pkgsrc-Bugs archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Re: pkg/35287: New pkgsrc entries for Perforce (p4d/p4/p4web)



The following reply was made to PR pkg/35287; it has been noted by GNATS.

From: Marc Tooley <netbsd-helper%quake.ca@localhost>
To: Amitai Schlair <schmonz%netbsd.org@localhost>
Cc: gnats-bugs%netbsd.org@localhost, Marc Tooley 
<netbsd-helper%quake.ca@localhost>
Subject: Re: pkg/35287: New pkgsrc entries for Perforce (p4d/p4/p4web)
Date: Thu, 18 Jan 2007 09:40:14 -0800

 Hello Amitai,
 
 Hey that's no problem whatsoever! I had a feeling that was a sticking 
 point; and, you're absolutely correct. It is a non-standard, hairy 
 extension in order to get rc.subr expanded to operate on arbitrarily 
 large numbers of daemons that are otherwise identical and need to be 
 handled identically.
 
 I actually wrote it for my job: it's necessary to have something like 
 this when you need every version of Perforce (ever) running 
 simultaneously.
 
 Please do feel free to apply whatever you wish: I'm just interested in 
 getting the new, actual binaries available to the public.
 
 I'll also post to tech-pkg@. I would love to know what the accepted form 
 of this kind of thing is.
 
 -Marc
 
 On Wednesday 17 January 2007 21:15, Amitai Schlair wrote:
 > I'm looking at your changes now, and the rc.d script is
 > intimidatingly detailed to read through and feel confident in (not
 > least because sometimes it says "/usr/pkg/sbin" and sometimes it says
 > "/usr/local/sbin"... :-). My feelings are mixed: on the one hand, I
 > like the idea of managing multiple p4d instances with a single script
 > -- and something similar could help me solve PR pkg/30957 -- but on
 > the other hand, the rc.d script winds up being distressingly complex.
 >
 > Part of the problem, as you've encountered firsthand, is that rc.subr
 > doesn't provide free goodies for managing multiple instances with one
 > script. Maybe it should. Until then, as much as I like the idea in
 > the abstract, this approach feels wrong to me. And I say this as a
 > guy who's written his fair share of hairy, non-traditional rc.d
 > scripts (check out mail/qmail-run sometime).
 >
 > Your effort here was obviously significant and I'd like to see it
 > come to fruition. I propose that you bring up the issues you faced in
 > writing this rc.d script in a post on tech-pkg@, and I'll pitch in
 > where I can, and if and when a consensus is reached, I'll be glad to
 > act on it. Meanwhile, for purposes of this PR, how about I apply the
 > rest of your update to 2006.1?
 



Home | Main Index | Thread Index | Old Index