[Buildroot] [PATCH 1/5] coreutils: belongs to system tools, not development

[Buildroot] [PATCH 1/5] coreutils: belongs to system tools, not development

[Gustavo Zacarias]

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
---
 package/Config.in | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)

diff --git a/package/Config.in b/package/Config.in
index e2ca8ee..a65ceac 100644
--- a/package/Config.in
+++ b/package/Config.in
@@ -86,9 +86,6 @@ source "package/bison/Config.in"
 source "package/bsdiff/Config.in"
 source "package/bustle/Config.in"
 source "package/ccache/Config.in"
-if BR2_PACKAGE_BUSYBOX_SHOW_OTHERS
-source "package/coreutils/Config.in"
-endif
 source "package/cvs/Config.in"
 if BR2_PACKAGE_BUSYBOX_SHOW_OTHERS
 source "package/diffutils/Config.in"
@@ -959,6 +956,7 @@ source "package/acl/Config.in"
 source "package/attr/Config.in"
 if BR2_PACKAGE_BUSYBOX_SHOW_OTHERS
 source "package/bootutils/Config.in"
+source "package/coreutils/Config.in"
 endif
 source "package/cpuload/Config.in"
 source "package/dsp-tools/Config.in"
-- 
1.8.3.2

[Buildroot] [PATCH 2/5] coreutils: bump to version 8.21

[Gustavo Zacarias]

Set PERL=missing so as to avoid generating manpages with help2man that breaks
on many occasions (and it's pointless).

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
---
 package/coreutils/coreutils-uname.patch | 173 --------------------------------
 package/coreutils/coreutils.mk          |  13 +--
 2 files changed, 3 insertions(+), 183 deletions(-)
 delete mode 100644 package/coreutils/coreutils-uname.patch

