Another Bariatric Surgery Complication

This is another case of the very frequent military hospital bariatric surgery complication. This patient probably has an internal hernia with torsion/volvulus of the small bowel mesentery. There is at least venous congestion and probably venous ischemia of the bowel given the mesenteric edema. One of the mesenteric vessels is almost the same size as the superior mesenteric vein. Bowel obstruction was also present which was probably closed-loop given the configuration.

I generally do not find the sagittal images very helpful on these cases and this bears that out. The coronal images are quite striking, with twisting of the mesentery and “whorl sign”


DietPi on OrangePi Zero for HomeAssistant

Well, it was a good run with DietPi running HomeAssistant on my Orange Pi Zero.  The install mentioned previously was a bare metal install and I have moved to docker, hopefully permanantly.  The DietPi for the Raspberry Pi I’m running it on will almost certainly never go away, but if you bought an OPZero there is hope.

In essence, the dietpi folks have created a script that debloats a Raspbian install.


Read about it here:

Essentially, OrangePi has condemned themselves to extinction.

Regardless of the above, DietPi can still run on any Debian device!

  • Any system can run DietPi. All you need a generic Debian image and use of our automated DietPi PREP system:
    #1285 (comment)

In the interest of further reducing the stretch and strain on our support infrastructure.

  • All OrangePi and BananaPi images will be dropped from our lineup.

Although these boards offer a low priced SBC product, the experience of these boards is never guaranteed due to their kernel development being behind the other SBC devices. In most cases requiring years of open source development to fix stability and achieve any acceptable level of expected use, with no cost to the board and/or chip manufacture.

As our work is focused outside of kernel development (its not what we do), we are only willing to support devices which provide a stable kernel experience and a high quality product in the first instance.


So a new adventure begins, DietPi-izing a Raspbian install and putting docker on it!

Building my first Docker Container “Mini-Mosquitto” MQTT Broker

I’ve previously written about being stymied by docker so after getting several containers running properly, I decided to try my hand at creating a mosquitto broker.  As with everything else you find on the internet, the veracity and integrity of installed binaries is incumbent you, the user.  This leads me to believe that in simple cases, it is probably better to create your own microservices.

There are plenty of brokers out there built on mosquitto.  I started with the very complete:

pascaldevink/rpi-mosquitto:latest    206.8 MB

Which is a great container EXCEPT, it does not have the mosquitto-clients which are necessary for mosquitto troubleshooting.  It does however, have websocket full implementation which is not exactly trivial.

I wanted small, so I started with Alpine Linux.  This ends up being around 5MB, which is their reported size.  Size reported by docker is actually smaller:

arm32v6/alpine:latest    4 MB

With mosquitto installed via the following Dockerfile, I ended up with a tiny image, comparable to the trafex image that is probably the most popular Pi MQTT broker.

jumblies/mini-mosquitto:latest 4.6 MB


FROM arm32v6/alpine

RUN apk add --no-cache mosquitto-clients mosquitto

EXPOSE 1883 
# unnecessary exposure to since no websocket 9001

ENTRYPOINT ["/bin/sh" "-c", "/usr/sbin/mosquitto -d"]

COPY config /mqtt/config

ENTRYPOINT ["/usr/sbin/mosquitto"]

I left out the maintainer info, since I doubt it will be distributed and it will probably only be built once.  This does run as root.

To run it:

docker run -it -d -p 1883:1883 --name mini-mosquitto jumblies/mini-mosquitto:latest

To use clients:

watch all messages:
docker exec -ti mini-mosquitto mosquitto_sub -h -v -t '#'

publish a message:
docker exec -ti mini-mosquitto mosquittpub -h -t "caviar" -m "tangerine speedo" -r

Adult presentation Pulmonary Sequestration

Kick Butt Pics





I apologize for the windowing but wanted to demonstrate the aberrant vessel arising from the aorta and the sagittal appearance of the bronchopulmonary foregut malformation a.k.a. pulmonary sequestration.

Findy McFindings


Wordy McWordface


The abnormalities in the right lower lobe identified on abdominal CT are an isolated finding. When viewed on multiplanar reformats, this does not appear consistent with bronchiectasis and is diagnostic for a pulmonary sequestration. There is a large systemic arterial feeding vessel extending off the left side of the descending thoracic aorta which crosses anterior to the descending thoracic aorta between the gastroesophageal junction and aorta just superior to the esophageal hiatus and feeds the bronchopulmonary foregut malformation. The venous drainage appears to be pulmonary venous rather than systemic venous and although this is not a constant indicator of intra-lobar versus extra-lobar sequestrations, it suggests an intra-lobar sequestration. Infection is unlikely but difficult to completely exclude. These malformations can acquire communication with the tracheobronchial tree



Right lower lobe pulmonary sequestration, likely intralobar. Superimposed infection is in the differential considerations although the fluid levels do not necessarily indicate the presence of infection or communication with the tracheobronchial tree. Follow-up pulmonary or thoracic surgery consultation is recommended.