Haha, I wasn't ready to post this to HN. It's not anywhere close to ready. But thank you aghuang. (I want to post this again when we clear Beta phase.)
These are still super early alpha. All MIT/Apache licensed. No telemetry.
I'm going to try to drive these to 100% parity but it's going to take a minute. And I'll need lots of help. (Anyone want to join me? I'll pay stipends and bug bounties.)
I'm going to do Microsoft Office, AutoCAD, and Solidworks too. Probably a lot of stuff to collapse.
It's not just SaaSocalypse. Everything can be open source multiplatform Rust now.
Party like it's 1999 and the Internet / tech scene belongs to us again. We can literally deenshittify the world without breaking a sweat.
Just wanted to say at least for the non-creative apps I am already doing that at https://azul.rs
But its a lot harder if you want to maintain at least some level of quality. For example, your PDF export exports texts as paths and correct PDF encoding and font subsetting took months of trial and error (I maintain the printpdf crate).
So you have to actually use your apps, test extensively, etc. - don‘t underestimate the effort required for QA and testing.
I'd love to see something like this for Solidworks, but think you're in for a boatload of trouble there. OpenCASCADE is less capable than Parasolid, and a true Rust-based parametric geometry kernel capable of boolean joins, fillets, etc. needed for anything resembling a modern feature set is wickedly difficult (that's one of the reasons why everyone licenses existing ones and don't try to roll their own).
Don't let that discourage you, though! A fast, local, fully-featured 3D parametric CAD program that takes advantage of modern hardware parallelism that can be extended/scripted is just basically one of the holy grails of software and even exploratory efforts might uncover some great paths forward to making that a reality!
Indeed. If you are going to burn tokens to clone something, then assembling a team of people who could help you burn those tokens in service of a bRep kernel (let alone a Parasolid kernel) would be an interesting undertaking.
With photoshop, illustrator, etc., there’s a fair amount of code out there. No LLM has however ever seen the source code —- or I suspect even much detailed discussion —- of a bRep kernel better than Opencascade, but with the right people to decide what maths to work on, maybe there is progress to be made. This knowledge is held by few people, many of whom will now have retired!
I've played around with most of the solutions in this space (OnShape, Fusion360, FreeCAD, Solvespace), but honestly at this point would probably just end up paying for the Solidworks hobbyist license; it's like $4/month and I have a Windows desktop available. That said, there's been some consensus built over the past few years about what the next-gen CAD software should probably look like, ranging from modern scripting capabilities to version control and intermediate shape representations.
A big of the attraction of re-building this kind of foundational software isn't the base capability (commercial 3D CAD has barely even seen incremental improvements for more than a decade, in large part because it's just pretty well solved in the existing programs), but that a modern kernel would be a major step forward for unblocking implementation of some of those other capabilities.
There is Truck - "Rust CAD Kernel" (https://github.com/ricosjp/truck) being developed by a Japanese CAE company, but not sure how full featured it is currently.
It started about 6 months back, kernel build from scratch in go.
I changed directions now and decided to rewrite in rust, take the good parts of occt and the good parts of my own kernel and build a new one child of both.
So far it did not get traction, I guess I should had named it SolidCraft :D
The videos of the features are very well done, and it looks like there is a unified version, which is not what any of the other products have. I look forward to the beta and beyond!
Despite whatever else happens with AI, I am excited about the near future where we have an open source version of basically every app. I hope this is the beginning of the end of the subscription model. Keep it up!
I'm doing a run at rebuilding a few of Apple's bundled apps (iPhoto, iTunes, Mail) as they were back around 2015, the replacements are much worse in the ways I care about so at this point rebuilding as GTK + Rust seems like a worthwhile experiment. I only started last week though so I don't have a sense of how long it will take yet.
do you mean to equate stealing software crafted by people who made sure it worked by "cracking" it to "pirating" decades of human effort by paying anthropic thousands of dollars to generate broken versions of it?
i am not sure i follow. it certainly isn't technically the same.
I'd love to see a better Office software and AutoCad software! Especially ones that can be manipulated more programatically by agents or at the CLI, like your PrintCraft and similar projects. I think that would be awesome to interact with dwg and xlsx files using a CLI!
A bit late for the comments, but please try to assure the quality of the software before going for new projects. These looks very promising! But they are not usable if they don't really work. There should be very low tolerance for bugs.
re: ms office, you should take a look at visigrid for the excel bits.
It's also rust based and very polished with a cli tool for automation. Can't speak to feature parity as I'm not an accountant or someone who traffics in spreadsheets but it suits my needs to the point I don't reach for a company subscription.
That's the bar for an AI generated app that I'd like to see your Adobe CC/CS apps hit.
I am holding onto my copy of Affinity 2 and never letting go... mainly because Canva never bothered porting the third version to iPad and the whole point of me buying Affinity was that I could open the same project on iPad and Windows.
Are we looking at the same stock? I am confused where this data is coming from. What I see is stock slightly higher than it was week ago, fell compared to day ago, and the change is 1-2%, way less than you 50% fall claim.
As I understood the message, the GP post stated their predicted price for Adobe, and the target price for the options they purchased, not any change in price that has already occurred.
Adobe's moat is so small, so it's a matter of time until someone frustrated with their pricing would rather just rebuild the entire Creative Suite. I have also rebuilt the features from Photoshop I want using Rust recently (primarily because the company I work for started blocking Photopea, but I digress...) and it's been a great to work with. I can imagine the same would hold true for Lightroom, Illustrator, and Indesign.
Yep. You've jumped right to the end of this story.
I think the same thing about the endless proliferation of Web-dev frameworks. Why bother? Just have AI write it in native JS or WebAssembly and be done with it.
Yep. I was on a long plane trip and settled in to do a bunch of work in InDesign, only to have the POS try to re-validate my subscription at that moment, without even a one-day workaround to no Internet connection.
But then I remembered that the flight had free Internet. So I log in... only to have these morons try to TEXT my phone a 2FA code. On the plane.
After I got back I checked with our IT department and found that our Adobe licenses are paid up for TWO YEARS.
Same. I hate paying Creative Suite, so during few hours I vibecoded Photoshop clone for myself in rust, with only the features I am using. Have been using it for the last 9 months, no problem. Super fast start, full PSD support, layers, texts etc.
”Etc” is doing a lot of heavy lifting here. I’m sure it works great for you, but something simple as “layers” is for me smart objects with plugin support and Camera Raw integration.
Photoshop has a lot of bad things going for it, but it’s still the only software that does what it does, and I’ve tried all the competitors that have popped up over the years.
If you have the time those are just 1-5 hour hurdles to get over, most of that is waiting for the agent to obtain the background info and for you to test the resulting outputs.
It's not that different from having Claude go to a website and sniff all the API calls it makes as you're interacting with it and then using that info to create a website that you prefer interacting with and all the layers and interactive components that you like.
