From owner-cvs-sbin Wed Sep 27 10:15:45 1995 Return-Path: owner-cvs-sbin Received: (from root@localhost) by freefall.freebsd.org (8.6.12/8.6.6) id KAA07020 for cvs-sbin-outgoing; Wed, 27 Sep 1995 10:15:45 -0700 Received: (from ache@localhost) by freefall.freebsd.org (8.6.12/8.6.6) id KAA07011 ; Wed, 27 Sep 1995 10:15:39 -0700 Date: Wed, 27 Sep 1995 10:15:39 -0700 From: "Andrey A. Chernov" Message-Id: <199509271715.KAA07011@freefall.freebsd.org> To: CVS-commiters, cvs-sbin Subject: cvs commit: src/sbin/startslip startslip.c Sender: owner-cvs-sbin@FreeBSD.org Precedence: bulk ache 95/09/27 10:15:38 Modified: sbin/startslip startslip.c Log: Log user name with each info and total connection time elapsed (for accounting purposes). Log few additional events like dialing number and connection established. Set speed POSIXly. From owner-cvs-sbin Fri Sep 29 12:35:14 1995 Return-Path: owner-cvs-sbin Received: (from root@localhost) by freefall.freebsd.org (8.6.12/8.6.6) id MAA28407 for cvs-sbin-outgoing; Fri, 29 Sep 1995 12:35:14 -0700 Received: (from ache@localhost) by freefall.freebsd.org (8.6.12/8.6.6) id MAA28395 ; Fri, 29 Sep 1995 12:35:11 -0700 Date: Fri, 29 Sep 1995 12:35:11 -0700 From: "Andrey A. Chernov" Message-Id: <199509291935.MAA28395@freefall.freebsd.org> To: CVS-commiters, cvs-sbin Subject: cvs commit: src/sbin Makefile Sender: owner-cvs-sbin@FreeBSD.org Precedence: bulk ache 95/09/29 12:35:11 Modified: sbin Makefile Log: Build secure init if available and allowed From owner-cvs-sbin Fri Sep 29 12:38:42 1995 Return-Path: owner-cvs-sbin Received: (from root@localhost) by freefall.freebsd.org (8.6.12/8.6.6) id MAA28491 for cvs-sbin-outgoing; Fri, 29 Sep 1995 12:38:42 -0700 Received: from aslan.cdrom.com (aslan.cdrom.com [192.216.223.142]) by freefall.freebsd.org (8.6.12/8.6.6) with ESMTP id MAA28486 ; Fri, 29 Sep 1995 12:38:39 -0700 Received: from localhost.cdrom.com (localhost.cdrom.com [127.0.0.1]) by aslan.cdrom.com (8.6.12/8.6.9) with SMTP id MAA24705; Fri, 29 Sep 1995 12:38:22 -0700 Message-Id: <199509291938.MAA24705@aslan.cdrom.com> X-Authentication-Warning: aslan.cdrom.com: Host localhost.cdrom.com didn't use HELO protocol To: "Andrey A. Chernov" cc: CVS-commiters@freefall.freebsd.org, cvs-sbin@freefall.freebsd.org Subject: Re: cvs commit: src/sbin Makefile In-reply-to: Your message of "Fri, 29 Sep 1995 12:35:11 PDT." <199509291935.MAA28395@freefall.freebsd.org> Date: Fri, 29 Sep 1995 12:38:22 -0700 From: "Justin T. Gibbs" Sender: owner-cvs-sbin@FreeBSD.org Precedence: bulk >ache 95/09/29 12:35:11 > > Modified: sbin Makefile > Log: > Build secure init if available and allowed On a similar subject, shouldn't "MAKE_EBONES" go away? Shouldn't the build be based solely on the existance of the directory? -- Justin T. Gibbs =========================================== Software Developer - Walnut Creek CDROM FreeBSD: Turning PCs into workstations =========================================== From owner-cvs-sbin Fri Sep 29 12:45:44 1995 Return-Path: owner-cvs-sbin Received: (from root@localhost) by freefall.freebsd.org (8.6.12/8.6.6) id MAA28767 for cvs-sbin-outgoing; Fri, 29 Sep 1995 12:45:44 -0700 Received: from halloran-eldar.lcs.mit.edu (halloran-eldar.lcs.mit.edu [18.26.0.159]) by freefall.freebsd.org (8.6.12/8.6.6) with SMTP id MAA28759 ; Fri, 29 Sep 1995 12:45:40 -0700 Received: by halloran-eldar.lcs.mit.edu; (5.65/1.1.8.2/19Aug95-0530PM) id AA19773; Fri, 29 Sep 1995 15:44:59 -0400 Date: Fri, 29 Sep 1995 15:44:59 -0400 From: "Garrett A. Wollman" Message-Id: <9509291944.AA19773@halloran-eldar.lcs.mit.edu> To: "Justin T. Gibbs" Cc: "Andrey A. Chernov" , CVS-commiters@freefall.freebsd.org, cvs-sbin@freefall.freebsd.org Subject: Re: cvs commit: src/sbin Makefile In-Reply-To: <199509291938.MAA24705@aslan.cdrom.com> References: <199509291935.MAA28395@freefall.freebsd.org> <199509291938.MAA24705@aslan.cdrom.com> Sender: owner-cvs-sbin@FreeBSD.org Precedence: bulk < said: > On a similar subject, shouldn't "MAKE_EBONES" go away? Shouldn't the > build be based solely on the existance of the directory? No, certainly not! How would you make no-crypto distributions, then? Or are you simply suggesting that everything should be controlled by a single option? -GAWollman -- Garrett A. Wollman | Shashish is simple, it's discreet, it's brief. ... wollman@lcs.mit.edu | Shashish is the bonding of hearts in spite of distance. Opinions not those of| It is a bond more powerful than absence. We like people MIT, LCS, ANA, or NSA| who like Shashish. - Claude McKenzie + Florent Vollant From owner-cvs-sbin Fri Sep 29 12:46:27 1995 Return-Path: owner-cvs-sbin Received: (from root@localhost) by freefall.freebsd.org (8.6.12/8.6.6) id MAA28826 for cvs-sbin-outgoing; Fri, 29 Sep 1995 12:46:27 -0700 Received: from irbs.irbs.com (irbs.com [199.182.75.129]) by freefall.freebsd.org (8.6.12/8.6.6) with ESMTP id MAA28818 ; Fri, 29 Sep 1995 12:46:20 -0700 Received: (from jc@localhost) by irbs.irbs.com (8.6.12/8.6.6) id PAA27305; Fri, 29 Sep 1995 15:45:45 -0400 From: John Capo Message-Id: <199509291945.PAA27305@irbs.irbs.com> Subject: Re: cvs commit: src/sbin Makefile To: gibbs@freefall.freebsd.org (Justin T. Gibbs) Date: Fri, 29 Sep 1995 15:45:45 -0400 (EDT) Cc: ache@freefall.freebsd.org, CVS-commiters@freefall.freebsd.org, cvs-sbin@freefall.freebsd.org In-Reply-To: <199509291938.MAA24705@aslan.cdrom.com> from "Justin T. Gibbs" at Sep 29, 95 12:38:22 pm X-Mailer: ELM [version 2.4 PL24] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 491 Sender: owner-cvs-sbin@FreeBSD.org Precedence: bulk Justin T. Gibbs writes: > > >ache 95/09/29 12:35:11 > > > > Modified: sbin Makefile > > Log: > > Build secure init if available and allowed > > On a similar subject, shouldn't "MAKE_EBONES" go away? Shouldn't the > build be based solely on the existance of the directory? That's the way it used to work. Some of us build different configurations from the same source tree. It was a pain to move the directory out of the way and then back. John Capo IRBS Engineering From owner-cvs-sbin Fri Sep 29 12:51:29 1995 Return-Path: owner-cvs-sbin Received: (from root@localhost) by freefall.freebsd.org (8.6.12/8.6.6) id MAA29343 for cvs-sbin-outgoing; Fri, 29 Sep 1995 12:51:29 -0700 Received: from aslan.cdrom.com (aslan.cdrom.com [192.216.223.142]) by freefall.freebsd.org (8.6.12/8.6.6) with ESMTP id MAA29337 ; Fri, 29 Sep 1995 12:51:26 -0700 Received: from localhost.cdrom.com (localhost.cdrom.com [127.0.0.1]) by aslan.cdrom.com (8.6.12/8.6.9) with SMTP id MAA24789; Fri, 29 Sep 1995 12:50:59 -0700 Message-Id: <199509291950.MAA24789@aslan.cdrom.com> X-Authentication-Warning: aslan.cdrom.com: Host localhost.cdrom.com didn't use HELO protocol To: "Garrett A. Wollman" cc: "Justin T. Gibbs" , "Andrey A. Chernov" , CVS-commiters@freefall.freebsd.org, cvs-sbin@freefall.freebsd.org Subject: Re: cvs commit: src/sbin Makefile In-reply-to: Your message of "Fri, 29 Sep 1995 15:44:59 EDT." <9509291944.AA19773@halloran-eldar.lcs.mit.edu> Date: