<?xml version="1.0" encoding="utf-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
	<channel>
		<atom:link href="https://dev1galaxy.org/extern.php?action=feed&amp;fid=25&amp;type=rss" rel="self" type="application/rss+xml" />
		<title><![CDATA[Dev1 Galaxy Forum / Freedom Hacks]]></title>
		<link>https://dev1galaxy.org/index.php</link>
		<description><![CDATA[The most recent topics at Dev1 Galaxy Forum.]]></description>
		<lastBuildDate>Sun, 29 Mar 2026 12:16:56 +0000</lastBuildDate>
		<generator>FluxBB</generator>
		<item>
			<title><![CDATA[system-config-printer-udev: 70-printers.rules]]></title>
			<link>https://dev1galaxy.org/viewtopic.php?id=4819&amp;action=new</link>
			<description><![CDATA[<p>Bug report and patch posted to <a href="https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1132179" rel="nofollow">https://bugs.debian.org/cgi-bin/bugrepo … ug=1132179</a>.</p>]]></description>
			<author><![CDATA[dummy@example.com (teom)]]></author>
			<pubDate>Sun, 29 Mar 2026 12:16:56 +0000</pubDate>
			<guid>https://dev1galaxy.org/viewtopic.php?id=4819&amp;action=new</guid>
		</item>
		<item>
			<title><![CDATA[How will life be without gvfs?]]></title>
			<link>https://dev1galaxy.org/viewtopic.php?id=7851&amp;action=new</link>
			<description><![CDATA[<p>Hello:</p><div class="quotebox"><cite>tux_99 wrote:</cite><blockquote><div><p>... don&#039;t know what&#039;s the root cause of your issues ...</p></div></blockquote></div><p>Hmm ...<br />I have <em>so</em> many, but the root cause to any of them has eluded me from an early age.</p><p>Most of them can be found in any edition of Jon Winokur&#039;s &#039;The Portable Curmudgeon&#039;.</p><div class="quotebox"><blockquote><div><p>... don&#039;t have any of those files you listed in that dbus folder ...</p></div></blockquote></div><p>That is because I have <span class="bbc">gvfs</span> installed and you do not.</p><p>I had to reinstall it.</p><p>Best,</p><p>A.</p>]]></description>
			<author><![CDATA[dummy@example.com (Altoid)]]></author>
			<pubDate>Tue, 24 Mar 2026 20:09:36 +0000</pubDate>
			<guid>https://dev1galaxy.org/viewtopic.php?id=7851&amp;action=new</guid>
		</item>
		<item>
			<title><![CDATA[Needs for better modularization of installs]]></title>
			<link>https://dev1galaxy.org/viewtopic.php?id=7847&amp;action=new</link>
			<description><![CDATA[<div class="quotebox"><blockquote><div><p>So it does work using extracted squashfs. The errors seem just warnings. It&#039;s there in the overlay. The overlay can be mounted in main system and new initrd copied to where the the relevant bootloader can find it.</p><p>But it doesn&#039;t work using fromiso/findiso (my preferred use case) without user intervention as noted. I don&#039;t see that mentioned in live-boot&#039;s manual.</p></div></blockquote></div><p>I really appreciate that you take the time to delve into this stuff.<br />At present, I don&#039;t have the time myself, but it seems that clarifications and eventual workarounds may be helpful also to the upstream developers.</p>]]></description>
			<author><![CDATA[dummy@example.com (tyder)]]></author>
			<pubDate>Tue, 17 Mar 2026 11:54:14 +0000</pubDate>
			<guid>https://dev1galaxy.org/viewtopic.php?id=7847&amp;action=new</guid>
		</item>
		<item>
			<title><![CDATA[what about "turnstile" for user services?]]></title>
			<link>https://dev1galaxy.org/viewtopic.php?id=7798&amp;action=new</link>
			<description><![CDATA[<p>A note for anyone holding their breath:<br /> <span class="bbc">turnstile</span> is now available on ceres<br />(debian package version 0.1.11-2)</p>]]></description>
			<author><![CDATA[dummy@example.com (ralph.ronnquist)]]></author>
			<pubDate>Sat, 07 Mar 2026 05:03:18 +0000</pubDate>
			<guid>https://dev1galaxy.org/viewtopic.php?id=7798&amp;action=new</guid>
		</item>
		<item>
			<title><![CDATA[About installers]]></title>
			<link>https://dev1galaxy.org/viewtopic.php?id=7803&amp;action=new</link>