Adobes moat is being the industry standard, and the wealth of documentation and tutorials and features like integrating with industry standards (printers, cameras, exotic formats). I can't imagine every company going through the hassle to have people relearn their complete stack when hiring new people because the have the n-th possible "adobe replacement" software.
They might if they pay adobe $1-5 million dollars a year. I'd its open source you hire the developer onto your company and have him work on the features you need in the app in house.
A company paying $1-5m for software that is reliable, provides support and created by experts in the area will consider that a deal. They go through standard training -- only in case the user hasn't used the software before -- and they're ready to go. By comparison, for your imaginary in-house developer, they'll need to do development (including shit like getting third-party library approvals), support, training etc along with all the overhead. You'll need to have a manager who may know nothing about computer graphics yet need to come up with plans and roadmaps. Users will learn from scratch. I very much doubt it's a one person job and it's worth it in the end.
You probably should actually talk to corporate IT to understand how things work.
Most people are going to use AI to interact with this. I think the notion that Adobe will always be the leader because of content and tutorials is wrong. Why would I watch a tutorial if I could just use MCP and ask LLM how to do something?
Pretty crazy. I think we’re already very numb to AI. Can you imagine seeing something like this pop up 3 years ago? Youd have been dumbfounded and it would have been major news.
I think there's some fatigue with all the brand new projects that have a skeleton of a project up that isn't really usable yet, where the original prompter doesn't bother to finish it off when it gets too hard. There's a lot of noise that doesn't translate to results so far.
I think this project is interesting, but it's more a wait and see what happens kind of thing.
Agreed. It’s hard to tell what apps will “survive” other then making a splash, especially since there is a constant reliance on llms with fluctuating costs and seemingly an over reliance on rust being the ideal language for everything.
Not surprised, I was super suspicious when I saw this post. Like I bet the underlying image/layout manipulation isn’t that technically challenging at this point, but there is just so much that goes into every single operation from the UI to validations to operation that even with the best use of AI I doubt you could replicate any of the image/video products without several months to a years worth of dedicated effort.
Yes, I would have believed it more if they were targeting one app. But doing all seven at once immediately made me think they likely put no care into this at all. AI can be impressive but it's not up to adobe apps without a lot of care.
I checked GitHub and has only two committers. One +550k lines, another +3k lines.
I didn't even bother downloading any of it to try it out - this is kind of what I expected. Maybe after a massive amount of token spend it'll be somewhat functional?
I loaded my smallest professional project today with fairly simple shapes and text and minimal large smart objects with few effects (essentially 1 page of an educational web system design).
It's not as bad as a few days ago, however it's extremely slow and lots of little bugs to the point I would never use it professionally. Weird things too like my entire mouse pointer is slow, drawing with a brush is slow and jaggy, editing a vector and the points don't highlight properly and ending edit suddenly it randomly moved somewhere on the screen.
Just very clunky bad feeling and not really usable compared to photoshop.
I'll keep an eye on it out of curiosity, but right now it's unusable in a professional context.
As many companies, we've started to use AI on a daily-basis. Yet, it didn't cut the need for testing (both automated and manual) and reviewing.
Now here we see a fully blown suite which is already alarming for basic testability.
Also, we had several apps targeting directly at Adobe in the past.
* Blackmagic DaVinci is the biggest, and it also bought several companies to start fusing a one-stop solution.
* Affinity also started with single app that got attention before being swallowed into Canva.
Leaving compatibility and ui-workflow (which is important to existing users),
It's much more sane to start and focus on a single app instead of a fully blown package.
From Adobe/Creative Cloud users, I do know Acrobat Pro is one of the problematic apps they have in terms (it freezes, guess a lot of legacy code, etc). And yet, even where there is some competition on PDF editors, Acrobat is still strong also due to UX and features.
I don't think manually QA testing a Photoshop clone is realistic for a single person.
This would either need to become a community effort, or rely on AI's computer use and vision capabilities to align the behavior to the Adobe reference software.
In any case, it's very interesting, and I'm curious if this can become good enough for professional use.
three days ago he said he one-shotted the entirety of photoshop. he's been promoting this all over reddit. seems odd to then say it's "very early stage"
This is the thing. Anyone can one shot an app and accept what Claude does. But to go over every single feature where it gets it wrong, and actually get it to do the correct thing, over all of those tiny details in the program - that’s where the 3 days turns into a year and maybe 3.
There’s a major difference between a polished UI and a polished program.
The fact that a model can get this far is still impressive, and it will probably one shot the program properly in < 3 years.
Really? I got the 0.7.0 version today and loaded a couple of my projects and tried using it in my normal day-to-day workflow.
1. Can't use brushes because they are slow and render jaggy polygons instead of the beautiful perfectly smooth brushes of Photoshop - literally can't draw a hand-print looking sketch circle as it looks like a weird digital polygon.
2. I use masks for texturing vector layers, literally cannot ctrl+click to select anything in my file because the texture clipped inside of the layer blocks lower layer selection even though it's completley clipped out of view. Makes it impossible to work on my files without going over to the layers and scrolling hundreds of layers and expanding folders to try to find it. This means I can't use it at all.
3. Try to do dumb little things like turn on one of my color modifying layers and it takes 2 seconds to sit there and wait until it activates instead of a few milliseconds like photoshop.
4. Can't activate my fx layer even though it shows that it IS activated and then others show not activated or something... something completely wrong about this, not even sure what.
5. tried to adjust one of my vector layers couldn't hold alt+hover to adjust the point handles.
6. Trying to type text, takes like 1second per character just to type (photoshop is instant).
7. Can't drag UI sections like Navigator/etc. over to the other panels and can't collapse/uncollapse panels.
8. Can't fill/edit the vector path drawing tool at all I don't think, maybe the UI is just weird but definitely a UX issue or just unusable? Not sure...
I'm not going to bother trying more things, but literally every single thing I tested had a bug or is janky to where it's unusable in a professional context.
I do see 100s of people are working on it now with 10s of thousands of stars on github and it does seem a hair better than my first test, so I'll keep watching the progress, but so far it's just too buggy/glitchy/missing features to work with for me (I have a pretty decent Windows pc - nvida rtx 4080 card, 128gb ram, etc).
Yeah, I got totally numb to AI slop. I looked at it and thought, this is surely so broken that nothing works. Came for the comments and the folks who actually had the time to invest in trying it out confirm that nothing works.
I'm sorry! I didn't mean to publish the repos so soon. This project is brand new. There's been a lot of excitement about this, but we have a long way to go.
If this is about the desktop apps and not the website, I'd love it if you could file an issue. We're going to work through everything.
Sorry me, I shouldn't be putting you down for developing something cool. The mix of grandiose with slop is a bit weird and dangerous tho a bit of humbleness would fit well as we know the effort you put to this isn't much yet. Unless you have years of experience in the video/image editing software.
Nobody is beating Adobe with "claude create for me a clone of adobe apps". Someone that knows what they are doing need to be behind it. It might be you but it would take years of experience if you don't have them yet.
