Nobody should need YOUR crash reports. First line defense should always be internal crash vulnerability identification and closure. Information from users can t

Nobody should need YOUR crash reports. First line defense should always be internal crash vulnerability identification and closure. Information from users can then take two forms: user initiated reporting and opt-in automated user reporting. Nowhere in that paragraph was their any technical reason that someone concerned with both getting crash fixes but avoiding crash reporting themselves should be told they can't have them both. Also, opt-in silent updates don't inconvenience anyone. The user gets whichever situation they are most comfortable with. Opt-in crash logs and opt-in silent updates should both be logged for easy user perusal. -- None of this inconveniences developers or users. -- The problem with any centralization of non-opt-in unlogged data gathering is it creates completely unnecessary hazards and trust problems. Once an organization is pulling data, that data is now available to (1) incentivize the organization to use it in ways the subjects many never have considered, (2) incentivize the company to increase the amount of data they pull for good or ill, (3) is potentially a target for security breaches, and (4) is a potential target for governments to coercively surveil users by accessing the data or encouraging more data to be pulled. None of the problems in the last paragraph will be apparent to users until they have already suffered damage. So concern about non-opt-in unlogged data gathering is extremely prudent.