ManageIQ switched recently to use only haml files instead of html.
While using haml still might benefit the majority of the files in the project that has been short anyway even before the haml conversion, the situation with longer and more complicated files is different.
There are several challenges in using haml among which:
- haml is whitespace sensitive/indentation sensitive which I find hard to rely on during development = it’s very fragile.
- no closing tags require to track really closely the tabs to see where sections actually end (as opposed to closing tags in html that mark that part clearly) - easily getting lost after 2-3 hierarchies down. so readability of the files is harder.
- it adds an extra step that converts haml to html - potential penalty on large/complicated files
- harder to convert more advanced use cases that involve angular directives, etc. (and less documentation/examples online for that)
- some of the haml files don’t contain RoR code which was the main use case for using haml.
So given the above, I’d like to propose to allow the usage of both haml and html in the project - specifically allowing html for the longer and more complicated use cases.