Re: [Qlandkartegt-users] Compiling Error ‘qFormatLogMessage’ was not declared in this scope

[Bas C. <seb...@xs...>]

On 2015-09-30 09:38, Oliver Eichler wrote:
> is that really necessary? I thought the benefit of Linux is to get
> updates fast and frequently. I have a certain understanding that it
> takes some time for the latest and greatest to hit the binary
> distributions. So patching a Qt5.5 feature is kind of ok. But Qt5.3 is
> over 1.5 years old. And Qt5.4 roughly a year.
> 
> To be honest I hate #ifdef because the code gets harder to read and to
> understand. You can't avoid it completely. Especially for the portable
> stuff. But for transient things like library versions I would like to
> keep it to a bare minimum. Is there no other way?

I'd like to provide QMapShack updates to Debian stable users too, so 
they're not forced to compile it themselves and have a newer version 
available than 0.6.0. Because of the Qt 5.4 requirement for QMapShack 
1.3.1, I cannot update the backport and have to leave it at 1.3.0.

It will take about two years for the next Debian stable release to be 
available, which will include Qt 5.4 (and possibly newer) and the 
QMapShack updates it enables. A strict interpretation of the Basic Rules 
for Backports [0] I should remove the QMS 1.3.0 backport because it is 
not in testing, 1.3.1 is, this brings QMS users back to the stone ages 
version and feature wise so to speak. I won't interpret the backports 
rules that strictly, because keeping the backport at 1.3.0 is preferable 
over forcing users back to 0.6.0. Backporting security fixes is 
complicated by not having the same version in testing, hopefully no 
security issues will be found in QMS to keep this a theoretical concern.

I'm not sure what the Qt 5.4 availability in other distributions is, but 
everyone still at 5.3 is stuck with QMS 1.3.0 just like Debian stable 
users.

[0] http://backports.debian.org/Contribute/#index7h3

Kind Regards,

Bas