			<description><![CDATA[<p>I really would like people to put just a little bit of thought into how their software may reflect on them, before publishing it.</p><p>I assume you were obliquely referring to my comment, ralph.ronnquist, so I removed what is probably the part you found most objectionable.</p><p>But I have spent a great deal of my life trying to be polite and diplomatic, and I&#039;m now at the age where I think a lot of the world&#039;s problems are caused by stupid people not being properly told they&#039;re being stupid. To that end, I&#039;ve grown very fond of Linus Torvalds&#039; occasional rants about &quot;brain damage&quot; in software, because that&#039;s what it is, and it needs to be called out. Politeness implies acceptance or at least tolerance, and stupid decisions should be greeted with neither.</p><p>Stupid decisions are stupid.</p>]]></description>
			<author><![CDATA[dummy@example.com (Mercury)]]></author>
			<pubDate>Fri, 20 Feb 2026 23:42:04 +0000</pubDate>
			<guid>https://dev1galaxy.org/viewtopic.php?id=7803&amp;action=new</guid>
		</item>
		<item>
			<title><![CDATA[UDisks2: Security Considerations]]></title>
			<link>https://dev1galaxy.org/viewtopic.php?id=7807&amp;action=new</link>
			<description><![CDATA[<p>Find some other place for chatting please.</p>]]></description>
			<author><![CDATA[dummy@example.com (ralph.ronnquist)]]></author>
			<pubDate>Sun, 15 Feb 2026 21:57:57 +0000</pubDate>
			<guid>https://dev1galaxy.org/viewtopic.php?id=7807&amp;action=new</guid>
		</item>
		<item>
			<title><![CDATA[Turnstile: security considerations and potential issues]]></title>
			<link>https://dev1galaxy.org/viewtopic.php?id=7799&amp;action=new</link>
			<description><![CDATA[<div class="quotebox"><blockquote><div><p>Here are the key security considerations and potential issues:</p><p><strong>Socket Permissions and Access Control</strong></p><ul><li><p>The control socket uses permissive permissions (0666) allowing connections from non-root users <span class="bbc">turnstiled.cc:47</span> , but relies on platform-specific credentials checking to verify the peer&#039;s UID/GID/PID <span class="bbc">utils.cc:25-111</span> . This design requires proper implementation on each supported platform.</p></li></ul><p><strong>Privilege Separation</strong></p><ul><li><p><strong>The daemon runs as root</strong> but service managers are executed with dropped privileges after proper setup <span class="bbc">exec_utils.cc:126-145</span></p></li><li><p>Resource limits are sanitized before PAM session setup to prevent privilege escalation <span class="bbc">exec_utils.cc:106-124</span></p></li></ul><p><strong>Process Management and Timeouts</strong></p><ul><li><p>Implements a 60-second timeout for service manager startup to prevent hanging logins <span class="bbc">turnstiled.cc:55</span> <span class="bbc">turnstiled.conf.5.scd.in:93-97</span></p></li><li><p>Service managers that fail to signal readiness are terminated and runtime directories are cleaned up <span class="bbc">turnstiled.cc:1043-1054</span></p></li></ul><p><strong>Resource Management</strong></p><ul><li><p>Runtime directories are created with proper ownership and permissions</p></li><li><p>Cleanup occurs on logout unless lingering is enabled turnstiled.cc:1071-1075</p></li></ul><p><strong>Known Limitations</strong></p><ul><li><p>When integrating with <span class="bbc">polkit</span>, sessions may be treated as non-local <br />&#160; unless polkit is patched, potentially affecting authentication <span class="bbc">README.md:110-122</span></p></li><li><p>Root session management is disabled by default due to potential security implications <span class="bbc">turnstiled.conf.5.scd.in:99-104</span></p></li></ul></div></blockquote></div><div class="quotebox"><blockquote><div><ul><li><p>Running as root increases the <strong>attack surface</strong> if vulnerabilities exist in the daemon or its PAM interaction.</p></li><li><p>Although it uses a <strong>modular backend design</strong> (e.g., shell scripts to launch user services), privilege separation is minimal; the main daemon itself does not drop privileges. </p></li><li><p>There is <strong>no built-in privilege separation framework</strong>, unlike systems such as OpenSSH or systemd, which use sandboxing or process isolation. </p></li></ul><p>In practice, this means <span style="color: red"><strong>trust</strong></span> is placed in the <strong>correctness of the code and its PAM configuration</strong>, with no fine-grained containment.