August 2026 - wlug - list.wlug.org.nz

California Lawmakers Unanimously Pass Open-Source Exemption From Age-Verification Law
by Lawrence D'Oliveiro 31 Aug '26

31 Aug '26

The US state of California passed a bill last year <https://www.tomshardware.com/software/operating-systems/california-introduc…>, due to take effect from next year, requiring all “operating system providers” to collect (self-reported) age information from users of their platforms, and pass that on to app developers. This caused consternation among developers of open-source operating systems, who (by design) do not have the degree of control over, and insight into, their user base that owners of proprietary platforms have available. Now, the California state legislators have agreed that this is an unreasonable burden on purveyors of software “under license terms that permit a recipient to copy, redistribute, and modify” (exactly what FLOSS licences allow), so they have passed an amendment excluding these from the requirement. There are a couple of other exclusions added in this amendment, one seemingly (we hope) drawing a distinction between Linux distro repos (exempt) and Apple/Google/Microsoft-style app stores (covered), and another to do with stores only offering addons to run within host apps. Both of these seem a little less clear to me. <https://www.tomshardware.com/software/linux/california-lawmakers-unanimousl…>

1 0

Undersea cable to Antarctica through Drake Passage is viable, researchers find
by Lawrence D'Oliveiro 25 Aug '26

25 Aug '26

There is an old joke about not underestimating the bandwidth of a truckload of hard drives cruising down the motorway. Well, to the folks in Antarctica, this isn’t a joke. Their only direct connection to the Internet is via satellite, and the only other way to transfer data in bulk between stations there and the rest of the world is via suitcases full of hard drives. Well, it looks like that might change soon, after the completion of a feasibility study into running the first ever undersea cable to Antarctica has come out looking favourable <https://www.tomshardware.com/networking/researcgers-find-a-drake-passage-ca…>. Of course, there are a few ticklish political issues involved -- aren’t there always?

1 0

AliExpress was silently running audio in your browser to fingerprint and track your device
by Peter Reutemann 25 Aug '26

25 Aug '26

'As cookies become a less reliable way to track people online, AliExpress may be showing how far companies will go to fill that gap. Researchers found code on the site's homepage that ran silent audio processes in the browser. Tied to Alibaba's security systems, the scripts tap a device's own audio hardware to generate a signal and measure the tiny, device-specific ways it comes back – producing something close to a fingerprint that doesn't need a single cookie to work. It's the kind of tracking a user would likely never notice. ' -- source: https://www.techspot.com/news/113581-aliexpress-silently-running-audio-brow… Cheers, Peter

1 0

Canonical is Funding a PhD to Automate C to Rust Translation
by Peter Reutemann 24 Aug '26

24 Aug '26

'Canonical has committed funding to a three-year PhD project focused on building a system that can automatically translate large C codebases into Rust. And they are not alone; UK Research and Innovation is matching their funding for the project, which is set to run through the University of Bristol's Programming Languages Research Group. The PhD is aiming to build an all-encompassing platform that can take a C repository running into hundreds of thousands of lines and translate it into Rust that's "safe, behaviourally correct and maintainable Rust." The work will be carried out by a student, Alex Wood, who will be supported by Professor Meng Wang leading the work, with Dr. Cristina David and Canonical's Jon Seager serving as co-supervisors.' -- source: https://developers.slashdot.org/story/26/08/23/0458226 Cheers, Peter

1 0

PINE64 Halts Linux Device Production Amid DRAM and EMMC Shortages
by Peter Reutemann 20 Aug '26

20 Aug '26

'BrianFagioli writes: PINE64 has some very bad news for Linux hardware fans. PINE64 says it has no plans to produce additional Linux devices in the near future because of the ongoing DRAM and eMMC shortage. Future production will depend on component pricing after mid 2027, while existing PineNote and PineTab2 inventory could run out in roughly three months. The company says PineTime, PineVoice, and Pinecil production will continue as usual, but the outlook for its Linux hardware is much less certain. "That's it for this post, all in all it's a bit bleak for the Linux devices, but in the meantime PineStore still have a great lineup of MCU powered devices which will continue to expand with more exciting devices," the company said.' -- source: https://hardware.slashdot.org/story/26/08/19/2134209 Cheers, Peter

1 0

Ubuntu Reverts Rust cp After It Breaks Live Image Builds
by Peter Reutemann 18 Aug '26

