Hacker News new | past | comments | ask | show | jobs | submit login
Game Builder: Create 3D games with friends, no experience required (blog.google)
495 points by sohkamyung on June 14, 2019 | hide | past | favorite | 138 comments

Similar to how the hottest chat app is google docs for the new generation [1], This is definitely a step in the write direction to snowball that momentum of 'virtual chat + collaborative playtime'

[1] https://www.theatlantic.com/technology/archive/2019/03/hotte...

Some of this is the penetration of G Suite and cheap Chromebooks into the education market. Its more or less the same play that Apple used back in the 1980's and 1990's.

Anyone for Roblox :)

Yes. I remember (nostalgic times) when Roblox + Google docs (for group documents) and Teamspeak where the main way we communicated

I must have gotten old somehow, because I remember Roblox as being that game I put on the iPad for my 5 year old nephews.

Roblox has kind of gone through a number of iterations. There was a time where it was really deep by default (and not much point and click) and knowing a good amount of Lua was useful. It has been around since 2005-2006, long before the iPad existed. I was .. 10 at time of release, it was really fun :).

I (and a lot of my friends) got their first real start with programming using Roblox back in 2006. Most of us now work professionally in software or security, directly as a result of our experience with making games in Roblox, or hacking the game.

Writing exploits for Roblox gave me experience with everything from binary analysis to web security and cryptography.

I've heard similar stories from other people for games like Runescape or Minecraft. It's really astounding to me the ways in which a dedicated enough teenager can take advantage of games to turn them into something really educational.

Also, my handle in Roblox was Aeacus! Hit me up if we ever played together :) I'm sure there's a ton of overlap between HN and the smallish Roblox scripting community of 2006.

The two are not mutually exclusive.

i think your giving it too much credit, this seems like a side project by some aspiring game developer at google,

It's published by Area 120, which is Google's internal "startup incubator". So it's sort of like a side project, but full time for a small group of people.

Wife and daughter have spent the last two months working on their own private minecraft mod with MCreator, this should tickle their interest. the puzzle block programming paradigm promptly pissed off my preteen to the point she's asked "wouldn't it be easier to type this?"

Seems she can implement cards using javacript.

I was ready to be skeptical but this is pretty cool. If it were trying to replace Minecraft or games like it then it would be a tough sell, and you would have a hard time breaking into that community. But it looks like it's more powerful for the individual contributors, with the ability to make custom blocks/entities on the fly rather than just using the ones someone else coded up and made into a mod.

I think it’s a great market to tackle- as both a game/level designer tool and also an educational intro to scripting/coding.

IMO it’s very reminiscent of Microsoft’s Kodu and to a lesser extent Media Molecule’s Little Big Planet and Dreams games.

I wish the team the best of luck! Hopefully it keeps chugging along for some time; I’m wary because particularly these kinds of not quite game engine and not quite game projects have a bad history of abandonment within 3-5 years (I’ve researched this particular space for a while).

> Hopefully it keeps chugging along for some time

That's my biggest question, especially with Google. Meanwhile, Kodu is still getting updates 10 years later! [0] (Disclosure: I was an intern on Kodu years back, so I'm probably biased ;) )

[0] https://www.kodugamelab.com/archive/new-kodu-release-15470/

I thought Kodu was pretty much frozen at this point, and I think Project Spark failed. Is it someone at MS working on the code?

There's still a dev fixing bugs and improving perf full-time, but it's not getting drastic changes. More people are using Kodu now than ever before.

Yes but they (Microsoft) also shut down Project Spark, the ‘spiritual successor’ to Kodu.

touché. What a missed opportunity Project Spark was. I feel like a book could be written about that.

Looks amazing!!

I wonder how well it will cope with non-basic games, as in non-platformers, non-shooters.

I've noticed that many frameworks are great for basic tutorial project usage but don't do well once you start on a real project

That's where you need to build up the fundamentals that you've learned. I'm excited to see what people are going to build with this once they've gotten comfortable with the engine. Creativity will be the limit!

I just had a play with it. It seems very well thought out. It has a very nice tutorial that introduces all the basic concepts. I am looking forward to trying out the multiplayer aspect, as that seems to be the most compelling thing about it.

Being able to write some code to make a custom block and then just drop it right into the world, working the same for everyone else, is pretty cool. Hopefully that is how it works.

