Re: Fix comment typo in /src/backend/command/cluster.c

On 01/28/2014 08:58 AM, David Fetter wrote:
> On Tue, Jan 28, 2014 at 02:51:22PM +0900, Michael Paquier wrote:
>> On Tue, Jan 28, 2014 at 2:29 PM, David Fetter <david(at)fetter(dot)org> wrote:
>>> On Tue, Jan 28, 2014 at 04:48:35PM +1300, Gavin Flower wrote:
>>>> I came across that abbreviation in a first years Maths course
>>>> "Principles of Mathematics" in 1968 at the University of Auckland..
>>>
>>> By my rough count (ack -l '\biff\b' |wc -l), it's used to mean
>>> equivalence 81 times in the source tree. Should we have a glossary of
>>> such terms?
>> And what about directly replacing those expressions in the comments of
>> the code with some more understandable language? This would be more
>> suited for non-native English speakers than maintaining a glossary
>> that you can surely find here and there after some googling.
>
> I'm interested to find 29 instances of "if and only if" in the source,
> which should be the same thing.
>
> Please find attached a mechanically done patch which expands the
> remaining instances of "iff" to the longer form, all of which are in
> comments. The patched source passes make -j8, but I have not tested
> it further.

"iff" is well-known abbreviation, I don't see a need to expunge it from
the source code. There might be places where some other wording or
spelling it out as "if and only if" would be better, but a mechanical
search/replace is not warranted.

FWIW, many other languages use a similar abbreviation for the same
thing, so it's not impossible for a non-native English speaker with
basic math education to guess. In Finnish, it's "joss", which stands for
"jos ja vain jos", and a quick look at Wikipedia shows a similar
construct in many other languages.

- Heikki