From owner-freebsd-ports Tue Dec 16 23:42:06 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id XAA19707 for ports-outgoing; Tue, 16 Dec 1997 23:42:06 -0800 (PST) (envelope-from owner-freebsd-ports) Received: from home.gtcs.com (home.gtcs.com [206.54.69.238]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id XAA19680 for ; Tue, 16 Dec 1997 23:41:58 -0800 (PST) (envelope-from bruce@gtcs.com) Received: from gtcs.com (localhost.gtcs.com [127.0.0.1]) by home.gtcs.com (8.8.5/8.8.5) with ESMTP id AAA12328; Wed, 17 Dec 1997 00:41:25 -0700 (MST) Disposition-Notification-To: bgingery@gtcs.com X-Comment1: in most cases both the Return-Receipt and Delivery-Notification X-Comment2: requests are part of an ongoing poll to determine what clients X-Comment3: and MTAs respond to the headers. Message-Id: <199712170741.AAA12328@home.gtcs.com> Date: Wed, 17 Dec 1997 00:41:23 -0700 (MST) From: bgingery@gtcs.com Reply-To: bgingery@gtcs.com Subject: Re: Recommended added target for bsd.port.mk To: fenner@parc.xerox.com cc: freebsd-ports@freebsd.org In-Reply-To: <97Dec16.192102pst.177480@crevenia.parc.xerox.com> MIME-Version: 1.0 Content-Type: TEXT/plain; CHARSET=US-ASCII Sender: owner-freebsd-ports@freebsd.org X-Loop: FreeBSD.org Precedence: bulk On 16 Dec, Bill Fenner wrote: -} bgingery@gtcs.com wrote: -} >Whoops - time to upgrade my bsd.port.mk, I guess. -} -} Nope, I was making a suggestion, not offering something that already -} exists. Ahhh. Okay. The exact performance isn't particularly important, although no real reason to repeat the "describe" target, per se. Personally, I *very* seldom use "make describe" but *do* use my shell alias "desc" which is merely a "more pkg/DESCR", and the (currently also a shell alias) "installed" which does what I described, passing `make package-name` to pkg_info -c. I also use `make package-name` for easy package deletes, which could be a "make delete" target, very easily, too. I do that seldom enough that I've not even made an alias for it. OTOH, to be consistent with handling pre-req's, a "make delete" should be interactive only (under normal operation) and remove packages which are registered as dependent on the port, as well. Then, too, there are enough bsd.port.mk targets now, that a "make help" is almost in order, even if it merely extracts the lines that describe the targets from itself. I know a few people have expressed a wish to re-do the whole ports handling methods, yet it's quite powerful as is. Personally, I'd rather see a "make installbin" invoke a fetch of the binary package and install it, adding to the process, than a replacement of what's already working so well. Yet, also, with someone recently, I mentioned the prospects of storing the ports tree in a series of zip-files by type ..e.g. /usr/ports/archivers.zip (etc.) or even /usr/ports/archivers/zip.zip (etc.) and only extracting what's needed for making a port. That would require changes to bsd.port.subdir.mk, though, to handle the compressed port tree. "make all ", or "make " from the parent directory. The *reason* for using .zip rather than .tar.gz was twofold. New "port tarballs" could easily be distinguised from package tars by their "zip" extension. The space taken would decrease significantly. Individual *known* files could be extracted by the bsd.port.subdir.mk other targets, easily for mass processing. without the overhead of handling all of the patches or scripts that might be in an existing port. This would be especially handy for targets like "fetch-list" and "describe" and "readme". I'm already using a Perl script for resolving *some* URLs on my website for infrequently accessed files. It merely returns HTTP headers and unzips to stdout of the file requested (if approved). Bruce Gingery