Originally Posted by daemonfowl
is this manageable for another machine connected to my server with no need to connect elsewhere ?
It is not uncommon for users to manage their own package repositories, but recognize that this duplicates a lot of work the project has already done.
or is it a necessary to fetch missing ports that can't be saved ?
I don't know what you are trying to say here.
is *every needed piece of software* saved in distfiles once compilation comes to an end ?
No, the contents of /usr/ports/distfiles
is as I described earlier. This is the directory where source is downloaded.
The actual building of ports is done in /usr/ports/pobj
. This is where you will find the intermediate object files + other ilk used to build the application's binaries, manpages, configuration files, etc
Once a package has been built, the contents of /usr/ports/pobj
can be deleted, but it is far simpler to just leave /usr/ports/pobj
if you want to later study the source code for any particular application.
I suspect one of your concerns is that building ports take a lot of diskspace. This is true. It is even more true if you build lots
of ports -- to the order of gigabytes of diskspace.