diff --git a/package/coreutils/coreutils-uname.patch b/package/coreutils/coreutils-uname.patch
deleted file mode 100644
index faa8908..0000000
--- a/package/coreutils/coreutils-uname.patch
+++ /dev/null
@@ -1,173 +0,0 @@
-On linux platforms, grok /proc/cpuinfo for the CPU/vendor info.
-
-Prob not suitable for upstream seeing as how it's 100% linux-specific
-http://lists.gnu.org/archive/html/bug-coreutils/2005-09/msg00063.html
-
-Patch originally by Carlos E. Gorges <carlos@techlinux.com.br>, but
-heavily reworked to suck less.
-
-To add support for additional platforms, check out the show_cpuinfo()
-func in the linux/arch/<ARCH>/ source tree of the kernel.
-
---- coreutils/src/uname.c
-+++ coreutils/src/uname.c
-@@ -50,6 +50,11 @@
- # include <mach-o/arch.h>
- #endif
- 
-+#if defined(__linux__)
-+# define USE_PROCINFO
-+# define UNAME_HARDWARE_PLATFORM
-+#endif
-+
- #include "system.h"
- #include "error.h"
- #include "quote.h"
-@@ -138,6 +143,117 @@
-   exit (status);
- }
- 
-+#if defined(USE_PROCINFO)
-+
-+# if defined(__s390__) || defined(__s390x__)
-+#  define CPUINFO_FILE    "/proc/sysinfo"
-+#  define CPUINFO_FORMAT  "%64[^\t :]%*[ :]%256[^\n]%c"
-+# else
-+#  define CPUINFO_FILE    "/proc/cpuinfo"
-+#  define CPUINFO_FORMAT  "%64[^\t:]\t:%256[^\n]%c"
-+# endif
-+
-+# define PROCINFO_PROCESSOR      0
-+# define PROCINFO_HARDWARE_PLATFORM 1
-+
-+static void __eat_cpuinfo_space(char *buf)
-+{
-+	/* first eat trailing space */
-+	char *tmp = buf + strlen(buf) - 1;
-+	while (tmp > buf && isspace(*tmp))
-+		*tmp-- = '\0';
-+	/* then eat leading space */
-+	tmp = buf;
-+	while (*tmp && isspace(*tmp))
-+		tmp++;
-+	if (tmp != buf)
-+		memmove(buf, tmp, strlen(tmp)+1);
-+	/* finally collapse whitespace */
-+	tmp = buf;
-+	while (tmp[0] && tmp[1]) {
-+		if (isspace(tmp[0]) && isspace(tmp[1])) {
-+			memmove(tmp, tmp+1, strlen(tmp));
-+			continue;
-+		}
-+		++tmp;
-+	}
-+}
-+
-+static int __linux_procinfo(int x, char *fstr, size_t s)
-+{
-+	FILE *fp;
-+
-+	char *procinfo_keys[] = {
-+		/* --processor --hardware-platform */
-+		#if defined(__alpha__)
-+			"cpu model", "system type"
-+		#elif defined(__arm__)
-+			"Processor", "Hardware"
-+		#elif defined(__avr32__)
-+			"processor", "cpu family"
-+		#elif defined(__bfin__)
-+			"CPU", "BOARD Name"
-+		#elif defined(__cris__)
-+			"cpu", "cpu model"
-+		#elif defined(__frv__)
-+			"CPU-Core", "System"
-+		#elif defined(__i386__) || defined(__x86_64__)
-+			"model name", "vendor_id"
-+		#elif defined(__ia64__)
-+			"family", "vendor"
-+		#elif defined(__hppa__)
-+			"cpu", "model"
-+		#elif defined(__m68k__)
-+			"CPU", "MMU"
-+		#elif defined(__mips__)
-+			"cpu model", "system type"
-+		#elif defined(__powerpc__) || defined(__powerpc64__)
-+			"cpu", "machine"
-+		#elif defined(__s390__) || defined(__s390x__)
-+			"Type", "Manufacturer"
-+		#elif defined(__sh__)
-+			"cpu type", "machine"
-+		#elif defined(sparc) || defined(__sparc__)
-+			"type", "cpu"
-+		#elif defined(__vax__)
-+			"cpu type", "cpu"
-+		#else
-+			"unknown", "unknown"
-+		#endif
-+	};
-+
-+	if ((fp = fopen(CPUINFO_FILE, "r")) != NULL) {
-+		char key[65], value[257], eol, *ret = NULL;
-+
-+		while (fscanf(fp, CPUINFO_FORMAT, key, value, &eol) != EOF) {
-+			__eat_cpuinfo_space(key);
-+			if (!strcmp(key, procinfo_keys[x])) {
-+				__eat_cpuinfo_space(value);
-+				ret = value;
-+				break;
-+			}
-+			if (eol != '\n') {
-+				/* we need two fscanf's here in case the previous
-+				 * length limit caused us to read right up to the
-+				 * newline ... doing "%*[^\n]\n" wont eat the newline
-+				 */
-+				fscanf(fp, "%*[^\n]");
-+				fscanf(fp, "\n");
-+			}
-+		}
-+		fclose(fp);
-+
-+		if (ret) {
-+			strncpy(fstr, ret, s);
-+			return 0;
-+		}
-+	}
-+
-+	return -1;
-+}
-+
-+#endif
-+
- /* Print ELEMENT, preceded by a space if something has already been
-    printed.  */
- 
-@@ -250,10 +344,14 @@ main (int argc, char **argv)
-   if (toprint & PRINT_PROCESSOR)
-     {
-       char const *element = unknown;
--#if HAVE_SYSINFO && defined SI_ARCHITECTURE
-+#if ( HAVE_SYSINFO && defined SI_ARCHITECTURE ) || defined(USE_PROCINFO)
-       {
-         static char processor[257];
-+#if defined(USE_PROCINFO)
-+        if (0 <= __linux_procinfo (PROCINFO_PROCESSOR, processor, sizeof processor))
-+#else
-         if (0 <= sysinfo (SI_ARCHITECTURE, processor, sizeof processor))
-+#endif
-           element = processor;
-       }
- #endif
-@@ -306,9 +404,13 @@ main (int argc, char **argv)
-       if (element == unknown)
-         {
-           static char hardware_platform[257];
-+#if defined(USE_PROCINFO)
-+          if (0 <= __linux_procinfo (PROCINFO_HARDWARE_PLATFORM, hardware_platform, sizeof hardware_platform))
-+#else
-           size_t s = sizeof hardware_platform;
-           static int mib[] = { CTL_HW, UNAME_HARDWARE_PLATFORM };
-           if (sysctl (mib, 2, hardware_platform, &s, 0, 0) >= 0)
-+#endif
-             element = hardware_platform;
-         }
- #endif
diff --git a/package/coreutils/coreutils.mk b/package/coreutils/coreutils.mk
index 17c81f7..e0e9d80 100644
--- a/package/coreutils/coreutils.mk
+++ b/package/coreutils/coreutils.mk
@@ -4,7 +4,7 @@
 #
 ################################################################################
 