> Nobody is beating Adobe with "claude create for me a clone of the adobe suite".
They're not trying to "beat Adobe" and they certainly didn't wake up one day and tell Claude to make the entire Adobe suite of products.
However, in the not-too-distant future, it'll be the norm to pay $0 for apps and buy AI tokens when needed. That's what Canva did with Affinity--you pay $0 for Affinity Photo, Designer and Publisher that are just as good and in some cases better than Adobe's apps.
Access to advanced features and AI requires a subscription.
Probably easy enough for Adobe to find a few potential things and still shut it down with the threat of lawfare, I'd have thought. They're very wealthy.
> ArtCraft is a form of "fair source" software. It's like free beer in that we give you the code, but there are a few small limitations on what you can do with it.
As seen on the main Artcraft repo. There's other tools that are MIT-licensed, although given that people are noting that this is AI-created the notion of copyright on this is suspect to begin with. Remember kids: all AI slop is born public domain because machines don't have a soul for the purpose of copyright law. And probably also the 13th amendment. Don't upload your brain, kids.
Anyway, let me get up on my soapbox and talk about these "have your cake and eat it too" licenses, that try to be source-available while privileging one particular vendor as the Blessed Owner with special monetization powers. To put it shortly, this doesn't work, because it mixes all the bad parts of FOSS with all the bad parts of proprietary software.
The business model of proprietary software does not work with even source-available terms. Proprietary software is a game of becoming essential to your customers and then taking them for everything they've got. All the successful proprietary companies are, effectively, selling legacy software that was useful to a group of companies a long time ago, but is now so deeply integrated into their IT systems they don't even know they have it until the licensing police show up. "You can use it for free but can't sell it" is anathema to this: it is specifically the "use" part that you're worried about.
Likewise, being able to modify the software - even only for internal uses - means that the proprietary owner has to compete against their own customers for ongoing payments and revenue. Again, if you've dug your talons into a big org to the point where you're indispensable, you don't want your customers making their own forks of your code. You want them paying $$$$/mo for enterprise support.
What FOSS is doing is basically promising to never do these things. If your goal is to make money, then adopting FOSS is like a cat going vegan. And the half-measure source-available licenses don't help, because you wind up conceding all the things that make you money just so you can claim ownership over things most people weren't interested in.
also, from their CLAUDE.md:
> Clean-room. Never read, disassemble or copy anything inside the Illustrator bundle (names/listings only). Never copy Adobe icons, artwork, presets or wording beyond feature names. Behaviour comes from public docs and black-box observation of the running app with synthetic documents only (screenshots by window id, stored under plan/illustrator/screenshots/, never committed). Never copy GPL/AGPL code (Inkscape, lib2geom…).
There is no such thing as clean-room vibe coding. Every line of FOSS code has been trained on. If you use any AI coding model, you are generating code that is going to be similar to copylefted code. There is no "MIT/Apache only" model out there you can use.
Also, no, switching to GPL doesn't help. Actually, it is my opinion that you cannot meaningfully comply with GPL while vibe-coding, because you have changed the definition of "corresponding source code" from a bunch of .rs files to a number of prompts being handed off to a cloud model. Remember: according to GPL, the source code is not "the thing you put into the compiler", it's "the thing the developers prefer to edit to alter the program". Otherwise, you could say "the source code is this uncommented ASM listing" and sidestep the GPL entirely. What we're doing with AI is the same idea, except instead of going down one level of abstraction, we're going up one.
It is exactly the same for books as it is for code: as per the Anthropic case, training is fair use. But that case says nothing about the generated outputs. All the Anthropic case did was say, "the legal system refuses to ban AI outright because copyright is not a union and artists are not entitled to collective protection". Individual outputs can still be infringing.
At inference time, there's a whole new legal question: are the specific outputs of the model substantially similar to anything that the LLM had access to? Even if AI training itself was infringement, a specific generated output of that model could be noninfringing anyway. And vice versa, AI training being fair use doesn't form a blanket defense against a specific AI copyright infringement. If you ask the model to create an art program, it could still generate something substantially similar to the code of an existing art program.
In the case of regurgitation[0], the fact that an LLM was involved is immaterial. The only thing the law cares about is if a copy was made somewhere along the chain and if the alleged copy was substantially similar to the purported original. The only thing LLMs do for this analysis is make access to the original far easier to allege, because LLMs are trained on everything.
[0] Inclusive of both model weights and context regurgitation
They usually have terms and conditions to make it YOUR FAULT if you publish something that has copyright, and they have paid loads of legal fees after-the-event for stolen books in their training data. It's never been fair use, just very wealthy people getting away with criminal acts.
Lovable have the following terms, for example -
"You assume full responsibility for your use of AI Output and agree not to rely on it for critical or high-risk functions (including medical, legal, financial, or safety-related purposes) without appropriate safeguards. AI Output may be similar or identical to content generated for other users who submit similar prompts. We do not guarantee that AI Output will be unique, free of third-party rights, accurate, suitable for any particular purpose, or that it will build, repair, complete, or improve any application or project or achieve any particular result."
It's wild that anyone still uses Adobe software these days. Adobe is one of the most famous companies for enshittification, and the open-source alternatives are generally very good.
Personally, I switched to open-source alternatives way back in the late 1900s, and they've only continued to improve ever since. Adobe has been lagging behind for over a quarter of a century, in the ways which matter to me, and they keep getting farther behind as they turn to worse and worse methods of extorting money from people.
YMMV, of course.
I can get behind almost anything which reduces lock-in for Adobe's remaining customers, but from what I see in other comments, this probably ain't it. It seems very vibey and buggy. But maybe after more work is put in, it could be good.
i personally like affinity better than the FOSS photoshop alternatives, mostly for UX reasons. what i still haven’t found a good replacement for is Lightroom. i use darktable but the way the interface is laid out is antithetical to getting into an efficient flow state. and i say this despite the fact that i consider the output strictly superior to Lightroom (or in fact anything i’ve ever used).
These things have very different personalities somehow.
Over the years I have tried Capture One — it's fine, I dunno, I just didn't settle with it, even though I shoot a lot of portraits which is its comfort zone. (I actually started out with it, free thing from SanDisk)
I really like DxO Photolab but I know this app enrages lots of people; I particularly like the tone curve handling, FilmPack integration, etc.
Lately I have been learning ART, the RawTherapee fork, and I have to say that despite its very idiosyncratic interface, I like this quite a lot.
It’s actually quite simple, give me any open source alternative and I can give you ten major features it lacks that Photoshop has.
If you just use Photoshop as a pixel editing tool, sure, but it does a lot more than that.
I’ve tried getting rid of creative suite for years, but when it comes down to it no competitor has actually come close.
Sure, anyone can list ten features Photoshop has that GIMP, for example, lacks. But it's just as easy to list ten features GIMP has that Photoshop lacks. They're not trying to be the same program.
