LinuxQuestions.org
Page 14 of 197 « First 41213 14 15162464114 Last »
Show 50 post(s) from this thread on one page

LinuxQuestions.org (/questions/)
-   Slackware (https://www.linuxquestions.org/questions/slackware-14/)
-   -   Requests for -current (14.2-->15.0) (https://www.linuxquestions.org/questions/slackware-14/requests-for-current-14-2-15-0-a-4175620463/)

saxa 03-13-2018 08:19 PM
adwaita-icon-theme 3.28.0
https://download.gnome.org/sources/a...-3.28.0.tar.xz

Maybe is it also worth wait a bit for gstreamer 1.14.x series.

saxa 03-13-2018 08:21 PM
gnome-keyring-3.28.0.1
https://download.gnome.org/sources/g....28.0.1.tar.xz

atelszewski 03-14-2018 12:58 AM
Hi,

Code:
http://slackware.osuosl.org/slackware-current/source/xap/mozilla-firefox/.mozconfig
Leftover possibly?

--
Best regards,
Andrzej Telszewski

gmgf 03-14-2018 01:01 AM
libinput-1.10.3:

https://cgit.freedesktop.org/wayland...?h=1.10-branch
https://www.freedesktop.org/software...-1.10.3.tar.xz

gmgf 03-14-2018 02:22 AM
Quote:
Originally Posted by gmgf (Post 5829877)

hexchat-2.14.0:

(hexchat use 'meson', and need 'luajit' for build, also, now) ;)

http://luajit.org/

(hexchat build with the SBO 'luajit' version)
https://slackbuilds.org/slackbuilds/...jit.SlackBuild

https://dl.hexchat.net/hexchat/hexchat-2.14.0.tar.xz
Finally it build on current without luajit, like this:

Code:
mkdir build
cd build
meson setup \
  --prefix=/usr \
  --libdir=lib${LIBDIRSUFFIX} \
  --libexecdir /usr/libexec \
  --bindir /usr/bin \
  --sbindir /usr/sbin \
  --includedir /usr/include \
  --datadir /usr/share \
  --mandir /usr/man \
  --sysconfdir /etc \
  --localstatedir /var \
  -Dwith-lua=false \
  .. || exit 1
  ninja || exit 1
  DESTDIR=$PKG ninja install || exit 1
cd ..

Thom1b 03-14-2018 02:23 AM
curl-7.59.0 is released with three security fixes.

Quote:
=======================================================

Project curl Security Advisory, March 14th 2018 -
[Permalink](https://curl.haxx.se/docs/adv_2018-9cd6.html)

VULNERABILITY
-------------

curl can be fooled into writing a zero byte out of bounds.

This bug can trigger when curl is told to work on an FTP URL, with the setting
to only issue a single CWD command (`--ftp-method singlecwd` or the libcurl
alternative `CURLOPT_FTP_FILEMETHOD`).

curl then URL-decodes the given path, calls strlen() on the result and deducts
the length of the file name part to find the end of the directory within the
buffer. It then writes a zero byte on that index, in a buffer allocated on the
heap.

If the directory part of the URL contains a "%00" sequence, the directory
length might end up shorter than the file name path, making the calculation
`size_t index = directory_len - filepart_len` end up with a huge index
variable for where the zero byte gets stored: `heap_buffer[index] = 0`. On
several architectures that huge index will wrap and work as a negative value,
thus overwriting memory *before* the intended heap buffer.

By using different file part lengths and putting %00 in different places in
the URL, an attacker that can control what paths a curl-using application uses
can write that zero byte on different indexes.

We are not aware of any exploit of this flaw.

INFO
----

This bug was introduced in December 2004 in [this
commit](https://github.com/curl/curl/commit/6e1e9caa32da0995).

The Common Vulnerabilities and Exposures (CVE) project has assigned the name
CVE-2018-1000120 to this issue.

CWE-122: Heap-based Buffer Overflow

AFFECTED VERSIONS
-----------------

- Affected versions: curl 7.12.3 to and including curl 7.58.0
- Not affected versions: curl < 7.12.3 and curl >= 7.59.0

libcurl is used by many applications, but not always advertised as such.

THE SOLUTION
------------

In curl version 7.59.0, curl rejects FTP URLs that contain any "control
characters". That is byte values below ascii 32.

