Out of all the big orgs, Amazon is probably the one that has the least regard for keeping technically interesting projects alive, and the certainly will bulldoze it for some dumb reason when the next re-org comes.
That said, the acquisition by Amazon is interesting in that it seems to overlap with their existing DB offerings? Which makes me wonder what exactly it is that they are getting out of this.
I can't think of many other languages/frameworks where one of the worst places to install it is from the primary sponsor/maintainer.
Of the few examples I have I my head, I'd even say that the fate of a product that has been acquired by Google is probably even better than those of Google's internally developed products. e.g. Waze is still alive and kicking 13 years after acquisition under its own brand and hasn't been completely swallowed by Google Maps. The Nest brand also stuck around for quite some time.
Amazon will drive these people to quit within a couple years over filling out MBRs and threats about how the MBR isn't good enough (MBR = monthly business report, pure bullshit theatre that drives the whole company mad 2/4 weeks every month).
Google would drive them to quit over a longer timeframe, with no threats or harrassment, just because every time they try to ship something cool, someone else has a reason not to do it and that kills you eventually.
Yeah, I've never personally worked at any of those places, but collaborated on a few open source projects with Googlers and Xooglers and the slow grinding wheels of their "processes" that wear people down over time were very apparent. Nevertheless, it has been a breeding ground for many interesting technologies, even if it often suffocates them over the long term.
Wouldn't be possible nowadays to LLM your way out of this?
Really no idea, just asking.
Honestly, the whole thing felt LLM to start with, tons of people lacking context and throwing words and summaries around.
Couldn't have said it any better myself having worked at AWS for close to 8y. I was there during their peak growth years and remember very well how some of these processes that were once not as maddening, devolved.
If I had 5 million I’m going to retire and never code for money again.
I do want to make small video games though, make some music. Pay for a friend’s kids college.
I wouldn’t waste a single extra hour making more money. Usually the golden handcuffs fall off after a year or two. That’s why Heroku went to crap, all the core people left.
Heck AWS is the sole reason all these projects needed to go through these license changes to prevent AWS from completely destroying their business models.
Weirdly I’d put Microsoft above modern-Google, and that’s still a low bar.
You might think that's got a very low ceiling. But, even though it's a bad example in many ways, OpenAI showed that the ceiling is very high. And if you're morally flexible even higher.
When a bigger entity (e.g. AWS) decides to undercut the original creator/vendor (e.g. redis,elastic search), MIT code des not help.
I know some lawyers who specialize in these choices. Everyone thinks their choice doesn't smell, but there are the resources to make informed choices.
There the question turns to "you may sue me, if it breaks" as reasoning. In reality sueing will rarely work, but having a business contract satisfies the company's board and insurance about using the software over an "AS IS"-license alone.
I don't know how you came to OpenAI as an example given that they famously succeeded while going closed-source with the release of ChatGPT.
I can't speak to how successful it is, but Daniel/Curl have been running with this model[0][1] for years. Probably not too different from 'drh and SQLite[2][3].
[0] https://curl.se/support.html
[1] https://rock-solid.curl.dev/
You really underestimate the value of support contracts.
I'm curious what you mean by this. I would have said the exact opposite - AWS tends to keep projects around for a very long time. They haven't acquired very many open-source projects, but the few that they have are all still running as far as I know.
100 items max per Transaction BatchGet 100 items, 16 MB max low write limits on same key Item size 400 KB max etc.
1. You're selling a system to a customer to use within their own AWS account, that you will have no access to, and it needs a transactional datastore (not just an object bucket) of some kind. The fact that it costs nothing by default (particularly valuable when the customer is trying to deploy a proof-of-concept), scales more-or-less perfectly without anybody touching it, requires zero day-to-day maintenance by you or the customer, and all it will ever ask is that you throw money at it, is very, very much a feature. One example I'm familiar with in the wild is Teleport: https://goteleport.com/docs/reference/deployment/backends/#d...
