You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Janos and I thought about the future of the econ-project-templates given the experience of the 2023 EPP iteration:
Problems
Students do not understand cookiecutter dialogue
Students do not understand the "concept cookiecutter" (e.g., students don't grasp the
difference between project and cookiecutter)
Connecting the downloaded project to GitHub often fails (especially if the remote is not
empty) and is perceived as very complicated even if it works (e.g., clone link vs URL)
Example project is out of date, especially data management (should be done by Hans-Martin and
Tim)
When cookiecutter fails, or you want to edit previous input, it is very annoying to
start over (fails, for example when the folder already exists, the environment already exists)
There is the replay option, but it is unknown, and does not seem to work perfectly
Observations
Cookiecutter replaces something students like to do: copying files and adjusting names
with something they do not understand and do not like to do
Students are usually not informed enough to make decisions about e.g. licenses, make
initial commit, which types of pre-commits, creating the environment or not. It seems
like fewer choices would make students feel better while doing the setup.
The Julia, Stata and R examples are not used and they are another part of the code
base that needs to be kept up-to-date (we could just reference the pytask
documentation)
Solution Ideas
Replace cookiecutter dialogue by minimal form:
project slug (-> rename to project name)
author name
author email
Form checks inputs in real-time, after filling it out, there appears a download button
Reference to videos like "how to create environment" appears
Workflow:
Students create repo on github
Students clone empty repo
Students download files from example project into empty repo
Students push added project to remote
Additional (Potential) Developer Benefits
Removing the example projects for other programming languages makes it easier to
maintain
Removing cookiecutter complexity makes it easier to test and maintain the example
project
The text was updated successfully, but these errors were encountered:
As discussed off-line, this generally sounds like a good plan. Thanks for bringing it up!
Let's tackle it at some point this year.
Probably this should be converted into a template repo then. And we could have a simple Python script as part of the repo, which does the adjustments like
Janos and I thought about the future of the econ-project-templates given the experience of the 2023 EPP iteration:
Problems
difference between project and cookiecutter)
empty) and is perceived as very complicated even if it works (e.g., clone link vs URL)
Tim)
start over (fails, for example when the folder already exists, the environment already exists)
Observations
with something they do not understand and do not like to do
initial commit, which types of pre-commits, creating the environment or not. It seems
like fewer choices would make students feel better while doing the setup.
base that needs to be kept up-to-date (we could just reference the pytask
documentation)
Solution Ideas
Additional (Potential) Developer Benefits
maintain
project
The text was updated successfully, but these errors were encountered: