Date: Mon, 31 Aug 1998 00:14:24 -0700 (PDT) From: asami@FreeBSD.ORG (Satoshi Asami) To: ac199@hwcn.org Cc: ports@FreeBSD.ORG, obrien@NUXI.com Subject: Re: cvs commit: ports/x11-wm/afterstep/pkg PLIST (more bsd.port.mk mods) Message-ID: <199808310714.AAA08628@silvia.hip.berkeley.edu> In-Reply-To: <19980831002849.A5522@zappo> (message from Tim Vanderhoek on Mon, 31 Aug 1998 00:28:49 -0400)
next in thread | previous in thread | raw e-mail | index | archive | help
* The bsd.man.mk doesn't need to handle MANLANG, MANxPREFIX, strange * manpages such as zzz.3qt, and it doesn't have to repeat the whole * mess for generate-plist. It doesn't allow absolute manpage paths, * and it doesn't have a '-' option. It doesn't do syntax checking * nearly as well. Ok. * > Ok, fixed. This is kinda yucky though. What's going to happen if the * > user uses "pkg_add -p"? * * Well, I have very little sympathy for a user who is trying to do this. * It's simply not supported for binary packages... It is. Otherwise, we might as well remove the -p option from pkg_add. Individual ports (I know, a lot of them) not supporting it is one thing, but the ports framework (i.e., bsd.port.mk and pkg_* tools) not supporting it is a whole different thing. * However, this new patch tries to avoid it where possible. Yeah, relative symlinks are cool. * I don't know. Since bsd.man.mk seems to go out of its way to allow * inter-section linking, I figured there must be some. <shrug> I guess. * Well, I was outvoted by a ratio of 2:1 on this (actually, it was also * a count of 2:1, but anyways... :). The arrows have been toasted * altogether. Cool. :) * Just for the record, ls(1) prints arrows from target -> source. * Doesn't that strike anyone else as being totally backwards? <sigh> But if "A is a symbolic link *to* B", would you write "A -> B" or "B -> A"? :> * - all (_all_) links are now relative symlinks. This means it is possible * to get something like "check.1 -> ../../../../../../ten/two/zed.1" It's ok. * - if we have "MLINKS=a.1 b.1", then we will "rm -f b.1 b.1.gz" first. No problem. * - a repeated target file may be specified by just a "-". Eg. * "MLINKS=sanity.1 insane.3 - moon.4 - wolf.5". This is * useful for ports like devel/ORBacus and print/mp. It looks nicer * in practice than in example. Could be. * Speed was a consideration, and to further increase the speed, it seems * necessary to begin optimizing make(1)... Cool. * If you want patches to test against current ports, I have them for ten * ports. I can't get them right now since they're on bento which isn't * allowing my ssh login at the moment. The ports are listed in * /a/tmp/hoek/plist/ports/mlink.ports (or was that "mlinks.ports"?), * though. I can extract patches once bento starts behaving again... No problem. I've just tested a couple that I converted myself (archivers/bzip2 and mail/elm), it seems to work fine. I'll commit it after a couple of days if it doesn't melt down. Satoshi To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-ports" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199808310714.AAA08628>