2. You have a huge OLTP workload that fits Dynamo's KV patterns (e.g. Amazon.com shopping carts, which is what it was originally built for). You don't care how much DynamoDB costs (in either dollars or engineering limitations) because any alternative would melt your face off if you even tried.
Most of the pain that comes from Dynamo is people who try to use it as a primary datastore in place of a relational database just to get the serverless pricing model. It's not worth giving up the flexibility on greenfield systems. It does become worth it to give up the flexibility when your system is mature and you don't have genuine flexibility anymore anyway.
The constraints are what let this happen. Unconstraining it might make a better generalist product but part of what you're opting into with DDB is the dumb "put an item in get an item out semantics" and the other side is knowing that it will still work if that volume increases dramatically.
For the same price, you can run a much more capable PSQL instance with way better features, but now you're on the hook for it being up 24/7.
I wish I could get a role to work on OLTP databases. PostgreSQL seems to be a fascinating topic so that's on my plate.
One blog says it is charged by the amount of change logs (but the official doc doesn't say so), which makes sense. But how do I estimate the amount? My hunch is: Put + Write + (1~100) * TransitWrite + Update + Delete + (1~25) * BatchWrite.
For example, I want to know how to calculate the charge of incremental export. One blog says it is charged by the amount of change logs (but the official doc doesn't say so), which makes sense. But how do I estimate the amount? My hunch is: Put + Write + (1~100) * TransitWrite + Update + Delete + (1~25) * BatchWrite.
Well ...
Either way that news is awful. Europeans selling out to US corporations and then wondering why Europe is not competitive in software engineering anymore. Funniest thing is Draghi keeps on saying that; well, recently we also heard that the USA has full access to all national police databases in the EU now. It seems as if Washington remote-proxy-controls the EU. Quite amazing to see, too. The amount of bribe money flowing must be legendary.
Quote from the article:
"As the CWI representative on the DuckDB Foundation... When DuckLabs spun out of CWI, we created this foundation, which holds all IP of open-source DuckDB, and will continue to do so." - Peter Boncz
AWS did not acquire the DuckDB technology itself, which is MIT-licensed open source and governed by the DuckDB Foundation which holds most of the related IP [1].
What they did acquire is DuckLabs, the Amsterdam-based services and development company behind the technology which is owned by and employs the creators and major contributors.
And MotherDuck is a US-based venture-funded commercial company, whose cloud-based data platform is centered around DuckDB but has been significantly expanded recently, including Python pipelines, an agentic context layer, and a visualisation layer.
Currently it says “Today, DuckLabs remains independent and fully owned by the original creators. We've deliberately chosen not to take venture capital, so we can focus on sustainable engineering, correctness, and keeping DuckDB open and MIT-licensed for everyone. “
Future will tell if that ends up being true.
Hah. Never have I ever heard a sentence that described Amazon less.
Not that I like either actor in this particular play, but this isn't exactly an example of community oriented goodwill on Amazon's part.
Open search wasn’t precisely born out of the kindness of their heart and their defence of FOSS values… it’s more like a predatory practice to capture market.
They used the software in a way that complied with the license. When the license changed to no longer suit them for future releases, they forked from the original license.
Remove the names here, and in context, every single person familiar with open source projects and licensing would think this is a nonissue. They used the software as the license was intended, and exercised the rights given to them by that license. That freedom to fork if they don't like where the project is going, or things change and the project is no longer viable for them with the new license, is one of the key freedoms enshrined in the whole FOSS movement.
Yes I think motherduck is going to get ... well motherducked bad by AWS
>Independence: We are fully founder-owned and not externally funded. Our priorities align with the long-term health of the project and its users.
I was really surprised when I first came across DuckDB at how good at it is for its target use cases. It is a game changer for me for the "local analytics" space, and its ability to scale up to a large degree helps a lot.
It is simply awesome to be able to point DuckDB at a mess of CSV and other files and have an instant database on top of it that I can run regular SQL over, and it is fast and just works.
I hope they just let the team carry on and don’t contaminate it with all the other craziness going on.
