serhii.net

“Hugo uses Go’s html/template and text/template libraries as the basis for the templating.” ¹

I tried to use go as “language” in code blocks to highlight Hugo templates and it seems to work nicely!

The result of

```go
{{ range (.Paginate ($pages_l.GroupByDate "2006-01-02")).PageGroups  }}
```

is

{{ range (.Paginate ($pages_l.GroupByDate "2006-01-02")).PageGroups  }}

(I generated the first code listing using the \{-{-< highlight go >\}\} Hugo shortcode)

Previously I had the posts split by days (“Day 1234”), now for every former h2-header I have a separate post, but still want to split them by days.

Hugo can group posts by stuff, including by dates. ¹

This kinda works with pagination. ²

Now my list.html template for Diensttagebuch uses this to iterate through days/groups:

{{ $pages_k := where .RegularPagesRecursive ".Parent.Title" "Days" }} 
{{ $pages_j := where $pages_k "Params.draft" "ne" true}} 
{{ $pages_l := where $pages_j "Params.hidden" "ne" true}} 
{{ range (.Paginate ($pages_l.GroupByDate "2006-01-02")).PageGroups  }}

With the important bit being here, this iterates by day, not by month as in the examples: $pages_l.GroupByDate "2006-01-02"

Then the “day” header itself is {{.Key}}, to get the day of the month + month-year I do this:

<span class="day">{{ dateFormat "02" .Key }}</span>
{{ dateFormat "Jan 2006" .Key }}

Then iterating through the individual posts inside each “day” is:

{{ range .Pages }}
    <a href="{{ .RelPermalink }}">{{.Title}}</a>
    <span class="description">
    {{ .Content }}
    </span>
{{ end }}

nvidia-smi has a python library: nvsmi · PyPI

import nvsmi

nvsmi.get_gpus()
nvsmi.get_available_gpus()
nvsmi.get_gpu_processes()

RabbitMQ is a message broker / scheduler that allows sending/receiving messages.

RabbitMQ is a message broker: it accepts and forwards messages. You can think about it as a post office: when you put the mail that you want posting in a post box, you can be sure that the letter carrier will eventually deliver the mail to your recipient. In this analogy, RabbitMQ is a post box, a post office, and a letter carrier.

The major difference between RabbitMQ and the post office is that it doesn’t deal with paper, instead it accepts, stores, and forwards binary blobs of data ‒ messages.

#GAN that generates handwritten text: amzn/convolutional-handwriting-gan: ScrabbleGAN: Semi-Supervised Varying Length Handwritten Text Generation (CVPR20) #pytorch

Python uninstalling requirements.txt

You can do python -m pip uninstall -r requirements.txt

python3 bdist_wheel errors

Errors with bdist_wheel missing as a command when installing python packages got fixed with the help of SO¹, needed to do python3 -m pip install wheel

After writing whatever: str or Path or whataver: Union[str, Path] for the N-th time I googled how to do this better. Well, ¹

from typing import Union
from pathlib import Path

pathlike = Union[str, Path]

whatever: pathlike = some_function()

def f_paths(path_one: pathlike):

From FreeCodeCamp:¹

• git branch shows all branches
• git push --all pushes all local branches to remote.
• git push origin some-branch:my-feature pushes the local branch some-branch to a remote branch called my-feature

I should really try this sometime. Having a reproducible OS install would make life much easier. On my radar a long time, but a person I was interviewing last week was the final drop I guess.

• A really nice post about this: Moving into NixOS | Gabriel Arazas
• Sample configs: Configuration Collection - NixOS Wiki
  • Someone’s i3 config: dotfiles/i3-config.nix at master · aaronjanse/dotfiles

    bindsym ${mod}+d exec ${nixpkgs.rofi}/bin/rofi -show run
    

    Quite interesting.
• Official guide: NixOS - Guides - How Nix works
• Building a Reproducible Data Science Environment with Nix

  Nix is a 100% reproducible package manager, for all languages and all things. This means your python environment, your R environment, your models, your entire computer can be completely reproduced, all using the magic of nix. In this article, we will walk through setting up a simple, reproducible, and failproof data science stack with nix, including importing packages not found on nixpkgs and caching the builds online

From SO, to find the disk space taken by files with a certain extension/type:¹

find ./photos/john_doe -type f -name '*.jpg' -exec du -ch {} + | grep total$