You are here

Planet Debian

Subscribe to Feed Planet Debian
Planet Debian - https://planet.debian.org/
Përditësimi: 5 months 3 orë më parë

Written too much javascript and python feels less liberal.

Pre, 02/11/2018 - 1:16pd
Written too much javascript and python feels less liberal. Dict is not an Object.

Junichi Uekawa http://www.netfort.gr.jp/~dancer/diary/201811.html.en Dancer's daily hackings

October 2018 report: LTS, Monkeysphere, Flatpak, Kubernetes, CD archival and calendar project

Enj, 01/11/2018 - 9:12md
Debian Long Term Support (LTS)

This is my monthly Debian LTS report.

GnuTLS

As discussed last month, one of the options to resolve the pending GnuTLS security issues was to backport the latest 3.3.x series (3.3.30), an update proposed then uploaded as DLA-1560-1. I after a suggestion, I've included an explicit NEWS.Debian item warning people about the upgrade, a warning also included in the advisory itself.

The most important change is probably dropping SSLv3, RC4, HMAC-SHA384 and HMAC-SHA256 from the list of algorithms, which could impact interoperability. Considering how old RC4 and SSLv3 are, however, this should be a welcome change. As for the HMAC changes, those are mandatory to fix the targeted vulnerabilities (CVE-2018-10844, CVE-2018-10845, CVE-2018-10846).

Xen

Xen updates had been idle for a while in LTS, so I bit the bullet and made a first discovery of the pending vulnerabilities. I sent the result to the folks over at Credativ who maintain the 4.4 branch and they came back with a set of proposed updates which I briefly review. Unfortunately, the patches were too deep for me: all I was able to do was to confirm consistency with upstream patches.

I also brought up a discussion regarding the viability of Xen in LTS, especially regarding the "speculative execution" vulnerabilities (XSA-254 and related). My understanding is upstream Xen fixes are not (yet?) complete, but apparently that is incorrect as Peter Dreuw is "condident in the Xen project to provide a solution for these issues". I nevertheless consider, like RedHat that the simpler KVM implementation might provide more adequate protection against those kind of attacks and LTS users should seriously consider switching to KVM for hosing untrusted virtual machines, even if only because that code is actually mainline in the kernel while Xen is unlikely to ever be. It might be, as Dreuw said, simpler to upgrade to stretch than switch virtualization systems...

When all is said and done, however, Linux and KVM are patches in Jessie at the time of writing, while Xen is not (yet).

Enigmail

I spent a significant amount of time working on Enigmail this month again, this time specifically working on reviewing the stretch proposed update to gnupg from Daniel Kahn Gillmor (dkg). I did not publicly share the code review as we were concerned it would block the stable update, which seemed to be in jeopardy when I started working on the issue. Thankfully, the update went through but it means it might impose extra work on leaf packages. Monkeysphere, in particular, might fail to build from source (FTBFS) after the gnupg update lands.

In my tests, however, it seems that packages using GPG can deal with the update correctly. I tested Monkeysphere, Password Store, git-remote-gcrypt and Enigmail, all of which passed a summary smoke test. I have tried to summarize my findings on the mailing list. Basically our options for the LTS update are:

  1. pretend Enigmail works without changing GnuPG, possibly introducing security issues

  2. ship a backport of GnuPG and Enigmail through jessie-sloppy-backports

  3. package OpenPGP.js and backport all the way down to jessie

  4. remove Enigmail from jessie

  5. backport the required GnuPG patchset from stretch to jessie

So far I've taken that last step as my favorite approach...

Firefox / Thunderbird and finding work

... which brings us to the Firefox and Thunderbird updates. I was assuming those were going ahead, but the status of those updates currently seems unclear. This is a symptom of a larger problem in the LTS work organization: some packages can stay "claimed" for a long time without an obvious status update.

We discussed ways of improving on this process and, basically, I will try to be more proactive in taking over packages from others and reaching out to others to see if they need help.

A note on GnuPG

As an aside to the Enigmail / GnuPG review, I was struck by the ... peculiarities in the GnuPG code during my review. I discovered that GnuPG, instead of using the standard resolver, implements its own internal full-stack DNS server, complete with UDP packet parsing. That's 12 000 lines of code right there. There are also abstraction leaks like using "1" and "0" as boolean values inside functions (as opposed to passing an integer and converting as string on output).

A major change in the proposed patchset are changes to the --with-colons batch output, which GnuPG consumers (like GPGME) are supposed to use to interoperate with GnuPG. Having written such a parser myself, I can witness to how difficult parsing those data structures is. Normally, you should always be using GPGME instead of parsing those directly, but unfortunately GPGME does not do everything GPG does: signing operations and keyring management, for example, has long been considered out of scope, so users are force to parse that output.

Long story short, GPG consumers still use --with-colons directly (and that includes Enigmail) because they have to. In this case, critical components were missing from that output (e.g. knowing which key signed which UID) so they were added in the patch. That's what breaks the Monkeysphere test suite, which doesn't expect a specific field to be present. Later versions of the protocol specification have been updated (by dkg) to clarify that might happen, but obviously some have missed the notice, as it came a bit late.

In any case, the review did not make me confident in the software architecture or implementation of the GnuPG program.

autopkgtest testing

As part of our LTS work, we often run tests to make sure everything is in order. Starting with Jessie, we are now seeing packages with autopkgtest enabled, so I started meddling with that program. One of the ideas I was hoping to implement was to unify my virtualization systems. Right now I'm using:

Because sbuild can talk with autopkgtest, and autopkgtest can talk with qemu (which can use KVM images), I figured I could get rid of schroot. Unfortunately, I met a few snags;

  • #911977: how do we correctly guess the VM name in autopkgtest?
  • #911963: qemu build fails with proxy_cmd: parameter not set (fixed and provided a patch)
  • #911979: fails on chown in autopkgtest-qemu backend
  • #911981: qemu server warns about missing CPU features

So I gave up on that approach. But I did get autopkgtest working and documented the process in my quick Debian development guide.

Oh, and I also got sucked down into wiki stylesheet (#864925) after battling with the SystemBuildTools page.

Spamassassin followup

Last month I agreed we could backport the latest upstream version of SpamAssassin (a recurring pattern). After getting the go from the maintainer, I got a test package uploaded but the actual upload will need to wait for the stretch update (#912198) to land to avoid a versioning conflict.

Salt Stack

My first impression of Salt was not exactly impressive. The CVE-2017-7893 issue was rather unclear: first upstream fixed the issue, but reverted the default flag which would enable signature forging after it was discovered this would break compatibility with older clients.

But even worse, the 2014 version of Salt shipped in Jessie did not have master signing in the first place, which means there was simply no way to protect from master impersonation, a worrisome concept. But I assumed this was expected behavior and triaged this away from jessie, and tried to forgot about the horrors I had seen.

phpLDAPadmin with sunweaver

I looked next at the phpLDAPadmin (or PHPLDAPadmin?) vulnerabilities, but could not reproduce the issue using the provided proof of concept. I have also audited the code and it seems pretty clear the code is protected against such an attack, as was explained by another DD in #902186. So I asked Mitre for rejection, and uploaded DLA-1561-1 to fix the other issue (CVE-2017-11107). Meanwhile the original security researcher acknowledged the security issue was a "false positive", although only in a private email.

I almost did a NMU for the package but the security team requested to wait, and marked the package as grave so it gets kicked out of buster instead. I at least submitted the patch, originally provided by Ubuntu folks, upstream.

Smarty3

Finally, I worked on the smart3 package. I confirmed the package in jessie is not vulnerable, because Smarty hadn't yet had the brilliant idea of "optimizing" realpath by rewriting it with new security vulnerabilities. Indeed, the CVE-2018-13982 proof of content and CVE-2018-16831 proof of content both fail in jessie.

I have tried to audit the patch shipped with stretch to make sure it fixed the security issue in question (without introducing new ones of course) abandoned parsing the stretch patch because this regex gave me a headache:

'%^(?<root>(?:<span class="createlink"><a href="/ikiwiki.cgi?do=create&amp;from=blog%2F2018-11-01-report&amp;page=%3Aalpha%3A" rel="nofollow">?</a>:alpha:</span>:[\\\\]|/|[\\\\]{2}<span class="createlink"><a href="/ikiwiki.cgi?do=create&amp;from=blog%2F2018-11-01-report&amp;page=%3Aalpha%3A" rel="nofollow">?</a>:alpha:</span>+|<span class="createlink"><a href="/ikiwiki.cgi?do=create&amp;from=blog%2F2018-11-01-report&amp;page=%3Aprint%3A" rel="nofollow">?</a>:print:</span>{2,}:[/]{2}|[\\\\])?)(?<path>(?:<span class="createlink"><a href="/ikiwiki.cgi?do=create&amp;from=blog%2F2018-11-01-report&amp;page=%3Aprint%3A" rel="nofollow">?</a>:print:</span>*))$%u', "who is supporting our users?"

I finally participated in a discussion regarding concerns about support of cloud images for LTS releases. I proposed that, like other parts of Debian, responsibility of those images would shift to the LTS team when official support is complete. Cloud images fall in that weird space (ie. "Installing Debian") which is not traditionally covered by the LTS team.

Hopefully that will become the policy, but only time will tell how this will play out.

Other free software work irssi sandbox

I had been uncomfortable running irssi as my main user on my server for a while. It's a constantly running network server, sometimes connecting to shady servers too. So it made sense to run this as a separate user and, while I'm there, start it automatically on boot.

