By Simon Goodley Immature technology could cripple enterprise application integration company and erstwhile darling of Silicon Valley, CrossWorlds Software Inc. Well into the first year of the Burlingham, California-based company’s marketing efforts, CrossWorlds is still struggling to present an impressive reference site. Ross Altman, research director at analyst firm the Gartner Group says, It is […]
By Simon Goodley
Immature technology could cripple enterprise application integration company and erstwhile darling of Silicon Valley, CrossWorlds Software Inc. Well into the first year of the Burlingham, California-based company’s marketing efforts, CrossWorlds is still struggling to present an impressive reference site. Ross Altman, research director at analyst firm the Gartner Group says, It is difficult at this point to find a referenceable CrossWorlds’ account or a definitive demo of ‘plug and play’ enterprise application integration. UK mobile telecommunications firm Orange Plc, was CrossWorlds’ first European customer, buying the software in September 1997. Orange wanted to link its front-office package from Vantive Corp to its SAP AG installation, and planned to complete the project by the third quarter of this year. But it still looks some way from being finished. One of the problems was that the 16-bit API supported by Vantive version 4 wouldn’t work with the 32 bit CrossWorlds software, so Orange was forced to upgrade to a newer version. Orange project manager Jeremy Brown says, During the end of 1997 and early 1998, CrossWorlds spent a lot of time trying to get the software to work with Vantive 4. Eventually it gave up. The project restarted when the Vantive version 7 migration was imminent, which was spring to summer 1998. Things then got worse. According to Brown, There was then the issue of architectural testing. Orange tried unplugging CrossWorlds’ connectors and shutting down servers to see how resilient the system was. The expectation was that the system would not lose transactions. But it did. Orange was the first partner to do that level of testing on the CrossWorlds software. It was quite intensive resilience testing, and that’s what highlighted the problems. Because of the failed resilience tests, the project missed the planned implementation date in 1998. Orange says it is still working on the project, and now hopes to implement the CrossWorlds software in the first quarter of 1999. Bart Foster, VP marketing at CrossWorlds concedes, It’s true to say that there were challenges at Orange. Santa Clara-based networking company Bay Networks Inc, now part of Northern Telecom Ltd, also embarked on a six month project to link SAP with Clarify Inc’s front office software using CrossWorlds. Dave Maillet, IS Manager at Bay, says, We went through a lot of pain with CrossWorlds. We went through a lot of regression testing and found a lot of bugs. Bay Networks gave CrossWorlds a lot of help and now the product works, but there are still problems. We took a chance and suffered a whole lot. Maillet continues, The biggest problem with CrossWorlds is that it’s slow. CrossWorlds is a single threaded environment so you have to do one task at a time. The product could be improved on performance. The users won’t wait three seconds to see a contract. Foster again admits to problems, but remains optimistic. Threading can be an issue, he says, but the situation will be improved dramatically, if not completely fixed, with release of version 2 of the server in January. Nigel Southern, sales director at Diagonal Plc – a UK-based CrossWorlds partner, says, I think that the product has some fantastic concept features and ideals embedded in its architecture, and there is a definite requirement for this product in the market place. However, I think that currently it is a little early to bet your enterprise on it. It’s definitely worth watching, but right now its not there … probably because it hasn’t been in R&D long enough. CrossWorlds’ marketing has been someway in advance of its product delivery. CrossWorlds argues that is just suffering from normal teething troubles. But Bill Coleman, CEO of BEA Software Inc, says, I think that the jury is still out, but CrossWorlds certainly over-hyped its value proposition before it had a chance to perfect it. Dissenting voices are widespread – for whatever motive – and there are not currently enough supporters or satisfied customers to shout them down. So it may not even be a question of whether the technology works, will work, or even if CrossWorlds can hone its business strategy to more effectively take advantage of this huge new market. It may simply be a question of whether CrossWorlds is losing too much credibility to thrive or even survive. As another technology VP comments, CrossWorlds is the laughing stock of the Valley. For a company that is so image conscious, that must seem like a killer blow in itself.
A longer version of this article will appear in the December issue of Software Futures.