The manpage of Portupgrade is quite good, so you will probably be able to figure out how it works without this tutorial too.
The only difficulty I had with Portupgrade was that certain Ruby files (that Portupgrade needs) were moved into a newly created package, so that Portupgrade broke itself by upgrading Ruby 🙁
This is kinda off-topic, but a lot of trolls have been saying the FreeBSD approach to SMP has been a nightmare and it has turned the kernel into an unmantainable nightmare. I myself am not a C coder and am completely clueless in how clean some code is. Has anyone actually taken a look at the code and checked how clean the implementation is? Is it as bad as some people say it has become? The only thing that worries me is that it took around 5 years of development on the 5 branch to finally get a stable release. This could be true because of the above statement or it could be true because there were lots of things to change due to the long history of FreeBSD.
There are massive changes between the two. I would suggest begining to look to see if you can track down some stuff from around the time when 5.1, iirc it was the first since I don’t remember a 5.0, was just announce for a nice run down of how things have changed between 4x and 5x.
Many things besides threading has changed. New scheduler. New threading system. Massive rework in the locking. Far better ACPI support. A lot nicer init system. There are many other things that have been reworked too, but can’t think of many off hand.
The manpage of Portupgrade is quite good, so you will probably be able to figure out how it works without this tutorial too.
The only difficulty I had with Portupgrade was that certain Ruby files (that Portupgrade needs) were moved into a newly created package, so that Portupgrade broke itself by upgrading Ruby 🙁
The article is too long. May be it should be called as a practical tutorial for newbees
This is kinda off-topic, but a lot of trolls have been saying the FreeBSD approach to SMP has been a nightmare and it has turned the kernel into an unmantainable nightmare. I myself am not a C coder and am completely clueless in how clean some code is. Has anyone actually taken a look at the code and checked how clean the implementation is? Is it as bad as some people say it has become? The only thing that worries me is that it took around 5 years of development on the 5 branch to finally get a stable release. This could be true because of the above statement or it could be true because there were lots of things to change due to the long history of FreeBSD.
There are massive changes between the two. I would suggest begining to look to see if you can track down some stuff from around the time when 5.1, iirc it was the first since I don’t remember a 5.0, was just announce for a nice run down of how things have changed between 4x and 5x.
Many things besides threading has changed. New scheduler. New threading system. Massive rework in the locking. Far better ACPI support. A lot nicer init system. There are many other things that have been reworked too, but can’t think of many off hand.
# su
# cvsup -g -L 2 /path/to/ports-supfile && portsdb -Fu
# portversion -l “<”
(if required:)
# portupgrade -arR