I created the following file in /etc/systemd/system/irssi@.service, based on this gist:

[Unit] Description=IRC screen session After=network.target [Service] Type=forking User=%i ExecStart=/usr/bin/screen -dmS irssi irssi ExecStop=/usr/bin/screen -S irssi -X stuff '/quit\n' NoNewPrivileges=true [Install] WantedBy=multi-user.target

A whole apparmor/selinux/systemd profile could be written for irssi of course, but I figured I would start with NoNewPrivileges. Unfortunately, that line breaks screen, which is sgid utmp which is some sort of "new privilege". So I'm running this as a vanilla service. To enable, simply enable the service with the right username, previously created with adduser:

systemctl enable irssi@foo.service systemctl start irssi@foo.service

Then I join the session by logging in as the foo user, which can be configured in .ssh/config as a convenience host:

Host irc.anarc.at Hostname shell.anarc.at User foo IdentityFile ~/.ssh/id_ed25519_irc # using command= in authorized_keys until we're all on buster #RemoteCommand screen -x RequestTTY force

Then the ssh irc.anarc.at command rejoins the screen session.

Monkeysphere revival

Monkeysphere was in bad shape in Debian buster. The bit rotten test suite was failing and the package was about to be removed from the next Debian release. I filed and worked on many critical bugs (Debian bug #909700, Debian bug #908228, Debian bug #902367, Debian bug #902320, Debian bug #902318, Debian bug #899060, Debian bug #883015) but the final fix came from another user. I was also welcome on the Debian packaging team which should allow me to make a new release next time we have similar issues, which was a blocker this time round.

Unfortunately, I had to abandon the Monkeysphere FreeBSD port. I had simply forgotten about that commitment and, since I do not run FreeBSD anywhere anymore, it made little sense to keep on doing so, especially since most of the recent updates were done by others anyways.

Calendar project

I've been working on a photography project since the beginning of the year. Each month, I pick the best picture out of my various shoots and will collect those in a 2019 calendar. I documented my work in the photo page, but most of my work in October was around finding a proper tool to layout the calendar itself. I settled on wallcalendar, a beautiful LaTeX template, because the author was very responsive to my feature request.

I also figured out which events to include in the calendar and a way to generate moon phases (now part of the undertime package) for the local timezone. I still have to figure out which other astronomical events to include. I had no response from the local Planetarium but (as always) good feedback from NASA folks which pointed me at useful resources to top up the calendar.

Kubernetes

I got deeper into Kubernetes work by helping friends setup a cluster and share knowledge on how to setup and manage the platforms. This led me to fix a bug in Kubespray, the install / upgrade tool we're using to manage Kubernetes. To get the pull request accepted, I had to go through the insanely byzantine CLA process of the CNCF, which was incredibly frustrating, especially since it was basically a one-line change. I also provided a code review of the Nextcloud helm chart and reviewed the python-hvac ITP, one of the dependencies of Kubespray.

As I get more familiar with Kubernetes, it does seem like it can solve real problems especially for shared hosting providers. I do still feel it's overly complex and over-engineered. It's difficult to learn and moving too fast, but Docker and containers are such a convenient way to standardize shipping applications that it's hard to deny this new trend does solve a problem that we have to fix right now.

CD archival

As part of my work on archiving my CD collection, I contributed three pull requests to fix issues I was having with the project, mostly regarding corner cases but also improvements on the Dockerfile. At my suggestion, upstream also enabled automatic builds for the Docker image which should make it easier to install and deploy.

I still wish to write an article on this, to continue my series on archives, which could happen in November if I can find the time...

Flatpak conversion

After reading a convincing benchmark I decided to give Flatpak another try and ended up converting all my Snap packages to Flatpak.

Flatpak has many advantages:

  • it's decentralized: like APT or F-Droid repositories, anyone can host their own (there is only one Snap repository, managed by Canonical)

  • it's faster: the above benchmarks hinted at this, but I could also confirm Signal starts and runs faster under Flatpak than Snap

  • it's standardizing: many of the work Flatpak is doing to make sense of how to containerize desktop applications is being standardized (and even adopted by Snap)

Much of this was spurred by the breakage of Zotero in Debian (Debian bug #864827) due to the Firefox upgrade. I made a wiki page to tell our users how to install Zotero in Debian considering Zotero might take a while to be packaged back in Debian (Debian bug #871502).

Debian work

Without my LTS hat, I worked on the following packages:

Other work

Usual miscellaneous:

Antoine Beaupré https://anarc.at/tag/debian-planet/ pages tagged debian-planet

Linux Audio Miniconf 2018 report

Enj, 01/11/2018 - 6:12md

The audio miniconference was held on the 21st in the offices of Cirrus Logic in Edinburgh with 15 attendees from across the industry including userspace and kernel developers, with people from several OS vendors and a range of silicon companies.

Community

We started off with a discussion of community governance lead by Takashi Iwai. We decided that for the ALSA-hosted projects we’ll follow the kernel and adopt the modified version of the contributor covenant that they have adopted, Sound Open Firmware already has a code. We also talked a bit about moving to use some modern hosted git with web based reviews. While this is not feasible for the kernel components we decided to look at doing this for the userspace components, Takashi will start a discussion on alsa-devel. Speaking of the lists Vinod Koul also volunteered to work with the Linux Foundation admin team to get them integrated with lore.kernel.org.

Liam Girdwood presenting virtualization (photo: Arun Raghavan) Kernel

Liam Girdwood then kicked off the first technical discussion of the day, covering virtualization. Intel have a new hypervisor called ACRN which they are using as part of a solution to expose individual PCMs from their DSPs to virtual clients, they have a virtio specification for control. There were a number of concerns about the current solution being rather specific to both the hardware and use case they are looking at, we need to review that this can work on architectures that aren’t cache coherent or systems where rather than exposing a DSP the host system is using a sound server.

We then moved on to AVB, several vendors have hardware implementations already but it seems clear that these have been built by teams who are not familiar with audio hardware, hopefully this will improve in future but for now there are some regrettable real time requirements. Sakamoto-san suggested looking at FireWire which has some similar things going on with timestamps being associated with the audio stream.

For SoundWire, basic functionality for x86 systems is now 90% there – we still need support for multiple CPU DAIs in the ASoC core (which is in review on the lists) and the Intel DSP drivers need to plumb in the code to instantiate the drivers.

We also covered testing, there may be some progress here this year as Intel have a new hypervisor called ACRN and some out of tree QEMU models for some of their newer systems both of which will help with the perennial problem that we need hardware for a lot of the testing we want to do. We also reviewed the status with some other recurring issues, including PCM granularity and timestamping, for PCM granularity Takashi Iwai will make some proposals on the list and for timestamping Intel will make sure that the rest of their driver changes for this are upstreamed. For dimen we agreed that Sakamoto-san’s work is pretty much done and we just need some comments in the header, and that his control refactoring was a good idea. There was discussion of user defined card elements, there were no concerns with raising the number of user defined elements that can be created but some fixes needed for cleanup of user defined card elements when applications close. The compressed audio userspace is also getting some development with the focus on making things easier to test, integrating with ffmpeg to give something that’s easier for user to work with.

Charles Keepax covered his work on rate domains (which we decided should really be much more generic than just covering sample rates), he’d posted some patches on the list earlier in the week and gave a short presentation about his plans which sparked quite a bit of discussion. His ideas are very much in line with what we’ve discussed before in this area but there’s still some debate as to how we configure the domains – the userspace interface is of course still there but how we determine which settings to use once we pass through something that can do conversions is less clear. The two main options are that the converters can expose configuration to userspace or that we can set constraints on other widgets in the card graph and then configure converters automatically when joining domains. No firm conclusion was reached, and since substantial implementation will be required it is not yet clear what will prove most sensible in practical systems.

Userspace

Sakamoto-san also introduced some discussion of new language bindings. He has been working on a new library designed for use with GObject introspection which people were very interested in, especially with the discussion of testing – having something like this would simplify a lot of the boilerplate that is involved in using the C API and allow people to work in a wider variety of languages without needing to define specific bindings or use the individual language’s C adaptations. People also mentioned the Rust bindings that David Henningsson had been working on, they were particularly interesting for the ChromeOS team as they have been adopting Rust in their userspace.

We talked a bit about higher level userspace software too. PulseAudio development has been relatively quiet recently, Arun talked briefly about his work on native compressed audio support and we discussed if PulseAudio would be able to take advantage of the new timestamping features added by Pierre-Louis Bossart. There’s also the new PipeWire sound server stack, this is a new stack which was originally written for video but now also has some audio support. The goal is to address architectural limitations in the existing JACK and PulseAudio stacks, offering the ability to achieve low latencies in a stack which is more usable for general purpose applications than JACK is.

DSPs

Discussions of DSP related issues were dominated by Sound Open Firmware which is continuing to progress well and now has some adoption outside Intel. Liam gave an overview of the status there and polled interest from the DSP vendors who were present. We talked about how to manage additions to the topology ABI for new Sound Open Firmware features including support for loading and unloading pieces of the DSP topology separately when dynamically adding to the DSP graph at runtime, making things far more flexible. The issues around downloading coefficient data were also covered, the discussion converged on the idea of adding something to hwdep and extending alsa-lib and tinyalsa to make this appear integrated with the standard control API. This isn’t ideal but it seems unlikely that anything will be. Techniques for handling long sequences of RPC calls to DSPs efficiently were also discussed, the conclusion was that the simplest thing was just to send commands asynchronously and then roll everything back if there are any errors.

Summary

Thanks again to all the attendees for their time and contributions and to Cirrus Logic for their generosity in hosting this in their Edinburgh office. It was really exciting to see all the active development that’s going on these days, it’ll be great to see some of that bear fruit over the next year!

Group photo broonie https://blog.sirena.org.uk Technicalities

My Work on Debian LTS (October 2018)

Enj, 01/11/2018 - 4:13md

after some nice family vacation in Tuscany, I did four hours of work on the Debian LTS project as a paid contributor at the end of this month. Thanks to all LTS sponsors for making this possible.

