----- Original Message -----
> From: "Andres Salomon" <dilin...@queued.net>
> To: "1005083" <1005...@bugs.debian.org>
> Cc: "Timothy Pearson" <tpear...@raptorengineering.com>
> Sent: Tuesday, February 15, 2022 4:21:27 PM
> Subject: Re: Chromium 99 update
>
> Hi,
>
> So the good news is, I'm not opposed to these patches in theory,
> especially if you're working with upstream to get them merged. The bad
> news is the timing. I just took over chromium, and I've been working to
> get fewer debian-specific patches rather than more. We've gone from
> 70-something patches to 50-something patches, dropping irrelevant (and
> often undocumented) ones and getting some upstream. I still have a lot
> more patches to work through before I'm in a place where I'd be
> comfortable considering patches for a new platform. So I'd be open to
> continuing this discussion in the chromium 100/101 time frame.
Great to hear and definitely looking forward to continuing discussion as you
have time for Chromium 100 or so. I'll continue to update the patchset in the
interim and build (where practical) on the Raptor build farm -- the latter will
be made a bit easier as newer versions are backported to Bullseye.
> BTW, do you have links to any upstream gerrit or bug entries so I can
> see why upstream doesn't want to support a new architecture?
There is a long and complex history here, much of it out of the public eye, but
the long and short of it is that Google simply doesn't want to put effort into
supporting an architecture it doesn't use / ship hardware products (Chromebook
/ Pixel) for. Personally I have suspicions that it goes deeper, into the owner
control aspects of the POWER systems currently shipping (e.g. Widevine will
never work by design, and it will be harder to track users in the future
without various vendor components (Intel ME/AMD PSP/ARM Trustzone) subverting
the OS, but the latter is still pure speculation on my part.
At the end of the day it's likely a pure business decision from Google -- one
less architecture to support means fewer employees working on the Chromium
project overall and lower costs / faster code churn.
Here are some of the public bugs and Google's responses:
https://bugs.chromium.org/p/chromium/issues/detail?id=925171
https://bugs.chromium.org/p/chromium/issues/detail?id=1029662
https://bugs.chromium.org/p/chromium/issues/detail?id=1076455
Note I am willing to try to re-enage with upstream and see if there has been an
opportunity for merge created with e.g. Gentoo and Void shipping Chromium with
these patches enabled. If Debian (and by extension Ubuntu?) were to also add
the patches, that would help with the overall business case in terms of the
real-world need to have the patches accepted by upstream.
Thoughts welcome!
