crush depth

Boot Config

Did a pile of server updates today, and for some reason one specific VPS refused to boot with this error:

Boot error

It turned out I hadn't updated the boot code to handle the latest zpool version. There's a note about this buried at the bottom of the zpool administration chapter of the FreeBSD documentation:

This would be fine, except I've literally never had to do this once on any number of physical machines or virtual machines over more than a decade of continual upgrades. I don't understand why it was suddenly a problem on this one machine.

I mounted a FreeBSD ISO into the virtual machine, booted from it, temporarily imported and mounted the zroot zpool to /mnt and ran:

# gpart bootcode -b /boot/pmbr -p /boot/gptzfsboot -i 1 vtbd0

Exported the zpool again, and rebooted. Machine came up.

New PGP Keys

It's that time of year again.

Fingerprint                                       | Comment
2680 A50E FD03 2007 FABE 8C87 B0E4 322E EE81 8BDE | 2022 personal
3CCE 5942 8B30 462D 1045 9909 C560 7DA1 46E1 28B8 | 2022 maven-rsa-key
8EAE 10EA 3E8D F42F E58F CFD2 7863 6912 E065 EAD3 | 2021 jenkins-maven-rsa-key

Keys are published to the keyservers as usual.

Cold Spot ABS

I'm attempting to print a part like this:


The part must be printed in the orientation shown due to the anisotropic properties of 3D printed parts: The part needs to be printed such that the print layers are oriented perpendicular to the expected load.

This wouldn't normally be an issue, but the part is 25cm long. The build volume of the Prusa i3 MK3S is 250 × 210 × 210 mm. That left me exactly one way to print the part: Lay it on the print bed lengthwise.

Because I wasn't completely certain that the part dimensions were correct, I printed a small slice of the end of the model for testing:


Unfortunately, the printed part was pretty severely warped. The two highlighted lines are supposed to be parallel!


The part was printed in eSun ABS. Now, ABS has the reputation of being difficult to print. It reportedly has a tendency to warp and shrink in exactly the manner shown above. However, my personal experience of ABS has been the complete opposite to date! I've not, so far, had a single part warp in any observable way. I think I've had good results so far due to printing exclusively in an enclosure, using a 110°C heated bed, and tuning the material settings carefully. Therefore, having a part come out of the enclosure looking like this was a bit of a surprise.

My first thought was to question where I was printing on the bed. I don't normally print in that lower right corner. I tried moving the slice of the model to the middle of the print bed and printing again... No warping!

Late last year, I attached a small PVC duct to the back of the enclosure so that I could vent ABS fumes out of the window during prints. Given that we're coming into the winter months in the UK, I wondered if the warping was actually caused by a cold draft from outside getting into the enclosure and ruining the first print. I moved the part back into the corner and printed again... It warped. Not only did it warp, but it suspiciously warped in exactly the same manner as the first print; one of the corners (not both) lifted from the heat bed. Coincidentally, the corner that lifted was the corner that was placed in the same location on the heat bed both times (the bottom right corner).

I tried printing a version of the part with "mouse ears" - small brims placed under each corner. Once again, the part warped, and specifically only warped on one corner. The entire brim on that corner of the part actually lifted off the heat bed, whilst the other corner stayed firmly down.


In the above image, the marked lines are supposed to be parallel. Severe warping and buckling can be observed on the right side of the part, and absolutely no warping is present on the left side. The part is symmetrical; if it warped on one side, it should warp on the other.

At this point, I was pretty much convinced that something was unusual about that specific part of the heat bed.

I found a discussion thread on the Prusa forums where someone had produced a thermal image of the heat bed running at temperatures very close to the temperatures I'm using for ABS:


The original poster notes:

Heat distribution is not perfectly even, but it doesn't look too bad. The cold spots are where the magnets sit.


These images are without a steel sheet attached. When a steel sheet is added, it should spread out the heat more evenly towards the perimeters of the bed.

Note that in the image, both the bottom left and bottom right areas of the heat bed are noticeably colder than the rest. This might not matter for larger parts, but for a part with a small corner area, it clearly does!

I suspect moving the part to the next grid square over (in the horizontal band that says "Do not print") will fix the problem.

Using Changelog


All of this is in the documentation, but here's a friendly version.

First, grab the changelog jar file and wrapper script. At the time of writing the current version is 4.1.0. You can get both from Maven Central:

Download them with wget:

$ wget
$ wget
$ wget
$ wget

I recommend verifying the signatures, for security reasons. You can skip this step if you like to live dangerously.

