We see that compiling from ports takes a very long time in older hardware (it makes sense). I have read that it is possible to compile a port on a faster machine using the command:
|
Code:
|
# portinstall -p name_of_port |
The -p switch creates packages (.tbz) for all the ports that the above command needed to compile. And as we know local packages are faster to install :
|
Code:
|
# pkg_add name_of_package.tbz |
What is the prerequisite so that somebody with two PCBSD installations can accomplish this little trick and upgrade the slower machine using these faster created packages? As far as I understand the process the critical points are
a) the compilers
b) the kernel sources (if compiling some driver)
c) the architecture (Intel or AMD, older or newer, x64 or not)
This is very interesting and I'm sure lots (including me) are willing to hear an answer from more experienced members.