Rendered at 23:29:45 GMT+0000 (Coordinated Universal Time) with Cloudflare Workers.
ayaros 20 hours ago [-]
As a very long time user and lover of SketchUp, I'm still peeved about the direction Trimble took in neglecting the SketchUp desktop modeler in favor of a web-based UI. Their licensing was also terrible... first they discontinued the free version of SketchUp, and then later on they discontinued perpetual licensing, meaning the only way to get the desktop modeler was to pay several hundred per year, which wasn't worth it to me. If a vibecoded alternative displaces it in the market, I won't feel one bit sorry for them.
The kinds of large scale projects I like to make in SketchUp really began to feel constrained at a certain point. A modeler with SketchUp's incredible UI that can take full advantage of any computer's available graphics hardware to render large amounts of geometry more efficiently would be a wonderful thing.
I'm very possessive of my SketchUp models, and so the idea of storing them on any kind of cloud server I don't have control over is abhorrent to me. I was not happy to see Trimble putting so much effort into making a web-based modeler.
fsloth 9 hours ago [-]
"he only way to get the desktop modeler was to pay several hundred per year, which wasn't worth it to me. "
This comes up so often it's good to explain what happened there.
They were not out to extract value from consumer.
They retargeted the offering to professional engineering and architecture firms.
SketchUp was originally an offering for the AEC market. Then Google bought it and essentially pivoted it to consumer offering. Then Trimble bought it and pivoted it back.
I do realize this hurts the person who is a non-professional user but AFAIK SketchUp sales have never been better.
So it's not a pivot out of stupidity or malice, it's just a pivot.
The main customers - engineering and architecture companies - seem to be fine with it.
If you were an engineer CAD manager it would be odd to pay hundreds per seat a year for a product consumers can get for a pittance. So it's about pricing power I guess as well.
And do agree as a consumer there should be better options.
Loquebantur 5 hours ago [-]
The problem here is what you euphemistically frame as "pricing power". It's a cornered market with every appearance of artificially high prices.
Monopoly rent extraction targeting a captive audience
Business customers don't pay for "having more buttons than mere consumers". Their business depending on such a crucial tool makes them reliant on swift and extensive support, negligible down times and so on. They are being cornered by ecosystem lock-in and workflow dependencies.
You could sell the same software at low prices to both "classes" of customers and charge extra for the services relevant for businesses. But then the fiction of "professional" whatever would be gone on the business side. And more importantly, companies like SketchUp would have to admit, they are merely SaaS with bad scaling prospects.
fsloth 4 hours ago [-]
”companies like SketchUp…”
I think you mean ”companies like Trimble and Autodesk”.
”You could sell the same software at low prices to both "classes" of customers”
Slightly disagree. How do you brand the product then? ”Design your next hobby woodworking project” and ”Design next award winning apartment complex” are promises that speak to totally different markets.
So I would argue it would not make sense if you care about the brand.
Thats why these open source alternatives are so great - they are actually targeting a market no-one is interested to serve commercially and can have hobbyist specific brand without having to voice cringy boasts about being a perfect clone of something for the users of the said clone.
Expansion of software patterns to adjacent underserved markets via open source is really nice.
Loquebantur 2 hours ago [-]
You engage in whataboutism.
The point here is, the actually largest share of consumers doesn't get served by the "market".
It's a systemic failure and people fail to grasp the exemplary nature of it.
voakbasda 8 hours ago [-]
The problem with this strategy is that anyone who used it during its consumer period find the second pivot as inexcusable. I would not trust a company to steward software upon which my business relies after they move to drop their products’ existing customer base in search of greener pastures.
stronglikedan 7 hours ago [-]
> The problem with this strategy is that anyone who used it during its consumer period find the second pivot as inexcusable.
Unfortunately, that's not a problem with the strategy. That was taken into account by the business when the pivot was made, and it was determined to be an insignificant amount of users. The strategy worked without any problems, given the increase in sales. Anyone who used it during the consumer period is no longer the intended customer.
fsloth 8 hours ago [-]
I would rephraze this that if you are not a product's intended target market chances of having nice product experience are suspect, and yes, it's a very reasonable choice to find some other product.
" in search of "
Trimble didn't need to search anything. They knew very well what the market is.
They could have a handled the consumer market retreat better.
PorciiVorbesc 6 hours ago [-]
>SketchUp was originally an offering for the AEC market.
I used it a lot since the first demo version came out on CDs of computer magazines of the era, and I wholeheartedly disagree with that.
The first release was definitely not feasible for professional use in the AEC market. It was just a easy to use 3D modelling tool for hobbyists and tinkerers who wanted to CAD or model something simple at home with the mouse, without the cost and steep learning curve of parametric pro tools like AutoCAD, 3DSMax or Maya.
It was the first 3D modelling tool I used that made 3d modelling simple and fun that a child could enjoy. But definitely not AEC levels, at least not the first release.
>Then Google bought it and essentially pivoted it to consumer offering.
When Google bought it, it was already popular in the consumer market. They bought it because they needed a simple tool that anyone could use to populate Google maps with 3D buildings of their neighbourhood via crowdsourcing, and nothing was simpler back then then sketchup.
>They retargeted the offering to professional engineering and architecture firms.
More correct would be they pivoted there because that's only userbase that will pay for SW. Unless you have the backing of someone like Google or Autodesk being you to finance such a loss-leader, then you're forced to move away from free SW for casual home users to paying pro-sumers to survive as a company.
twelvechairs 4 hours ago [-]
Architects would use it originally and to a small extent professional engineers because they wanted to communicate in 3d and tools at the time were so bad. It just couldnt do the drawings.
The big problem for architects is it never got to be able to do architectural drawings while other software came in with easier 3d modelling and visualisation (revit which autodesk bought and built out fully for industry use, rhino etc. Even blender).
Google did very little with it and Trimble too. Google i can understand ignoring it after Google earth stopped being a focus. Trimble I dont really understand what they want it for
Also it was notable for being a major commercial success for the ruby programming language, outside ruby on rails.
fsloth 4 hours ago [-]
”Trimble I dont really understand what they want it for”
Architecture and engineering.
It’s used in design workflows from apartmentns to nuclear power plants.
Sure, it’s not evolved much. But I guess that stability is a value for some users.
fsloth 5 hours ago [-]
”More correct would be they pivoted there because that's only userbase that will pay for SW”
Negative. This was part of Trimbles strategy in AEC software purchases. They bought Tekla for Tekla Structures on 2011 for example.
Ofc they want to make money but it’s not like they _first_ bought SketchUp randomly and then tried to find someone to pay for it.
japanuspus 16 hours ago [-]
Been through the same journey -- and have been super annoyed each year when I paid the several hundred dollar license for a tool I didn't use much, but really needed when I did use it.
My exit has been FreeCAD with an MCP-bridge, letting me talk to an agent that implements my requests directly in the FreeCAD file I have open. Works super well with FreeCAD which has all the features I need, but where they can be hard to discover.
anentropic 11 hours ago [-]
Out of curiosity, how do you tell the agent what you want - just describing it in words? Or draw something and scan it in?
khafra 10 hours ago [-]
I can't tell you how the parent commentor does it, but I saw an excellent reproduction of some historical building done by Astra in headless blender, using photographs, blueprints, and contemporary descriptions. Frontier models are smart enough to incorporate just about any kind of evidence you can give them about what you want, so give them everything.
abirch 9 hours ago [-]
I've used Claude (Opus) and Gemini (Argon) for Blender for some 3D printing. The work reasonably well, I'll have to try Astra next.
W0lfEagle 15 hours ago [-]
Same. Working in FreeCAD with agents is actually pretty awesome.
VBprogrammer 15 hours ago [-]
It's interesting, I have an interest in wood working, for a long time the defacto application for planning the majority of woodworking project was SketchUp. I can't say the last time I remember someone discussing it, probably because they are all using knock-off old versions and don't want to mention it.
taylorfinley 15 hours ago [-]
Sharing my own 'sketchup' specifically made for woodworking [0] for like-minded folks. Instead of extruding rectangles you use carpentry tools on dimensional lumber from a materials library.
You can set up a digital twin of your actual workshop, with config-driven models of table/band/miter/hand saws, router tables, drill press, chisels and rasps, all sorts of fun.
I recently added a jig system too, so you can build jigs that you can actually use, both in your shop and in the tool. It's all drivable by a human or an MCP, and has video and plan authoring and sharing built in.
I've been building this as a labor of love since getting laid off in April, it's all completely free to use and people have made some really damn cool plans with it.
I've been working on addressing a similar need, but for CNC machines and from a very different perspective and UI approach --- small, specialized tools which allow making one thing/doing one task:
This looks really cool, and the box you made looks great! I like the look of bamboo end grain, personally.
Sawdust has support for laser cutters [0] and has procedures for generating finger joint boxes, with svg export. It can also do more advanced forms via mcp, like this arcade cabinet design [1]. G-code is high on my to-do list because LightWorks isn't getting updates on linux any more would be great to send a design straight to my diode machine.
Sawdust also has support for 3d printing, both for including printable components in designs as well as printing scale models of your build [2]. No slicing built in, but full .step output for any slicer you like, with build volume aware scaling so you can print it on your machine.
Ages ago, a fine cabinet-maker I used to hang out with (he was my great aunt's bridge partner's husband) told me, "End grain is like a belly button, you know everyone has one, but unless it's someone special, you don't want to see it." --- agree that bamboo does change that calculus a bit, and I've always wondered what he would have thought of a bamboo box with cigarette and lighter and ash tray integrated into it which was made of short little segments of bamboo glued together (but never remembered to arrange for him to see it).
thatonehack 8 hours ago [-]
Awesome. I’m not into wood working as a hobby but I do have a bunch of different tools including a few different saws; I will be checking this out! Looks awesome and I love your attitude about it being a labor of love…… the best products out there that isn’t jammed down everyone’s throats!
JKCalhoun 10 hours ago [-]
Very cool. And I am working on some wood projects in the garage at the moment…
If you ever have/get any, a photo of any of the things on your site, actually built, would be a cool addition—make it real.
variodot 10 hours ago [-]
I work on a similar tool and felt the need to prove the engine myself, check out https://shopspec.io/gallery
I have a few projects that are in progress while continuing to build out ShopSpec. This is so that I can address any rough edges as they come up. For instance, material handling and safety during sheet breakdown.
taylorfinley 4 hours ago [-]
I'm building a tortoise table next weekend for our new sulcata hatchling, and am planning on making a youtube video showing the whole process, warts and all. Should be fun!
If you go to the plans section in the 'my shop' tab [0] and click a plan, e.g. this Pinocchio puppet [1] that someone requested via the public feature/plan request board, you can upload photos of anything you build and share a comment about how it went. Credit where due, I modeled this on Instructables and Thingiverse and similar project sharing sites. I for one cannot wait to see some pics of what people build!
Very cool! Some feedback on the landing page - discard if you don't care:
1) The sawhorse explosion feels physically inconsistent, in that all of the pieces go "straight" out at once, leading to things like the nails being dragged sideways through the wood. Since you have the build instructions, you could reverse it a few pieces at a time to fix that.
