We don’t live in a perfect world, and the priority should be preventing Google from having complete control over the internet. If you have any better ideas for how to do that, please let us know.
Can’t speak for everyone else but I just think that getting people to change their OS is much harder than getting people to change their web browser, so it makes sense to start there.
Chromium is in that awkward position where it is technically an open-source community-driven project but in practice operates like a closed-source one. Google has complete control over decisions and can make changes that serve their profits (see: JPEGXL debacle).
Of course, you can make the same critique of Firefox, but at the very least they are a nonprofit foundation and don’t have enough market share to abuse their power.
1. Of course there is no such thing as flawless data analysis, but they have data and people hired to analyse it and we… don’t.
2. RH makes their money of off companies that pay for support. Do you think that if a company can’t use CentOS they’re going to pirate RHEL? Really?
3. Conclusions from leadership are never perfect, but that does not mean the default conclusion should be that they are awful. RH seems to be doing pretty well at the moment.
Opportunity cost is a central concept in economics. When deciding between decisions A and B, companies will consider the revenue they would gain from either decision. If companies accept the revenue from decision A, they are doing it at the cost of the revenue they would have made from decision B. This is what I mean by “lost revenue”. By choosing to shut down git.centos.org, RH loses the revenue they would have gotten from keeping the repository open and gains the revenue they believe they will get from shutting it down. My apologies if this sounds condescending, but you keep insisting that there’s no such thing as “lost revenue” and I genuinely don’t know how to respond to that without just explaining opportunity cost.
RH presumably has access to far more detailed statistics on usage and adoption of CentOS and RHEL than we do. Many of the execs at RH have been at the company for decades, and so will have been observing these statistics over the years and through various transition periods.
I think it takes a lot of hubris and contempt for RH staff to know this and think that clearly you can see something they don’t. RH have concluded from the data they have that clones are siphoning away their profits. Until RH starts taking a financial hit from this (so far they are doing great!) I’m inclined to believe they know what they’re doing.
I don’t understand why people keep claiming that the advantages of these clones (expanding the ecosystem) outweigh the disadvantages (lost revenue) to Red Hat. Surely, RH have access to far better statistics on this issue than we do, and they have obviously concluded that this is not the case.
Wow, another blog post on the same two changes everyone else has been talking about for two weeks, rehashing the exact same arguments.
I genuinely cannot think of a company that contributes more to open-source projects than Red Hat. They to do this through hiring hundreds of talented engineers and adhering to an incredibly generous upstream-first policy, where most of the work they do gets made available for everyone else. No matter what Linux distribution you use, you are running a huge amount of code written by Red Hat.
Unfortunately, because we don’t live in a utopia, they need to be able to pay their engineers to do this work, and clones that simply redistribute their packages pose a serious threat to their ability to do this. Previously, Red Hat went out of their way to make the exact RHEL sources available to these clones. Now, Red Hat is still publishing all the code they write through Stream, but they do not have a specific repo for the exact RHEL patch set. I don’t see how this isn’t completely reasonable.
If you led a company making a product, and were responsible for thousands of employees, and some other company simply took that product and repackaged it, released it for free, while (in the case of Rocky and Oracle) running ads to compete with you in the process, are you telling me you would go out of your way to publish the exact patchsets to a special repository to make it as easy as possible for them to do this?