A fork is a point in time clone, after which the original and the fork do their own things (which may include the fork getting new/changed code from the original and/or vice versa). A fork is not a perpetual downstream dependency.

Oracle claims OL is not a fork. If they are truly only providing additional hardware support and are otherwise perpetually staying up to date with RHEL source (which they claim to be built from), then they aren't really a fork. They definitely claim to have started with RHEL, not CentOS. Since they claim 100% compatibility with RHEL, that makes sense. And there's no real reason to get the source from anywhere but RHEL, since it's freely available.


________________________________
From: [log in to unmask] <[log in to unmask]> on behalf of Stephan Wiesand <[log in to unmask]>
Sent: Monday, February 3, 2020 15:12
To: David Sommerseth <[log in to unmask]>
Cc: [log in to unmask] <[log in to unmask]>
Subject: Re: EL 8

Caution:  EXTERNAL email

> On 3. Feb 2020, at 21:11, David Sommerseth <[log in to unmask]> wrote:
>
> On 01/02/2020 17:12, Nico Kadel-Garcia wrote:
>> *No one* calls it "Oracle 8". It's still RHEL 8. Oracle now owns and
>> can still use the Red Hat trademarks.
>
> No, not at all.  It was IBM who acquired Red Hat; but IBM has so far kept Red
> Hat as a separate company/brand with its own organization.
>
> And Oracle Linux is essentially a fork of CentOS, so it's even one step
> further down on the "downstream ladder".
>
>   Fedora -> RHEL -> CentOS/SL -> Oracle Linux

Sigh, we had this meme before, and by repeating it like a mantra it won't become less false.

If Oracle Linux was a CentOS fork, how could they possibly outperform CentOS in getting stuff out by *months*?

And that's not the only respect in which OL compares quite favourably. Just look for updateinfo.xml in repo metadata, or the SRPMs for the latest updates.