It was reported on MSO that it is possible for a user to upvote or downvote more than once on a single post (resulting in reputation changes for the author for each vote).
After some data examination, we were able to confirm that this vulnerability exists (it is happening due to race conditions s...
Seems weird. A race condition that allows for duplicate voting? I guess this is what happens when you try to enforce data integrity in the application instead of the database?
@PaulWhite Technical debt is also an accumulation of risk as well as baggage. It's not debt in the sense of an overdraft or mortgage. It's more like owing money to the mob - if you don't pay it down they eventually come round and break your legs.
It's fun sometimes to pretend that real databases were designed at all, rather than just coming into existence as raw data storage in a rush to get some sort of product to market.
If you're a startup, you don't have the money to pay experts. When you are successful, you focus on expansion and not tech debt. It's gotta take a near-death experience to make it a priority
anecdotally of course. I haven't done any research
@PaulWhite Don't talk to me about NOLOCK - finally got one of the teams to clean up 1,400 NOLOCK hints in their ETL codebase, and it took the enterprise architect threatening to withhold signoff to get them to do it. Note that this happened on an IFRS17 project feeding their general ledger and external reporting.
Muppets.
Oh well, where there's muck, there's brass. Keeps us in a job, I suppose.
this is the user table, which only has 431 foreign key relationships. Oh and all tables (nearly all) have the last 5 columns listed, which is great fun.
and the reason for the table having 431 foreign key relationships directed at it.
I'm not sure, but I think deleting a user might be a fairly expensive process.
to print it out at 100% zoom level would take 336 48x36 inch plotter sheets.
and they say size doesn't matter
I think it's likely a good thing they didn't mention any of this in the job interview
@HannahVernon Needlessly so, I think. It's the preliminary ETL for a general ledger feed on an IFRS17 programme. It's the worst case of enterprise fizzbuzz I've ever seen.
And really badly implemented. £12 or 15m depending on who you ask, and none of the decision makers have any background in large technology projects.
It's a candidate for the most incompetently managed project I've ever seen. And I mean incompetent in true sense of the word. Even the consultants from a nameless big-5 firm have no background in large technology projects.
Dunning-Kruger effect in action.
They don't know what they're doing and lack the cognitive faculties to realise they don't know what they're doing.
I sure hope that this UPDATE statement I'm looking at, in all its 1.5 MB, 22800 lines of code glory, with 4150 parameter markers, was generated by some monstrous tool and not written by some equally monstrous human.
I sensed a great disturbance in the force. It was as if millions of chickens cried out in terror and were smothered in mayonnaise. My 11 herbs and spices must be cautious