Stephen Sugden

> but the processed of application would not (since none of its imports had changes in the raw digest)? The processed digest of application should depend on the processed digests...

> what are you thoughts? Even though I wasn't asked, I would like to contribute some input on this topic, as it's something I've been spending quite some time on...

> Your first case could be solved by not only creating the sha of the actual line content but also the details I just want to mention that this is...

> would using that sha of the whole statement + field sha + line index help? Yes, I meant for `sha` in my suggestion to refer to the SHA of...

A further note about `own_hash` above: it would be a **really good idea** if the various `Field` types worked off of **frozen** strings representing unaltered field lines. Then they could...

I can reproduce this issue with erb-formatter 0.7.3 and syntax_tree 6.2.0 Given the following snippet, the block will be removed. ```erb # The entire block will be removed resulting in...

> My concern is, what if we can't implement `has_many :through`? Should we implement `has_one :through` instead? I'm worried that users might assume `has_many :through` works the same way as...

@Jay0921 I want to say thanks for taking this on! Unfortunately I'm about to go on vacation until August, so with the usual pressure of wrapping things up on my...

Also here because CI workflows suddenly stopped working. Thankfully we didn't deploy anything broken 😅 I'd like to echo the other suggestions that an explicit path to a hidden file/directory...