I move over a backlog of 4h from October to November (so I will work 12h on Debian LTS in November 2018).

Furthermore, I have signed up for Debian ELTS work with another 4h (as a start, more availability planned for upcoming months).

This is my list of work done in October 2018:

  • Upload of poppler (DLA 1562-1 [1]), fixing 4 CVEs
  • Discuss my research on CVE-2018-12689 in phpldapadmin from August 2018 with Antoine Beaupré who identified the published exploit as 'false positive' (for details, see his monthly LTS report for Octobre 2018).

light+love
Mike

References

[1] https://lists.debian.org/debian-lts-announce/2018/10/msg00024.html

sunweaver http://sunweavers.net/blog/blog/1 sunweaver's blog

FLOSS Activities October 2018

Enj, 01/11/2018 - 11:45pd
Changes Issues Review Administration
  • Debian: investigate MSA issue, investigate Fastly outage, investigate buildd issue, forwarded mail bounce info
  • Debian wiki: unblacklist IP range, clean up stray tmp files, whitelist email addresses, ask on IRC about accounts with bouncing email, ping accounts with bouncing email, disable accounts with bouncing email
  • Debian derivatives census: merge patches, deploy changes, clean up cruft, delete giant source packages
Communication Sponsors

All work was done on a volunteer basis.

Paul Wise http://bonedaddy.net/pabs3/log/ Log

Time for an official MIME type for patches?

Enj, 01/11/2018 - 8:15pd

As part of my involvement in the Nikita archive API project, I've been importing a fairly large lump of emails into a test instance of the archive to see how well this would go. I picked a subset of my notmuch email database, all public emails sent to me via @lists.debian.org, giving me a set of around 216 000 emails to import. In the process, I had a look at the various attachments included in these emails, to figure out what to do with attachments, and noticed that one of the most common attachment formats do not have an official MIME type registered with IANA/IETF. The output from diff, ie the input for patch, is on the top 10 list of formats included in these emails. At the moment people seem to use either text/x-patch or text/x-diff, but neither is officially registered. It would be better if one official MIME type were registered and used everywhere.

To try to get one official MIME type for these files, I've brought up the topic on the media-types mailing list. If you are interested in discussion which MIME type to use as the official for patch files, or involved in making software using a MIME type for patches, perhaps you would like to join the discussion?

As usual, if you use Bitcoin and want to show your support of my activities, please send Bitcoin donations to my address 15oWEoG9dUPovwmUL9KWAnYRtNJEkP1u1b.

Petter Reinholdtsen http://people.skolelinux.org/pere/blog/ Petter Reinholdtsen - Entries tagged english

Montreal's Debian &amp; Stuff - November 2018

Enj, 01/11/2018 - 5:00pd

November's wet, My socks are too, Away from keyboard; still on the net, Let's fix /usr/bin/$foo.

November can be a hard month in the Northen Hemisphere. It tends to be dark, rainy and cold. Montreal sure has been dark, rainy and cold lately.

That's why you should join us at our next Debian & Stuff later this month. Come by and work on Debian-related stuff - or not! Hanging out and chatting with folks is also perfectly fine. As always, everyone's welcome.

The date hasn't been decided yet, so be sure to fill out this poll before November 10th. This time we'll be hanging out at Koumbit.

What else can I say; if not for the good company, the bad poutine from the potato shack next door or the nice craft beer from the very hipster beer shop a little bit further down the street, you should drop by to keep November from creeping in too far.

Louis-Philippe Véronneau https://veronneau.org/ Louis-Philippe Véronneau

Review: In Pursuit of the Traveling Salesman

Enj, 01/11/2018 - 4:25pd

Review: In Pursuit of the Traveling Salesman, by William J. Cook

Publisher: Princeton University Copyright: 2012 ISBN: 0-691-15270-5 Format: Kindle Pages: 272

In Pursuit of the Traveling Salesman is a book-length examination of the traveling salesman problem (TSP) in computer science, written by one of the foremost mathematicians working on solutions to the TSP. Cook is Professor of Applied Mathematics and Statistics at Johns Hopkins University and is one of the authors of the Concorde TSP Solver.

First, a brief summary of the TSP for readers without a CS background. While there are numerous variations, the traditional problem is this: given as input a list of coordinates on a two-dimensional map representing cities, construct a minimum-length path that visits each city exactly once and then returns to the starting city. It's famous in computer science in part because it's easy to explain and visualize but still NP-hard, which means that not only do we not know of a way to exactly solve this problem in a reasonable amount of time for large numbers of cities, but also that a polynomial-time solution to the TSP would provide a solution to a vast number of other problems. (For those familiar with computability theory, the classic TSP is not NP-complete because it's not a decision problem and because of some issues with Euclidean distances, but when stated as a graph problem and converted into a decision problem by, for example, instead asking if there is a solution with length less than n, it is NP-complete.)

This is one of those books where the quality of the book may not matter as much as its simple existence. If you're curious about the details of the traveling salesman problem specifically, but don't want to read a lot of mathematics and computer science papers, algorithm textbooks, or books on graph theory, this book is one of your few options. Thankfully, it's also fairly well-written. Cook provides a history of the problem, a set of motivating problems (the TSP doesn't come up as much and isn't as critical as some NP-complete problems, but optimal tours are still more common than one might think), and even a one-chapter tour of the TSP in art. The bulk of the book, though, is devoted to approximation methods, presented in roughly chronological order of development.

Given that the TSP is NP-hard, we obviously don't know a good exact solution, but I admit I was a bit disappointed that Cook spent only one chapter exploring the exact solutions and explaining to the reader what makes the problem difficult. Late in the book, he does describe the Held-Karp dynamic programming algorithm that gets the work required for an exact solution down to exponential in n, provides a basic introduction to complexity theory, and explains that the TSP is NP-complete by reduction from the Hamiltonian path problem, but doesn't show the reduction of 3SAT to Hamiltonian paths. Since my personal interest ran a bit more towards theory and less towards practical approximations, I would have appreciated a bit more discussion of the underlying structure of the problem and why it's algorithmically hard. (I did appreciate the explanation of why it's not clear whether the general Euclidean TSP is even in NP due to problems with square roots, though.)

That said, I suppose there isn't as much to talk about in exact solutions (the best one we know dates to 1962) and much more to talk about in approximations, which is where Cook has personally spent his time. That's the topic of most of this book, and includes a solid introduction to the basic concept of linear programming (a better one than I ever got in school) and some of its other applications, as well as other techniques (cutting planes, branch-and-bound, and others). The math gets a bit thick here, and Cook skips over a lot of the details to try to keep the book suitable for a general audience, so I can't say I followed all of it, but it certainly satisfied my curiosity about practical approaches to the TSP. (It also made me want to read more about linear programming.)

If you're looking for a book like this, you probably know that already, and I can reassure you that it delivers what it promises and is well-written and approachable. If you aren't already curious about a brief history of practical algorithms for one specific problem, I don't think this book is sufficiently compelling to worth seeking out anyway. This is not a general popularization of interesting algorithms (see Algorithms to Live By if you're looking for that), or (despite Cook's efforts) particularly approachable if this is your first deep look at computer algorithms. It's a niche book that delivers on its promise, but probably won't convince you the topic is interesting if you don't see the appeal.

Rating: 7 out of 10

Russ Allbery https://www.eyrie.org/~eagle/ Eagle's Path

Debian LTS work, October 2018

Mër, 31/10/2018 - 11:26md

I was assigned 15 hours of work by Freexian's Debian LTS initiative and carried over 4 hours from September. I worked all 19 hours.

I released security updates for the linux (DLA 1529-1) and linux-4.9 (DLA 1531-1) packages. I prepared and released another stable update for Linux 3.16 (3.16.60), but have not yet included this in a Debian upload. I also released a security update for libssh (DLA 1548-1).

Ben Hutchings https://www.decadent.org.uk/ben/blog Better living through software

RHL'19 St-Cergue, Switzerland, 25-27 January 2019

Mër, 31/10/2018 - 10:06md

(translated from original French version)

The Rencontres Hivernales du Libre (RHL) (Winter Meeting of Freedom) takes place 25-27 January 2019 at St-Cergue.

Swisslinux.org invites the free software community to come and share workshops, great meals and good times.

This year, we celebrate the 5th edition with the theme «Exploit».

Please think creatively and submit proposals exploring this theme: lectures, workshops, performances and other activities are all welcome.

RHL'19 is situated directly at the base of some family-friendly ski pistes suitable for beginners and more adventurous skiers. It is also a great location for alpine walking trails.

Why, who?

RHL'19 brings together the forces of freedom in the Leman basin, Romandy, neighbouring France and further afield (there is an excellent train connection from Geneva airport). Hackers and activists come together to share a relaxing weekend and discover new things with free technology and software.

If you have a project to present (in 5 minutes, an hour or another format) or activities to share with other geeks, please send an email to rhl-team@lists.swisslinux.org or submit it through the form.

If you have any specific venue requirements please contact the team.

You can find detailed information on the event web site.

Please ask if you need help finding accommodation or any other advice planning your trip to the region.

Daniel.Pocock https://danielpocock.com/tags/debian DanielPocock.com - debian

Free software activities in October 2018

Mër, 31/10/2018 - 4:47md

Here is my monthly update covering what I have been doing in the free software world during October 2018 (previous month):

We intend to maintain changes to these modules under their original open source licenses and applying only free and open fixes and updates. You can find out more at goodformcode.com.

  • My activities as the current Debian Project Leader are covered in my monthly "Bits from the DPL" email to the debian-devel-announce mailing list.

  • I created Github-esque ribbons to display on Salsa-hosted websites. (Salsa being the collaborative development server for Debian and is the replacement for the now-deprecated Alioth service.)