I see the same argument all the time about Windows vs Linux. Linux lacks this, Linux lacks that, etc... but somehow, people fail to notice that Windows lacks a whole lot of important things that Linux has. Because a Windows user won't be aware of what they're missing. But for me, coming from a Linux background, Windows is missing so much I would view it as unusable.
Changing from one to the other is painful, because of the differences. It's not a drop-in replacement you can keep using exactly the same way. But that doesn't make either one better or worse. They're different, and that's a good thing. But different requires learning, and learning is hard.
Anyway, you're correct that $FreeProgram isn't a drop-in replacement for $ProprietaryProgram. But that's not a valid argument on the relative value of the two programs.
The interesting parts are what is lacking though, not just any feature. A life without smart objects is terrible in photo editing as an example, at least once you’ve seen the upsides.
That and Camera Raw actually being an incredibly capable raw conversion engine.
But I’m not trying to convince anyone here, just saying it makes sense people are paying for Adobe software.
Network effects like presets and plugin ecosystem, tutorials and books, team knowledge (you can ask a colleague how to do things).
Switching costs like lightroom catalogues, learning new programs, if it works why fixit.
And the price is not that high if your company pays for it. Open stuff would have to be much better and Adobe much more buggy before most people consider switching.
It's a bit like saying you never hear complaints from piano players, only dentists. But dentists also play piano, just like everyone else. Dentists just know more about teeth than the average pianist.
I only know two or three developers who have a strong concept of what Photoshop does or can do, and all of those only use a fraction of what it does (a couple of layer blend modes, healing brush, soft proofing, maybe image slicing).
There are people who live inside photoshop the way developers live inside VSC or Zed or emacs.
FWIW I do hear these complaints from people who live in Adobe apps but they are very pragmatic about how they approach it. For example, Lightroom is not without its faults but photographers will switch to Capture One and delay leaving the suite so they don't have to relearn everything at once.
Some of those photographers then use their new baseline familiarity with Capture One to test out Affinity for further edits.
But you're just never going to have the time to burn your bridges with the platform in one go. I don't think that is very different from the way developers work; in practice nobody with any serious obligations can simply switch language and toolchain in one go.
So there is attrition, but there is never going to be a threshold over which people cross and get a widespread revolt. I'm not even sure that happened with the switch to Creative Cloud, because they padded that transition by making it easier on photographers and with the other single app plan (that I think no longer exists?)
My own feeling is that there are serious cracks in the way Photoshop works. For example the "smart objects" bodge when they could do what Affinity Photo does with live filter layers, not to mention (when comparing with Affinity) Photoshop's weaker layer blend controls, lack of cross-model curves, lack of a frequency separation tool, culture of over-reliance on plugins.
But Photoshop has things Affinity does not, yet — a mature DAM/Bridge integration, some of the more advanced warp tools, etc., and tiny details like better perlin noise generation.
So it's like the situation with Word, isn't it — 95% of people use only a tiny fraction of what Word can do, but they don't all use the same tiny fraction.
People have their reasons to switch or not switch; even among hobbyist photographers who complain about the cost of everything else, the price of the photography plan does not seem to be one of them (thanks in part to Adobe's strategic reliance on Amazon Prime Day/Deal Days bundles)
The most interesting part to me is their PDF stuff, as that has been at times somewhat lackluster on Linux, and there's this huge huge huge existing base of non-artists that could benefit from something like that.
I'm not sure if it really makes sense to copy the UI and visual language of the source that verbatim (and I believe that Adobe will rightfully retaliate on specifically that), but for the time being, that's neat.
I bet the industry didn't expect that their IP laundering tools would eventually launder their own IP.
The projects proposed couldn't be further away from the FOSS bubble with its ideas, values and spirit.
It is - effectively - exactly the same in spirit as Adobe. It's just business people fighting each other.
Which is nice don't get me wrong, but it's a different kind of nice.
Phrased differently and more inflammatory, the F in Foss does not mean fraud and hopefully never will. This however is just that. Defrauding someone out of their unjustified monopoly through GPU compute instead of merit.
Just because Adobe sucks doesn't mean that someone capable of doing that and doing it this blatantly would be trustworthy and "on our side".
I tried their photoshop "equivalent" and it's a piece of garbage. Essentially nothing works. I'm sure it will improve but as of right now I would be pretty embarrassed to put something out that works so poorly.
I'm struggling with the audience? The main thing about Adobe right now is millions of hours of youtube videos that show me how to do stuff. I would be completely lost with this software
If you have an LLM and MCP why do you need a tutorial? Why would I watch a tutorial when the LLM can tell me how to use a feature? There is not a moat just because there are tutorials. That moat is gone. I don't need to watch tutorials anymore.
There are some incredibly smart people on HN, but I wasn't quite ready for this project to face the firing squad. It's still much too new for the world's harshest critics.
With lots of additional effort, I expect this to succeed much like bun. Fingers crossed.
Obviously you don't need to pay for Photoshop anymore. Does that mean your product is the replacement? No. Just an example.
The fact that you or anyone can get to 95-99% very quickly. The rest can be added over time at a very high pace and some "actual effort".
Unless Photoshop can add some sort of proprietary features then why use them? (they have been doing this with AI features mostly, but it should still all be copyable)
the "photoshop equivalent" is really broken at a fundamental level. I can't do ANYTHING I tried to do. I'm sure it sort of works but the basics aren't there. Have you tried using it yourself?
I mean Claude Code is one of the most buggy piece of software in existence, so I wouldn't use that as a measurement.
There's also massive marketing forces in play here that can't be ignored.
Bro, you don't get to ask for calm and patience when you make claims like "software is over. But for real. I just one-shotted Photoshop. The whole thing." and then post your repo in various subreddits and HN comments. You did ask for this, whether you have the self-awareness to realize it or not.
But look on the bright side, this is a chance to prove all of the haters wrong. Given your enthusiasm about AI and its capabilities it shouldn't be an issue.
I see, I was wondering why you did not post it to HN and only mentioned it in comments.
Makes perfect sense to me, since this is obviously an early alpha version. It already looks very impressive to me, and I look forward to the project proving these doubters wrong.
I hope there will not be legal trouble with Adobe.
Indeed — what I am reading of this strongly suggests a misunderstanding about exactly what it is that makes Photoshop what it is.
I think it might be possible to slop-code an Excel clone because — truly minor interactivity quibbles aside — it can be tested with a truly exhaustive suite. The clone can be proved to do what it is supposed to do, over whatever subset one would choose to implement.
But Photoshop is not an application in the same way that Excel is an application.
(Not least because while powerful, Excel, like Word, is an implementation of an existing idea. Photoshop is the OG; it has all the history of the OG)
Photoshop is also a surface and a medium. It has a feel, it has responsiveness. It's like oil painting or watercolour: its own thing, with its own pen/brush dynamics, its own interactivity.
Part of the reason people don't want to switch away to Affinity (even though Affinity is free as in beer, highly capable and in some ways better) is just that to a set of very hardcore users, it does not have this feel. They just don't like how it feels.
