Requests for -current (14.2-->15.0)

TommyC7 · 07-14-2018, 12:41 PM

Ah, I see.

Thank you rworkman, have a good one.

cwizardone · 07-14-2018, 03:29 PM

PulseAudio-12.1

The change log,

Quote:

Hi all,

Some regressions slipped into the 12.0 release, so here comes 12.1 that
tries to make things better:

* Fixed a crash when switching to the A2DP bluetooth profile
* Fixed the plugin search path in module-ladspa-sink
* Fixed file permissions for the pipes created by module-pipe-sink and
module-pipe-source

The tarball:
https://www.freedesktop.org/software...io-12.1.tar.xz
SHA1: 3d8ecff70b7b47d82fd8a262755ce24dd956bef1
MD5: 2511b7ce7193eecf029c91de96eee547

--
Tanu

https://lists.freedesktop.org/archiv...ly/030259.html

USUARIONUEVO · 07-14-2018, 04:11 PM

utf8proc-2.1.1
https://github.com/JuliaStrings/utf8.../v2.1.1.tar.gz

ivandi · 07-14-2018, 06:10 PM

There are some "lazy" doinst.sh that handle the .new files in bulk. For example dovecot. So I modified the proposed patch from post #1646. It wont check if the new file is listed in doinst.sh, exclude some important files and remove also the .orig files.

removepkg-purge.diff:

Code:

--- removepkg.orig	2018-06-19 19:39:00.000000000 -0400
+++ removepkg	2018-07-14 18:35:30.661000000 -0400
@@ -342,6 +342,28 @@
    fi
    sed -n "/$TRIGGER/,/^$/p" < $ADM_DIR/packages/$PKGNAME | \
     fgrep -v "FILE LIST:" | sort -u > $TMP/delete_list$$
+   if [ "$PURGE" = true ]; then
+    if grep "\.new$" $TMP/delete_list$$ > $TMP/new_files_list$$ 2> /dev/null ; then
+     for NF in $(cat $TMP/new_files_list$$) ; do
+      CF="$(dirname "$NF")/$(basename "$NF" .new)"
+      if [ -r "$ROOT/$CF" ] ; then
+       [ "$CF" == "etc/passwd" ] && continue
+       [ "$CF" == "etc/shadow" ] && continue
+       [ "$CF" == "etc/group" ] && continue
+       [ "$CF" == "etc/gshadow" ] && continue
+       [ "$CF" == "etc/inittab" ] && continue
+       [ "$CF" == "etc/rc.d/rc.inet1.conf" ] && continue
+       echo "$CF" >> $TMP/cfg_files_list$$
+       echo "$CF.orig" >> $TMP/cfg_files_list$$
+      fi
+     done
+     if [ -r $TMP/cfg_files_list$$ ] ; then
+      mv $TMP/delete_list$$ $TMP/delete_list$$.$$
+      cat $TMP/delete_list$$.$$ $TMP/cfg_files_list$$ | sort -u > $TMP/delete_list$$
+      rm -f $TMP/delete_list$$.$$ $TMP/cfg_files_list$$ $TMP/new_files_list$$
+     fi
+    fi
+   fi
    # Pat's new-new && improved pre-removal routine.
    cat_except $ADM_DIR/packages $PKGNAME | sort -u > $TMP/required_list$$
    if [ -r $ADM_DIR/scripts/$PKGNAME ]; then
@@ -415,7 +437,8 @@
   -preserve | --preserve) PRESERVE=true; shift;;
   -terse | --terse) TERSE=0; shift;;
   -warn | --warn) WARN=true; shift;;
-  -* | --*) echo "Usage: $(basename $0) [-copy] [-keep] [-preserve] [-warn] packagename ..."; exit 1;;
+  -purge | --purge) PURGE=true; shift;;
+  -* | --*) echo "Usage: $(basename $0) [-copy] [-keep] [-preserve] [-warn] [-purge] packagename ..."; exit 1;;
   *) break
  esac
 done

removepkg.8-purge.diff:

Code:

--- removepkg.8.orig	2018-06-19 19:28:00.000000000 -0400
+++ removepkg.8	2018-07-14 18:37:18.442000000 -0400
@@ -33,6 +33,9 @@
 [
 .B \--warn
 ]
+[
+.B \--purge
+]
 .BI packagename
 .SH DESCRIPTION
 .B removepkg
@@ -94,6 +97,11 @@
 .B \--warn packagename
 Generate a report to the standard output about which files and directories
 would be removed, but does not actually remove the package.
+.TP
+.B \--purge
+Remove the files coresponding to the .new files in the package. These files are usually handled by
+.B doinst.sh
+upon install. For example if the package contains /etc/foo.conf.new then /etc/foo.conf and /etc/foo.conf.orig will be removed.
 .SH " "
 It's possible to remove a package from a filesystem
 other than / by supplying