Fri, 29 Sep 1995 12:50:58 -0700 From: "Justin T. Gibbs" Sender: owner-cvs-sbin@FreeBSD.org Precedence: bulk ><D.org> said: > >> On a similar subject, shouldn't "MAKE_EBONES" go away? Shouldn't the >> build be based solely on the existance of the directory? > >No, certainly not! How would you make no-crypto distributions, then? >Or are you simply suggesting that everything should be controlled by a >single option? > >-GAWollman Non crypto distributions are already handled by the make release target. >-- >Garrett A. Wollman | Shashish is simple, it's discreet, it's brief. ... >wollman@lcs.mit.edu | Shashish is the bonding of hearts in spite of distance. >Opinions not those of| It is a bond more powerful than absence. We like peopl >e >MIT, LCS, ANA, or NSA| who like Shashish. - Claude McKenzie + Florent Vollant -- Justin T. Gibbs =========================================== Software Developer - Walnut Creek CDROM FreeBSD: Turning PCs into workstations =========================================== From owner-cvs-sbin Fri Sep 29 12:56:37 1995 Return-Path: owner-cvs-sbin Received: (from root@localhost) by freefall.freebsd.org (8.6.12/8.6.6) id MAA29697 for cvs-sbin-outgoing; Fri, 29 Sep 1995 12:56:37 -0700 Received: from aslan.cdrom.com (aslan.cdrom.com [192.216.223.142]) by freefall.freebsd.org (8.6.12/8.6.6) with ESMTP id MAA29676 ; Fri, 29 Sep 1995 12:56:33 -0700 Received: from localhost.cdrom.com (localhost.cdrom.com [127.0.0.1]) by aslan.cdrom.com (8.6.12/8.6.9) with SMTP id MAA24827; Fri, 29 Sep 1995 12:55:51 -0700 Message-Id: <199509291955.MAA24827@aslan.cdrom.com> X-Authentication-Warning: aslan.cdrom.com: Host localhost.cdrom.com didn't use HELO protocol To: John Capo cc: gibbs@freefall.freebsd.org (Justin T. Gibbs), ache@freefall.freebsd.org, CVS-commiters@freefall.freebsd.org, cvs-sbin@freefall.freebsd.org Subject: Re: cvs commit: src/sbin Makefile In-reply-to: Your message of "Fri, 29 Sep 1995 15:45:45 EDT." <199509291945.PAA27305@irbs.irbs.com> Date: Fri, 29 Sep 1995 12:55:51 -0700 From: "Justin T. Gibbs" Sender: owner-cvs-sbin@FreeBSD.org Precedence: bulk >Justin T. Gibbs writes: >> >> >ache 95/09/29 12:35:11 >> > >> > Modified: sbin Makefile >> > Log: >> > Build secure init if available and allowed >> >> On a similar subject, shouldn't "MAKE_EBONES" go away? Shouldn't the >> build be based solely on the existance of the directory? > >That's the way it used to work. Some of us build different >configurations from the same source tree. It was a pain to move >the directory out of the way and then back. What's the use of most of what is in secure without a built eBones? In fact our secure telnet did absolutely nothing for us until it was recently fixed to work with eBones. Yet, many makefiles use the "MAKE_EBONES" variable to make decisions on how to build things. If anything, there sould be a single variable referenced in /usr/src/Makefile for non exportable material that simply exludes the right subdirectories. > >John Capo >IRBS Engineering -- Justin T. Gibbs =========================================== Software Developer - Walnut Creek CDROM FreeBSD: Turning PCs into workstations =========================================== From owner-cvs-sbin Fri Sep 29 13:08:55 1995 Return-Path: owner-cvs-sbin Received: (from root@localhost) by freefall.freebsd.org (8.6.12/8.6.6) id NAA00136 for cvs-sbin-outgoing; Fri, 29 Sep 1995 13:08:55 -0700 Received: from grunt.grondar.za (grunt.grondar.za [196.7.18.129]) by freefall.freebsd.org (8.6.12/8.6.6) with ESMTP id NAA00126 ; Fri, 29 Sep 1995 13:08:19 -0700 Received: from grumble.grondar.za (grumble.grondar.za [196.7.18.130]) by grunt.grondar.za (8.6.12/8.6.9) with ESMTP id WAA15213; Fri, 29 Sep 1995 22:07:57 +0200 Received: from localhost (localhost [127.0.0.1]) by grumble.grondar.za (8.6.12/8.6.9) with SMTP id WAA22398; Fri, 29 Sep 1995 22:07:56 +0200 Message-Id: <199509292007.WAA22398@grumble.grondar.za> X-Authentication-Warning: grumble.grondar.za: Host localhost didn't use HELO protocol To: "Justin T. Gibbs" cc: "Andrey A. Chernov" , CVS-commiters@freefall.freebsd.org, cvs-sbin@freefall.freebsd.org Subject: Re: cvs commit: src/sbin Makefile Date: Fri, 29 Sep 1995 22:07:55 +0200 From: Mark Murray Sender: owner-cvs-sbin@FreeBSD.org Precedence: bulk > >ache 95/09/29 12:35:11 > > > > Modified: sbin Makefile > > Log: > > Build secure init if available and allowed > > On a similar subject, shouldn't "MAKE_EBONES" go away? Shouldn't the > build be based solely on the existance of the directory? Yehbo! Please do this - get the crap out of src/Makefile! M -- Mark Murray 46 Harvey Rd, Claremont, Cape Town 7700, South Africa +27 21 61-3768 GMT+0200 Finger mark@grumble.grondar.za for PGP key From owner-cvs-sbin Fri Sep 29 13:12:35 1995 Return-Path: owner-cvs-sbin Received: (from root@localhost) by freefall.freebsd.org (8.6.12/8.6.6) id NAA00412 for cvs-sbin-outgoing; Fri, 29 Sep 1995 13:12:35 -0700 Received: from grunt.grondar.za (grunt.grondar.za [196.7.18.129]) by freefall.freebsd.org (8.6.12/8.6.6) with ESMTP id NAA00392 ; Fri, 29 Sep 1995 13:12:08 -0700 Received: from grumble.grondar.za (grumble.grondar.za [196.7.18.130]) by grunt.grondar.za (8.6.12/8.6.9) with ESMTP id WAA15225; Fri, 29 Sep 1995 22:11:48 +0200 Received: from localhost (localhost [127.0.0.1]) by grumble.grondar.za (8.6.12/8.6.9) with SMTP id WAA22433; Fri, 29 Sep 1995 22:11:47 +0200 Message-Id: <199509292011.WAA22433@grumble.grondar.za> X-Authentication-Warning: grumble.grondar.za: Host localhost didn't use HELO protocol To: "Garrett A. Wollman" cc: "Justin T. Gibbs" , "Andrey A. Chernov" , CVS-commiters@freefall.freebsd.org, cvs-sbin@freefall.freebsd.org Subject: Re: cvs commit: src/sbin Makefile Date: Fri, 29 Sep 1995 22:11:47 +0200 From: Mark Murray Sender: owner-cvs-sbin@FreeBSD.org Precedence: bulk > < said: > > > On a similar subject, shouldn't "MAKE_EBONES" go away? Shouldn't the > > build be based solely on the existance of the directory? > > No, certainly not! How would you make no-crypto distributions, then? > Or are you simply suggesting that everything should be controlled by a > single option? Right now we have two wrongs making a right. We _have_ to have a make with no MAKE_EBONES to make the clean distribution, followed by a nasty hack to make the tainted distribution. Why can we not have "make distribution" do it properly the first time? Sure we will need to fix the make files, but that can of worms has been brewing for a while... M -- Mark Murray 46 Harvey Rd, Claremont, Cape Town 7700, South Africa +27 21 61-3768 GMT+0200 Finger mark@grumble.grondar.za for PGP key From owner-cvs-sbin Fri Sep 29 13:19:11 1995 Return-Path: owner-cvs-sbin Received: (from root@localhost) by freefall.freebsd.org (8.6.12/8.6.6) id NAA00728 for cvs-sbin-outgoing; Fri, 29 Sep 1995 13:19:11 -0700 Received: from aslan.cdrom.com (aslan.cdrom.com [192.216.223.142]) by freefall.freebsd.org (8.6.12/8.6.6) with ESMTP id NAA00708 ; Fri, 29 Sep 1995 13:18:57 -0700 Received: from localhost.cdrom.com (localhost.cdrom.com [127.0.0.1]) by aslan.cdrom.com (8.6.12/8.6.9) with SMTP id NAA25022; Fri, 29 Sep 1995 13:17:42 -0700 Message-Id: <199509292017.NAA25022@aslan.cdrom.com> X-Authentication-Warning: aslan.cdrom.com: Host localhost.cdrom.com didn't use HELO protocol To: Mark Murray cc: "Garrett A. Wollman" , "Justin T. Gibbs" , "Andrey A. Chernov" , CVS-commiters@freefall.freebsd.org, cvs-sbin@freefall.freebsd.org Subject: Re: cvs commit: src/sbin Makefile In-reply-to: Your message of "Fri, 29 Sep 1995 22:11:47 +0200." <199509292011.WAA22433@grumble.grondar.za> Date: Fri, 29 Sep 1995 13:17:42 -0700 From: "Justin T. Gibbs" Sender: owner-cvs-sbin@FreeBSD.org Precedence: bulk >> <B >SD.org> said: >> >> > On a similar subject, shouldn't "MAKE_EBONES" go away? Shouldn't the >> > build be based solely on the existance of the directory? >> >> No, certainly not! How would you make no-crypto distributions, then? >> Or are you simply suggesting that everything should be controlled by a >> single option? > >Right now we have two wrongs making a right. We _have_ to have a make >with no MAKE_EBONES to make the clean distribution, followed by a nasty >hack to make the tainted distribution. Why can we not have "make >distribution" do it properly the first time? Sure we will need to fix >the make files, but that can of worms has been brewing for a while... > >M Actually, you have to have "MAKE_EBONES" in you're make.conf in order to do a full make release which makes both the secure and non-secure distributions. Its all screwed up if you ask me. >-- >Mark Murray >46 Harvey Rd, Claremont, Cape Town 7700, South Africa >+27 21 61-3768 GMT+0200 >Finger mark@grumble.grondar.za for PGP key -- Justin T. Gibbs =========================================== Software Developer - Walnut Creek CDROM FreeBSD: Turning PCs into workstations =========================================== From owner-cvs-sbin Fri Sep 29 13:19:14 1995 Return-Path: owner-cvs-sbin Received: (from root@localhost) by freefall.freebsd.org (8.6.12/8.6.6) id NAA00734 for cvs-sbin-outgoing; Fri, 29 Sep 1995 13:19:14 -0700 Received: from halloran-eldar.lcs.mit.edu (halloran-eldar.lcs.mit.edu [18.26.0.159]) by freefall.freebsd.org (8.6.12/8.6.6) with SMTP id NAA00714 ; Fri, 29 Sep 1995 13:19:04 -0700 Received: by halloran-eldar.lcs.mit.edu; (5.65/1.1.8.2/19Aug95-0530PM) id AA19932; Fri, 29 Sep 1995 16:18:30 -0400 Date: Fri, 29 Sep 1995 16:18:30 -0400 From: "Garrett A. Wollman" Message-Id: <9509292018.AA19932@halloran-eldar.lcs.mit.edu> To: Mark Murray Cc: "Garrett A. Wollman" , "Justin T. Gibbs" , "Andrey A. Chernov" , CVS-commiters@freefall.freebsd.org, cvs-sbin@freefall.freebsd.org Subject: Re: cvs commit: src/sbin Makefile In-Reply-To: <199509292011.WAA22433@grumble.grondar.za> References: <199509292011.WAA22433@grumble.grondar.za> Sender: owner-cvs-sbin@FreeBSD.org Precedence: bulk < said: > Right now we have two wrongs making a right. We _have_ to have a make > with no MAKE_EBONES to make the clean distribution, followed by a nasty > hack to make the tainted distribution. Why can we not have "make > distribution" do it properly the first time? Sure we will need to fix > the make files, but that can of worms has been brewing for a while... Not everyone's build environment is based on `make release' and `make distribute'. -GAWollman -- Garrett A. Wollman | Shashish is simple, it's discreet, it's brief. ... wollman@lcs.mit.edu | Shashish is the bonding of hearts in spite of distance. Opinions not those of| It is a bond more powerful than absence. We like people MIT, LCS, ANA, or NSA| who like Shashish. - Claude McKenzie + Florent Vollant From owner-cvs-sbin Fri Sep 29 13:22:16 1995 Return-Path: owner-cvs-sbin Received: (from root@localhost) by freefall.freebsd.org (8.6.12/8.6.6) id NAA00874 for cvs-sbin-outgoing; Fri, 29 Sep 1995 13:22:16 -0700 Received: from grunt.grondar.za (grunt.grondar.za [196.7.18.129]) by freefall.freebsd.org (8.6.12/8.6.6) with ESMTP id NAA00844 ; Fri, 29 Sep 1995 13:21:43 -0700 Received: from grumble.grondar.za (grumble.grondar.za [196.7.18.130]) by grunt.grondar.za (8.6.12/8.6.9) with ESMTP id WAA15252; Fri, 29 Sep 1995 22:21:01 +0200 Received: from localhost (localhost [127.0.0.1]) by grumble.grondar.za (8.6.12/8.6.9) with SMTP id WAA22557; Fri, 29 Sep 1995 22:21:00 +0200 Message-Id: <199509292021.WAA22557@grumble.grondar.za> X-Authentication-Warning: grumble.grondar.za: Host localhost didn't use HELO protocol To: "Justin T. Gibbs" cc: John Capo , ache@freefall.freebsd.org, CVS-commiters@freefall.freebsd.org, cvs-sbin@freefall.freebsd.org Subject: Re: cvs commit: src/sbin Makefile Date: Fri, 29 Sep 1995 22:21:00 +0200 From: Mark Murray Sender: owner-cvs-sbin@FreeBSD.org Precedence: bulk > What's the use of most of what is in secure without a built eBones? > In fact our secure telnet did absolutely nothing for us until it > was recently fixed to work with eBones. Yet, many makefiles use > the "MAKE_EBONES" variable to make decisions on how to build things. > If anything, there sould be a single variable referenced in > /usr/src/Makefile for non exportable material that simply exludes the > right subdirectories. Hear! Hear! Much fixing needs to be done in the makefiles - Which it seems Andrey is starting to do... Under the src/release area, "make distribute" is much neglected. M -- Mark Murray 46 Harvey Rd, Claremont, Cape Town 7700, South Africa +27 21 61-3768 GMT+0200 Finger mark@grumble.grondar.za for PGP key From owner-cvs-sbin Fri Sep 29 13:22:51 1995 Return-Path: owner-cvs-sbin Received: (from root@localhost) by freefall.freebsd.org (8.6.12/8.6.6) id NAA00931 for cvs-sbin-outgoing; Fri, 29 Sep 1995 13:22:51 -0700 Received: from aslan.cdrom.com (aslan.cdrom.com [192.216.223.142]) by freefall.freebsd.org (8.6.12/8.6.6) with ESMTP id NAA00922 ; Fri, 29 Sep 1995 13:22:46 -0700 Received: from localhost.cdrom.com (localhost.cdrom.com [127.0.0.1]) by aslan.cdrom.com (8.6.12/8.6.9) with SMTP id NAA25052; Fri, 29 Sep 1995 13:21:50 -0700 Message-Id: <199509292021.NAA25052@aslan.cdrom.com> X-Authentication-Warning: aslan.cdrom.com: Host localhost.cdrom.com didn't use HELO protocol To: "Garrett A. Wollman" cc: Mark Murray , "Justin T. Gibbs" , "Andrey A. Chernov" , CVS-commiters@freefall.freebsd.org, cvs-sbin@freefall.freebsd.org Subject: Re: cvs commit: src/sbin Makefile In-reply-to: Your message of "Fri, 29 Sep 1995 16:18:30 EDT." <9509292018.AA19932@halloran-eldar.lcs.mit.edu> Date: Fri, 29 Sep 1995 13:21:50 -0700 From: "Justin T. Gibbs" Sender: owner-cvs-sbin@FreeBSD.org Precedence: bulk >< said: > >> Right now we have two wrongs making a right. We _have_ to have a make >> with no MAKE_EBONES to make the clean distribution, followed by a nasty >> hack to make the tainted distribution. Why can we not have "make >> distribution" do it properly the first time? Sure we will need to fix >> the make files, but that can of worms has been brewing for a while... > >Not everyone's build