Re: This is going to break port building without poudriere!

From: Alexander Leidinger <Alexander_at_Leidinger.net>
Date: Fri, 26 Jan 2024 11:09:40 UTC
Am 2024-01-26 11:42, schrieb Alex Dupre:
> Alexander Leidinger wrote:
>> Did I misunderstand that Luca wants to convert master/slave ports like
>> my php case into subpages aware ports to cut down on package build 
>> times?
> Honestly PHP would be the worst example to be converted, PHP extensions
> can be built individually, they don't build the core PHP every time.
> There are other master/slave ports where a big percentage of the code 
> is
> compiled in all builds.

The extract/configure vs build ratio is bad compared to a "build 
everything" case. Other ports may cut down more on the build time. The 
reason I used the php port as an example is the amount of extensions it 
provides = the mount of packages it could generate when converted to 
subpackages, and because it fits the "use master/slave ports as a last 
resort" case (= use subpackages instead of master/slave = no slave port 
to depend upon = no slave port to depend upon). Without making it 
explicit in the docs that this is a bad idea and why, it would mean 
someone will provide a patch or introduce a smiliar cased port and the 
fear of suboptimal port builds will be realized.

To my understanding the current implemntation does *not* separate the 
concept of individual packages which can be build vs origin (except we 
keep slave ports for subpackages, which defeats the purpose to cut down 
on build time during package building). This understanding can off 
course be wrong, so any example which makes me understand that this is 
not the case is welcome.

Bye,
Alexander.

-- 
http://www.Leidinger.net Alexander@Leidinger.net: PGP 0x8F31830F9F2772BF
http://www.FreeBSD.org    netchild@FreeBSD.org  : PGP 0x8F31830F9F2772BF