On Sat, Jan 10, 2026 at 11:23:56AM +0100, Peter N. M. Hansteen via mailop wrote: [...] > My impression, or at least the way I read what was in that communication, is > that > > * It's a large code base that has been evolving over a *long* time > > * The system is complicated enough and with enough factors (in the hundreds) > involved that feeding the same message through the system several times is > likely to produce different results each time > > * The messages the system produces for external parties to see are unspecific > at best and may in fact point to factors other than those that actually > determined the pass/no pass decision
These are perfectly reasonable and expected of any decent spam-filtering system. There are a lot of signals under consideration, the strengths of which need to adjust dynamically in the face of changing threats, and you don't want to reveal to your adversaries exactly how you detected them. > * The code has passed through many hands, and I at least get the impression > that nobody currently there can honestly say they understand all aspects > of the system And many of us have probably worked on such a mudball at some point in our careers. It often correlates with a large dysfunctional corporation with far too many tentacles and no clear mission beyond Number Go Up. Institutional knowledge is constantly defending a war of attrition from stack ranking, blind allocation, a culture of prioritising new features over keeping existing infrastructure working, and going all-in on the latest CEO fad. Not all of Big Tech is like this of course, but it's a common theme. I'm weird and actually *enjoy* working on such mudballs, at least when I'm allowed to fix and improve them. (CV available on request, unless you're Google.) > Despite all of this, they trust the system absolutely, claiming that it > has a negligible false positive rate. At Google, the system cannot fail, it can only be failed. Any apostates who disagree get managed out, leaving behind only true believers. You can't fix something if you can't see that it's broken. [...] > I had thought up a really snappy and harsh one-liner to end this one, but > I'll save that for another occasion. I've got one which is a mere two words, the first of which is Anglo-Saxon in origin and would likely get me thrown off this list. Besides, I'm saving it for the next time I'm approached by one of their recruiters. _______________________________________________ mailop mailing list [email protected] https://list.mailop.org/listinfo/mailop