environment is based on `make release' and `make >distribute'. > >-GAWollman Then you must be doing something local to the makefiles anyway, so you can re-add the feature or fix the release target to do what you need it to. I'd rather see your work brought back into the release makefile since the "release" target does much more than many people need. It would be nice to give other options. >-- >Garrett A. Wollman | Shashish is simple, it's discreet, it's brief. ... >wollman@lcs.mit.edu | Shashish is the bonding of hearts in spite of distance. >Opinions not those of| It is a bond more powerful than absence. We like peopl >e >MIT, LCS, ANA, or NSA| who like Shashish. - Claude McKenzie + Florent Vollant -- Justin T. Gibbs =========================================== Software Developer - Walnut Creek CDROM FreeBSD: Turning PCs into workstations =========================================== From owner-cvs-sbin Fri Sep 29 13:25:48 1995 Return-Path: owner-cvs-sbin Received: (from root@localhost) by freefall.freebsd.org (8.6.12/8.6.6) id NAA01121 for cvs-sbin-outgoing; Fri, 29 Sep 1995 13:25:48 -0700 Received: from irbs.irbs.com (irbs.com [199.182.75.129]) by freefall.freebsd.org (8.6.12/8.6.6) with ESMTP id NAA01104 ; Fri, 29 Sep 1995 13:25:31 -0700 Received: (from jc@localhost) by irbs.irbs.com (8.6.12/8.6.6) id QAA27867; Fri, 29 Sep 1995 16:25:28 -0400 From: John Capo Message-Id: <199509292025.QAA27867@irbs.irbs.com> Subject: Re: cvs commit: src/sbin Makefile To: gibbs@freefall.FreeBSD.org (Justin T. Gibbs) Date: Fri, 29 Sep 1995 16:25:27 -0400 (EDT) Cc: jc@irbs.com, gibbs@freefall.FreeBSD.org, ache@freefall.FreeBSD.org, CVS-commiters@freefall.FreeBSD.org, cvs-sbin@freefall.FreeBSD.org In-Reply-To: <199509291955.MAA24827@aslan.cdrom.com> from "Justin T. Gibbs" at Sep 29, 95 12:55:51 pm X-Mailer: ELM [version 2.4 PL24] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 1342 Sender: owner-cvs-sbin@FreeBSD.org Precedence: bulk Justin T. Gibbs writes: > > >Justin T. Gibbs writes: > >> > >> >ache 95/09/29 12:35:11 > >> > > >> > Modified: sbin Makefile > >> > Log: > >> > Build secure init if available and allowed > >> > >> On a similar subject, shouldn't "MAKE_EBONES" go away? Shouldn't the > >> build be based solely on the existance of the directory? > > > >That's the way it used to work. Some of us build different > >configurations from the same source tree. It was a pain to move > >the directory out of the way and then back. > > What's the use of most of what is in secure without a built eBones? > In fact our secure telnet did absolutely nothing for us until it > was recently fixed to work with eBones. Yet, many makefiles use > the "MAKE_EBONES" variable to make decisions on how to build things. > If anything, there sould be a single variable referenced in > /usr/src/Makefile for non exportable material that simply exludes the > right subdirectories. > MAKE_EBONES or somesuch should control builds with eBones. Builds with DES should be controlled by something else. Being able to build with eBones but turn it off at run time would be nice. A system built with eBones expects it to be used. passwd defaults to changing the kerberos password for instance. I think we need some more knobs. John Capo IRBS Engineering From owner-cvs-sbin Fri Sep 29 13:32:34 1995 Return-Path: owner-cvs-sbin Received: (from root@localhost) by freefall.freebsd.org (8.6.12/8.6.6) id NAA01381 for cvs-sbin-outgoing; Fri, 29 Sep 1995 13:32:34 -0700 Received: from GndRsh.aac.dev.com (GndRsh.aac.dev.com [198.145.92.241]) by freefall.freebsd.org (8.6.12/8.6.6) with ESMTP id NAA01369 ; Fri, 29 Sep 1995 13:32:18 -0700 Received: (from rgrimes@localhost) by GndRsh.aac.dev.com (8.6.12/8.6.12) id NAA00663; Fri, 29 Sep 1995 13:31:54 -0700 From: "Rodney W. Grimes" Message-Id: <199509292031.NAA00663@GndRsh.aac.dev.com> Subject: Re: cvs commit: src/sbin Makefile To: mark@grondar.za (Mark Murray) Date: Fri, 29 Sep 1995 13:31:54 -0700 (PDT) Cc: gibbs@freefall.freebsd.org, ache@freefall.freebsd.org, CVS-commiters@freefall.freebsd.org, cvs-sbin@freefall.freebsd.org In-Reply-To: <199509292007.WAA22398@grumble.grondar.za> from "Mark Murray" at Sep 29, 95 10:07:55 pm X-Mailer: ELM [version 2.4 PL24] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 721 Sender: owner-cvs-sbin@FreeBSD.org Precedence: bulk > > > >ache 95/09/29 12:35:11 > > > > > > Modified: sbin Makefile > > > Log: > > > Build secure init if available and allowed > > > > On a similar subject, shouldn't "MAKE_EBONES" go away? Shouldn't the > > build be based solely on the existance of the directory? > > Yehbo! Please do this - get the crap out of src/Makefile! Yuko... crap out of src/Makefile and into src/sbin/Makefile == CRAP in top level make files. This only, IMHO, creates more crap. If this ``crap'' goes anyplace for the above case it goes in src/sbin/init/Makefile. -- Rod Grimes rgrimes@gndrsh.aac.dev.com Accurate Automation Company Reliable computers for FreeBSD From owner-cvs-sbin Fri Sep 29 13:35:10 1995 Return-Path: owner-cvs-sbin Received: (from root@localhost) by freefall.freebsd.org (8.6.12/8.6.6) id NAA01594 for cvs-sbin-outgoing; Fri, 29 Sep 1995 13:35:10 -0700 Received: from grunt.grondar.za (grunt.grondar.za [196.7.18.129]) by freefall.freebsd.org (8.6.12/8.6.6) with ESMTP id NAA01563 ; Fri, 29 Sep 1995 13:34:48 -0700 Received: from grumble.grondar.za (grumble.grondar.za [196.7.18.130]) by grunt.grondar.za (8.6.12/8.6.9) with ESMTP id WAA15282; Fri, 29 Sep 1995 22:34:08 +0200 Received: from localhost (localhost [127.0.0.1]) by grumble.grondar.za (8.6.12/8.6.9) with SMTP id WAA22635; Fri, 29 Sep 1995 22:34:07 +0200 Message-Id: <199509292034.WAA22635@grumble.grondar.za> X-Authentication-Warning: grumble.grondar.za: Host localhost didn't use HELO protocol To: "Garrett A. Wollman" cc: Mark Murray , "Justin T. Gibbs" , "Andrey A. Chernov" , CVS-commiters@freefall.freebsd.org, cvs-sbin@freefall.freebsd.org Subject: Re: cvs commit: src/sbin Makefile Date: Fri, 29 Sep 1995 22:34:07 +0200 From: Mark Murray Sender: owner-cvs-sbin@FreeBSD.org Precedence: bulk > < said: > > > Right now we have two wrongs making a right. We _have_ to have a make > > with no MAKE_EBONES to make the clean distribution, followed by a nasty > > hack to make the tainted distribution. Why can we not have "make > > distribution" do it properly the first time? Sure we will need to fix > > the make files, but that can of worms has been brewing for a while... > > Not everyone's build environment is based on `make release' and `make > distribute'. Granted. Then the logic should be "if the eBones directory exists, build eBones _unless_ asked not to", rather than "the directory must be present _and_ permission must be granted to build eBones". This would mean that secure and eBones can be added to the system just by dropping them in. M -- Mark Murray 46 Harvey Rd, Claremont, Cape Town 7700, South Africa +27 21 61-3768 GMT+0200 Finger mark@grumble.grondar.za for PGP key From owner-cvs-sbin