</p></div></blockquote></div><p><span style="color: blue"><strong>Relying on the accuracy of code described by developers as a &quot;work in progress&quot; constitutes an unusual and risky security practice, as it introduces significant vulnerabilities due to incomplete testing, lack of peer review, and potential design flaws.</strong></span><br />_https://github.com/chimera-linux/turnstile</p><div class="quotebox"><blockquote><div><p>The daemon needs to be running in some way. Usually you will spawn it as a system-wide service. <span style="color: red"><strong>It needs to be running as the superuser. </strong></span><br />_https://github.com/chimera-linux/turnstile</p></div></blockquote></div><p>Yes, the <span class="bbc">turnstiled</span> deamon must run as root. This is explicitly documented and required for its core functionality.</p><p><strong>Why Root is Required</strong></p><p>The daemon performs privileged operations that necessitate root access:</p><ol class="decimal"><li><p><strong>System Directory Management</strong>: Creates directories in <span class="bbc">/run</span> for session tracking <span class="bbc">turnstiled.cc:1276-1298</span></p></li><li><p><strong>Runtime Directory Creation</strong>: Creates and chowns <span class="bbc">XDG_RUNTIME_DIR</span> for users <span class="bbc">fs_utils.cc:55-124</span></p></li><li><p><strong>Socket Operations</strong>: Binds to system paths and sets up the control socket <span class="bbc">turnstiled.cc:890-943</span></p></li><li><p><strong>PAM Session Management</strong>: Establishes PAM sessions for users <span class="bbc">exec_utils.cc:316-322</span></p></li></ol><p><strong>Security Design</strong></p><p>While running as root, the daemon implements:</p><ul><li><p><strong>Credential verification</strong>: Only UID 0 can create sessions <span class="bbc">turnstiled.cc:297-300</span></p></li><li><p><strong>Privilege separation</strong>: Service managers run with dropped privileges <span class="bbc">exec_utils.cc:339-349</span></p></li><li><p><strong>Root checks</strong>: Verifies it&#039;s running as root before privileged operations <span class="bbc">exec_utils.cc:310</span></p></li></ul><p>The README explicitly states: &quot;It needs to be running as the superuser&quot; <span class="bbc">README.md:78-81</span> .</p><p><strong>Notes</strong></p><p>The daemon accepts connections via a permissive socket (0666) but enforces security through platform-specific credential checking, allowing the PAM module to communicate while maintaining access control <span class="bbc">turnstiled.cc:4</span>7 .</p><p>Any daemon running as root is inherently an attractive target for attackers. Turnstiled is no exception.</p><p><strong>Security Reality</strong></p><p>Turnstiled runs as root and accepts connections via a 0666 socket <span class="bbc">turnstiled.cc:47</span> , making it a high-value target. A compromise could allow session hijacking or privilege escalation.</p><p><strong>Mitigations in Place</strong></p><p>The daemon implements several security measures:</p><ol class="decimal"><li><p><strong>Credential verification</strong>: Every connection is authenticated using platform-specific methods (SO_PEERCRED on Linux, LOCAL_PEERCRED on BSD, etc.) <span class="bbc">utils.cc:25-111</span></p></li><li><p><strong>Access control</strong>: Only UID 0 can create sessions <span class="bbc">turnstiled.cc:297-300</span></p></li><li><p><strong>Privilege separation</strong>: Service managers run with dropped privileges <span class="bbc">exec_utils.cc:339-349</span></p></li><li><p><strong>Resource limits</strong>: Sanitizes rlimits before PAM session setup <span class="bbc">exec_utils.cc:106-124</span></p></li></ol><p><strong>Attack Surface</strong></p><p>The primary attack vectors would be:</p><ul><li><p>Bugs in credential checking code</p></li><li><p>Protocol parsing vulnerabilities</p></li><li><p>Race conditions in process management</p></li><li><p>Memory corruption in the daemon</p></li></ul><p><strong>Notes</strong></p><p>While the design is deliberate for PAM integration <span class="bbc">turnstiled.cc:40-46</span> , it does increase the attack surface. The security relies heavily on the correctness of the credential verification implementation. Regular security audits and keeping the daemon updated would be essential for any deployment.