  • Started a highly work-in-progress "Debbugs Enhancement Suite" Chrome browser extension to enhance various parts of the bugs.debian.org web interface.

  • Even more hacking on the Lintian static analysis tool for Debian packages:

    • New features:

      • Warn about packages that use PIUPARTS_* in maintainer scripts. (#912040)
      • Check for packages that parse /etc/passwd in maintainer scripts. (#911157)
      • Emit a warning for packages that do not specify Build-Depends-Package in symbol files. (#911451)
      • Check for non-Python files in top-level Python module directories. [...]
      • Check packages missing versioned dependencies on init-system-helpers. (#910594)
      • Detect calls to update-inetd(1) that use --group without --add, etc. (#909511)
      • Check for packages that encode a Python version number in their source package name. [...]
    • Bug fixes:

    • Misc:

      • Also show the maintainer name on the tag-specific reporting HTML. [...]
      • Tidy a number of references regarding the debhelper-compat virtual package. [...]
Reproducible builds

Whilst anyone can inspect the source code of free software for malicious flaws almost all software is distributed pre-compiled to end users.

The motivation behind the Reproducible Builds effort is to ensure no flaws have been introduced during this compilation process by promising identical results are always generated from a given source, thus allowing multiple third-parties to come to a consensus on whether a build was compromised.

This month:

  • I attended the Tandon School of Engineering (part of New York University) to speak and work with students from the Application Security course on the topic of reproducible builds.

  • Wrote and forwarded patch for Fontconfig to ensure the cache filenames are determinstic. [...]

  • I sent two previously-authored patches for GNU mtools to ensure the Debian Installer images could become reproducible. (1 & 2)

  • Submitted 11 Debian patches to fix reproducibility issues in fast5, libhandy, lmfit-py, mp3fs, opari2, pjproject, radon, sword, syndie, wit & zsh-antigen. I also submitted an upstream pull request for python-changelog.

  • Made a large number of changes to our website, including adding step-by-step instructions and screenshots on how to signup to our project on Salsa and migrating the TimestampsProposal page on the Debian Wiki to our website.

  • Fixed an issue in disorderfs — our FUSE-based filesystem that deliberately introduces non-determinism into directory system calls in order to flush out reproducibility issues — where touch -m and touch -a were not working as expected (#911281). In addition, ensured that failing an XFail test should in-itself be a failure [...].

  • Made the following changes to diffoscope, our in-depth and content-aware diff utility that can locate and diagnose reproducibility issues to:

    • Add support for comparing OCaml files via ocamlobjinfo. (#910542)

    • Add support for comparing PDF metadata using PyPDF2. (#911446)

    • Support gnumeric 1.12.43. [...]

    • Use str.startswith(...) over str.index(...) == 0 in the Macho comparator to prevent tracebacks if text cannot be found on the line. (#910540).

    • Add note on how to regenerate debian/tests/control.in and regenerate debian/tests/control with no material changes to add the regeneration comment itself. (1, 2)

    • Prevent test failures when running under stretch-backports by checking the OCaml version number. (#911846)

    • I also added a Salsa ribbon to the diffoscope.org website. [...]

  • Categorised a huge number of packages and issues in the Reproducible Builds "notes" repository and kept isdebianreproducibleyet.com up to date [...].

  • Worked on publishing our weekly reports. (#180, #181, #182 & #183)

  • Lastly, I fixed an issue in our Jenkins-based testing framework that powers tests.reproducible-builds.org to suppress some warnings from the cryptsetup initramfs hook which were causing some builds to be marked as "unstable". [...]


Debian


Debian bugs & patches filed
  • debbugs: Correct "favicon" location in <link/> HTML header. (#912186)

  • ikiwiki: "po" plugin can insert raw file contents with [[!inline]] directives. (#911356)

  • kitty: Please update homepage. (#911848)

  • pipenv: Bundles a large number of third-party libraries. (#910107)

  • mailman: Please include List-Id header on confirmation mails. (#910378)

  • fswatch: Clarify Files-Excluded entries. (#910330)

  • fuse3: Please obey nocheck build profile. (#910029)

  • gau2grid: Please add a non-boilerplate long description. (#911532)

  • hiredis: Please backport to stretch-backports. (#911732)

  • Please remove unnecessary overrides in fuse3 (#910030), puppet-module-barbican (#910374), python-oslo.vmware (#910011) & python3-antlr3(#910012)

  • python3-pypdf2: Python 3.x package ships non-functional Python 2.x examples. (#911649)

  • mtools: New upstream release. (#912285)

I also a filed requests with the stable release managers to update lastpass-cli (#911767) and python-django (#910821).


Debian LTS

This month I have worked 18 hours on Debian Long Term Support (LTS) and 12 hours on its sister Extended LTS project.

  • Multiple "frontdesk" shifts, triaging upstream CVEs, liasing with the Security Team, etc.

  • Issued DLA 1528-1 to prevent a denial-of-service (DoS) vulnerability in strongswan, a virtual private network (VPN) client and server where verification of an RSA signature with a very short public key caused an integer underflow in a length check that resulted in a heap buffer overflow.

  • Issued DLA 1547-1 for the Apache PDFBox library to fix a potential DoS issue where a malicious file could have triggered an extremely long running computation when parsing the PDF page tree.

  • Issued DLA 1550-1 for src:drupal7 to close remote code execution and an external URL injection exploit in the Drupal web-based content management framework as part of Drupal's SA-CORE-2018-006 security release.

  • Issued ELA-49-1 for the Adplug sound library to fix potential DoS attack due to double-free vulnerability.


Uploads
  • redis:

    • 5.0~rc5-2 — Use the Debian hiredis library now that #907259 has landed. (#907258)
    • 5.0.0-1 — New upstream release.
    • 5.0.0-2 — Update patch to sentinel.conf to ensure the correct runtime PID file location (#911407), listen on ::1 interfaces too for redis-sentinel to match redis-server, & run the new LOLWUT command in the autopkgtests.
  • python-django:

    • 1.11.16-1 — New upstream bugfix release.
    • 1.11.16-2 — Fix some broken README.txt symlinks. (#910120)
    • 1.11.16-3 — Default to supporting Spatialite 4.2. (#910240)
    • 2.1.2-1 — New upstream security release.
    • 2.1.2-2 — Default to supporting Spatialite 4.2. (#910240)
  • libfiu:

  • 0.96-5 — Apply patch from upstream to write fiu_ctrl.py atomically to avoid a.parallel build failure. (#909843)

  • 0.97-1 — New upstream release.
  • 0.97-2 — Mangle return offset sizes for 64-bit variants to prevent build failures on 32-bit architectures. (#911733)

  • adminer (4.6.3-2) — Use continue 2 to avoid a switch/continue warning in PHP 7.3, thus preventing an autopkgtest regression. (#911825)

  • bfs (1.2.4-1) — New upstream release.

  • django-auto-one-to-one (3.1.1-1) — New upstream release.

  • lastpass-cli (1.3.1-5) — Add ca-certificates to Depends.

  • python-redis (2.10.6-5) — Fix debian/watch file.

  • python-daiquiri (1.5.0-1) — New upstream release.


I also sponsored uploads of elpy (1.25.0-1) and hiredis (0.14.0-1).

FTP Team


As a Debian FTP assistant I ACCEPTed 95 packages: barrier, cct, check-pgactivity, cloudkitty-dashboard, cmark-gfm, eclipse-emf, eclipse-jdt-core, eclipse-platform-team, eclipse-platform-ua, eclipse-platform-ui, eos-sdk, equinox-p2, fontcustom, fonts-fork-awesome, fswatch, fuse3, gau2grid, gitlab, glom, grapefruit, grub-cloud, gsequencer, haskell-base-compat-batteries, haskell-invariant, haskell-parsec-numbers, haskell-reinterpret-cast, haskell-resolv, haskell-shelly, haskell-skylighting-core, haskell-wcwidth, hollywood, intelhex, javapoet, libgpg-error, libjsoncpp, libnbcompat, lintian-brush, llvm-toolchain-snapshot, mando, mat2, mini-httpd-run, modsecurity, mtree-netbsd, neutron-tempest-plugin, ngspice, openstack-cluster-installer, pg-checksums, pg-cron, pg-dirtyread, pg-qualstats, pg-repack, pg-similarity, pg-stat-kcache, pgaudit, pgextwlist, pgfincore, pgl-ddl-deploy, pgmemcache, pgpool2, pgrouting, pgsql-ogr-fdw, pgstat, pipenv, postgresql-hll, postgresql-plproxy, postgresql-plsh, puppet-module-barbican, puppet-module-icann-quagga, puppet-module-icann-tea, puppet-module-rodjek-logrotate, pykwalify, pyocd, python-backports.csv, python-fastfunc, python-httptools, python-redmine, python-tld, python-yaswfp, python3-simpletal, r-cran-eaf, r-cran-emoa, r-cran-ggally, r-cran-irace, r-cran-parallelmap, r-cran-popepi, r-cran-pracma, r-cran-spp, radon, rust-semver-parser-0.7, syndie, unicycler, vitetris, volume-key, weston & zram-tools.

I additionally filed 14 RC bugs against packages that had potentially-incomplete debian/copyright files against fontcustom, fuse3, intelhex, libnbcompat, mat2, modsecurity, mtree-netbsd, puppet-module-barbican, python-redmine, r-cran-eaf, r-cran-emoa, r-cran-pracma, radon & syndie.

Chris Lamb https://chris-lamb.co.uk/blog/category/planet-debian lamby: Items or syndication on Planet Debian.

