Comments on: Red Hat: Go big or play it safe?

Red Hat apparently feels like it needs to constrain its ambition to pacify stronger competitors. Why?

[dragonbite]

"Could Red Hat be crushed by Oracle for entering the database market, for example?"

Can also be true by Red Hat *NOT* getting a database market entrant since Oracle may start pulling away when they get Solaris + Oracle under one roof! Will that leave Red Hat in the dust?

[cesarr77]

RedHat is making a HUGE mistake and will pay the consequences.

MySQL should be a RedHat customer migrating users from expensive Solaris boxes and databases to commodity hardware with RedHat Linux and MySQL.

We have the technology that RedHat needs to accelerate database migrations. I can only hope they work with a company like Monty Program Ab and us to really help customers migrate to a cheaper "stack".

For more information about our migration technology visit: www.ants.com/acs.

Cesar.

[MarkyGoldstein]

They all own some share of Red Hat already.

[Matt Asay]

Very true that they own part of Red Hat, but this doesn't mean Red Hat should avoid stepping on their toes.

[royrusso]

"Play it Safe". Doing otherwise would require: Upper-management with an appetite to take risk and destroying middle-management fiefdoms. Neither of those will happen at RHT in your lifetime.

[Mr. Dee]

Matt, its easy for you to write these things because you don't work or run the Company. Sorry, but Redhat is in the business of becoming a major player, they want to become what Microsoft, Oracle is to the Enterprise which will undoubtedly upset partners but make others happy. Still I just see Redhat remaining a consistent Company, but an ineffective one where growth carries them to the top. Microsoft, IBM, Oracle and others will ensure they stay where they are. I would be surprised if Redhat gets so frustrated and ends up merging with Novell.

[Matt Asay]

I'm trying to understand what you just said, Mr. Dee. I think you tried to disagree with me, but I can't see anywhere that you actually disagree with me. Which would mean that you agree with me, but you seem to be trying to avoid that....So, what exactly did you mean?

[Mr. Dee]

They want to be a big player, but they can't because that opportunity will be thwarted by the major players. They will attempt desperate measures to become big, but it will not work to get them where they ultimately want to be. If it works out in their favor they will say, great for us, if it doesn't they will say no problem, we will just work with what we have. At the same time, I am confused by their long term strategy, which probably is why my previous post confused you.

[realwariola]

Red Hat will become major player sooner or later. But it doesnt have to be in the DB to do that. It can support all DB on its platform for now so that it can grow its JBoss business supporting all DB equally. And MySQL is toast anyway nowadays and MariaDB is the way forward. Red Hat aggressively growing its JBoss middleware by acquiring Amentra to do its consulting and re-architecture JBoss to support multiple runtime and frameworks with JBoss 5 which will be release soon.

On the other end, RHEL business will be generating growth by the acquisition of Qumranet and the KVM support will be debuting this September. This will pit Red Hat against VMware and Microsoft more effectively than ever. They are focusing building end-to-end complete Virtualization stack with VDI, Management tools, Thin provisioning, integrated policy auditing (RH IPA) etc under the brand RHEV.

So all ways are pointing into the right path for Red Hat and they will become major player sooner or later.

[Matt Asay]

Unless, of course, Oracle's Linux initiatives pay off. Or IBM decides to buy Novell/SUSE. Or Canonical undermines its model. No one else is standing still, but you seem to think Red Hat doing so is a good strategy. I disagree.

[MisterGee]

It is time for Matt's weekly who will buy Red Hat column. Does it drive more traffic than your other speculation? If you look at how Red Hat OS revenue breaks down, a huge portion comes from HW OEM relationships. Red Hat being bought by a HW vendor makes little sense (although HP does need a Java stack to call their own since they killed Bluestone). You then have to look at either a non-server hardware vendor (storage, network, etc), or a software company. Cisco? It has been reported that Cisco was part owner of Qumranet before Red Hat bought it.

At this point why does anyone have to buy Red Hat? There is no imperative. Red Hat is mutually assured destruction. Any server vendor that buys them also drives away a chunk of sales. It seems like speculation just because other companies are being bought.

[Matt Asay]

Actually, if I wanted traffic I'd post about Apple. I write about Red Hat and acquisition speculation/product mix changes because this is the way Red Hat will have to grow.

Yes, Red Hat does a lot of its business through hardware OEMs. So does Microsoft, but it has never shied away from stepping on those partners when it has had to (nor have Red Hat's hardware OEMs shied away from stepping on its toes, either, for that matter). Red Hat is actively trying to expand beyond its hardware dependence: witness its work with Synnex. It's a major initiative there.

It would seem that you, too, would like Red Hat to keep to its comfortable place while ORacle et al constrain its ability to grow. I think that's unwise. I'm surprised that you don't, too.

[kragil]

Matt, buying a FOSS company is tricky for proprietary software/hardware companies. At the moment RH is all FOSS and that would change and some employees/execs might just jump ship and start White West Linux and instantly have a lot of interest.

RH is GPL all the way and that actually is their best growth protection.

[Matt Asay]

It's also Red Hat's best growth prevention. GPL will not work (on a big scale) in applications, or really anything beyond its core (by its CTO's own admission). I'm a HUGE fan of Red Hat, but I think it will struggle to scale far beyond $1B without moving well beyond the OS and its current licensing model. I'd love to be wrong but I've yet to see any examples that would suggest Red Hat's model will work beyond the OS and app server.