A [patch for CVE-2018-1000120](https://curl.haxx.se/CVE-2018-1000120.patch) is available
Quote:
LDAP NULL pointer dereference
=============================

Project curl Security Advisory, March 14th 2018 -
[Permalink](https://curl.haxx.se/docs/adv_2018-97a2.html)

VULNERABILITY
-------------

curl might dereference a near-NULL address when getting an LDAP URL.

The function `ldap_get_attribute_ber()` is called to get attributes, but it
turns out that it can return `LDAP_SUCCESS` and still return a `NULL` pointer
in the result pointer when getting a particularly crafted response. This was a
surprise to us and to the code.

libcurl-using applications that allow LDAP URLs, or that allow redirects to
LDAP URLs could be made to crash by a malicious server.

We are not aware of any exploit of this flaw.

INFO
----

The bug is only present in curl versions built to use OpenLDAP.

This bug was introduced in May 2010 in [this
commit](https://github.com/curl/curl/commit/2e056353b00d09).

The Common Vulnerabilities and Exposures (CVE) project has assigned the name
CVE-2018-1000121 to this issue.

CWE-476: NULL Pointer Dereference

AFFECTED VERSIONS
-----------------

- Affected versions: curl 7.21.0 to and including curl 7.58.0
- Not affected versions: curl < 7.21.0 and curl >= 7.59.0

libcurl is used by many applications, but not always advertised as such.

THE SOLUTION
------------

In curl version 7.59.0, curl checks the pointer properly before using it.

A [patch for CVE-2018-1000121](https://curl.haxx.se/CVE-2018-1000121.patch) is available.
Quote:
RTSP RTP buffer over-read
=========================

Project curl Security Advisory, March 14th 2018 -
[Permalink](https://curl.haxx.se/docs/adv_2018-b047.html)

VULNERABILITY
-------------

curl can be tricked into copying data beyond end of its heap based buffer.

When asked to transfer an RTSP URL, curl could calculate a wrong data length
to copy from the read buffer. The memcpy call would copy data from the heap
following the buffer to a storage area that would subsequently be delivered to
the application (if it didn't cause a crash). We've managed to get it to reach
several hundreds bytes out of range.

This could lead to information leakage or a denial of service for the
application if the server offering the RTSP data can trigger this.

We are not aware of any exploit of this flaw.

INFO
----

This bug was introduced in January 2010 in [this
commit](https://github.com/curl/curl/commit/bc4582b68a673d3) when RTSP support
was first added.

The Common Vulnerabilities and Exposures (CVE) project has assigned the name
CVE-2018-1000122 to this issue.

CWE-126: Buffer Over-read

AFFECTED VERSIONS
-----------------

- Affected versions: curl 7.20.0 to and including curl 7.58.0
- Not affected versions: curl < 7.20.0 and curl >= 7.59.0

libcurl is used by many applications, but not always advertised as such.

THE SOLUTION
------------

In curl version 7.59.0, curl makes sure that this code never gets told to copy
more data than it is allowed to read from the buffer.

A [patch for CVE-2018-1000122](https://curl.haxx.se/CVE-2018-1000122.patch) is available.

cwizardone 03-14-2018 10:20 AM
Samba-4.8.0

Release notes, https://www.samba.org/samba/history/samba-4.8.0.html

Thom1b 03-14-2018 01:56 PM
new bind versions are released with security fix.

Quote:
Posting date: 14 March 2018
Program Impacted: BIND
Versions affected: 9.0.x -> 9.8.x, 9.9.0->9.9.11-P1,
9.10.0->9.10.6-P1, 9.11.0->9.11.2-P1,
9.12.0->9.12.0-P1
Description:

"update-policy local;", which is a permission cluster provided
as a shortcut for operators who use Dynamic DNS (DDNS), was
misleadingly named in that its original implementation did not
actually enforce a requirement that the updates it allows originate
locally.

A full description of "update-policy local;" is included in
Section 6.2 of the BIND Administrator Reference Manual, but to
briefly summarize:

When "update-policy local;" is set for a zone in named.conf,
named will create and use an automatically generated session
key (named "local-ddns" by default and stored in local storage
on the server) and will permit updates to the zone to any
client successfully authenticating using that key.

Since the key is generated and stored locally, in usual cases
this should equate to only allowing local updates unless an
operator deliberately copies the local-ddns key elsewhere.