18 Aug '26

'Ubuntu’s transition to Rust Coreutils has encountered another compatibility issue, this time affecting the essential Unix cp command. In the latest Ubuntu Foundations team update, Canonical engineer Benjamin Drung reported that the team validated fixes from Zellic’s Rust Coreutils audit against version 0.8 on Ubuntu 26.04 LTS. However, Ubuntu has reverted cp to GNU Coreutils and submitted an upstream PR for the underlying fix. The issue is tracked in Launchpad as bug #2158691 and involves Ubuntu’s coreutils-from package, which determines whether a command uses GNU Coreutils or the Rust-based uutils implementation. According to the bug report, Ubuntu briefly reinstated the Rust Coreutils version of cp in coreutils-from 0.0.0~ubuntu26. This change caused failures in livecd-rootfs, one of the packages responsible for building Ubuntu live images. The bug was labeled Critical, and the immediate solution was to restore GNU cp. The technical issue involved how uutils cp handled combinations of archive and symlink-related options, particularly cases like cp -afL. In GNU cp, the -a option is equivalent to -dR --preserve=all, which includes recursive copying and attribute preservation. The Rust implementation, however, handled some mutually exclusive flags too aggressively. When options like -a and -L were combined, the recursive behavior implied by -a could be lost, resulting in errors such as “-r not specified; omitting directory,” even though archive mode should include recursive copying.' -- source: https://linuxiac.com/ubuntu-reverts-rust-cp-after-it-breaks-live-image-buil… Cheers, Peter

2 1

Linux Kernel 7.2 Has Been Officially Released with Many New Features
by Peter Reutemann 18 Aug '26

18 Aug '26

'Linux Kernel 7.2 has just been officially released with a slew of new features, reports the blog 9to5Linux. Highlights of Linux 7.2 "include cache-aware load-balancing support, initial HDMI 2.1 FRL support to the AMDGPU driver, support for devres-based management of ACPI notify handlers, initial CRI platform support for the Intel Xe driver, and Rust support for the IBM System/390 (S/390) architecture." Linux kernel 7.2 also introduces a "Fair(er)" GPU scheduler, support for the 'zerocopy' library to Rust support to make zero-cost memory manipulation effortless, new hwcaps for the 2025 dpISA extensions on the AArch64 (ARM64) architecture, and enables large folios by default for the Btrfs file system. It also brings Intel CPU model number support for Panther Lake R processor series, improvements to the kernel's swap subsystem, support for multi-size transparent huge pages (mTHPs) to the khugepaged kernel thread, and support for compressed files to the SMB filesystem. On top of that, Linux 7.2 improves the new NTFS filesystem introduced in Linux kernel 7.1, adds devicetree updates for 64-bit NXP/Freescale and Qualcomm platforms, introduces MPTCP signaling support for IPv6 addresses, adds GRO/GSO support for PPPoE, and brings more SMP load-balancing updates... [T]he TCP authentication option has been implemented, and there are also some Thunderbolt networking improvements. Also worth mentioning is that the KVM subsystem has received support for AMD's "guest-mode execution trap" and Intel's "mode-based execution control" (MBEC) features, the NFS file system's default block size was bumped to 4MB on systems with at least 16GB RAM, and support for the Intel Trusted Domain Extensions (TDX) feature has been added. ' -- source: https://linux.slashdot.org/story/26/08/16/2349224 Cheers, Peter