If this patch is accepted I think it would be nice to modify slackpkg so that "slackpkg clean-system" runs "removepkg --purge"

Cheers

rworkman · 07-14-2018, 06:54 PM

Quote:

Originally Posted by ivandi

There are some "lazy" doinst.sh that handle the .new files in bulk. For example dovecot. So I modified the proposed patch from post #1646. It wont check if the new file is listed in doinst.sh, exclude some important files and remove also the .orig files.

removepkg-purge.diff:

Code:

--- removepkg.orig	2018-06-19 19:39:00.000000000 -0400
+++ removepkg	2018-07-14 18:35:30.661000000 -0400
@@ -342,6 +342,28 @@
    fi
    sed -n "/$TRIGGER/,/^$/p" < $ADM_DIR/packages/$PKGNAME | \
     fgrep -v "FILE LIST:" | sort -u > $TMP/delete_list$$
+   if [ "$PURGE" = true ]; then
+    if grep "\.new$" $TMP/delete_list$$ > $TMP/new_files_list$$ 2> /dev/null ; then
+     for NF in $(cat $TMP/new_files_list$$) ; do
+      CF="$(dirname "$NF")/$(basename "$NF" .new)"
+      if [ -r "$ROOT/$CF" ] ; then
+       [ "$CF" == "etc/passwd" ] && continue
+       [ "$CF" == "etc/shadow" ] && continue
+       [ "$CF" == "etc/group" ] && continue
+       [ "$CF" == "etc/gshadow" ] && continue
+       [ "$CF" == "etc/inittab" ] && continue
+       [ "$CF" == "etc/rc.d/rc.inet1.conf" ] && continue
+       echo "$CF" >> $TMP/cfg_files_list$$
+       echo "$CF.orig" >> $TMP/cfg_files_list$$
+      fi
+     done
+     if [ -r $TMP/cfg_files_list$$ ] ; then
+      mv $TMP/delete_list$$ $TMP/delete_list$$.$$
+      cat $TMP/delete_list$$.$$ $TMP/cfg_files_list$$ | sort -u > $TMP/delete_list$$
+      rm -f $TMP/delete_list$$.$$ $TMP/cfg_files_list$$ $TMP/new_files_list$$
+     fi
+    fi
+   fi
    # Pat's new-new && improved pre-removal routine.
    cat_except $ADM_DIR/packages $PKGNAME | sort -u > $TMP/required_list$$
    if [ -r $ADM_DIR/scripts/$PKGNAME ]; then
@@ -415,7 +437,8 @@
   -preserve | --preserve) PRESERVE=true; shift;;
   -terse | --terse) TERSE=0; shift;;
   -warn | --warn) WARN=true; shift;;
-  -* | --*) echo "Usage: $(basename $0) [-copy] [-keep] [-preserve] [-warn] packagename ..."; exit 1;;
+  -purge | --purge) PURGE=true; shift;;
+  -* | --*) echo "Usage: $(basename $0) [-copy] [-keep] [-preserve] [-warn] [-purge] packagename ..."; exit 1;;
   *) break
  esac
 done

removepkg.8-purge.diff:

Code:

--- removepkg.8.orig	2018-06-19 19:28:00.000000000 -0400
+++ removepkg.8	2018-07-14 18:37:18.442000000 -0400
@@ -33,6 +33,9 @@
 [
 .B \--warn
 ]
+[
+.B \--purge
+]
 .BI packagename
 .SH DESCRIPTION
 .B removepkg
@@ -94,6 +97,11 @@
 .B \--warn packagename
 Generate a report to the standard output about which files and directories
 would be removed, but does not actually remove the package.
+.TP
+.B \--purge
+Remove the files coresponding to the .new files in the package. These files are usually handled by
+.B doinst.sh
+upon install. For example if the package contains /etc/foo.conf.new then /etc/foo.conf and /etc/foo.conf.orig will be removed.
 .SH " "
 It's possible to remove a package from a filesystem
 other than / by supplying

If this patch is accepted I think it would be nice to modify slackpkg so that "slackpkg clean-system" runs "removepkg --purge"

I can't comment on whether the patch will be accepted, but I can say that slackpkg will not have such modifications. At *best* it would accept a "slackpkg clean-system --purge" option to do that, but removing config files upon package removal would not be desirable for me.

jakedp · 07-14-2018, 09:01 PM

Quote:

Originally Posted by rworkman

I can't comment on whether the patch will be accepted, but I can say that slackpkg will not have such modifications. At *best* it would accept a "slackpkg clean-system --purge" option to do that, but removing config files upon package removal would not be desirable for me.

Or me, or many. There is many situations where it is not desirable. One could write a script to zcat the slackpkg that finds config files to rm. It should be a user intervention in any case.

jakedp · 07-15-2018, 02:15 PM

Catfish 1.4.6, which with this release is now officially a part of XFCE!

