What do you think about the move of RHEL ?

[PMc]

Linux is more than RHEL.
Raise your hand if you ever audited the source code.

Why should one do that?
How many local patches do you need in the source code to make it work appropriately? Here it is some 25.

But then, companies need to buy a product where they can sue the manufacturer, and consumers need only a smartphone. This is a different time now.

 

[Alex Seitsinger]

I'm happy to say that I've spent the last thirteen days, fourteen hours, seven minutes, and probably like 1000 milliseconds mulling over this unexpected and (honestly) unprofessional change to their business model. I am flabbergasted that those limey linux losers have the audacity to do what they did to everyone they did it to.

 

[bakul]

Your sacrifice is noted.

 

[recluce]

There is also the situation, apparently not rare, where a wealthy company would have a bunch of CentOS servers and one or two RH ones, and use those to get RH support if they wanted it. (I haven't worked in such a company, and my evidence is just stories from friends and acquaintances).

I am working for a company (19,000 employees, 14,000 seats in IT, 3,000+ servers (40% Windows, 10% AIX, 50% RHEL/CentOS), which has about 90% paid RHEL licenses and 10% CentOS installations. CentOS is used for temporary development VMs, appliances etc. Everywhere where the internal license-provisioning process is too painful, really. We are now reconsidering if we should not replace RHEL along with moving away from CentOS. So RedHat could only lose at my place of work - are we that unique?

I have been here for 3 years now and finally some FreeBSD servers are entering the mix!

 

[ralphbsz]

... which has about 90% paid RHEL licenses and 10% CentOS installations.

Why not call your RedHat (now IBM) sales person, and ask for a 10% discount on RHEL, if you commit to move all Linux machines to RHEL?

We are now reconsidering if we should not replace RHEL along with moving away from CentOS.

Replace RHEL with what? I you insist on having high-end support, then AFAIK the only real alternative is SUSE. Not clear whether this is significantly different from RedHat. Then there is second tier support, and simply giving up on support and creating your own support mechanism in-house.

 

[LibreQuest]

I don't understand why people are saying fedora could be affected when fedora is basically rhel testing. It's like a couple of years upstream of rhel.

 

[PMc]

Replace RHEL with what? I you insist on having high-end support,

That post was about servers. What good does high-end support? Consider AIX: if you find a bug in AIX, you have approx. one year of discussion until the high-end support agrees that there is a bug. One year, while you look at the Berkeley sources, where the same bug exists and where they have copied their stuff from.

I don't know why companies need this kind of support. But then, I also don't know why companies usually insist in employing only guranteed 100% skillfree personnel that is restricted to ITIL buerocracy and cannot read any sources.

 

[rmomota]

Why should one do that?
How many local patches do you need in the source code to make it work appropriately? Here it is some 25.

But then, companies need to buy a product where they can sue the manufacturer, and consumers need only a smartphone. This is a different time now.

Why should one do that?
Well, not only one but many and many start by one, am I correct?

 

[rmomota]

I don't understand why people are saying fedora could be affected when fedora is basically rhel testing. It's like a couple of years upstream of rhel.

and that's why it will be affected... somehow.
I don't know how it will be managed in future but Fedora Open Source project seems to be in the target of affected distros.
But what people want? what users want, what users need?
I need a system, a free system that is secure and user friendly with all applications I need for daily use.
Open source or close source, if it's free, it just works and it's secure it's alright to me.
I want that freedom to buy any computer and to have the decision to install whatever system I want and I have a choice of buying one, with professional support etc, or get a free one supported by community.
what I really want?.. a fully working system.

 

[LibreQuest]

and that's why it will be affected... somehow.
I don't know how it will be managed in future but Fedora Open Source project seems to be in the target of affected distros.
But what people want? what users want, what users need?
I need a system, a free system that is secure and user friendly with all applications I need for daily use.
Open source or close source, if it's free, it just works and it's secure it's alright to me.
I want that freedom to buy any computer and to have the decision to install whatever system I want and I have a choice of buying one, with professional support etc, or get a free one supported by community.
what I really want?.. a fully working system.

You bring up a great point. I've never had an issue paying for my operating system. The first os I ever purchased was Mandrake Linux and continued to be a power pack user. I don't have an issue paying red hat for a license if I were a red hat user. I wouldn't mind paying for FreeBSD. For me it's about how the system works.

Going by the gpl you would only need to buy a copy of red hat to get the source code. So I don't get the issue here.

 

[subnetspider]

I think I made a great decision to migrate all my Debian and Ubuntu servers to FreeBSD instead of CentOS, which I considered doing 2 years ago.
This is not directly related to the RHEL drama, of course, but I feel like I still dodged a bullet here. ^ ^

I wouldn't mind paying for FreeBSD.

Well, with FreeBSD, you actually have the choice of whether or not to pay (or to donate).

 

[LibreQuest]

I think I made a great decision to migrate all my Debian and Ubuntu servers to FreeBSD instead of CentOS, which I considered doing 2 years ago.
This is not directly related to the RHEL drama, of course, but I feel like I still dodged a bullet here. ^ ^


Well, with FreeBSD, you actually have the choice of whether or not to pay (or to donate).

Indeed, I do donate.

 

[recluce]

That post was about servers. What good does high-end support? Consider AIX: if you find a bug in AIX, you have approx. one year of discussion until the high-end support agrees that there is a bug. One year, while you look at the Berkeley sources, where the same bug exists and where they have copied their stuff from.

I don't know why companies need this kind of support. But then, I also don't know why companies usually insist in employing only guranteed 100% skillfree personnel that is restricted to ITIL buerocracy and cannot read any sources.

We need contracted support for everything production-related to fulfill regulatory requirements in our industry, even if it is sometimes bloody useless. AIX is a great example of that. We use CentOS where things are non-production (and/or transient) and our internal process to get a RHEL license assigned is just to cumbersome.

That being said, once you have to take care of 14,000 users and have several hundred people in IT, "seats of the pants" management does not work so great. At that point you need frameworks like ITIL (even though that is not a great example) to keep things organized and IT reasonably effective and efficient.

 

[taiwan740]

I think for some scenarios, making it closed source does make sense. Particularly when it comes to things like government systems, its not ideal having all your loopholes out there in plain text.

As for the commercial side of it, well, even Kelis charges for her milkshake training.

 

[kpedersen]

I think for some scenarios, making it closed source does make sense. Particularly when it comes to things like government systems, its not ideal having all your loopholes out there in plain text.

Whether source code or disassembly, it doesn't really matter to the sorts of hackers trying to get into government systems. In many ways the assembly is probably more readable than things like Rust / C++

Plus some of the more interesting security flaws are exhibited only once the binaries are compiled.