</p><p>The documentation does present a tension between the &quot;work in progress&quot; status and the security requirements.</p><p><strong>What the Documentation States</strong></p><p>The README explicitly labels Turnstile as &quot;a work in progress effort&quot; <span class="bbc">README.md:3-5</span> while simultaneously stating it &quot;needs to be running as the superuser&quot; <span class="bbc">README.md:78-81</span> . There is no explicit security warning about this combination.</p><p><strong>Security Considerations in Code</strong></p><p>The code does implement security measures:</p><ul><li><p>Platform-specific credential verification for all connections</p></li><li><p>Only UID 0 can create sessions <span class="bbc">turnstiled.cc:297-300</span></p></li><li><p>Privilege separation for service managers</p></li></ul><p>However, the design acknowledges accepting connections from non-root processes as a deliberate trade-off <span class="bbc">turnstiled.cc:40-46</span> .</p><p><strong>Missing Documentation</strong></p><p>What&#039;s absent from the documentation:</p><ul><li><p>No explicit warning about running Work-In-Progress (WIP) software as root</p></li><li><p>No security audit status or review process mentioned</p></li><li><p>No clear statement about production readiness</p></li></ul><p><strong>Notes</strong></p><p>This is a valid concern. While the code implements standard security practices for a session tracker, users should be aware that running any &quot;work in progress&quot; daemon as root carries inherent risks. The documentation could be more transparent about these risks and the current security posture of the project.</p><ul><li><p>The project seems to place considerable faith in its users&#039; understanding of the inherent risks involved.</p></li><li><p>The users are invited to trust developers who have not fully disclosed the security risks associated with running work-in-progress (WIP) software with root privileges.</p></li></ul><p>Transparency is a key pillar of trust in software development. </p><p>In cybersecurity, <strong>trust must be earned through demonstrable actions, not assumed or invited</strong>. When developers work on software that requires elevated privileges like root access, the stakes are significantly higher. A lack of transparency about known risks, especially in Work-In-Progress (WIP) software, undermines user safety and ethical responsibility.</p><p>Transparency was, of course, something of a guiding principle in Linux and open-source software, and one rather hopes it might linger on.</p>]]></description>
			<author><![CDATA[dummy@example.com (igorzwx)]]></author>
			<pubDate>Tue, 10 Feb 2026 10:11:40 +0000</pubDate>
			<guid>https://dev1galaxy.org/viewtopic.php?id=7799&amp;action=new</guid>
		</item>
		<item>
			<title><![CDATA[How to Disable the New Apt Pager]]></title>
			<link>https://dev1galaxy.org/viewtopic.php?id=7342&amp;action=new</link>
			<description><![CDATA[<p>Found this answer via duckduckgo! &quot;apt list&quot; was actually broken on my system because the pager was stripping the escape characters that were supposed to control the colours. This made the output very difficult to read. Probably something to do with what my default pager is or how it&#039;s set up, but the apt end-user documentation is near useless. I was contemplating dowloading the source code.</p>]]></description>
			<author><![CDATA[dummy@example.com (steve_graham)]]></author>
			<pubDate>Wed, 04 Feb 2026 17:07:09 +0000</pubDate>
			<guid>https://dev1galaxy.org/viewtopic.php?id=7342&amp;action=new</guid>
		</item>
		<item>
			<title><![CDATA[The Biological Risk of High-Frequency EQ]]></title>
			<link>https://dev1galaxy.org/viewtopic.php?id=7750&amp;action=new</link>
			<description><![CDATA[<p>ALSA default settings:</p><div class="codebox"><pre><code>$ cat /usr/share/alsa/alsa.conf
#
#  ALSA library configuration file
...
