Re: [Emacs-diffs] master b88e7c8: Make transpose-regions interactive (Bug#30343)

[Stefan Monnier]

I haven't really participated so far since it's definitely a bikeshed's
color kind of discussion, but since I'm to blame for it, I'll try and
explain why I brought it up:

I said I "dislike seeing" such code, without any further explanation
because indeed, there's no strong technical reason behind it (I could
come up with various technical reasons, but they'd just be
rationalizations).  I guess it just reflects my "strong types"
background, where I like my code to be sanity checked by a static
analyzer.  For Elisp we don't have any strong static analyzer available,
so we use the next best thing: our byte-compiler's warnings.

With "Elisp embedded in C strings", we don't have any such tool
support, so it makes me feel "unprotected" (we can't even be sure that
parentheses are balanced without run-time testing).

I didn't bring it up with the aim to try and change that specific chunk
of code, but rather to get the opinion of others, and probably also to
try and influence future code.

>> Yes, they are all weak reasons, but so are the proposed reasons in
>> favor of the change.
> Yes, agreed on both counts.

Clearly.


        Stefan