Does anyone else remember an old "game" for making games? It would have been late 90s when I played it, and I remember we bought it on a CD. You could build little top down games in it, I think the example was a football game.

My family's software company published a product called "Game Builder" in the early-mid 90s that could be used to create adventure games. Here's a listing of it from 1999: https://web.archive.org/web/19990224063815/http://www.mvpsof...

Ha: "In our testing, children as young as eight years old could easily write games with Game Builder Lite..." I'm guessing that would have been me in 1992.

GameMaker? Now developed by Yoyo Games


These are too early to be the one you’re looking for, but there have been a number of these, including these two I had myself back in the day:



I remember GameMaker from the 80s on the Apple ][e!

Also had Adventure Construction Set by EA (before they went all sports-crazy): https://en.wikipedia.org/wiki/Adventure_Construction_Set

I also had a cracked copy of Pinball Construction Set too: https://en.wikipedia.org/wiki/Pinball_Construction_Set

Klick & Play?

Or maybe the latter iteration of it (which I grew up with), The Games Factory.

Man, I lost many hours to Klik & Play. Most of the games I made were terrible Arkanoid clones.

Games factory? Thanks to that game, me and my friends made many games without programming back in middle school.


I had that (and its precursor, Klik'n'Play). They were a good bridge to programming; by the time you're making something approaching a commercial game in them, your event grid is huge and you're doing a ton of work with hidden objects which you're basically using as a weird kind of declarative programming language. At which point it's easier to just write actual code in an actual language.

DIV games studio?

DIV game studio was really cool. Especially the online community forum where you could share your creations. I still remember that clay animation game. It looked so impressive


Seems nice. Anyone has an opinion on why Google built this? Do you think it is related to Stadia? Or just a company hobby turning well?

From another comment, it was released by their internal startup incubator. It's potentially not related to anything else at all inside Google, just a project a small team was able to put together with Google's money.

Thank you for the information. I did not see it.

While I suspect there's no deep strategy behind it, one could argue that there's the possibility that long-term, future social networks are going to pop up inside things like Fortnite or Minecraft, and Google would want to capture as much of that space as possible (and sell adverts into it).

Google is always focused on AI. I am pretty sure there will be AI among the players who will behave like humans. Maybe your best friend in game will be an AI.

I find this comment fascinating. I wonder how you came to think this was a tool for AI... I can't see it because it very clearly just looks like Minecraft with some basic scripting. No one is going to be building advanced ai in something like that.

But perhaps an ai could be trained by it.

> Find 3D models on Google Poly and use them in your game instantly.

Google Poly seems to be a place for uploading things made with TiltBrush or with a program called Blocks. First time hearing about the latter.

Will it be possible to load models from OBJ, COLLADA or other formats?

Edit: Originally I asked what about Google SketchUp also, but it turns out SketchUp was sold to another company in 2012.

You can already upload any model to Google Poly. I remember submitting one made in Blender. I thing I had to export it as OBJ.

Coincidentally enough I've spent half the day writing a .obj loader, and in the process found out that Windows 10's "3D Viewer" app can open .obj files.

Edit: More than you ever wanted to know about the .obj and .mtl formats:



> Originally I asked what about Google SketchUp also

I'm still using Google's Sketchup 7.0, it runs well enough on Windows 10. The new company turned it into an expensive Pro tool with a 'lite' web only version :(

Any word on when it will get Linux support?

It's working for me using Steam Play (Proton 4.2-7) on Xubuntu 19.04.

I thought Steam could automatically load up Wine, or something like that?

You are correct, they have developed Proton [1] which is based on Wine.

[1] https://github.com/ValveSoftware/Proton

But it's not automatically. I think the developer has to enable that. Probably because the developer is still intended (or wants) to verify that it really works properly in Proton.

Or not sure. But I have not found a way to run arbitrary Windows Steam games on Linux with Proton. Not sure how I would do that.

Yes, you can force it on all games if you want. I'm currently playing many games on the latest version of Proton.

It might be named a bit confusingly though, I'm at work and don't remember exactly but it might be worded something like "Override proton version for all games" which also enables it for games not explicitly marked as compatible.

I thought if you enabled the beta version, you could try any Windows game using Proton.

I tried Proton a while back. My experience was hit and miss with my particular collection.

have you managed to get it working on proton?