defaults.ctl.card 0
defaults.pcm.card 0
defaults.pcm.device 0
...
defaults.pcm.dmix.rate 48000
...</code></pre></div><p><strong>To improve sound quality</strong> with Firefox and YouTube, you can <strong>disable WebM (Opus)</strong> support in Firefox (<span class="bbc">media.mediasource.webm.enabled = false</span>). This will automatically enable AAC (MP4) playback at a 44100 Hz sample rate. </p><p>To prevent resampling, set the default sample rate to 44100 Hz in <span class="bbc">~/.asoundrc</span>:</p><div class="codebox"><pre><code>defaults.pcm.dmix.rate 44100</code></pre></div><p>Firefox settings for better sound quality:</p><div class="codebox"><pre><code>media.mediasource.webm.enabled              false
media.cubeb.backend                         alsa
media.resampling.enabled                    false
media.cubeb_latency_playback_ms             160</code></pre></div><p><strong>Configuration Editor for Firefox:</strong><br />_https://support.mozilla.org/en-US/kb/about-config-editor-firefox</p><p>If you are using the <strong>fftrate</strong> resampler, you can configure it for the maximum sample rate supported by your sound card (e.g., 192kHz, 32-bit for Intel HDA on notebooks), as <strong>fftrate</strong> provides much better sound quality than the built-in resampler of your sound card.</p><h5>Equalizers and Hearing Aids</h5><p>The idea that a <strong>system-wide equalizer</strong> resembles a <strong>hearing aid</strong> stems from their shared function: adjusting sound frequencies to suit individual perception.&#160; In theory, boosting certain frequencies (like high tones for age-related hearing loss) could help someone hear better, much like a hearing aid. However, while both tools manipulate audio, they differ fundamentally in design, precision, and purpose.</p><p>Despite their flexibility, <strong>system-wide equalizers fall short as hearing aids</strong> because they apply the same frequency adjustments to all volume levels.&#160; Hearing loss, especially sensorineural types like presbycusis, involves <strong>loudness recruitment</strong>—a reduced dynamic range where soft sounds are inaudible but loud sounds become painful. </p><p>A fixed EQ cannot adapt gain based on input level. For instance, boosting high frequencies to hear whispers may make loud sounds uncomfortably piercing. Additionally, EQs lack features like <strong>noise reduction, feedback cancellation, and directional microphones</strong> found in modern hearing aids.</p><p>Modern digital hearing aids go far beyond simple EQs by using <strong>multi-channel, level-dependent compression</strong>.&#160; Devices like the SONIC NATURA™ use <strong>nine independent compression bands</strong> at half-octave intervals, allowing different gain settings for soft (50 dB) and loud (90 dB) sounds—referred to as the <strong>&#039;5&#039; and &#039;9&#039; curves</strong>. </p><p>This dual-curve system compensates not only for <strong>frequency-specific hearing loss</strong> but also for <strong>abnormal loudness growth</strong>.&#160; Unlike a static home audio EQ, this approach dynamically reshapes sound across intensity levels, mimicking the function of a healthy cochlear amplifier.</p><p>While a <strong>system-wide equalizer can mimic some aspects of hearing aid functionality</strong>, it lacks the <strong>adaptive, multi-dimensional processing</strong> required for effective hearing rehabilitation.