However, in June 2017 disclosed CVE-2017-3143, a flaw in TSIG
authentication which enabled an attacker who was able to send
and receive messages to an authoritative DNS server and who had
knowledge of a valid TSIG key name for the zone and service being
targeted to manipulate BIND into accepting an unauthorized dynamic
update. In our disclosure for CVE-2017-3143 we warned of its
potential interaction with the behavior of update-policy local.
By policy, however, ISC prefers security releases to contain
only the minimal fix required to prevent the exploitable condition.
Therefore, security patch releases for CVE-2017-3143 fixed only
the TSIG authentication flaw without changing the behavior of
the "update-policy local;" feature.

Beginning with the March 2018 maintenance releases of BIND
(9.9.12, 9.10.7, 9.11.3, and 9.12.1) the behavior of "update-policy
local" is now changed so that updates are permitted under the
policy only when they are received from locally configured
addresses AND use the local session key.

Impact:

We think it is unlikely that many operators are deliberately
relying on the non-local option of the previous behavior (and
we recommend against it) but if any are, please see the "Workarounds"
section of this advisory for advice on how to replicate the
previous behavior. For all other operators (those who were not
relying on the non-local side-effect of the previous behavior)
the new behavior should represent an improvement in DDNS security
if you use the local update policy.

Workarounds:

The change in "update-policy local;" behavior which debuts in
the March 2018 maintenance releases should improve security by
properly restricting updates to only those that are received
from locally configured addresses AND are authenticated using
the local session key. However, in the event that an operator
was deliberately relying on the non-local option of the previous
behavior, behavior equivalent to the previous behavior of
"update-policy local;" can be produced by using this syntax:
"update-policy { grant local-ddns zonesub any; };"

Solution:

Software versions which enforce the corrected, more restrictive
behavior are now available from our downloads page,
http://www.isc.org/downloads/all.

- BIND 9 version 9.9.12
- BIND 9 version 9.10.7
- BIND 9 version 9.11.3
- BIND 9 version 9.12.1

Do you still have questions? Questions regarding this advisory
should go to security-officer@isc.org. To report a new issue,
please encrypt your message using security-officer@isc.org's PGP
key which can be found here:
https://www.isc.org/downloads/softwa...y/openpgp-key/.
If you are unable to use encrypted email, you may also report new
issues at: https://www.isc.org/community/report-bug/.

dugan 03-14-2018 09:08 PM
vim-gvim's slack-desc still says it's build against GTK2, even though that hasn't been true in quite a while.

gmgf 03-15-2018 03:38 AM
Quote:
Originally Posted by gmgf (Post 5830436)
gvfs.SlackBuild for current with meson:

Code:
mkdir -p build
cd build
meson setup \
  --prefix=/usr \
  --libdir=lib${LIBDIRSUFFIX} \
  --libexecdir    /usr/libexec \
  --mandir        /usr/man \
  --sysconfdir    /etc \
  --localstatedir  /var \
  -Dsystemduserunitdir=false \
  -Dtmpfilesdir=false \
  -Ddnssd=false \
  -Dgoogle=false \
  -Dgoa=false \
  -Dnfs=false \
  -Dlogind=false \
  -Dman=true \
  -Ddeprecated_programs=true \
  .. || exit 1
  ninja || exit 1
  DESTDIR=$PKG ninja install || exit 1
cd ..

gmgf 03-15-2018 12:51 PM
xfce4-terminal-0.8.7.2

https://git.xfce.org/apps/xfce4-terminal/
http://archive.xfce.org/src/apps/xfc....8.7.2.tar.bz2

Ne01eX 03-15-2018 06:17 PM
1 Attachment(s)
Blue-1.13.tar.bz2 (skin for MPlayer)

Also, may be interest: Blue-multilingual-1.5.tar.bz2, and more others. ;-)

Imho, this is very usability:

http://www.mplayerhq.hu/images/skins/ICY-01.jpg

As compromise, my variant MPlayer.SlackBuild into attachment. :-)

About ICY:
1. Support buttons, not present into Blue (such as "URL" etc).
2. Argument of your choice. :-D
3. New, modern look for Slackware MPlayer.
4. Same Blue. :-D

gmgf 03-16-2018 03:10 AM
libtirpc-1.0.3:

https://sourceforge.net/projects/lib...ibtirpc/1.0.3/
https://sourceforge.net/projects/lib...r.bz2/download

gmgf 03-16-2018 07:08 AM
xf86-video-ati-18.0.1 and xf86-video-amdgpu-18.0.1:

https://cgit.freedesktop.org/xorg/dr...f86-video-ati/
ftp://ftp.x.org/pub/individual/drive...-18.0.1.tar.gz

https://cgit.freedesktop.org/xorg/dr...-video-amdgpu/
ftp://ftp.x.org/pub/individual/drive...-18.0.1.tar.gz

