Skip to content

Blockerflow for Mac 1.1.0

Choose a tag to compare

View all tags
@khaleel-git khaleel-git released this 12 Oct 13:59
· 26 commits to master since this release
mac-v1.1.0
0994a04

Important

macOS may say "Blockerflow is damaged and can't be opened"

The app is not damaged. Blockerflow for Mac is distributed without an Apple Developer licence, so it is not signed or notarized by Apple, and macOS blocks apps it cannot verify.

After you drag Blockerflow into Applications, open Terminal and run this one command:

xattr -cr /Applications/Blockerflow.app

Then open Blockerflow again. You only need to do this once per install. You can also try right click, then Open, then Open again.

Support: help me get an Apple Developer ID (€99 a year) to sign Blockerflow for macOS and iOS, so it opens without warnings. Donate via PayPal

What's new in 1.1.0

  • Block adult images (new, on by default). Checks the screen for pictures and video the web filter cannot judge, such as a video in WhatsApp or a picture in Photos. While a browser, messenger, Photos, Preview or a video player is in front, it takes a screenshot about every 1.5 seconds and a model on your Mac checks it. Nothing is saved or sent. When something explicit is found it closes the tab (Chrome, Brave, Edge, Vivaldi, Opera, Arc, Safari) or hides the app, and shows the block screen.
  • Block suggestive images (lingerie and swimwear; beach and gym photos get blocked too).
  • Needs macOS 14 or newer and the Screen Recording permission (macOS asks at the first start). After allowing it, quit and reopen Blockerflow.
  • Turning either switch off needs your accountability partner, like Adult Content.
  • Black-and-white pictures are no longer skipped as if they were blank pages.
  • The oldest supported macOS is now 12.3. The download grew from 25 MB to 69 MB because it now carries the image model and the ONNX Runtime library.

How well it works: on one Mac (Apple silicon, Chrome) it blocked adult pictures from a search and left ordinary pictures alone, but one ordinary-looking search was blocked in an earlier test, so expect some false alarms and some misses. Keep the web filter on too. Not tested yet: closing the tab in each browser (the app hides the browser if it cannot close the tab) and Intel Macs.

Licence

The image check uses NudeNet 320n (https://github.com/notAI-tech/NudeNet), trained with Ultralytics YOLOv8, both AGPL-3.0, and ONNX Runtime (MIT). The AGPL-3.0 text: https://www.gnu.org/licenses/agpl-3.0.html.

Assets 3
Loading