2) A couple places the AI-written text sneaks through in ways that are nonsensical or irrelevant to the reader - like "drag to orbit. The page scrolls past." on the 3d-print birdhouse diagram. The AI writing is more noticeable in the "learn woodworking" section but also seems to generally make more sense. (Personally, I'd be paranoid about potentially giving unsafe advice, but I assume you check it all before posting. I didn't notice any safety issues here.)
I've bookmarked it to try the tool out.
taylorfinley 4 hours ago [-]
Thank you, the feedback is much appreciated!
I'm embarrassed that "drag to orbit. The page scrolls past." has escaped my notice til now, yikes. I have an entire skill [0] devoted to eliminating this kind of ui slop, sadly it missed this one.
The landing page is very old at this point and definitely needs a freshen-up. The exploded view feature was built before fastener support, so you are absolutely right about that. Sawdust already can do much better, here's the same Sawhorse project's guide [1], take a look at the video there to see the fasteners driven in with respect for actual physical collisions :)
If you do try the product out you can easily submit feedback through the MCP or via the ui chat, I get paged for every feature request and love working on things users actually want. Would highly value any additional feedback.
Edit: the fix for the explode (and removing the slop ui label) is deploying now, thanks again for the feedback!
mft_ 14 hours ago [-]
I’ve just built a kitchen that was 100% modelled in Sketchup 2017 - the free forever (AAUI?) version still available in a few places. The Mac version I use occasionally crashes, but otherwise it’s perfectly usable.
WillAdams 12 hours ago [-]
Until you upgrade to a new Mac which doesn't have the Intel chip which that version is compiled for and which lacks Rosetta.
mft_ 2 hours ago [-]
I'm running it well on an M1. I thought all Apple Silicon Macs can download/use Rosetta currently?
WillAdams 48 minutes ago [-]
Rosetta is going away in the current or next Mac OS version:
It's one of the only software's I've felt genuinely zero remorse about pirating after purchasing a perpetual license and having them pull it out after. Adding to that the 3d warehouse forced backward-incompatibility on objects that you can trivially just open and export to an older version with zero differences to get it to work in most cases.
RobotToaster 12 hours ago [-]
I sometimes wonder what would've happened to SketchUp if google hadn't sold it.
lopis 12 hours ago [-]
Would probably have become an addition to Google Graveyard
RobotToaster 5 hours ago [-]
Sadly that may have been a better fate.
9 hours ago [-]
diego_moita 10 hours ago [-]
> I'm still peeved about the direction Trimble took
There is a name for it: "enshitification" is the strategy of maximizing short term gains, even if that results in product suicide on the long run.
criddell 8 hours ago [-]
Does enshitification cover setting the price to something sustainable?
Trimble bought Sketchup 14 years ago. In the world of commercial software, that's a decent run and they seem to be going strong. They just aren't interested in the nickel-and-dime world of selling to and dealing with consumers.
pmkary 15 hours ago [-]
In 2014 I was working on a building energy software for Iran's building regulations which differs significantly from the rest of the world. (most of the world rely on simulation on an engine called EnergyPlus, Iran used to have a static envelope-based heat transfer calculation as well). Back then the U.S National Renewable Energy Lab (NREL) used to have this plugin for SketchUp where you could model a building and the plugin would let you add energy details to the model.
I wanted to have that data model and SketchUp had this amazing cool thing where all plugins (which were written in Ruby) used to be running in the same space and so I could access the runtime of any other plugin. Here the team of OpenStudio were kind enough to share some of their internal APIs with me and so I could take their model and append my own data on top and let my users reuse all of their models from OpenStudio.
And then SketchUp web came and that whole plugin world was crashed. This forced me to go down the path of using a union-of-all-industry-file-formats (like TIFF for pictures) for building energy called gbXML, back then no one supported the file and so it was hopeless (and hard to implement) and eventually trying to create a CAD, which given the state of tech at 2016~2017 was hopeless too, and so it failed miserably.
Seeing this now, I was wondering how hard would it be to recreate that project with its own full CAD from KetchUp...
aembleton 14 hours ago [-]
> Seeing this now, I was wondering how hard would it be to recreate that project with its own full CAD from KetchUp...
Would be worth trying with AI to help you develop it. Probably not too hard if you could spend a few weekends on it.
fsloth 8 hours ago [-]
" Probably not too hard .. spend a few weekends"
It's not too hard but I can tell "few" means "few hundred" at least.
nico 23 hours ago [-]
This is awesome. At the same time I wonder if we are entering an era in which it’s just better to build your own tools
Need a 3D modeler? Use Claude or codex to build the 3D model first, then build the modeler around it as you iterate towards what you want to build
So many times I’ve had to spend hours learning a software, reading docs, doing tutorials, only to later be able to do a small basic idea of what I actually wanted to do in the first place
Now I can just build only the software I need, for the thing I need it for (or, ala REA, maybe either modify or reverse engineer some existing software, to better fit my workflows)
In that sense, it would also be cool to be able to drive KetchUp with an agent. I’m a big fan of what pen.dev has done (although now their tool is not free/open anymore)
Aurornis 22 hours ago [-]
I have several personal tools I’ve built and maintained.
It’s always more work than it looks. You can get to mostly usable tools quickly if your request isn’t too difficult, but every time I use them I run into another thing that needs to be fixed, changed, clarified, or cleaned up.
It’s fun at first, but over time you realize you’re spending more time building the tools than using the tools to build the thing you originally wanted to work on.
I think personal software is best for highly specific tools or for things you need to use repeatedly for a very long time.
For something like a SketchUp competitor, I wouldn’t want to try to maintain my own. The number of edge cases and features is bigger than you imagine up front, and the vibecoded version that might appear a few hours after your prompt is going to be very far from the version you want to use.
Gigachad 21 hours ago [-]
I’ve been using the vibecoded Lightroom clone and it’s 5000 commits and around 20 releases in yet still has countless issues and missing features. I feel like it will get there eventually but the idea that you should just vibe up your own Lightroom rather than using an existing one is absurd with today’s tech. It’s still a massive amount of work even if it is way faster than ever before.
The vibecoded apps all rely on an army of users exploring the program and reporting all the bugs and edge cases. You don’t want to be doing that on your own.
4ggr0 14 hours ago [-]
this is something which is starting to annoy me, and the effects can be seen on HN. someone posting a post or comment talking about their new vibecoded project, then a couple other people chime in, all with their own, separate, vibecoded projects, trying to reach the same goal.
it seems like the open-source space is losing some aspects of collaboration. people used to search for already existing projects on some git-hosters, maybe fork them or put effort into the existing thing. now half a dozen people each run their own race using AI, each solution misses 30% of the features and has x-amounts of bugs.
work together ffs :( i mean, 20 people can use AI to vibecode the same project, why does everyone now feel the need to do everything solo. (because they want to be the prolific founder and cool dude who did the most in the project* instead of being one of many contributors, i assume...)
*or rather, their AI.
Gigachad 14 hours ago [-]
I've observed the same, so many forks of forks of forks happening right now rather than group efforts. But I think this is coming from the fact that PR review is incredibly costly to the maintainer while generating code is fast. I've submitted PRs to open source projects which claim to welcome contributions but the PRs just never get reviewed (this was an issue pre AI too). If you wanted to build something, constantly fighting an uphill battle to get PRs merged is less appealing than just creating your own project where you can work full speed on whatever you want.
I'm not too sure what the future here looks like and if this is even a problem, but things certainly are changing. These days submitting a well written bug report or feature suggestion, or validating existing reports is more useful than sending a PR.
4ggr0 13 hours ago [-]
maybe the focus should shift a bit to using AI for PR reviews. adversarial reviews, summarizing it for a human. we'll see where time takes us...
Gigachad 11 hours ago [-]
What is the point though. The maintainer can generate the same code you can. And they can iterate on it quickly without having to go through a review process.
forestry 10 hours ago [-]
Day jobs and lives away from the PC.
utopiah 10 hours ago [-]
> work together ffs
Right, and that's something I already argued for since the very beginning of GenAI, namely that suddenly being very costly, economically and ecologically, is not that bad if you build upon what exists and collaborate on progress.
... except that's exactly WHY vibecoded project exist : they do NOT have to give a shit with existing codebase and agree on weird rules. They just start from "scratch". In fact I'm convinced there is the Ikea effect at play, namely that the PM of the vibe coded project is emotionally attached, even proud, of the result they think they provided somehow, despite them not actually doing it. Obviously for-profit GenAI companies have a big incentive to maximize token use and will never promote workflows and tool that minimize the one thing they monetize.
So even though I agree in theory, in practice I think there is little to no overlap.
Rapzid 14 hours ago [-]
I believe the goal of most of these people is to make money.
andybak 13 hours ago [-]
Weird that they are all giving the source code away for free and failing on the whole to even ask for donations.
Rapzid 3 hours ago [-]
All?
ArtCraft is definitely a play at selling subscriptions.
The slew of "oh yeah, blah blah blah blah blah, that's why I built [X]." comments on every AI post now?
Click through and many you'll find "Pricing".
It's a gold rush.
andybak 27 minutes ago [-]
1. Go to any ArtCraft repo on GitHub
2. Click "Fork"
3. Profit
egeozcan 15 hours ago [-]
Why do you even maintain it if it's for you?
I have lot of tools that work for me.
When externalities break them, it's usually one prompt to fix them.
maitrungduc 9 hours ago [-]
and when it works on your machine, but fail when exporting on an old macos 14, you only realize when someone tell you...
IshKebab 16 hours ago [-]
My experience has been the opposite. I have two vibe coded tools that I use frequently and I spent a few weeks of occasional prompting and now they are basically done. I haven't touched the code for one of them for a month and I use it most days.
To be fair neither is quite as complex as a whole 3D modeller.
kristophph 15 hours ago [-]
I think this is the key:
> To be fair neither is quite as complex as a whole 3D modeller.
Once you cross a certain complexity in your project, it becomes quite work intensive even using AI to do the coding. All the back and forth and cleaning up, exploring ideas and starting from scratch again takes time and effort and maybe most importantly, extensive human thinking. Just to clarify, I am not against AI and use it extensively myself. But vieb-coding all your tools is just not really a feasable idea for most (time/token-cost constraint) if you tools are of a certain complexity.
IshKebab 13 hours ago [-]
Actually I think "complexity" is probably not quite the right measure. You can vibe code very complex things, as long as they are in the "easy for AI" zone - clear goals, ideally non-graphically, ideally there's a specification or even better a reference implementation, best an automated way of checking correctness. CAD is about the worst case (and very complicated).
But you could easily vibe code, e.g. an H.265 decoder, even though it is super complex.
altmanaltman 22 hours ago [-]
Isn't the whole argument that ai-driven code will reduce the time spent building because AI can do it now? And a SketchUp competitor by defintion will not be personal software just like SketchUp is not.
vips7L 22 hours ago [-]
AI can’t do it. You still have to hold its hand. You still have to sit there and ask it to do things. There is no fire and forget.
altmanaltman 22 hours ago [-]
For personal projects that don't matter? Sorry but that's your opinion at this point. I respect it but also fully disagree. Also even in that case, isn't it still reducing time and problems we have with pre-AI personal projects like maintainability and feature creep like you mentioned.
arlcode 17 hours ago [-]
For my personal projects I think it does matter even more. I've got some perculiar and weird quirks that may even go against best practices (that's what makes them personal projects).