These kind of arrangements generally get the investors their money back plus a modest return. The employees generally don’t get much, relative to the risk they took.
Buying the team and having them all quit is such a bad outcome for AWS, especially they would not keep any of the IP, which is owned by the DuckDB Foundation. I don't think they would be that short-sighted.
Is profit sharing, like what you're implying, actually happening? If so, in what quantity?
I can’t say that’s true for me and my company. It’s not all lies or anything, it’s a fragmented view that emphasizes conflict and the most visible 10% of what the company actually does.
What big corp is not a total mess?
The things I hear from recent AWS departures are very concerning.
On my team we were expected to work 90+ hours a week, and were told we were slackers if we were unwilling to work Sunday afternoons. But I've heard from others that some teams are a lot better.
It's both a great time to be gold mining in tech, and also a terrible time to be a bit employee.
Maybe this will be the rare case where founders survive corporate shenanigans and keep on doing their thing. But I'm bummed, as I've heard that optimism many more times than I've seen it happen.
From my own experience I can say it integrates far better into your Rust app than DuckDB does.
Well over 100 monthly contributors, too.
Indexes, transactions, a first-class storage format are all things that come included with DuckDB, that you won't have with Datafusion.
DataFusion is being used as a building block for a growing number of databases and data processing engines in Rust, as it offers the necessary primitives.
My guess is that Amazon wants official hosted versions and doesn’t want to go through something like the Redis fiasco with licensing. In that case, they ended up having to support their own development anyway (with Valkey), so they might as well just buy the team.
DuckDB team probably gets a nice package and pay bump, but it’s really unlikely they’re getting hundreds of millions from this.
Assuming historical returns, your money doubles roughly every 7 years, so within the rest of your lifetime, that 1 million should turn into at least 8. That's an extremely comfortable upper-middle-class lifestyle on the interest payments alone. If your children don't spend it all, your grandchildren would easily have private jet money by the time their parents retire.
If getting a million dollars wouldn't affect how much money you can leave your kids, you already have generational wealth.
Well said.
Where I live, one million dollars would allow me to pay off my house, open healthily sized investment accounts for my kids, pad my investment account, setup a trust and, overall, set my family up for a comfortable life in the future. I don't see how that isn't generational wealth.
Basically the bar is higher than "something you can pass down". It is enough that the next generation does not need to worry about making money either.
We can debate “live comfortably” if you want, but no 4.5 people are doing that from the investment proceeds of 1m.
What you are describing is kind of more like social mobility.
Generational wealth is definitely not "affect[ing] how much money you can leave your kids." That's an equivocation - if you're leaving your kids a dollar, another dollar will "affect how much money you can leave your kids."
edit: you need a million dollars to securely retire at all, and that's if your parents, kids, or you don't get sick. If they do, a million is not only not "generational wealth" but it may not even last you three years.
At a more sane expected return of 5% annually, you get $50k a year to live off of to just keep what you have (or rather, watch it slowly erode in value due to inflation).
That is basically a one person income, maybe two adults if you pinch pennies and live in a crappy apartment or a low-end house in a midwestern suburb.
You can get a lot more lucky with a million bucks than you can with 10k if you gamble, but there are no guarantees. Risk tolerance is the most impactful variable. For those in the low risk tolerance group, I think you'd need at least 2 mil these days. And that's with being frugal, as well as probably not having much left over for kids.
You would also have to train your kids to responsibly use the money without demonstrating it, as you’d just be hoarding it. People don’t have a good track record there, either.
I think the best you could do generationally with a million would be to try to invest moderately and draw down a small percentage (2-3%) to demonstrate fully considered use of the money. This would keep you in a middle class income but give you more ability to donate charitably, vacation together, let one spouse retire earlier, solve a financial crisis for a child, etc.
Letting them in on the thinking would give them a good chance to handle a high six/low seven figure inheritance, depending on how the investing goes.
Common definition is somewhere between 1.5 to 5 million per child.
You need more than $1.5m today should be $6m by the time they're 20. Depending on inflation, cost-of-living, tuition, etc. that might be enough.