-COREUTILS_VERSION = 8.18
+COREUTILS_VERSION = 8.21
 COREUTILS_SITE = $(BR2_GNU_MIRROR)/coreutils
 COREUTILS_SOURCE = coreutils-$(COREUTILS_VERSION).tar.xz
 COREUTILS_LICENSE = GPLv3+
@@ -53,20 +53,13 @@ COREUTILS_CONF_ENV = ac_cv_c_restrict=no \
 		gl_cv_func_working_mkstemp=yes \
 		gl_cv_func_working_utimes=yes \
 		gl_getline_needs_run_time_check=no \
-		utils_cv_localtime_cache=no
+		utils_cv_localtime_cache=no \
+		PERL=missing
 
 COREUTILS_CONF_OPT = --disable-rpath \
 		--disable-dependency-tracking \
 		--enable-install-program=hostname
 
-define COREUTILS_TOUCH_UNAME_C
-	# ensure uname.c file's timestamp does not change,
-	# so help2man does not run
-	touch -d '2010-01-01' $(@D)/src/uname.c
-endef
-
-COREUTILS_POST_PATCH_HOOKS += COREUTILS_TOUCH_UNAME_C
-
 define COREUTILS_POST_INSTALL
 	# some things go in root rather than usr
 	for f in $(COREUTILS_BIN_PROGS); do \
-- 
1.8.3.2

[Buildroot] [PATCH 2/5] coreutils: bump to version 8.21

[Peter Korsgaard]

>>>>> "Gustavo" == Gustavo Zacarias <gustavo@zacarias.com.ar> writes:

 > Set PERL=missing so as to avoid generating manpages with help2man that breaks
 > on many occasions (and it's pointless).

 > Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>

Committed, thanks.

-- 
Bye, Peter Korsgaard

[Buildroot] [PATCH 3/5] util-linux: add agetty->getty symlink

[Gustavo Zacarias]

When busybox isn't installed symlink agetty->getty to make the default
inittab work.

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
---
 package/util-linux/util-linux.mk | 11 +++++++++++
 1 file changed, 11 insertions(+)

diff --git a/package/util-linux/util-linux.mk b/package/util-linux/util-linux.mk
index 0925ab3..20b3d35 100644
--- a/package/util-linux/util-linux.mk
+++ b/package/util-linux/util-linux.mk
@@ -112,6 +112,17 @@ endif
 
