Original URL: http://www.theregister.co.uk/2006/04/26/brown_borland_tools/

Borland wants to be a Red Hat for developers

Lead generation

By Gavin Clarke

Posted in Developer, 26th April 2006 02:06 GMT

Borland's Java and Windows application development tools will once again lead the industry, setting the pace in both quality and innovation. So says Nigel Brown, who is leading Borland's integrated development environment (IDE) business to independence and out of the shadow of Borland's other application lifecycle management (ALM) tools. Brown, an 18-year Borland veteran, was recently named general manager of Borland's IDE business.

In an interview with The Register, he promised significant investment in research and development for Borland's IDEs along with significant increases in sales and marketing for the popular JBuilder and Delphi programming environments among other tools. Brown is also rehiring engineers and architects who were at the forefront of developing languages like Delphi but who had left Borland.

Borland has long been the victim of a brain drain to other software companies, most notably Microsoft, which spent years playing catch-up in terms of features and language design.

Brown vowed Borland's spun-out IDE business would recapture its lead. "Four to five years ago, what we put in Delphi one year, Microsoft copied one or two years later. Our goal is to get back to that scenario where we are leading. Also JBuilder - that drove Java to the limits. Our goal is to get back to that," he said.

He blamed Borland's surrender on a legacy of under-investment in tools thanks to corporate priorities that came to emphasise the bigger ALM picture, with more R&D pro-rata going into Borland's Caliber, StarTeam and Together software. "Over the last few years [tools] haven't been part of Borland's key investments... that has been ALM product. Our immediate goal is to get that [investment] to, or above, what [other] software companies put into their leading products."

This begs the question: where will Borland get the cash to sustain the R&D and sales and marketing? budgets. The company is asssessing suitors to help fund and develop the IDE business, which also features the company's database and application server.

One reason Borland has been forced to spin out tools is because of the price pressure the IDE market is experiencing. The increase in "free" tools has forced IDE providers to re-examine what exactly it is they offer in order to make any money.

In common with chief executive Todd Nielsen , Brown is keeping exact revenue plans pretty close to his chest but repeated his boss's wording on providing "services." Like Nielsen too, the words "Red" and "Hat" are employed and there is a fervent belief the new company can monetize its installed base of three million developers.

"There is a model with the likes of Red Hat making money from support and added services. Borland is a global company and the IDE business will be global," Brown said. "It will be a matter of how can we get more of the IDE products into developers hands for good money, and [being] less hung up on licensing model," Brown said. "When we spin off the IDE business we will be a big startup."

Another challenge is who will buy Borland's tools. One reason Borland moved into ALM and upped its pitch was to target executives and IT directors who - unlike developers - hold the keys to the large purchasing decisions. It seems Borland's IDE business will try to shift pricing to suit the budgets of developers, whose spending runs to several thousand dollars and can be put through on a credit card. "We won't be speaking to CTOs except in rare instances," Brown said.

Borland is evaluating subscriptions, but Brown noted there is unlikely to be a change in the short term to the current licensing-base approach. The IDE business also expects to use web-based marketing to reach developers.

Brown is placing his faith in the concept of "smaller means more efficient." He believes a smaller organization, independent of the large Borland machine, will have more resources to develop features and functions instead of just "writing infrastructure." Brown estimates that two thirds of resources for JBuilder went into building infrastructure and just one third into features. ®