Keyboard layouts and CLDR (was: Re: 0027, 02BC, 2019, or a new character?)

[Marcel Schneider via Unicode]

On Mon, 29 Jan 2018 16:07:11 -0700, Doug Ewell wrote:
> 
> Marcel Schneider wrote:
> 
> > Prior to this thread, I believed that the ratio of Windows users
> > liking the US-International vs Mac users liking the US-Extended was
> > like other “Windows implementation” vs “Apple implementation” ratios.
> 
> For many users, it may not be a question of what they like, but rather
> (a) what they are aware of, (b) what comes standard with their Windows
> installation, and (c) in the workplace, what their IT overlords have
> granted them permission to use.

c: Hierarchical relationships may be complicated in some places but generally 
there should be an open door, and suggestion boxes or pinwalls may also be 
available. As an “overuser” the IT manager must think and evaluate in the place 
of his employees/coworkers, but professional stress and the desire to unplug 
during week-ends could be mainly responsible of his unawareness. Though I 
thought that for professionalismʼs sake they should deploy an appropriate 
layout fork, and I canʼt see any point in not using MSKLC at that level.

b: There is often much reluctance to add only an extra font, and I hesitated a 
long time prior to installing Firefox or any other extraneous (third-party) 
software, but that is actually so common — Chrome browser doesnʼt ship with 
Windows neither but weʼre often encouraged to download it — that failing to 
update oneʼs keyboard is due to a lack of marketing and thus, resolves to point (a).

> 
> I use a modified version of John Cowan's "Moby Latin" layout on all my
> machines:

It would be interesting to know more about your modifications.

> 
> http://recycledknowledge.blogspot.com/2013/09/us-moby-latin-keyboard-for-windows.html

Sadly the downloads are still unavailable (as formerly discussed). But I saved 
in time, too (June 2015).

> 
> which allows me to type about 900 characters *in addition* to Basic
> Latin, with 100% backward compatibility with U.S. English (i.e. none of
> the apostrophe and quotation-mark shenanigans we are talking about). But
> (a) I happen to know about Moby Latin, (b) it doesn't ship with Windows,
[Addition on Mon, 29 Jan 2018 19:08:55 -0700:
> Of course that is not a "luxury." Knowing that third-party options are
> available, let alone free and easily installed ones, is the luxury. 
]
> and (c) I am able to install it (and even modify it). Many users do not
> have all or even any of these luxuries.

I agree. It is a blessing to be able to fine-tune oneʼs keyboard. Often those 
knowing about correct diacritics but not about keyboarding canʼt help clicking 
in charmaps and symbol dialogs at document length. That is utter time waste.
And thatʼs what is wrong about letting people waste their time while knowing 
better.

> 
> There is perhaps another factor: many Americans, who are probably the
> majority users of US-International though not the only ones, simply do
> not know or care about accents and other "foreign stuff." Even those who
> know a language other than English often write it in ASCII, and see it
> that way in marketing and other professionally created material. For
> example, menus in Mexican restaurants often list "albondigas" and
> "jalapenos."

French too, though being an accented language/script, is prone to omitting 
other localeʼs diacritics as far as they are supposed not to be on the AZERTY.
(It happened that a tilde was refused for lack of support, while we do have 
that dead key!) The more when accents merely indicate correct intonation, 
as in “albóndigas.” When reading “jalapenos” we reflexively add the tilde for 
spelling. Compare with locales writing consonants only. Writing in ASCII is 
also a sort of assimilation, as we all like to name things in our own language.

> 
> The non-phonetic spelling of English may further encourage English-only
> speakers to ignore the squiggles and dots that are necessary to indicate
> correct pronunciation of other languages.

That can be OK for common words, but gets dangerous when it comes to 
proper names. Thereʼs a slippery path from easy to sloppy. Itʼs still about i18n.
But I wasnʼt aware that by not using diacritics and having to do much effort 
to remember correct spelling, English-speaking people may really hate those 
diacritics even when occurring on foreign stuff.

> 
> Given that, interest among potential users of US-International to find a
> better solution is probably very low.

I was about to make a quick update of the US-Intl [or ABC-Intl], but 
if so, I could eventually save that for now.

> 
> > If so many people like it, why was Windows Intl not updated, then?
> 
> 1. I'd be surprised if there were "so many people," or much demand to
> update it. Microsoft might have a few other items on their backlogs.

However, resulting from the KLC files you kindly provided me, as of Latin 
keyboard layouts shipped with Windows, from Windows 7 to Windows 8, 
eight layouts were updated, including United Kingdom, Turkish, German 
and Inuktitut, both variants for each locale. Not US. For completenessʼ 
sake, weʼll also metion that five new layouts were added:
Azerbaijani (Standard)
English (India)
Hausa
Hawaiian
Latvian (Standard)

Soon, thanks to enriched CLDR, many many more should be released by 
the means of Windows Update.

> 
> 2. I don't speak for Microsoft, but there is often fear of making
> changes to existing standards, even changes that fill in holes in the
> standard. Users who type a formerly invalid sequence and get a valid
> character, instead of the beep or question mark they once got, and
> complain about the change, might seem to be a low-priority constituency,
> but you'd be surprised.

Indeed, I am!

> 
> > To like a particular layout does not mean to want to stick with it
> > when anything better comes up. Userʼs choice is always respected.
> 
> See above regarding what users might like if only they had a choice.

We wonʼt blame end-users sticking with the choices of other people theyʼre 
paid by, and who are those customers who could request changes that 
vendors would be ready to accept but who presumably donʼt for whatever 
reasons.

Regards,

Marcel