Re: Import of DragonFly Mail Agent

From: David Chisnall <theraven_at_FreeBSD.org>
Date: Tue, 25 Feb 2014 09:09:15 +0000
On 25 Feb 2014, at 08:09, Daniel Kalchev <daniel_at_digsys.bg> wrote:

> What we risk with "everything is a port" concept is that we live in a world that there is a lot of software to chose from, but from time to time, the software happens to be incompatible with FreeBSD in one way, or another. Another risk is the confusion of too much choice.

I think that, over the next few years, the hard line between base system and ports is going to become a little bit more of a gradient.  I would like us to end up with multiple tiers:

1) These packages are required for absolutely everything, don't even think about not installing them even in a minimal service jail.

2) These packages are required for a useable system.  They're in the default install, but if you're creating a jail you might not want them (e.g. nvi, some of the management tools) because you'll be doing all of your configuration with the version in the base system.

3) These packages are maintained by the FreeBSD project and are expected to integrate well with the base system.  Some of them are part of various recommended installs for different configurations (e.g. graphical workstation, web server, whatever), but you can have a working minimal install without any of them.  They will be supported for the duration of the release, including prompt security updates.  

4) These packages are third-party programs that have been tested with FreeBSD and packaged by members of the FreeBSD project, but are developed independently.  They will be supported on a best-effort basis for the release, but you may find that upgrading to a new version requires a newer release at some point.

5) These packages are provided by third parties, on third-party repositories, with no involvement from anyone in the FreeBSD project.  

Currently, the base system overlaps tiers 1-3, and ports overlaps tiers 3-4.  Tier 3 is the source of most bikesheds, because there are lots of things that would benefit from some FreeBSD-specific integration work, are essential to a large section of the FreeBSD userbase, but are completely irrelevant to another large section.  

David
Received on Tue Feb 25 2014 - 08:09:40 UTC

This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:40:47 UTC