There's going to be a lot of complaints about open-source restricting access.
It's going to keep happening because it just doesn't make sense for a lot of previous business models that supported and open-source project, something that was seen recently with tailwind.
In one of my projects, one that remains source-available, I had encountered an "open-source justice warrior" that made it their mission to smear the project because of the switch, grasping at straws to do everything they could to paint my intentions as malicious.
It's really too bad, and will only hurt the availability of free alternatives if one cannot provide the source under a "just don't commercially compete with the paid version of the product" license without getting branded as a scamming cash grabber
They still have linked their OpenCollective account, where they have raised $10K and still have a balance of $5K. [0]
It's not a lot in the great scheme of things, but, have they been using a platform that's seemingly built for communities and open source to bootstrap their business?
Because this is not a 'open core' situation. They just closed the repo and ran away. If they had that idea all along, I feel like it hasn't be very, let's say, ethical.
Wait, so a company shared their work with the public for however long, then decided to leave what was shared up ... but stop sharing ... and you're upset?!?
They did everything properly by the rules of OSS, decided it wasn't in their best interest to keep doing OSS, and left all their code available, as required by OSS. They were a textbook good participant.
Meanwhile, 99% of companies never open source anything: why aren't you complaining about how "unethical" they are?
How can people still not understand that OSS can be abused?
It doesn't matter that the previous code is still available. Nobody can technically delete it from the internet, so that's hardly something they did "right".
The original maintainers are gone, and users will have to rely on someone else to pick up the work, or maintain it themselves. All of this creates friction, and fragments the community.
And are you not familiar with the concept of OSS rugpulls? It's when a company uses OSS as a marketing tool, and when they deem it's not profitable enough, they start cutting corners, prioritizing their commercial product, or, as in this case, shut down the OSS project altogether. None of this is being a "textbook good participant".
> Meanwhile, 99% of companies never open source anything: why aren't you complaining about how "unethical" they are?
Frankly, there are many companies with proprietary products that behave more ethically and have more respect for their users than this. The fact that a project is released as OSS doesn't make it inherently better. Seeing OSS as a "free gift" is a terrible way of looking at it.
“Open core” is when part of the product is open-source and part is private.
Was a significant part of the product private before this announcement?
If not, someone can fork the repo and immediately launch a competitor (FOSS or paid). (Technically even if so, except it wouldn’t be immediate, and if they’d have to re-implement too much, it would be easier to start from scratch.)
Yes there were significant portions that were proprietary before this, including support for some services.
The parts that were open source might still be worth forking, but you would probably need to change every occurrence of the name to avoid trademark issues.
I evangelized localstack at my company a while back, but as we integrated it deeper into our CI test runs we started running into more and more things they don't support, and it feels impossible to get any attention from their support/devs despite being paying customers.
Their Cloud Pod and ephemeral instance features in particular feel pretty half-baked and not very useful at the moment.
Fun tangent: it's pretty easy to write a crack for the pro version; we actually used that for about a month as a pilot to confirm that it would do what we needed it to.
I too was excited about the idea originally but then started realizing that they will have an increasingly untenable service area to try and maintain and mimic and it was just never going to work out.
First minio and then localstack, as an open source maintainer I find that abandoning their community is bad faith. I totally get wanting to monetize but removing the free product entirely feels like such a betrayel.
Luckily, I've been vibing with Devin since this started having it build a cleanbox emulator on top of real s3 tuned for my specific use case. It's a lot less general but it's much faster and easy to add the sort of assertions I need in it. It's no localstack but for my limited use case it works.
More reason to run your infrastructure using open source software in your own datacenter. OpenStack has been around for closing in on two decades, running clouds and being mostly governance-drama-free.
It's not surprising that a proprietary ecosystem built on open source software locked up behind a gate doesn't make a worthwhile ecosystem for building open source tooling against.
Running OpenStack for this is a massive project cost compared to spinning up a few local services, and the operational mess is on a different planet from "I need to fake a handful of API calls on my laptop". Self-hosting still means updates, drivers, and k8s/OpenStack glue code. Nobody sane are doing that for local dev, use Minikube or Podman if you want DIY and still like weekends.
It didn't support the one thing I wanted but it was so easy to find the right place in the code, I was happy. Never got to continue it though or turn it into a PR
For S3 emulation, I'm using rustfs. It's very compact and fast to run, and you can just start it with `docker run` inside tests if you don't want to set up a full integration test harness.
