Slashdot Mirror
Red Hat Becomes First $2 Billion Open-Source Company (zdnet.com)
An anonymous reader quotes a report from ZDNet: Red Hat just became the first open-source company to make a cool 2 billion bucks. Not bad considering Red Hat became the first billion dollar Linux company only four years ago. Red Hat did it the old-fashioned way: They earned the money instead of playing upon the gullibility of venture capitalists. Red Hat's total revenue for its fourth quarter was $544 million. That's up 17 percent in U.S. dollars year-over-year, or 21 percent measured constant currency. Subscription revenue for the quarter was $480 million, up 18 percent in U.S. dollars year-over-year, or 22 percent measured in constant currency. Subscription revenue in the quarter was 88 percent of total revenue. Analysts estimated Red Hat would make $534 million. Looking ahead for its 2016 FY Red Hat expects to see between $2.380 billion to $2.420 billion. At this rate, Red Hat should easily become the first $3 billion open-source company.
While Red Hat's president and CEO Jim Whitehurst credits the "hybrid cloud infrastructures," Red Hat's subscription revenue can largely be ascribed to Red Hat's flagship product: Red Hat Enterprise Linux. Still, RHEL, which is now available on Microsoft Azure, is becoming a prominent cloud operating system.
While Red Hat's president and CEO Jim Whitehurst credits the "hybrid cloud infrastructures," Red Hat's subscription revenue can largely be ascribed to Red Hat's flagship product: Red Hat Enterprise Linux. Still, RHEL, which is now available on Microsoft Azure, is becoming a prominent cloud operating system.
I've long switched to debian-based distributions, but I remember buying boxed distros in the 90s to help support the commercialization of Linux. I'm still waiting for LOTD*, but a win's a win.
If you were me, you'd be good lookin'. - six string samurai
Spell it "SystemD" not "systemd"
That way it looks like an ASCII penis.
The free version of RHEL is CentOS or Scientific Linux, not Fedora. Fedora is, as you note, a upstream, fast-moving, bleeding-edge, "test" OS. If that's not what you want then you're using the wrong thing.
Red Hat made the decision to go with systemd for their own operating system. They didn't "force" it on anyone else. If you're unhappy with Debian / whoever else's decision to use systemd you'd better to talk to those maintainers. Obviously they see value in it or they wouldn't be using it.
Generally, when you say 'an X dollar company', people are referring to market cap, or the aggregate consensus value believed in by your investors.
XML is like violence. If it doesn't solve the problem, use more.
Most Products and Services in the IT world are that way. Dell Computers are more expensive not because they are better, but rather because you get "Enterprise" support. Do not underestimate the power of "Enterprise Support" in the world of CIOs and Directors of IT. They have a distinct aversion to taking the blame for bad decisions, and that "Enterprise" label allows them to shift blame to the vendors.
When you build the solution yourself, and it doesn't work, you get the blame. When you have Dell or someone else "Enterprise" build it, and it doesn't work, you can blame the vendor. That difference is worth the price for the people that care.
Agent K: A *person* is smart. People are dumb, stupid, panicky animals, and you know it.
Redhat's success really has little to do with open source, other than they could take advantage of the fact that someone else created the Linux kernel for them and they could build on that.
Who funds Linux development? RedHat: 11.2%
this post contain no useful information, no need to mod it down
Would be interesting to know where the other $1.7B goes.
Lot's of highly paid employees. Lots of them. Far more than they actually need, and hence the $1.7 Billion in overhead.
As an actual Director of IT, aversion to taking the blame for things that are my responsibility has nothing to do with it.
I have a limited team of people to work with, and I have a lot of things to manage, including the phone system, the data center, our office network connections, security, desktop support, and the list goes on.
What I don't have the time to do is take my small team and build brand spanking new solutions myself. I pay someone to provide 4 hour support so that when my tired ass needs to show up in the data center at 4am, there is a disk or even a whole new box waiting for me that I know is going to work. And often a technician to put it in for me.