AI can get the implementation done just fine but in my experience it needs guidance because it can't guess what features and details I want / don't want.
The difference to "before" is that it's now possible at all for me to get personal projects to a working state but in sum they still require a few hours every now and then supervising.
80fj0348i 21 hours ago [-]
> For personal projects that don't matter?
My personal projects matter a lot. For some of them I am trusting my private data in the code, so I need to hold it to a high security and privacy standard. For others, I have hundreds or thousands of users (game mods, helper tools for hobbies, etc.), and I don't want to ship low quality to them.
vips7L 22 hours ago [-]
It’s just my opinion that you have to ask the LLM to do things? Seems like reality to me tbh.
Aurornis 20 hours ago [-]
> For personal projects that don't matter?
If it doesn’t matter then do what you want. By definition, it doesn’t matter.
Some of us actually use these tools, so it does matter. You do have to consider things when it matters.
Aurornis 21 hours ago [-]
I was talking about AI developed tooling.
It’s amazing that you can get something usable quickly, but then every time you use the software you’re back in the AI working on your tool.
Using AI to build the tool becomes the project, replacing the project that led you to think about vibecoding a tool.
Teever 16 hours ago [-]
I'm coming to think that a hybrid of vibe-coding but with an existing substrate is the right way to go.
For instance in the context of this SketchUp clone that might mean using AI to generate an addon for Blender.
For some of my recent projects I'm using the community edition of Odoo to serve as that substrate because it already has an established community of modules to draw from, ten years or so of real world use, and I don't have to burn tokens generating a user account system with permissions or the data models / ORM to interface with them.
This seems to be serving me well, but we'll see. It could be that I end up in the same the tool becoming the project scenario that you describe, just a little bit later than if I had started from a blank slate.
socalgal2 21 hours ago [-]
My experience is it's short to get a working tool and long to finish. Many 3D modulers (blender for example) have literally 1000s of features. Yes, you might not need all of them but you probably need 100s, each with quirks and UI/UX decisions. Just iterating on camera controls can take days. You get an orbit camera working via a prompt. But then find it only zooms in/out so far because the LLM decided that was ideal. So you prompt to fix that. Then you find moving the camera is not zoom aware so it moves in meters per frame and when you're zoomed in things move too faster. So you prompt to fix that. Repeat 10000x. Every feature, extrude, bevel, chamfer, subdivide, select vertices, select faces, select edges, etc etc all require this handholding and finese.
At least that's my experience.
Small tools with few features is great. a 3D modeler is not that.
leoedin 9 hours ago [-]
LLMs also seem prone to making weird decisions that no human would have done. You can very quickly have a broad implementation - often with features that are great and that you didn't even know you wanted - but when you dig into the features, the weirdness reveals itself bit by bit. If you were developing by hand you would have used a sensible range for zoom from the start, but the LLM doesn't seem to work that way.
I suspect people saying "everyone will vibe code all their software" have never actually tried to vibe code a meaningfully complex application.
samiv 13 hours ago [-]
Yep having the same experience right now with some curve/spline tessellation into triangle strips. There are many degenerate edge cases and when you ask the "ai" to handle one it breaks something else etc. Makes me think that if I had just hand coded it myself I would intimately know the code and get it working faster than prompting. Gah.
throwup238 7 hours ago [-]
> Yep having the same experience right now with some curve/spline tessellation into triangle strips. There are many degenerate edge cases and when you ask the "ai" to handle one it breaks something else etc. Makes me think that if I had just hand coded it myself I would intimately know the code and get it working faster than prompting. Gah.
It gets so much worse. At some point you start hitting edge cases where there is no clear right answer and you have to come up with a kernel wide coherent approach to things like 1ulp differences. Parasolid is full of decades of kludges on top of kludges and giant switch statements trying to preserve backwards compatibility as they explored the space.
heyzk 23 hours ago [-]
I really appreciate this sentiment, and at the same time there are real opportunity costs to building and maintaining your own tools. Even in the era of tokens.
I think it comes down to where you want to spend your time and attention.
I definitely find it easier to create the subset of blender that I need than learning blender, even if it is throwaway.
mastermage 16 hours ago [-]
Interesting perspective for me I find it more interesting to learn the transferable skill set behind using this kind of software. Sure I can make a software that does exactly what I want when poked but then I do not learn any transferable skills.
wilted-iris 22 hours ago [-]
To offer a counterpoint, I’m currently spending a lot of time on the other side of this: having my agent use blender in the background and build on top of it. My agent can drive it well and I’ve benefited from its import capability, renderers, layering, and simulation. I get a lot from blender, even though I don’t ever see its UI.
13 hours ago [-]
21 hours ago [-]
t-writescode 22 hours ago [-]
I ... really..? What sort of stuff are you wanting out of Blender? If it's 3d printing, you probably literally just need to do the soda can or donut tutorial - not even the whole donut tutorial.
Like, I have to imagine that 80% of the "quick stuff" people need to do in Blender would be done with:
* create quad
* exude like 5 times
* done
Maaaybe
* perform boolean operation between object and a circle or 3
---
Like, I have less than 200 hours in Blender and I was pretty functional after like ... 30 minutes to an hour of a single tutorial.
1123581321 21 hours ago [-]
I don’t see how this could be faster (and not buggier) than generating a script to make what you want in Blender for you.
You’ve likely just switched hobbies to your Blender clone instead of what you wanted to make.
heyzk 22 hours ago [-]
And a fantastic learning experience
bhouston 23 hours ago [-]
> At the same time I wonder if we are entering an era in which it’s just better to build your own tools
I wonder if Claude Code and other similar tools start to maintain their own base apps that everyone keeps asking for? This would be like a "cache" that makes the requests cheaper/easier to deliver?
Then everyone's "own" apps is really just a customization layer on top of a solid base that already does all the heavy lifting.
So many people are building their own versions of:
- PDF editors
- Word processors
- 3D editors
- Video editors
- Pixel-based image editors
- Vector-based image editors
- CAD editors
etc.
In some of these there are a few flavors as well that are major architectural choices. But most of what people are doing is just playing around with UI and file formats. Much of the functionality is generally the same.
Gigachad 18 hours ago [-]
We have existing tools for this. You fork the repo of the closest thing and have AI tune it to exactly what you want.
socalgal2 20 hours ago [-]
I guess I'm too used to the collaboration features of word processors like Google Docs. If I build my own, those go out the window otherwise everyone has to sign up for my word processor and learn my word processor's UX.
For the rest, at least so far, I haven't had the collaboration experience but for word processors I can't go back to local files that can't easily be edited/viewed/commented on by multiple people or myself on every device I own.
dirkc 17 hours ago [-]
I'm pretty sure this already happens to some extend. Look at the default website output you get from the different models, there tends to be a 'default' look/approach it takes.
I think they all have been post-trained on specific tasks like building sites, apps, etc and that is a large part of the reason they are pretty good at it.
mastermage 16 hours ago [-]
Honestly i think that now is the chance for established software to go with the flow and I think the major way to go with the flow is create are adapt software to be extensible (so that everyone can let claude or opencode or whatever add their functionality of choice without having to rebuilt the wheel from scratch everytime) and secondly direct mcp access to the software itself.
ripplarity 19 hours ago [-]
yes but tons of new features introduced making things interesting
guhidalg 23 hours ago [-]
Sigh... are we really just going to re-invent plugins?
bhouston 22 hours ago [-]
We do need a safe layer for non-engineers to customize apps.
All these clones of Photoshop and similar tools are mostly dead ends, they are experiments to see if you can do it, but there will be a consolidation around regularly maintained and powerful and complete apps that likely also offer support.
These experiments are dead-ends for the most part long-term. Fun play things, but they will not gain significantly user bases.
t-writescode 22 hours ago [-]
Isn't this more of an argument to use common OSS choices? Like, encourage people to use Blender for 3d modelling; LibreOffice for documents; Gimp, Krita, or one of the many other art programs for ... art; OBS for simple video stuff; and so on and so forth?
stonogo 19 hours ago [-]
> We do need a safe layer for non-engineers to customize apps.
Why? Every time someone cripples software in the name of safety it just gets replaced with more capable software that isn't.
Gigachad 18 hours ago [-]
Plugins have long been troubled. It’s hard to make a plugin interface that is actually useful while also not breaking on every app update. And plugins have been such a massive vector for distributing malware.
wccrawford 22 hours ago [-]
I think that could be true once you're experienced enough to know what you want.
Before that, though, I think you have to learn what the experts do and what can be done and what you should want. And then you can start to make your own opinions.
By then, unless the major players have done it really wrong, you're not going to make something better than what already exists.
Right now, it's different, because we have that experience prior to the ability to just wish up an app and have it happen.
I think the value of adapting full tools to your own workflow far outweighs the maintenance costs nowadays. The tool above for instance took my agent a couple hours and it was good to go. It wasn’t a small slice by any means either: it supports edit/obj mode, extrude, loop cuts, modifiers like subdiv, save/load, undo/redo, bones, auto rig, posing, and more.
Mind you, this is a slice of an open source tool which was still outside of what I wanted from it. Cost wasn’t even part of the equation here.
nico 21 hours ago [-]
Very cool. I’m in the process of creating a video game and building all the tooling around it, both for me and the players. And also, for doing things like recording a quick sequence of gameplay, “sandwiching” it with an intro/otro and posting to YouTube, then ingesting the engagement data to show on the internal player… still early days, not enough data to start optimizing on it, but I’m really enjoying building exactly what I want as I go along
You mention that you have no plans to share. Any chance I could persuade you otherwise? I maintain Open Brush and Open Blocks and I'm very interested in getting something more complex working for 3d modelling and actually shipping it to the various app stores. I've got a few different approaches in mind but using Blender as a base it quite appealing.
I could start from scratch obviously - but seems a shame.
dyarosla 10 hours ago [-]
This really isn’t Blender as a base- more like uses Blender as a base reference.
I think both those apps are great and happy to know folks like you are maintaining them; this one though is simply more of a personal software than a path to an OSS project!
t-writescode 22 hours ago [-]
That's a pretty neat looking program.
Since Blender is OSS, you could probably use one of the existing VR views for Blender and write a bunch of in-app Python scripts to do what you want, if you wanted to also be able to output to Blender files and work within Blender already.
This looks like a pretty well-trodden area in Blender.
dyarosla 22 hours ago [-]
Yes I could fiddle and adjust - But see, i dont want that. I want a modeler that works how i want, some things that work like Blender and others that work a different way.
Adding scripts into an existing product does not serve this.
I think we are entering into a new era where we don’t need to accept fully featured software with all of its decisions should they not fit ours.
t-writescode 22 hours ago [-]
You might be surprised just how much you can rewrite of Blender's UI if you want to.