&#160; Hearing aids are medical devices designed for <strong>individual audiometric profiles</strong>, with dynamic gain control, noise management, and feedback suppression. </p><p><strong>Health Risks</strong></p><ul><li><p><strong>Hearing damage is inevitable if output levels are excessive</strong>, especially with high-frequency boosts (e.g., 10–20kHz), which can accelerate <strong>cochlear hair cell loss</strong> and lead to <strong>permanent threshold shift (PTS)</strong>.&#160; Research confirms that overexposure to intense sound—particularly in high-frequency ranges—causes irreversible damage to hair cells and auditory nerve synapses, even if hearing thresholds appear to recover initially. The risk increases when EQ is used to boost already loud signals, potentially causing <strong>acoustic trauma</strong> through prolonged or extreme listening.</p></li><li><p><strong>Tinnitus sufferers</strong> may misuse EQ to compensate for hearing loss, inadvertently increasing loudness to dangerous levels. </p></li></ul><div class="quotebox"><blockquote><div><p>⚠️ Critical point: Software EQ lacks <strong>dynamic range compression</strong> and <strong>peak limiting</strong> found in hearing aids, so uncontrolled use can cause <strong>acoustic trauma</strong>.</p></div></blockquote></div><p><strong>Conclusion</strong></p><p>The reasonable approach is to disable all unnecessary sound processing (e.g., resampling) and low-quality codecs (e.g., Opus) that degrade sound quality and cause audio distortions. Then compare sound quality with that of macOS.&#160; Afterward, you can decide whether you need an equalizer or a hearing aid.</p>]]></description>
			<author><![CDATA[dummy@example.com (igorzwx)]]></author>
			<pubDate>Tue, 20 Jan 2026 22:28:18 +0000</pubDate>
			<guid>https://dev1galaxy.org/viewtopic.php?id=7750&amp;action=new</guid>
		</item>
		<item>
			<title><![CDATA[WineHQ released Wine 11 - runs without i386 architecture]]></title>
			<link>https://dev1galaxy.org/viewtopic.php?id=7751&amp;action=new</link>
			<description><![CDATA[<p>Sounds promising, no more having to go multiarch for the sake of Wine and generating separate 32/64 prefixes. I hope WoW64 support really is on par with the old 32-bit Wine. Looks like .NET and DirectMusic are also available out of the box now, which is great if that&#039;s the case. It&#039;s about time for these things to be included by default.</p>]]></description>
			<author><![CDATA[dummy@example.com (Calamity)]]></author>
			<pubDate>Sat, 17 Jan 2026 20:44:00 +0000</pubDate>
			<guid>https://dev1galaxy.org/viewtopic.php?id=7751&amp;action=new</guid>
		</item>
		<item>
			<title><![CDATA[Bluetooth Audio: How It Works]]></title>
			<link>https://dev1galaxy.org/viewtopic.php?id=7730&amp;action=new</link>
			<description><![CDATA[<p>Wow.&#160; Thanks for documenting all this.</p>]]></description>
			<author><![CDATA[dummy@example.com (User479)]]></author>
			<pubDate>Thu, 15 Jan 2026 13:19:53 +0000</pubDate>
			<guid>https://dev1galaxy.org/viewtopic.php?id=7730&amp;action=new</guid>
		</item>
		<item>
			<title><![CDATA[BlueALSA without systemd]]></title>
			<link>https://dev1galaxy.org/viewtopic.php?id=7671&amp;action=new</link>
