> Red Hat is a fantastic company, and a pioneer in successfully commercializing open source. However, beyond Red Hat the effort has largely been a failure from a business standpoint.
SUSE started around a year before RedHat, in 1992. Sure, we haven't had the same success and growth, but I would disagree that RedHat pioneered the model. I wouldn't even argue SUSE pioneered it (arguably it was Cygnus in 1989 -- though it was eventually acquired by RedHat in 2000). And even earlier you could argue that RMS pioneered it with his distribution of GNU Emacs in the early 80s (which had a description of the business model as the final chapter in the manual).
[I work at SUSE and am obviously biased, but it is a little bit annoying that people always ignore that we still exist...]
Purely anecdotal and I've never worked for or on a team whose main responsibility was Linux administration, but at two places I worked in the past, our Linux guys didn't like SUSE. It wasn't a technical problem, and more marketing could probably help fix the problem.
I have a SUSE Linux certification, and at two companies folks from the Linux team commented that SUSE makes them uncomfortable because they're unsure of how to pronounce the name. It's like GIF or Porsche but worse. Half of them pronounced it Suzy, like the person's name. Some pronounced it closer to Zeus with a -e at the end. Some pronounced it "Soos" like it has a silent -e. They preferred Red Hat or Ubuntu not because of technical superiority but because they were embarrassed to attempt pronouncing SUSE. It's likely the only time they heard it pronounced was from coworkers who didn't know how to pronounce it either.
There was a video explaining how to pronounce SUSE several years ago (both funny and a little worrying that such a video was necessary)[1]. It is pronounced strangely since it was a German name and so you have to put on your best German impression to get the "s" sound right.
All jokes aside, I think most folks would agree that SUSE's marketing really needs to be reworked very significantly.
The German pronunciation would be pretty close to "zooza", so the pronunciation in that video isn't German, and a German accent would hinder you from pronouncing it that way (in German you can't have the /s/ sound at the beginning of a word followed by a vowel). Either way, "sooza" and "zooza" are not exactly hard for English speakers: they're just completely non-inferrable from the "SUSE" spelling.
Yeah, most of my German colleagues say "zooza" -- but saying it this way makes it difficult for English speakers to infer what you're talking about. So I say it as "sooza" even if it's _slightly_ incorrect.
Respectfully, this is something I've never been able to understand about open source projects. Why stick with names that are difficult for most humans, and causes folks to forget your great company still exists? Why not run a small marketing and imaging campaign to sort that out. You already got a great green chameleon logo :-)
> Are there still people who pronounce linux lie'-nucks?
In German it's the stressed "ee" variant. I think with made up words that don't appear to belong to a certain language, yoiu just tend to pronounce it like it was a word in your native language.
Google in contrast just looks English because the "oo" is much more common in english, just like the "le" at the end (apart from some regions where you append "le" to nouns to make them sound small/cute).
I try to use his original pronunciation wherever and whenever I can, but most of the time I hear “Lin-ucks” with a short “I”, and I naturally respond in kind without even thinking about it.
But then there are occasionally articles like this one which remind me that I’m saying it wrong, and that I know better.
As a non-native English speaker, yet one who's been in North America and dreaming in English for 2+ decades, Ooboontoo was the one and only pronunciation that came to my mind (I suppose because it's a fairly phonetic interpretation).
To me, "You-Ban-Too" is a completely "you have to squint and tilt your head and really want to see it" pronunciation/interpretation :->
It's partly because of the flattening of certain vowels that happens in Am English. Like how con is similar to Kahn in Am English, bun has a little ah to its uh. Hence u (you) bun (bahn) tu (Too).
Right, fortunately Ubuntu has enough marketing behind it that most people who know Linux already know how to pronounce it. SUSE's marketing is more limited in that regard.
In 2006, Ubuntu was a ridiculous name. Then Canonical's marketing kicked in.
> SUSE makes them uncomfortable because they're unsure of how to pronounce the name.
Amusing the irony in this considering most people pronounce "Ubuntu" and indeed "Linux" incorrectly.
FWIW, it was the first distro I ever tried back in 1999 and I always pronounced it "Suzy" but have heard a few variants. Everyone pronounces different words differently so it doesn't really matter.
There's a fascinating solution to this problem, used by Nikon[1]:
[...] Nikon Corporation (that's Nikon Japan) officially and consciously blesses all regional pronunciations of their company name.
This is policy, agreed upon decades ago—likely in the 1950s—in some meeting room somewhere in Tokyo. Nee'kon is correct. Neye-kon is correct. Nick-on is correct. Many others are correct.