And you get battle-tested software and a huge collection of existing plugins, operational modes, export formats, all that. Fully bespoke software for everything you want is a way toward losing interoperability between people and platforms.
dyarosla 21 hours ago [-]
It doesnt have to be one or the other: you can have custom software that still works within existing/interoperable formats, as I’m doing here (im not planning to reinvent OBJ/FBX/BLEND)
But building up as opposed to building down or rewriting is freeing.
I can pull in ideas from Houdini and mix them in. I can use Blender’s source as reference for specific implementations should I need. I can keep my tooling specific to the projects Im working on and design it to fit that as opposed to build a workflow within a software meant to be general purpose.
IMO, this kind of full customization is going to become far more commonplace.
solarkraft 18 hours ago [-]
With strong building blocks, perhaps. But I don’t want to maintain millions of lines of AI generated CAD kernel (or pay the token and wait time). I’m also happy not having to make design choices in that space considering I have 0 experience with it. It’s still a liability.
geraneum 21 hours ago [-]
> I wonder if we are entering an era in which it’s just better to build your own tools
Not necessarily. But only if what you need is not available in open source or not achievable by forking open source. Otherwise it’s a waste of money (token) and attention. This is a significantly smaller pool that “build [all] your own tools”.
Although, AI mega corpos would love if we do constantly re-build everything. You can tell by the tendency of the agents to write everything from scratch instead of importing the code that does the same thing or better.
utopiah 9 hours ago [-]
Typically what takes time isn't learning a shortcut here or there, rather it's to understand how a proficient user thinks and THUS why the tool work in a "weird" way for newcomers. It's easy to click and paint, why would pro user layers? It's so complicated... then you watch how pro do it, and it's because behind the complicated steps comes a lot of power.
I have no doubt as a newcomer you can make software that will do what you want, but most likely the outcome will no be good, not because of yourself or the tool, but rather because the expertise to map the problem space to the solution space is not trivial, and that's precisely what requires a new way to think, and thus tools that aren't so easy to learn. It's not poor UX.
"So many times I’ve had to spend hours learning a software, reading docs, doing tutorials, only to later be able to do a small basic idea of what I actually wanted to do in the first place
Now I can just build only the software I need, for the thing I need it for"
Or another way to put it for devs “A language that doesn't affect the way you think about programming is not worth knowing.”
― Alan J. Perlis
__mp 5 hours ago [-]
I also have my own tool. I vibed a browser-based OpenSCAD harness. It’s a simple SPA with a chat loop. It stores everything in the user’s browser and requires the users to bring their own tokens (URL with API Key or OpenRouter with PKCE)
I host it here: https://vibe3d.spoerri.dev - my partner enjoys it because she can bring an idea to print pretty fast.
Maybe a similar idea could be interesting for KetchUp as well.
pettijohn 19 hours ago [-]
I've been using SolidCraft with MCP and Codex to model things. Pretty wild. I gave it a photo of a pencil sketch and a paragraph; a few minutes later I had a 3D model. https://solidcraft.io/mcp/
canterburry 7 hours ago [-]
This was only possible because a well thought out, working example could be used as reinforcement learning for the LLM to benchmark it's own development against and improve its output. Had this not already exists in it's current form, I doubt the LLM could have built this on its own.
So, I guess yes, if there is someone's hard work you can copy and steal and make your own, I guess that's true. But try to create something that no one has ever built yet.
cael450 21 hours ago [-]
I have always been skeptical of the reality of this, especially for businesses. But then one person on our CSM team spent a week building an alternative to Gainsight, saving the company like $90k a year. Granted, it required a technical person to do it even with vibe coding, but it doesn’t take much of his time to maintain it.
For an example. Lots of people are now vibe coding their tools.
baby 21 hours ago [-]
Looking at CRM tools these days and they are so complicated that I’m thinking of just using my tool and add features as I need them.
Would love to hear advice from people who use CRM tools on the job.
revolvingthrow 14 hours ago [-]
Better as in you'll be happier with the end product after a dozen iterations and two dozen small tweaks? Sure.
I think people underestimate what a time investment building your own tool that's of at least medium complexity is, though. It might be worth it if it's something you'll use regularly for a long time, but I don't foresee it being the norm. Even assuming no bugs ever or no performance issues ever and all the architectural decisions made were appropriate (pretty far from current reality), merely describing what you want and how it should behave and look like is a long process, and one that isn't all that enjoyable once you go through it once or twice already.
The existing software that has been around for several years and you kinda dislike is probably much more reliable than your own homebrew, too. Doesn't matter in some cases, matters hugely in others. Learning that your custom video gallery organizer just vanished half your family videos probably isn't very enjoyable, but at least it's easy to recognize it happening. Something more abstract like CAD or 3d printing slicer, ehhh...
22 hours ago [-]
0xb0565e486 23 hours ago [-]
Skip the 3D modeler and have the agent do the operations directly instead.
A good visualizer might be useful thought! If that part of the work needs to be validated.
dyarosla 21 hours ago [-]
Ask any 3D creative and you’ll know that AI is not yet sufficient to generate 3D models at the quality that most industries need them at.
superpapotas 18 hours ago [-]
you can try fabivo, its not for all uses, but i think the ai agent is great.
burgerone 17 hours ago [-]
Why are people suddenly entirely unwilling to invest effort into their projects...
Wat is the point of personal projects if in the end you learned nothing
gamblor956 21 hours ago [-]
Once upon a time, not so long ago, people gained skills by learning about things instead of outsourcing all the hard parts.
By outsourcing the modeling to Claude or whatever, you don't learn how to build 3d models on your own. You've become dependent on the LLMs, and you're trivially replaceable by someone who can build 3d models on their own...because they can also learn to tell Claude to do all the work.
t-writescode 21 hours ago [-]
I mean, there is value to artisan work (paying for it), and there’s value in quick stuff; but I’m not sure LLMs can do either in the case of 3d art.
Their models are horribly topoligized, which will matter for most use cases for models; and if you’re just using it to make a cube, it’d be better to learn the related program to do that.
People tend to ask LLMs to make full art pieces (a decade of specialized knowledge and muscle memory), not “draw me a square”.
In 3d art, there is a gulf between the easy stuff and the hard stuff.
Tl; dr: for 3d art, I agree.
secretsatan 20 hours ago [-]
This just sounds so horribly inefficient though i’m sure the makers of LLMs love you.
sergio 21 hours ago [-]
This is so awesome and something I definitely need.
I'm a very casual user of SketchUp (I use it once every few months), but I have a single file which is a detailed model of my house, which I have invested many hours in creating, and that only works on the full Desktop app. So the free web app is not an option for me. And a subscription doesn't make financial sense so I have to ask for a new trial every time I have a home renovation project and I need to modify my model. I'm aware that eventually they might refuse a new trial (they haven't yet) and that day I will just stop using it and learn Blender.
However, I'm now realizing that at least for home renovation projects, an essential feature is the SketchUp 3D Warehouse, where you can download user shared models of a TV, a toaster, a chair, etc. Specially for a casual user, you don't need to have the exact model of your real chair, just a chair will do. So the warehouse is an essential part of the value.
I'm sure there are alternatives for free models other than SketchUp 3D Warehouse, but having it directly in SketchUp is super convenient.
jphorism 19 hours ago [-]
+1 to the 3D Warehouse. I have used it extensively over the years for trialing furniture; IKEA seems to have most of their models designed to scale there.
jkl5xx 21 hours ago [-]
Just have Claude generate assets like that with blender or something. It gets you good enough 3d models pretty quick
elxr 24 hours ago [-]
Tell your agent to add some showcases of the features on the landing page.
Impressive how good that rust UI feels in a browser though. Even the text rendering is impressive. Maybe rust really is the future of cross platform web+mobile apps.
mastermage 16 hours ago [-]
I am generally happy to pay for my software that I use. However I do not have the money to shell out for some of the professional grade software just for a hobby. And I also do not enjoy paying hundreds in subscription fees acumulated for things i use once a month.
So these AI Clones popping up are realy helping me, as I do not need the full featureset of the original usualy and these AI Clones are close enough for my hobby applications. Additionaly these tools all run on linux which makes my move to Linux etch closer into the present. Some of the tools were a reason I was not switching.
I will still continue to pay for the things i truly get value out of and are in my opinion reasonably priced. To support the creators.
lifeisloving 16 hours ago [-]
Are you using them? They all seem to be broken and UI (so they can post on HN/twitter) is the only thing that matters.
Seems unlikely that people are going to spend their Claude sub every month maintaining the thing either. They'll move onto the next cheap LLM project and dedicate their tokens to that.
Im sure that's not every project, but Id rather just have something reliable with a reputation. AI clones are killing open source's reputation for me, I just dont trust anything now.
Not to mention how easy it would be to get a malicious PR approved into a public repo who's maintainer relies solely on the competency of a cheap llm.
mastermage 16 hours ago [-]
there are some I actualy use semi regularly these are mainly GeoLibre and Open CAD Studio. Completely fine for all I need with it. I am currently trying SolidCraft but still deliberating.
ArtCraft is also pretty cool but I have my Affinity Photo 2 Licenses so I do not realy need these currently.
pseudony 16 hours ago [-]
So, no money for the modeling tool but enough for a range of dewault/milwaukee/makita tools and the raw resources (wood, screws, glue) etc ? :)
If it was crocheting, I would get it, but woodworking is not an inexpensive hobby where the software becomes the top line item.
mastermage 16 hours ago [-]
I am not doing woodworking though? Privately I do 3D Printing.
This here I would also use at my job whera I am a bit of a Jack of all trades so once a quarter I need to sketch some model of something to see if it would fit or not. Example would be right now i need to sketch a simple collision protection for a outdoor cable tray (for Solar Pannel Strings). Gonna not take me longer than an hour. I do not use it enough to justify spending 40 Bucks a Month 4 times a year.
pseudony 15 hours ago [-]
Ah, cool. Never mind then :)
Its just that often times, even us programmers seem reticent to pay for software even though we happily throw dollars across the counter for a sugary coffee somewhere.
I know of sketch up primarily from DYI magazines, hence the assumption.
mikae1 13 hours ago [-]
What is the name of this? It says both KetchUp and BenchUp and they're used randomly?
EDIT: latest release switched to BenchUp fro KetchUp. Let me assume it's and attempt to duck infringement.
beeforpork 7 hours ago [-]
'duck infringement' -- parse error at first, because I read this like a noun phrase, like 'duck typing': 'if it looks like a duck and behaves like a duck, it is a duck'. It is kinda fitting for infringement, too... :-)
jitl 22 hours ago [-]
I absolutely love sketchup, and seeing it passed between various developers and never improved while gradually getting less and less free was very sad. it was awesome when owned by google and totally free. super happy to see a spiritual successor!
twelvechairs 21 hours ago [-]
Classic case of high value to end users but low revenue, because ease of use doesnt bring money.
Gained a huge user base quickly then sold to Google who wanted it apparently to help people toy around in Google earth, then they did little with it and eventually sold to Trimble, mostly a surveying company, who similarly had hopes it would help expand their users and market reach, which it hasnt really its mostly still evades commercial revenues and continues to be used by the same kind of users as originally
i directed pi to their documentation + github repo and created a pretty awesome skill that allows me to take measurements on paper, snap a pic, and then have it generate the drawing. can also snap a general pic and then provide a model number of an item in the pic, and it will maintain scale. this stuff used to take me an hour+!
dizhn 13 hours ago [-]
Skill public?