Not at all. Linux support is developers effort and if you are serious about it you won't use wine.

It can actually load arbitrary games with a port of Wine https://github.com/ValveSoftware/Proton/

Some games run better on wine than on their native platforms.

I really believe gaming is the #1 thing holding back linux in the home. The #1 thing in enterprise is probably cheap sysadmins.

Try Steam's new built-in "Proton" system. It's made almost every Windows-only game in my Steam library run flawlessly with a double-click.

With recent developments like DMVK / Vulkan and such, even the latest DirectX games run amazingly well (I'm able to run high-end games like Heroes of the Storm, Witcher 3, etc at native speeds on High/Ultra quality settings).

Got ahead of myself and forgot to check the Steam forum for it... and it was asked and answered already. TLDR; No plans.


This is great, but I wish there was a stand alone installation. I get that most gamers would be happy to use Steam, and I am myself, but I would love to use this in some of the coding clubs I run. I can imagine the look on the network guy's face when I go to him and ask for Steam to be installed on a school network.

Why do I need steam for this? Seems very odd.

Steam provides a CDN, updater, friends list/social infrastructure, really terrible but functional user generated content sharing infrastructure, and cloud saves to game developers, so it's pretty appealing even if you're a team inside Google since the alternative is building your own installer and updater - Google has no competitor to Steam on Windows or any other platform of note other than Android.

Also, Gamers actively get angry with developers now if a game releases on something other than Steam, so it's probably better for PR and adoption rates to just target Steam and be done with it.

To be more specific, gamers get annoyed when they have to install yet another "game manager" program to play a game. This is especially true when that game manager is well know to lack features and be insecure.

Installing a game standalone doesn't annoy anyone I know (though, to be fair, I'm now older with few friends)

And to generalize it a bit - users get annoyed when they want to just buy a product, and the producer pulls some vendor lock-in bullshit trick. They're right to be annoyed, because not only it is inconvenient, such moves are meant to exploit them.

"Game manager" is a especially visible case because, with the exception of Steam, Battle.Net, they're almost universally crap. Some of them are so bad I wouldn't want to install one even if it was the only game platform in existence!

You normally have to create an account to get the new game manager to work as well. Now I'm trusting yet another company with my account details, and had to "accept" a bunch of terms and conditions all totally unrelated to the game I want to play.

FFXIV from Steam requires the download and usage of the abysmal launcher it has. As an added bonus, when you buy the game on Steam, you're locked in to Steam releases going forward: can't buy Shadowbringers from FFXIV's site, you must buy it from Steam.

Every Ubisoft game since Black Flag that I've bought from Steam has required Uplay.

I'm sure there are other examples, but Steam isn't preventing users from being hit with another "game manager" download. In some cases, it facilitates it.

and steam networking which solve so much issues hiding away traversals, upnp, and whatever other trick it's needed to get two peers to talk as needed

Sounds like the lambdaMOO of the present. To drop code objects into a life multiplayer community is an ancient concept. Fun today as it was back then.

The build and instant play feature reminds me of Trackmania for some reason.

It was (is) a great game, but I still cannot believe why it never became famous.

I remember Trackmania; it appeared well known to me when I played it around 2010, mostly because of the free (as in beer) version.

Are you from the US? It was fairly popular in Europe and also played in the ESL, but I hardly ever saw players from anywhere else.

haha I am from India, one of the somewhat 100 players from India when it came out. I was ranked 15 when trackmania 2 came out.

Yes, you are right all the best players were from Europe.

It looks very cool and also very familiar. I used to work with the tech that turned into CREY (https://www.playcrey.com/), and this seems to be in the same vein. CREY might be interesting for anyone who wants to check out comparable offerings.

"with friends, no experience required"

...but what if I've got experience but no friends?

I played a game like this when I was a kid called Atmosphir.


How is content moderation done on thisnplatform? Is it all manual? It isneasy to create offensive or illegal contetm when you have tools for free-form expression. This problem is the reason why some game developers stopped adding even simple features like uploading custom tag images in multiplayer games.

Oh I can't wait to try this one out. The trailer makes it look good. Most importantly, it has workshop support! https://steamcommunity.com/app/929860/workshop/

Anyone know if Google plans on open sourcing it, so we can port it to different platforms?