I used an SQS-on-top-of-Redis emulation before, but I can't recommended it now (no updates for 6 years).
Complete coincidence but today I was looking for an AWS mock for E2E tests. Not the whole AWS footprint but just a few services and looked at LocalStack for the first time.
It took Claude to put together a service (with web interface and everything) for those 2 services 15 mins.
I’m not claiming my experience is translated universally but perhaps if your core competency is something like LocalStack you need to think about alternative business ideas.
Well LLMs are trained on code like those from Localstack, and a lot of them can be emulated to first order as CRUD operations, so its rather unsurprising. It does mean that things do become difficult for pure tech SaaS businesses like this one, and as also seen with Tailwind.
IMO, the trajectory was set back when they removed the option for monthly payments. Minimum US$450 to play made it a non-starter for my projects (even with commercial ambitions). They changed this just as I started to integrate (~2024, I think) so I kept to the free capabilities. Have been waiting for the other show to drop and here we are.
Edit: looks like they’ve reintroduced monthly billing within the last few months. I guess that’s a sort of win, even if not for the OSS community. But I’d still be reluctant to get into bed with them at this stage.
There's going to be a lot of complaints about open-source restricting access.
It's going to keep happening because it just doesn't make sense for a lot of previous business models that supported and open-source project, something that was seen recently with tailwind.
In one of my projects, one that remains source-available, I had encountered an "open-source justice warrior" that made it their mission to smear the project because of the switch, grasping at straws to do everything they could to paint my intentions as malicious.
It's really too bad, and will only hurt the availability of free alternatives if one cannot provide the source under a "just don't commercially compete with the paid version of the product" license without getting branded as a scamming cash grabber
They still have linked their OpenCollective account, where they have raised $10K and still have a balance of $5K. [0]
It's not a lot in the great scheme of things, but, have they been using a platform that's seemingly built for communities and open source to bootstrap their business?
Because this is not a 'open core' situation. They just closed the repo and ran away. If they had that idea all along, I feel like it hasn't be very, let's say, ethical.
--
Wait, so a company shared their work with the public for however long, then decided to leave what was shared up ... but stop sharing ... and you're upset?!?
They did everything properly by the rules of OSS, decided it wasn't in their best interest to keep doing OSS, and left all their code available, as required by OSS. They were a textbook good participant.
Meanwhile, 99% of companies never open source anything: why aren't you complaining about how "unethical" they are?
How can people still not understand that OSS can be abused?
It doesn't matter that the previous code is still available. Nobody can technically delete it from the internet, so that's hardly something they did "right".
The original maintainers are gone, and users will have to rely on someone else to pick up the work, or maintain it themselves. All of this creates friction, and fragments the community.
And are you not familiar with the concept of OSS rugpulls? It's when a company uses OSS as a marketing tool, and when they deem it's not profitable enough, they start cutting corners, prioritizing their commercial product, or, as in this case, shut down the OSS project altogether. None of this is being a "textbook good participant".
> Meanwhile, 99% of companies never open source anything: why aren't you complaining about how "unethical" they are?
Frankly, there are many companies with proprietary products that behave more ethically and have more respect for their users than this. The fact that a project is released as OSS doesn't make it inherently better. Seeing OSS as a "free gift" is a terrible way of looking at it.
Because this thread isn't about those other companies.
“Open core” is when part of the product is open-source and part is private.
Was a significant part of the product private before this announcement?
If not, someone can fork the repo and immediately launch a competitor (FOSS or paid). (Technically even if so, except it wouldn’t be immediate, and if they’d have to re-implement too much, it would be easier to start from scratch.)
Yes there were significant portions that were proprietary before this, including support for some services.
The parts that were open source might still be worth forking, but you would probably need to change every occurrence of the name to avoid trademark issues.
yes, there were a large number of AWS products and features that were only available with a subscription
I evangelized localstack at my company a while back, but as we integrated it deeper into our CI test runs we started running into more and more things they don't support, and it feels impossible to get any attention from their support/devs despite being paying customers.
Their Cloud Pod and ephemeral instance features in particular feel pretty half-baked and not very useful at the moment.
Fun tangent: it's pretty easy to write a crack for the pro version; we actually used that for about a month as a pilot to confirm that it would do what we needed it to.
