-
Want to receive the latest contracting news and advice straight to your inbox? Sign up to the ContractorUK newsletter here. Every sign up will also be entered into a draw to WIN £100 Amazon vouchers!
The Generic API
- Thread starter wurzel
- Start date
- Joined
- Jul 6, 2005
- Messages
- 36,223
- Likes received
- 5,502
- Thanks received
- 1,182
- Location
- Covid City
Actually, he was. Were you there?
- Joined
- Jul 6, 2005
- Messages
- 63,203
- Likes received
- 5,553
- Thanks received
- 1,318
We don't pick you up on your Russian typo's so lay off our English ones.
******* MAKE ME STOP!!!
- Joined
- Sep 11, 2008
- Messages
- 27,898
- Likes received
- 550
- Thanks received
- 161
- Location
- Euroland
If it was in Wognum, NL, yes. What a f**king pile of tulipe.
Metadata and operational data all crammed in the same tables, nobody that knew the data model (basically there wasn't one), and some twunt who said 'we don't use referential integrity in oracle because it makes the performance worse'. FFS, the whole point of a RDBMS is referential integrity, but would Bob call Oracle and check his assertions with them? Of course not.
I didn't even get paid for half that contract because clientco AND agent went bust. The only good thing is that Bob didn't get paid either.
Last edited:
10:45
- Joined
- Apr 24, 2008
- Messages
- 17,844
- Likes received
- 990
- Thanks received
- 182
Intelligent Finance used that when I was there. They ended up putting 15 millions lines of code round it.
- Joined
- Jul 6, 2005
- Messages
- 63,203
- Likes received
- 5,553
- Thanks received
- 1,318
- Joined
- Oct 28, 2006
- Messages
- 36,581
- Likes received
- 3,172
- Thanks received
- 1,186
Yet another Oracle Consulting project of legends - kerching!
- Joined
- Jan 28, 2010
- Messages
- 16,178
- Likes received
- 516
- Thanks received
- 117
- Location
- A bit too close to Wandsworth
Yeah, kind of, except it wasn't "sparse" data.
To be fair to said database vendor they advised in the manuals that each item (an in-flight proccess in a workflow system) have a small number of attributes corresponding to process state and keys in a properly designed relational schema.
Unfortunately the people who implemented the system at clientco didn't read the manual and every data item ended up as an atrribute of the associated process.
You can guess how the performance scaled with a few million processes and subprocesses (which would duplicate the data items they required) each having several hundred associated pieces of data.