Just wish it had gamepad support!

that should be fairly trivial to implement if they are indeed using unity. i agree it would be cool to have, additionally it would allow for local multiplayer which is always still the best way to enjoy these things imho =]

I just tried it out, edited some code. It's alright. The problem I have with it is the graphics quality. It's very plain. No shader access. Close to no lighting.

I decided not to invest more time into it, it needs some work to be a game engine.

Haven't tried it yet. I think it will be a great prototyping tool too.

Wait, so can you publish those games on say Steam later on?

It seems you can't yet.

Anyone know what engine this is using?

Considering google cloud's partnership with Unity, probably Unity.

yup, it's Unity

Can you share "cards" with other people or just within your own projects?

Reminds me of Macromedia Flash and SWiSH Max if anyone remembers that tool.

OMG I had so much fun with SWiSH. Haven't thought about it for a few years now ... brings back some nice memories :)

This very much reminds me of Little Big Planet in a good way.


Predictable comments are unsubstantive. Please don't post unsubstantive comments here.


Projects from Google 120, to my understanding can/are spun-out as separate entities. Therefore it can’t be shut down by Google management.

this refrain has become the "first post!!1" of the HN community, every time google is posted about.

what does "lolz teh g00gz will shut it down next week!" add to the conversation?

It's shorthand for stating a dislike of Google's reputation for short term thinking around products - whenever there's a new Google launch many people's immediate reaction will be "Looks ace, but is it worthwhile investing the time and energy necessary?". That's a complaint Google should take seriously. It's how lots of people in the tech community view Google. As you say, it seems to be posted every time Google launch something.

There was a great comment by jsafi here on HN a while ago suggesting why Google are like this - https://news.ycombinator.com/item?id=19553840

It's hard to imagine a way they could change it besides stopping launching anything that doesn't have a long term plan in the company.

The reputation of google products being shutdown now preceeds anything they release. It is tarnished, and would never be repaird till they pubically announce that they intend to keep it running (which, of course, they will not).

Therefore, anybody that's not just curious, but is instead going to be invested in using any product of theirs would do well to find an alternative, and invest in finding ways to export to said alternative.

Googles' products can no longer be trusted to remain, and i urge anyone to never rely on it remaining.

A lot, it lets people know that something cherished might be suddenly ripped out from underneath them

Gooooooogle Graveyard, a resting place for great ideas


A fair warning to people who would otherwise get excited. A reminder to people to not tie to Google services, and always have a backup plan because they will get shut down when you least expect it.

It's also a way of sending a market signal. If enough people start hesitating with adoption and promotion of Google services, they might finally eventually notice something is wrong and change their ways.

I’d really like some “x becomes beginner friendly”-filter on HackerNews. Stockholm syndrome perhaps.

Not sure who the audience for this is. Kids I guess? Game developers want engines and tools that make development easier without losing power or flexibility. Players want new interesting gaming experiences.

I'm not sure, I could see myself playing this with friends.

I enjoy playing games with friends that involve challenges,.but often that just translates to outmaneuvering some enemy in some game.

This seems like a great experience where we can really experience each other's creations. Sort of like Minecraft, but imo it's hard to embrace something that "belongs" to someone else, as things tend to be in minecraft.

People who have fun creating, but aren't making a profession out of it. Basically, the digital version of the casual creatives that run D&D games (and create content for those games) for their friends.

Which includes kids, sure, but isn't just kids.

Yes it looks like it's for kids. Basically everything a kid would want from Minecraft, minus the licensed characters I guess.

i dunno i think the survival/mining/crafting aspect of minecraft is a big part of what makes it popular. also i think the "card programming model" and every variation of that sort of thing is such a bad idea. either support real programming or don't, don't make users deal with some junky programming emulation system

> either support real programming

This prominently advertises live (type it in and its an active part of the game with no build/reload/whatever step) JavaScript coding, so it supports real coding.

> don't make users deal with some junky programming emulation system

Excel supports real programming, but people get a lot done it relying mainly on its wide variety of "junky programming emulation" features. People who don't see themselves as programmers often see what programmers see as "junky programming emulation" as accessible tools that don't have (and with familiarity end up lowering) the psychological/emotional barriers actual coding has for lots of people.

