An #OSS tool you use in your #homelab has a new update! The release notes talk about new "Enterprise license" and "reorganizing" existing features into free and paid versions. What do you do?

Bonus difficulty: Assume the tool stops working if you try to block spying/telemetry.

#boostswelcome #enshittification #opensourcesoftware #opensource #selfhosting #selfhosted #homelab
Replace it immediately
45.5%
Keep the old version forever
14.3%
Upgrade until it eats a required feature
26%
Fork it and hang on
14.3%
Poll ended at .
As far as keeping the old version, for my specific use of this specific tool it does not have strong security implications. The tool itself can bring a very large exposure, but my usage is effectively read-only.
Looks like it is getting replaced. (I'm disappointed in the 44 of you that want to give a hostile company tons of free labor. Do better. โ€‹โ€‹)

If anyone else uses
#Botkube and missed the announcement, sorry to be the bearer of bad news: github.com/kubeshop/botkube/releases/tag/v1.9.0

I already implemented about 60% of the "event watcher" pipeline in
#nodered so maybe I'll just finish that out. I needed nodered anyway because Botkube's webhook "support" is total amateur hour. If I had to guess, I would say it is an internal structure they just throw at the endpoint with no documentation. It changes, sometimes dramatically, from version to version.

Anyone know of a better event monitor for
#k8s #homelab #monitoring? Things like "{sts/deploy} not progressing" or "crash-loop backoff". I've got alertmanager, but it has gaps.

(edit to say '44' since there are 40 people who want to kick it down the road until it becomes an emergency)
Release v1.9.0 ยท kubeshop/botkube

What's Changed Botkube is introducing new plugins with advanced functionality that will be part of the Botkube Team and Enterprise packages. These advanced plugins require cloud services provided b...

GitHub

@dis How does upgrading equal free labor for the company? It doesn't say contribute.

This will keep happening until we find ways to fund open source. https://mastodon.social/@jacob@jacobian.org/111914181286275856

@jgillich

"Fork it" and (to a lesser degree) "stay at the old version" both involve doing maintenance for free on someone else's commercial software.

That isn't "funding OSS", that is "volunteering for a VC firm".
@dis How does the fork benefit the company? I'd think the opposite, it adds competition for their (now paid) offering.
@jgillich
How do you fix a bug in the fork without contributing it to the commercial offering?

How many users are going to whip out a credit card to call upstream for support when they can just open a ticket with the "good" version for free?

At least 22 people volunteered to single-handedly take over all development and support tasks. (44 to my view, but I could be convinced that "sit on your thumbs" is more like "wait for it to break".)

For a real life example, check out what
#Mirantis did when they destroyed Lens. github.com/MuhammedKalkan/OpenLens/
GitHub - MuhammedKalkan/OpenLens: OpenLens Binary Build Repository

OpenLens Binary Build Repository. Contribute to MuhammedKalkan/OpenLens development by creating an account on GitHub.

GitHub

@dis I don't think happens as much as you think.Take Nextcloud and ownCloud, they are now very different projects, I don't think they exchange patches anymore. Maybe initially, but eventually projects diverge too much for that to be possible.

The company also likely funded most of the development before the fork, so IMO this isn't that big of a deal. They have a right to use the code under the terms of the license. Using a copyleft license is a solution if you want to prevent this.

@dis Mirantis was pretty smart about dismantling Lens, it was a slow burn. First, they build additional services, then they take existing features and make them proprietary, then they silently stop publishing source code. By the time people noticed it was proprietary, it was already too late to fork.

But we also have to acknowledge that most users are perfectly happy to leech off other people's free work, and don't actually want to fork themselves.

@dis
I don't like how Mirantis went about this, and I actually started building my own Lens replacement in response (https://github.com/getseabird/seabird). But we have to acknowledge that someone's gotta pay to make development happen. Relying on volunteers alone is not sustainable, maintainer burnout is a serious issue in open source. https://blog.tidelift.com/maintainer-burnout-is-real

Heck, Mastodon gGmbH couldn't even afford to pay their CTO. Open source is just as unsustainable as ever, few projects actually have good funding.

GitHub - getseabird/seabird: Native Kubernetes desktop IDE designed for seamless cluster exploration

Native Kubernetes desktop IDE designed for seamless cluster exploration - getseabird/seabird

GitHub
@dis For now I'm voting "Big oof". Would have to think real hard about that.
@dis It depends. Are there any good alternatives? Then I'll try to replace it, if not then I'll stick to the old version.
@dis Support upstream fork, also immediately make local patchset to remove telemetry checks and spam their telemetry with garbage data.
@AMS The problem I have with that branch of answers is the unspoken part: "Take over most of the maintenance and support tasks for free, while allowing them to claim any actual fixes and features for the paid version." I just don't get how so many people seem ok with doing work for a company they hate, for FREE.

As far as telemetry, I should have said "it requires full internet access, so the effort of identifying and blocking spyware without hobbling the tool is very large."
@dis Yeah, the company getting free labor still is a huge problem. I have no moral issue with trying to keep local patches to remove license checks and send garbage telemetry though until I find a replacement. That's what they get for pulling that shit.