User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.6; rv:6.0) Gecko/20110812
            Thunderbird/6.0
MIME-Version: 1.0
References: <o28feg4sv686.dlg@nililand.de>           
            <35FEA381-6EA8-4464-9E0F-C504C2D92641@gmail.com>
            <CANQYN6yH95UQCFKCmLNEaVa0NdtpdUQiy2m5xKMF=gyLrS+LDg@mail.gmail.com>
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
Message-ID:  <4E59601C.5010209@morningstar2.co.uk>
Date:         Sat, 27 Aug 2011 22:22:36 +0100
Reply-To: Mailing list for the LaTeX3 project
              <LATEX-L@listserv.uni-heidelberg.de>
Sender: Mailing list for the LaTeX3 project <LATEX-L@listserv.uni-heidelberg.de>
From: Joseph Wright <joseph.wright@MORNINGSTAR2.CO.UK>
Subject: Re: l3keys: mirror boolean key
To: LATEX-L@listserv.uni-heidelberg.de
In-Reply-To:  <CANQYN6yH95UQCFKCmLNEaVa0NdtpdUQiy2m5xKMF=gyLrS+LDg@mail.gmail.com>
Precedence: list
Status: R

On 27/08/2011 13:16, Bruno Le Floch wrote:
> On 8/27/11, Will Robertson <wspr81@gmail.com> wrote:
>> On 26/08/2011, at 3:04 AM, Ulrike Fischer wrote:
>>
>>> Now I want to define a "mirrored" key
>>>
>>> no-sym
>>>
>>> so that "no-sym = true" sets  \l_chessfss_sym_bool to false.
>>>
>>> Is there some easy way?
>>
>> Not directly as far as I know (using
>>
>>   .code:n = { \bool_set_false:N \l_chessfss_sym_bool }
>>
>> works of course but it's not the best interface), so I think this would be a
>> good addition. (I seem to remember kvoptions has a similar feature.) Which
>> do you think is best to name it?
>>
>>   no-sym .bool_unset:N = \l_chessfss_sym_bool
>>   no-sym .bool_set_false:N = \l_chessfss_sym_bool
>>   no-sym .bool_set_reverse:N = \l_chessfss_sym_bool
> 
> I find .bool_set_reverse:N more descriptive (cf. \reverse_if:N), as
> no-sym = false is valid user input, I guess?

Early-on in the life of l3keys, we did have '.bool_set_<something>:N'
(where I forget exactly what <something> was). The reasons we did not
keep it then were:
 a) The idea was to keep things simply initially;
 b) The other variable-setting functions all mirror the base expl3
    names.

Now, boolean keys are a bit odd, as they are actually a special kind of
choice (that's how they are implemented). So it may well be best to
provide this ability. I guess I favour '.bool_set_inverse:N'.
-- 
Joseph Wright