> accessible tools that don't have...the psychological/emotional barriers actual coding has for lots of people

I don't think I've ever seen this expressed before; that is, people (in general) having "psychological/emotional barriers" to "actual coding"?

Do you have any links or other references to this; studies, documents, heck - even a general audience article? It would all be interesting to read about. I've seen people struggle with learning (and often giving up) "coding" and "software development", and I am interested in what these barriers may be, etc - and what ways around them, or lessening them, there are.

I've run into this topic tangentially when it comes to children's education (particularly Papert's Mindstorms seems to touch on the issue), but it would be interesting to understand it more when it comes to older individuals.

> I don't think I've ever seen this expressed before; that is, people (in general) having "psychological/emotional barriers" to "actual coding"?

Are you a developer, by any chance? There's a whole academic field of study[1] dedicated to overcome the barriers faced by people trying to develop software artifacts without the need of being trained in a formal system.[2]

[1] https://en.wikipedia.org/wiki/End-user_development [2] http://acypher.com/wwid/FrontMatter/index.html#Introduction

You totally missed minecraft?

Might be me?

Would like to build games and toys, but no longer have the time to learn a proper game-engine.

It’s a shame that making simple games doesn’t require coding like it used to. For myself, and a lot of others I’m sure, games were a huge motivator to teach myself programming. Now kids won’t be as motivated to do that

Why gatekeep though? If we make it easy for people to make games, who cares if they end up learning to program?

No one actually cares about the code in the games they play. Programming should be a means, not the end. (See also: https://prog21.dadgum.com/87.html)

I would have killed to have tools like Unity when I got into game programming. All of my hobby time was eaten up trying to wrangle OpenGL and C++ and crappy open source engines, and cobbling together a shitty physics engine, so I never got around to the actual game part. When I went back years later to prototype ideas in Unity, it was a breath of fresh air.

However, in my struggles to get off the ground, I did learn a lot more about font rendering than I ever wanted to, and it has come up unexpectedly useful.

> I would have killed to have tools like Unity when I got into game programming. All of my hobby time was eaten up trying to wrangle OpenGL and C++ and crappy open source engines, and cobbling together a shitty physics engine, so I never got around to the actual game part.

I'm still in that boat, and I am an 8 YoE "Software Engineer". I recently got into boardgame.io library (also made by a Googler, coincidentally) and it is a fantastic library.

I've attempted to make min-max CLI board game simulations before but failed. Also tried (still somewhat active) to make a 2D artillery game and got close, but just kept getting bogged down in framework (actually migrating from framework to framework) and other chores. Currently it uses LibGDX.

With this boardgame lib I just wrote game logic. It was fun, for a change! Plus it has multiplayer and integrates with React really well for a UI so it's actually useful.

I just wanted the CLI boardgame sims to find the best moves/races in games I play IRL. With this I can actually have other people play and maybe show it off at some point.

Wait, you can’t leave us hanging about how font rendering was unexpectedly useful. What kind of fonts did you deal with later in life and why?

It's not that exciting of a story. ;)

I had to write my own font rendering code for lack of much in the way of general purpose libraries for doing it in 3-D engines back around 2005. Your options were, more or less, to pay through the nose for Scaleform, to roll your own, or to use something really ugly. So I had to learn some basic typography to tie Freetype into a text rendering engine, which got me practical experience with editing font files, anti-aliasing, hinting, kerning, processing digraphs, Unicode, the works.

Web 2.0 started more or less right after that, so interest in digital typography in general exploded. I ended up working on some applications of text layout in non-left-to-right and pictograph languages for 2D canvas, SVG, and web based image manipulation in general. So out of sheer coincidence my hobby work messing with OpenGL/Freetype was suddenly directly relevant.

pretty exciting to me. lots of interesting projects. thanks for sharing.

Quite the contrary - tools like these kinds of gamemaker programs, which factor out knowledge -that previously required a huge amount of individual energy expense- into the social domain, make that strong visceral+interactive portion of computer art more likely to be experienced by more people, leading to the greater likelihood of wanting to learn how to go deeper into those experiences teased by the gamemaker programs. The obvious analogy is how the usefulness of view-source in the earlier days of the internet lead more people into pursuing programming than when you needed some sole purpose programing language compiler or interpreter to see the power of computer code in action.

More people being able to do more stuff is just better. Gatekeeping is rubbish.

