There are some things you just don't expect. You don't, for example, expect to find a red-hot chilli pepper in your lasagne and salad. But that's just what happened to me.
And the IT equivalent, I suspect, was a few years ago when Oracle CEO Larry Ellison sashayed onto a conference platform and held up a network computer (NC) for the Oracle faithful to kneel down to. Big Iron software meets the equivalent of the Sinclair Z80. The audience looked as though it had just found something suspicious in the Oracle salad.
Technical knowhow welded to strong leadership can take a promising start-up company to a fairly high business plateau.
But unless the basic technical knowledge is supplemented by strong business knowledge shared across a healthy organisation, a company can haemorrhage talent and money faster than Microsoft can make it. That Oracle should have done just this is still no surprise.
I always found it immensely entertaining to listen to Ellison at conferences. His drive had taken Oracle into a seemingly bullet-proof place in the database market - but he did the equivalent of a racing-turn, leaving most of his employees, including those at a very senior level, totally bemused and trailing in the road way behind him.
Making even a small tactical turn in a small company takes all the evangelical skill its most respected leaders can muster. It can be done very skilfully, as others have proved, but in the case of Oracle it simply was not.
Ellison thought that standing in front of his Oracle users and preaching would be enough to persuade them. It persuaded many. But it only takes one miscalculation to get it wrong and Ellison's was a common one - price.
He promised a sub-$1,000 NC and has not managed it. And as the prices of sub-Pentiums plummet nearer to the £500 mark, Oracle has lost the race to provide the money-spinner to rival Microsoft for ever.
It is also sad that Netscape, which owns one-third of Network Computer Incorporated (NCI), is adding to its woes with the long drawn-out saga of the NC and its lack of direction. And the boss of Oracle's NCI subsidiary has resigned after 14 years. This must take us back to start-up time.
Observers claim Jerry Baker left because he is technically gifted but does not have the experience to run a growing, fledgling company.
Baker, chairman and chief executive of NCI, will be replaced by Oracle executive vice president David Roux. It is also claimed that NCI struggled over the logistics of its merger with Navio, the software company that Oracle bought from Netscape last August.
Oracle has bitten off too much which is not core business, allowed executives with technical knowhow but little strategic business sense too much leeway, and has been too slow to develop the NC from a nice idea to the money-maker that Ellison intended it to be.
This must also cast a shadow over the Oracle and Cable & Wireless merger talks. I have no doubt that they are going ahead but every doubt that they will be successful.
Oracle is trying to rival Microsoft's Web TV offering. It seems as if the company is attempting to become a super box-shifter.
But no one is going to persuade me that set-top boxes and digital TVs are ever going to be its core business. That is the chilli in the salad.
Chris Bunch of Microsoft partner Cloudreach gives his take on this year's Inspire conference
We pull out the key information from Big Blue's quarterly results
Telford-based firm moves into the Nordics with Getac
Desktop 3D printer shipments see first ever year-on-year decline