This was explained to me when I was a Nikon employee by the Executive Vice President of Nikon Inc. himself, in the flesh. I can still picture the conversation, 25 years later. And to anticipate the usual objections: Nikon Corp. is the ONLY, the SOLE, the UNITARY authority on this question. "Nikon" is not a natural word in the Japanese language. The company that is now Nikon (then Nippon Kogaku, K.K.) made it up. It's a brand name—like Kodak or Xerox. A neologism, and maybe even a portmanteau (that's debated). It was likely first conceived in the Latin alphabet, not in Kanji. Nikon Corp. gave birth to this name, and they own it. They get to make the rules about it. And their rule is clear: all regional pronunciations are equally correct. If you doubt this, just listen to their own regional advertising, produced by the company's various subsidiaries around the world. You'll hear several different, completely official, pronunciations. Oy. I feel better now.
Addendum: Of course, it goes without saying that however they pronounce it in Scotland is wrong.
I get that it's a German company and that they can name it what they want, but I also think it's part of the problem. The only reason I know the correct pronunciation is because I was in a class with a Swiss person who made reference to it.
More generally, it's an example of why I think that, for US companies, "weird" names are a bad idea. People don't like to sound stupid and something as simple as a name that's difficult to pronounce, or sounds odd can limit adoption of your product.
I’ve been a Linux user and Administrator since late 1998, and various forms of *nix before then.
I’ve known of SUSE for a long time, but it was never the name that stopped me from doing anything with it. I just never used a product or worked on a project where it was used, and so I was never exposed to it.
TIL, I’ve been pronouncing it wrong for probably something close to a decade.
Isn't it pronounced 'sues'? I doubt it's the name but it always makes me wonder what Suse's position is. Is there anything specifically better than Redhat or Fedora?
It may have bigger market in Europe but when I heard it sets the default file system to btrfs when others are jumping off that ship, I wonder what their plan is.
German doesn't have a concept of a "silent -e" for their words. Like how Porsche is pronounced Porsch-uh, SUSE is pronounced "Soos-uh". I don't work for/with SUSE but I also accept when people say "Suzy" because at least it acknowledges that the -e is supposed to make a noise.
I think you're being a bit over-sensitive. They said a pioneer, which does not preclude other pioneers. Also, what's the bar for "successfully" commercializing open source? In a world where anything less than a unicorn is considered failure - and bear in mind the OP is from a promoter of that ethos - Suse's trajectory might not qualify. Even if it did, it would be one of the few exceptions to the general rule that commercializing open source largely has been a failure. Open source as a development method is alive and well, but as a business in its own right not so much.
P.S. Eight years at Red Hat, ending a year and a half ago. Glad not to be under IBM's inept thumb as my erstwhile colleagues soon will be.
It's very popular in Europe, RedHat is very popular in the US.
I think the IBM acquisition is a fantastic opportunity for SUSE to start their engines here in the US. They have a fantastic value proposition: start with openSUSE Leap for free, then upgrade to a paid SUSE contract (can be done on place). This should appeal to CentOS and Fedora users who consider using RHEL, but don't because it's too hard to switch. They can also appeal to people who don't need the IBM integration but feel like they're not getting as much attention as before (priorities will likely change at RedHat). They can also appeal to developers with their rolling Tumbleweed release (which can also convert to openSUSE Leap/SUSE).
If SUSE had publicly traded stocks, I'd consider buying in right now as they have a really good opportunity over the next few years to steal customers from RedHat/IBM. They're big Linux contributors (Greg Kroah-Hartman used to work on the project), so they are _very_ attractive.
In France, I've never worked for or even heard about a company that uses Suse. It's usually Debian if the sysadmin have all latitude, RedHat if certifications or having a company behind the product mattered for the managers.
At this point SUSE is just as niche in Europe as it is anywhere else.
In an enterprise setting, why would you throw your lot in with a company that seems to change owners every few years rather than the biggest name in open source that also now happens to be owned by IBM?
It's not something that RH says you could so there you go.
You can't just pull one system out and put a new one, specially Fedora, this would be equivalent of pulling a rug from underneath someone standing on it, who knows you might be lucky, just try it.
SUSE on the other hand encourages switching after the introduction of LEAP 15, in other words it's supposed to be trivial to do.
I was delighted to find that SuSE has started using btrfs to do the same root filesystem snapshotting that Solaris does with ZFS! Suddenly it was back on my radar after not paying much attention to it since the 1990's (I went into Debian land and stayed there, pretty much).