Automated testing that focusses on pixel-precise rendering misses the point. It may actually matter less that it renders PSDs pixel-accurately than that it actually feels right.
If you are not convinced by my argument, look at "dodge and burn retouching". For better or worse, high end fashion photo retouching is done in an extremely precise, wildly technical, tiny subset of Photoshop tools.
If you want to replace Photoshop it has to be responsive at that level. And since that level is partly subjective experience, it's going to take some people.
Office apps are strictly based on written spec. You can if spend good amount of token could easily build something very close not but 100% with that spec. PS is a different animal.
Worked with Excel’s spec for years the actual behavior has plenty of quirks that aren’t covered by the spec, so cloning it isn’t as easy as it sounds and I believe its the same with Adobe products
That "feel" is how drawing algorithms work, how responsive the interface is and so on. You can do automatic testing for that as well.
I played a bit with porting NES games with an older Claude. It just wrote tests not just for pixel perfectness, but for frames, timing and so on.
Sure Adobe will fight anyone that tries to do that openly, but historically such attempts failed in a grand scheme of things. Stopping a single project like this, even if you manage to trace down the devs, is a years long litigation - and in that time such project gets enough adoption to demolish the incumbent. And when one project dies, five other spawn in its place.
> what I am reading of this strongly suggests a misunderstanding about exactly what it is that makes Photoshop what it is.
do you mean marketing and market manipulation by encouraging piracy by young people and student so they develop habits with the software then force companies to buy the subscription ?
Every piece of beta software that has ever existed, has bugs. AI doesn't change this. Software development has largely changed to build with AI first, get basic things working, iterate and fix. This is just the reality older developers just have to accept.
Bugs are ever present. And beta software should be treated with charity. But Photoshop, let alone the entire suite, is a beast. An effort, like this says it is, aiming for parity fundamentally misunderstands the reality of both the software they’re trying to replicate and the capability of AI assistance now and in the foreseeable suitable future
The question is how much of that beast is really necessary for a true MVP. My bet is that you can have a stable core of Photoshop, and an option for designers to easily add/port plugins for whatever is necessary for their specific workflows.
We see that already with CMSes and CRMs. Sure Wordpress is a beast, but the vast majority of that complexity comes from it being universal.
I've seen non-technical people building their CMSes and even CRMs for small businesses within a fraction of time it would take them to set up the off the shelf product. The biggest challenge there is security of course - but Photoshop clones don't have to deal with it too much.
My bet is that in single months we will have a stable clone of the Photoshop core, and with an ability to easily modify it, designers will just build their own custom plugins to fit their workflows better than PS ever could.
Unlike CMS/CRMs the cost of switch/testing out is minimal. You don't need to switch the whole company to it, you don't have security issues, and if it's too buggy you just switch back.
Also, this is a way for graphic designers to maintain their relevance - if they can have full controls of their tools, they can do the work faster and better and maintain their creative freedom.
Very late to this but for what it's worth I stumbled upon this and tested their release from earlier today and it seems basically fine. Everything I've tried works except for popping out panels (like the history panel, layers panel, etc) which doesn't appear to be supported. Not sure what was broken for you since you neglected to elaborate, but if things weren't working for you right up front then it must be iterating quickly. Not saying this to brag or to take a stance on it or anything, just observing.
Yes, I build and ran the photoshop clone on my Mac just now, and it seemed mostly fine.
For the problems I did face, I am not sure if they were bugs in the software or just me not being familiar with Photoshop with the Mac's touchpad controls.
I'd think this project kind of settles whether one can build large, working software with AI, for those that were still unconvinced.
I am surprised this has not seen more discussion on HN. The post is flagged.
... brought to you by storyteller.ai, which now has a defunct hero video, but apparently still 10 mio users (at $12+ monthly). I very much applaud them for pouring their (est.) $1.4B ARR into Open Source.
Given the vibes i have very little hope, but i have to have some hope, because at this point I've been waiting my whole life for an open source InDesign competitor (no, scribus is not competitive)
Hopefully the suite will not be abandoned in one month, and it'll become actually production-ready one day. Although this seems like slop for now (at least it's rust, not electron), enough contributions may (hopefully) make it worth something.
206 comments
[ 1.6 ms ] story [ 37.8 ms ] threadThese are still super early alpha. All MIT/Apache licensed. No telemetry.
I'm going to try to drive these to 100% parity but it's going to take a minute. And I'll need lots of help. (Anyone want to join me? I'll pay stipends and bug bounties.)
I'm going to do Microsoft Office, AutoCAD, and Solidworks too. Probably a lot of stuff to collapse.
It's not just SaaSocalypse. Everything can be open source multiplatform Rust now.
Party like it's 1999 and the Internet / tech scene belongs to us again. We can literally deenshittify the world without breaking a sweat.
But its a lot harder if you want to maintain at least some level of quality. For example, your PDF export exports texts as paths and correct PDF encoding and font subsetting took months of trial and error (I maintain the printpdf crate).
So you have to actually use your apps, test extensively, etc. - don‘t underestimate the effort required for QA and testing.
Don't let that discourage you, though! A fast, local, fully-featured 3D parametric CAD program that takes advantage of modern hardware parallelism that can be extended/scripted is just basically one of the holy grails of software and even exploratory efforts might uncover some great paths forward to making that a reality!
With photoshop, illustrator, etc., there’s a fair amount of code out there. No LLM has however ever seen the source code —- or I suspect even much detailed discussion —- of a bRep kernel better than Opencascade, but with the right people to decide what maths to work on, maybe there is progress to be made. This knowledge is held by few people, many of whom will now have retired!
A big of the attraction of re-building this kind of foundational software isn't the base capability (commercial 3D CAD has barely even seen incremental improvements for more than a decade, in large part because it's just pretty well solved in the existing programs), but that a modern kernel would be a major step forward for unblocking implementation of some of those other capabilities.
There’s plenty of work that’s not appropriate for, but works for me.
It started about 6 months back, kernel build from scratch in go. I changed directions now and decided to rewrite in rust, take the good parts of occt and the good parts of my own kernel and build a new one child of both.
So far it did not get traction, I guess I should had named it SolidCraft :D
You may want those in new high-performance infinite precision: https://github.com/timschmidt/hyperreal
Exact curves: https://github.com/timschmidt/hypercurve
Meshes: https://github.com/timschmidt/hypermesh
and more in the ~22 crate stack.
I'll be pinging you about this in the future
we remember different 1999s. i remember everyone pirating software made by large companies at that moment in time
i am not sure i follow. it certainly isn't technically the same.
Someone will make up new laws just to get rid of that specifically and that will likely even work, if one guy is behind _all_ the recreations.
Maybe you could somehow distribute that, so that it's at least a game of whack-a-mole for the defender.
These are not monetized in any way. They don't have telemetry, paywalls, or anything.
