To be fair, Microsoft has played its part in keeping Novell down, with launches such as Windows NT and Active Directory that competed heavily with Novell's mainstay Netware directory. Novell has also had plenty of self-made problems, principally its inability to communicate a story that non-technical management — the holders of the purse strings in many companies — could understand.
Recently Novell has been addressing this inability to present a coherent picture of what it does. The acquisition of Suse Linux in 2003 added some much-needed momentum to the struggling company, but some of that has now tailed off — thanks in part to a high-profile tie-up with Microsoft, which many in the open source community saw as treachery from a company that, for a while, had been happy to bang the open source drum.
This week, at its annual BrainShare user conference in Salt Lake City, Novell has come up with new strategy. Unveiled by chief technology officer Jeff Jaffe, the strategy is all about making IT infrastructure more agile and is named after a Madagascan relative of the Mongoose — the Fossa. Open-source advocates will quickly realise that Fossa, along with being a cat-like and very agile mammal, is also a play on the Free and Open Source Software (FOSS) acronym.
ZDNet.com.au's sister site ZDNet.co.uk sat down with Jaffe at BrainShare to discuss what Fossa really means, and whether the company has backed away from its previously pro-open-source position.
Q: What exactly is the Fossa project — is it more than a vision or strategy? Will we see a Fossa product suite, for instance?
A: We talked about seven technology pillars and really each of them is at a different level of maturity. The seven we talked about were virtualisation, Linux, orchestration, policy, identity, compliance and collaboration. In most of these areas, and most notably Linux, there is a well-defined community roadmap and a lot of the value-add we will be providing [will include] making it business critical; interoperability; and some of the longer-term road-map items.
Some of the areas such as orchestration and identity management — we are really building on Novell technologies and strategies that we have provided blueprints on for years. And with these technologies we want to build up the community and get everyone to identify with our vision. We want to get open identity services, we want to get every object in the world tied to our identity management.
I think it's a combination of things. I would say in some cases we are done, in some cases we are along the way. It is a full product suite, but there is a lot of innovation to come.
To what degree are those identity products open source at the moment?
Very little. We have some open-source projects; but it's still growing. From the point of view of where the customer wants go with agility, we need it all, but in practice it's going to mature at a different rate.
How much of this is about addressing the vision problem that Novell had in the past — in that you had good products but have not traditionally been great at communicating how they all fit together. You have also lost some momentum from not being as vociferous about open source since former chief executive Jack Messman left the company. Is Fossa an attempt to address all those issues?
We have for several years positioned ourselves as a mixed-source company. We are very excited about Linux but we have never said we are an open-source company. I think Ron [chief executive Ron Hovsepian] brings terrific customer perspective, and under his leadership we have made the mixed-source message much more explicit.
Mixed source as a technology message sounds a little confusing, but as a customer message it's reality. Customers have open source, they have proprietary, and they want them to work together and no one is helping. I think we have been pretty consistent about the mixed-source theme and I don't think our long-term vision is any different. Even embedded in the name "Fossa" is open source.
I also think that for the past year-and-a-half we have been clear that our strategy is to manage resources. The way we manage resources is in two domains. One is on a single system via enterprise Linux, and the other way is across systems via systems management products.
But we have never answered the question: what greater goal are you achieving by managing resources? And that is what we are trying to do today. Our Fossa project answers that question because there is an overarching goal that chief information officers want in terms of agility and, by focusing on that like no other company in the world, we think we can address that.Could you describe how your virtualisation products fit together and whether there are in gaps in the offerings you have at the moment, which you feel you might need to plug through internal development or acquisition?
On the other hand you do want to choose at least one technology as the preferred technology and in this particular case what makes most sense for us, as we think it makes the most sense for customers, is to select the open-source technology. For any key piece of technology, customers don't want vendor lock-in, they want openness, preferably open source.
We have been very big supporters of Xen and that will probably be our preferred strategy for the future. Supporting Xen has many different aspects to it: we are big participants in the Xen community, we were the first to embed Xen in Linux with SLES 10. When it comes to making Xen industrial strength, then we probably have the lead on any of the Xen community, simply because with OEX 2, with Netware running on SLES using Xen, we have many more real workloads, real customers that are supporting that today.
The other piece is the management piece, and we saw a gap that we filled with Platespin.
So do you think that you now have all the virtualisation products your customers require?
I wouldn't go that far. I think there are some important areas for investment. I think to make virtualisation really take off, there is a need for better tools. I think one of the other mega-trends that is going to drive virtualisation into enterprises, even more so than today, is multicore. As you get more and more powerful computers, you will begin to outstrip the need of the applications, and you have more and more wasted capacity unless you utilise it properly.