SAT solvers for fun and fairness

Mar, 30/10/2018 - 10:55md

Trøndisk 2018, the first round of the Norwegian ultimate series (the frisbee sport, not the fighting style) is coming up this weekend! Normally that would mean that I would blog something about all the new and exciting things we are doing with Nageru for the stream, but for now, I will just point out that the stream is on plastkast.no and will be live from 0945–1830 CET on Saturday (group stage) and 1020–1450 (playoffs) on Sunday.

Instead, I wanted to talk about a completely different but interesting subproblem we had to solve; how do you set up a good schedule for the group stages? There are twelve teams, pre-seeded and split into two groups (call them A0–A5 and B0–B5) that are to play round-robin, but there are only two fields—and only one of them is streamed. You want a setup that maximizes fairness in the sense that people get adequate rest between matches, and also more or less equal number of streamed games. Throw in that one normally wants the more exciting games last, and it starts to get really tricky to make something good by hand. Could we do it programmatically?

My first thought was that since this is all about the ordering, it sounded like a variant of the infamous travelling salesman problem. It's well-known that TSP is NP-hard (or NP-complete, but I won't bother with the details), but there are excellent heursitic implementations in practice. In particular, I had already used OR-Tools, Google's optimization toolkit, to solve TSP problems in the past; it contains a TSP solver that can deal with all sorts of extra details, like multiple agents to travel (in our case, multiple fields), subconstraints on ordering and so on. (OR-Tools probably doesn't contain the best TSP solver in the world—there are specialized packages that do even better—but it's much better than anything I could throw together myself.)

However, as I tried figuring out something, and couldn't quite get it to fit (there are so many extra nonlocal constraints), I saw that the OR-Tools documentation had a subsection on scheduling problems. It turns out this kind of scheduling can be represented as a so-called SAT (satisfiability) problem, and OR-Tools also has a SAT solver. (SAT, in its general forms, is also NP-hard, but again, there are great heuristics.) I chose the Python frontend, which probably wasn't the best idea in the world (it's poorly documented, and I do wonder when Python will take the step into the 90s and make spelling errors in variables into compile-time errors instead of throwing a runtime exception four hours into a calculation), but that's what the documentation used, and the backend is in C++ anyway, so speed doesn't matter.

The SAT solver works by declaring variables and various constraints between them, and then asking the machine to either come up with a solution that fits, or to prove that it's not possible. Let's have a look of some excerpts to get a feel for how it all works:

We know we have 15 rounds, two fields on each, and every field should contain a match. So let's generate 30 such variables, each containing a match number (we use the convention that match 0, 2, 4, 6, etc. are on the stream field and 1, 3, 5, 7, etc. are played in parallel on the other field):

matchnums = [] for match_idx in range(num_matches): matchnums.append(model.NewIntVar(0, num_matches - 1, "matchnum%d" % (match_idx)))

So this is 30 variables, and each go from 0 to 29, inclusive. We start a fairly obvious constraint; we can only play each match once:

model.AddAllDifferent(matchnums)

The SAT solver might make this into a bunch of special constraints underneath, or it might not. We don't care; it's abstracted away for us.

Now, it's not enough to just find any ordering—after all, we want to find an ordering with some constraints. However, the constraints are rarely about the match numbers, but more about the teams that play in those matches. So we'll need some helper variables. For instance, it would be interesting to know which teams play in each match:

home_teams = [] away_teams = [] for match_idx in range(num_matches): home_teams.append(model.NewIntVar(0, num_teams - 1, "home_team_match%i" % (match_idx))) away_teams.append(model.NewIntVar(0, num_teams - 1, "away_team_match%i" % (match_idx))) model.AddElement(matchnums[match_idx], home_teams_for_match_num, home_teams[match_idx]) model.AddElement(matchnums[match_idx], away_teams_for_match_num, away_teams[match_idx])

AddElement() here simply is an indexing operation; since there's no difference between home and away teams for us, we've just pregenerated all the matches as A0 vs. A1, A0 vs. A2, etc. up until A4 vs. A6, A5 vs. A6 and then similarly for the other gruop. The “element” constraint makes sure that e.g. home_team_match0 = home_teams_for_match_num[matchnum0]. Note that even though I think of this is as an assignment where the home team for match 0 follows logically from which match is being played as match 0, it is a constraint that goes both ways; the solver is free to do inference that way, or instead first pick the home team and then deal with the consequences for the match number. (E.g., if it picks A5 as the home team, the match number most certainly needs to be 14, which corresponds to A5–A6.)

We're not quite done with the helpers yet; we want to explode these variables into booleans:

home_team_in_match_x_is_y = [[ model.NewBoolVar('home_team_in_match_%d_is_%d' % (match_idx, team_idx)) for team_idx in range(num_teams) ] for match_idx in range(num_matches)] for match_idx in range(num_matches): model.AddMapDomain(matchnums[match_idx], match_x_has_num_y[match_idx])

and similarly for away team and match number.

So now we have a bunch of variables of the type “is the home team in match 6 A4 or not?”. Finally we can make some interesting constraints! For instance, we've decided already that the group finals (A0–A1 and B0–B1) should be the last two matches of the day, and on the stream field:

model.AddBoolOr([match_x_has_num_y[28][0], match_x_has_num_y[28][15]]) model.AddBoolOr([match_x_has_num_y[26][0], match_x_has_num_y[26][15]])

This is a hard constraint; we don't have a solution unless match 0 and match 15 are the last two (and we earlier said that they must be different).

We're going to need even more helper variables now. It's useful to know whether a team is playing at all in a given round; that's the case if they are the home or away team on either field:

plays_in_round = {} for team_idx in range(num_teams): plays_in_round[team_idx] = {} for round_idx in range(num_rounds): plays_in_round[team_idx][round_idx] = model.NewBoolVar('plays_in_round_t%d_r%d' % (team_idx, round_idx)) model.AddMaxEquality(plays_in_round[team_idx][round_idx], [ home_team_in_match_x_is_y[round_idx * 2 + 0][team_idx], home_team_in_match_x_is_y[round_idx * 2 + 1][team_idx], away_team_in_match_x_is_y[round_idx * 2 + 0][team_idx], away_team_in_match_x_is_y[round_idx * 2 + 1][team_idx]])

Now we can establish a few other very desirable properties; in particular, each team should never need to play two matches back-to-back:

for round_idx in range(num_rounds - 1): for team_idx in range(num_teams): model.AddBoolOr([plays_in_round[team_idx][round_idx].Not(), plays_in_round[team_idx][round_idx + 1].Not()])

Note that there's nothing here that says the same team can't be assigned to play on both fields at the same time! However, this is taken care of by some constraints on the scheduling that I'm not showing for brevity (in particular, we established that each round must have exactly one game from group A and one from group B).

Now we're starting to get out of the “hard constraint” territory and more into things that would be nice. For this, we need objectives. One such objective is what I call ”tiredness”; playing matches nearly back-to-back (ie., game - rest - game) should have a penalty, and the solution should try to avoid it.

tired_matches = [] for round_idx in range(num_rounds - 2): for team_idx in range(num_teams): tired = model.NewBoolVar('team_%d_is_tired_in_round_%d' % (team_idx, round_idx)) model.AddMinEquality(tired, [plays_in_round[team_idx][round_idx], plays_in_round[team_idx][round_idx + 2]]) tired_matches.append(tired) sum_tiredness = sum(tired_matches)

So here we have helper variables that are being set to the minimum (effectively a logical AND) of “do I play in round N” and “do I play in round N + 2”. Tiredness is simply a sum of those 0–1 variables, which we can seek to minimize:

model.Minimize(sum_tiredness)

You may wonder how we went from a satisfiability problem to an optimization problem. Conceptually, however, this isn't so hard. Just ask the solver to find any solution, e.g. something with sum_tiredness 20. Then simply add a new constraint saying sum_tiredness <= 19 and ask for a re-solve (or continue). Eventually, the solver will either come back with a better solution (in which case you can tighten the constraint further), or the message that you've asked for something impossible, in which case you know you have the optimal solution. (I have no idea whether modern SAT solvers actually work this way internally, but again, conceptually it's simple.)

As an extra bonus, you do get incrementally better solutions as you go. These problems are theoretically very hard—in fact, I let it run for fun for a week now, and it's still not found an optimal solution—and in practice, you just take some intermediate solution that is “good enough”. There are always constraints that you don't bother adding to the program anyway, so there's some eyeballing involved, but still feels like a more fair process than trying to nudge it by hand.

We had many more objectives, some of them contradictory (e.g., games between more closely seeded opponents are more “exciting”, and should be put last—but they should also be put on the stream, so do you put them early on the stream field or late on the non-stream field?). It's hard to weigh all the factors against each other, but in the end, I think we ended up with something pretty nice. Every team gets to play two or three times (out of five) on the stream, only one team needs to be “tired” twice (and I checked; if you ask for a hard maximum of once for every team, it comes back pretty fast as infeasible), many of the tight matches are scheduled near the end… and most importantly, we don't have to play the first matches while I'm still debugging the stream. :-)

You can see final schedule here. Good luck to everyone, and consider using a SAT solver next time you have a thorny scheduling problem!

Steinar H. Gunderson http://blog.sesse.net/ Steinar H. Gunderson

Enabling Wake-on-Lan with the N34 Mini PC

Mar, 30/10/2018 - 8:58md

There is a room at the top of my house which was originally earmarked for storage (the loft is full of insulation rather than being a useful option). Then I remembered I still had my pico projector and it ended up as a cinema room as well. The pico projector needs really low light conditions with a long throw, so the fact the room only has a single small window is a plus.

I bought an “N34” mini PC to act as a media player - I already had a spare DVB-T2 stick to Freeview enable things, and the Kodi box downstairs has all my DVDs stored on it for easy streaming. It’s a Celeron N3450 based box with 4G RAM and a 32GB internal eMMC (though I’m currently running off an SD card because that’s what I initially used to set it up and I haven’t bothered to copy it onto the internal device yet). My device came from Amazon and is branded “Kodlix” (whose website no longer works) but it appears to be the same thing as the Beelink AP34.