One time, I remember mentioning looking at the price tag for some service that would cost me $10K a year and thinking I could totally do that myself. I mentioned that to someone else and he pretty much said, there is no way that you can do that for ten thousand dollars if you actually add up the costs of building and maintaining that. He reminded me of a rule I have: Only do something yourself if you know you are the best person to do it. In the end, I might save $10K only to cause myself sleepless nights and reduce the availability of my systems.
Yes, people get ripped off all the time. You won't see me touch Oracle with a ten foot pole unless I absolutely, positively, need the high end features it provides. I will use an open source DB or a cheaper commercial one. But I will still throw down money for support, because they know this stuff better than I do and can get it to work faster than I can. That makes my company able to make money and the people at it able to do their jobs. Enterprise support is not something you buy just because of the label, but if they really do provide good Enterprise support, you put that shit on your budget and you don't look back.
Why? Debian doesn't source from Red Hat. You're saying that because Red Hat does it, EVERYONE has to? Come on!
They really did not. Especially not on Debian. Everyone had their say, publicly, and the majority of Debian maintainers chose systemd over sysvinit/OpenRC.
The free version of RHEL is CentOS or Scientific Linux, not Fedora. Fedora is, as you note, a upstream, fast-moving, bleeding-edge, "test" OS. If that's not what you want then you're using the wrong thing.
CentOS was created without Redhat's help when Redhat discontinued and tried to kill the open source version of their product. The gave everyone fedora so they could technically be in compliance with the open source license but they had enough glue that was not open source so that centos and fedora had to jump thru a bunch of hoops to be usable at all. So, no, CentOS is not the officially sanctioned open source version of Redhat. Redhat tried (and luckily failed) to kill off the open source version of their product.
We recently bought a set of servers from ___, Enterprise VMWare destined servers, and the SSDs? LiteOn. Which failed to deliver the performance needed for the job. Flat out didn't work. Granted, the company _____ replaced the drives, eventually, after we proved they were not capable. The problem I have, is I would NEVER have spec'ed LiteOn Drives for anything even close to "Enterprise".
And in the end, we wasted nearly 4 Man Months of time trying to fix the problem.
And my boss, buys Enterprise, even when I can PROVE that they are exactly the same, off the shelf consumer products, for twice the price. Me, I would buy two for the price of "Enterprise" and keep one on the shelf as a Spare. Knowing where you can get the perfomance you need, at a price that isn't "Enterprise" often allows you to stretch your IT budget AND provide the support your organization needs.
I'll pay for support, I'll even pay a lot for support. But I won't pay for "Enterprise" that is only "consumer" with a new label.
Here is BACKBLAZE's article on Drives t hat kind of supports my view ... https://www.backblaze.com/blog...
Agent K: A *person* is smart. People are dumb, stupid, panicky animals, and you know it.
The mailing list is public, go and check it. Alternately, this site was put together as a summary of the various positions and options. If I can characterize it, there was a strong desire to move away from sysvinit due to lack of features, bugs, and difficulty of maintenance. Systemd at the time was seen as the best of the alternatives, offering more features and easier maintenance, at the cost of compatibility with non-Linux systems.
Not that I have any great insight into the minds of developers, but I suspect that the decision might have gone otherwise if it were re-held today. I think there would be a stronger consensus against sysvinit, as even fewer people are interested in maintaining those scripts. OpenRC has had more time to mature, and as far as I know Upstart development has basically ended. Interestingly, OpenRC and systemd share a number of features, particularly in their heavy use of C libraries, for which OpenRC receives no criticism and systemd no end to criticism. Either way it looks like that, dependency resolution, cgroup support, and parallel startup have made everybody's minimum feature list. I'm sure it would have been an even more different story if cgroups/process tracking had been a part of sysvinit/POSIX to begin with, but as I understand they were mostly codifying existing practices rather than trying to actually create a good standard.
