SeachessBulletin report (2023)
Table of contents
This report summarises the publication of the bulletin in 2023.
An attempt to draw some insights from the publication of the weekly bulletins in 2023. See also previous years:
Note that previous reports may not show the same numbers as this report. This is due to retrospective data fixes.
Throughout this report I'll use a few terms worth defining upfront:
- Item: An article, video, tutorial, tool or other resource read or assessed.
- Source: A place where to find items like newsletters or social media.
Let's start with some general numbers to frame the rest of the report.
| year | read_items | published_items | read_sources | published_sources |
|---|---|---|---|---|
| 2020 | 1040 | 306 | 33 | 20 |
| 2021 | 1104 | 306 | 34 | 22 |
| 2022 | 649 | 306 | 27 | 22 |
| 2023 | 447 | 294 | 19 | 17 |
This year I've read less than last year, continuing the trend downwards. The proportion between items read and items published is not great, I should reduce the amount of published items or increase my reading habits.
| year | mean | std | min | max |
|---|---|---|---|---|
| 2020 | 27.37 | 9.75 | 1 | 45 |
| 2021 | 21.23 | 7.05 | 9 | 44 |
| 2022 | 12.48 | 6.04 | 5 | 43 |
| 2023 | 9.31 | 3.87 | 3 | 25 |
# HTTPS vs HTTP
I believe all web resources should be accessible via HTTPS. However, some places only publish their content using HTTP.
| year | read_https | read_http | published_https | published_http |
|---|---|---|---|---|
| 2020 | 1017 | 23 | 300 | 6 |
| 2021 | 1091 | 13 | 305 | 1 |
| 2022 | 642 | 7 | 304 | 2 |
| 2023 | 444 | 3 | 291 | 3 |
# Sources
My regular sources are where I find most of my readings. These are the top 10 sources and the total items I found thanks to them. Note that the empty source signals either an ad-hoc source like following a link from another item.
| source | 2023 | 2022 | 2021 | 2020 |
|---|---|---|---|---|
| mastodon | 149 | 7 | 0 | 0 |
| 112 | 126 | 289 | 474 | |
| lobsters | 57 | 148 | 218 | 182 |
| work | 33 | 29 | 83 | 1 |
| tldr | 25 | 45 | 51 | 50 |
| family | 19 | 20 | 16 | 0 |
| weekrust | 14 | 28 | 70 | 70 |
| github | 7 | 14 | 22 | 1 |
| webtoolsweekly | 7 | 12 | 21 | 31 |
| friend | 6 | 20 | 20 | 0 |
The biggest change this year is that I stopped using Twitter and started using Mastodon. This is how Twitter changed over time:
| source | 2023 | 2022 | 2021 | 2020 |
|---|---|---|---|---|
| 0 | 133 | 136 | 32 |
In contrast with the top 10 sources for items I chose for the bulletins:
| source | 2023 | 2022 | 2021 | 2020 |
|---|---|---|---|---|
| mastodon | 91 | 4 | 0 | 0 |
| 73 | 56 | 76 | 108 | |
| lobsters | 32 | 74 | 69 | 45 |
| tldr | 23 | 26 | 23 | 9 |
| work | 22 | 13 | 14 | 0 |
| family | 18 | 5 | 9 | 0 |
| weekrust | 10 | 13 | 10 | 6 |
| github | 7 | 9 | 13 | 0 |
| webtoolsweekly | 5 | 8 | 12 | 4 |
| friend | 3 | 12 | 3 | 0 |
This year both lists are pretty similar. As expected, Mastodon has replaced completely Twitter which I stopped using at the end of last year.
Another way of thinking about sources is to look at how often domains repeat. It's a weird one though given that a single domain doesn't necessarily equate to a single publisher. Good examples of this are github.com, youtube.com or medium.com.
| domain | 2023 | 2022 | 2021 | 2020 |
|---|---|---|---|---|
| github.com | 124 | 129 | 142 | 121 |
| www.youtube.com | 4 | 4 | 6 | 13 |
| pythonspeed.com | 4 | 1 | 4 | 1 |
| surma.dev | 3 | 0 | 1 | 0 |
| dev.to | 3 | 9 | 32 | 29 |
| antonz.org | 3 | 2 | 1 | 0 |
Note that I'm excluding domains with 2 or less repetitons for practical reasons. This is the summary for 2023:
| 2023 | count | percentage |
|---|---|---|
| 1 | 282 | 94.00% |
| 2 | 12 | 4.00% |
| 3 | 3 | 1.00% |
| 4 | 2 | 0.67% |
| 124 | 1 | 0.33% |
These are the top 10 domains from items published in the bulletin:
| domain | 2023 | 2022 | 2021 | 2020 |
|---|---|---|---|---|
| github.com | 95 | 90 | 61 | 34 |
| www.youtube.com | 3 | 4 | 2 | 9 |
| dev.to | 2 | 2 | 4 | 4 |
| jvns.ca | 2 | 1 | 2 | 0 |
| mitchellh.com | 2 | 0 | 0 | 0 |
| pythonspeed.com | 2 | 0 | 2 | 1 |
| samwho.dev | 2 | 0 | 0 | 0 |
| simonwillison.net | 2 | 1 | 0 | 1 |
| surma.dev | 2 | 0 | 0 | 0 |
Once again, GitHub stays as the dominant domain. The rest is scattered despite the raise of platforms like dev.to. medium.com in particular is no surprise because I avoid as much as I can reading anything in that platform due to their cookie littering behaviour.
| 2023 | count | percentage |
|---|---|---|
| 1 | 182 | 95.29% |
| 2 | 7 | 3.66% |
| 3 | 1 | 0.52% |
| 95 | 1 | 0.52% |
# Topics
Once I read an article I tag it with a set of tags that I think represent the topic of the article. The top 10 tags are:
| tag | 2023 | 2022 | 2021 | 2020 |
|---|---|---|---|---|
| programming_language/rust | 144 | 153 | 188 | 192 |
| topic/tool | 60 | 92 | 125 | 94 |
| topic/web | 37 | 66 | 98 | 29 |
| programming_language/python | 32 | 26 | 42 | 40 |
| topic/database | 31 | 50 | 44 | 51 |
| topic/software_development | 25 | 42 | 48 | 43 |
| topic/data | 23 | 56 | 98 | 53 |
| programming_language/typescript | 22 | 28 | 16 | 3 |
| topic/security | 20 | 28 | 86 | 33 |
| programming_language/webassembly | 20 | 40 | 62 | 44 |
And the top 10 that where published in the bulletin:
| tag | 2023 | 2022 | 2021 | 2020 |
|---|---|---|---|---|
| programming_language/rust | 99 | 74 | 44 | 37 |
| topic/tool | 47 | 63 | 61 | 32 |
| topic/web | 25 | 33 | 37 | 5 |
| topic/database | 21 | 35 | 20 | 14 |
| programming_language/python | 19 | 18 | 18 | 18 |
| topic/data | 16 | 30 | 25 | 14 |
| programming_language/typescript | 15 | 18 | 6 | 1 |
| topic/programming_language | 15 | 10 | 13 | 4 |
| programming_language/javascript | 14 | 24 | 22 | 6 |
| topic/library | 14 | 10 | 8 | 6 |
Rust stays as my main interest and databases and data have slown down a bit.
For a third year in a row, WebAssembly doesn't show often enough in the bulletin although it has been, with Rust, my main focus for years.
# Conclusions
This year has been quite similar to 2022 in topics and sources (excluding the swap between Twitter and Mastodon). However I read substantially fewer items than previous years. I have decided to reduce the number of links per issue from 6 to 4.