January 19, 2009
As Giovanni Asproni says it “estimation is a fundamental activity of every project”. Or at least it should be. The meaning of “estimation” varies however often for the different participants to software development projects. For the customer, “estimation” means an almost accurate amount of money he will spend and time it will need before he gets what he wants. For the project manager, it is an early commitment to the customer and from the developer to do something for a deadline. For the developer, it is the amount of time that he thinks he could do the stuff, if everything goes right.
The experienced developer will often add a 50% cushion to his estimates, knowing that not everything goes right. You can have an additional meaning in the case of outsourcing. For the salesperson, the estimation is the maximum price he can ask, while still having a lot of chances to sign a deal. A lot of young engineers will remind with astonishment hearing in a pre-sales meeting a salesperson answering to the customer that everything he wants is indeed possible… and for a cheap price too ;o)
For a developer, estimation is often equaled to fixing deadlines. He could then feel trapped by the original commitment that created expectations for management and customers. Customers have often two fixed expectations: delivery date and functionalities. Therefore, the effort is the only variable left to adjust the project activity. You can often witness a poor transition from estimation to planning with project managers that dream to have a baby in one month, if only they could manage a team of nine women. Because initial expectations cannot be achieved in many projects, the finger in the air used for estimating get then pointed downwards for blaming people: first from customers to project managers, then from project managers to developers. So print Giovanni’s article and give it to your customer before your next project planning meeting.
January 12, 2009
Borland discreetly announces this week the resignation of both Tod Nielsen, CEO since 2005, going to VMware as COO, and Peter Morowski, SVP of research and development, leaving the company to pursue other opportunities. Borland will also reduce its workforce by approximately 130 employees, or approximately 15 percent of its regular full-time staff. For the fourth quarter of 2008, Borland is expecting to report total revenue in the range of $38.5 million to $40 million, down at least 10% from previous quarter. Considering that operating costs of revenues for the third quarter were around $45 million and that a lot of them are fixed costs, this will be another quarter with an important operating loss for Borland. The current CFO will act as CEO.
In the past year Borland has often blamed its IDE division to be the cause of this problems, slowing its ambition to be an ALM leader with huge financial losses. In May, Borland finally managed to sell the CodeGear unit to Embarcadero. The financial results of CodeGear are now difficult to estimate, as Embarcadero is a private company. What is left at Borland is a mixed set of products, resulting of some in-house innovation and external acquisitions (Togethersoft, Segue), grouped under the “Application Lifecycle Management” banner. None of the components of this set is considered as a top leader in its specific market and good integration between existing products is always difficult to realize. Trying to sell these type products is more difficult, because you have to reach a higher level in the enterprise than for individual developer products. It is also the type of project that most companies will postpone in a period of difficult economic conditions. Furthermore, this put Borland in competition with bigger fishes, like IBM that purchased rival Telelogic last year, and companies like MKS that have their roots in the ALM market.
It would be a bad thing if Borland fails just after having celebrated its 25th birthday, but if it was already struggling in times where the economy was good, we could fear that survival would be even more difficult in the hard times that are ahead of us. Researching for this post, I found that Borland had already cut 40% of its workforce in 1995 after the resignation of its founder Philippe Kahn, so maybe we should believe that it could be just “another deep crisis” in Borland history, a company briefly knew under the name of Inprise ;o)
Borland full press release
January 6, 2009
Methods & Tools is a free e-newsletter for software developers, testers and project managers. Winter 2008 issue’s content:
* Fingers in the Air: a Gentle Introduction to Software Estimation
* Behavior Driven Database Design
* Optimizing the Contribution of Testing to Project Success
* Service Components and Compositions
45 pages of software development knowledge that you can download from http://www.methodsandtools.com/mt/download.php?winter08