As professionals, we focus a great deal on integrity, both professional/personal, as well as data integrity.� However, I think the two are intertwined inexorably.� But, that's a post for a different time.
For now, about data integrity, particularly data feed integrity.
I've asked business partners, colleagues, peers and vendors, and get the same answer, I don't think it's just me.� Developers and system integrators have gotten so lazy and used to reliable communications, that it seems developers no longer consider data integrity components of value when developing their applications and reading/writing external resources.� This, naturally, creates a ripe target for vulnerability exploits through the use of fuzzers, and also a soft-underbelly that seems rather easily missed by testers.
So, pardon the curmugeonly stance, but I'm crying fowl1.� Developers MUST get their ducks in a row when it comes to data integrity!� Batch totals, hash totals, trial sums, header/trailer records, file locking and semaphores, validation on access of any resource, record counts, signed hash, all of these are necessary, yet it seems a lost art.
What is your experience?� Do you also find developers and integrators woefully deficient in ensuring processing file integrity?
1- yes, spelling is intentionally munged.� Please don't comment and cry foul. Or fowl.