 UTIL_LINUX_POST_INSTALL_TARGET_HOOKS += UTIL_LINUX_INSTALL_PAMFILES
 
+# Install agetty->getty symlink to avoid breakage when there's no busybox
+ifeq ($(BR2_PACKAGE_UTIL_LINUX_AGETTY),y)
+ifeq ($(BR2_PACKAGE_BUSYBOX),)
+define UTIL_LINUX_GETTY_SYMLINK
+	ln -sf agetty $(TARGET_DIR)/sbin/getty
+endef
+endif
+endif
+
+UTIL_LINUX_POST_INSTALL_TARGET_HOOKS += UTIL_LINUX_GETTY_SYMLINK
+
 $(eval $(autotools-package))
 $(eval $(host-autotools-package))
 
-- 
1.8.3.2

[Buildroot] [PATCH 3/5] util-linux: add agetty->getty symlink

[Peter Korsgaard]

>>>>> "Gustavo" == Gustavo Zacarias <gustavo@zacarias.com.ar> writes:

 > When busybox isn't installed symlink agetty->getty to make the default
 > inittab work.

Committed, thanks.

-- 
Bye, Peter Korsgaard

[Buildroot] [PATCH 4/5] zsh: alpha sort menu option

[Gustavo Zacarias]

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
---
 package/Config.in | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/package/Config.in b/package/Config.in
index a65ceac..e70a58c 100644
--- a/package/Config.in
+++ b/package/Config.in
@@ -932,7 +932,6 @@ source "package/at/Config.in"
 if BR2_PACKAGE_BUSYBOX_SHOW_OTHERS
 source "package/bash/Config.in"
 source "package/dash/Config.in"
-source "package/zsh/Config.in"
 endif
 source "package/dialog/Config.in"
 source "package/dtach/Config.in"
@@ -949,6 +948,9 @@ source "package/time/Config.in"
 source "package/which/Config.in"
 endif
 source "package/xmlstarlet/Config.in"
+if BR2_PACKAGE_BUSYBOX_SHOW_OTHERS
+source "package/zsh/Config.in"
+endif
 endmenu
 
 menu "System tools"
-- 
1.8.3.2

[Buildroot] [PATCH 4/5] zsh: alpha sort menu option

[Peter Korsgaard]

>>>>> "Gustavo" == Gustavo Zacarias <gustavo@zacarias.com.ar> writes:

 > Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
 > ---
 >  package/Config.in | 4 +++-
 >  1 file changed, 3 insertions(+), 1 deletion(-)

 > diff --git a/package/Config.in b/package/Config.in
 > index a65ceac..e70a58c 100644
 > --- a/package/Config.in
 > +++ b/package/Config.in
 > @@ -932,7 +932,6 @@ source "package/at/Config.in"
 >  if BR2_PACKAGE_BUSYBOX_SHOW_OTHERS
 >  source "package/bash/Config.in"
 >  source "package/dash/Config.in"
 > -source "package/zsh/Config.in"

You are right, certainly - But I think it is more userfriendly to make
an exception and keep all the shells together.

What do others think?

-- 
Bye, Peter Korsgaard

[Buildroot] [PATCH 4/5] zsh: alpha sort menu option

[Thomas Petazzoni]

Dear Peter Korsgaard,

On Tue, 03 Dec 2013 09:09:36 +0100, Peter Korsgaard wrote:
> >>>>> "Gustavo" == Gustavo Zacarias <gustavo@zacarias.com.ar> writes:
> 
>  > Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
>  > ---
>  >  package/Config.in | 4 +++-
>  >  1 file changed, 3 insertions(+), 1 deletion(-)
> 
>  > diff --git a/package/Config.in b/package/Config.in
>  > index a65ceac..e70a58c 100644
>  > --- a/package/Config.in
>  > +++ b/package/Config.in
>  > @@ -932,7 +932,6 @@ source "package/at/Config.in"
>  >  if BR2_PACKAGE_BUSYBOX_SHOW_OTHERS
>  >  source "package/bash/Config.in"
>  >  source "package/dash/Config.in"
>  > -source "package/zsh/Config.in"
> 
> You are right, certainly - But I think it is more userfriendly to make
> an exception and keep all the shells together.
> 
> What do others think?