If you're looking at the other software we've built, that's a source available (soon to be open source) Higgsfield and OpenRouter/FAL.
https://www.reddit.com/r/SaaS/comments/1wwnhbt/ive_created_o...
https://www.reddit.com/r/OpenAI/comments/1wxmunv/ive_created...
https://www.reddit.com/r/GeminiAI/comments/1wwo52q/ive_creat...
https://www.reddit.com/r/aiwars/comments/1wvz0lj/100_open_so...
https://www.reddit.com/r/vibecoding/comments/1wvz4ua/100_ope...
It's also rust based and very polished with a cli tool for automation. Can't speak to feature parity as I'm not an accountant or someone who traffics in spreadsheets but it suits my needs to the point I don't reach for a company subscription.
That's the bar for an AI generated app that I'd like to see your Adobe CC/CS apps hit.
I'll keep an eye on your repos!
https://github.com/storytold
Nope. I just downloaded and installed PhotoCraft [1] and launched it. No sign-up, dialog boxes, licensing agreement.
The installer for LightCraft [2] isn't ready yet; building it from source took about 5 minutes.
[1]: https://getartcraft.com/apps/photocraft
[2]: https://getartcraft.com/apps/lightcraft
The Adobe apps are 100% open source, desktop, MIT, no telemetry.
https://github.com/storytold/photocraft etc.
It’s not a simple port, is it? The iPad apps have significantly different UI.
Yes, even hardware acceleration works.
Adobe to $120 price target at least, could be even lower.
I think the same thing about the endless proliferation of Web-dev frameworks. Why bother? Just have AI write it in native JS or WebAssembly and be done with it.
What about their service?
But then I remembered that the flight had free Internet. So I log in... only to have these morons try to TEXT my phone a 2FA code. On the plane.
After I got back I checked with our IT department and found that our Adobe licenses are paid up for TWO YEARS.
Despicable.
Photoshop has a lot of bad things going for it, but it’s still the only software that does what it does, and I’ve tried all the competitors that have popped up over the years.
It's not that different from having Claude go to a website and sniff all the API calls it makes as you're interacting with it and then using that info to create a website that you prefer interacting with and all the layers and interactive components that you like.
A company paying $1-5m for software that is reliable, provides support and created by experts in the area will consider that a deal. They go through standard training -- only in case the user hasn't used the software before -- and they're ready to go. By comparison, for your imaginary in-house developer, they'll need to do development (including shit like getting third-party library approvals), support, training etc along with all the overhead. You'll need to have a manager who may know nothing about computer graphics yet need to come up with plans and roadmaps. Users will learn from scratch. I very much doubt it's a one person job and it's worth it in the end.
You probably should actually talk to corporate IT to understand how things work.
I think this project is interesting, but it's more a wait and see what happens kind of thing.
Every Crafting App is written from scratch in Rust and held to the same rules."
Yes. We are numb to AI (slop). Maybe if this project is still being developed in a year people might take it seriously.
Just curious, because I've worked within design/film etc. for years.
I did open the PhotoCraft and loaded a few of my simpler layout projects. And literally nothing works.
Let's see:
1. Tried to select text, like the selector is off from the cursor.
2. Tried to edit text, takes like 3 seconds to drag the slider for size, nothing happens (text doesn't change size).
3. Absolutely everything is insanely laggy, selecting, switching, etc.
4. Can't seem to expand/collapse folders.
5. Everything shows shortcut keys, shortcut keys don't do anything.
6. Tried to move something, kind of got frozen for a while. Eventually disappeared I think?
7. Clicked something, selected a side panel (don't know why), hid the other panel, can't stretch, collapse the panels or get it back...
This is actual joke software, I'm confused why you are impressed?
I checked GitHub and has only two committers. One +550k lines, another +3k lines.
This was on a fresh DNG, not some big existing project.
Perhaps with massive amount of spend it can be brought up to parity to work even for complex Photoshop projects.
It's not as bad as a few days ago, however it's extremely slow and lots of little bugs to the point I would never use it professionally. Weird things too like my entire mouse pointer is slow, drawing with a brush is slow and jaggy, editing a vector and the points don't highlight properly and ending edit suddenly it randomly moved somewhere on the screen. Just very clunky bad feeling and not really usable compared to photoshop.
I'll keep an eye on it out of curiosity, but right now it's unusable in a professional context.
Now here we see a fully blown suite which is already alarming for basic testability.
Also, we had several apps targeting directly at Adobe in the past.
* Blackmagic DaVinci is the biggest, and it also bought several companies to start fusing a one-stop solution.
* Affinity also started with single app that got attention before being swallowed into Canva.
Leaving compatibility and ui-workflow (which is important to existing users), It's much more sane to start and focus on a single app instead of a fully blown package.
From Adobe/Creative Cloud users, I do know Acrobat Pro is one of the problematic apps they have in terms (it freezes, guess a lot of legacy code, etc). And yet, even where there is some competition on PDF editors, Acrobat is still strong also due to UX and features.
This would either need to become a community effort, or rely on AI's computer use and vision capabilities to align the behavior to the Adobe reference software.
In any case, it's very interesting, and I'm curious if this can become good enough for professional use.
Creative apps don’t fit in either of those categories.
A house of mirrors.
The fact that this does everything it does is super impressive.
There’s a major difference between a polished UI and a polished program.
The fact that a model can get this far is still impressive, and it will probably one shot the program properly in < 3 years.
1. Can't use brushes because they are slow and render jaggy polygons instead of the beautiful perfectly smooth brushes of Photoshop - literally can't draw a hand-print looking sketch circle as it looks like a weird digital polygon.
2. I use masks for texturing vector layers, literally cannot ctrl+click to select anything in my file because the texture clipped inside of the layer blocks lower layer selection even though it's completley clipped out of view. Makes it impossible to work on my files without going over to the layers and scrolling hundreds of layers and expanding folders to try to find it. This means I can't use it at all.
3. Try to do dumb little things like turn on one of my color modifying layers and it takes 2 seconds to sit there and wait until it activates instead of a few milliseconds like photoshop.
4. Can't activate my fx layer even though it shows that it IS activated and then others show not activated or something... something completely wrong about this, not even sure what.
5. tried to adjust one of my vector layers couldn't hold alt+hover to adjust the point handles.
6. Trying to type text, takes like 1second per character just to type (photoshop is instant).
7. Can't drag UI sections like Navigator/etc. over to the other panels and can't collapse/uncollapse panels.
8. Can't fill/edit the vector path drawing tool at all I don't think, maybe the UI is just weird but definitely a UX issue or just unusable? Not sure...
I'm not going to bother trying more things, but literally every single thing I tested had a bug or is janky to where it's unusable in a professional context.
I do see 100s of people are working on it now with 10s of thousands of stars on github and it does seem a hair better than my first test, so I'll keep watching the progress, but so far it's just too buggy/glitchy/missing features to work with for me (I have a pretty decent Windows pc - nvida rtx 4080 card, 128gb ram, etc).
