>You are wrong, please read the PKGBUILD for tango-cpp, and see that
>omniorb needs to be < 4.3.0.

No, it is you who is wrong.

PKGBUILD is not upstream. Upstream's CMakeLists.txt only contains a comment, 
citing a minimum version for omniorb:
" omniORB>=4.2.4 is required for compilation in C++17 (or better) mode. " [a]

>Also this is used in production

Just because some admins don't update a shared library in production is not a 
good enough reason to not allow AUR users to use the latest stable omniorb with 
the latest stable tango-cpp.

>Finally please do read again the guidelines about naming, or take as
>example packages such as python310, which is clearly not version 310
>of python

Please don't cite a wrong example. AUR/python310 also violates Arch package 
naming guidelines.

But what the latter does not violate but omniorb425 does is: python310 does not 
conflict with the main python package, whereas omniorb425 does conflict with 
omniorb.

If you really want to (unnecessarily) stick to this old version, resubmit a 
non-conflicting package. I suggest the name to be omniorb-4.2, to allow the 
(hypothetical) possibility of updating the third version point (4.2.6 to 4.2.7 
if there ever will be such).

On 9 October 2023 13:09:12 GMT+02:00, Antonio Bartalesi 
<[email protected]> wrote:
>You are wrong, please read the PKGBUILD for tango-cpp, and see that
>omniorb needs to be < 4.3.0.
>Also this is used in production and having a new version "on the
>horizon" does not justify breaking the current installation.
>Finally please do read again the guidelines about naming, or take as
>example packages such as python310, which is clearly not version 310
>of python

[a]: https://gitlab.com/tango-controls/cppTango/-/blob/9.4.2/CMakeLists.txt#L15

Reply via email to