From owner-freebsd-hackers Thu Jun 19 14:15:31 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id OAA19764 for hackers-outgoing; Thu, 19 Jun 1997 14:15:31 -0700 (PDT) Received: from sand.sentex.ca (sand.sentex.ca [206.222.77.6]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id OAA19756 for ; Thu, 19 Jun 1997 14:15:23 -0700 (PDT) Received: from gravel (gravel.sentex.ca [205.211.165.210]) by sand.sentex.ca (8.8.5/8.8.3) with SMTP id RAA04444; Thu, 19 Jun 1997 17:25:58 -0400 (EDT) Message-Id: <3.0.2.32.19970619171704.02693930@sentex.net> X-Sender: mdtancsa@sentex.net X-Mailer: QUALCOMM Windows Eudora Pro Version 3.0.2 (32) Date: Thu, 19 Jun 1997 17:17:04 -0400 To: "Jordan K. Hubbard" From: Mike Tancsa Subject: Re: make world error in RELENG_2_2 Cc: freebsd-hackers@FreeBSD.ORG In-Reply-To: <24525.866752787@time.cdrom.com> References: Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Sender: owner-hackers@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk At 01:39 PM 6/19/97 -0700, Jordan K. Hubbard wrote: >> If there is nothing wrong, why were so many people having problems all of a >> sudden? I would say the lack of the login.conf in /etc was a 'problem'... > >That was a problem. Our problem. However, in every single case that >I've seen the world target falling over so far, it has been: > > a) A 2.2.x tree splatted on top of a previous generation src > tree (and that's bad). > > b) A bogus cvsup file, missing some crucial section of the > source code (src-all is a good target and highly recommended). > > c) Stale depends or includes. You have sort of proved my point... As I said before, I try and follow the mailling lists as closely as possbile for any caveats / changes to the cvsup process... My sup file consists of the following *default tag=RELENG_2_2 *default host=cvsup.freebsd.org *default prefix=/usr *default base=/usr *default release=cvs delete use-rel-suffix src-all src-eBones src-secure Since there is precious little docs on the make world process, I have generally assumed that make world will do everything that is necessary to, well, make world (except for /etc, which is clear why you dont want to do that)... I had to learn the hard way about deleteing /usr/include, and the stale depends issue. Again, I assumed that make world would take care of these issues (and so did other users)... Perhaps a different target name, or a change in functionality of the target world would help here... My problems of a few Saturdays ago were due to c)... >Also note that RELENG_2_2 recently switched from SHARED=symlinks >to SHARED=copies by default. This is not a 'problem' to be fixed >so much as a change to simply be cognizant of, and if you've any >doubts then blow away /usr/include and start over with a make includes; >in all cases, the solution is fairly simple when you think about it. Its simple now :-) Look, I am really hesitant to be critical about the FreeBSD project because its such an amazing OS. (I am still amazed at what is produced strickly by volunteer efforts... ) Although I would guess the majority of people dont rebuild their OS from scratch, this area seems like it could use a little more clarification in either the docs, or in an FAQ. Better yet, as part of the CVSUP process, perhaps checkout a file that contains notes on any critical changes that would effect the build process... like blow away /usr/include and rebuild first etc... Even a simple section in the handbook after the CVSUP docs about the make world process would help alot... I would volunteer, but I certainly do not consider myself an expert... ---Mike ********************************************************************** Mike Tancsa (mike@sentex.net) * To do is to be -- Nietzsche Sentex Communications Corp, * To be is to do -- Sartre Cambridge, Ontario * Do be do be do -- Sinatra (http://www.sentex.net/~mdtancsa) *