Yeah, I got totally numb to AI slop. I looked at it and thought, this is surely so broken that nothing works. Came for the comments and the folks who actually had the time to invest in trying it out confirm that nothing works.
If this is about the desktop apps and not the website, I'd love it if you could file an issue. We're going to work through everything.
Nobody is beating Adobe with "claude create for me a clone of adobe apps". Someone that knows what they are doing need to be behind it. It might be you but it would take years of experience if you don't have them yet.
They're not trying to "beat Adobe" and they certainly didn't wake up one day and tell Claude to make the entire Adobe suite of products.
However, in the not-too-distant future, it'll be the norm to pay $0 for apps and buy AI tokens when needed. That's what Canva did with Affinity--you pay $0 for Affinity Photo, Designer and Publisher that are just as good and in some cases better than Adobe's apps.
Access to advanced features and AI requires a subscription.
As seen on the main Artcraft repo. There's other tools that are MIT-licensed, although given that people are noting that this is AI-created the notion of copyright on this is suspect to begin with. Remember kids: all AI slop is born public domain because machines don't have a soul for the purpose of copyright law. And probably also the 13th amendment. Don't upload your brain, kids.
Anyway, let me get up on my soapbox and talk about these "have your cake and eat it too" licenses, that try to be source-available while privileging one particular vendor as the Blessed Owner with special monetization powers. To put it shortly, this doesn't work, because it mixes all the bad parts of FOSS with all the bad parts of proprietary software.
The business model of proprietary software does not work with even source-available terms. Proprietary software is a game of becoming essential to your customers and then taking them for everything they've got. All the successful proprietary companies are, effectively, selling legacy software that was useful to a group of companies a long time ago, but is now so deeply integrated into their IT systems they don't even know they have it until the licensing police show up. "You can use it for free but can't sell it" is anathema to this: it is specifically the "use" part that you're worried about.
Likewise, being able to modify the software - even only for internal uses - means that the proprietary owner has to compete against their own customers for ongoing payments and revenue. Again, if you've dug your talons into a big org to the point where you're indispensable, you don't want your customers making their own forks of your code. You want them paying $$$$/mo for enterprise support.
What FOSS is doing is basically promising to never do these things. If your goal is to make money, then adopting FOSS is like a cat going vegan. And the half-measure source-available licenses don't help, because you wind up conceding all the things that make you money just so you can claim ownership over things most people weren't interested in.
also, from their CLAUDE.md:
> Clean-room. Never read, disassemble or copy anything inside the Illustrator bundle (names/listings only). Never copy Adobe icons, artwork, presets or wording beyond feature names. Behaviour comes from public docs and black-box observation of the running app with synthetic documents only (screenshots by window id, stored under plan/illustrator/screenshots/, never committed). Never copy GPL/AGPL code (Inkscape, lib2geom…).
There is no such thing as clean-room vibe coding. Every line of FOSS code has been trained on. If you use any AI coding model, you are generating code that is going to be similar to copylefted code. There is no "MIT/Apache only" model out there you can use.
Also, no, switching to GPL doesn't help. Actually, it is my opinion that you cannot meaningfully comply with GPL while vibe-coding, because you have changed the definition of "corresponding source code" from a bunch of .rs files to a number of prompts being handed off to a cloud model. Remember: according to GPL, the source code is not "the thing you put into the compiler", it's "the thing the developers prefer to edit to alter the program". Otherwise, you could say "the source code is this uncommented ASM listing" and sidestep the GPL entirely. What we're doing with AI is the same idea, except instead of going down one level of abstraction, we're going up one.
At inference time, there's a whole new legal question: are the specific outputs of the model substantially similar to anything that the LLM had access to? Even if AI training itself was infringement, a specific generated output of that model could be noninfringing anyway. And vice versa, AI training being fair use doesn't form a blanket defense against a specific AI copyright infringement. If you ask the model to create an art program, it could still generate something substantially similar to the code of an existing art program.
In the case of regurgitation[0], the fact that an LLM was involved is immaterial. The only thing the law cares about is if a copy was made somewhere along the chain and if the alleged copy was substantially similar to the purported original. The only thing LLMs do for this analysis is make access to the original far easier to allege, because LLMs are trained on everything.
[0] Inclusive of both model weights and context regurgitation
Lovable have the following terms, for example - "You assume full responsibility for your use of AI Output and agree not to rely on it for critical or high-risk functions (including medical, legal, financial, or safety-related purposes) without appropriate safeguards. AI Output may be similar or identical to content generated for other users who submit similar prompts. We do not guarantee that AI Output will be unique, free of third-party rights, accurate, suitable for any particular purpose, or that it will build, repair, complete, or improve any application or project or achieve any particular result."
Personally, I switched to open-source alternatives way back in the late 1900s, and they've only continued to improve ever since. Adobe has been lagging behind for over a quarter of a century, in the ways which matter to me, and they keep getting farther behind as they turn to worse and worse methods of extorting money from people.
YMMV, of course.
I can get behind almost anything which reduces lock-in for Adobe's remaining customers, but from what I see in other comments, this probably ain't it. It seems very vibey and buggy. But maybe after more work is put in, it could be good.
(Disclosure: I'm the developer)
Over the years I have tried Capture One — it's fine, I dunno, I just didn't settle with it, even though I shoot a lot of portraits which is its comfort zone. (I actually started out with it, free thing from SanDisk)
I really like DxO Photolab but I know this app enrages lots of people; I particularly like the tone curve handling, FilmPack integration, etc.
Lately I have been learning ART, the RawTherapee fork, and I have to say that despite its very idiosyncratic interface, I like this quite a lot.
I’ve tried getting rid of creative suite for years, but when it comes down to it no competitor has actually come close.
I see the same argument all the time about Windows vs Linux. Linux lacks this, Linux lacks that, etc... but somehow, people fail to notice that Windows lacks a whole lot of important things that Linux has. Because a Windows user won't be aware of what they're missing. But for me, coming from a Linux background, Windows is missing so much I would view it as unusable.
Changing from one to the other is painful, because of the differences. It's not a drop-in replacement you can keep using exactly the same way. But that doesn't make either one better or worse. They're different, and that's a good thing. But different requires learning, and learning is hard.
Anyway, you're correct that $FreeProgram isn't a drop-in replacement for $ProprietaryProgram. But that's not a valid argument on the relative value of the two programs.
But I’m not trying to convince anyone here, just saying it makes sense people are paying for Adobe software.
Network effects like presets and plugin ecosystem, tutorials and books, team knowledge (you can ask a colleague how to do things).
Switching costs like lightroom catalogues, learning new programs, if it works why fixit.
And the price is not that high if your company pays for it. Open stuff would have to be much better and Adobe much more buggy before most people consider switching.
It's a bit like saying you never hear complaints from piano players, only dentists. But dentists also play piano, just like everyone else. Dentists just know more about teeth than the average pianist.
There are people who live inside photoshop the way developers live inside VSC or Zed or emacs.