MDKDIO 03-16-2018 10:36 AM
dmidecode 3.1

http://savannah.nongnu.org/projects/dmidecode/

Download can be found here
http://download.savannah.gnu.org/rel...ode-3.1.tar.xz

gmgf 03-16-2018 12:53 PM
hdparm-3.55:

https://sourceforge.net/projects/hdparm/files/hdparm/
https://sourceforge.net/projects/hdp...ar.gz/download

cmake-3.10.3:

https://blog.kitware.com/cmake-3-10-...-for-download/
https://cmake.org/files/v3.10/cmake-3.10.3.tar.gz

USUARIONUEVO 03-16-2018 05:07 PM
libvorbis-1.3.6
https://ftp.osuosl.org/pub/xiph/rele...s-1.3.6.tar.xz

USUARIONUEVO 03-17-2018 12:15 AM
python-setuptools-38.6.0
https://pypi.python.org/packages/95/...ols-38.6.0.zip

alex14641 03-17-2018 08:12 AM
bluez 5.49
Change log: http://www.bluez.org/release-of-bluez-5-49/
Source: http://www.kernel.org/pub/linux/blue...ez-5.49.tar.xz

Darth Vader 03-17-2018 10:39 AM
Again, https://github.com/i-rinat/libvdpau-va-gl

What the heck is your ethical/political/religious issues with this poor VDPAU/VA-API bridge which lacks from Slackware? It is just a really small library!

Dear BDFL, please be aware that some people have 15W CPUs in their ultra-modern Intel based computers or laptops and they may want to use MPlayer or Flash Player!

gmgf 03-17-2018 03:06 PM
pygobject-3.28.1:

http://ftp.gnome.org/pub/gnome/sourc...ct-3.28.1.news
http://ftp.gnome.org/pub/gnome/sourc...-3.28.1.tar.xz

bash-completion-2.8:

https://github.com/scop/bash-completion/releases
https://github.com/scop/bash-complet...ion-2.8.tar.xz

atelszewski 03-17-2018 03:27 PM
Hi,

Shouldn't rc.consolekit come before rc.networkmanager in rc.M as per discussion in here?

--
Best regards,
Andrzej Telszewski

USUARIONUEVO 03-17-2018 03:59 PM
Quote:
Originally Posted by atelszewski (Post 5832163)
Hi,

Shouldn't rc.consolekit come before rc.networkmanager in rc.M as per discussion in here?

--
Best regards,
Andrzej Telszewski
And this is needed ?

Code:
# Start HAL:
if [ -x /etc/rc.d/rc.hald ]; then
  sh /etc/rc.d/rc.hald start
fi
no hal package from slackware .. or im lost something ?

orbea 03-17-2018 04:51 PM
I don't think its needed, but its also doesn't cause any harm since it doesn't do anything without hal.

USUARIONUEVO 03-17-2018 05:56 PM
Quote:
Originally Posted by orbea (Post 5832180)
I don't think its needed, but its also doesn't cause any harm since it doesn't do anything without hal.
I comment cause i think same as other user , some clean ups in rc.M time.
Arround HAL , is not part of slackware and is not present in slackbuilds ... then , im not see reasons to have code arround that.

ngc891 03-17-2018 10:27 PM
Gnome 3.28.0 is out, so it may be a good time to put the Gnome libraries used in slackware-current in sync with the release, to make the integration of Gnome software in Slackware easier. All sources are available here. Here are the status of these packages in Slackware:

Code:
Already up-to-date:
NetworkManager          1.10.6
atkmm                  2.24.2
glibmm                  2.54.1
gtkmm3                  3.22.2
libcroco                0.6.12
libgsf                  1.14.42
libgudev                232
libnotify              0.7.7
librsvg                2.42.3
libsecret              0.18.5
libsigc++              2.10.0
network-manager-applet  1.8.10
pangomm                2.40.1


