Conversation
Prior to this commit the `trytls/runme` script had a handful of small problems: 1. It assumed Python 2.7 was used for the `pip` install. 2. The upstream `trytls` package errors on Python 3.8 due to use of a now removed deprecated `platform` function. This commit fixes both of these problems. The first, by detecting the Python lib dir that was created. The second, by using a fork of the upstream project that has a fix for Python 3.8+ systems in place. A TODO note is left to remind us to switch back to the upstream once a fix PR lands. These changes are sufficient to get the `trytls` runner working again on a Python 3.10 system.
|
So I've never really gotten my hands dirty with shell scripts (preferring to write Python code instead in most cases), so a lot of what's going on here looks like line noise to me (though it might not quite be Perl). Even for Python, there are alll the problems of keeping up with dependencies and Python versions and the environment. For these reason, in my view we should be reducing the use of Python and sh code in the repo over time. Specifically for trytls, I'm not sure what value it provides at these point. Presumably bogo will cover a lot of the same surface area. We have an |
This line of thought resonates with me and I would vote for removal as well. The upstream trytls project hasn't had a commit in 7 years. If there were tests unique to Trytls I suspect it'd be less long term maintenance work to lift them into the BOGO suite or gin something up ourselves. I'll close this PR and open a follow-up that removes the trytls bits 👍 |
Description
Prior to this commit the
trytls/runmescript had a handful of small problems:pipinstall.trytlspackage errors on Python 3.8 due to use of a now removed deprecatedplatformfunction.This commit fixes both of these problems. The first, by detecting the Python lib dir that was created. The second, by using a fork of the upstream project that has a fix for Python 3.8+ systems in place. A TODO note is left to remind us to switch back to the upstream once a fix PR lands (ouspg/trytls#314).
These changes are sufficient to get the
trytlsrunner working again on a Python 3.10 system.Run Results
Here's the output from `trytls` with these updates applied:
It looks like three tests are ERRORing:
Perhaps someone more familiar with the project can speak to whether these ERROR results are expected, or if further investigation is merited!