BHSPitMonkey 18 hours ago [-]
To this day I still use the 2017 freeware version (through WINE) for drafting all my home projects. It's just so darned intuitive to me (for anything without too many curves, at least) and I can't be bothered to learn anything else. The web version works in a pinch but it has always felt slower and less capable to me.
Going to have to try this!
0x0203 5 hours ago [-]
Would you be willing to share how you got it to run in wine? I've tried many times with the old 2017 installer I had from back then, and have never had any luck. I've gotten it to run once or twice, but it's never been actually usable. I've kinda given up, so maybe newer versions of wine work better now?
redm 11 hours ago [-]
Same for https://hew3d.com/, I'm not sure which is better but both are open source.
egeozcan 15 hours ago [-]
Hey, I already have an open source drawing tool called ketchup :)
I guess name collisions will become way more common in an era where people slop their own tools. We just don't have enough words on this planet.
Problem existed even before AI, with framework names. Headlines like "Planecrash in Ohio gains 50% in one year", "ShowHN: I use car accident to get new customers", "Made my wife happy with broken arm"
itomato 8 hours ago [-]
clankers should have enough sense to look for collisions
yboris 24 hours ago [-]
Great work!
Feedback: every time I draw a rectangle it asks me to name the object; seems like the wrong default behavior.
pippy360 24 hours ago [-]
>every time I draw a rectangle it asks me to name the object;
yeah this seems insane. The push/pull also doesn't work the way I expect it to. Sketchups main selling point was how easy it was to use and pick up
LinguaBrowse 20 hours ago [-]
Yeah, this falls well short of Sketchup's intuitiveness.
Every interaction I try on KetchUp has some kind of gotcha. It doesn't snap to grid by default, and as you say, it's insane that I have to name every rectangle. I can't extrude a face with a single click-and-drag of the push/pull tool; instead I've got to select it first with the arrow tool (which strangely does not match how the Move tool behaves), then extrude. Also, clicking and roving through the File/Edit/View/Tools menubar doesn't interact as expected.
Sorry to say, this ain't SketchUp. It seems like yet another flashy vibe-code that has fulfilled an 80% quality level but crumbled on the final 20% of polish that really makes or breaks the user experience. Really feels like the author didn't test-drive it for usability.
cfeduke 6 hours ago [-]
I just ran across Sawhorse[1] which is an open-source SketchUp alternative built with a focus on woodworkers earlier today. I haven't used it yet, but likely will.
I love it. I do 3D printing as a hobby, but I really dislike the design phase. I usually need simple shapes: boxes or enclosures for electronics. I can find most of what I need on Thingiverse and similar platforms, but there’s always that 20% of things I need to design myself to adapt them to my specific needs.
SketchUp 8 was the easiest software for me to learn, and it just clicks with me. I really wanted to learn FreeCAD, but there's something about the way it works that I just can't get along with, so I keep going back to SketchUp.
Later, I tried all the alternatives: Onshape, Fusion, Tinkercad, and the new SolidCraft, but SketchUp still feels so simple and intuitive to me.
I've had better luck with OpenSCAD for simple things, and I still need to try it out with LLMs. That's on my TODO list!
hasley 17 hours ago [-]
I have the same problem with FreeCAD: I really want to uae FreeCAD. But after some weeks or months I cannot even remember simple things like how to add a new object (or hole) to a face of an object.
dceddia 20 hours ago [-]
I've had some good luck with using Claude (Opus 5.5) for making models for 3D printing just by describing the thing I want (with measurements if you've got em) and having it create an OpenSCAD file.
It's also weirdly good at taking an existing STL file of a functional print and writing some python to modify it.
One example was a badge holder that I liked the design of, but only held 2 badges. I wanted 3, plus a slot on the back for a key. Gave it measurements of the key, it grafted a key slot on the back, and added the third badge slot. In typical 3D printing fashion, of course I had to print like 5 iterations of the thing until the sizing of everything actually worked.
oliyoung 20 hours ago [-]
Love love this. The SketchUp workflow has always clicked for me, but its increasingly difficult to access it as it's ownership has changes and they've monetise it more
This is great for small scale STL design too
MojzisekJan 6 hours ago [-]
I’m working on something similar called Rysmo, 3D modeler focused on architecture. I’d be curious to hear what you think!
The link says KetchUp but the actual website says "BenchUp" I'm confused about the naming situation.
xiange 20 hours ago [-]
I tried it out, and it still differs from SketchUp’s habits/workflow. If the goal is to replace SketchUp, would it be better to keep the habits consistent with SketchUp?
wow this is a really well-built application, it even comes with its own MCP server and LLM integrations, nice work.
suddenlybananas 15 hours ago [-]
I appreciate the website telling me that Mac, Windows and Linux is a list with 3 elements.
tannhaeuser 15 hours ago [-]
Not to mention your interaction with the app proceeds through the stage of Draw - Complete - Deliver.
andyferris 21 hours ago [-]
I actually vibe-coded (kinda vibe, mostly hand holding) a "2D Sketchup" for myself for a large yard renovation/construction project we've been doing for about 12 months. It was mostly successful - as others noted, you spend a lot of time investing in the tool rather than drawing your project.
janice1999 23 hours ago [-]
It would benefit from curated examples in the File menu to demonstrate the features and give first time users something to interact with.
sakopov 16 hours ago [-]
I'm about 40 minutes in trying to figure out how to scale a rectangle to a certain dimension. I never used SketchUp. Is it also this unintuitive?
tjelen 13 hours ago [-]
Just start typing numbers, eg. typing 5 will focus the text input in lower right corner and then you can finish by typing "5m,3.50m" - that will set the sides of the rectangle to specific exact values. Almost all tools support a similar kind of direct numerical input this way..
spectaclepiece 16 hours ago [-]
Really loving these new software tools often written in Rust.
Been using Solidcraft which is another great example. Lower memory usage and better performance than Fusion 360 and improving quickly.
The power agentic development gives to a single person with love for the craft of software development creates better products than a large team with corporate steering.
This was always the case but types of software like advanced CAD modelling was outside the scope of what a single developer could accomplish. Now it isn't and it's a great thing that will create a lot of high quality new tools for humanity.
sorobunfon 5 hours ago [-]
couple things I don't immediately get from the UI:
1) how do I make a thing exactly N units long? like, a 10cm long stick
2) can I just put dovetails in random places, and have them make the object split in two? Would be wonderful for 3d printing
reacharavindh 16 hours ago [-]
One more item on my todo list likely got a booster.. instead of building it from scratch I might build mine on top of this.
Seriously, the subscription frontier models scratch my control freak nerves and lets me build my computing set up just the way I want and not a single bit more. I’m currently making a minimal Linux distribution for myself based on Gentoo with the idea to develop any and all graphical applications except Firefox _for_ myself using LLMs. Nothing on my system I don’t need.. It sounds absurd but is still possible in 2026 with these subsidised LLMs. They are not my private data, so I don’t mind having the agents build it, test it, and automate it. If they want to train on my calculator app and CAD built upon other code from the Internet they are already trained on, they are welcome to :-)
This right here is a new age of personal computing.
Xeoncross 21 hours ago [-]
Small feature request: I can't figure out how to use it with a trackpad. Seems to be mouse-only. I use sketchup with a trackpad plus one of the different modifier keys for pan, zoom and orbit.
I can't acknowledge why this OSS is so fucused. There are models like this in HuggingFace.
m1keil 23 hours ago [-]
Is KetchUp free?
Yes. KetchUp is open source under the MIT license and written in Rust.
Omg why is rust relevant to the question whether this is free or not.
gpm 22 hours ago [-]
Free-as-in-freedom FSF style implies the natural follow on question to "am I allowed to change it" is "how easy will it be for me to change it"...
PS. I think you might also want to review the comment guidelines on this site... e.g. "Please don't post shallow dismissals, especially of other people's work. A good critical comment teaches us something. "
m1keil 15 hours ago [-]
My comment teaches you that rust propoganda gets on people's nerves.
gpm 7 hours ago [-]
Snowflakes who are bothered by someone mentioning a programming language when discussing an open source software project are not a market that matters to anyone.
astr0n0m3r 21 hours ago [-]
The comment was fine and infinitesimally critical considering the FAQs were vibe-coded as well.
dcoulmbe 12 hours ago [-]
OP here. I will assume you never built anything in Rust. I didn't get the trend too until I started building stuff with it.
ChatGPT:
Rust's biggest advantage is that it provides C/C++-level performance with much stronger memory safety guarantees, without requiring a garbage collector.
astr0n0m3r 37 minutes ago [-]
Rustaceans are beginning to sound like a vegans.
shaunpud 23 hours ago [-]
I run it on Arch
faangguyindia 21 hours ago [-]
It matters because if it used Electron, people may say it will use more RAM.
dizhn 13 hours ago [-]
The MIT license could use a Rust rewrite.
ActionHank 9 hours ago [-]
Half the refs call it BenchUp, half call it KetchUp.
Vibed out project has already gone through a rename and it's 2 weeks old?
Also, serious question, why would I trust this software?
PostOnce 20 hours ago [-]
Lot of upvotes and positive comments for a piece of software that can't extrude a cube without glitching.
No point filing a bug report; if the AI can't write it right, the AI probably can't fix it, and no one is willing to do the work.
tortasaur 11 hours ago [-]
> if the AI can't write it right, the AI probably can't fix it
This doesn't match my experience at all. Frontier models are quite good at finding the cause of and fixing bugs, if they have access to the right tooling. LLMs can easily fix bugs that they introduced with the level of detail in the average bug report.
hollowturtle 11 hours ago [-]
Is this a slopified version of SketchUp or? I'm not interested in free AI clones, quality wise they're way worse and all it takes to see that is that after a bunch of clicks eveything blows up(artcraft is such an example). I'm in this positioning where I don't trust anymore any new software, the only next software I'll pay a licenze for will be one that has clear AI disclosure and how they avoid slop, or no AI usage at all. I want quality, we are getting back steaming shit
dcoulmbe 10 hours ago [-]
The slop you are talking about is the slop non-developers produce. I have been a developer for the past 20+ years. I know how to scale and stress-test an application. For instance (real data), my slop runs the exact same file at 240 FPS VS 45 on SketchUp.
There are good and bad slops.
hollowturtle 10 hours ago [-]
So what's the full disclosure here? Do you know every line of code that has been pushed or not? Btw 240fps? Browser's fps are capped to monitor refresh rate
hollowturtle 10 hours ago [-]
Follow up: I confirm what I said, the moment you start clicking around you experience all the wrong things
arshxyz 6 hours ago [-]
The landing page was made with Astra and not Claude or a Chinese model. How can I tell? It's got the north-east pointing arrows that Astra loves for some reason
theknarf 12 hours ago [-]
Is it "KetchUp" or "BenchUp", the website can't seem to agree with itself?
xlayn 1 days ago [-]
That name is really close to ask to be sued... remember the Mikerowsoft vs Microsoft.
dcoulmbe 21 hours ago [-]
yeah, it didn't think much of it until I started seeing a lot of stars on Github.. I think I'll rename it Dovetail3D
globular-toast 17 hours ago [-]
Seems to have been renamed to BenchUp now?