$ gpg --verify com.io7m.changelog.cmdline-4.1.0-main.jar.asc
gpg: assuming signed data in 'com.io7m.changelog.cmdline-4.1.0-main.jar'
gpg: Signature made Tue 12 Jan 2021 20:11:22 GMT
gpg:                using RSA key B32FA649B1924235A6F57B993EBE3ED3C53AD511
gpg:                issuer ""
gpg: using pgp trust model
gpg: Good signature from " (2021 maven-rsa-key) <>" [full]
Primary key fingerprint: B32F A649 B192 4235 A6F5  7B99 3EBE 3ED3 C53A D511
gpg: binary signature, digest algorithm SHA512, key algorithm rsa3072

$ gpg --verify
gpg: assuming signed data in ''
gpg: Signature made Tue 12 Jan 2021 20:11:22 GMT
gpg:                using RSA key B32FA649B1924235A6F57B993EBE3ED3C53AD511
gpg:                issuer ""
gpg: using pgp trust model
gpg: Good signature from " (2021 maven-rsa-key) <>" [full]
Primary key fingerprint: B32F A649 B192 4235 A6F5  7B99 3EBE 3ED3 C53A D511
gpg: binary signature, digest algorithm SHA512, key algorithm rsa3072

You need Java 11 or newer installed. Check that you have an up-to-date version:

$ java -version
openjdk version "15.0.2" 2021-01-19
OpenJDK Runtime Environment (build 15.0.2+7)
OpenJDK 64-Bit Server VM (build 15.0.2+7, mixed mode)

Now, place both the jar file and the sh script somewhere on your $PATH. Personally, I use a $HOME/bin directory for this sort of thing.

$ echo $PATH

$ mkdir -p $HOME/bin
$ mv com.io7m.changelog.cmdline-4.1.0-main.jar $HOME/bin/

$ mv $HOME/bin/changelog
$ chmod +x $HOME/bin/changelog

$ which changelog

The changelog wrapper script requires two environment variables, CHANGELOG_HOME and CHANGELOG_JAR_NAME, to be defined. This allows the wrapper script to find the jar file and run it. The CHANGELOG_HOME environment variable tells the wrapper which directory it needs to look into for the jar file, and the CHANGELOG_JAR_NAME variable specifies the name of the jar file.

$ export CHANGELOG_HOME="$HOME/bin"
$ export CHANGELOG_JAR_NAME="com.io7m.changelog.cmdline-4.1.0-main.jar"

Now, you should be able to run changelog:

$ changelog
changelog: Main: INFO: Usage: changelog [options] [command] [command options]

      Set the minimum logging verbosity level.
      Default: info
      Possible Values: [trace, debug, info, warn, error]

  Use the "help" command to examine specific commands:

    $ changelog help help.

  Command-line arguments can be placed one per line into a file, and the file
  can be referenced using the @ symbol:

    $ echo help > file.txt
    $ echo help >> file.txt
    $ changelog @file.txt

    change-add              Add a change to the current release.
    help                    Show detailed help messages for commands.
    initialize              Initialize the changelog.
    release-begin           Start the development of a new release.
    release-current         Display the version number of the current release.
    release-finish          Finish a release.
    release-set-version     Set the version number of the current release.
    version                 Retrieve the program version.
    write-atom              Generate an atom feed.
    write-plain             Generate a plain text log.
    write-xhtml             Generate an XHTML log.



See the documentation for information on how to use the various changelog subcommands.

Editing Manually

Given that changelogs are simple XML files, it's possible to manually edit them in any text editor. However, the changelog tool is very strict about the format of changelogs, and will summarily reject any changelog that isn't valid according to the published XSD schema.

Modern IDEs such as Intellij IDEA contain excellent support for real-time XML editing, including autocompletion of XML elements, and real-time schema errors as you type. This makes it very difficult to accidentally create an invalid changelog file.

First, download the schema file and save it somewhere. For this example, I saved it in /tmp/schema.xsd, but you'll probably want to keep it somewhere a little more permanent than your temporary directory!

$ wget
$ mv schema.xsd /tmp

If you were to open an existing project that contained a README-CHANGES.xml file that was created with the changelog tool at this point, you would see something like this:


Note that the XML namespace is highlighted in red, because the IDE doesn't know where to find the schema for that namespace. If you click the namespace and then ALT+Enter, you'll get a drop-down menu that will allow you to set up the schema file:


If you select "Manually setup external resource", you'll get a file picker that you can use to select the schema.xsd file you downloaded earlier:


Now, when the schema file is selected, the XML namespace is no longer highlighted in red. Additionally, when you type, you'll be offered autocompletions of XML elements:


If you make mistakes editing the changelog file, you'll get nice validation errors as you type.

Here's "you tried to create an element that's not in the schema":


Here's "you typed garbage into the attributes of a known element":


Here's "you tried to reference an undeclared ticket system":


You can also, at any point, right-click the editor and select "Validate" to see the complete list of errors:


New PGP Keys

It's that time of year again.

Fingerprint                                       | Comment
2C83 11D5 E344 626A E76E B866 CB39 C234 E824 F9EA | 2021 personal
B32F A649 B192 4235 A6F5 7B99 3EBE 3ED3 C53A D511 | 2021 maven-rsa-key (RSA key to sign Maven Central releases)
FDE9 3549 7094 0B4D 5637 2157 53A8 207B C03F 0ACE | 2021 jenkins-maven-rsa-key
3E10 4C4C AE29 C040 1A26 6247 177C 64B8 7BA5 CD8A | 2021 android signing

Keys are published to the keyservers as usual.