Ideas:

 * Instead of "Shell and utilities", have two menus: "Shell" and
   "Utilities".

 * In the "Shell and utilities" menu, create two sub-sections,
   separated by comments:

comment "Shells"
... some packages ...

comment "Utilities"
... some packages ...

Best regards,

Thomas
-- 
Thomas Petazzoni, CTO, Free Electrons
Embedded Linux, Kernel and Android engineering
http://free-electrons.com

[Buildroot] [PATCH 4/5] zsh: alpha sort menu option

[Gustavo Zacarias]

On 12/03/2013 05:50 AM, Thomas Petazzoni wrote:

> Ideas:
> 
>  * Instead of "Shell and utilities", have two menus: "Shell" and
>    "Utilities".
> 
>  * In the "Shell and utilities" menu, create two sub-sections,
>    separated by comments:
> 
> comment "Shells"
> ... some packages ...
> 
> comment "Utilities"
> ... some packages ...

Yes why not?
There aren't enough options to justify sub menus.
While we are on this topic the situation for Graphic libraries and
applications has gone out of control, we really need to split it up a
bit because comments for sections are getting mixed up with comments for
dependencies - it _REALLY_ looks ugly.
Any takers?
Regards.

[Buildroot] [PATCH 4/5] zsh: alpha sort menu option

[Peter Korsgaard]

>>>>> "Gustavo" == Gustavo Zacarias <gustavo@zacarias.com.ar> writes:

Hi,

 > While we are on this topic the situation for Graphic libraries and
 > applications has gone out of control, we really need to split it up a
 > bit because comments for sections are getting mixed up with comments for
 > dependencies - it _REALLY_ looks ugly.

Yes, I agree.

 > Any takers?
 > Regards.



-- 
Bye, Peter Korsgaard

[Buildroot] [PATCH 4/5] zsh: alpha sort menu option

[Peter Korsgaard]

>>>>> "Thomas" == Thomas Petazzoni <thomas.petazzoni@free-electrons.com> writes:

 >> You are right, certainly - But I think it is more userfriendly to make
 >> an exception and keep all the shells together.
 >> 
 >> What do others think?

 > Ideas:

 >  * Instead of "Shell and utilities", have two menus: "Shell" and
 >    "Utilities".

 >  * In the "Shell and utilities" menu, create two sub-sections,
 >    separated by comments:

 > comment "Shells"
 > ... some packages ...

 > comment "Utilities"
 > ... some packages ...

Sounds good to me.

-- 
Bye, Peter Korsgaard

[Buildroot] [PATCH 5/5] sysklogd: install to /sbin

[Gustavo Zacarias]

Install syslogd/klogd to /sbin instead of /usr/sbin to really win over
busybox and avoid the ugly noise when booting without busybox installed.
Also drop uninstall hook and manpage installation.

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
---
 package/sysklogd/sysklogd.mk | 18 ++----------------
 1 file changed, 2 insertions(+), 16 deletions(-)

diff --git a/package/sysklogd/sysklogd.mk b/package/sysklogd/sysklogd.mk
index 8ffdc0f..615b084 100644
--- a/package/sysklogd/sysklogd.mk
+++ b/package/sysklogd/sysklogd.mk
@@ -27,28 +27,14 @@ define SYSKLOGD_BUILD_CMDS
 endef
 
