OSS Industry Leaders speak to vendors and venture capitalist (Red Hat and MySQL, 2 Articles)

Open source startups can hit profitability sooner than it took closed source incumbents, as long as they steer clear of rivals' costly business practices.

That's according to MySQL chief executive Marten Mickos, who told vendors and venture capitalists open source companies could cut their costs by not ploughing money into expensive sales and marketing activities.

However, start-ups shouldn't expect a fast track to growth just because they have adopted open source code or development methodologies.

"Maybe we [MySQL] come from Scandinavia, but open source is not socialism, it's not a party. Open source is not a business model. Open source is a smarter way to produce the goods and distribute the goods. It doesn't give you a biz model automatically," Mickos said.

Speaking at this week's Open Source Business Conference (OSBC), Mickos picked what he believes are the four most promising business models: advertising, licensing your product with a closed-source product using and OEM model, maintenance, and charging for enterprise-class features.

"Maybe today you have to be bigger to become profitable, but maybe you will get there sooner as an open source company then a closed source company. Closed source companies spend enormous amounts of money on sales and marketing," Mikos said.

Industry pundits estimate the traditional enterprise vendor model sees 75 per cent of revenue ploughed straight back into sales and marketing.

"How much did Linux vendors spend promoting Linux compared to IBM on OS/2? How much is MySQL spending on marketing in the database sector – not much. There's a very strong benefit for open source companies and they have a chance to rise sooner on this profitability curve," Mickos said. ®

OSBC Red Hat chief executive Matthew Szulik today told startups to forget the "romance" of open source and build businesses that compete with proprietary vendors on services and value.

Opening the Open Source Business Conference (OSBC) in San Francisco, Szulik said [advocates of] open source no longer need to legitimize its existence by talking about lower costs or the actual technology components. The challenge now is scaling to meet customer demand.

Szulik urged startups to rip up the prevailing Silicon Valley business plan of exit strategy via Google acquisition, and focus on delivering value - especially to enterprise customers.

"It's about how do you create a culture - that's very, very different to how do you create a product and sell that to Google," he said. "There's more than enough venture capital to fund early stage companies, but [the question is] how do you grow beyond a $1bn company?"

He spoke of trillions of dollars of opportunity in liberating IT infrastructure and data locked up in enterprise legacy applications and platforms. That means potentially migrating applications, building composite applications or adding a web-front end to legacy systems.

"The enterprise can't simply unhook billions of dollars they've invested in Cobol and legacy environments. That's an enormous opportunity for some clever mind," he said.

With large enterprises already using Linux and open source in mission-critical deployments and their infrastructure, open source has arrived. So the discussion has shifted to what Szulik called new sources of value - or service levels.

"The software industry wasn't trained to create a high degree of services. If it were, 45 per cent of your budgets wouldn't be tied to services and maintenance. What makes open source work, is not the one and zero but developing service levels to compete with the proprietary space [vendors] and scale beyond that."

According to Szulik, building a scalable business is "something Red Hat continues to have to learn".

Continuing the theme of unlocking enterprise data from legacy and proprietary systems, he made his first veiled reference to Microsoft, on Office and the global move towards open standards around document XML formats such as ODF.

According to Szulik "someone" (i.e. Microsoft) wants remuneration for storing and retrieving personal information "at a time when search and indexing is evolving".

He also picked up on the debate re-ignited by Microsoft last week when again laid claim to unspecified patents that Linux and open source infringe. According to Szulik, open source is not a "renegade industry" and respects intellectual property. But companies will derive little business value from devising and registering patents, he argues.

"There's little emperical evidence between patents and innovation, especially in the infrastructure space... we see patents really being a challenge to innovation. The industry moves much, much faster than any remedying process. If you receive a patent injunction it will be years, and you will be out of business, before there's a remedy," Szulik said. ®

"How much did Linux vendors spend promoting Linux compared to IBM on OS/2? How much is MySQL spending on marketing in the database sector – not much.

He forgets that it took Red Hat *years* before they were profitable. I understand his reasoning, but I'm not sure I agree with it.

Smart guy. Just like Torvalds, to him it's just a development model, the way to produce the goods.

Its no coincidence that Thorvalds and Mickos put out a good product and have achieved success in a far faster and more profound manner than stallman who is so kooky that even most in the OSS community shudder every time he gets up to blowviate.

we see patents really being a challenge to innovation

I have to agree with that, too. Look at RIM and the Blackberry. They seriously innovated, then some patent troll that had a paper patent that could be skewed to cover RIM's invention slammed RIM for innovating.

FYI, a "paper patent" is a patent where the "inventor" never builds a prototype, never builds a product, never takes it to market. He just sits on the paper hoping someone else independently invents something close to it and makes it profitable so he can sue. It's easy money requiring no effort -- the only invention on his part is the patent itself.

If we can't get rid of software patents, at least the patent laws need to be changed back to what they were, requiring actual creation of the invention prior to the application. You don't get the same thing in copyright, I can't tell the Copyright Office I'd like a copyright on a work I may never write so I can sue someone who publishes something close to what I may have written in the future.

In addition, failure to show a good-faith attempt to market the invention should be an affirmative defense in court, with the burden of proof on the plaintiff.

achieved success in a far faster and more profound manner than stallman

Now where is that Hurd kernel again? Stallman started it 17 years ago and it's only up to version 0.2! And he doesn't even have to write the microkernel, as he's using Mach.

I know it's microkernel-based, so it's harder, but that's no excuse. Tanenbaum wrote Minix from scratch and released it with textbooks in the late 80s only as a teaching tool. It wasn't much, just the bare skeleton of an OS to show students how it's done. He didn't start trying to make it a real production OS until probably after 2000, and now Minix 3 is already ahead of Hurd. And Tanenbaum only had a student and a couple programmers working on it.

Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.