It finished building. I confirm that it runs correctly.

Did that segfault issue get fixed yet? The one that happens when limits are set low (512). Doesn't affect anything on default sys limits as far as I konw. I've been paying attention to discussions while I was away.


Am 29.08.26 um 16:35 schrieb [email protected]:
Welcome back! Thanks for the praise, though I think landry did most of
the work that I just adjusted for librewolf. Sounds good; I too look
forward to continue working with you and the ports team.

I missed the removal of Xtst, and as for freebl3, it was pulled in from
previous a port-lib-depends-check suggestion and wasn't sure to
remove it when it was no longer suggested in 154.x.

The patch applied cleanly. I am currently building qt6-6.11.2 and some
reverse dependencies, so I might be a while before I can compile this. Sorry.

I think it looks like it should do just fine though. Might miss
offline translation personally, but honestly seems like a sane
default. Don't see any grammar, spelling, or other issues.

If you already built this subsequent iteration this then I would concur
getting this into cvs.

Thanks for the update.

Hope this helps. Happy hacking everybody! May you have a good one.
--
yaydn

På lørdag 29. august 2026 kl. 16:51, skrev Leah Rowe
<[email protected]>:

Ok, here is the new patch attached.

I'm happy for this version to be reviewed and, I hope, merged. The
difference versus the previous patch is that it reintroduces and expands
the disablement options via GenerativeAI, thus:

+      "GenerativeAI": {
+        "Enabled": false,
+        "Chatbot": false,
+        "LinkPreviews": false,
+        "TabGroups": false,
+        "Locked": true
+      },

It still introduces the individual toggles via AIControls as in
www/mozilla-firefox, turning everything off and locking the toggles as
in the Firefox port, however:

I re-added the GenerativeAI disablement config because the
JustTheBrowser project does both:

https://raw.githubusercontent.com/corbindavenport/just-the-browser/8054bf40874b5fc5e51cdb1af2a188cd5c9e2426/firefox/policies.json


Just The Browser does the same as OpenBSD's firefox port, and was in
fact the basis for that setting being introduced recently in OpenBSD's
firefox port, disabling the AI via individual toggle options. However,
it still also disables via the GenerativeAI toggle, which OpenBSD's
firefox port removed in favour of just disabling via AIControls.

Therefore, this is now a slight divergence in the LibreWolf port versus
the Firefox port. The Firefox port currently disables only via
AIControls, whereas my LibreWolf port does, with this patch, disable it
via both AIControls and GenerativeAI toggles.

This is probably excessive, but I agree with the policy that we should
not enable AI in the browser. Doing so needlessly multiplies the amount
of electricity your web browser uses, frivolously for a feature that
most people don't use. If someone wants to use an AI, they always can,
but there's no reason why it should be integrated directly into the browser.



Am 29.08.26 um 15:29 schrieb Leah Rowe:
don't merge it. i'm going to make another change to the patch, and
re-submit.

Am 29.08.26 um 14:21 schrieb Leah Rowe:
Hi everyone,

I hereby announce that I am resuming my regular work maintaining the
LibreWolf port in OpenBSD. I announced two months ago that I was
taking a brief hiatus from the work, and I left yaydn exclusively in
charge of maintaining the port.

I would like to continue working with yaydn, as this person has done
stellar work while I was away, so maybe we can work something out.

Anyway:

Here is a patch (sorry, I still haven't learned CVS. I patched the
GitHub mirror).

It fixes the following:

* Brings AI policy in line with www/mozilla-firefox. LibreWolf
already turned off AI the same way OpenBSD's ff port does, but the
OpenBSD ff also locked those toggles. I've locked them accordingly in
the LibreWolf port, as applied in the ff port.

* Minor update to mozconfig (MOZ_APP_REMOTINGNAME is now properly
exported as per upstream policy)

* Redundant WANTLIB entries removed from the Makefile

I say again:

I'm back. I shall continue maintaining this port full-time now. I'm
early. I said I'd be back in October, but that I might be back in
late August or so. The "might" clause has been executed :)

--
Company director, Minifree Ltd
Registered in England, No. 9361826 | VAT No. GB202190462
Registered Office: 19 Hilton Road, Canvey Island, Essex SS8 9QA, UK


--
Company director, Minifree Ltd
Registered in England, No. 9361826 | VAT No. GB202190462
Registered Office: 19 Hilton Road, Canvey Island, Essex SS8 9QA, UK

Attachment: OpenPGP_0x5C654067D383B1FF.asc
Description: OpenPGP public key

Attachment: OpenPGP_signature.asc
Description: OpenPGP digital signature

Reply via email to