In any case, Debian moving away from sysvinit wasn't any more influenced by Red Hat than it was by Canonical. All options were on the table, and each of them had their proponents. There was somewhat more backlash against systemd than other options, but I don't think politics played much of a hand in the decision -- politics didn't have to maintain the code afterwards. And despite the popular clamor for sysvinit, most everyone else seems to have dropped it happily, so ignoring the populace seems to have been the right decision in that respect.
Those who advocate genocide deserve every protection afforded by law, and none afforded by common human decency.
All of what you say is accurate, however, RH don't paywall the code, just the binaries. This is how CentOS and friends exist - RH publish the source RPMs to comply with the GPL, and from the beginning ensured everyone knew they were there. They also announced that if you wanted to use them to build your own products, you would have to strip out all of RH's logos and other IP (hence the "prominent North American Enterprise Linux vendor" of CentOS).
CentOS is now pretty tight with RH as well, as they've directly sponsored CentOS since 2014 and have hired the project leaders.
Bullshit
There needs to be a mod option for "dumbass" but troll will do.
There is nothing in the GPL that demands that RH do anything other than what they are doing for RHEL.
You can paywall GPL code, idiot.
Fedora is the test version.
If RHEL is out of license compliance, Fedora would do nothing to solve that.
shh
I never said it wasn't within their rights nor did I say they are out of compliance. They release the source of the parts of their OS that they are required to do so. The average user is not going to have the skills nor desire to compile their own OS. Nothing that you said negates the fact that where previously they were releasing a free open source version of redhat in parallel with their commercial version, when they switched and no longer released a free open source version that the open source community who was used to using redhat had to scramble to get something usable again. redhat went one step further than most companies and even makes them stripe out any reference to the name redhat which is probably considerable work. CentOS is fine now but it was a rough at the very beginning. I actually switched to mandrake/mandriva when they did that and have since switched to ubuntu. Previously I had owned purchased copies of redhat but I don't trust them anymore.
1) License tracking is license tracking, regardless of whether it's Redhat, Microsoft, or Oracle. The major appeal of Free (and Open Source) software is not having to report to anyone.
It's not a licence, it's support. Depending on whether you have signed a contract that restricts you further (at your own choice) or not, you can install as many copies as you like (but you would need to do some work to get updates to more instances, like run reposync on an instance that has a subscription).
2) We don't use Redhat support for anything other than software updates. The few times we used Redhat support for problem solving, the support personnel knew less about the problem space than we did. This is typical of most paid support.
We, a linux-centric team within a big telco, also only use RH support for updates for the same reason (we can almost always resolve issues quicker than it would take to get a ticket escalated to soneone with skils. However, in a few instances RH support has identified issues where someone on our team made a mistake which then alloeed a quicker resolution.
However, other less-experienced teams in the rest of the company need RH support more than we do.
Also, when dealing with other vendors (server, storage etc.), it helps to be able to tell them 'Red Hat has vetted our config, your solution claims to be certified on Red Hat on this hardware combination, our support contract now requires you to fix our stirage issue'
3) With Redhat, there is no freedom to fire up a new virtual (or physical) server without significant additional cost, just like with non-Free software.
For VMs, that depends on the subscription. You could easily buy unlimited VM subs and run 10 VMs for less in subs than the 5 VM sub.
In fact, Red Hat Smart Virtualisation (RHEV, aka supported ovirt, plus unlimited VM subs) is cheaper than the cheapest way to get vSphere Enterprise (VSPP), and you get supported VMs for free.
Who are you paying for your vietualusation infrasructure software? Did you factor his in to your comparison?
For non-production physical machines, we spin up RHEL instances in minutes (via kickstart), and only take out a sub if it ends up being a permanent installation or requiring support or updates. However we are busy phasing out the use of physical servers wherever possible and replacong them with VMs on RHEV.