Getting Linux onto it turned out to be a hassle. GRUB does not want to play with the EFI BIOS; it can be operated sometimes if manually called from the EFI Shell, but it does not work as the default EFI image to load. Various forum posts recommended the use of rEFInd, which mostly works fine.

Other than that Debian Stretch worked without problems. I had to pull in a backports kernel in order to make the DVB-T2 stick work properly, but the hardware on the N34 itself was all supported out of the box.

The other issue was trying to get Wake-on-Lan to work. The room isn’t used day to day so I want to be able to tie various pieces together with home automation such that I can have everything off by default and a scene configured to set things up ready for use. The BIOS has an entry for Wake-on-Lan, ethtool reported Supports Wake-on: g which should mean MagicPacket wakeup was enabled, but no joy. Looking at /proc/acpi/wakeup gave:

/proc/acpi/wakeup contents Device S-state Status Sysfs node HDAS S3 *disabled pci:0000:00:0e.0 XHC S3 *enabled pci:0000:00:15.0 XDCI S4 *disabled BRCM S0 *disabled RP01 S4 *disabled PXSX S4 *disabled RP02 S4 *disabled PXSX S4 *disabled RP03 S4 *disabled pci:0000:00:13.0 PXSX S4 *disabled pci:0000:01:00.0 RP04 S4 *disabled PXSX S4 *disabled RP05 S4 *disabled PXSX S4 *disabled RP06 S4 *disabled pci:0000:00:13.3 PXSX S4 *disabled pci:0000:02:00.0 PWRK S4 *enabled platform:PNP0C0C:00

pci:0000:01:00.0 is the network card:

01:00.0 Ethernet controller [0200]: Realtek […] Ethernet Controller [10ec:8168] (rev 0c)

I need this configured to allow wakeups which apparently is done via sysfs these days:

echo enabled > /sys/bus/pci/devices/0000\:01\:00.0/power/wakeup

This has to be done every boot so I just tied it into /etc/network/interfaces.

All of this then enables Home Assistant to control the Kodi box:

Home Assistant Kodi WoL configuration wake_on_lan: media_player: - platform: kodi name: Kodi (Cinema) host: kodi-cinema.here port: 8000 username: kodi password: !secret kodi_cinema_pass enable_websocket: false turn_on_action: service: wake_on_lan.send_magic_packet data: mac: 84:39:be:11:22:33 broadcast_address: 192.168.0.2 turn_off_action: service: media_player.kodi_call_method data: entity_id: media_player.kodi_cinema method: System.Shutdown

My Home Assistant container sits on a different subnet to the media box, and I found that the N34 wouldn’t respond to a Wake-on-Lan packet to the broadcast MAC address. So I’ve configured the broadcast_address for Home Assistant to be the actual IP of the media box, allowed UDP port 9 (discard) through on the firewall and statically nailed the ARP address of the media box on the router, so it transmits the packet with the correct destination MAC:

ip neigh change 192.168.0.2 lladdr 84:39:be:11:22:33 nud permanent dev eth0

I’ve still got some other bits to glue together (like putting the pico projector on a SonOff), but this gets me started on that process.

(And yes, the room is a bit cosier these days than when that photograph was taken.)

Jonathan McDowell https://www.earth.li/~noodles/blog/ Noodles' Emptiness

I was a podcast guest on The REPL

Pre, 26/10/2018 - 10:00md

Daniel Compton hosted me on his Clojure podcast, The REPL, where I talked about Debian, packaging Leiningen, and the Clojure ecosystem in Debian. It's got everything: spooky abandoned packages, anarchist collectives, software security policies, and Debian release cycles. Absolutely no shade was thrown at other distros.

Give it a listen:

Your browser does not support the audio element.

Download: MP3

More Q&A

After the podcast was published, Ivan Sagalaev wrote me with a great question about how the different versions of Clojure in Ubuntu 18.04 work:

First of all, THANK YOU for making sudo apt install leiningen work! It's so much better and more consistent than sourcing bash scripts :-)

I have a quick question for you. After installing leiningen and clojure on Ubuntu 18.04 I see that lein repl starts with clojure 1.8.0, while the clojure package itself seems to be independent and is version 1.9.0. How is it possible? I frankly haven't even seen lein downloading its own clojure.jar...

I replied:

Leiningen is "ahead-of-time (AOT) compiled", which is a fancy way of saying that the Leiningen you download from Ubuntu is pre-built. This means it is already compiled to Java bytecode, which can be run directly by Java. I ship the binary Leiningen package as an "uberjar", which means all its dependencies are also included inside the Leiningen jar.

Leiningen depends on and is built with Clojure 1.8, so the Leiningen uberjar in Debian also depends on Clojure 1.8. The "clojure" package in 18.04 defaults to installing Clojure 1.9, but that can be installed simultaneously with the "clojure1.8" package that Leiningen depends on in order to build. You can change your default Clojure to 1.8 using alternatives.

When you launch lein repl, by default the Clojure 1.8 runtime that's compiled in is used. If you run lein repl in the root of a Clojure 1.9 project, Leiningen will download Clojure 1.9 from Clojars and launch a 1.9 repl. If you want to use the Clojure 1.9 shipped with Debian, you can change :local-repo to point at /usr/share/maven-repo, but be careful to also set :offline? to true so you don't try to install things into the system maven repo by accident.

Elana Hashman https://hashman.ca/ hashman.ca

smartmontools

Pre, 26/10/2018 - 11:46pd

I don't do much Debian stuff these days (too busy) but I have adopted some packages over the last year. This has happened if a package that I rely on is lacking person-power and was at risk of being removed from Debian. I thought I should write about some of them. First up, smartmontools.

smartmontools let you query the "Self-Monitoring, Analysis and Reporting Technology" (S.M.A.R.T.) information in your computer's storage devices (hard discs and solid-state equivalents), as well as issue S.M.A.R.T. commands to them, such as instructing them to execute self-tests.

I rescued smartmontools for the Debian release in 2015, but I thought that was a one-off. Since I've just done it again I'm now considering it something I (co-)maintain1.

S.M.A.R.T. can, in theory, give you advance warning about a disc that is "not well" and could stop working. In practice, it isn't very good at predicting disc failures2 — which might explain why the package hasn't received more attention — but it can still be useful: last year it helped me to detect an issue with excessive drive-head parking I was experiencing on one of my drives.

  1. Personally I think the notion of single-maintainers for packages is old and destructive, and I think it should be the exception rather than the norm. Unfortunately it's still baked into a lot of our processes, policies and tools. ↩

jmtd https://jmtd.net/log/ Jonathan Dowland's Weblog

Migrated website from ikiwiki to Hugo

Enj, 25/10/2018 - 8:42md

So, I’ve been using ikiwiki for my website since 2011. At the time, I was hosting the website on a tiny hosting package included in a DSL contract - nothing dynamic possible, so a static site generator seemed like a good idea. ikiwiki was a good social fit at the time, as it was packaged in Debian and developed by a Debian Developer.

Today, I finished converting it to Hugo.

Why?

I did not really have a huge problem with ikiwiki, but I recently converted my blog from wordpress to hugo and it seemed to make sense to have one technology for both, especially since I don’t update the website very often and forget ikiwiki’s special things.

One thing that was somewhat annoying is that I built a custom ikiwiki plugin for the menu in my template, so I had to clone it’s repository into ~/.ikiwiki every time, rather than having a self-contained website. Well, it was a submodule of my dotfiles repo.

Another thing was that ikiwiki had a lot of git integration, and when you build your site it tries to push things to git repositories and all sorts of weird stuff – Hugo just does one thing: It builds your page.

One thing that Hugo does a lot better than ikiwiki is the built-in server which allows you to run `hugo server´ and get a local http URL you can open in the browser with live-reload as you save files. Super convenient to check changes (and of course, for writing this blog post)!

Also, in general, Hugo feels a lot more modern. ikiwiki is from 2006, Hugo is from 2013. Especially recent Hugo versions added quite a few features for asset management.

  • Fingerprinting of assets like css (inserting hash into filename) - ikiwiki just contains its style in style.css (and your templates in other statically named files), so if you switch theming details, you could break things because the CSS the browser has cached does not match the CSS the page expects.
  • Asset minification - Hugo can minimize CSS and JavaScript for you. This means browers have to fetch less data.
  • Asset concatenation - Hugo can concatenate CSS and JavaScript. This allows you to serve only one file per type, reducing the number of round trips a client has to make.

There’s also proper theming support, so you can easily clone a theme into the themes/ directory, or add it as a submodule like I do for my blog. But I don’t use it for the website yet.

Oh, and Hugo automatically generates sitemap.xml files for your website, teaching search engines which pages exist and when they have been modified.

I also like that it’s written in Go vs in Perl, but I think that’s just another more modern type of thing. Gotta keep up with the world!

Basic conversion

The first part to the conversion was to split the repository of the website: ikiwiki puts templates into a templates/ subdirectory of the repository and mixes all other content. Hugo on the other hand splits things into content/ (where pages go), layouts (page templates), and static/ (other files).

The second part was to inject the frontmatter into the markdown files. See, ikiwiki uses shortcuts like this to set up the title, and gets its dates from git:

[[!meta title="My page title"]]

on the other hand, Hugo uses frontmatter - some YAML at the beginning of the markdown, and specifies the creation date in there:

--- title: "My page title" date: Thu, 18 Oct 2018 21:36:18 +0200 ---

You can also have lastmod in there when modifying it, but I set enableGitInfo = true in config.toml so Hugo picks up the mtime from the git repo.

