So, I have a lot of knowledge invested in T4. I wrote several code generators of varying complexity and, as a rule, I love it with the caution of code-generating code, which is difficult for unit test (but I even worked on it a little). Now, though, I have heard more and more about the Rosyln project. I heard a lot of things that he supposedly does. Syntax analysis, use as a scripting language, etc., as well as some code generation (although I think this happens at runtime, not at compile time)
How do these two very different technologies add up? Also, today T4 is stable, what should normally be the βdefaultβ for code generators?
In what ways do they differ, and what are the pros and cons with any method?
Earlz
source share