5.2-BETA and related ports issues
eik at FreeBSD.org
Sat Nov 29 18:55:37 PST 2003
Mark Linimon wrote:
> On Sat, 29 Nov 2003, Dag-Erling Smørgrav wrote:
>>Do you actually review ports Makefiles?
> You _are_ kidding here, right?
> Yes, the ports team does read over the ports Makefile. Yes, the
> bento cluster attempts to find all the problems that can be found
> by automated processes. Yes, my own code attempts to lint
> inconsistencies in the ports collection. Yes, other automated
> scripts attempt to continually build /usr/ports/INDEX looking
> for inconsistencies, and maintainers whose email address bounces,
> and maintainers who might not have seen PRs about their ports
> (maintainers who are not committers, in this case). Yes,Bill
> Fenner's script attempts to find all unfetchable ports.
> I've seen other scripts that attempt to portlint the entire ports
> hierarchy but I don't know if they run periodically. They probably
> ought to.
> There are currently, as of this moment, 9722 ports in the tree.
> It's not humanly possible to read over every line of every single
> Makefile plus every single pkg-plist and grok them all. The ports
> folks rely on many eyes for help with this, just as with every
> other line of code in FreeBSD.
> There are certainly a lot of bugs in the ports tree. If people
> use send-pr for its intended purpose, specific bugs are more likely
> to get fixed than by casual discussions on mailing lists.
Please: I *know* that I violate the prefix. We don't need more QA people
here (we do, but thats not the point in this discussuion), we need more
people with which you can discuss open questions. Apperantly nobody
cared about PR conf/56736 nor PR bin/56748, and I didn't had the
capacity to push them. I guess I'll back out the installation into /etc,
but hey, how do I have to yell "feature needed"?
Please don't blame ports QA. Blame a misguided porter that didn't get
support from the base system maintainers...
More information about the freebsd-current