I wrote a small script to automatize those steps, but it was obviously not perfect (also, it inserted lastmod, which it should not have).

One thing it took me some time to figure out was that index.mdown needs to become _index.md in the content/ directory of Hugo, otherwise no pages below it are rendered - not entirely obvious.

The theme

Converting the template was surprisingly easy, it was just a matter of replacing <TMPL_VAR BASEURL> and friends with { .Site.BaseURL } and friends - the names are basically the same, just sometimes there’s .Site at the front of it.

Then I had to take care of the menu generation loop. I had my bootmenu plugin for ikiwiki which allowed me to generate menus from the configuration file. The template for it looked like this:

<TMPL_LOOP BOOTMENU> <TMPL_IF FIRSTNAV> <li <TMPL_IF ACTIVE>class="active"</TMPL_IF>><a href="<TMPL_VAR URL>"><TMPL_VAR PAGE></a></li> </TMPL_IF> </TMPL_LOOP>

I converted this to:

{{ $currentPage := . }} {{ range .Site.Menus.main }} <li class="{{ if $currentPage.IsMenuCurrent "main" . }}active{{ end }}"> <a href="{{ .URL }}"> {{ .Pre | safeHTML }} <span>{{ .Name }}</span> </a> {{ .Post }} </li> {{ end }}

this allowed me to configure my menu in config.toml like this:

[menu] [[menu.main]] name = "dh-autoreconf" url = "/projects/dh-autoreconf" weight = -110

I can also specify pre and post parts and a right menu, and I use pre and post in the right menu to render a few icons before and after items, for example:

[[menu.right]] pre = "<i class='fab fa-mastodon'></i>" post = "<i class='fas fa-external-link-alt'></i>" url = "https://mastodon.social/@juliank" name = "Mastodon" weight = -70

Setting class="active" on the menu item does not seem to work yet, though; I think I need to find out the right code for that…

Fixing up the details

Once I was done with that steps, the next stage was to convert ikiwiki shortcodes to something hugo understands. This took 4 parts:

The first part was converting tables. In ikiwiki, tables look like this:

[[!table format=dsv data=""" Status|License|Language|Reference Active|GPL-3+|Java|[github](https://github.com/julian-klode/dns66) """]]

The generated HTML table had the class="table" set, which the bootstrap framework needs to render a nice table. Converting that to a straightforward markdown hugo table did not work: Hugo did not add the class, so I had to convert pages with tables in them to the mmark variant of markdown, which allows classes to be set like this {.table}, so the end result then looked like this:

{.table} Status|License|Language|Reference ------|-------|--------|--------- Active|GPL-3+|Java|[github](https://github.com/julian-klode/dns66)

I’ll be able to get rid of this in the future by using the bootstrap sources and then having table inherit .table properties, but this requires saas or less, and I only have the CSS at the moment, so using mmark was slightly easier.

The second part was converting ikiwiki links like [[MyPage]] and [[my title|MyPage]] to Markdown links. This was quite easy, the first one became [MyPage](MyPage) and the second one [my title](my page).

The third part was converting custom shortcuts: I had [[!lp <number>]] to generate a link LP: #<number> to the corresponding launchpad bug, and [[!Closes <number>]] to generate Closes: #<number> links to the Debian bug tracker. I converted those to normal markdown links, but I could have converted them to Hugo shortcodes. But meh.

The fourth part was about converting some directory indexes I had. For example, [[!map pages="projects/dir2ogg/0.12/* and ! projects/dir2ogg/0.12/*/*"]] generated a list of all files in projects/dir2ogg/0.12. There was a very useful shortcode for that posted on the Hugo documentation, I used a variant of it and then converted pages like this to {{< directoryindex path="/static/projects/dir2ogg/0.12" pathURL="/projects/dir2ogg/0.12" >}}. As a bonus, the new directory index also generates SHA256 hashes for all files!

Further work

The website is using an old version of bootstrap, and the theme is not split out yet. I’m not sure if I want to keep a bootstrap theme for the website, seeing as the blog theme is Bulma-based - it would be easier to have both use bulma.

I also might want to update both the website and the blog by pushing to GitHub and then using CI to build and push it. That would allow me to write blog posts when I don’t have my laptop with me. But I’m not sure, I might lose control if there’s a breach at travis.

Julian Andres Klode https://blog.jak-linux.org/post/ Posts on Blog of Julian Andres Klode

MQTT enabling my doorbell

Enj, 25/10/2018 - 8:05md

One of the things about my home automation journey is that I don’t always start out with a firm justification for tying something into my setup. There’s not really any additional gain at present from my living room lights being remotely controllable. When it came to tying the doorbell into my setup I had a clear purpose in mind: I often can’t hear it from my study.

The existing device was a Byron BY101. This consists of a 433MHz bell-push and a corresponding receiver that plugs into a normal mains socket for power. I tried moving the receiver to a more central location, but then had issues with it not reliably activating when the button was pushed. I could have attempted the inverse of Colin’s approach and tried to tie in a wired setup to the wireless receiver, but that would have been too simple.

I first attempted to watch for the doorbell via a basic 433MHz receiver. It seems to use a simple 16 bit identifier followed by 3 bits indicating which tone to use (only 4 are supported by mine; I don’t know if other models support more). The on/off timings are roughly 1040ms/540ms vs 450ms/950ms. I found I could reliably trigger the doorbell using these details, but I’ve not had a lot of luck with reliable 433MHz reception on microcontrollers; generally I use PulseView in conjunction with a basic Cypress FX2 logic analyser to capture from a 433MHz receiver and work out timings. Plus I needed a receiver that could be placed close enough to the bell-push to reliably pick it up.

Of course I already had a receiver that could decode the appropriate codes - the doorbell! Taking it apart revealed a PSU board and separate receiver/bell board. The receiver uses a PT4318-S with a potted chip I assume is the microcontroller. There was an HT24LC02 I2C EEPROM on the bottom of the receiver board; monitoring it with my BusPirate indicated that the 16 bit ID code was stored in address 0x20. Sadly it looked like the EEPROM was only used for data storage; only a handful of values were read on power on.

Additionally there were various test points on the board; probing while pressing the bell-push led to the discovery of a test pad that went to 1.8v when a signal was detected. Perfect. I employed an ESP82661 in the form of an ESP-07, sending out an MQTT message containing “ON” or “OFF” as appropriate when the state changed. I had a DS18B20 lying around so I added that for some temperature monitoring too; it reads a little higher due to being inside the case, but not significantly so.

All of this ended up placed in the bedroom, which conveniently had a socket almost directly above the bell-push. Tying it into Home Assistant was easy:

binary_sensor: - platform: mqtt name: Doorbell state_topic: "doorbell/master-bedroom/button"

I then needed something to alert me when the doorbell was pushed. Long term perhaps I’ll add some sounders around the house hooked in via MQTT, and there’s a Kodi notifier available, but that’s only helpful when the TV is on. I ended up employing my Alexa via Notify Me:

notify: - name: alexa platform: rest message_param_name: notification resource: https://api.notifymyecho.com/v1/NotifyMe data: accessCode: !secret notifyme_key

and then an automation in automations.yaml:

- id: alexa_doorbell alias: Notify Alexa when the doorbell is pushed trigger: - platform: state entity_id: binary_sensor.doorbell to: 'on' action: - service: notify.alexa data_template: message: "Doorbell rang at {{ states('sensor.time') }}"

How well does this work? Better than expected! A couple of days after installing everything we were having lunch when Alexa chimed; the door had been closed and music playing, so we hadn’t heard the doorbell. Turned out to be an unexpected delivery which we’d otherwise have missed. It also allows us to see when someone has rang the doorbell when we were in - useful for seeing missed deliveries etc.

(Full disclosure: When initially probing out the mains doorbell for active signals I did so while it was plugged into the mains. My ‘scope is not fully isolated it seems and at one point I managed to trip the breaker on the mains circuit and blow the ringer part of the doorbell. Ooops. I ended up ordering an identical replacement (avoiding the need to replace the bell-push) and subsequently was able to re-use the ‘broken’ device as the ESP8266 receiver - the receiving part was still working, just not making a noise. The new receiver ended up in the living room, so the doorbell still sounds normally.)

  1. I have a basic ESP8266 MQTT framework I’ve been using for a bunch of devices based off Tuan PM’s work. I’ll put it up at some point. 

Jonathan McDowell https://www.earth.li/~noodles/blog/ Noodles' Emptiness

Review: Move Fast and Break Things

Enj, 25/10/2018 - 6:52pd

Review: Move Fast and Break Things, by Jonathan Taplin

Publisher: Little, Brown and Company Copyright: April 2017 Printing: 2018 ISBN: 0-316-27574-3 Format: Kindle Pages: 288

Disclaimer: I currently work for Dropbox, a Silicon Valley tech company. While it's not one of the companies that Taplin singles out in this book, I'm sure he'd consider it part of the problem. I think my reactions to this book are driven more by a long association with the free software movement and its take on copyright issues, and from reading a lot of persuasive work both good and bad, but I'm not a disinterested party.

Taplin is very angry about a lot of things that I'm also very angry about: the redefinition of monopoly to conveniently exclude the largest and most powerful modern companies, the ability of those companies to run roughshod over competitors in ways that simultaneously bring innovation and abusive market power, a toxic mix of libertarian and authoritarian politics deeply ingrained in the foundations of Silicon Valley companies, and a blithe disregard for the social effects of technology and for how to police the new communities that social media has created. This is a book-length rant about the dangers of monopoly domination of industries, politics, on-line communities, and the arts. And the central example of those dangers is the horrific and destructive power of pirating music on the Internet.

If you just felt a mental record-scratch and went "wait, what?", you're probably from a community closer to mine than Taplin's.

