For a really fun time, search the `jinja` and `mar...
# random
b
For a really fun time, search the
jinja
and
markupsafe
repos for issues mentioning
soft_unicode
Oddly someone argues that since semver gets broken by people, they too can break semver. In a very roundabout way
c
what? there’s no logic in that
b
Specifically they broke semver, but argue that it's the users fault for not pinning dependencies and references Additionally, please read hynek.me/articles/semver-will-not-save-you
Which honestly doesn't sit right with me because even if you follow what they say and
use a tool like pip-tools to pin your dependencies and control when you get updates
Then I'm still broken when I try to update, and there's no offered help on solving that
The fact that https://peps.python.org/pep-0440/#compatible-release exists says, roundabout, that a
2.1
shall not break from
2.0
, etc...
c
Ah, interesting read, and some valid points but I’ve not yet read the whole thing.. but surely semver or not if you don’t pin you’re toast sooner or later. And upgrading will always be painful to some degree regardless of semver as well, imo. So yea.. 🤷
b
"This fire is already burning half the city, so what's me adding this stack of logs to it" 🧐
c
If things break because you rely on a bug, then that breakage is on you, not the maintainer.. but it’s still as painful.
(when fixed)
b
In this case,
jinja
didn't specify a good version specifier.
jinja
depends on
markupsafe>=0.23
. That says "I'm compatible with every version of
markupsafe
above `0.23`", which isn't true. It isn't compatible with
2.1.0
most notably
It's also awkward because the same people own flask, jinja, and markupsafe