People talking about generational wealth aren't saying, "what if they live a frugal life in the Topeka suburbs."
Also wealth does not start at the ability to live an extravagant life without a salary... You're wealthy way before that point.
Technically any asset transferred to the next generation is “generational wealth”.
But in the context of a startup sale, when you say generational wealth almost nobody assumes you mean the ability to pass a few thousand down.
The common understanding is that you have enough money that future generations do not need to worry about money, assuming they maintain an average or slightly above average lifestyle and use the money responsibly.
Speaking for yourself I guess
Buying a house in desirable areas is going to be in the ~ $1M range, college is going to be hundreds of thousands, etc.
MotherDuck, a semi-competitor of theirs has raised $100mil in funding. DuckLabs has reportedly taken no external funding (so all ownership is with the founders) and was profitable with 30+ employees.
On the sidelines, other companies behind beloved open source products, like Astral, Astro, Bun are getting bought left and right.
With that as a backdrop, I think they should have been able to get quite a good payout.
The DuckDB foundation owns some equity in MotherDuck, which is a data lakehouse platform based on DuckDB, and all three have been moving closely together in making DuckDB better locally as well as in the role of a query engine that really threatens a lot of amazon's role in data lakehouses. DuckDB in the hands of a good team means that you can use AWS almost only for storage, instead of using any of the managed services.
I know the projects will remain open, and ostensibly still contributed to in the same direction, but surely AWS thinks of this as another piece of a product suite to compete with Databricks.
Nontheless, I think the open embedded query engine approach DuckDB is spearheading is larger than one project, and I remain excited about the broader ecosystem (especially datafusion).
"DuckDB has been an in-process database since day one. But people have asked us – very persistently – for a client/server mode, and we have finally caved"
This was already satisfied by numerous projects and products, and feels like a "me too" attempt to capture AI-based workflows. DuckDB always felt like "SQLite for Analytic Data" but I fear these changes and now acquiring the org leading technical direction is where they deviate for good. AWS is so unnecessary for what DuckDB can (and should, IMO) be; MongoDB jumps to mind as a cautionary tale.
Are databases not a solved problem? Why are there lots of different databases? Why is one faster than the other? What's different between them?
My attempt at an answer: no, databases are not solved.
Specifically, different databases are better or worse for different use-cases.
For example, Postgres is a great "all around database" - you can use it for a lot of different things. As a "relational" database, it's really good if you have a table full of users, a table full of order, and you want to see all orders made by a user with ID=123. You need to answer questions like that a lot (eg every time someone on a website loads a page) and you need the answer fast (hundreds of miliseconds at most)
However, say your use-case is more like... you've got 100 billion rows of billing data ("joe was charged $123.45 on 2026-03-07 for a shirt, blue, size 11, brand foobar") in one table. You don't care much about joe, but you want to be able to find out how much was billed, total, in 2026-03 for blue shirts (or all year for brand foobar, or all time, for size 11). Postgres would struggle with data of that volume - you'd need a really big expensive database. A "columnar" database like duckdb (or clickhouse) might be able to answer those questions better.
Anyway, different databases are better/worse for:
- Large piles of data that you need to query in seconds
- Huge (petabytes) of data that you need to query in minutes, but can query in parallel
- Many related piles (like a standard relational database)
- Cases where you're mostly getting or retrieving single items (key-value stores)
- Huge piles of data that represent a long stream of events in time (time-series datbases)
- Piles of data that look and act more like files (object stores)
- When you need strict transactions
- When your need is very write-heavy
- When your need is very read-heavy
- and probably many others - I'm not even a huge data guy :)
So it all depends on your use-case. There are still cases that are not served well by any existing database - eg "filtering billions of rows, in milliseconds, by an arbitrary portion of several dozen very-high-cardinality columns" (to use an example that came up recently for me IRL) :)
Check this curated list of databases from Carnegie Mellon University: https://dbdb.io/
There's no single set of requirements and desired properties that people have for databases.
