mostly towards oracle linux, cloudlinux and others that charge support and fees to the os.
Perhaps, but that’s not consistent with what they say they’re targeting. “Simply rebuilding code without adding value or changing it in any way” isn’t what Oracle, at least (I’m not familiar with Amazon’s incarnation), are doing–but it is what Rocky and Alma (and CentOS before them) are/were doing.
I think what they’re doing is permitted by the GPL–that license says, in short, that if they give you the binary they must also give you (on request) the source. And they’re doing that–if they’ve given you the binary, you have access to their portal from which you can get the source. I don’t think it meets the spirit of the license, but it probably meets the letter.
Indeed it doesn’t–but it also doesn’t say anywhere that they have to give it to you in the first place.
…but this is what I’m seeing. Apparently they’ve decided to follow the Bud Light School of Marketing.
I’m thinking by now that the title of this post is somewhat contrary to the content!
My 2 cents
For the latest version of AmazonLinux, they are now using Fedora. Unsure if AWS has procured a license or subscription for this or not.
I read an article earlier today that RHEL has complied with GPL, although some may debate whether they complied in the practical/legal sense of the license or with the spirit of the license.
Either way, I kind of do see RHEL’s point of the rebuilders using their software and that they aren’t receiving enough back from them either in terms of maintenance contributions or in financial terms.
RHEL’s response in pulling the repos is a bit of a knee-jerk reaction, but considering how the costs of producing and maintaining the software has been increasing over the past couple of years and especially over the past 20 months, I am hardly surprised.
So do I, but it seems to overlook the fact that they’ve built their business on other people’s software. And I’m aware that they do contribute back to many upstream projects, but I wonder how Linus, for example, feels about the equities here–and I can’t believe they contribute anything of significance to all the projects they use.
And really, that’s true of any F/OSS project–most people are just going to use it without giving back anything of substance. If you aren’t OK with that, you really shouldn’t be working in the F/OSS world. And it sounds like RedHat has decided they aren’t OK with this any more.
if someone did not find my sarcasm, I am sorry, MArko
AFAIK, the GPL doesn’t allow limits on redistributing (republishing) the source code…
Which is what Red-Hat is stipulating in their Portal. according to the Alma report…
My 2 cents
Yes, I think you’re right there–I was thinking about a different issue.
Despite the ‘fact’ that e.g. Rocky Linux, Alma Linux and soon maybe others provide a ‘path forward’, the very important aspects of "Sustainability and reliability’ of such ‘paths’ are very uncertain and raise red flags all over the place.
So, based on lessons learned, I would argue not to await any unstable ‘RHEL’ path forward and spend/waist resources, time and effort on that (and get disappointed again), but focus solely on sustainable and reliable way’s for the future of NSx. Reliability, sustainability create our required Continuity so to speak.
I never understood why a lot of users here advocated for “some” RHEL copy as basis for NS8, when the writing was already written on the wall: Don’t trust Red Hat! And some of these distros touting a “bug for bug” copy of RHEL, as if people want bugs!
If I or a client of mine needed a platform to run say a big ticket enterprise product like SAP ERP software, I might choose to use a RHEL original or eg clone from Oracle to run the Database, as it’s fully “supported” for that use case.
Ubuntu was popular for a while, until Canonical dropped it’s mask, especially with that snap-crap. The snap “client” software may be open source, but the server end is completly closed source, and the client is of NO use without the server side! Also, software distributed with snap often displays bad concepts, limits and plenty of other issues, which using apt / yum do not entail! Not enough space, limited possibilities to outsource the space management / usage and much more!
I was one of the few since the beginning suggesting & advocating Debian, and I still believe in Debian, one of the first Linux, older than Red Hat, and now, the last trustable, reliable Linux, suitable for server operation.
My 2 cents
Hi @ All,
Thanks, my whole point. 3 red flags: ‘wants’, ‘bypass’, restrictions’.
As I said
While the community debates whether this violates the GPL, we firmly believe that such agreements violate the spirit and purpose of open source.
No one can prevent redistribution of GPL software.
Our legal advisors have reassured us that we have the right to obtain the source to any binaries we receive
RHEL can try to make things difficult. They have to redistribute the code source
I just switch to Debian and I am very happy I did it at the right time.
Doubtless inertia has been a big factor; Neth, and SME before it, and e-smith before that, have always been EL-based.
The “bug for bug compatible” is common hyperbole, the actual meaning of which is something that functions just like the original. It’s not a matter of anyone wanting bugs, it’s a matter of wanting something that works just like RHEL.
If we’re going to continue to be EL-based, I’m continuing to think Oracle Linux is the best of the lot–they have plenty of resources to support it, they appear to have a good business unit behind it, and they’ve made recent public commitments to keeping it open. They could, of course, do a 180 on the latter (as RH has done), but they’ve already been using this as a major competitive advantage against RH, so it seems unwise of them to throw it away. But I quite recognize the irony of considering Oracle the “good guys.”
NS8 in its current form works with Debian as well. Whether the recent events would be reason enough to ditch EL as a platform is another issue, though I don’t think I’d really argue against it.
I’m not stating Red was bad per se. In fact, for several years it was really quite good.
Blue is a different animal.
Some people sell you lemon juice. Other sell you squeezed lemon pips for a surcharge.
It may have been funny in the beginning, but RHEL clone is more correct, less long.
OracleDB on OracleLinux may be considered best of breed combo, I’m not so sure one can say that of DB2 on AIX anymore…
Both would be typical of very large enterprise environments…
However, Larry being Larry, it would NOT surprise me if Oracle get’s to do the “Hat” trick…
Office, MySQL, now the whole Linux Team…
(For the unawares, Short form: The whole “Office” team quit practically in corpore at Oracle, leaving Oracle to hand Office to the Apache team. The same happened with MySQL, they forked MariaDB and said “fork you, Larry!”.)
My 2 cents
In general, why would we care at all if NSx and beyond are vendor agnostic?
I would generally agree with the above, given that the discussed OSes all have decent enough hardware support in the current versions for server hardware.
A reality check reveals all the “gotchas” in small detail differences between underlying OSes, and the choice of Hypervisor if virtualized, and even the age of test hardware used…
Only in practice are there differences between theory and practice - or how did that old saying go?
All interested in the development of NS8 from a Beta to a fine OS for networking have been following the issues that have cropped up - with the dev team doing great work to strike each one down which crops up!
I do trust in our great dev team to continue this innovative coding and make NS8 as good or better than NS7!!!
My 2 cents
Come to think of it, SUSE is not doing bad at all too
here’s a recap of concerns: