Qt5 WebEngine and Python3

Bernhard Rosenkraenzer bero at lindev.ch
Sun Sep 27 16:36:27 BST 2020


On Sunday, September 27, 2020 16:59 CEST, Neal Gompa <ngompa13 at gmail.com> wrote: 
> On Sun, Sep 27, 2020 at 4:53 AM Adriaan de Groot <groot at kde.org> wrote:
> > The
> > results are in the FreeBSD port as one big patch to the Python bits of the
> > qt5-webengine tarball; apply them after configure.

Thanks, that will be very useful. I suspect Google will release its own fork of python2 when it finally goes away (because they think that's less effort than fixing all their builders -- not just chromium/blink, but also Android etc), but I hope distros won't go along with this insanity.

We were thinking along the lines of doing what we're doing with the other web rendering engine that requires build tools from decades ago (we just build a minimal version of autoconf 2.13 inside the firefox package so we don't end up propagating that broken obsolete thing), but of course this is better.

Of course turning it back into KHTML (replacing the whole build system with cmake, getting rid of some "We don't want to use Qt and KF, so we reimplement some classes instead"-isms, ...) would be even better, but keeping that in sync with upstream releases wouldn't be feasible (even worse than fixing QtWebEngine every time a new Chromium is released), and I doubt they would accept any patch that eliminates googleisms upstream.

> "Demon offspring of our own web thing" is an utterly appropriate way
> to describe this. :)
> 
> If I am allowed to grumble for a second, I wish Qt hadn't dropped
> QtWebKit. Unlike QtWebEngine, QtWebKit was much more nicely integrated
> into the stack and didn't suffer from nearly as much insanity.
> 
> It's a shame that it's gone and we're stuck with this. :(

It isn't gone, there's a more or less active fork, last commit 6 days ago.
https://github.com/qtwebkit/qtwebkit/tree/qtwebkit-dev
But of course since it has been dropped from Qt, pretty much everything that used to use it has moved on, and I doubt we want to depend on both.

ttyl
bero



More information about the Distributions mailing list