I work with deep learning models everyday, and in order to do this efficiently I ssh into multiple computers on my network at the same time.
For example:
- Computer A: Running a training script
- Computer B: Running an inference script
- Computer C: Creating synthetic data for the next training session.
- Host Computer: Coding
The code that I am running on Computer A, B, and C is pushed to GitHub from my Host Computer and pulled on each target computer. My Host Computer is used only by myself, so it wouldn't be that inconvenient to use token authentication on my Host Computer. However, Computer A, B, and C are all computers that are shared by several members on my team, and there can be multiple people connecting to any given computer via ssh at any given time.
Consequently, it is necessary that everyone authenticates git operations with their own username and password, so using a credential helper on Computer A, B, and C is not an option. (I don't want one of my team members to accidentally make a push using my git account.)
However, starting August 13, 2021 I will be forced to use token authentication. I can't use a credential helper on Computer A, B, and C. At the bare minimum, it is necessary for at least the username to be prompted for each git operation. Will I need to copy and paste my token from a text file every time I do a git operation on Computer A, B, and C? Since I'm working with 4+ computers at any given time, I have to push and pull from GitHub many times everyday. Copying and pasting an access token from a text file every time is too inconvenient for me. I need a workaround.
question from:
https://stackoverflow.com/questions/65858829/git-password-deprecation-is-my-use-case-still-supported 与恶龙缠斗过久,自身亦成为恶龙;凝视深渊过久,深渊将回以凝视…