From: Charles Oliver Nutter <charles.nutter@...>
Date: 2009-02-26T04:29:56+09:00
Subject: [ruby-core:22496] Re: YASNP (Yet Another Selector Namespace Proposal)

Florian Gilcher wrote:
> 
> On Feb 25, 2009, at 7:46 PM, Charles Oliver Nutter wrote:
> 
>> Florian Gilcher wrote:
>>> On Feb 25, 2009, at 6:17 AM, Yehuda Katz wrote:
>>>> I'm also in favor of discussing this, but all I hear so far in 
>>>> opposition is vague FUD... no specific examples of problems that 
>>>> could be caused. It would be a lot easier to have a lively 
>>>> discussion about specific concerns, and I'd love to have it!
>>>>
>>>> -- Yehuda
>>> Is FUD the new "the opposition doesn't convince me"? I hear it a lot. 
>>> I think it's overused. Also, I find it insulting or at least a 
>>> "verbal" fauxpas as well as misplaced[1].
>>
>> I think the FUD here is "oh no, what if this happens or that happens 
>> and code gets more complicated!". You can what if any feature to 
>> death, but without considering the probability of real issues it 
>> really is FUD.
>>
>> And of course the "what iffers" may not be weighing those scenarios 
>> against the positive uses of namespaces.
> 
> That may be true or not, but it is still against the rules of a good 
> discussion to generally calling it FUD. Disprove the other or ignore 
> him. Accept if he doesn't have your point of view. Tell him that he 
> doesn't convince you. For me, that comment was ad hominem, because it 
> also has generic implications about the others state of mind and behaviour.
> 
>>
>>> The problem is rather obvious: suddently, I can have multiple 
>>> definitions of a method (or none!) depending on context. Nowadays, i 
>>> can just inspect/test that behaviour in IRB/rdb and be sure that it 
>>> is about the same everywhere (safe for the unusual catch that it 
>>> could be un/redefined along the way). With selector namespaces, that 
>>> catch would get commonplace.
>>
>> You can't really control what any library decides to do today, so I 
>> don't think you lose any more control with namespaces. And 
>> monkeypatching code because you want *others* to be affected has 
>> always been a losing proposition...since they may have monkeypatched 
>> code to affect *you* already.
> 
> As i pointed out more then once: I am on the side of having namespaces, 
> i just wanted to illustrate that it is not impossible to find things get 
> more complex. But i was also not speaking about control but about making 
> sense of a current and future state of a system at a given moment.
> 
>>
>>> BTW: how will IRB handle file-based switches?
>>
>> It would need to have modifications to its parser logic to support 
>> namespacing, just as it needed modifications to allow stabby lambdas 
>> through.
>>
>> - Charlie
> 
> Well, thats for sure. Will it be able to switch namespaces, will it not? 
> I wanted to raise the question, not imply that it is hard to answer.

Yeah, I don't see it as being too difficult, though IRB does force the 
issue of "turning namespaces off" at some point, since you may want to 
do that in a long IRB session. And if you can turn namespaces 
off...maybe they should be explicitly scoped :)

- Charlie