{
  "$type": "site.standard.document",
  "bskyPostRef": {
    "cid": "bafyreiewkp74wzyglen2hhvvjrgp26jy7pmv7gitqhnm4nifyj4druddoi",
    "uri": "at://did:plc:haakkg7y3xdghcdmprxeexso/app.bsky.feed.post/3mhxifze3mzt2"
  },
  "path": "/t/why-i-dont-like-proton-my-thoughts/36561#post_13",
  "publishedAt": "2026-03-26T09:26:16.000Z",
  "site": "https://discuss.privacyguides.net",
  "textContent": "It’s not always irrational imo but in this case it is since they’re outsourcing the most privacy-sensitive part of the feature to Google Play services.\n\nPush notifications are different imo. With FCM, Google is in the delivery path, but whether Google sees the actual notification content depends on how the app implements it. If an app sends a normal FCM notification or data payload, then Google necessarily handles that payload in transit. If the app is privacy-conscious, it can instead send only a wake-up ping/token through FCM and fetch the real content from its own server afterward. So relying on Play services for push can be a pragmatic compromise. It’s not ideal of course but it’s not too bad either. There are also non-Google alternatives (like UnifiedPush) of course.\n\nDocument scanning is different imo. Here Play services is not just relaying a small wake-up signal. It gets camera access and runs the scanner flow itself. Google says the scan flow is on-device, which is better than cloud OCR, but you’re handing raw document images to a proprietary Google component. That being said we don’t know if they’re using ML Kit but I suspect they do.\n\nEdit: It might also be that I’m totally misunderstanding something about how ML Kit works so take it with a grain of salt",
  "title": "Why I don't like Proton (my thoughts)"
}