below the canvas, as you already have an add page button in the bottom left, you should have more vertical space for the design as a result.
- You've got several different export sharing options in the header side by side. Might be worth collapsing that into a menu to minimise clutter. Both tools like canva and more professional tools layout their UIs like this. By all means do don't it just because they are, but there is some rational here why they would.
- There are two places in the header that talk about the save state on the left and the right (one saying saved, the other saying saved in the browser). Consider having consolidating that.
- do you really want to place the copy "Sign up to keep it forever" next to the sign up button? do you have any reason this is likely to stand out from the rest of the UI features there? I reckon if not get rid of it. Or if it's that important get rid of anything competiting for its attention which right now is a lot of things.
- you have a ellipses for a menu with a ton of options, in most software this is aligned to the left. If you've removed other things frmo the the head it may make sense to break that up in submenus so those pieces of functionality is more discoverable.
- Your commment functionality requires sign up to use, keeping it there may help lead to signs ups, but if you're designing this as an open source tool, maybe you're less interested in growth hacking, it might make sense to remove it from the header as well when the user isn't signed in (at the same time keeping it there signals to people weighing up your product that that feature exists.
- you have 2 layers of headers, is it worth utilising horitonzal space for some of your buttons and put them in the right hand side?
Hopefully you can feed the above back into your agent on a loop lol.
The license choice is a little bit weird, but it is OSI-approved anyway: https://opensource.org/license/aal
I was horrified looking at the commit history. The whole project was authored by Claude and it spawned out of a single commit. Really, check it out
https://github.com/shaongitbd/doory/commit/eac7e3b
I'm pretty sure a good portion of Canva is vibe-coded too. You could argue Canva has tests, QA, etc. But somehow I feel we just tend to associate quality to commercial products.
:) this is the truth, every single developer I knew moved away from coding by hand every single one.
Well you don‘t know me, but you can still feel free to count me against your stats
Well seems like you don't actually know any good developer
No-one is writing code anymore tbh
old workflow is dead
tbf we've been trying to kill it for a long time, look at every dev product ever
The Attribution Assurance License is a particularly bold choice for a repo of code the author presumably didn't write
Interestingly, Claude has generated a LICENSE file which demands commits & changes are GPG signed. But the repo's license file has no GPG metadata (nor does the rest of the app), so it's impossible for anyone to commit to, extend, or meaningfully change this code without violating the repo's license.
In short, nobody can commit to, change, or meaningfully adjust this repository while adhering to its license.
Sorry to break it to you, its from our internal project (only the editor part) which had way less feature. And you should watch commit history of open source projects more often every single one usages Claude/codex, people no longer code by hand its over.
Vibe coded or not , proper engineering , testing , fixing , iteration is still needed . Your don't have one , full of crashes and bugs.
Actually, the rate of AI usage among the free and open source projects I most respect is quite low.
do you mean people no longer code without a LLM proxy? i think they are still doing it "by hand".
That's not true. Look at Emacs, look at GCC
Pretty standard at this point, especially for non-critical open source projects like this...
Great but imho Canva alternatives that stand a chance should be made API and MCP-first, not just an afterthought.
That's the thing I miss with Canva right now. Yes there are API & MCP options (4 different ways even I think, it's pretty messy), but it's just not sufficient. Ideally you can fully vibe code a poster or a video from the context of your repo, which has all the context and assets, then edit the result online in their point-and-click editor.
100% agree with this. It should be an mcp or a chatgpt plugin
It's open source.
But if it was just made with one prompt if you considered that feature important why would you use this as a base and not just start your own project for the clout.
If something was easy for someone to spin up there's little reason to build on their work rather than just make your own thing you fully control.
I think the poster is saying that adding an MCP feature to it is trivial, which doesn't say anything about the complexity in building the tool in the first instance.
Just making an observation that people vibe coding these projects don’t want to contribute to shared efforts, they want to be the guy who did the thing.
Awesome. Can you get MCP working for me?
Your Claude Opus 5.5 can.
^ This
> Ideally you can fully vibe code a poster or a video from the context of your repo
Ideally for who?
Maybe I'm reading too much into it, but this seems to be a tool for people who at least have a slight intention of producing something coherent, consistent and generally nice looking. What you're suggesting is to turn this "tool for a human" into a slop producer, so it's able to produce as much content as fast as possible, with as little human input as possible. Couldn't you fork this and implement this "slop factory feature" yourself instead of asking this tool to be re-built from the ground up to cater to this?
I do find it funny how many people on HN consider a product needing a bot layer a necessity as if the producer of the product derives some kind of benefit from allowing anonymous bot traffic to their site or their app; as if bots have some added value as opposed to be nerd parasites on the internet's store of value.
I do find it funny how many people on HN think products are needed.
They're not, that's what AI is killing. The true successor of Canva isn't some vibed clone, it's just a bag of tools, without branding, and this will have to come from outside because obviously the company behind Canva isn't willing to reduce Canva to what it truly is, once you cut out the branding, which is... just a bag of tools.
"Why people pick Doory"
Is this evidence-based? There's one github issue and six commits.