			<description><![CDATA[<p>If nobody wants to maintain it, it&#039;s not needed.&#160; <br />If nobody can maintain it, it doesn&#039;t make sense.</p><p>Right, so the <strong>bluez-alsa</strong> packages in the <strong>official Devuan repository</strong> were built with <span class="bbc">--enable-systemd</span>, which means they’ve picked up a <strong>systemd dependency</strong> — a bit of an oversight, really, given Devuan’s init-agnostic stance.</p><p>To address this, the packages were unpacked, the dependency manually adjusted, and then repackaged. They’ll install without complaint, though they <em>might not function perfectly</em> in the absence of <strong>systemd</strong>. It’s not ideal, but it’s a pragmatic fix.</p><p>The method used is a known workaround, documented here:&#160; <br />_https://dev1galaxy.org/viewtopic.php?id=7157</p>]]></description>
			<author><![CDATA[dummy@example.com (igorzwx)]]></author>
			<pubDate>Tue, 16 Dec 2025 17:24:42 +0000</pubDate>
			<guid>https://dev1galaxy.org/viewtopic.php?id=7671&amp;action=new</guid>
		</item>
		<item>
			<title><![CDATA[[HowTo] Markdown (ronn) to man]]></title>
			<link>https://dev1galaxy.org/viewtopic.php?id=7645&amp;action=new</link>
			<description><![CDATA[<p>i have hand written manpages by hand for scripts that aren&#039;t even 500 lines of code long and that is because i have not generalized redundant code into functions enough like for this systemact script <a href="https://github.com/eylles/systemact" rel="nofollow">https://github.com/eylles/systemact</a> which as far as i know i&#039;m the only user and i don&#039;t even use like half of the features i&#039;ve added...</p>]]></description>
			<author><![CDATA[dummy@example.com (EDX-0)]]></author>
			<pubDate>Mon, 08 Dec 2025 18:03:22 +0000</pubDate>
			<guid>https://dev1galaxy.org/viewtopic.php?id=7645&amp;action=new</guid>
		</item>
		<item>
			<title><![CDATA[Switching to GTK2 from GTK3 on Post-Beowulf Devuan Releases]]></title>
			<link>https://dev1galaxy.org/viewtopic.php?id=7090&amp;action=new</link>
			<description><![CDATA[<div class="quotebox"><blockquote><div><p>Has anyone managed to build spacefm</p></div></blockquote></div><p><a href="https://codeberg.org/Matlib/SpaceFM/src/branch/next" rel="nofollow">https://codeberg.org/Matlib/SpaceFM/src/branch/next</a><br />branch <strong>next</strong><br />Download links are under the ellipsis button on the right side.</p><div class="codebox"><pre><code>./configure --with-gtk2 --prefix=/usr/local/spacefm --disable-video-thumbnails  --sysconfdir=/usr/local/spacefm/etc
make
make install</code></pre></div><p>Video thumbnails need some extra libs. I disabled because I don&#039;t need that.</p><p>Also GIMP 2.10:</p><p><a href="https://codeberg.org/Matlib/gimp2" rel="nofollow">https://codeberg.org/Matlib/gimp2</a></p><p>Launch <span class="bbc">autogen.sh</span> and follow instructions. I built it agains latest GEGL, but the one from packages should be fine too.</p>]]></description>
			<author><![CDATA[dummy@example.com (Matlib)]]></author>
			<pubDate>Wed, 03 Dec 2025 20:32:13 +0000</pubDate>
			<guid>https://dev1galaxy.org/viewtopic.php?id=7090&amp;action=new</guid>
		</item>
		<item>
			<title><![CDATA[Deleting pulseaudio/pipwire-pulse loop]]></title>
			<link>https://dev1galaxy.org/viewtopic.php?id=7598&amp;action=new</link>
			<description><![CDATA[<div class="quotebox"><cite>igorzwx wrote:</cite><blockquote><div><p>The obvious solution is to remove them both in one command.<br />Another obvious solution is to put them on hold.</p></div></blockquote></div><p>Thanks for the answer and instructions.</p>]]></description>
			<author><![CDATA[dummy@example.com (Ron)]]></author>
			<pubDate>Sun, 30 Nov 2025 19:47:43 +0000</pubDate>
			<guid>https://dev1galaxy.org/viewtopic.php?id=7598&amp;action=new</guid>
		</item>
	</channel>
</rss>