 define SYSKLOGD_INSTALL_TARGET_CMDS
-	$(INSTALL) -D -m 0500 $(@D)/syslogd $(TARGET_DIR)/usr/sbin/syslogd
-	$(INSTALL) -D -m 0500 $(@D)/klogd $(TARGET_DIR)/usr/sbin/klogd
-	$(INSTALL) -D -m 0644 $(@D)/sysklogd.8 $(TARGET_DIR)/usr/share/man/man8/sysklogd.8
-	$(INSTALL) -D -m 0644 $(@D)/syslogd.8 $(TARGET_DIR)/usr/share/man/man8/syslogd.8
-	$(INSTALL) -D -m 0644 $(@D)/syslog.conf.5 $(TARGET_DIR)/usr/share/man/man5/syslog.conf.5
-	$(INSTALL) -D -m 0644 $(@D)/klogd.8 $(TARGET_DIR)/usr/share/man/man8/klogd.8
+	$(INSTALL) -D -m 0500 $(@D)/syslogd $(TARGET_DIR)/sbin/syslogd
+	$(INSTALL) -D -m 0500 $(@D)/klogd $(TARGET_DIR)/sbin/klogd
 	if [ ! -f $(TARGET_DIR)/etc/syslog.conf ]; then \
 		$(INSTALL) -D -m 0644 package/sysklogd/syslog.conf \
 			$(TARGET_DIR)/etc/syslog.conf; \
 	fi
 endef
 
-define SYSKLOGD_UNINSTALL_TARGET_CMDS
-	rm -f $(TARGET_DIR)/usr/sbin/syslogd
-	rm -f $(TARGET_DIR)/usr/sbin/klogd
-	rm -f $(TARGET_DIR)/usr/share/man/man8/sysklogd.8
-	rm -f $(TARGET_DIR)/usr/share/man/man8/syslogd.8
-	rm -f $(TARGET_DIR)/usr/share/man/man5/syslog.conf.5
-	rm -f $(TARGET_DIR)/usr/share/man/man8/klogd.8
-	rm -f $(TARGET_DIR)/etc/syslog.conf
-endef
-
 define SYSKLOGD_CLEAN_CMDS
 	$(MAKE) -C $(@D) clean
 endef
-- 
1.8.3.2

[Buildroot] [PATCH 5/5] sysklogd: install to /sbin

[Peter Korsgaard]

>>>>> "Gustavo" == Gustavo Zacarias <gustavo@zacarias.com.ar> writes:

 > Install syslogd/klogd to /sbin instead of /usr/sbin to really win over
 > busybox and avoid the ugly noise when booting without busybox installed.
 > Also drop uninstall hook and manpage installation.

Committed, thanks.

-- 
Bye, Peter Korsgaard

[Buildroot] [PATCH 1/5] coreutils: belongs to system tools, not development

[Peter Korsgaard]

>>>>> "Gustavo" == Gustavo Zacarias <gustavo@zacarias.com.ar> writes:

 > Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
 > ---
 >  package/Config.in | 4 +---
 >  1 file changed, 1 insertion(+), 3 deletions(-)

 > diff --git a/package/Config.in b/package/Config.in
 > index e2ca8ee..a65ceac 100644
 > --- a/package/Config.in
 > +++ b/package/Config.in
 > @@ -86,9 +86,6 @@ source "package/bison/Config.in"
 >  source "package/bsdiff/Config.in"
 >  source "package/bustle/Config.in"
 >  source "package/ccache/Config.in"
 > -if BR2_PACKAGE_BUSYBOX_SHOW_OTHERS
 > -source "package/coreutils/Config.in"
 > -endif

Committed, thanks.

I'm really starting to think we should get rid of
BR2_PACKAGE_BUSYBOX_SHOW_OTHERS, or atleast make it default y.

Comments, anybody?

-- 
Bye, Peter Korsgaard

[Buildroot] [PATCH 1/5] coreutils: belongs to system tools, not development

[Jeremy Rosen]

> Committed, thanks.
> 
> I'm really starting to think we should get rid of
> BR2_PACKAGE_BUSYBOX_SHOW_OTHERS, or atleast make it default y.
> 
> Comments, anybody?
> 


I personally don't know exactly what busybox provides, this option is
a simple way to make sure I don't compile any useless stuff...

