Thinking about Emacs coaching goals with Prot
| emacs, communityI want to get better at learning with other people's help, so I'm going to experiment with engaging Prot as an Emacs coach. Our first session is this week. Time to lay the groundwork!
If I meet with Prot twice a month for three months, that's a budget of €60 (~CAD 100), which is a reasonable size for an experiment especially since I still have the budget set aside from the Google Open Source Peer Bonus and lovely folks already donated to cover the costs for EmacsConf. When I schedule something with someone, the accountability makes it easier to get stuff done and out the door. For this, a real person is much better than AI because:
- I get to take advantage of Prot's very large context window, and he knows stuff about the Emacs, the community, and me that I might not remember to mention
- He can ask real questions and prod at things that are unclear or contradictory, unlike the confirmation bias of LLMs
- He might point out things that wouldn't occur to me to ask about
- It triggers my "I promised someone I'd do this" thing
- I get to support an individual worth supporting rather than contributing to the concentration of wealth and information in for-profit entities
My motivations:
I want to make better use of my focused time during the rest of the schoolyear. For the next three months, my schedule will be fairly predictable and I'll have regular chunks of focused time. Over the past two months, I've averaged around 10 hours of Emacs-related stuff per week (including 1.5 hours or so for Emacs News). I'm currently thinking about language learning and speech input. EmacsConf is on the horizon and will probably ramp up after September, but I can also think ahead of workflow improvements or ways to collaborate with other people. I might put together an Emacs News Highlights presentation. Also, I'm always looking out for ways to build the community.
Summer break during July and August will shake things up again, but I might be able to find some focused time early morning or evening. I'd like to be in a good position to make the most of those time fragments.
- I want to improve my Emacs Lisp development workflow and learn more about libraries and techniques that might be useful. I'm beginning to have more time to sharpen the saw and I'm curious about all the cool stuff that I missed or skimmed over the past ten years. What are some useful setups for completion, debugging, navigation, etc.?
- Current: I sporadically use the extra awesomeness in seq, pcase, lispy, erefactor, ert, buttercup, and undercover, but not consistently. I'd like to reduce the friction and make these habitual.
- Areas of friction / improvement:
- writing tests, especially for things that are more interactive
- navigating code that might be scattered in literate config files or in Emacs Lisp files
- forgetting to restart or to make sure all code is saved; running tests via Emacs batch mode will help, as will
package-isolateandrestart-emacs
- I want to improve my workflows for writing, making videos, and streaming. If I get better at sharing what I'm working on, I might be able to connect with more people and bounce ideas around. Also, accountability might help me nudge this over the threshold. I probably still need to work in stops and starts, so I want to reduce the friction. I'm curious about other people's workflows for sharing. I like joining meetups, but I tend to share stuff only if no one else has anything planned, because I have my blog and my YouTube channel in case I want to share anything with a wider group of people. I just have to actually post things.
Current: ~1.5 Emacs posts a week aside from Emacs News, attending meetups, sporadically adding short video demos to posts
Average number of Emacs-related posts that aren't Emacs News
(let* ((start "2026-02-01") (end "2026-03-31") (posts (my-blog-posts start end (lambda (o) (and (member "emacs" (alist-get 'categories o)) (not (member "emacs-news" (alist-get 'categories o))))))) (count (length posts))) (my-weekly-average count start end))- Goal: 2-3 non-News posts a week, one video a month, one stream or meetup a month; maybe also beyond looking at the numbers, it might be interesting to build more momentum around a topic, set up trails/navigation, cultivate more of a digital garden
- Areas of friction / improvement:
- Resisting "one more tweak"
- Streaming: Still need to get the hang of talking to myself or having half-conversations with chat: can be worked around by scheduling a session with Prot and opening it to the public
- Hiding private information or setting up a separate Emacs for demonstration
- Harvesting videos/clips/notes afterwards
- I want to move more of my configuration into files and libraries that other people can reuse, like sachac/learn-lang and sachac/speech-input. I can also separate the function definitions from the configuration in my code so that people can reuse the functions if they want.
- Areas of friction / improvement
- renaming things when I want to move them to a library
- duplicating small functions (ex: simplify string)
- figuring out how to make it possible for someone else to start using my stuff
- Areas of friction / improvement
Starting questions for Prot:
- Meta: what are people finding useful for coaching and behaviour change, like learning new keyboard shortcuts or workflows?
- Your literate config exports to individual .el files. I could probably do something similar to separate my functions from my personal config in order to make it easier for people to reuse parts of my config. Is it worth doing so? Do people tell you that they use those private Emacs Lisp files by loading them, or do they mostly rely on your published packages?
- Do you have some tweaks to make it easier to jump to function definitions considering a literate configuration?
- What's your general process for migrating things from your config to a repository or package?
Could be fun. Let's experiment!