2 ms·
Little of both. We were pretty sure it would already be impossible, or nearly so, for a malicious user to probe our infrastructure this way, but when it's so s
by throwanem 10y ago
Little of both.
We were pretty sure it would already be impossible, or nearly so, for a malicious user to probe our infrastructure this way, but when it's so simple to be even more sure, why not?
Similarly, we'd already observed a low but nonzero rate of users inadvertently providing such addresses - not during signup or onboarding so much, but in recipient lists they submitted. Since we used the same recipient checking code everywhere, why not cut that back to zero, too?