I'm in favor of keeping it or provide another way to find duplicates

> --
> Bye, Peter Korsgaard
> _______________________________________________
> buildroot mailing list
> buildroot at busybox.net
> http://lists.busybox.net/mailman/listinfo/buildroot
> 

[Buildroot] [PATCH 1/5] coreutils: belongs to system tools, not development

[Gustavo Zacarias]

On 12/03/2013 05:08 AM, Jeremy Rosen wrote:

> I personally don't know exactly what busybox provides, this option is
> a simple way to make sure I don't compile any useless stuff...
> 
> I'm in favor of keeping it or provide another way to find duplicates

Sometimes the busybox version is short of features and you need the full
version even when keeping busybox around.
With that being said you sometimes want to avoid the opposite too. On
this patchset i've found there are many of the big packages required
(and needed fixing) to get a nice (no warnings or error messages)
booting system with the default skeleton, OTOH: coreutils, util-linux,
some shell (bash), sysvinit, sysklogd, linux-pam (inherent for util-linux).
And with those we still get a warning from util-linux's hostname not
being feature complete (oh the irony, busybox hostname is, we need
net-tools to replace it).
And we're still lacking ifup/ifdown :)
Regards.

[Buildroot] [PATCH 1/5] coreutils: belongs to system tools, not development

[Peter Korsgaard]

>>>>> "Gustavo" == Gustavo Zacarias <gustavo@zacarias.com.ar> writes:

 > On 12/03/2013 05:08 AM, Jeremy Rosen wrote:
 >> I personally don't know exactly what busybox provides, this option is
 >> a simple way to make sure I don't compile any useless stuff...
 >> 
 >> I'm in favor of keeping it or provide another way to find duplicates

 > Sometimes the busybox version is short of features and you need the full
 > version even when keeping busybox around.

Exactly, like we're hiding bash/zsh even though the differences from
busybox ash are pretty huge. Same for syslogd/sysvinit or vim.

 > With that being said you sometimes want to avoid the opposite too. On
 > this patchset i've found there are many of the big packages required
 > (and needed fixing) to get a nice (no warnings or error messages)
 > booting system with the default skeleton, OTOH: coreutils, util-linux,
 > some shell (bash), sysvinit, sysklogd, linux-pam (inherent for util-linux).
 > And with those we still get a warning from util-linux's hostname not
 > being feature complete (oh the irony, busybox hostname is, we need
 > net-tools to replace it).
 > And we're still lacking ifup/ifdown :)

Which relate to the point about these options not getting a lot of
testing.

-- 
Bye, Peter Korsgaard

[Buildroot] [PATCH 1/5] coreutils: belongs to system tools, not development

[Thomas Petazzoni]

Dear Peter Korsgaard,

On Tue, 03 Dec 2013 09:01:43 +0100, Peter Korsgaard wrote:
> >>>>> "Gustavo" == Gustavo Zacarias <gustavo@zacarias.com.ar> writes:
> 
>  > Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
>  > ---
>  >  package/Config.in | 4 +---
>  >  1 file changed, 1 insertion(+), 3 deletions(-)
> 
>  > diff --git a/package/Config.in b/package/Config.in
>  > index e2ca8ee..a65ceac 100644
>  > --- a/package/Config.in
>  > +++ b/package/Config.in
>  > @@ -86,9 +86,6 @@ source "package/bison/Config.in"
>  >  source "package/bsdiff/Config.in"
>  >  source "package/bustle/Config.in"
>  >  source "package/ccache/Config.in"
>  > -if BR2_PACKAGE_BUSYBOX_SHOW_OTHERS
>  > -source "package/coreutils/Config.in"
>  > -endif
> 
> Committed, thanks.
> 
> I'm really starting to think we should get rid of
> BR2_PACKAGE_BUSYBOX_SHOW_OTHERS, or atleast make it default y.
> 
> Comments, anybody?

We've always said that it was important to keep this option so that
newcomers don't enable things like coreutils, bash and so on, and use
what I'd consider the default "Buildroot experience", i.e Busybox.