QuantumNomad_ 24 hours ago [-]
Better rename it to “I Can’t Believe it’s Not Cat Soup 3D”
Jasp3r 14 hours ago [-]
Sketchy McSketchFace
rogerclark 20 hours ago [-]
The name also makes no sense anymore. Isn't it obvious that "SketchUp" is itself just a thin pun that references the word "ketchup"? They're just calling it nearly the exact same thing except that now "ketchup" on its own no longer has anything to do with drawing or sketching anything!!
fraywing 21 hours ago [-]
I see projects like this as both cool and terrifying for the economy
polski-g 7 hours ago [-]
Very cool product, if it supports a CutList feature (like OpenCutList) SketchUp 2017 would finally be dead to me.
wg0 13 hours ago [-]
Curious how much is the token cost? Genuine question.
999900000999 21 hours ago [-]
I loved SketchUp!
When I was first getting started with Unity I made all my early models in Sketchup. Thanks for sharing.
syntaxing 20 hours ago [-]
All thats missing is a MCP server!
rapnie 10 hours ago [-]
Ironic typo in the repo description "100% vibe-coded, only the finest slop, with great emphase on rendering speed and UI/UX".
absqueued 22 hours ago [-]
Whoever built this, thank you!
dcoulmbe 12 hours ago [-]
:)
singularity2001 16 hours ago [-]
so these tools are for Manuel adjustments give me scenarios in which just telling the agent to do this adjustment doesn't cut it
dofm 23 hours ago [-]
Is OBJ a viable form of geometry export here? Not clear what else it can export.
emmelaich 22 hours ago [-]
export menu lists stl and obj.
dofm 22 hours ago [-]
Ahh. OK. Both mesh formats, perhaps not so useful for buildings CAD but it is what it is.
williamgators 10 hours ago [-]
AI is freaking amazing
trevvr 23 hours ago [-]
You spelled length wrong
nicolamanzini 4 hours ago [-]
thanks a lot.
lapinovski 12 hours ago [-]
finally! I do simple woodworking projects and this was exactly what I was looking for! I don't want extra bits that's unrelated constantly pushed to my face or everything being locked behind a paywall that's not even related. Not awkward unnamed actions and have a dedicated side bar that's not trying to be 'smart'.
I just want to measure 3d simple blocks. And rest of the features are ADDITIONAL instead being forced on me. Love it.
dcoulmbe 12 hours ago [-]
thanks! that means the world to me! :)
charcircuit 23 hours ago [-]
A pun doesn't make a name not trademark infringement.
selcuka 19 hours ago [-]
They seem to have renamed it to BenchUp. Not sure whether that would be sufficient.
rogerclark 20 hours ago [-]
It's not even a pun! The original name was barely a pun at all... this just the original name with any semblance of pun removed! Agh!
charcircuit 19 hours ago [-]
There is a popular condiment called ketchup. There is a clear pun between Ketchup and KetchUp.
navane 14 hours ago [-]
It's not a pun if you use the exact word. SketchUp was the pun.
dcoulmbe 12 hours ago [-]
The original idea was KetchUp (SketchUp) without subscription (S)
opan 23 hours ago [-]
I like the idea and the name from the title, but if this is using LLMs, I'm not interested... Good luck to you anyway, I'm sure there's an audience for this. Used to know a guy who used SketchUp quite a bit. He modeled his basement living room with its weird overhangs and then a U desk to fit exactly into a corner by the stairs, then built the desk. I think he had no prior CAD experience.
charcircuit 23 hours ago [-]
Pretty much all software is built using LLMs. That's a rather harsh filter.
mapontosevenths 23 hours ago [-]
It's pretty easy to avoid. You just don't use anything Apple, Microsoft, Google, Meta, Adobe, ServiceNow, Mozilla, Linux, etc.
There are literally at least 5 softwares you can still use... As long as you stick to BeOS.
dizhn 13 hours ago [-]
I haven't looked at the software itself yet. I am VERY interested in it.
It's like Bootstrap, then React all over again in terms of all web sites looking the same and the "Claude look" is much more suitable for a literary site than a software product.
21 hours ago [-]
21 hours ago [-]
21 hours ago [-]
fireflywedding 22 hours ago [-]
Does KetchUp plan to support importing existing SketchUp (.skp) files? For people with a library of existing models, import compatibility would matter more than export — STL/OBJ lose the editable structure that makes those files reusable.
The kinds of large scale projects I like to make in SketchUp really began to feel constrained at a certain point. A modeler with SketchUp's incredible UI that can take full advantage of any computer's available graphics hardware to render large amounts of geometry more efficiently would be a wonderful thing.
I'm very possessive of my SketchUp models, and so the idea of storing them on any kind of cloud server I don't have control over is abhorrent to me. I was not happy to see Trimble putting so much effort into making a web-based modeler.
This comes up so often it's good to explain what happened there.
They were not out to extract value from consumer.
They retargeted the offering to professional engineering and architecture firms.
SketchUp was originally an offering for the AEC market. Then Google bought it and essentially pivoted it to consumer offering. Then Trimble bought it and pivoted it back.
I do realize this hurts the person who is a non-professional user but AFAIK SketchUp sales have never been better.
So it's not a pivot out of stupidity or malice, it's just a pivot.
The main customers - engineering and architecture companies - seem to be fine with it.
If you were an engineer CAD manager it would be odd to pay hundreds per seat a year for a product consumers can get for a pittance. So it's about pricing power I guess as well.
And do agree as a consumer there should be better options.
Monopoly rent extraction targeting a captive audience
Business customers don't pay for "having more buttons than mere consumers". Their business depending on such a crucial tool makes them reliant on swift and extensive support, negligible down times and so on. They are being cornered by ecosystem lock-in and workflow dependencies.
You could sell the same software at low prices to both "classes" of customers and charge extra for the services relevant for businesses. But then the fiction of "professional" whatever would be gone on the business side. And more importantly, companies like SketchUp would have to admit, they are merely SaaS with bad scaling prospects.
I think you mean ”companies like Trimble and Autodesk”.
”You could sell the same software at low prices to both "classes" of customers”
Slightly disagree. How do you brand the product then? ”Design your next hobby woodworking project” and ”Design next award winning apartment complex” are promises that speak to totally different markets.
So I would argue it would not make sense if you care about the brand.
Thats why these open source alternatives are so great - they are actually targeting a market no-one is interested to serve commercially and can have hobbyist specific brand without having to voice cringy boasts about being a perfect clone of something for the users of the said clone.
Expansion of software patterns to adjacent underserved markets via open source is really nice.
The point here is, the actually largest share of consumers doesn't get served by the "market".
It's a systemic failure and people fail to grasp the exemplary nature of it.
Unfortunately, that's not a problem with the strategy. That was taken into account by the business when the pivot was made, and it was determined to be an insignificant amount of users. The strategy worked without any problems, given the increase in sales. Anyone who used it during the consumer period is no longer the intended customer.
" in search of "
Trimble didn't need to search anything. They knew very well what the market is.
They could have a handled the consumer market retreat better.
I used it a lot since the first demo version came out on CDs of computer magazines of the era, and I wholeheartedly disagree with that.
The first release was definitely not feasible for professional use in the AEC market. It was just a easy to use 3D modelling tool for hobbyists and tinkerers who wanted to CAD or model something simple at home with the mouse, without the cost and steep learning curve of parametric pro tools like AutoCAD, 3DSMax or Maya.
It was the first 3D modelling tool I used that made 3d modelling simple and fun that a child could enjoy. But definitely not AEC levels, at least not the first release.
>Then Google bought it and essentially pivoted it to consumer offering.
When Google bought it, it was already popular in the consumer market. They bought it because they needed a simple tool that anyone could use to populate Google maps with 3D buildings of their neighbourhood via crowdsourcing, and nothing was simpler back then then sketchup.
>They retargeted the offering to professional engineering and architecture firms.
More correct would be they pivoted there because that's only userbase that will pay for SW. Unless you have the backing of someone like Google or Autodesk being you to finance such a loss-leader, then you're forced to move away from free SW for casual home users to paying pro-sumers to survive as a company.
The big problem for architects is it never got to be able to do architectural drawings while other software came in with easier 3d modelling and visualisation (revit which autodesk bought and built out fully for industry use, rhino etc. Even blender).
Google did very little with it and Trimble too. Google i can understand ignoring it after Google earth stopped being a focus. Trimble I dont really understand what they want it for
Also it was notable for being a major commercial success for the ruby programming language, outside ruby on rails.
Architecture and engineering.
It’s used in design workflows from apartmentns to nuclear power plants.
Sure, it’s not evolved much. But I guess that stability is a value for some users.
Negative. This was part of Trimbles strategy in AEC software purchases. They bought Tekla for Tekla Structures on 2011 for example.
Ofc they want to make money but it’s not like they _first_ bought SketchUp randomly and then tried to find someone to pay for it.
My exit has been FreeCAD with an MCP-bridge, letting me talk to an agent that implements my requests directly in the FreeCAD file I have open. Works super well with FreeCAD which has all the features I need, but where they can be hard to discover.
You can set up a digital twin of your actual workshop, with config-driven models of table/band/miter/hand saws, router tables, drill press, chisels and rasps, all sorts of fun.
I recently added a jig system too, so you can build jigs that you can actually use, both in your shop and in the tool. It's all drivable by a human or an MCP, and has video and plan authoring and sharing built in.
I've been building this as a labor of love since getting laid off in April, it's all completely free to use and people have made some really damn cool plans with it.
[0] https://sawdust.diy
- Flatten aboard: https://studio.flet.dev/run/46tp8GPq9g/
- Make rabbeted box w/ optional sliding lid: https://community.carbide3d.com/t/rabbet-box-designer-on-fle... (working on a hinged option now)
drawing on what I learned putting together: https://github.com/WillAdams/gcodepreview and https://willadams.gitbook.io/design-into-3d
Sawdust has support for laser cutters [0] and has procedures for generating finger joint boxes, with svg export. It can also do more advanced forms via mcp, like this arcade cabinet design [1]. G-code is high on my to-do list because LightWorks isn't getting updates on linux any more would be great to send a design straight to my diode machine.
Sawdust also has support for 3d printing, both for including printable components in designs as well as printing scale models of your build [2]. No slicing built in, but full .step output for any slicer you like, with build volume aware scaling so you can print it on your machine.
[0] https://sawdust.diy/knowledge-base/how-to-use-the-laser-work...
[1] https://sawdust.diy/whats-new/laser-fabrication-enclosures-a...
[2] https://sawdust.diy/whats-new/woodworking-models-can-be-expo...
If you ever have/get any, a photo of any of the things on your site, actually built, would be a cool addition—make it real.
I have a few projects that are in progress while continuing to build out ShopSpec. This is so that I can address any rough edges as they come up. For instance, material handling and safety during sheet breakdown.
