Re: [Q] Tools for Ada Quality and Style

[dewar@cs.nyu.edu (Robert Dewar)]

Bob said

"I believe you'll search in vain for any words of mine that say the
way to implement this process improvement is exclusively or even
chiefly through a prettyprinter."

Oops, sorry I must have slipped in copying the attribution, it's easy
to do sometimes when quoting.

However, I must say, I don't like ANY tools that mess with what I type.
Even tiny ones! If you rely on even tiny tools to clean up your sources,
it means that you are tolerating at least for a limited time, untidy
sources, and I think the best path to uniformity is to foster an
attitude that just can't *stand* to look at code that is not properly
formatted. 

But that's really a small taste matter, basically I think we are pretty
much in agreement.

For GNAT itself, the compiler enforces many layout rules, and it is not
even possible to compile a GNAT unit that does not obey the basic style
rules. This has been very useful in getting a high degree of uniformity
in the GNAT sources themselves.