What made you change your mind about this?

Best regards,

Thomas
-- 
Thomas Petazzoni, CTO, Free Electrons
Embedded Linux, Kernel and Android engineering
http://free-electrons.com

[Buildroot] [PATCH 1/5] coreutils: belongs to system tools, not development

[Peter Korsgaard]

>>>>> "Thomas" == Thomas Petazzoni <thomas.petazzoni@free-electrons.com> writes:

Hi,

 >> Comments, anybody?

 > We've always said that it was important to keep this option so that
 > newcomers don't enable things like coreutils, bash and so on, and use
 > what I'd consider the default "Buildroot experience", i.e Busybox.

True.

 > What made you change your mind about this?

It mainly dates back from when those busybox alternatives didn't get
much testing, so were likely to fail. We default to having busybox
enabled, but I don't think we really need to make it more difficult to
use the alternatives.

E.G. if we support bash/coreutils/.., then we should really support
them.

-- 
Bye, Peter Korsgaard

[Buildroot] [PATCH 1/5] coreutils: belongs to system tools, not development

[Thomas Petazzoni]

Dear Peter Korsgaard,

On Tue, 03 Dec 2013 11:05:19 +0100, Peter Korsgaard wrote:

>  > We've always said that it was important to keep this option so that
>  > newcomers don't enable things like coreutils, bash and so on, and use
>  > what I'd consider the default "Buildroot experience", i.e Busybox.
> 
> True.
> 
>  > What made you change your mind about this?
> 
> It mainly dates back from when those busybox alternatives didn't get
> much testing, so were likely to fail. We default to having busybox
> enabled, but I don't think we really need to make it more difficult to
> use the alternatives.
> 
> E.G. if we support bash/coreutils/.., then we should really support
> them.

I definitely agree with this last part, but I don't see how "really
supporting them" conflicts with the idea of hiding them by default to
avoid having newcomers confused by these.

If the amount of conflicts in package/Config.in is your concern, then I
believe I agree with Gustavoz suggestion of moving the dependency on
BUSYBOX_SHOW_OTHERS down to the individual package/<foo>/Config.in.

Best regards,

Thomas
-- 
Thomas Petazzoni, CTO, Free Electrons
Embedded Linux, Kernel and Android engineering
http://free-electrons.com

[Buildroot] [PATCH 1/5] coreutils: belongs to system tools, not development

[Peter Korsgaard]

>>>>> "Thomas" == Thomas Petazzoni <thomas.petazzoni@free-electrons.com> writes:

Hi,

 >> E.G. if we support bash/coreutils/.., then we should really support
 >> them.

 > I definitely agree with this last part, but I don't see how "really
 > supporting them" conflicts with the idea of hiding them by default to
 > avoid having newcomers confused by these.

It's mainly about when the featureset significantly differs between
busybox and the "full" variant. The question if the busybox version is
"good enough" unfortunately is quite system/use case dependent.

-- 
Bye, Peter Korsgaard

[Buildroot] [PATCH 1/5] coreutils: belongs to system tools, not development

[Gustavo Zacarias]

On 12/03/2013 05:59 AM, Thomas Petazzoni wrote:

>> Committed, thanks.
>>
>> I'm really starting to think we should get rid of
>> BR2_PACKAGE_BUSYBOX_SHOW_OTHERS, or atleast make it default y.
>>
>> Comments, anybody?
> 
> We've always said that it was important to keep this option so that
> newcomers don't enable things like coreutils, bash and so on, and use
> what I'd consider the default "Buildroot experience", i.e Busybox.
> 
> What made you change your mind about this?

/me thinks he got some patch conflict or something like that :)
We can always make it per-package in package/x/Config.in -> depends on
BR2_PACKAGE_BUSYBOX_SHOW_OTHERS to avoid those issues.
For me it's default y whenever i use BR, however i have no strong
opinion either way.
Regards.