Red Hat doesn't have any btrfs engineers, they all moved to Facebook -- so they don't currently have anyone in-house to do btrfs work. SUSE does (one of the upstream maintainers is a SUSE guy). Facebook also runs btrfs.
Red Hat's plan for "replacing" btrfs is to write a bunch of additional tools around devicemapper/LVM. My experience with using LVM is that this is not going to end well -- if you thought btrfs was unstable you've seen nothing yet.
To be clear, btrfs does have its issues, but linking to RHEL no longer supporting something they never actually fully supported is a bit of weird thing to argue is a negative of btrfs. Red Hat doesn't get to by the sole arbiter of what is and isn't good software (and it should be noted that Project Atomic's snapshotting using ostree is simply not as good as using btrfs and the transactional-update integration into openSUSE Kubic -- btrfs snapshots are atomic for a start and rollbacks are a single syscall away).
The article touts the business model as the underlying failure of Open Source. However I think this is missing the forest for the trees. If you add up all the value that Open Source delivered by giving millions, yes, let's go for millions, of developers access to high quality tools that previously were only available as expensive, proprietary products then clearly the model has been a runaway success. The only lament would appear to be that the value was not captured by a single corporation.
I think the pendulum is now swinging back to the pre-PC days of large corporations, centralized control of technologies where only groups with large resources can participate. However that will only be a temporary phase until the next generation of mammals are able to come up with the tools to create competition once more and the whole cycle will repeat.
The article didn't claim that OSS wasn't successful in the broader sense, just that in the business sense, the Red Hat business model has failed. It further points to a model that does work, SaaS.
It seems to be closer to saying that the Red Hat model - essentially being a professional scapegoat, with some consulting on the side - hasn't succeeded to keep up with the returns venture capitalism expects.
The flip side is that companies that take over the operations part of the value chain and not just corporate liability are fully able to compete with closed source companies, and that kind of hosted service is where corporate open source is likely to go.
The article /does/ discuss that, and mentions that Open Source in the sense of software technology has been a huge success, but also that as a /business model/ is far, far, behind the proprietary models. It also mentions that there are another OS-based models that are successful, such as SaaS, cloud, etc.
That's how they were used at my last employer. Yes, their OpenShift platform is very good, and they have a lot of good products in their ecosystem, but the biggest value they provided was a vendor to blame if things went south.
Maybe Red Hat and IBM aren't as different as we like to think.
Funnily enough, nobody sues IBM when their product is an obscenely expensive barely functional mount of manure
Even more funny, those who propose the acquisitions of said tools are not held responsible when things go south for their complete inadequacy for their job.
And even more funny would be to find someone who has enough power to sue IBM, which probably has more lawyers in its legal department than the plaintiff has total employees.
Not once have I had an insurmountable problem with any Linux platform.
The same is not true with Microsoft, Oracle and IBM from experience, even with paid up top tier support. Even 60,000 employee defence companies have no real power. I've seen someone forced to throw a registry fix out to 5000 desktops (over hundreds of disparate sites) because MSFT wouldn't fix a fucking bug.
>> We had made the product so easy to use and so important, that we had out-engineered ourselves. Great for the open source community, not so great for us.
I can relate to this.
The other sad truth is that if you do not 'out-engineer yourself', then your open source project will likely not even be able to get any users to begin with; it's extremely competitive.
It took me over 5 years of work; constantly refining my project in order to reach almost 5000 stars on GitHub and 50K downloads per week.
Those are pretty big numbers for a framework targeted entirely at developers but in terms of monetization, it's barely enough to cover a single salary and my project is one of the lucky ones because it's back-end and it's somewhat more opinionated.
Front end frameworks and non-opinionated back end tools (e.g. the ones that just implement existing standards/protocols) have near zero monetization opportunities. For front end frameworks, you need a huge success like Vue.js or React in order to monetize at all. It seems that Evan You from Vue.js gets enough money to cover his salary - But Vue has over 100K stars on GitHub and is one of the top 10 repos in the world. If Evan was a musician or an actor, he'd have a multi-million dollar salary for sure.
The problem of investment in the world of open source software is a difficult one. What company wants to pay for improvements when they know competitors are going to get all the benefits of these investments too. It's one of these bad equilibrium, prisoner's dilemma, kind of situation.
In practice, the funding often comes from those people or companies that more urgently need the improvements or those that are pushed by end users who are willing to pay a premium to avoid lock-in. There is a longer tail of users who free ride and pay nothing for improvements.
