Re: [boost] Re: Re: Re: FORMAL review of "Output Formatters" - results

From: John Torjo (john.lists_at_[hidden])
Date: 2004-10-25 10:43:24

>>
>>>- Should the library use locales and/or iword/qword?
>>
>>                                                 ^ pword, that is
>>This is an implementation detail.
> 
> 
> How? The difference between:
> 
>   locale::global(locale(), new vector_io_facet(.......))
> 
> and
> 
>   cout << vector_io_format(......)
> 
> is in interface, not implementation.
> 
> 

What I meant was that you can have an interface similar to:

set_format( std::cout, "[%,%]" /* or whatever */);

In this case, whether the implementation uses locales or not can be an 
implementation detail.

Of course, you can choose to make it part of the interface as well.

> 
> 
> Hmm... are you about to write some code? Isn't this a bit contrary to review
> manager role -- he's supposed to be non involved in the development of the
> library that's being reviewed.

I assumed this to be a *different* lib (as Jonathan proposed).

Certainly if you think it should be part of OutFmt lib, then I will 
_not_ write anything.

In fact, I will step back, and _not_ write anything - period.

The above was supposed to handle IO, while Reece's lib should handle 
Output-only.

> 
> 
>>And just to make it clear once again:
>>The "Output Formatters" library *should not* provide Input facilities
>>(Reece, please take note!)
> 
> 
> Now, I don't want to start a revolution, and I don't have anything against
> you in personal, but there are some points:
> 
> 1. The library is rejected. Only another review can make it enter Boost.

Yes

> 2. We don't have a rule that re-review must be managed by the same person
> who managed the first review. We don't have any "library won't be

True, sorry about that.

> re-reviewed unless do change this and that" resolution. So I don't think

True.

> it's OK to require Reece to remove input facilities. I don't think input

Fine, anyway, if he wants to provide them, it's his decision.

Doing a recount of who whether input facilities are wanted or not - 
there were 4 votes for "No", 1 in-between (Paul Bristow), and 1 for "Yes"

> 3. It's a bit problematic that you start to take part in design
> discussion/development.

Fine, I will step back.

Best,
John

-- 
John Torjo,    Contributing editor, C/C++ Users Journal
-- "Win32 GUI Generics" -- generics & GUI do mix, after all
-- http://www.torjo.com/win32gui/
-- v1.5 - tooltips at your fingertips (work for menus too!)
    + bitmap buttons (work for MessageBox too!)
    + tab dialogs, hyper links, lite html