Re: [ANN] Lua 5.3.0 (work2) now available (number coercion holy crusade)

[Coroutines <coroutines@...>]

On Mon, Mar 24, 2014 at 5:41 PM, Rena <hyperhacker@gmail.com> wrote:
>
> On Mar 24, 2014 6:34 PM, "Hisham" <h@hisham.hm> wrote:
>>
>> On 24 March 2014 10:47, Roberto Ierusalimschy <roberto@inf.puc-rio.br>
>> wrote:
>> >> Most common breakage seem to be (%d+) captures. [...]
>> >
>> > That is my experience too. (Not sure about the "most common", but sure
>> > about a "quite common".)
>>
>> I just hacked together this super-ugly patch [1] (Roberto's patch plus
>> a hack to make tonumber work) just to run the LuaRocks test suite with
>> coercions off and see what happened.
>>
>> The results: it ran successfully! Looks like my strongly-typed brain
>> hasn't failed me and I've been treating Lua as if it never had
>> string-to-number coercions in the first place.
>>
>> I just checked the code and all my (%d+)'s were already religiously
>> followed by uses of tonumber.
>>
>> If the addition of integers will make people review string conversions
>> in their code anyway, perhaps now is a good time to at least deprecate
>> automatic coercion, giving people a #define so they can test their
>> code without it and prepare for the future.
>>
>> I haven't tried to run the LuaRocks testsuite with 5.3 yet because it
>> uses a number of rockspecs which are not marked to be 5.3-compatible
>> and therefore will fail with an "unsupported version" error message.
>> Porting them all is a larger task.
>>
>> -- Hisham
>> who really likes a disciplined, typed world
>>
>> [1] https://gist.github.com/hishamhm/9750528
>>
>
> I feel like format strings are one place where coercion is not a bad thing.
> It's all about making concise code that's easy to read and understand. When
> you write string.format("%d", x) it's clear that you intend x to represent a
> number, so having an implicit tonumber() around it makes sense. (Plus if x
> is, say, a table, an implicit conversion can throw an "expected number, got
> table" error,while if I'm remembering correctly tonumber(x) would return
> nil, making it harder to track down the bug.)
>
> In fact I don't see why such coercion is a bad thing in most cases, since
> context usually exists. You know that the operands to + must be numbers and
> .. must be strings.

For the record, I also like coercion.  It's something I welcome in a
scripting language.

   -- but jeez if it weren't there I would be oh too happy to set fire
to the .. operator.  Like all good, red-blooded American boys I like
to space out my operators and having ' .. ' between strings is such an
awful waste of space.  At least make it single-char like $ or
something.

(sorry for diverting the focus of the thread)