No, that’s just standard AI generated nonsense marketing prose
Canva has always felt like The Print Shop to me: https://archive.org/details/The_Print_Shop_1984_Broderbund
This has been posted here many times https://theprintshop.club (emulation that can even print)
Good idea, but the real value is in having high-quality built-in templates. Right now, they look just like what I could get by asking Claude.
"yeah, that's what we did too."
I don't understand why someome capable of self hosting would use a "Canva alternative" instead of something normal like Inkscape/Figma/...
Canva is so bad. It's impossible to align stuff properly. Text rendering is awful. Official builtin templates have horrible kerning. It's genuinely one of the worst graphic editors I've used, and I'm not even remotely a professional designer. I don't want a "Canva alternative", I want a normal editor.
So with this future of vibe coded tools , how to choose which one to use?
AI doesn't change this. Would I rather use some hobby project or the real tool? Depends on my budget and how much I care about functionality/quality – sometimes the hobby project, often the real tool. How long each took to develop is not a factor in my purchasing decision.
The one that is actually supported and not abandoned.
I'm not expecting the people who take the vibecode shortcut to have the discipline and commitment and talent to steward their projects in the manner of the successful ones I've come to trust in the past.
Thats the point. They wont care about the craft , design decision ,best engineering princinples.
You can fix bugs you care about piloting a $20 sub in abandoned projects relatively easily.
Yeah, I would and do, but many people like shiny and just jump around for some reason. And the ones that have a reason and make products popular overnight, have as reason that they need likes and upvotes as they are influencers.
People who have jobs, lives, other stuff to do in other words, keep sending me 'the latest new bla' many times per day, every day. All products that will be gone tomorrow.
I mean, AI made my own canva because my wife uses it and hates their support, bizarre amounts of long standing critical showstopper bugs and attitude to long standing members. So I hope they do get vibecoded to death, however, I would not release such a project until after a year of testing and fixing, which is where our version is about. Canva in the basis, like Miro etc is a one shot with Opus 5.5: it works fine-ish. But you will need months of grind to get it so people will want to use it.
Edit: in the vibecode age, 2000 commits would be the minimum to consider a project maintained. And then I do have a script to analyse the 'effort' of an average commit because some people fake commits with crap to inflate the number.
Which one?
There are many too many new projects (OSS and closed) that have AI in them and existing projects (Linux, etc.) that are getting tainted by AI.
I think this is a lost cause.
Vibe coded or not , Software quality comes with proper maintenance and engineering principles.. i dont believe full vibers will do those.
Your own
Yeah ... looks like that is the answer these days.
Pick the one you like and vibe fix bugs you run into.
This also works with paid closed source software today.
Ask claude
Brought to you by Claude.
A lot of what makes Canva stand out comes from its IP coupled to their ecosystem/product (templates, assets and stock photography), its otherwise a pretty simple tool with basic functionality (but with good UX) compared with more professional software like photoshop or illustrator, or Figma. I'm sure you can built an open source equivalent in terms of software, but I don't think it's really possible for open source to produce a true replacement or alternative.
THAT ASIDE, it might be worth focusing on UX. This looks vibe coded, maybe it isn't.
- The UI is broken in a few places (the tools text in the shapes menu)
- Consider prioritising more important tools based on available space in the ui, the header is a bit cluttered.
- Some of the tool menus are awkward to use, like in Canva those are in a side bar and when they open the have panels that don't disappear when their parent looses focus. Probably worth making sure you're not using their exact icons as well as these look pretty similar.
- Your add page button is creating an awkward scroll which you could probably limit by exposing add page in a way that didn't do that. Maybe remove the add page button below the canvas, as you already have an add page button in the bottom left, you should have more vertical space for the design as a result.
- You've got several different export sharing options in the header side by side. Might be worth collapsing that into a menu to minimise clutter. Both tools like canva and more professional tools layout their UIs like this. By all means do don't it just because they are, but there is some rational here why they would.
- There are two places in the header that talk about the save state on the left and the right (one saying saved, the other saying saved in the browser). Consider having consolidating that.
- do you really want to place the copy "Sign up to keep it forever" next to the sign up button? do you have any reason this is likely to stand out from the rest of the UI features there? I reckon if not get rid of it. Or if it's that important get rid of anything competiting for its attention which right now is a lot of things.
- you have a ellipses for a menu with a ton of options, in most software this is aligned to the left. If you've removed other things frmo the the head it may make sense to break that up in submenus so those pieces of functionality is more discoverable.
- Your commment functionality requires sign up to use, keeping it there may help lead to signs ups, but if you're designing this as an open source tool, maybe you're less interested in growth hacking, it might make sense to remove it from the header as well when the user isn't signed in (at the same time keeping it there signals to people weighing up your product that that feature exists.
- you have 2 layers of headers, is it worth utilising horitonzal space for some of your buttons and put them in the right hand side?
Hopefully you can feed the above back into your agent on a loop lol.
This is IMO much more precise, kindly and practical feedback than this product or this announcement really warrants.
Think of the apps you could have written for yourself in the time you spent writing this ;-)
fully vibed low effort clone atually. it is now crashing while testing.
> This looks vibe coded, maybe it isn't
Claude is the author on the commits.