bug#30513: Unicode Character Name is misspelled (MATHEMATICAL ITALIC SMALL LAMDA)

[Marcin Borkowski]

On 2018-02-23, at 21:15, Drew Adams <drew.adams@oracle.com> wrote:

>> > I would not make the point that this adds too many chars
>> > for `ucs-names' or for `C-x 8 RET'.
>> >
>> > I would make the point that we should not be inventing
>> > character names and then associating such inventions with
>> > what has heretofore been a pretty faithful reflection of
>> > the Unicode standard.
>> 
>> How about this one?
>> 
>> ︘
>> PRESENTATION FORM FOR VERTICAL RIGHT WHITE LENTICULAR BRAKCET
>>                                                          ^^
>> And what about this one?  Press C-x 8 RET MATHEMATICAL ITALIC SMALL TAB
>> and try to answer the puzzle: where has the "MATHEMATICAL ITALIC SMALL H"
>> gone?
>> 
>> (The answer, rot'd-13 so that I don't spoil it for Unicode wannabe
>> detectives;-): CYNAPX PBAFGNAG, U+210E.)
>> 
>> IOW, I would argue that _some_ kind of system to help the user
>> overcome the inherent Unicode problems might be a good idea.
>
> Agreed: some help would help. ;-)  But not at the cost of
> changing `ucs-names'.
>
> You snipped most of my post, including the part that said
> that although we should leave the set of Unicode names as
> Unicode defines them, so that `ucs-names' remains faithful
> to the standard, we can certainly add Emacs constructs (e.g.
> commands, completion functions, whatever), to help users
> use alternate names of our own invention, including spelling
> corrections.
>
> The fault is not with `ucs-names'.  The fault, if there
> be any, is with the ways we currently _make use of it_
> for users.

I agree.

> We could offer additional or alternative ways for users to
> make use of it.  We could, for example, change `insert-char'
> to respect a user option that expresses just how much such
> help to provide, e.g., the degree of spelling help,
> correction, abbreviation, or whatever.
>
> If we do that then we should at least allow one of the
> option values to mean that no such help is to be offered,
> in which case `insert-char' would offer only the official
> names.
>
> There are other uses of `ucs-names', beyond `insert-char',
> at least in 3rd-party libraries.  We should definitely not
> assume that all uses of `ucs-names' should benefit or be
> troubled by any Emacs-specific "improvements" we might
> want to offer for the available char names.

+1.  For one interesting use of ucs-names, see my blog post here:
http://mbork.pl/2017-10-02_Converting_TeX_sequences_to_Unicode_characters

Best,

--
Marcin Borkowski