If you go to the plans section in the 'my shop' tab [0] and click a plan, e.g. this Pinocchio puppet [1] that someone requested via the public feature/plan request board, you can upload photos of anything you build and share a comment about how it went. Credit where due, I modeled this on Instructables and Thingiverse and similar project sharing sites. I for one cannot wait to see some pics of what people build!
[0] https://sawdust.diy/myshop
[1] https://sawdust.diy/plans/8088707d-db0a-4d2a-a73a-686f2c5e2c...
I signed up inside a minute of opening it :)
1) The sawhorse explosion feels physically inconsistent, in that all of the pieces go "straight" out at once, leading to things like the nails being dragged sideways through the wood. Since you have the build instructions, you could reverse it a few pieces at a time to fix that. 2) A couple places the AI-written text sneaks through in ways that are nonsensical or irrelevant to the reader - like "drag to orbit. The page scrolls past." on the 3d-print birdhouse diagram. The AI writing is more noticeable in the "learn woodworking" section but also seems to generally make more sense. (Personally, I'd be paranoid about potentially giving unsafe advice, but I assume you check it all before posting. I didn't notice any safety issues here.)
I've bookmarked it to try the tool out.
I'm embarrassed that "drag to orbit. The page scrolls past." has escaped my notice til now, yikes. I have an entire skill [0] devoted to eliminating this kind of ui slop, sadly it missed this one.
The landing page is very old at this point and definitely needs a freshen-up. The exploded view feature was built before fastener support, so you are absolutely right about that. Sawdust already can do much better, here's the same Sawhorse project's guide [1], take a look at the video there to see the fasteners driven in with respect for actual physical collisions :)
If you do try the product out you can easily submit feedback through the MCP or via the ui chat, I get paged for every feature request and love working on things users actually want. Would highly value any additional feedback.
[0] https://gist.githubusercontent.com/birep/bb865b2799d3f86d55f...
[1] https://sawdust.diy/share/882f2473-d283-46d0-8f38-7b289253d6...
Edit: the fix for the explode (and removing the slop ui label) is deploying now, thanks again for the feedback!
https://developer.apple.com/news/?id=w5ngl9k2
There is a name for it: "enshitification" is the strategy of maximizing short term gains, even if that results in product suicide on the long run.
Trimble bought Sketchup 14 years ago. In the world of commercial software, that's a decent run and they seem to be going strong. They just aren't interested in the nickel-and-dime world of selling to and dealing with consumers.
I wanted to have that data model and SketchUp had this amazing cool thing where all plugins (which were written in Ruby) used to be running in the same space and so I could access the runtime of any other plugin. Here the team of OpenStudio were kind enough to share some of their internal APIs with me and so I could take their model and append my own data on top and let my users reuse all of their models from OpenStudio.
And then SketchUp web came and that whole plugin world was crashed. This forced me to go down the path of using a union-of-all-industry-file-formats (like TIFF for pictures) for building energy called gbXML, back then no one supported the file and so it was hopeless (and hard to implement) and eventually trying to create a CAD, which given the state of tech at 2016~2017 was hopeless too, and so it failed miserably.
Seeing this now, I was wondering how hard would it be to recreate that project with its own full CAD from KetchUp...
Would be worth trying with AI to help you develop it. Probably not too hard if you could spend a few weekends on it.
It's not too hard but I can tell "few" means "few hundred" at least.
Need a 3D modeler? Use Claude or codex to build the 3D model first, then build the modeler around it as you iterate towards what you want to build
So many times I’ve had to spend hours learning a software, reading docs, doing tutorials, only to later be able to do a small basic idea of what I actually wanted to do in the first place
Now I can just build only the software I need, for the thing I need it for (or, ala REA, maybe either modify or reverse engineer some existing software, to better fit my workflows)
In that sense, it would also be cool to be able to drive KetchUp with an agent. I’m a big fan of what pen.dev has done (although now their tool is not free/open anymore)
It’s always more work than it looks. You can get to mostly usable tools quickly if your request isn’t too difficult, but every time I use them I run into another thing that needs to be fixed, changed, clarified, or cleaned up.
It’s fun at first, but over time you realize you’re spending more time building the tools than using the tools to build the thing you originally wanted to work on.
I think personal software is best for highly specific tools or for things you need to use repeatedly for a very long time.
For something like a SketchUp competitor, I wouldn’t want to try to maintain my own. The number of edge cases and features is bigger than you imagine up front, and the vibecoded version that might appear a few hours after your prompt is going to be very far from the version you want to use.
The vibecoded apps all rely on an army of users exploring the program and reporting all the bugs and edge cases. You don’t want to be doing that on your own.
it seems like the open-source space is losing some aspects of collaboration. people used to search for already existing projects on some git-hosters, maybe fork them or put effort into the existing thing. now half a dozen people each run their own race using AI, each solution misses 30% of the features and has x-amounts of bugs.
work together ffs :( i mean, 20 people can use AI to vibecode the same project, why does everyone now feel the need to do everything solo. (because they want to be the prolific founder and cool dude who did the most in the project* instead of being one of many contributors, i assume...)
*or rather, their AI.
I'm not too sure what the future here looks like and if this is even a problem, but things certainly are changing. These days submitting a well written bug report or feature suggestion, or validating existing reports is more useful than sending a PR.
Right, and that's something I already argued for since the very beginning of GenAI, namely that suddenly being very costly, economically and ecologically, is not that bad if you build upon what exists and collaborate on progress.
... except that's exactly WHY vibecoded project exist : they do NOT have to give a shit with existing codebase and agree on weird rules. They just start from "scratch". In fact I'm convinced there is the Ikea effect at play, namely that the PM of the vibe coded project is emotionally attached, even proud, of the result they think they provided somehow, despite them not actually doing it. Obviously for-profit GenAI companies have a big incentive to maximize token use and will never promote workflows and tool that minimize the one thing they monetize.
So even though I agree in theory, in practice I think there is little to no overlap.
ArtCraft is definitely a play at selling subscriptions.
The slew of "oh yeah, blah blah blah blah blah, that's why I built [X]." comments on every AI post now?
Click through and many you'll find "Pricing".
It's a gold rush.
2. Click "Fork"
3. Profit
I have lot of tools that work for me.
When externalities break them, it's usually one prompt to fix them.
To be fair neither is quite as complex as a whole 3D modeller.
> To be fair neither is quite as complex as a whole 3D modeller.
Once you cross a certain complexity in your project, it becomes quite work intensive even using AI to do the coding. All the back and forth and cleaning up, exploring ideas and starting from scratch again takes time and effort and maybe most importantly, extensive human thinking. Just to clarify, I am not against AI and use it extensively myself. But vieb-coding all your tools is just not really a feasable idea for most (time/token-cost constraint) if you tools are of a certain complexity.
But you could easily vibe code, e.g. an H.265 decoder, even though it is super complex.
AI can get the implementation done just fine but in my experience it needs guidance because it can't guess what features and details I want / don't want.
The difference to "before" is that it's now possible at all for me to get personal projects to a working state but in sum they still require a few hours every now and then supervising.
My personal projects matter a lot. For some of them I am trusting my private data in the code, so I need to hold it to a high security and privacy standard. For others, I have hundreds or thousands of users (game mods, helper tools for hobbies, etc.), and I don't want to ship low quality to them.
If it doesn’t matter then do what you want. By definition, it doesn’t matter.
Some of us actually use these tools, so it does matter. You do have to consider things when it matters.
It’s amazing that you can get something usable quickly, but then every time you use the software you’re back in the AI working on your tool.
Using AI to build the tool becomes the project, replacing the project that led you to think about vibecoding a tool.
For instance in the context of this SketchUp clone that might mean using AI to generate an addon for Blender.
For some of my recent projects I'm using the community edition of Odoo to serve as that substrate because it already has an established community of modules to draw from, ten years or so of real world use, and I don't have to burn tokens generating a user account system with permissions or the data models / ORM to interface with them.
This seems to be serving me well, but we'll see. It could be that I end up in the same the tool becoming the project scenario that you describe, just a little bit later than if I had started from a blank slate.
At least that's my experience.
Small tools with few features is great. a 3D modeler is not that.
I suspect people saying "everyone will vibe code all their software" have never actually tried to vibe code a meaningfully complex application.
It gets so much worse. At some point you start hitting edge cases where there is no clear right answer and you have to come up with a kernel wide coherent approach to things like 1ulp differences. Parasolid is full of decades of kludges on top of kludges and giant switch statements trying to preserve backwards compatibility as they explored the space.
I think it comes down to where you want to spend your time and attention.
https://www.winestockwebdesign.com/Essays/Lisp_Curse.html
Like, I have to imagine that 80% of the "quick stuff" people need to do in Blender would be done with:
* create quad * exude like 5 times * done
Maaaybe
* perform boolean operation between object and a circle or 3
---
Like, I have less than 200 hours in Blender and I was pretty functional after like ... 30 minutes to an hour of a single tutorial.
You’ve likely just switched hobbies to your Blender clone instead of what you wanted to make.
I wonder if Claude Code and other similar tools start to maintain their own base apps that everyone keeps asking for? This would be like a "cache" that makes the requests cheaper/easier to deliver?
Then everyone's "own" apps is really just a customization layer on top of a solid base that already does all the heavy lifting.
So many people are building their own versions of:
- PDF editors
- Word processors
- 3D editors
- Video editors
- Pixel-based image editors
- Vector-based image editors
- CAD editors
etc.
In some of these there are a few flavors as well that are major architectural choices. But most of what people are doing is just playing around with UI and file formats. Much of the functionality is generally the same.
For the rest, at least so far, I haven't had the collaboration experience but for word processors I can't go back to local files that can't easily be edited/viewed/commented on by multiple people or myself on every device I own.
I think they all have been post-trained on specific tasks like building sites, apps, etc and that is a large part of the reason they are pretty good at it.
All these clones of Photoshop and similar tools are mostly dead ends, they are experiments to see if you can do it, but there will be a consolidation around regularly maintained and powerful and complete apps that likely also offer support.
These experiments are dead-ends for the most part long-term. Fun play things, but they will not gain significantly user bases.
Why? Every time someone cripples software in the name of safety it just gets replaced with more capable software that isn't.
Before that, though, I think you have to learn what the experts do and what can be done and what you should want. And then you can start to make your own opinions.
By then, unless the major players have done it really wrong, you're not going to make something better than what already exists.
Right now, it's different, because we have that experience prior to the ability to just wish up an app and have it happen.
I think the value of adapting full tools to your own workflow far outweighs the maintenance costs nowadays. The tool above for instance took my agent a couple hours and it was good to go. It wasn’t a small slice by any means either: it supports edit/obj mode, extrude, loop cuts, modifiers like subdiv, save/load, undo/redo, bones, auto rig, posing, and more.
Mind you, this is a slice of an open source tool which was still outside of what I wanted from it. Cost wasn’t even part of the equation here.
One of the tools (built for me to score options of different assets): https://id-le.com/review/intro-logo-animation
Video: https://youtube.com/shorts/tlI9_XBuqVk
Repo: https://github.com/nicobrenner/id-le
Online: https://id-le.com/
I could start from scratch obviously - but seems a shame.