Fri Sep 29 13:39:29 1995 Return-Path: owner-cvs-sbin Received: (from root@localhost) by freefall.freebsd.org (8.6.12/8.6.6) id NAA01746 for cvs-sbin-outgoing; Fri, 29 Sep 1995 13:39:29 -0700 Received: from aslan.cdrom.com (aslan.cdrom.com [192.216.223.142]) by freefall.freebsd.org (8.6.12/8.6.6) with ESMTP id NAA01738 ; Fri, 29 Sep 1995 13:39:17 -0700 Received: from localhost.cdrom.com (localhost.cdrom.com [127.0.0.1]) by aslan.cdrom.com (8.6.12/8.6.9) with SMTP id NAA25196; Fri, 29 Sep 1995 13:36:54 -0700 Message-Id: <199509292036.NAA25196@aslan.cdrom.com> X-Authentication-Warning: aslan.cdrom.com: Host localhost.cdrom.com didn't use HELO protocol To: Mark Murray cc: "Justin T. Gibbs" , "Garrett A. Wollman" , "Andrey A. Chernov" , CVS-commiters@freefall.freebsd.org, cvs-sbin@freefall.freebsd.org Subject: Re: cvs commit: src/sbin Makefile In-reply-to: Your message of "Fri, 29 Sep 1995 22:28:33 +0200." <199509292028.WAA22596@grumble.grondar.za> Date: Fri, 29 Sep 1995 13:36:54 -0700 From: "Justin T. Gibbs" Sender: owner-cvs-sbin@FreeBSD.org Precedence: bulk >> Actually, you have to have "MAKE_EBONES" in you're make.conf in order >> to do a full make release which makes both the secure and non-secure >> distributions. Its all screwed up if you ask me. > >Hey? I understand that for a "make release" to work you must _not_ have >MAKE_EBONES in your make.conf. Ask Jordan - I stuffed up part of the last >release schedule (2.0.5) by having this in my /etc/make.conf while Jordan >did not. We were "last-minute-lets-get-this-in" engineering the eBones >code and the @#$% thing was not working at WC. Well, you need the des, and kerberos libraries in your chroot tree in order to make the distribution. The only way that those libraries will be built during the first part of the release (a make world) is if you have MAKE_EBONES in your personal make.conf. The make.conf in the chroot area is installed from the src and does not have MAKE_EBONES in it. Since, until a two days ago, the stuff in eBones didn't reference the locally built version of the kerberos libraries, a make bootstrap in eBones would fail in the chroot area for lack of the des, and krb libraries. Its all a mess. >I agree that it may be useful to be able to set something in make.conf >to _not_ make eBones even if the directory is present, but our present >logic is "broken as designed" IMHO. Yup. >M >-- >Mark Murray >46 Harvey Rd, Claremont, Cape Town 7700, South Africa >+27 21 61-3768 GMT+0200 >Finger mark@grumble.grondar.za for PGP key -- Justin T. Gibbs =========================================== Software Developer - Walnut Creek CDROM FreeBSD: Turning PCs into workstations =========================================== From owner-cvs-sbin Fri Sep 29 13:42:58 1995 Return-Path: owner-cvs-sbin Received: (from root@localhost) by freefall.freebsd.org (8.6.12/8.6.6) id NAA01893 for cvs-sbin-outgoing; Fri, 29 Sep 1995 13:42:58 -0700 Received: from grunt.grondar.za (grunt.grondar.za [196.7.18.129]) by freefall.freebsd.org (8.6.12/8.6.6) with ESMTP id NAA01870 ; Fri, 29 Sep 1995 13:42:31 -0700 Received: from grumble.grondar.za (grumble.grondar.za [196.7.18.130]) by grunt.grondar.za (8.6.12/8.6.9) with ESMTP id WAA15312; Fri, 29 Sep 1995 22:42:08 +0200 Received: from localhost (localhost [127.0.0.1]) by grumble.grondar.za (8.6.12/8.6.9) with SMTP id WAA22735; Fri, 29 Sep 1995 22:42:07 +0200 Message-Id: <199509292042.WAA22735@grumble.grondar.za> X-Authentication-Warning: grumble.grondar.za: Host localhost didn't use HELO protocol To: "Rodney W. Grimes" cc: mark@grondar.za (Mark Murray), gibbs@freefall.freebsd.org, ache@freefall.freebsd.org, CVS-commiters@freefall.freebsd.org, cvs-sbin@freefall.freebsd.org Subject: Re: cvs commit: src/sbin Makefile Date: Fri, 29 Sep 1995 22:42:07 +0200 From: Mark Murray Sender: owner-cvs-sbin@FreeBSD.org Precedence: bulk > > Yehbo! Please do this - get the crap out of src/Makefile! > > Yuko... crap out of src/Makefile and into src/sbin/Makefile == CRAP in top > level make files. This only, IMHO, creates more crap. > > If this ``crap'' goes anyplace for the above case it goes in > src/sbin/init/Makefile. Er, yeah. all the "decision making" is best farmed out to the appropriate app, where possible. I am not suggesting src/sbin/Makefile, but the makefile for the app itself. With a little effort, it should be possible to clean up release/Makefile quite a bit (and speed it up...). M -- Mark Murray 46 Harvey Rd, Claremont, Cape Town 7700, South Africa +27 21 61-3768 GMT+0200 Finger mark@grumble.grondar.za for PGP key From owner-cvs-sbin Fri Sep 29 13:52:54 1995 Return-Path: owner-cvs-sbin Received: (from root@localhost) by freefall.freebsd.org (8.6.12/8.6.6) id NAA02130 for cvs-sbin-outgoing; Fri, 29 Sep 1995 13:52:54 -0700 Received: from grunt.grondar.za (grunt.grondar.za [196.7.18.129]) by freefall.freebsd.org (8.6.12/8.6.6) with ESMTP id NAA02123 ; Fri, 29 Sep 1995 13:52:22 -0700 Received: from grumble.grondar.za (grumble.grondar.za [196.7.18.130]) by grunt.grondar.za (8.6.12/8.6.9) with ESMTP id WAA15327; Fri, 29 Sep 1995 22:52:09 +0200 Received: from localhost (localhost [127.0.0.1]) by grumble.grondar.za (8.6.12/8.6.9) with SMTP id WAA22804; Fri, 29 Sep 1995 22:52:08 +0200 Message-Id: <199509292052.WAA22804@grumble.grondar.za> X-Authentication-Warning: grumble.grondar.za: Host localhost didn't use HELO protocol To: "Justin T. Gibbs" cc: Mark Murray , "Garrett A. Wollman" , "Andrey A. Chernov" , CVS-commiters@freefall.freebsd.org, cvs-sbin@freefall.freebsd.org Subject: Re: cvs commit: src/sbin Makefile Date: Fri, 29 Sep 1995 22:52:08 +0200 From: Mark Murray Sender: owner-cvs-sbin@FreeBSD.org Precedence: bulk > >Hey? I understand that for a "make release" to work you must _not_ have > >MAKE_EBONES in your make.conf. Ask Jordan - I stuffed up part of the last > >release schedule (2.0.5) by having this in my /etc/make.conf while Jordan > >did not. We were "last-minute-lets-get-this-in" engineering the eBones > >code and the @#$% thing was not working at WC. > > Well, you need the des, and kerberos libraries in your chroot tree in > order to make the distribution. The only way that those libraries will > be built during the first part of the release (a make world) is if you > have MAKE_EBONES in your personal make.conf. The make.conf in the chroot > area is installed from the src and does not have MAKE_EBONES in it. > Since, until a two days ago, the stuff in eBones didn't reference the > locally built version of the kerberos libraries, a make bootstrap in > eBones would fail in the chroot area for lack of the des, and krb > libraries. Its all a mess. Not so. Look at the src/release/Makefile release.2: target. There is a positively disgusting hack to get eBones bootstrapped into the make release. If MAKE_EBONES is in your personal make.conf, things break later when you try to make world without it, as you have a nasty mix of secure, eBones and a make.conf with _no_ MAKE_EBONES. Trust me on this one, and ask Jordan and Gary too - they ended up phoning me here in South Africa to fix this same problem :-O. I agree though that this make system is ropey as hell... > >I agree that it may be useful to be able to set something in make.conf > >to _not_ make eBones even if the directory is present, but our present > >logic is "broken as designed" IMHO. > > Yup. > > >M > >-- > >Mark Murray > >46 Harvey Rd, Claremont, Cape Town 7700, South Africa > >+27 21 61-3768 GMT+0200 > >Finger mark@grumble.grondar.za for PGP key > > -- > Justin T. Gibbs > =========================================== > Software Developer - Walnut Creek CDROM > FreeBSD: Turning PCs into workstations > =========================================== -- Mark Murray 46 Harvey Rd, Claremont, Cape Town 7700, South Africa +27 21 61-3768 GMT+0200 Finger mark@grumble.grondar.za for PGP key From owner-cvs-sbin Fri Sep 29 14:02:10 1995 Return-Path: owner-cvs-sbin Received: (from root@localhost) by freefall.freebsd.org (8.6.12/8.6.6) id OAA02534 for cvs-sbin-outgoing; Fri, 29 Sep 1995 14:02:10 -0700 Received: from aslan.cdrom.com (aslan.cdrom.com [192.216.223.142]) by freefall.freebsd.org (8.6.12/8.6.6) with ESMTP id OAA02523 ; Fri, 29 Sep 1995 14:01:59 -0700 Received: from localhost.cdrom.com (localhost.cdrom.com [127.0.0.1]) by aslan.cdrom.com (8.6.12/8.6.9) with SMTP id OAA25322; Fri, 29 Sep 1995 14:01:10 -0700 Message-Id: <199509292101.OAA25322@aslan.cdrom.com> X-Authentication-Warning: aslan.cdrom.com: Host localhost.cdrom.com didn't use HELO protocol To: Mark Murray cc: "Justin T. Gibbs" , "Garrett A. Wollman" , "Andrey A. Chernov" , CVS-commiters@freefall.freebsd.org, cvs-sbin@freefall.freebsd.org Subject: Re: cvs commit: src/sbin Makefile In-reply-to: Your message of "Fri, 29 Sep 1995 22:52:08 +0200." <199509292052.WAA22804@grumble.grondar.za> Date: Fri, 29 Sep 1995 14:01:09 -0700 From: "Justin T. Gibbs" Sender: owner-cvs-sbin@FreeBSD.org Precedence: bulk >Not so. Look at the src/release/Makefile release.2: target. There is a >positively disgusting hack to get eBones bootstrapped into the >make release. If MAKE_EBONES is in your personal make.conf, things >break later when you try to make world without it, as you have a >nasty mix of secure, eBones and a make.conf with _no_ MAKE_EBONES. >Trust me on this one, and ask Jordan and Gary too - they ended up >phoning me here in South Africa to fix this same problem :-O. I don't think this is true anymore. I don't see how having secure binaries could prevent you from building non-secure binaries. It just doesn't make any sense. David complained to me last week about make release not working unless the libraries are in the chroot tree. Look at release.2: release.2: @cd ${.CURDIR} ; $(MAKE) ckRELEASEDIR cd ${.CURDIR}/../etc && make distrib-dirs DESTDIR=${RD}/trees/bin cd ${.CURDIR}/.. ; make distribute DISTDIR=${RD}/trees cd ${.CURDIR}/../eBones && ( \ make bootstrap ;\ The make bootstrap will fail (or used too until I did the LDADD fixes to eBones) since the libraries aren't *installed* in the bootstrap target until after all of eBones is built. I think it was kprop that fell over first. >I agree though that this make system is ropey as hell... Yup. >-- >Mark Murray >46 Harvey Rd, Claremont, Cape Town 7700, South Africa >+27 21 61-3768 GMT+0200 >Finger mark@grumble.grondar.za for PGP key -- Justin T. Gibbs =========================================== Software Developer - Walnut Creek CDROM FreeBSD: Turning PCs into workstations =========================================== From owner-cvs-sbin Fri Sep 29 14:06:56 1995 Return-Path: owner-cvs-sbin Received: (from root@localhost) by freefall.freebsd.org (8.6.12/8.6.6) id OAA02896 for cvs-sbin-outgoing; Fri, 29 Sep 1995 14:06:56 -0700 Received: from grunt.grondar.za (grunt.grondar.za [196.7.18.129]) by freefall.freebsd.org (8.6.12/8.6.6) with ESMTP id OAA02860 ; Fri, 29 Sep 1995 14:06:38 -0700 Received: from grumble.grondar.za (grumble.grondar.za [196.7.18.130]) by grunt.grondar.za (8.6.12/8.6.9) with ESMTP id XAA15359; Fri, 29 Sep 1995 23:05:10 +0200 Received: from localhost (localhost [127.0.0.1]) by grumble.grondar.za (8.6.12/8.6.9) with SMTP id XAA22916; Fri, 29 Sep 1995 23:05:10 +0200 Message-Id: <199509292105.XAA22916@grumble.grondar.za> X-Authentication-Warning: grumble.grondar.za: Host localhost didn't use HELO protocol To: John Capo cc: gibbs@freefall.FreeBSD.org (Justin T. Gibbs), ache@freefall.FreeBSD.org, CVS-commiters@freefall.FreeBSD.org, cvs-sbin@freefall.FreeBSD.org Subject: Re: cvs commit: src/sbin Makefile Date: Fri, 29 Sep 1995 23:05:09 +0200 From: Mark Murray Sender: owner-cvs-sbin@FreeBSD.org Precedence: bulk > MAKE_EBONES or somesuch should control builds with eBones. Builds > with DES should be controlled by something else. This is on its way. A new DES library will be placed in secure, and this library will be used by eBones. > Being able to build with eBones but turn it off at run time would > be nice. A system built with eBones expects it to be used. passwd > defaults to changing the kerberos password for instance. Hmm.. this can be done - there are options to the apps that (I suppose) could be made dependant on am environment variable. Too much work. Too little time. :-( > I think we need some more knobs. I agree, but with great reservation. I would HATE to see too many user configurable hacks in the global system makefiles. I would rather the apps handle this sort of thing intelligently. M -- Mark Murray 46 Harvey Rd, Claremont, Cape Town 7700, South Africa +27 21 61-3768 GMT+0200 Finger mark@grumble.grondar.za for PGP key From owner-cvs-sbin Fri Sep 29 14:16:50 1995 Return-Path: owner-cvs-sbin Received: (from root@localhost) by freefall.freebsd.org (8.6.12/8.6.6) id OAA03259 for cvs-sbin-outgoing; Fri, 29 Sep 1995 14:16:50 -0700 Received: from grunt.grondar.za (grunt.grondar.za [196.7.18.129]) by freefall.freebsd.org (8.6.12/8.6.6) with ESMTP id OAA03246 ; Fri, 29 Sep 1995 14:16:23 -0700 Received: from grumble.grondar.za (grumble.grondar.za [196.7.18.130]) by grunt.grondar.za (8.6.12/8.6.9) with ESMTP id XAA15381; Fri, 29 Sep 1995 23:15:55 +0200 Received: from localhost (localhost [127.0.0.1]) by grumble.grondar.za (8.6.12/8.6.9) with SMTP id XAA22975; Fri, 29 Sep 1995 23:15:54 +0200 Message-Id: <199509292115.XAA22975@grumble.grondar.za> X-Authentication-Warning: grumble.grondar.za: Host localhost didn't use HELO protocol To: "Justin T. Gibbs" cc: Mark Murray , "Garrett A. Wollman" , "Andrey A. Chernov" , CVS-commiters@freefall.freebsd.org, cvs-sbin@freefall.freebsd.org Subject: Re: cvs commit: src/sbin Makefile Date: Fri, 29 Sep 1995 23:15:54 +0200 From: Mark Murray Sender: owner-cvs-sbin@FreeBSD.org Precedence: bulk > I don't think this is true anymore. I don't see how having secure > binaries could prevent you from building non-secure binaries. It just > doesn't make any sense. It didn't make sense. It also didn't work. :-( > David complained to me last week about make release not working unless > the libraries are in the chroot tree. Look at release.2: > > release.2: > @cd ${.CURDIR} ; $(MAKE) ckRELEASEDIR > cd ${.CURDIR}/../etc && make distrib-dirs DESTDIR=${RD}/trees/bin > cd ${.CURDIR}/.. ; make distribute DISTDIR=${RD}/trees > cd ${.CURDIR}/../eBones && ( \ > make bootstrap ;\ Sure - these are just putting things in the ${DISTDIR} (or should be - with recent changes it is quite possible that this is broken). > The make bootstrap will fail (or used too until I did the LDADD fixes to > eBones) since the libraries aren't *installed* in the bootstrap target > until after all of eBones is built. I think it was kprop that fell over > first. Ah yes, the LDADD changes. If those are not _correctly_ set to the appropriate ebones/lib/libfoo/obj/ then it will fail. At 2.0.5 release time, this was working. I own up to breaking this when I redid the make files during the Great Repository Copy (tm). > >I agree though that this make system is ropey as hell... > > Yup. :-) -- Mark Murray 46 Harvey Rd, Claremont, Cape Town 7700, South Africa +27 21 61-3768 GMT+0200 Finger mark@grumble.grondar.za for PGP key From owner-cvs-sbin Fri Sep 29 14:25:01 1995 Return-Path: owner-cvs-sbin Received: (from root@localhost) by freefall.freebsd.org (8.6.12/8.6.6) id OAA03550 for cvs-sbin-outgoing; Fri, 29 Sep 1995 14:25:01 -0700 Received: from GndRsh.aac.dev.com (GndRsh.aac.dev.com [198.145.92.241]) by freefall.freebsd.org (8.6.12/8.6.6) with ESMTP id OAA03526 ; Fri, 29 Sep 1995 14:24:53 -0700 Received: (from rgrimes@localhost) by GndRsh.aac.dev.com (8.6.12/8.6.12) id OAA00787; Fri, 29 Sep 1995 14:24:10 -0700 From: "Rodney W. Grimes" Message-Id: <199509292124.OAA00787@GndRsh.aac.dev.com> Subject: Re: cvs commit: src/sbin Makefile To: mark@grondar.za (Mark Murray) Date: Fri, 29 Sep 1995 14:24:09 -0700 (PDT) Cc: wollman@lcs.mit.edu, mark@grondar.za, gibbs@freefall.freebsd.org, ache@freefall.freebsd.org, CVS-commiters@freefall.freebsd.org, cvs-sbin@freefall.freebsd.org In-Reply-To: <199509292034.WAA22635@grumble.grondar.za> from "Mark Murray" at Sep 29, 95 10:34:07 pm X-Mailer: ELM [version 2.4 PL24] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 1774 Sender: owner-cvs-sbin@FreeBSD.org Precedence: bulk > > > < said: > > > > > Right now we have two wrongs making a right. We _have_ to have a make > > > with no MAKE_EBONES to make the clean distribution, followed by a nasty > > > hack to make the tainted distribution. Why can we not have "make > > > distribution" do it properly the first time? Sure we will need to fix > > > the make files, but that can of worms has been brewing for a while... > > > > Not everyone's build environment is based on `make release' and `make > > distribute'. > > Granted. Then the logic should be "if the eBones directory exists, build > eBones _unless_ asked not to", rather than "the directory must be present > _and_ permission must be granted to build eBones". > > This would mean that secure and eBones can be added to the system just > by dropping them in. At this point in this conversation I would suggest to go read the FreeBSD mail archives for similial discussions and the rational that was used when makeing the decision for that state of the defaults. It has something to do with the path of least surprize to protect people from ``unknowingling'' building a set of binaries that could not be exported. You will find almost _all_ software that has these types of restrictions in it requires 2 things to be done, the addition of the non exportable code source files and the turning of a knob to enable it. I would also suggest that the ad hoc, without white paper, and review changing of the FreeBSD source tree in this area just stinks. Insignificant discussion before action has again occured :-(. -- Rod Grimes rgrimes@gndrsh.aac.dev.com Accurate Automation Company Reliable computers for FreeBSD From owner-cvs-sbin Fri Sep 29 14:35:21 1995 Return-Path: owner-cvs-sbin Received: (from root@localhost) by freefall.freebsd.org (8.6.12/8.6.6) id OAA03844 for cvs-sbin-outgoing; Fri, 29 Sep 1995 14:35:21 -0700 Received: from grunt.grondar.za (grunt.grondar.za [196.7.18.129]) by freefall.freebsd.org (8.6.12/8.6.6) with ESMTP id OAA03811 ; Fri, 29 Sep 1995 14:34:56 -0700 Received: from grumble.grondar.za (grumble.grondar.za [196.7.18.130]) by grunt.grondar.za (8.6.12/8.6.9) with ESMTP id XAA15418; Fri, 29 Sep 1995 23:34:44 +0200 Received: from localhost (localhost [127.0.0.1]) by grumble.grondar.za (8.6.12/8.6.9) with SMTP id XAA23138; Fri, 29 Sep 1995 23:34:43 +0200 Message-Id: <199509292134.XAA23138@grumble.grondar.za> X-Authentication-Warning: grumble.grondar.za: Host localhost didn't use HELO protocol To: "Rodney W. Grimes" cc: mark@grondar.za (Mark Murray), wollman@lcs.mit.edu, gibbs@freefall.freebsd.org, ache@freefall.freebsd.org, CVS-commiters@freefall.freebsd.org, cvs-sbin@freefall.freebsd.org Subject: Re: cvs commit: src/sbin Makefile Date: Fri, 29 Sep 1995 23:34:43 +0200 From: Mark Murray Sender: owner-cvs-sbin@FreeBSD.org Precedence: bulk > It has something to do with the path of least surprize to protect people > from ``unknowingling'' building a set of binaries that could not be > exported. You will find almost _all_ software that has these types of > restrictions in it requires 2 things to be done, the addition of the > non exportable code source files and the turning of a knob to enable it. Unfortunately this has indirectly resulted in eBones not being as clean as it might be. Your logic is clear and sensible, but the make system that has been deduced from it is not. The release/Makefile release.2: target is an example. M -- Mark Murray 46 Harvey Rd, Claremont, Cape Town 7700, South Africa +27 21 61-3768 GMT+0200 Finger mark@grumble.grondar.za for PGP key From owner-cvs-sbin Fri Sep 29 14:42:58 1995 Return-Path: owner-cvs-sbin Received: (from root@localhost) by freefall.freebsd.org (8.6.12/8.6.6) id OAA04234 for cvs-sbin-outgoing; Fri, 29 Sep 1995 14:42:58 -0700 Received: from GndRsh.aac.dev.com (GndRsh.aac.dev.com [198.145.92.241]) by freefall.freebsd.org (8.6.12/8.6.6) with ESMTP id OAA04228 ; Fri, 29 Sep 1995 14:42:42 -0700 Received: (from rgrimes@localhost) by GndRsh.aac.dev.com (8.6.12/8.6.12) id OAA00821; Fri, 29 Sep 1995 14:42:07 -0700 From: "Rodney W. Grimes" Message-Id: <199509292142.OAA00821@GndRsh.aac.dev.com> Subject: Re: cvs commit: src/sbin Makefile To: mark@grondar.za (Mark Murray) Date: Fri, 29 Sep 1995 14:42:07 -0700 (PDT) Cc: mark@grondar.za, wollman@lcs.mit.edu, gibbs@freefall.freebsd.org, ache@freefall.freebsd.org, CVS-commiters@freefall.freebsd.org, cvs-sbin@freefall.freebsd.org In-Reply-To: <199509292134.XAA23138@grumble.grondar.za> from "Mark Murray" at Sep 29, 95 11:34:43 pm X-Mailer: ELM [version 2.4 PL24] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 821 Sender: owner-cvs-sbin@FreeBSD.org Precedence: bulk > > > It has something to do with the path of least surprize to protect people > > from ``unknowingling'' building a set of binaries that could not be > > exported. You will find almost _all_ software that has these types of > > restrictions in it requires 2 things to be done, the addition of the > > non exportable code source files and the turning of a knob to enable it. > > Unfortunately this has indirectly resulted in eBones not being as clean > as it might be. Your logic is clear and sensible, but the make system > that has been deduced from it is not. The release/Makefile release.2: > target is an example. I have no comments on release/Makefile. -- Rod Grimes rgrimes@gndrsh.aac.dev.com Accurate Automation Company Reliable computers for FreeBSD From owner-cvs-sbin Fri Sep 29 15:03:31 1995 Return-Path: owner-cvs-sbin Received: (from root@localhost) by freefall.freebsd.org (8.6.12/8.6.6) id PAA05277 for cvs-sbin-outgoing; Fri, 29 Sep 1995 15:03:31 -0700 Received: (from joerg@localhost) by freefall.freebsd.org (8.6.12/8.6.6) id PAA05258 ; Fri, 29 Sep 1995 15:03:20 -0700 Date: Fri, 29 Sep 1995 15:03:20 -0700 From: Joerg Wunsch Message-Id: <199509292203.PAA05258@freefall.freebsd.org> To: CVS-commiters, cvs-sbin Subject: cvs commit: src/sbin/slattach slattach.c Sender: owner-cvs-sbin@FreeBSD.org Precedence: bulk joerg 95/09/29 15:03:19 Modified: sbin/slattach slattach.c Log: Call the -u UCMD command on exit. The command is called with the current unit number, and the "new" unit number of -1. This allows the script to actually deconfigure the SLIP interface (e.g. by running "ifconfig delete") which has been impossible previously. This is most likely a candidate for 2.1, too. From owner-cvs-sbin Fri Sep 29 17:43:45 1995 Return-Path: owner-cvs-sbin Received: (from root@localhost) by freefall.freebsd.org (8.6.12/8.6.6) id RAA09604 for cvs-sbin-outgoing; Fri, 29 Sep 1995 17:43:45 -0700 Received: from time.cdrom.com (time.cdrom.com [192.216.222.226]) by freefall.freebsd.org (8.6.12/8.6.6) with ESMTP id RAA09590 ; Fri, 29 Sep 1995 17:43:33 -0700 Received: from localhost (localhost [127.0.0.1]) by time.cdrom.com (8.6.12/8.6.9) with SMTP id RAA09675; Fri, 29 Sep 1995 17:42:45 -0700 To: Mark Murray cc: John Capo , gibbs@freefall.FreeBSD.org (Justin T. Gibbs), ache@freefall.FreeBSD.org, CVS-commiters@freefall.FreeBSD.org, cvs-sbin@freefall.FreeBSD.org Subject: Re: cvs commit: src/sbin Makefile In-reply-to: Your message of "Fri, 29 Sep 1995 23:05:09 +0200." <199509292105.XAA22916@grumble.grondar.za> Date: Fri, 29 Sep 1995 17:42:45 -0700 Message-ID: <9673.812421765@time.cdrom.com> From: "Jordan K. Hubbard" Sender: owner-cvs-sbin@FreeBSD.org Precedence: bulk > I agree, but with great reservation. I would HATE to see too many > user configurable hacks in the global system makefiles. I would rather > the apps handle this sort of thing intelligently. Registry! Registry! Would you like to see everything from .emacs to .fvwmrc *go away*? Would you like to see an end to the question of how an arbitrary port stores its startup configuration and/or adds itself to the list of apps started at boot time? Would you like to be able to find out, at a glance, just what version of a given piece of software (or entire subsystem) you were running? Would you like to see sysctl() done CORRECTLY? Registry! Registry! The only problem is, you need the rest of the world to go along. Not easy. Implmenting an NT style registry (with substantial improvements over NT - they didn't get it exactly right nor does it deal with the multiple user issue very well) would be easy. Getting applications weenies to use it would not. The latter reason is mainly why I haven't implemented it yet.. :-( Jordan From owner-cvs-sbin Fri Sep 29 22:59:57 1995 Return-Path: owner-cvs-sbin Received: (from root@localhost) by freefall.freebsd.org (8.6.12/8.6.6) id WAA18108 for cvs-sbin-outgoing; Fri, 29 Sep 1995 22:59:57 -0700 Received: from grunt.grondar.za (grunt.grondar.za [196.7.18.129]) by freefall.freebsd.org (8.6.12/8.6.6) with ESMTP id WAA18089 ; Fri, 29 Sep 1995 22:59:46 -0700 Received: from grumble.grondar.za (grumble.grondar.za [196.7.18.130]) by grunt.grondar.za (8.6.12/8.6.9) with ESMTP id HAA17103; Sat, 30 Sep 1995 07:39:25 +0200 Received: from localhost (localhost [127.0.0.1]) by grumble.grondar.za (8.6.12/8.6.9) with SMTP id HAA26736; Sat, 30 Sep 1995 07:38:09 +0200 Message-Id: <199509300538.HAA26736@grumble.grondar.za> X-Authentication-Warning: grumble.grondar.za: Host localhost didn't use HELO protocol To: "Jordan K. Hubbard" cc: Mark Murray , John Capo , gibbs@freefall.freebsd.org (Justin T. Gibbs), ache@freefall.freebsd.org, CVS-commiters@freefall.freebsd.org, cvs-sbin@freefall.freebsd.org Subject: Re: cvs commit: src/sbin Makefile Date: Sat, 30 Sep 1995 07:38:08 +0200 From: Mark Murray Sender: owner-cvs-sbin@FreeBSD.org Precedence: bulk I was complaining about hacks in _makefiles_. This is an example of doing things OK. In fact it it GREAT. When do we start? ;-) M > > I agree, but with great reservation. I would HATE to see too many > > user configurable hacks in the global system makefiles. I would rather > > the apps handle this sort of thing intelligently. > > Registry! Registry! > > Would you like to see everything from .emacs to .fvwmrc *go away*? > Would you like to see an end to the question of how an arbitrary port > stores its startup configuration and/or adds itself to the list of > apps started at boot time? > Would you like to be able to find out, at a glance, just what version > of a given piece of software (or entire subsystem) you were running? > Would you like to see sysctl() done CORRECTLY? > > Registry! Registry! > > The only problem is, you need the rest of the world to go along. Not > easy. Implmenting an NT style registry (with substantial improvements > over NT - they didn't get it exactly right nor does it deal with the > multiple user issue very well) would be easy. Getting applications > weenies to use it would not. > > The latter reason is mainly why I haven't implemented it yet.. :-( > > Jordan -- Mark Murray 46 Harvey Rd, Claremont, Cape Town 7700, South Africa +27 21 61-3768 GMT+0200 Finger mark@grumble.grondar.za for PGP key From owner-cvs-sbin Sat Sep 30 20:23:25 1995 Return-Path: owner-cvs-sbin Received: (from root@localhost) by freefall.freebsd.org (8.6.12/8.6.6) id UAA29506 for cvs-sbin-outgoing; Sat, 30 Sep 1995 20:23:25 -0700 Received: (from peter@localhost) by freefall.freebsd.org (8.6.12/8.6.6) id UAA29497 ; Sat, 30 Sep 1995 20:23:16 -0700 Date: Sat, 30 Sep 1995 20:23:16 -0700 From: Peter Wemm Message-Id: <199510010323.UAA29497@freefall.freebsd.org> To: CVS-commiters, cvs-sbin Subject: cvs commit: src/sbin/nfsiod nfsiod.c Sender: owner-cvs-sbin@FreeBSD.org Precedence: bulk peter 95/09/30 20:23:15 Modified: sbin/nfsiod nfsiod.c Log: Catch the case where the children can die too soon causing wait3() to return ECHILD, which causes the while loop to run forever. (eg: when accidently running nfsiod on a slow system with a kernel without NFS support...) Obtained from: NetBSD; Frank van den Linden