LQ) Tell us a little about yourself, what got you interested in Linux and what your role at RH is.
JH) I started on the Mac, being a multimedia person back in the day. Market forces painted me into the Wintel corner, and I had to learn so much cruft just to keep my machine running at all, that I find a job doing it for others.
I found myself in tech support. I was testing Linux for a large company I used to work for. It was Red Hat Linux 5.2, and it was awesome.We wanted to see if it could replace Solaris and other Unices in the back end.
One dark and stormy night, the IS cops nuked my dual boot, for suspicion of running pirated software, including all my e-mail and files for work. After several attempts at explaining what the GPL was I gave up and went back to my pen to stew in my own juices.
I was incredibly unhappy with Microsoft, the (lack of) quality in their software, and it's "blind leading the naked" advocacy. I was deciding whether to write a disgruntled rambling manifesto to the CEO, or launch a bootp server and take over the whole infrastructure and hold it hostage until they rebuilt my machine when a recruiter called me.
Two weeks worth of notice and I was free.
LQ) Tell us a little about the just released Fedora project (How do you see it impacting RH, how does it compare to Cooker or even Debian, what went into it's release, etc).
// edited by request
JH) Fedora is what Red Hat Linux was. Kind of theNew York to New Amsterdam. It's a project with rolling releases, not a product with predictable release dates, support, services, etc.
We decided that the rapid release cycle of our retail product was not conducive to retail selling paradigms, nor was slowing down innovation to suit that. We also decided it was time we made it a little easier for folks to develop with us, and to build on our products.
I would compare it to Debian. What's in the release is what would have gone in Red Hat Linux 10, if there was one. The main difference will be the amount and level of influence contributors and thrid parties will have on the project's direction.
LQ) How will Fedora effect older but popular releases, such as RHL7.3 (ie. will they be picked up and supported beyond the RH EOL date)?
JH) End of life is the official end of Red Hat technical and product support for things like 7.*. Fedora is not the migration path for those products if you are interested in support, 5-7 years of errata, ISV/IHV certification, training, and stability of the API/ABI. Fedora is for developers, contributors, beta testers, hobbyists, and enthusiasts.
So, to clarify, the Fedora project will not be releasing (for example) security updates for *any* older RHL releases
JH) Fedora will focus on Fedora. Individuals may find that errata and updates issued for Fedora are suitable for backporting, recompilation or may well work as is on other products and releases. It's all free software. One of the benefits of the project we merged with was their ability to handle outside contributions.
Red Hat won't guarantee or support official errata for any EOL'd product, and Fedora won't have guaranteed resolution times, or contractual priority states for Red Hat engineers.
It's all stated better than I can sum up at the FAQ and project homepage.
LQ) It would seem to me that Fedora marks Red Hat's exit from the desktop Linux market. Is Linux on the desktop just not a profitable market or were there other reasons?
JH) We haven't yet gotten into the desktop market. Red Hat Enterprise Linux WS is the "corporate desktop" to us. We will be looking at home entertainment, usability and productivity issues and decide what our desktop strategy will be.
Fedora is more about what product you will find in our retail channel (or not). Retail==desktop. I think you would agree the best desktop play is in OEM pre-loads than retail. Who really walks into Best Buy and grabs a copy of Windows XP? The boxes are there for show. Most folks get an OS on the machine they buy.
Does RH have any additional OEM plans?
JH) To do as much as our customers want or need. We are always willing to work out pre-load agreements. We have growing relationships with IBM, Dell, and HP. We have emerging business with IHVs based in Asia and Europe as well.
LQ) Some people have been able to get RHEL into companies by starting with RHL as a base, proving it's worth and then justifying the expense for support/etc. What would the recommended path be now?
JH) There's many paths. One might be to run RHEL in the lab, and buy the subscription and support before running in production. Folks can still get exposed to it in our training. Fedora's not meant to be the acquisition vehicle that Red Hat Linux was, but we will work out such a vehicle. For example, we've done demo pricing for companies evaluating RHEL, and may do so in the future. We're not selling the bits. We sell the convenience, services, and overall Red Hat experience.
Folks interested in the try before you buy aren't usually the large
enterprise customers. Those folks are interested in the professional boxed set and it's pricing. We will find a better way to address that gap.
