Re: HELLO changes

[Dave Love]

Juri Linkov <address@hidden> writes:

> Good, but you have mixed uppercases and lowercases in capital letters of
> language names: you changed most of the names to lowercase, but kept
> uppercase for Czech, Estonian, German, Slovak, Slovenian, Swedish.  You
> should choose either uppercase or lowercase and stick to it in all names.

Why?  It is clearly wrong to capitalize according to my Oxford
English, even according to the only relevant English reference I have
(Hart's Rules).  How exactly is ICU wrong and what authority should be
used instead?  If it's wrong, could you convince the maintainers of
the locale data with which I'm trying to be consistent?

> Also you added RTL name for Hebrew, but text sample is in LTR, because
> it was decided to write it in logical-order until BIDI display
> will be supported.  So the language name should be changed to LTR.

I don't think it's helpful to present a backwards name to users.
(These names are actually ones I collected to be able to present the
names of language environments to users in their native form, but I
haven't added the code for that yet.)

> I think it's too early to remove Unicode section, because it still
> produces different visual results than text in non-Unicode.

I think you're confusing Unicode characters and a particular Emacs
representation of them.  I've put a fair amount of infrastructure in
place so that users can be mostly protected from the distinction.  The
mule-unicode charsets are exemplified by the maths I added.  The
iso8859 parts of mule-unicode may or may not be displayed differently
from the iso8859 charsets.  Indeed, you probably want them displayed
the same if you're using coding systems which mix, say,
latin-iso8859-1 and mule-unicode.  If you want to see how a specific
internal charset is displayed, you can use list-charset-chars.

> This is good that you've removed FORTRASH, because it can't be mentioned
> without insult to all hackers.

I'm sorry I've been so insulting, especially by maintaining GNU
Fortran.

> Since this name was added for fun, it should
> be replaced by C.  Moreover, C is the default locale!

That seems a good half pun, but rms is the pun master.

> For example, Finnish has now ASCII-only informal greeting "Hei".
> I propose to add more formal greeting "Hyvää päivää" (Good afternoon)
> which has a-umlaut.  For Estonian I propose to add "Tere päevast"
> (Good afternoon) with a-umlaut, and "Tere õhtust" (Good evening)
> with o-tilde.

> For example, Swedish sample text could also have "hej då"
> or something other word containing a-ring letter and other non-ASCII letters.
> For Danish "En øl værsgo" could be added with o-slash and ae, and so on.

I'm not sure how more examples of unlauts and tildes help.  Are
multiple Latin-1 examples really useful, especially multiple examples
of, say, Nordic languages?  I don't feel strongly about it.

> Note also that all changes in HELLO file should be duplicated
> in `sample-text' properties of every `set-language-info-alist'.

I'm not sure they serve the same purpose.  The sample text would be
better showing the exemplar characters (or whatever the correct term
is) of the language.

-- 
ALGOL 60 is alive and well and living in FORTRAN 90.  -- Tony Hoare
Fortran 90 is a very pleasant language.  -- Richard O' Keefe