Some of those photographers then use their new baseline familiarity with Capture One to test out Affinity for further edits.
But you're just never going to have the time to burn your bridges with the platform in one go. I don't think that is very different from the way developers work; in practice nobody with any serious obligations can simply switch language and toolchain in one go.
So there is attrition, but there is never going to be a threshold over which people cross and get a widespread revolt. I'm not even sure that happened with the switch to Creative Cloud, because they padded that transition by making it easier on photographers and with the other single app plan (that I think no longer exists?)
My own feeling is that there are serious cracks in the way Photoshop works. For example the "smart objects" bodge when they could do what Affinity Photo does with live filter layers, not to mention (when comparing with Affinity) Photoshop's weaker layer blend controls, lack of cross-model curves, lack of a frequency separation tool, culture of over-reliance on plugins.
But Photoshop has things Affinity does not, yet — a mature DAM/Bridge integration, some of the more advanced warp tools, etc., and tiny details like better perlin noise generation.
So it's like the situation with Word, isn't it — 95% of people use only a tiny fraction of what Word can do, but they don't all use the same tiny fraction.
People have their reasons to switch or not switch; even among hobbyist photographers who complain about the cost of everything else, the price of the photography plan does not seem to be one of them (thanks in part to Adobe's strategic reliance on Amazon Prime Day/Deal Days bundles)
I'm not sure if it really makes sense to copy the UI and visual language of the source that verbatim (and I believe that Adobe will rightfully retaliate on specifically that), but for the time being, that's neat.
I bet the industry didn't expect that their IP laundering tools would eventually launder their own IP.
Can't believe HN sentiment is taking adobe's side over open source, just because the latter used AI.
It is - effectively - exactly the same in spirit as Adobe. It's just business people fighting each other.
Which is nice don't get me wrong, but it's a different kind of nice.
Phrased differently and more inflammatory, the F in Foss does not mean fraud and hopefully never will. This however is just that. Defrauding someone out of their unjustified monopoly through GPU compute instead of merit.
Just because Adobe sucks doesn't mean that someone capable of doing that and doing it this blatantly would be trustworthy and "on our side".
> Software is over. But for real. I just one-shotted Photoshop. The whole thing.
You're telling me it's not the entirety of photoshop?
99% parity will take a while, but I'm sure it'll be measured in months and not years.
Feel free to check in on this claim. It's measurable.
There are some incredibly smart people on HN, but I wasn't quite ready for this project to face the firing squad. It's still much too new for the world's harshest critics.
With lots of additional effort, I expect this to succeed much like bun. Fingers crossed.
Obviously you don't need to pay for Photoshop anymore. Does that mean your product is the replacement? No. Just an example.
The fact that you or anyone can get to 95-99% very quickly. The rest can be added over time at a very high pace and some "actual effort".
Unless Photoshop can add some sort of proprietary features then why use them? (they have been doing this with AI features mostly, but it should still all be copyable)
fact? what fact? have people lost their fucking minds here?
And then four days ago:https://news.ycombinator.com/item?id=49922904
And you've mentioned ArtCraft numerous times https://hn.algolia.com/?q=by%3Aechelon+artcraft
Obviously you've been trying to get traction on this for a while. You can't claim that you "one-shotted photoshop" and not expect people to notice.
Calm down and be patient. Let the community token spend have time to diffuse into the products.
This effort needs a community to climb the quality and parity gradients. But I'm not ready to really show it off until we've hit some thresholds.
I've got the attention I needed to bootstrap this, so I can go back to work.
But look on the bright side, this is a chance to prove all of the haters wrong. Given your enthusiasm about AI and its capabilities it shouldn't be an issue.
Makes perfect sense to me, since this is obviously an early alpha version. It already looks very impressive to me, and I look forward to the project proving these doubters wrong.
I hope there will not be legal trouble with Adobe.
scbench.ai suggests this will happen on projects where the human is no longer in the loop, and it's a good sized project to find out.
I think it might be possible to slop-code an Excel clone because — truly minor interactivity quibbles aside — it can be tested with a truly exhaustive suite. The clone can be proved to do what it is supposed to do, over whatever subset one would choose to implement.
But Photoshop is not an application in the same way that Excel is an application.
(Not least because while powerful, Excel, like Word, is an implementation of an existing idea. Photoshop is the OG; it has all the history of the OG)
Photoshop is also a surface and a medium. It has a feel, it has responsiveness. It's like oil painting or watercolour: its own thing, with its own pen/brush dynamics, its own interactivity.
Part of the reason people don't want to switch away to Affinity (even though Affinity is free as in beer, highly capable and in some ways better) is just that to a set of very hardcore users, it does not have this feel. They just don't like how it feels.
Automated testing that focusses on pixel-precise rendering misses the point. It may actually matter less that it renders PSDs pixel-accurately than that it actually feels right.
If you are not convinced by my argument, look at "dodge and burn retouching". For better or worse, high end fashion photo retouching is done in an extremely precise, wildly technical, tiny subset of Photoshop tools.
If you want to replace Photoshop it has to be responsive at that level. And since that level is partly subjective experience, it's going to take some people.
I played a bit with porting NES games with an older Claude. It just wrote tests not just for pixel perfectness, but for frames, timing and so on.
Sure Adobe will fight anyone that tries to do that openly, but historically such attempts failed in a grand scheme of things. Stopping a single project like this, even if you manage to trace down the devs, is a years long litigation - and in that time such project gets enough adoption to demolish the incumbent. And when one project dies, five other spawn in its place.
do you mean marketing and market manipulation by encouraging piracy by young people and student so they develop habits with the software then force companies to buy the subscription ?
I want to leave adobe for good and this LightCraft thing does what no other alternative can, Presets from lightroom.
if the creator fixes Masks and uses local ai possibilities for masking and such, this could very fast result in me canceling my sub.
So far yes its a vibe-coded mess but your crazy telling me adobes hella broken software is better.
We see that already with CMSes and CRMs. Sure Wordpress is a beast, but the vast majority of that complexity comes from it being universal.
I've seen non-technical people building their CMSes and even CRMs for small businesses within a fraction of time it would take them to set up the off the shelf product. The biggest challenge there is security of course - but Photoshop clones don't have to deal with it too much.
My bet is that in single months we will have a stable clone of the Photoshop core, and with an ability to easily modify it, designers will just build their own custom plugins to fit their workflows better than PS ever could.
Unlike CMS/CRMs the cost of switch/testing out is minimal. You don't need to switch the whole company to it, you don't have security issues, and if it's too buggy you just switch back.
Also, this is a way for graphic designers to maintain their relevance - if they can have full controls of their tools, they can do the work faster and better and maintain their creative freedom.
For the problems I did face, I am not sure if they were bugs in the software or just me not being familiar with Photoshop with the Mac's touchpad controls.
I'd think this project kind of settles whether one can build large, working software with AI, for those that were still unconvinced.
I am surprised this has not seen more discussion on HN. The post is flagged.