Bugfix level:
at-spi2-atk    2.26.1  -> 2.26.2
gtk+3          3.22.28 -> 3.22.29


Minor level:
adwaita-icon-theme              3.26.1  -> 3.27.90
at-spi2-core                    2.26.2  -> 2.28.0
atk                            2.26.1  -> 2.28.1
dconf                          0.26.1  -> 0.27.1
gcr                            3.26.0  -> 3.28.0
glib2                          2.54.3  -> 2.56.0
glib-networking                2.54.1  -> 2.56.0
gnome-keyring                  3.20.1  -> 3.28.0.1
gobject-introspection          1.54.1  -> 1.56.0
gsettings-desktop-schemas      3.24.1  -> 3.28.0
gvfs                            1.34.2.1 -> 1.36.0
libsoup                        2.60.3  -> 2.62.0
pango                          1.40.14  -> 1.42.0
pygobject3                      3.26.1  -> 3.28.0
vte                            0.50.2  -> 0.52.0


Major level:
gmime  2.6.23 -> 3.2.0

All the other packages are not (yet?) shipped in Slackware.

dugan 03-17-2018 11:28 PM
I was seeing what I could do about updating the vulkingsdk SlackBuild (so far it looks like you no longer need to build spirv-tools separately). And then I noticed this line in vulkan-sdk.SlackBuild:

Code:
ln -s /usr/include/SPIRV $PKG/usr/include/spirv
Really?

I would have expected it to have been done using the usual pattern, using the parentheses and the subshell.

USUARIONUEVO 03-18-2018 02:45 PM
dnsmasq-2.79
http://www.thekelleys.org.uk/dnsmasq...sq-2.79.tar.xz

Cython-0.28.1
https://pypi.python.org/packages/be/...-0.28.1.tar.gz

python-setuptools-39.0.1
https://pypi.python.org/packages/72/...ols-39.0.1.zip

dchmelik 03-18-2018 08:03 PM
science/mathematics/graphics packages
 
I'd like a functional contemporary (DOS or Windows 3.1+, Win3 keystrokes) tabbed text-editor such as Notepadqq. Slackware doesn't have a non-buggy (and not just for programmers) one! (KDE Advanced Text Editor has broken tab bars since at most KDE3, but nothing else is almost as good.)

Any my SlackBuilds (apart from games, unless you want Rogue from later Unix/BSD games packages, would be cool...) would be neat to see (re)taken over by Slackware (then I could do other builds) but there's question how useful they are for oldschool users & programmers/scientists/system-administrators, and size (MB, package list KB.)

mine (on SlackBuilds.org, SBo) include
  • GRX (graphics library for pure tty framebuffer & SVGAlib, X Window System, X. I maintain v2, but they haven't ported v3 well yet)
  • KuickShow (was part of Slackware's KDE3, still works w/KDE4 to 5)
  • Mathics (w/dependencies, subset of top mathematics software, Wolfram Mathematica)
There's much software I want to add/compile/try, can't think of it all, but here are a few just thinking of recently after arising (so should update or post again.)
  • anything from official Slackware team's unofficial (I call 'quasi-official') builds
  • anything from SBo 'system' (especially tty-/*sh-specific, archiving/compression, fonts,) 'development,' 'academic,' 'office,' 'graphics' (GIMP plugins? Karbon? Krita? Inkscape? MyPaint? MtPaint? Pinta? Scribus?,) 'audio,' (some) 'misc'
  • any more asm or C or tty/*sh tools (hex editors, libraries, *sh cool utilities/hacks & conveniences.)
  • any more classic Unix (*BSD/'openSolaris') or similar GNU software
  • any fully type-safe language (not merely type-strong) (Swift? SBo.)
  • any additional KDE stuff than usually included
  • fpc (SBo. haven't programmed for it, but sounds fun/classic.)
  • FreeBASIC (and/or QB64, whatever. SBo. I'll be criticized, but anyone my age might've used such, and criticize Python, similarly...)
  • Grafx2 (SBo. command-line graphics editor.)
  • KDE preconfiguration option to disable users' files' auto-indexing (as is, crashing my new installations since '00s/KDE4, because of multi-GB/-TB '/home,' so many times) (would be nice if once per OS installation/usage, it emailed KDE.org we're opting out, requesting end, or only opt-in setup, for critically horrible default.)
  • JDK installer (or also OpenJDK) & NetBeans (SBo) (yes, I know one must accept license, but a good installer could run one through.)
  • MonoDevelop, etc.
  • PROLOG (and/or any mathematics/logic software... more TeX editors?)
  • Win3-/CDE-type X program manager, i.e., icon group boxes (still with Apple-type or Win3-type launcher panel, taskbar, system tray, optional style)
  • Xen (on SBo. Must work alongside KVM.)
  • ZFS (on SBo. Don't currently use it, but had NetBSD & Slackware shared '/home'... I understand if there are licensing issues. Installer maybe?)
stuff I'd like to see people build for Slackware 14.2 to 15 SBo (some too advanced for me, some lower-priority or I have no attachment to being the one doing these.)
  • BOINC
  • Buzztrax (tracker-synthesizer like 1990s' Buzz.)
  • Caffe (for some advanced black & white image colorizer artificial intelligence.)
  • CDE (even though not so great and I use KDE4, dislike that and all the rest.)
  • MLVWM (if can be upgraded from 32-bit)
  • Mycroft.AI
  • TDE (KDE3.5 fork... still dislike KDE4+)