Even businesses that support open source, at any moment in the future, have strong incentives to use network effects to steer their customers into more subjugated positions and adopt proprietary non standard or non open systems, or to sell themselves to businesses that will do that.
The coagulation of tech companies in a few large players is a sign that this may be happening right now.
I wonder if there are ways to promote wider investment in open source and open standards technology that benefit more people and get us into better economic Nash equilibria.
Seems weird to claim Red Hat failed. Even if their revenue is tiny compared to Microsoft or Amazon they are still on a chart measuring market cap by billions of dollars.
it's incredible to see in the same period of time just how massive both Microsoft and Amazon have been able to perform (market wise at least) and specifically the hockey stick market cap of Amazon in that time frame...
I see Red Hat's business model as a product of the times - there was this cheap, underutilized, but difficult-to-use competitor to the OS they really wanted to use, and a lot of corporations that really wanted to spin-up big IT departments and web services quickly. They filled the market need that existed in the late 1990s; it's unclear if this model fills any major market needs that exist now, when turnkey SaaS solutions exist for a lot.
This doesn't mean there won't be other successful open-source companies. It seems like each generation of open-source projects fills a different economic need, all specific to the business landscape at the time:
In the early-mid 2000s we had open-source as a retention bonus for good engineers and PR campaign for the companies involved (Hadoop, Nutch, Protobufs, Snappy, and other Google open-source projects, Cassandra, HBase, Hive, HipHop, and other Facebook open-source projects).
We also saw open-source as a means of securing cooperation between different corporate interests (WebKit) or securing critical partnerships & goodwill by assuring the source code would be available (Chromium, Android).
Starting in the early 2010s, we saw companies releasing an open-source product and then making money off a hosted version of it (Elastic, Datastax).
Nowadays in the crypto world we see organizations releasing an open-source software system, reserving some of the currency for the development team, and profiting off the capital appreciation of the currency. (Ethereum etc.)
It's worth nting that the exemplar of the prorietary shrinkwrap consumer softare market is also a one-off: Microssoft.
There is no other pure-play software company of comparable market success.
The consumer personal-computing operating system market resisted incursions by CPM, other DOS vendors, other GUI alternatives, x86 commercial unices, Linux, FreeeBSD (and others), and BeOS. Apple's share remains a scant fraction.
Likewise, the Office applications, and small-business server market remains dominated by Microsoft, especially file, print, mail, and directory services (AD).
Contenders in the software space tend to be small, heavily reliant on professional services (IBM, Oracle, Salesforce, consulting firms), specialised (Adobe, AutoDesk)<, or hardware-integrated (Apple, also Sun/Oracle and some others).
The SaaS market has impinged somewhat; Gmail, Google Docs, etc., but still only marginally.
This from someone whose watched and criticised Microsoft for decades, and strongly prefers Linux. I've not had a Microsoft desktop personally or professsionally since the mid-2000s, and even then it was under protest and largely avoided.
That article never actually explained why Red Hat was successful where others weren't. It got lost on showing how _not_ successful they were, even as it said how exceptional their results were.
Agreed. My theory: the market will only sustain a finite amount of revenue by open source software creators. That amount grows 20-30% per year as competing business models collapse and as customers learn the value proposition, but there is a ceiling. Red Hat has been brilliant about identifying the market segments that were ripe for an open source player, and entering those markets (often via acquisition).
There might be room for another Red Hat now that Red Hat will no longer be independent. IBM may have just legitimized Red Hat as a top-tier player.
That is, of course, if IBM still has the clout and capitalization to be considered a top-tier player. 20 years ago, it would have been considered funny to say "IBM is great and all, but they're no Apple". My oh my have times changed.
I think this is a great article with a provocative title. I don't know whether there will ever be another Red Hat or not, it's a meaningless question. But if you want to build a business on open source, this is a critical article to read because Red Hat's model is a hard way to do it and there are other models out there.
Elastic, Mongodb, and Docker all have this business model. I guess the difference is in their case they developed it first and then open sourced - they've got a much more credible claim to superior support.
SUSE started around a year before RedHat, in 1992. Sure, we haven't had the same success and growth, but I would disagree that RedHat pioneered the model. I wouldn't even argue SUSE pioneered it (arguably it was Cygnus in 1989 -- though it was eventually acquired by RedHat in 2000). And even earlier you could argue that RMS pioneered it with his distribution of GNU Emacs in the early 80s (which had a description of the business model as the final chapter in the manual).
[I work at SUSE and am obviously biased, but it is a little bit annoying that people always ignore that we still exist...]