1 0

Some Personal Spam Trivia
by Lawrence D'Oliveiro 17 Aug '26

17 Aug '26

I run my own mail server for my domain. Volumes are low enough that running a greylist (blocking all the drive-by spammers) gets rid of 90% of the crap. As for the rest, I had set up lists of blocked senders and recipients (addresses I have declared invalid on my domain) in Postfix. Then, a few months ago, I decided I would like to collect statistics on how often these list entries are hit, and I couldn’t be bothered to scan logfiles to do it. So a few months ago I implemented a “socketmap server” <https://www.postfix.org/socketmap_table.5.html> which responds to lookup requests from Postfix for those lists, gets the information from back-end database tables in MariaDB, and also logs entries in other database tables to record the requests. This makes it easy for me to add/remove entries in those blocklists, without having to restart Postfix (or my socketmap server). And I can pull back statistics on database lookups at any time. The first thing I’ve noticed so far, the most commonly-used rejection entry in the blocked-sender table is the one that refuses mail from the entire “.shop” TLD. I have never done any business with any site with such a domain; has anybody else done so? My feeling is, no reputable business has such a site. As for blocked recipients, I am fond of giving out different recipient addresses to different online forums and businesses and anybody else who needs to contact me. This way, I can spot immediately if an address I gave to one entity starts being used for mail coming from an entirely different entity. One or two online sites were too trusting in their early days about letting web scrapers glean email addresses from their user pages. I did later go in and enter new addresses on these sites, and invalidate the old ones, and they haven’t repeated the same mistake. But one of those old addresses keeps popping up, in some interestingly-corrupted forms. The address I originally entered took the form “from-«forum»@«domain”. I still get attempts to send mail to that (now invalid) address, but also to “u003cfrom-«forum»@«domain”. And more recently also to “cfrom-«forum»@«domain” -- addresses which I have never used anywhere. That “u003c” looks like a corruption of “U+003C”, which would be how you would write the Unicode code point for the “<” character (which is of course commonly used to delimit email addresses and URLs and such). I suspect the “cfrom” version is a further corruption of the “u003cfrom” one. I also get email sent to an address I haven’t used since my Apple Macintosh days, which would have been over a quarter century ago. I guess this goes to show that the spammers’ address lists can only grow over time, they can never remove any address as the lists are passed on, just in case it is still valid, because by the nature of their business they can never pay any attention to bounces that are supposed to indicate invalid addresses. And that passing on of lists introduces its own corruptions, as well. I wonder how big the spammers’ email lists are these days? I have this feeling the number of entries in typical lists might have exceeded the entire population of the Earth by now ...

1 0

Canonical Confirms Ubuntu Fixes for DirtyClone Linux Kernel Flaw
by Peter Reutemann 13 Aug '26

13 Aug '26

'Canonical has published additional details about DirtyClone, the recently disclosed Linux kernel vulnerability that can allow a local user to gain root privileges on affected systems. DirtyClone, tracked as CVE-2026-43503, was publicly disclosed by JFrog on June 25, 2026, and has a CVSS 3.1 score of 8.8 (high severity). Canonical notes the issue was responsibly disclosed to Linux kernel maintainers earlier, with the CVE record published on May 23. The first Ubuntu security updates addressing this vulnerability were released on June 2. Just a reminder that DirtyClone is a local privilege escalation flaw, allowing a local user to gain root access. For container deployments running third-party workloads, Canonical notes the vulnerability could enable container escapes, although no proof-of-concept exploit for this scenario has been published. DirtyClone affects the same components as the previously disclosed Dirty Frag and Fragnesia vulnerabilities. Systems where administrators have already blocked the affected kernel modules as mitigation for those issues are also protected against DirtyClone. For Ubuntu users, the fix is provided through Linux kernel image packages. The following kernel versions include the fix: - Ubuntu 26.04 LTS: 7.0.0-22.22 - Ubuntu 20.04 LTS with 5.15 kernel: 5.15.0-181.191~20.04.1 - Ubuntu 22.04 LTS: 5.15.0-181.191 - Ubuntu 24.04 LTS: 6.8.0-124.124 - Ubuntu 25.10: 6.17.0-35.35 Importantly, Ubuntu 20.04 LTS systems still using the 5.4 kernel remain affected. Older releases, including 14.04 LTS, 16.04 LTS, and 18.04 LTS, are also listed as affected in Canonical’s table.' -- source: https://linuxiac.com/canonical-confirms-ubuntu-fixes-for-dirtyclone-linux-k… Cheers, Peter

1 0

I Blame Python
by Lawrence D'Oliveiro 12 Aug '26

12 Aug '26

So MATLAB users are moving away from the language (and toolkit?) in favour of other tools <https://www.infoworld.com/article/4207496/matlab-programming-language-sinki…>. Yes, proprietary languages and toolkits are a lot less attractive these days, given the embarrassment of riches available in the open-source world. MATLAB was probably good enough at the time for its original use (matrix manipulation), but take advice from someone who had to do programming work in that language (for pay, yet): *stay well away* from the GUI-toolkit API, if you can help it. It was a long time ago, but I still get the shakes thinking about it ...

2 2