Daedra 03-18-2018 11:07 PM
Another friendly yearly request. Please remove compiz from the tree so that SBo can take it over and maintain it along with all the companion packages. Or update the very out dated 0.8.8 version to the compiz-reloaded version of 0.8.14.

Thank you

atelszewski 03-19-2018 12:49 AM
Hi,

Are we going to have XDG Base Directory Specification implemented?

Is $XDG_RUNTIME_DIR support implementation hard to achieve?

--
Best regards,
Andrzej Telszewski

Didier Spaier 03-19-2018 01:07 AM
Quote:
Originally Posted by atelszewski (Post 5832569)
Hi,

Are we going to have XDG Base Directory Specification implemented?

Is $XDG_RUNTIME_DIR support implementation hard to achieve?

--
Best regards,
Andrzej Telszewski
You can already have it as a user setting:
Code:
didier[~]$ cat .profile
export PAGER=/usr/bin/most
export GROFF_ENCODING=UTF-8
export XDG_CACHE_HOME=/dev/shm/$(whoami)
mkdir -p /dev/shm/$(whoami)
chmod 700 /dev/shm/$(whoami)
export XDG_RUNTIME_DIR=$XDG_CACHE_HOME
didier[~]$
To make this a system default just write this as /etc/skel/.profile

However I doubt that it be "the Slackware way" to populate /etc/skel at the distribution level, traditionally it has been up to the administrator, with a very minor exception if I remember correctly.

atelszewski 03-19-2018 01:13 AM
Hi,

Have a look at specs for $XDG_RUNTIME_DIR.
It's not that obvious.

--
Best regards,
Andrzej Telszewski

gmgf 03-19-2018 03:52 AM
poppler-0.63.0:

https://poppler.freedesktop.org/
https://poppler.freedesktop.org/poppler-0.63.0.tar.xz

Didier Spaier 03-19-2018 05:40 AM
1 Attachment(s)
Quote:
Originally Posted by atelszewski (Post 5832575)
Have a look at specs for $XDG_RUNTIME_DIR.
It's not that obvious.
Well, let's quote the relevant part of the XDG Base Directory Specification:
Quote:
$XDG_CACHE_HOME defines the base directory relative to which user specific non-essential data files should be stored. If $XDG_CACHE_HOME is either not set or empty, a default equal to $HOME/.cache should be used.

$XDG_RUNTIME_DIR defines the base directory relative to which user-specific non-essential runtime files and other file objects (such as sockets, named pipes, ...) should be stored. The directory MUST be owned by the user, and he MUST be the only one having read and write access to it. Its Unix access mode MUST be 0700.

The lifetime of the directory MUST be bound to the user being logged in. It MUST be created when the user first logs in and if the user fully logs out the directory MUST be removed. If the user logs in more than once he should get pointed to the same directory, and it is mandatory that the directory continues to exist from his first login to his last logout on the system, and not removed in between. Files in the directory MUST not survive reboot or a full logout/login cycle.

The directory MUST be on a local file system and not shared with any other system. The directory MUST by fully-featured by the standards of the operating system. More specifically, on Unix-like operating systems AF_UNIX sockets, symbolic links, hard links, proper permissions, file locking, sparse files, memory mapping, file change notifications, a reliable hard link count must be supported, and no restrictions on the file name character set should be imposed. Files in this directory MAY be subjected to periodic clean-up. To ensure that your files are not removed, they should have their access time timestamp modified at least once every 6 hours of monotonic time or the 'sticky' bit should be set on the file.