LQ) RH recently had it's best quarter ever (Congrats BTW!) and posted a profit of about 3M. What is RH doing to make sure it remains profitable and what do you see as the biggest obstacles to long term success and profitability?
JH) We're continuing to execute on what's gotten us to this point. Don't fix it if ain't broke, but we are always looking to reduce the cost of selling, operation costs, and things like that. The real crux of the success has been in driving open source up the solution stack, like expanding into applications, systems monitoring and management. And moving into other verticals. Increasing government adoption for example.
LQ) What factors do you think allowed RH to succeed where others have failed?
JH) A certain amount has to be attributed to being an early player, and getting the product out there any which way we could. I think we hired the right people at the right time, and of course the IPO helped fill the "war chest". The founding Red Hatter's really paved the way.
LQ) What do you consider RH's biggest strength? Biggest weakness?
JH) Brand recognition, and reputation for being quality focused. I think the biggest weakness is being the Linux market leader, if that makes sense. It makes us the target of Linux haters and the supporters of other distributions. We have been called the "Microsoft of Linux" which is so unfair. For one thing, there are better metaphors for success.
We have a reputation for being aloof, that is deserved, but not true. We have resources we have to use wisely, and focusing on one direction means we can't focus on other things. But that doesn't mean we don't care about the home desktop, or multimedia support. We are in a Microsoft world in some respects, which puts an expectation on us to replace everything they've done in 25 years of being a thriving monopoly, in ten years in the shadow of a monopoly.
LQ) What direction do you see RH taking, both short and long term?
JH) Short term: enterprise verticals, security, virtualization, management, government adoption, globalization. Open source is spreading like brish fire in other countries. Emerging economies see it as the great equalizer.
Long term: to take over the world and drive our slave army before us doing our evil bidding. Wait, that was supposed to be internal only.
LQ) What are your comments on some people calling RH the "MS of Linux"?
JH) Heh. I didn't see this before I hinted at it above. But I think it's a short cut to thinking for many folks. I mean, if you're in the car business it all comes back to Ford, right? So they say "Hmmm... I need a software success barometer..."
I think folks also use it to remind us that they will not tolerate another Microsoft.
I call us the Starbucks of Linux, which is probably seen as only slightly less evil, but it fits better to me. Yes you may have a local roaster that you patronize religiously, but the fact is, if you are in the middle of some strange city, you know what Starbucks will have for you.
You are paying for the comfort that this triple shot soy half caf will taste the same as it did in Topeka as it does in Wichita. It's all about consistently setting and delivering expectations. The delta between the promise and experience is the brand, the delta between what you pay and what you get is the value. If you can be consistent in predicting those deltas, you have customers who get what they paid for rather than paying for what they got.
LQ) Seibt was recently quoted as saying: "Linux means two companies: Red Hat and SuSE, and nobody else." Is this how you see the market?
JH) For some customers too much choice is confusing, and if the other offerings aren't differentiated where you desire, you will gravitate toward the "name players". Frankly, the market is conditioned to reject choice. They've never had it.
Red Hat doesn't want to lead a market of one (or even two), we don't want a single vendor world any more than the next guy. I myself run other distros all the time. I'm a big fan of Knoppix, I like some of what Lindows does, I dig Slackware and Debian almost as much as I like getting flamed by their users. The power of the GPL is that the market is bound only by the motivation of suppliers.
LQ) If you couldn't use a Red Hat based distro what distro would you use?
JH) I'd use OSX. :-D
I'm not sure, I'm so used to it, I'd be lost in another world. I'm a web guy, so BSD would appeal to me.
LQ) RH recently joined LinuxQuestions.org. What do you anticipate the level of participation will be and what factors lead to the decision?
JH) As much as the folks I bribed to sign up with empty promises of cash and the accolades of peers and friends can give.
It's up to the participants, I find that the questions are handled more by the other Red Hat geeks than the official Red Hat presence, so I hope we can make sure that threads don't go off into oblivion, and that unanswered queries get some help.
LQ) Do you have any suggestions for us?
JH) I got lost the first couple of attempts to use the interface, so either it needs a little usability help, or I'm a dolt. All indications pointing to the latter.
I would also see about subdividing the distro forums into topics like "Installation", "networking" etc. Since the threads roll over pretty aggressively an archive or FAQ in each topic would be nice.
I don't know, it seems pretty good to me so far.