Repository navigation
docs(windows): PyPI is current, so fix the reason rather than the advice - #95
Merged
Merged
Conversation
WINDOWS_SETUP.md told operators the published release was "well behind this repo" and still carried 0.1.0. True until an hour ago; false now that 0.3.0 is published. The advice it justified is still right for a producer, so this changes the REASON rather than the recommendation. A producer keeps installing editable from its clone, because deployment on that box is `git pull` (twice today, that is how fixes reached the scheduled tasks) and because the Task Scheduler wrappers live in the repo under docs/examples/windows/. Pinning a machine that runs unattended daily captures to release cadence is the wrong trade when what a broken run misses is not re-fetchable. Read-only consumers have neither constraint, so their section now says `pip install cotdata` outright instead of promising it "once a current release is on PyPI". Also fixes three troubleshooting steps that told you to install the PyPI build over an editable producer install. That was already inconsistent with Step 3, and it is a live footgun now that the PyPI build exists and resolves: it would silently replace the clone-linked install, after which `git pull` would stop taking effect on the producer with nothing to indicate why. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
WINDOWS_SETUP.mdtold operators the published release was "well behind this repo" and still carried0.1.0. True this morning; false now that 0.3.0 is on PyPI.Changes the reason, not the recommendation
A producer should still install editable from its clone, but for reasons that survive PyPI being current:
git pull. Twice today that is how fixes reached the scheduled tasks. Pinning a machine that runs unattended daily captures to release cadence is the wrong trade when what a broken run fails to capture is not re-fetchable.docs/examples/windows/, so a producer needs the clone anyway.Read-only consumers have neither constraint, so that section now says
pip install cotdataoutright instead of promising it "once a current release is on PyPI".Also: three troubleshooting steps that would break a producer
They told you to run
pip install cotdata/uv pip install "cotdata[norgate]"whenimport cotdatafails. That was already inconsistent with Step 3's editable install, and it is a live footgun now that the PyPI build exists and resolves: it would silently replace the clone-linked install, after whichgit pullstops taking effect on the producer with nothing to indicate why. They now point back at Step 3.Swept the rest of
docs/andREADME.mdfor the same claim; the only remaining mention is in the 0.3.0 changelog entry, which is correct as history.README.md's quickstartpip install cotdataneeded no change and is newly true.🤖 Generated with Claude Code