Hail to thee all!

Making a routine sudo yay -Syyu, I got a crowded list of packages ready to update. But I refused the update. The reason: this package.
giu@ Obarun ~> sudo pacman -Ss qt5-webkit
[sudo] Mot de passe de giu : 
extra/qt5-webkit 5.212.0alpha4-1 [installé: 5.212.0alpha3-9]
    Classes for a WebKit2 based implementation and a new QML API
community/gambas3-gb-qt5-webkit 3.14.3-7 (gambas3)
    Qt5 toolkit webkit component
More precisely: 
giu@ Obarun ~> sudo pacman -Si qt5-webkit
Dépôt                    : extra
Nom                      : qt5-webkit
Version                  : 5.212.0alpha4-1
Description              : Classes for a WebKit2 based implementation and a new QML API
Architecture             : x86_64
URL                      : https://www.qt.io
Licences                 : GPL3  LGPL3  FDL  custom
Groupes                  : --
Fournit                  : --
Dépend de                : qt5-location  qt5-sensors  qt5-webchannel  libwebp  libxslt
                           libxcomposite  gst-plugins-base  hyphen  woff2
Dépendances opt.         : gst-plugins-good: Webm codec support
Est en conflit avec      : --
Remplace                 : --
Taille du téléchargement : 12,55 MiB
Taille installée         : 48,95 MiB
Paqueteur                : Antonio Rojas <arojas@ archlinux.org>
Compilé le               : mar. 10 mars 2020 08:05:40
Validé par               : Somme MD5  Somme SHA-256  Signature
Knowing that the version of Qt5-base installed is: qt5-base 5.14.1-3...
knowing that Marianarlt has warned (in a sticky) NOT TO upgrade anything Qt5 before the Obarun home-made Qt5-base in on par...
Knowing, yet, that the installed version of qt5-webkit is already 5.212.0alpha3-9, which is superior(?) to the version of Qt5-base, and that nothing bad happened...
(And unless these two version numberings have nothing in common, in a deceptive fancy way, which I suspect...)

Is it Safe/Unsafe to update this package?

In advance, thank you for any answer! Regards
having just updated this package, and having it's dependent apps function, suggests to me that it's safe to update..
but yes, it's wise to be cautious..or at least aware of this issue, if a problem occurs.
Hi ncmprhnsbl!

Thanks for the quick reply and feedback! Yes, I try to be cautious. Maybe a little bit too much in this particular case. Thanks again! :)
The rule I have deduced is that Qt5 pkg versions must match qt5-base by obarun or should be held back.
Basically you copy all the proposed qt5-xxx pkgs of the update list and add them at the end of this command with commas, no spaces in between:

# pacman -Su --ignore pkg1,pkg2,.... etch

Or add them on an Ignorepkg line in your pacman.conf, pacman will remind you that you are ignoring them each time you run -Su
Good (yet simple) tip!!! Thanks to this --ignore option, I'm now processing all my pending upgrades, BUT the dubious ones. By the way, there is a new version of qt5-base available in the obextra repo, namely: obextra/qt5-base 5.14.1-4.

Just to be perfectly cautious, I'm now first upgrading qt5-base, and only after will I also upgrade qt5-xcb-private-headers --which has also reached version 5.14.1-4 since yesterday.

Yet, before setting this thread as solved, could someone confirm (or disprove) what I suspect, i.e. that, possibly, (?!) the qt5-webkit package numbering is a particular case and is not indexed to the other qt5-packages version numbering (!?)

Thanks in advance for any answer on that question!
yes qt5-webkit is a particular case.
The most important is qt5-base. if you see a release coming from Arch higher than the one coming from Obarun repo, do not update and wait for the repositories Obarun update.
I haven't researched this well, but in the qt hierarchy there must be a key pkg below qt5-base that is essential for all the others. What if in the qt5-base you add a restriction of such a pkg/pkgs to be of a version xxx or less, so higher editioned pkgs can't come in? I've seen it for versions to say higher than, I can't remember an example of less than.
@ Éric

Thanks for your reply. I'll then set this thread as SOLVED and change its title into a more explicit one, for better results to future researches.

Powered by Obarun