Running docker-compose up mfr_requirements fails because it waits for user interaction. Removing -e lets pip install aiohttppretty without failing because pip doesn't know what to do.
Changes
Remove the -e flag from the requirements file
Side effects
This flag looks like its been here since the dependency was first added, but there is no reason stated. There may be a reason its there? If so we can look at --ignore-installed as an alternative
Coverage remained the same at 72.448% when pulling 6ba2bcef2779adde68a0ee6d1dd53a5dd00e9785 on birdbrained:bf/fix-req-install into f943de8bfcd334cbf85ffae47f4d73efb57226fa on CenterForOpenScience:develop.
Coverage remained the same at 72.448% when pulling 8c3f9c62265e6d955c407661987fed75e8f226c3 on birdbrained:bf/fix-req-install into f943de8bfcd334cbf85ffae47f4d73efb57226fa on CenterForOpenScience:develop.
My bad, I thought my newline was putting it in the body not the subject. (Normally I don't use commit -m but an editor to write commit messages. Serves me right for trying something new.)
Purpose
Running
docker-compose up mfr_requirements
fails because it waits for user interaction. Removing-e
lets pip install aiohttppretty without failing because pip doesn't know what to do.Changes
Remove the
-e
flag from the requirements fileSide effects
This flag looks like its been here since the dependency was first added, but there is no reason stated. There may be a reason its there? If so we can look at
--ignore-installed
as an alternativeQA Notes
Probably no.
Deployment Notes
Probably no. This only affects dev environments.