Hacker News Reader

Comment by augusto-moura | Hacker News Reader

Most of the divergences are already compile flags or command line flags on upstream. They are just disabling it by default. Is not a Brave-only thing, but different browsers have different settings disabled, the community that compiles the browser from source also pushes those features to be conditionally added. Eg. Debian, Arch or other distros, could have WEI disabled on their chromium distribution, forks like Edge and Vivaldi could disable GAIA, etc.

Just give a look at the divergence issues and the PRs solving them [1], most of them are just single file or small changes to compilation/default flags

[1]: https://github.com/brave/brave-browser/wiki/Deviations-from-...

View on HN

They have much more changes than just compile flags. Here's the repo where they maintain their patch set: https://github.com/brave/brave-core/tree/master/patches

If you open the patches and see their content, a good part of them are just one liners, or small file changes

Cumulatively they have 8879 lines of patches, the largest being 172

    /tmp/tmp.KVE7aSEvue/brave-core/patches $ wc -l *.patch | sort -rh | head
      8879 total
       172 third_party-blink-renderer-modules-webgl-webgl_rendering_context_base.cc.patch
       112 chrome-BUILD.gn.patch
       104 chrome-installer-linux-debian-build.sh.patch
        76 chrome-android-java-AndroidManifest.xml.patch
        72 components-vector_icons-aggregate_vector_icons.py.patch
        68 chrome-android-BUILD.gn.patch
        61 chrome-installer-linux-BUILD.gn.patch
        57 chrome-browser-resources-settings-privacy_page-personalization_options.html.patch
        55 net-socket-socks5_client_socket.cc.patch

Not insurmountable, but not as easily hand-waved, either.

One might say -- "but a lot of that's just context!"

Those are significant lines of code that these changes must play nicely with. There's actually more, these numbers do a disservice to everyone

Can you link some _substantial_ changes? I've clicked on random on 10 of the files and haven't seen more than 2 line change.

Cloned it locally, used 'wc' to do the work:

    /tmp/tmp.KVE7aSEvue/brave-core/patches $ wc -l *.patch | sort -rh | head
      8879 total
       172 third_party-blink-renderer-modules-webgl-webgl_rendering_context_base.cc.patch
       112 chrome-BUILD.gn.patch
       104 chrome-installer-linux-debian-build.sh.patch
        76 chrome-android-java-AndroidManifest.xml.patch
        72 components-vector_icons-aggregate_vector_icons.py.patch
        68 chrome-android-BUILD.gn.patch
        61 chrome-installer-linux-BUILD.gn.patch
        57 chrome-browser-resources-settings-privacy_page-personalization_options.html.patch
        55 net-socket-socks5_client_socket.cc.patch

All together, nearly 9000 lines of patches. The largest currently is 172

Sure, some of this is inflated by the context of nearby lines - but we don't get to ignore those, patches must play nice.

Interesting that the few that I randomly clicked on were css changes. Not a reskin indeed.

> Cookies are given a maximum lifetime of 7 days for cookies set through Javascript and 6 months for cookies set through HTTP

This might explain why my Steam account is always logged out.

It's non-sticky for me in Firefox too.

Yet it's always logged in for me.

If you put your session to resume where it left off, all cookies are stored as session cookies and last forever.

What do you mean by "put your session to resume where it left off"?

In chrome or chromium it would be in the settings somewhere… reopen tabs when you open the browser?

Any fork of chrome is probably doing the same thing with regards to session resuming and session cookies.

Settings - Get Started - "Continue where you left off"

https://ibb.co/tC5GVNx

I wonder why they disable scroll to text fragment, I always thought that was a neat feature

Some more context here: https://github.com/brave/brave-browser/issues/8342

You can leave a comment if it's something you'd like to see as an option

Reply on news.ycombinator.com