I too was excited about the idea originally but then started realizing that they will have an increasingly untenable service area to try and maintain and mimic and it was just never going to work out.
First minio and then localstack, as an open source maintainer I find that abandoning their community is bad faith. I totally get wanting to monetize but removing the free product entirely feels like such a betrayel.
Luckily, I've been vibing with Devin since this started having it build a cleanbox emulator on top of real s3 tuned for my specific use case. It's a lot less general but it's much faster and easy to add the sort of assertions I need in it. It's no localstack but for my limited use case it works.
More reason to run your infrastructure using open source software in your own datacenter. OpenStack has been around for closing in on two decades, running clouds and being mostly governance-drama-free.
It's not surprising that a proprietary ecosystem built on open source software locked up behind a gate doesn't make a worthwhile ecosystem for building open source tooling against.
Running OpenStack for this is a massive project cost compared to spinning up a few local services, and the operational mess is on a different planet from "I need to fake a handful of API calls on my laptop". Self-hosting still means updates, drivers, and k8s/OpenStack glue code. Nobody sane are doing that for local dev, use Minikube or Podman if you want DIY and still like weekends.
An emulator for integration testing against the major cloud providers seems like it should:
1. be table-stakes for a SDK from the cloud providers themselves
2. have the obvious home in a foundation like the CNCF; how else could you be "cloud native" afterall?
What are the alternatives? I primarily used it for S3 and SQS emulation.
I haven't used this yet, but there's also https://github.com/robotocore/robotocore. It came up on HN a few days ago: https://news.ycombinator.com/item?id=47420619.
It's not a complete replacement, but if you're in a Python ecosystem, you might find Moto to be of interest.
https://github.com/getmoto/moto
at least some parts of localstack seemed to be built on moto, based on a brief look at some point
I haven't evaluated it deeply yet, but I saw https://github.com/hectorvent/floci
I am about to test this one http://docs.getmoto.org/en/latest/index.html
I've fiddled around with https://docs.getmoto.org/en/latest/docs/server_mode.html
It didn't support the one thing I wanted but it was so easy to find the right place in the code, I was happy. Never got to continue it though or turn it into a PR
MinIO is a drop in replacement for S3. I plan on switching to this as soon as I can. For now, I just pinned localstack to 4.14.0
>MinIO
I have some bad news for you: https://news.ycombinator.com/item?id=47000041
It's comedic that they said that right after Minio pulled the same thing as localstack.
RustFS is a good and simple-to-usr alternative for MinIO.
For S3 emulation, I'm using rustfs. It's very compact and fast to run, and you can just start it with `docker run` inside tests if you don't want to set up a full integration test harness.
I used an SQS-on-top-of-Redis emulation before, but I can't recommended it now (no updates for 6 years).
Complete coincidence but today I was looking for an AWS mock for E2E tests. Not the whole AWS footprint but just a few services and looked at LocalStack for the first time.
It took Claude to put together a service (with web interface and everything) for those 2 services 15 mins.
I’m not claiming my experience is translated universally but perhaps if your core competency is something like LocalStack you need to think about alternative business ideas.
Well LLMs are trained on code like those from Localstack, and a lot of them can be emulated to first order as CRUD operations, so its rather unsurprising. It does mean that things do become difficult for pure tech SaaS businesses like this one, and as also seen with Tailwind.
Try proxymock. It's not open source but it is free to use.
Did localstack never get bit enough that a fork would emerge or am I missing an obvious one?
"You either die a hero, or you live long enough to see yourself become the villain"
So is local stack dead? Is this situation the lesser evil? Or is it not dead and we will see a villain rise?
Edit: I see now, they have commercial offerings: https://www.localstack.cloud/pricing
I am not sure if my corp will be willing to pay or tell us to find something else, but I use it everyday, our integration tests depend on local stack.
IMO, the trajectory was set back when they removed the option for monthly payments. Minimum US$450 to play made it a non-starter for my projects (even with commercial ambitions). They changed this just as I started to integrate (~2024, I think) so I kept to the free capabilities. Have been waiting for the other show to drop and here we are.
Edit: looks like they’ve reintroduced monthly billing within the last few months. I guess that’s a sort of win, even if not for the OSS community. But I’d still be reluctant to get into bed with them at this stage.