Announcement: https://bluesabre.org/2018/07/15/cat...-xfce-project/
Git: https://git.xfce.org/apps/catfish

gmgf · 07-16-2018, 02:17 AM

whois-5.3.2:

http://ftp.debian.org/debian/pool/ma...s_5.3.2.tar.xz

USUARIONUEVO · 07-16-2018, 06:33 PM

wooow

again , pulseaudio

12.2

In the forum i read some problems arround , that

changes --> https://lists.freedesktop.org/archiv...ly/030280.html

sources --> https://www.freedesktop.org/software...io-12.2.tar.xz

casualfred · 07-16-2018, 11:15 PM

I'm not quite sure where to post this, but I just wanted to report that updating xf86-video-ati from 18.0.1 to 20180711_f533b1f6 broke my multi-monitor setup. I use xrandr to set up my screens from a .xinitrc. I posted a bug report about this here: https://bugs.freedesktop.org/show_bug.cgi?id=107253 (I've never posted a bug report before).

My setup includes an onboard Intel graphics card and a Radeon R7 360 card with a monitor attached to each.

I also tried out their latest git commit (16 July 2018, d9a139bc...), but it had the same problem. 18.0.1 works well though!

Darth Vader · 07-17-2018, 12:48 AM

Quote:

Originally Posted by casualfred

I'm not quite sure where to post this, but I just wanted to report that updating xf86-video-ati from 18.0.1 to 20180711_f533b1f6 broke my multi-monitor setup. I use xrandr to set up my screens from a .xinitrc. I posted a bug report about this here: https://bugs.freedesktop.org/show_bug.cgi?id=107253 (I've never posted a bug report before).

My setup includes an onboard Intel graphics card and a Radeon R7 360 card with a monitor attached to each.

I also tried out their latest git commit (16 July 2018, d9a139bc...), but it had the same problem. 18.0.1 works well though!

Did you tried to use the open-source AMDGPU instead of RADEON driver?

A Sea Islands graphics card like yours is also pretty well supported by the open-source AMDGPU, which is without argue with higher performances.

https://wiki.gentoo.org/wiki/AMDGPU
https://wiki.archlinux.org/index.php/AMDGPU
https://help.ubuntu.com/community/AMDGPU-Driver

You need only some kernel parameters for switching the graphics stack

Code:

radeon.cik_support=0 amdgpu.cik_support=1

At my job, happens to have an entire office equipped with R7 260 (8 computers), and they works perfectly with the AMDGPU. True, under SLED, but still...

casualfred · 07-17-2018, 01:23 AM

Quote:

Originally Posted by Darth Vader

Did you tried to use the open-source AMDGPU instead of RADEON driver?
A Sea Islands graphics card like yours is also pretty well supported by the open-source AMDGPU, which is without argue with higher performances.

Cool! I was not aware of the existence of amdgpu - I'll look into it.

Darth Vader · 07-17-2018, 01:29 AM

Quote:

Originally Posted by casualfred

Cool! I was not aware of the existence of amdgpu - I'll look into it.

Have fun!

BTW, you may want a newer kernel together with your new AMDGPU driver, and there you will find ones well made

https://dusk.idlemoor.tk/linux-4.17/x86_64/

Darth Vader · 07-17-2018, 03:14 AM

Quote:

Originally Posted by casualfred

I'm not quite sure where to post this, but I just wanted to report that updating xf86-video-ati from 18.0.1 to 20180711_f533b1f6 broke my multi-monitor setup. I use xrandr to set up my screens from a .xinitrc. I posted a bug report about this here: https://bugs.freedesktop.org/show_bug.cgi?id=107253 (I've never posted a bug report before).

My setup includes an onboard Intel graphics card and a Radeon R7 360 card with a monitor attached to each.

I also tried out their latest git commit (16 July 2018, d9a139bc...), but it had the same problem. 18.0.1 works well though!

You just received a response and your bug report was marked as solved by no one other than Michel Dänzer

Quote:

Michel Dänzer 2018-07-17 08:00:19 UTC

Until this is fixed in xf86-video-intel, it should work with the modesetting driver instead.

*** This bug has been marked as a duplicate of bug 100086 ***

Apparently, your issue is really on xf86-video-intel which does not play well with the updated xf86-video-ati in some particular cases.

casualfred · 07-17-2018, 07:03 AM

Quote:

Apparently, your issue is really on xf86-video-intel which does not play well with the updated xf86-video-ati in some particular cases.

Yep, that's what they said. But I don't really care *which* driver is at fault here - the problem only happened when I updated xf86-video-ati (full details in my bug report), so I'm suggesting that maybe slackware-current might want to stick with 18.0.1 until they get both drivers working properly. Just my two cents anyway.