It still could introduce people to logical thinking. For me when I was kid I was too stupid and gave up too easily to code anything. It felt like an impossible task to learn everything to even get the basics up and running. And it was my dream to make games as a child. Even modding felt like magic beyond my grasp. I couldn't even make a game with RPGMAKER. Maybe this will help someone like my kid-self find that with a little bit of time they can do it.

But more kids are writing more code now than ever before, there are books on it and classes on it and video series and coding literacy is actually a cultural movement (maybe driven by a need to depress wages by oversaturating the market,) multiple game creation frameworks and the modding scene which is huge. Motivation is clearly not a problem.

And the sort of simple games that don't require coding in these frameworks are the sort of simple games that no one wants to make and no one wants to play anymore. Sure, you can make breakout or pong or a really, really terrible shooter[0] with just the GUI, but kids who want to "make games" don't want those, they want to make the next Minecraft or Fortnite or MOBA or whatever, which means they're going to have to learn to code eventually.

Not everyone has to start with helloworld.c and learning how to write their own renderer from scratch nowadays. But if and when they do want to learn to write code, then they have far more and better options available than previous generations, from asset packs and templates to Handmade Hero and everything in between.


“You can build your own cards with Javascript. Game Builder comes with an extensive API that allows you to script almost everything in the game. All the code is live, so just make a change and save it, no compiling required.”

My friend got into programming by modding Warcraft 3 maps.

In the beginning he was just dragging and dropping random units to the map.

In the end he was quite proficcient in the embedded script editor.

This was true for me as well, before I even learned what programming was, I was already creating custom maps and making scripts through trial and error and lots of visits to google and the wc3 forums. It was nostalgic many years later learning "real" programming as an adult. I think it was a good learning experience and I sure had fun with it.

'triggers' were recently resurrected by the new Overwatch map editor basically just drop down boxes with 100's of conditions... I just don't have the patience when I know I could write this code myself If I actually knew the API

I don’t know... this seems likely easy to get going, but limited by the cards. And then you can jump to JS, so it may be a good gateway?

The block-based visual programming system also looks similar to Scratch. I think kids still pick up valuable experience with basic logic, side-effects and weird cases like non-terminating loops.

I met a 10-year-old who made a cat-and-mouse game with Scratch. He was thrilled to explain to me that in a previous version of his game, touching the mouse would do infinite damage. He needed to throttle the damage function to 10 times/sec to get reasonable scores.

Looking at the examples in there -- what exactly in that is simple to you? It's fully 3D, which is hard as fuck.

I learned a lot of skills doing Doom WADs that were less close to real coding than what this lets you do (full Javascript API).

Yeah, I repeated that as a mantra when I was first learning to program and trying to make games. We all did, it was a common meme of our local game development scene.

And it's utter and complete nonsense. If what you want is to make a game, then these tools aren't half as bad, and at the level of simple games, unless you have some really novel mechanic, you're quite literally wasting time and effort coding it instead of using a game builder.

What drove this point home to me after more than a decade were two situations.

1) on a global game jam, the artist in our team got fed up waiting for us programmers after 3 or 4 hours, and did the prototype we're trying to build in 15 minutes flat, using Construct 2.

2) you too may be surprised to discover that some of the fun and highly acclaimed titles like Cook Serve Delicious or Nuclear Throne were made in Game Maker. I didn't even suspect it until I found it on-line (in the former case, because sources were available in Humble Bundle's Game Maker bundle).

There's value in hand-coding a game, and this is what I still like to do in simple cases, but it's often not the practical option. And if a game maker allows you to add custom code, then really there's no difference between this and Unity or Unreal Engine.

Also, a lot of professional game developers seem to have gotten started with Game Maker as children. One example that comes to mind is Matt Thorsen, creator of Celeste.

Looking at the site, it seems like they also allow for custom scripts if you want functionality that their templates don’t provide.

There are enough gateways into programming that if a child is even remotely predisposed to it, they'll come across it at the right time.

No tutorials on how to get started and the product could possibly be purged in another 3 years. Start with something like Gamemaker if you really want something that will be around for sometime.

based on the article, I think you missed the point. This is a competitor with Minecraft. Not a competitor to Unity.

> No tutorials on how to get started

The system indeed has a tutorial built in.

Applications are open for YC Winter 2021

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact