Path: csiph.com!feeder.erje.net!2.us.feeder.erje.net!news.glorb.com!usenet.stanford.edu!not-for-mail From: Bruce Korb Newsgroups: gnu.utils.bug Subject: Re: GNU sharutils generated archive license Date: Sun, 6 Aug 2017 17:25:18 -0700 Lines: 161 Approved: bug-gnu-utils@gnu.org Message-ID: References: <07492d131b1346b5b1e7998dd54cf235@exmbdft6.ad.twosigma.com> <20170701144908051778979@bob.proulx.com> <33bcde17-724e-58f0-f025-42fb72a9de57@gnu.org> NNTP-Posting-Host: lists.gnu.org Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: 7bit X-Trace: usenet.stanford.edu 1502065532 21147 208.118.235.17 (7 Aug 2017 00:25:32 GMT) X-Complaints-To: action@cs.stanford.edu To: Paul Eggert , "'bug-gnu-utils@gnu.org'" Envelope-to: bug-gnu-utils@gnu.org User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:52.0) Gecko/20100101 Thunderbird/52.1.0 In-Reply-To: Content-Language: en-US X-detected-operating-system: by eggs.gnu.org: GNU/Linux 2.2.x-3.x [generic] X-Received-From: 2001:4830:134:3::e X-BeenThere: bug-gnu-utils@gnu.org X-Mailman-Version: 2.1.21 Precedence: list List-Id: Bug reports for the GNU utilities List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Xref: csiph.com gnu.utils.bug:2214 On 07/11/17 14:07, Paul Eggert wrote: > Bruce Korb wrote: > >> if the person/entity most anxious about that far fetched fear of >> infringement (my opinion) would provide a patch with such a disclaimer, >> I would be happy to incorporate the thing. > > I was going to try to suggest something, but ran into so many problems So much history. Sorry. Cruft from before I started maintaining it and still more by twisted interconnected dependencies that are probably not needed but not worth digging into (mostly 'cuz it works for me). If you are really curious, run bootstrap with xtrace set. > * There are seemingly-duplicate files README.git and README-GIT. Plus, > README-alpha. They are confusing. Surely just one such file is needed. I'll look at that. However, all "foo.git" files are auto-edited and the oritinals (e.g. README.git) should be stripped from the distro. > * './bootstrap' failed because the Fedora package 'autogen' wasn't > installed. > > * './bootstrap' failed because the Fedora package > 'autogen-libopts-devel' wasn't installed. Fedora thinks it a good idea to keep them separate and not dependent. Other distros add a dependency so "autogen" sucks up the libopts. > * './bootstrap' took a veerrrrryy long time, seemingly stuck in > gnulib-tool, without outputting anything. Eventually something happened, > but wow. Autotool issue. "bootstrap" runs them all and the aforementioned "twisted interconnected dependencies" produce this issue. Wow, indeed. > * Eventually './bootstrap' failed with the following diagnostics at the > end. At this point I got lost and gave up. > > ,,, > /usr/bin/autopoint (GNU gettext-tools) 0.19.8.1 > Uses a versions archive in dirxz format. > autoreconf: Entering directory `.' > autoreconf: running: true --force > autoreconf: running: aclocal --force --warnings=all -I m4 > configure.ac:13: warning: macro 'AM_GNU_GETTEXT' not found in library > configure.ac:14: warning: macro 'AM_GNU_GETTEXT_VERSION' not found in > library > configure.ac:110: warning: AC_LANG_CONFTEST: no AC_LANG_SOURCE call > detected in body > ../../lib/autoconf/lang.m4:193: AC_LANG_CONFTEST is expanded from... > ../../lib/autoconf/general.m4:2729: _AC_RUN_IFELSE is expanded from... > ../../lib/m4sugar/m4sh.m4:639: AS_IF is expanded from... > ../../lib/autoconf/general.m4:2748: AC_RUN_IFELSE is expanded from... > ../../lib/m4sugar/m4sh.m4:639: AS_IF is expanded from... > ../../lib/autoconf/general.m4:2031: AC_CACHE_VAL is expanded from... > m4/libopts.m4:158: LIBOPTS_WITHLIB_REGEX is expanded from... > m4/libopts.m4:425: INVOKE_LIBOPTS_MACROS is expanded from... > configure.ac:110: the top level > configure.ac:110: warning: AC_LANG_CONFTEST: no AC_LANG_SOURCE call > detected in body I have complained about this. I've never seen or heard a reasonable solution. The recommendation: Ignore it. :( Anyway, I get a lot further: > make[2]: Entering directory '/u/gnu/proj/sharutils-bld/libopts' > make all-am > make[3]: Entering directory '/u/gnu/proj/sharutils-bld/libopts' > gcc -DHAVE_CONFIG_H -I. -I.. -I. -Wno-format-contains-nul -Wall -ggdb3 -gdwarf-2 -O0 -Wno-format-contains-nul -MT libopts_a-libopts.o -MD -MP -MF .deps/libopts_a-libopts.Tpo -c -o libopts_a-libopts.o `test -f 'libopts.c' || echo './'`libopts.c > In file included from parse-duration.c:30:0, > from libopts.c:16: > intprops.h:24:20: fatal error: verify.h: No such file or directory > #include > ^ > compilation terminated. > Makefile:1309: recipe for target 'libopts_a-libopts.o' failed > make[3]: *** [libopts_a-libopts.o] Error 1 > make[3]: Leaving directory '/u/gnu/proj/sharutils-bld/libopts' > Makefile:1244: recipe for target 'all' failed > make[2]: *** [all] Error 2 > make[2]: Leaving directory '/u/gnu/proj/sharutils-bld/libopts' > Makefile:1319: recipe for target 'all-recursive' failed > make[1]: *** [all-recursive] Error 1 > make[1]: Leaving directory '/u/gnu/proj/sharutils-bld' > Makefile:1258: recipe for target 'all' failed > make: *** [all] Error 2 whatever "verify.h" is supposed to be. Because gnulib-tool automatically triggers a bunch of header dependencies and I can never know which ones or where, I have to parse the gnulib-tool output for a list of header dependencies. Icky poo: >>update_include_list> sed 's@^INCLUDE-LIST$@#include "system.h"\ > #include \ > #include \ > #include \ > #include \ > #include \ > #include \ > #include \ > #include \ > #include \ > #include \ > #include \ > #include \ > \ > #include "base64.h"\ > #include "dirname.h"\ > #include "error.h"\ > #include "idcache.h"\ > #include "inttostr.h"\ > #include "locale.h"\ > #include "md5.h"\ > #include "pathmax.h"\ > #include "quotearg.h"\ > #include "strftime.h"\ > #include "unlocked-io.h"\ > #include "xalloc.h"\ > #include "xgetcwd.h"\ > #include "xstrtol.h"@' local.h.git If I don't do that, I either have to maintain the list by hand or suffer failures because some headers are not included. Rather than that, I force feed that list. I'd prefer idempotent self-sufficient headers, but instead I do this. Anyway, a long explanation for taking a long time. And I guess I need to stuff "verify.h" into the list. So I will make "bootstrap" complain more explicitly about missing tools. <<