Australia's Bureau of Meteorology Dumps Water Data Project

littlekorea writes "Australia's weather bureau has racked up bills of $38 million for a water data system, based on Red Hat Linux, MySQL and Java, that was originally scheduled to cost somewhere between $2 million and $5 million. The Bureau's supplier, an ASX-listed IT services provider SMS Management and Technology, did a good job of embedding itself in the bureau, with all changes having to be made by the original consultant that built it."

Comments

Re:Reminder

[pejyel]

It is rude to randomly redirect visitors to beta.slashdot. Even more so because beta sucks.

Providing a hard to find opt-out, adding /?nobeta=1 to the url, just upgrades the aggravation level from "rude" to "insulting and infuriating". The only acceptable option is, as always, opt-in.

I guess you need reminding. a lot.

that's why everybody's heading to this page to talk about what our options are if DICE refuses to hear us.

on topic thread here

[crutchy]

this article seems to imply that linux was the reason for the cost blowout... and not that it was managed by a government agency.

look at any project administered by any government agency around the world... how many are on budget? why is that? it has nothing to do with linux and everything to do with government waste

TFA disagrees with submission summary

[maynard]

According to TFA:

SMS Management and Technology won the IT initial development contract in June 2009 in a deal initially expected to be worth at least $2.5 million per year.

The audit office questioned the value of the project, which is estimated to have reached $38.5 million for associated systems and applications by 30 June 2013.

So, across four years what should have cost $10M wound up costing nearly $40M. However:

Within 18 months, however, four change orders had been processed, increasing the value of the deal to over $15.4 million in the first two years

Thus, change orders from a client who changed milestones mid-stream:

The milestones for the contract were not tightly specified, nor was the extent to which the industry partner staff would be integrated with or separated from internal bureau IT staff roles and deliverables.

Leading to a situation where, "The contract began to resemble a time and materials contract rather than a fixedâfee contract contingent on achieving milestones and deliverables." Meaning that the client kept changing their mind so often the consulting firm was required to baby a system they hadn't thought through to begin with and had thus grown into a monstrosity that served disparate and disorganized goals.

No wonder it went over budget.

But that has nothing to do with open source and everything to do with bad project management. Notice that they've solved the problem by choosing "...a replacement, based on an off-the-shelf software product."

Which, if it meets their needs - bully for them. But is more likely an imposed solution to a problem they hadn't clearly defined to begin with. Thus, it's likely they'll find themselves in the same situation. Not because open source software is bad, or the commercial software is bad, or the consulting firm was probably bad... but because the bureau of meteorology has no idea what it wants to do with this data.

The problem here is with undefined goals set by management. Until they face that fact they'll go round this merry-go-round again and again. And taxpayers will foot the bill.