Re: Knowing where a function has been used (bis) [Was: Re: Optimising Elisp code]

[Eric Abrahamsen]

Barry Margolin <barmar@alum.mit.edu> writes:

> In article <8636tcl05a.fsf@zoho.com>, Emanuel Berg <moasen@zoho.com> 
> wrote:
>
>> Barry Margolin wrote:
>> 
>> > I doubt that anyone makes decision about
>> > whether to split large functions into
>> > separate functions based on the overhead of
>> > function calling.
>> 
>> This is what the Gnus people told me when
>> I suggested not writing functions that are
>> longer and deeper than a novel by Dostoevsky -
>> they said, ~"Gnus is slow, Elisp is slow,
>> making neat modules of those defuns will make
>> it even slower."
>
> Sounds like a bullshit rationalization. The real reason is probably that 
> refactoring is a PITA.

As someone who's currently trying to refactor Gnus (replacing obarray
hashtables with actual hashtables), I can confirm that it is indeed a
PITA.