Hi Guix,

  I’ve been recently thinking about how Guix relates to sobriety, its
  ecological impact, and more specially, on how Guix envisions to afford
  possible limitations on available resources in the times to come. I’m
  thinking here mainly about bandwidth, storage resources and energy
  consumption, but not only.

  I wonder if this is something relevant, or even something we need to
  keep in mind as we move on: I haven’t found any related discussion
  about that matter in the logs. If sobriety is something we need to
  consider when contributing to Guix, I’d like to drop here a couple of
  ideas to help reducing the impact on the environment of the huge
  computing consumption carried on by Guix development and usage.

  First, tests: some take ages to execute. Do we really need this ?
  Upstream projects sometimes implement checking procedures which
  expands forever in massively parallel hardware, is this acceptable at
  all ? Try to build ‘python-lifelines’ to get a feeling of what I mean
  here. Proposal: we should maybe impose a limitation to the time tests
  take, as a rule.

  Then, package size: ‘emacs-calibredb’ is 5 GB. This is not an
  exception, but rather the norm, I’m afraid. Building packages,
  upgrading systems, etc. usually implies a huge storage plan. And this
  is getting worst. Proposal: first, make it explicit, ‘guix build/lint/shell’
  should always return the size of the package, similarly to many other
  guix commands. Let’s make the problem pop up, so we all become aware
  of where exactly we’re going.

  Related to the previous, bandwidth. Huge packages take a lot of
  bandwidth to install. Sure, this is the way it is, and intrinsic to
  functional package management, but still: do we really need to bring
  whole inputs to get a single binary ? Are we just happily adding
  inputs for the sake of building the package ? Proposal: new tooling
  and checking would be necessary to automatise things. In the meantime,
  please, dear Guix contributors, watch this kind of details.

  I’m sure there are many other aspects where we can do better. Ideas
  about this ?

Best,

--
Cayetano Santos
.
gpg: CCB8 1842 F9D7 058E CD67 377A BF5C DF4D F6BF 6682
key: meta.sr.ht/~csantosb.pgp

Attachment: signature.asc
Description: PGP signature

Reply via email to