I'm going to be clear up-front: this is a bad book. I'm not going to recommend that you read it; quite the contrary, I recommend actively avoiding it. It's poorly written, poorly argued, facile, and unfair, and I say that with a great deal of frustration because I agree with about 80% of its core message. This is the sort of book from an erstwhile ally that makes me cringe: it's a significant supply of straw men, weak arguments, bad-faith arguments, and motivated reasoning that make the case for economic reform so much harder. There are good arguments against capitalism in the form in which we're practicing it. Taplin makes only some of them, and makes them badly.

Despite that, I read the entire book, and I'm still somewhat glad that I did, because it provides a fascinating look at the way unexamined premises lead people to far different conclusions. It also provides a more visceral feel for how people, like Taplin, who are deeply and personally invested in older ways of doing business, reach for a sort of reflexive conservatism when pushing back against the obvious abuses of new forms of inequality and market abuse. I found a reminder here to take a look at my own knee-jerk reactions and think about places where I may be reaching for backward-looking rather than forward-looking solutions.

This is a review, though, so before I get lost in introspection, I should explain why I think so poorly of this book as an argument.

I suspect most people who read enough partisan opinion essays on-line will notice the primary flaw in Move Fast and Break Things as early as I did: this is the kind of book that's full of carefully-chosen quotes designed to make the person being quoted look bad. You'll get a tour of the most famous ill-chosen phrases, expressions of greed, and cherry-picked bits of naked capitalism from the typical suspects: Google, Facebook, and Amazon founders, other Silicon Valley venture capitalists and CEOs, and of course Peter Thiel. Now, Thiel is an odious reactionary and aspiring fascist who yearns for the days when he could live as an unchallenged medieval lord. There's almost no quote you could cherry-pick from him that would make him look worse than he actually is, so I'll give Taplin a free pass on that one. But for the rest, Taplin is not even attempting to understand or engage with the arguments that his opponents are making. He's just finding the most damning statements, the ones that look the ugliest out of context, and parading them before the reader in an attempt to provoke an emotional reaction.

There is a long-standing principle of argument that you should engage with your opponents' position in its strongest form. If you cannot understand the merits and strengths of the opposing position and restate them well enough that an advocate of the opposing view would accept your summary as fair, you aren't prepared to argue the point. Taplin does not even come close to doing that. In the debate over the new Internet monopolies and monopsonies, one central conflict is between the distorting and dangerous concentration of power and the vast and very real improvements they've brought for consumers. I don't like Amazon as a company, and yet I read this book on a Kindle because their products are excellent and the consumer experience of their store is first-rate. I don't like Google as a company, but their search engine is by far the best available. One can quite legitimately take a wide range of political, economic, and ethical positions on that conflict, but one has to acknowledge there is a real conflict. Taplin is not particularly interested in doing that.

Similarly, and returning to the double-take moment with which I began this review, Taplin is startlingly unwilling to examine the flaws of the previous economic systems that he's defending. He writes a paean to the wonderful world of mutual benefit, artistic support, and economic fairness of record labels! Admittedly, I was not deeply enmeshed in that industry the way that he was, and he restrains his praise primarily to the 1960s and 1970s, so it's possible this isn't as mind-boggling as it sounds on first presentation. But, even apart from the numerous stories of artists cheated out of the profits of their work by the music industry long before Silicon Valley entered the picture, Taplin only grudgingly recognizes that the merits he sees in that industry were born of a specific moment in time, a specific pattern of demand, supply, sales method, and cultural moment, and that this world would not have lasted regardless of Napster or YouTube.

In other words, Taplin does the equivalent of arguing against Uber by claiming the taxi industry was a model of efficiency, economic fairness, and free competition. There are many persuasive arguments against new exploitative business practices. This is not one of them.

More tellingly to me, there is zero acknowledgment in this book that I can recall of one of the defining experiences of my generation and younger: the decision by the music and motion picture industries to fight on-line copying of their product by launching a vicious campaign of legal terrorism against teenagers and college students. Taplin's emotional appeals and quote cherry-picking falls on rather deaf ears when I vividly remember the RIAA and MPAA setting out to deliberately destroy people's lives in order to make an example of them, a level of social coercion that Google and Facebook have not yet stooped to, at least at that scale. Taplin is quite correct that his ideological opponents are scarily oblivious to some of the destruction they're wreaking on social and artistic communities, but he needs to come to terms with the fact that some of his allies are thugs.

This is where my community departs from Taplin's. I've been part of the free software community for decades, which includes a view of copyright that is neither the constrained economic model that Taplin advocates as a way to hopefully support artists, nor the corporate libertarian free-for-all from which Google draws its YouTube advertising profits. The free software community stands mostly opposed to both of those economic models, while pursuing the software equivalent of artist collectives. We have our own issues with creeping corporate control of our communities, and with the balance to strike between expanding the commons and empowering amoral companies like Google, Facebook, and Amazon to profit off of our work. Those fights play out in software licensing discussions routinely. But returning to a 1950s model of commercial music (which looks a lot like the 1980s model of commercial software) is clearly not possible, or even desirable if it were.

And that, apart from the poor argumentative technique and the tendency to engage with the weakest of his opponents' arguments, is the largest flaw I see in Taplin's book: he's invested in a binary fight between the economic world of his youth, which worked in ways that he considers fair, and a new economic world that is breaking the guarantees that he considers ethically important. He's not wrong about the problem, and I completely agree with him on the social benefit of putting artists in a more central position of influence in society. But he's not looking deeply at examples of artistic communities that have navigated this better than his own beloved music industry (book publishing, for example, which certainly has its problems with Amazon's monopsony power but is also in some ways stronger than it has ever been). And he's not looking at communities that are approaching the same problem from a different angle, such as free software. He's so caught up on what he sees as the fundamental unfairness of artists not being paid directly by each person consuming their work that he isn't stepping back to look at larger social goals and alternative ways they could be met.

I'm sure I'm making some of these same mistakes, in other places and in other ways. These problems are hard and some of the players truly are malevolent, so you cannot assume good will and good faith on all fronts. But there are good opposing arguments and simple binary analysis will fail.

Taplin, to give him credit, does try to provide some concrete solutions in the last chapter. He realizes that you cannot put the genie of easy digital copies back in the bottle, and tries to talk about alternate approaches that aren't awful (although they're things like micropayments and subscription services that are familiar ground for anyone familiar with this problem). I agree wholeheartedly with his arguments for returning to a pre-Reagan definition of monopoly power and stricter regulation of Internet advertising business. He might even be able to convince me that take-down-and-stay-down (the doctrine that material removed due to copyright complaints has to be kept off the same platform in the future) is a workable compromise... if he would also agree to fines, paid to the victim, of at least $50,000 per instance for every false complaint from a media company claiming copyright on material to which they have no rights. (Taplin seems entirely unaware of the malevolent abuses of copyright complaint systems by his beloved media industry.) As I said, I agree with about 80% of his positions.

But, sadly, this is not the book to use to convince anyone of those positions, or even the book to read for material in one's own debates. It would need more thoughtful engagement of the strongest of the arguments from new media and technology companies, a broader eye to allied fights, a deep look at the flaws in the capitalist system that made these monopoly abuses possible, and a willingness to look at the related abuses of Taplin's closest friends. Without those elements, I'm afraid this book isn't worth your time.

Rating: 3 out of 10

Russ Allbery https://www.eyrie.org/~eagle/ Eagle's Path

learn.to/quote

Enj, 25/10/2018 - 2:00pd

The “properly quote eMail messages and on Usenet” documentation is hosted on a server that appears to not get too much care at the moment. I’ve dug out workable versions:

The original link, with its http://learn.to/quote/ redirection, which contained the links to the translations into Dutch and English, unfortunately no longer works.

I’m asking everyone to please honour these guidelines when posting in Usenet and responding to eMail messages, as not doing so is an insult to all the (multiple, in the case of Usenet and mailing lists) readers / recipients of your messages. Even if you have to spend a little time trimming the quote, it’s much less than the time spent by all readers trying to figure out a TOFU (reply over fullquote) message.

Ich bitte jeden darum, sich bitte beim Posten im Usenet und Verfassen von eMails sich an diese Richtilinien zu halten; dies nicht zu tun ist ein Affront wider alle (im Falle von Usenet und Mailinglisten viele) Leser bzw. Empfänger eurer Nachrichten. Selbst wenn man zum Kürzen des Zitats ein bißchen Zeit aufwenden muß ist das immer noch deutlich weniger als die Mühe, die jeder einzelne Leser aufwenden muß, herauszufinden, was mit einer als TOFU (Text oben, Vollzitat unten) geschriebenen eMail gemeint ist.

Mag ik iederéén verzoeken, postings in het Usenet en mailtjes volgens deze regels te schrijven? Als het niet te doen is vies tegen alle ontvanger’s en moeilijk om te lezen. Zelfs als je een beetje tijd nodig heb om het oorspronkelijke deel te korten is het nog steeds minder dan de moeite van alleman, om een TOFU (antwoord boven, fullquote beneden) boodschap proberen te begrepen.

Thorsten Glaser http://www.mirbsd.org/ debian tag cloud

Salsa ribbons

Mër, 24/10/2018 - 4:55md

Salsa is the name of the collaborative development server for Debian and is the replacement for the now-deprecated Alioth service.

To make it easier to show the world that you use Salsa, I've created a number of Github-esque ribbons that you can overlay on your projects' sites by copying & pasting the appropriate snippet into your HTML.

For example:

You can find them, with instructions, here:

lamby.pages.debian.net/salsa-ribbons

If you're not satisfied with one of the colours, the original source is available.

Chris Lamb https://chris-lamb.co.uk/blog/category/planet-debian lamby: Items or syndication on Planet Debian.

Faqet