If $XDG_RUNTIME_DIR is not set applications should fall back to a replacement directory with similar capabilities and print a warning message. Applications should use this directory for communication and synchronization purposes and should not place larger files in it, since it might reside in runtime memory and cannot necessarily be swapped out to disk
It seems me that the settings I adopted comply to this, but maybe I am wrong?

At least I didn't encounter any issue so far that I could link to these settings.

I set XDG_RUNTIME_DIR to be the same as XDG_CACHE_HOME because:
  1. It's allowed by the specification, if I understand it correctly
  2. Some apps use the latter for the same aims as others use the former. When that occurs this setting eases inter processes communication.
Let's take the example of the current state of /dev/shm/didier (tree attached). In it you find sockets for at-spi2 and speech-dispatcher. This helps orca, at-spi2 and speech-dispatcher communicating together, thus I can hear orca spell all that I am currently typing, with the help of espeak-ng and alsa.

Additionally as /dev/shm/didier is in RAM, after a reboot all caches will have been emptied, including those of Firefox and Thuderbird for instance. Auto-cleaning, so to speak.

PS. Currently:
Code:
didier[~]$ du -sh /dev/shm/didier
12M        /dev/shm/didier
didier[~]$

DarrenDrapkin 03-19-2018 05:56 AM
amarok with an mp3 library

orbea 03-19-2018 06:25 AM
Quote:
Originally Posted by Daedra (Post 5832558)
Another friendly yearly request. Please remove compiz from the tree so that SBo can take it over and maintain it along with all the companion packages. Or update the very out dated 0.8.8 version to the compiz-reloaded version of 0.8.14.

Thank you
The current compiz works well here, what would be improved in the new version?

Didier Spaier 03-19-2018 06:35 AM
Quote:
Originally Posted by orbea (Post 5832638)
The current compiz works well here, what would be improved in the new version?
Here I use compiz-0.9.13.1.r4138 which works pretty well and has the advantage that all features are in the same package.

Source directory:
http://slackware.uk/slint/x86_64/sli...source/compiz/

Alternatively you'd need all this, which is older:
http://slackware.uk/slint/x86_64/sli...ce/compiz-set/

atelszewski 03-19-2018 06:39 AM
Hi,

Quote:
Originally Posted by Didier Spaier (Post 5832629)
It seems me that the settings I adopted comply to this, but maybe I am wrong?
The hardest part:
Quote:
The lifetime of the directory MUST be bound to the user being logged in. It MUST be created when the user first logs in and if the user fully logs out the directory MUST be removed. If the user logs in more than once he should get pointed to the same directory, and it is mandatory that the directory continues to exist from his first login to his last logout on the system, and not removed in between. Files in the directory MUST not survive reboot or a full logout/login cycle.
--
Best regards,
Andrzej Telszewski

orbea 03-19-2018 06:53 AM
Quote:
Originally Posted by Didier Spaier (Post 5832639)
Here I use compiz-0.9.13.1.r4138 which works pretty well and has the advantage that all features are in the same package.
I'm a bit ignorant. Would you mind elaborating on what features are missing?

My personal limited use case for compiz is to stop screen tearing with some wine games that have a broken native vsync when not using a window manager (Where I would use compton instead).

Didier Spaier 03-19-2018 07:30 AM
1 Attachment(s)
Quote:
Originally Posted by orbea (Post 5832652)
Would you mind elaborating on what features are missing?
All features shipped in:
ccsm
compiz-bcop
compiz-plugins-main
compizconfig-python
libcompizconfig

plus extra plugins.

Attached is a pic of compiz running in MATE and the CCSM

As an aside in Mate mate-tweaks allows to easy switch the window manager, here choosing between marco, metacity and compiz, the formers with or whithout compositor.

Didier Spaier 03-19-2018 07:57 AM
Quote:
Originally Posted by atelszewski (Post 5832642)
The hardest part:
Quote:
The lifetime of the directory MUST be bound to the user being logged in. It MUST be created when the user first logs in and if the user fully logs out the directory MUST be removed. If the user logs in more than once he should get pointed to the same directory, and it is mandatory that the directory continues to exist from his first login to his last logout on the system, and not removed in between. Files in the directory MUST not survive reboot or a full logout/login cycle.
I just checked, both without and with a display manager (in runlevels 3 and 4, in Slackware parlance).
  • After a reboot there is no directory /dev/shm/didier.
  • It is created as soon as I log in as didier.
  • Its content is preserved after I log out.
  • Its content is preserved after a new log in.
It seems me that it is exactly the specified behavior.

