We Need to Talk About EXWM

System Crafters Store is live!

I’ve been getting requests for merch for a while, that day has finally arrived!

Check out the set of items I’ve put together with the System Crafters logo:


These are all high quality print-on-demand items so shipping can take up to 2 weeks but it’s usually faster than the estimate you get when you buy something.

Please let me know your feedback on everything: items, pricing, shipping costs, etc. I want the community to be proud to wear them!

This is just the first step; I plan to add new designs every 3 months or so, it will only get better from here!

We Need to Talk About EXWM

As I’ve mentioned before a few times, EXWM the maintainer ch11ng has been missing for over a year.

  • Last commits to EXWM and XELB were in July 2020
  • People are concerned: https://github.com/ch11ng/exwm/issues/845
  • Many lingering issues that people face daily, I’ve been burned frequently
  • I even tried StumpWM again!
  • Nothing beats EXWM, though.

What can be done about it?

Can the project be revived?

  • The project is in definite need of help to address pain points
  • It seems medranocalvo might be an admin on the repo
  • All it would really take is for an individual or small group to learn the code and try fixing things
  • FSF copyright assignment approval is a big challenge (can that be abandoned?)
  • Many people would be happy and relieved for EXWM to be maintained again

Should someone fork EXWM?

  • A hard fork with no intention to merge upstream can avoid FSF assignment issues
  • Could be an opportunity to rethink some things, redesign the code
  • Aside from learning the code, the challenge would be making it enticing (and smooth) to migrate
  • This could be the fastest route to a bright new future

A new alternative?

If we’re considering a fork, why not think bigger?

  • Can we fix some core issues with EXWM while planning for the future?
  • Minimal window manager core not written in Emacs Lisp but driven by it via RPC
  • Write it either in C or an embedded Scheme (Chibi, perhaps)
  • Create an abstract internal model; focus on X Window System and Wayland later
  • Make it possible to have a more traditional tiling window manager that is still Emacs-driven

Could be a huge undertaking, but possibly more future-proof!

Here’s what I’ll do

Solving this problem is important to me! I can’t live without my Emacs Desktop Environment.

  • I’m going to start investigating all possibilities in my (limited) free time
  • Great opportunity to stream the work on Twitch: https://twitch.tv/SystemCrafters
  • Start building a knowledgebase of how window management works with Xorg
  • Investigate various technical options, live coding regularly
  • As soon as a good direction is determined, get community involvement

For stream schedule: Mondays or Wednesdays? Tentatively 8am-10am Pacific Time.

What do you think?

Random Chat

Subscribe to the System Crafters Newsletter!
Stay up to date with the latest System Crafters news and updates! Read the Newsletter page for more information.
Name (optional)
Email Address