Months (2016)
Months (2015)
Tools
History
edit SideBar
|
Evaluation
- what is my current experimentation velocity?
- did it peak?
- if it did, was by a messy (or rather too complicated) code?
- messy (or rather too complicated) code repository?
- probably too complicated because of 2 highly major changes tried to be done at the same time
- could the peak be resulting from own cognitive limitations?
- something similar to the 5 +/-2 by applied to code software development?
- do I respect http://fabien.benetou.fr/Fabien/Principle ?
- did branching broke too much and if so, how can I rollback?
- mostly a problem of tracking the database content
Consequent next steps
- merge back
- what are the differences between the branches
- master : (looks identical to AllowingMultipleParents)
- NativeGenerators : Fork/Generate as generators but not properly working
- AllowingMultipleParents : trying to fix the models and database because of the new merge function
- are some branches actually not worth saving
- what about simply dropping all branches but master
- after the merge make sure the database is being tracked
- as changed to the models force the database to be tracked too
- it is not any more in .gitignore
- consider then what is the next exploration to try (following SRE principle)
Current situation
- possibility to fetch images and blend them
- like it was done in January 2015
- but via web interface
- so potentially able to get feedback from users
- possible to change size
- parent tracking broken
- model somehow not tracking multiple parents
- maybe data loss since last migrations
- not tracking multiple parents from merge
To do by priority
- step back, what would bring the most fruitful feedback to gather already
- make it more visual
- drag&drop of images on top of other in gallery
- impossible to
- merge generated images
- properly track novelty
|