What queries does it accept? How does it persist data? How does it manage replication and partitioning across multiple servers? Are questions with many answers and the right one varies by application.
nothing prevents to build single database which would cover all such answers. Its engineering, funding and distribution problems: no-one built it yet.
Even if you did it, I'd be surprised if a single code base is optimal across the spectrum from resource constrained microcontrollers all the way up to IBM Z Series mainframes and everything in between; along with the full spectrum of persistence from in-memory MRU (or similar) caching to full two phase global consensus as well as optimistic eventual consistency.
There's just too much scope to be the best in class at everything, and even if somehow you did it, the maintenance cost of all the options would be huge.
Best you'll get is as computing continues to increase, the cost of using less than the best becomes more reasonable. There's a lot of database applications where any reasonable database works, and there's a lot of databases that are reasonable in wide application. That may well lead to fewer databases being available, but it's unlikely to converge down to a single database. Just as most engineering domains don't converge down to a single solution for all applications.
is a marginal niche, but everything else can be handled by one DB. You think we have PG with extensions (including duckdb) already covers most of the ground, making something like that distributed is also feasible task (there are projects), architecting system to make core engine embeddable is also feasible task.
There are no technical reasons for having more than one codebase, only economical and coordination-related.
Every tool is a trade off between effort to create vs power of the solution.
Effort is generally expensive so most things settle on some general purpose local maximum. If you had infinite effort available, you could build bespoke hardware and software from the ground up to solve every problem. It would be faster and more power efficient than any solution available today.
CPUs win out over integrated circuits because the same CPU can be used for ~every software problem, so by using a CPU you benefit from everyone pooling their efforts to improve the general purpose CPU rather than their own specific niche. But when you reach a certain scale/requirements it makes sense to do something more specific. This is one reason why we have standardized GPUs. Still general purpose but more specialized than a CPU. Or think about how Bitcoin mining moved to ASICs, because they need to do one specific thing as fast and as power efficiently as possible.
So for databases, when you get to specific scale and requirements the same kind of specialization starts to make sense. DuckDB or Clickhouse for analytical loads, TigerBeetle for high scale transactional stuff, etc. And that scale is aggregated across ~all software users, i.e. scale of analytical workloads being big enough to support analytical DBs.
Also as time goes on and industries develop the cost to develop specific solutions can go down.
It’s not a solved problem because each iteration of technology doesn’t just fix the mistakes of the past, it’s an evolution to solve the problems of the present.
"Are cars not a solved problem? Why are there lots of different cars? Why is one faster than the other? What's different between them?"
I think the best way to approach the subject in an easy to grasp way is to ask Gemini or another frontier AI to teach you the basics, they will do a surprisingly good job and they'll be able to react to your questions with INFINITE patience.
What do you mean by solved problem? I don't own DuckDB (sqlite, postgresql, etc). If I think I can create something as good or better than DuckDB, should I give up doing so (and get filthy rich with an acquisition) because someone thinks databases are solved? Solved databases aren't mine.
Why do we need a source distribution for a well regarded MIT licensed project? Because it's not easy to contribute code to DuckDB if you don't work at DuckLabs. The CI used to take 5 hours for a simple bug fix last I looked (may have improved since I flagged it on social media).
There are two that I'm aware of:
* Haybarn: https://rusty.today/blog/duckdb-extension-distribution-gap/
* Pygmy-Goose: https://github.com/Pygmy-Goose/pygmy-goose
KuzuDB folks worked on something called GRainDB in 2022: https://vldb.org/cidrdb/2022/graindb-a-relational-core-graph...
But circa 2023 decided to write their own. Work continues as LadybugDB. One of my long term goals is to find an integration point with DuckDB's table implementation as the "node table". Conversely at some point DuckDB could implement all the join algorithms and LadybugDB's REL table in their code base.
For now, any talk of Graph on DuckDB is limited to DuckPGQ and the more recent entrant DuckGQL (both of which don't touch the storage layer which is the main reason why LadybugDB exists).