Of course if you never reboot you'll have to clean the house yourself but this is a desktop specification, not a server one.

orbea 03-19-2018 08:13 AM
@Didier Spaier Thanks for the details, but that doesn't really help me understand the use cases that aren't possible with the older compiz. :)

Also I don't think compiz would need to be removed for someone to update it at slackbuilds.org. My understanding is that it would be fine as long as the info file contains REQUIRES="%README%" and the README explicitly explains that it conflicts with the compiz package included in the main tree. You can look at pkgconf for a example which conflicts with and replaces pkg-config in the main tree.

Didier Spaier 03-19-2018 08:20 AM
Quote:
Originally Posted by orbea (Post 5832687)
@Didier Spaier Thanks for the details, but that doesn't really help me understand the use cases that aren't possible with the older compiz.
I am not sure I can give a good response. I have been told that the new one is more accessible to visually impaired users using the Orca screen reader or a Braille device but am not able to elaborate on that right now.

I will just add that as a packages provider it's less work for me to maintain just one package ;)

atelszewski 03-19-2018 08:31 AM
Hi,

Quote:
Originally Posted by Didier Spaier (Post 5832677)
  • Its content is preserved after I log out.
versus

Quote:
It MUST be created when the user first logs in and if the user fully logs out the directory MUST be removed.
--
Best regards,
Andrzej Telszewski

Didier Spaier 03-19-2018 09:17 AM
Well, I don't know what exactly is a "full log out".

Does it means that the user is no more recorded in /var/run/utmp? But not all apps write this file, and after having logged out and killed all processes owned by didier with "killall -u didier", "users" only output "root" but "grep didier /var/run/utmp" still matches.

In my understanding "fully log out" should practically mean that there is no remaining process owned by didier.

One could check that periodically in a cron job, that would then remove or propose to remove /dev/shm/didier and its content. However I think that would rather be the admin's duty to write it than done by something shipped in Slackware. I could be wrong of course, that's just an opinion and I am ready to hear other ones.

kjhambrick 03-19-2018 11:10 AM
Quote:
Originally Posted by Didier Spaier (Post 5832719)
Well, I don't know what exactly is a "full log out".

Does it means that the user is no more recorded in /var/run/utmp? But not all apps write this file, and after having logged out and killed all processes owned by didier with "killall -u didier", "users" only output "root" but "grep didier /var/run/utmp" still matches.

In my understanding "fully log out" should practically mean that there is no remaining process owned by didier.

One could check that periodically in a cron job, that would then remove or propose to remove /dev/shm/didier and its content. However I think that would rather be the admin's duty to write it than done by something shipped in Slackware. I could be wrong of course, that's just an opinion and I am ready to hear other ones.
Hoo Boy ...

That is a tough question to answer, let alone being a tough programming nut to crack :(

There was quite an uproar a while back when that infamous new init program changed the long standing default behavior of *IX ... killed background processes after user logs out

The XDG Base Directory Specification requirement that you and atelszewski posted is pretty clear ( It MUST be created when the user first logs in and if the user fully logs out the directory MUST be removed. )

But as you said, the hard part is what does 'fully logged out' really mean ?

ick.

-- kjh

atelszewski 03-19-2018 12:02 PM
Hi,

In my opinion, this happens when the last process, for the particular user, exits.

--
Best regards,
Andrzej Telszewski

gmgf 03-19-2018 12:23 PM
gparted-0.31.0:

https://sourceforge.net/projects/gpa...parted-0.31.0/
https://sourceforge.net/projects/gpa...ar.gz/download

Didier Spaier 03-19-2018 01:46 PM
Quote:
Originally Posted by atelszewski (Post 5832785)
In my opinion, this happens when the last process, for the particular user, exits.
Which could very well occur no sooner than when the machine is rebooted...

Often when I type userdel for a not logged in user, It fails because there are still running processes started on behalf of this user.


All times are GMT -5. The time now is 10:48 PM.
Page 14 of 197 « First 41213 14 15162464114 Last »
Show 50 post(s) from this thread on one page