-
Notifications
You must be signed in to change notification settings - Fork 279
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Setup #61
Setup #61
Conversation
Just some mental notes of what doesn't work yet
|
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Overall, it seems a pretty tidy transition as it should be. Thanks! Don't let the quantity of minor comments etc express otherwise. And yeah, in work etc. I prefer to share early and often in case there are significant issues that should be addressed early. Handily, I don't think there are any here.
I'll note here that this includes the changes in #60. If this is merged first it is accepting both changes. I don't have a problem with #60, just pointing it out.
MANIFEST.in
is for the sdist, IIRC. If you specify include_package_data = True
then it will make it into the wheel as well and thus the installation. IIRC... I think this may also replace [options.data_files]
. But it's bin awhile since I've fiddled with this and it's one of the areas where there are many ways to do it and most have gotchas.
Maybe https://pypi.org/project/appdirs/ and a CLI override option for config file locations? Though this strikes me as maybe a separate issue unless installation breaks the existing functionality? I should go see what it is.
Maybe src/plotman/plotman.py
should be core.py
or something? Module paths like plotman.plotman
strike me as a bit odd.
I would avoid encouraging people to activate the environment when running. There's plenty of room for personal preference on whether to activate or not but from the perspective of supporting people it ends up adding hidden global state in $PATH
that you have to debug. Whereas if people send you a traceback with venv/bin/plotman
then you know they are running the plotman in the env.
Looking good for a first share. I think this is one of the first things I poked @ericaltendorf about getting done with plotman. Good to see it happening!
Co-authored-by: Kyle Altendorf <sda@fstab.net>
@jkbecker, for getting the config from the 'right place', I think that can be a separate PR? This is already a restructure so keeping it focused is more valuable than normal. Also keeping it from getting held up on other changes is good. Unless I'm misunderstanding and the config location is particularly affected by this? From a quick search it seems to just be looking in cwd right now. |
The main reason I wanted to couple those topics is: Once plotman is properly installed, it can be launched with the That said, considering I'll probably work on that next anyways, we could separate it as well? There are some considerations when it comes to "correct" config file location handling.
Does this make sense? Am I missing a special case? Happy to deal with this in a separate PR if it makes the individual contributions more manageable, either way is fine for me. |
Agreed on all points. I'm not sure if you said this indirectly or not, but |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I'll commit the minor tweaks tomorrow after reviewing my own review at a more appropriate hour. @ericaltendorf do you want to discuss the effects of this on users before merging to develop? No more launching as ./plotman.py
and such. Or, is the addition to the README sufficient?
Thanks again @jkbecker, definitely happy to have this change. :]
I think the addition to the README should be sufficient, along with an announce on the plotman keybase channel. I guess improving the (FWIW what I've been doing so far is creating a sibling directory (e.g., |
typo, thx for noticing Co-authored-by: Kyle Altendorf <sda@fstab.net>
Co-authored-by: Kyle Altendorf <sda@fstab.net>
Co-authored-by: Kyle Altendorf <sda@fstab.net>
Co-authored-by: Kyle Altendorf <sda@fstab.net>
such as with --editable installation
I guess I may as well draft the keybase message here. Suggestions and corrections are of course appreciated. Or, you could handle this if you want, @jkbecker. Hi all. @jkbecker has kindly reworked the project structure of plotman for us so that it will be a properly installable Python package. This is being merged into the Now that I've mentioned how the |
Looking good @altendky, thank you! I prefer not to announce it myself, as I'm not previously associated with the project. Don't want this to look like an intrusion ;) ❗ One detail just came to mind: If people just read the README on GitHub and Or are we assuming (for now), that people installing the development branch are existing users and somewhat more experienced with how Plotman works? At the very least, #84 should be addressed before merging to main. |
It's not like I have some long standing involvement with plotman either. |
Hm, random idea: Maybe it would be a good idea to merge the current development branch progress to main, before merging setup into development? I feel like there are a lot of small improvements in development that could be helpful now, people are complaining about bad GUI plot handling almost every other hour for example. Merging setup onto development would just make merging develop on main a much bigger change, right? |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
The request for more commentary on config.yaml
is worthy of being a requested change I think. Even though I plan on doing it.
A setuptools-based installation script as well as the required re-structuring of the entire project.
See updated README.md for the new installation instructions once this is in effect.