I think both those apps are great and happy to know folks like you are maintaining them; this one though is simply more of a personal software than a path to an OSS project!
Since Blender is OSS, you could probably use one of the existing VR views for Blender and write a bunch of in-app Python scripts to do what you want, if you wanted to also be able to output to Blender files and work within Blender already.
This looks like a pretty well-trodden area in Blender.
Adding scripts into an existing product does not serve this.
I think we are entering into a new era where we don’t need to accept fully featured software with all of its decisions should they not fit ours.
And you get battle-tested software and a huge collection of existing plugins, operational modes, export formats, all that. Fully bespoke software for everything you want is a way toward losing interoperability between people and platforms.
But building up as opposed to building down or rewriting is freeing.
I can pull in ideas from Houdini and mix them in. I can use Blender’s source as reference for specific implementations should I need. I can keep my tooling specific to the projects Im working on and design it to fit that as opposed to build a workflow within a software meant to be general purpose.
IMO, this kind of full customization is going to become far more commonplace.
Not necessarily. But only if what you need is not available in open source or not achievable by forking open source. Otherwise it’s a waste of money (token) and attention. This is a significantly smaller pool that “build [all] your own tools”.
Although, AI mega corpos would love if we do constantly re-build everything. You can tell by the tendency of the agents to write everything from scratch instead of importing the code that does the same thing or better.
I have no doubt as a newcomer you can make software that will do what you want, but most likely the outcome will no be good, not because of yourself or the tool, but rather because the expertise to map the problem space to the solution space is not trivial, and that's precisely what requires a new way to think, and thus tools that aren't so easy to learn. It's not poor UX.
"So many times I’ve had to spend hours learning a software, reading docs, doing tutorials, only to later be able to do a small basic idea of what I actually wanted to do in the first place
Now I can just build only the software I need, for the thing I need it for"
Or another way to put it for devs “A language that doesn't affect the way you think about programming is not worth knowing.” ― Alan J. Perlis
I host it here: https://vibe3d.spoerri.dev - my partner enjoys it because she can bring an idea to print pretty fast.
Maybe a similar idea could be interesting for KetchUp as well.
So, I guess yes, if there is someone's hard work you can copy and steal and make your own, I guess that's true. But try to create something that no one has ever built yet.
https://news.ycombinator.com/item?id=50406927
For an example. Lots of people are now vibe coding their tools.
Would love to hear advice from people who use CRM tools on the job.
I think people underestimate what a time investment building your own tool that's of at least medium complexity is, though. It might be worth it if it's something you'll use regularly for a long time, but I don't foresee it being the norm. Even assuming no bugs ever or no performance issues ever and all the architectural decisions made were appropriate (pretty far from current reality), merely describing what you want and how it should behave and look like is a long process, and one that isn't all that enjoyable once you go through it once or twice already.
The existing software that has been around for several years and you kinda dislike is probably much more reliable than your own homebrew, too. Doesn't matter in some cases, matters hugely in others. Learning that your custom video gallery organizer just vanished half your family videos probably isn't very enjoyable, but at least it's easy to recognize it happening. Something more abstract like CAD or 3d printing slicer, ehhh...
A good visualizer might be useful thought! If that part of the work needs to be validated.
Wat is the point of personal projects if in the end you learned nothing
By outsourcing the modeling to Claude or whatever, you don't learn how to build 3d models on your own. You've become dependent on the LLMs, and you're trivially replaceable by someone who can build 3d models on their own...because they can also learn to tell Claude to do all the work.
Their models are horribly topoligized, which will matter for most use cases for models; and if you’re just using it to make a cube, it’d be better to learn the related program to do that.
People tend to ask LLMs to make full art pieces (a decade of specialized knowledge and muscle memory), not “draw me a square”.
In 3d art, there is a gulf between the easy stuff and the hard stuff.
Tl; dr: for 3d art, I agree.
I'm a very casual user of SketchUp (I use it once every few months), but I have a single file which is a detailed model of my house, which I have invested many hours in creating, and that only works on the full Desktop app. So the free web app is not an option for me. And a subscription doesn't make financial sense so I have to ask for a new trial every time I have a home renovation project and I need to modify my model. I'm aware that eventually they might refuse a new trial (they haven't yet) and that day I will just stop using it and learn Blender.
However, I'm now realizing that at least for home renovation projects, an essential feature is the SketchUp 3D Warehouse, where you can download user shared models of a TV, a toaster, a chair, etc. Specially for a casual user, you don't need to have the exact model of your real chair, just a chair will do. So the warehouse is an essential part of the value.
I'm sure there are alternatives for free models other than SketchUp 3D Warehouse, but having it directly in SketchUp is super convenient.
Impressive how good that rust UI feels in a browser though. Even the text rendering is impressive. Maybe rust really is the future of cross platform web+mobile apps.
I will still continue to pay for the things i truly get value out of and are in my opinion reasonably priced. To support the creators.
Seems unlikely that people are going to spend their Claude sub every month maintaining the thing either. They'll move onto the next cheap LLM project and dedicate their tokens to that.
Im sure that's not every project, but Id rather just have something reliable with a reputation. AI clones are killing open source's reputation for me, I just dont trust anything now.
Not to mention how easy it would be to get a malicious PR approved into a public repo who's maintainer relies solely on the competency of a cheap llm.
ArtCraft is also pretty cool but I have my Affinity Photo 2 Licenses so I do not realy need these currently.
If it was crocheting, I would get it, but woodworking is not an inexpensive hobby where the software becomes the top line item.
This here I would also use at my job whera I am a bit of a Jack of all trades so once a quarter I need to sketch some model of something to see if it would fit or not. Example would be right now i need to sketch a simple collision protection for a outdoor cable tray (for Solar Pannel Strings). Gonna not take me longer than an hour. I do not use it enough to justify spending 40 Bucks a Month 4 times a year.
Its just that often times, even us programmers seem reticent to pay for software even though we happily throw dollars across the counter for a sugary coffee somewhere.
I know of sketch up primarily from DYI magazines, hence the assumption.
EDIT: latest release switched to BenchUp fro KetchUp. Let me assume it's and attempt to duck infringement.
Gained a huge user base quickly then sold to Google who wanted it apparently to help people toy around in Google earth, then they did little with it and eventually sold to Trimble, mostly a surveying company, who similarly had hopes it would help expand their users and market reach, which it hasnt really its mostly still evades commercial revenues and continues to be used by the same kind of users as originally
i directed pi to their documentation + github repo and created a pretty awesome skill that allows me to take measurements on paper, snap a pic, and then have it generate the drawing. can also snap a general pic and then provide a model number of an item in the pic, and it will maintain scale. this stuff used to take me an hour+!
Going to have to try this!
https://github.com/egeozcan/ketchup
Problem existed even before AI, with framework names. Headlines like "Planecrash in Ohio gains 50% in one year", "ShowHN: I use car accident to get new customers", "Made my wife happy with broken arm"
Feedback: every time I draw a rectangle it asks me to name the object; seems like the wrong default behavior.
yeah this seems insane. The push/pull also doesn't work the way I expect it to. Sketchups main selling point was how easy it was to use and pick up
Every interaction I try on KetchUp has some kind of gotcha. It doesn't snap to grid by default, and as you say, it's insane that I have to name every rectangle. I can't extrude a face with a single click-and-drag of the push/pull tool; instead I've got to select it first with the arrow tool (which strangely does not match how the Move tool behaves), then extrude. Also, clicking and roving through the File/Edit/View/Tools menubar doesn't interact as expected.
Sorry to say, this ain't SketchUp. It seems like yet another flashy vibe-code that has fulfilled an 80% quality level but crumbled on the final 20% of polish that really makes or breaks the user experience. Really feels like the author didn't test-drive it for usability.
1. https://github.com/sawhorse3d/sawhorse
SketchUp 8 was the easiest software for me to learn, and it just clicks with me. I really wanted to learn FreeCAD, but there's something about the way it works that I just can't get along with, so I keep going back to SketchUp.
Later, I tried all the alternatives: Onshape, Fusion, Tinkercad, and the new SolidCraft, but SketchUp still feels so simple and intuitive to me.
I've had better luck with OpenSCAD for simple things, and I still need to try it out with LLMs. That's on my TODO list!
It's also weirdly good at taking an existing STL file of a functional print and writing some python to modify it.
One example was a badge holder that I liked the design of, but only held 2 badges. I wanted 3, plus a slot on the back for a key. Gave it measurements of the key, it grafted a key slot on the back, and added the third badge slot. In typical 3D printing fashion, of course I had to print like 5 iterations of the thing until the sizing of everything actually worked.
This is great for small scale STL design too
https://github.com/MojzisekJan/rysmo
And it uses a project called openskp [1] to work with SketchUp files, also AI coded to some extend, but looks reasonable at a glance.
[1] - https://github.com/iamahsanmehmood/openskp
Also in rustlang, but under GPL3.
Been using Solidcraft which is another great example. Lower memory usage and better performance than Fusion 360 and improving quickly.
The power agentic development gives to a single person with love for the craft of software development creates better products than a large team with corporate steering.
This was always the case but types of software like advanced CAD modelling was outside the scope of what a single developer could accomplish. Now it isn't and it's a great thing that will create a lot of high quality new tools for humanity.
1) how do I make a thing exactly N units long? like, a 10cm long stick
2) can I just put dovetails in random places, and have them make the object split in two? Would be wonderful for 3d printing
Seriously, the subscription frontier models scratch my control freak nerves and lets me build my computing set up just the way I want and not a single bit more. I’m currently making a minimal Linux distribution for myself based on Gentoo with the idea to develop any and all graphical applications except Firefox _for_ myself using LLMs. Nothing on my system I don’t need.. It sounds absurd but is still possible in 2026 with these subsidised LLMs. They are not my private data, so I don’t mind having the agents build it, test it, and automate it. If they want to train on my calculator app and CAD built upon other code from the Internet they are already trained on, they are welcome to :-)
This right here is a new age of personal computing.
;-)
Omg why is rust relevant to the question whether this is free or not.
PS. I think you might also want to review the comment guidelines on this site... e.g. "Please don't post shallow dismissals, especially of other people's work. A good critical comment teaches us something. "
ChatGPT: Rust's biggest advantage is that it provides C/C++-level performance with much stronger memory safety guarantees, without requiring a garbage collector.
Vibed out project has already gone through a rename and it's 2 weeks old?
Also, serious question, why would I trust this software?
No point filing a bug report; if the AI can't write it right, the AI probably can't fix it, and no one is willing to do the work.
This doesn't match my experience at all. Frontier models are quite good at finding the cause of and fixing bugs, if they have access to the right tooling. LLMs can easily fix bugs that they introduced with the level of detail in the average bug report.
There are good and bad slops.
When I was first getting started with Unity I made all my early models in Sketchup. Thanks for sharing.
I just want to measure 3d simple blocks. And rest of the features are ADDITIONAL instead being forced on me. Love it.
There are literally at least 5 softwares you can still use... As long as you stick to BeOS.
It's like Bootstrap, then React all over again in terms of all web sites looking the same and the "Claude look" is much more suitable for a literary site than a software product.