Agreed, agreed, agreed.
On machines, or within individual applications, where it's easy for me to
enter diacritics, I almost always do so. If it's easy and available
system-wide, as it has been on the Mac since 1984 for many pre-composed
chars, I'm even more likely to do so.
On Windows, I'm likely to skip the diacritics most of the time because the
US keyboard drivers (mandatory on US development machines that have to
resemble plain-vanilla customer setups), and the rest of the OS, simply make
it too much trouble.
I'm not at all happy about it, though. If it's a somewhat formal document, I
make the extra effort to use the proper chars, but it's not fun. Where I can
make it easy to do without screwing up a development machine, such as within
MS-Word, I also use the proper chars, but it's more fun to do so. Word has
what is effectively an "IME" or input pre-processor that allows you to set
up your own key combinations without monkeying with the system keyboard
driver and screwing up everything else.
I long for the day when the plain-vanilla consumer version of US Windows
will have a system-wide IME, just as all Japanese systems already do.
Japanese, as everyone on this list knows, has a lot more characters than
keys, and they resolve the problem with a smart IME. Once you have such an
IME, you don't have to limit it to character entry. You can create your own
mnemonic keystroke sequences to insert ANYTHING: paragraphs of boilerplate,
today's date, a bitmap of your signature, anything that could be pasted from
It's high time for us to have such an IME as a standard feature of US
Windows. Even English has "more characters than keys", and few educated
English-speakers limit themselves to English words. We need to be able to
easily enter the names and addresses of our customers correctly, both out of
respect and to avoid mistakes that may cost money. And we'd like to be able
enter abbreviations and have the full form automatically inserted the way
the Japanese Windows users can.
I understand that something like this is coming in NT 5, but it's hard to
tell what features will or won't be in NT 5. Does anyone know?
BTW, given the widespread use of Word for most non-email documents, it's
still not safe to assume that a typical US or UK document will be composed
entirely of US-ASCII chars. Regardless of how much effort the author may
take to use proper diacritics, he'll probably use curly quotes and
apostrophes by default, inserted automatically in matching left and right
pairs by the "IME" in MS Word.
From: Alain <email@example.com>
To: Unicode List <firstname.lastname@example.org>
Date: Wednesday, August 26, 1998 1:44 PM
Subject: Latin 1 Western-Eurocentric?
>A 11:15 98-08-25 -0700, Glen Perkins a écrit :
>>[...] Latin-1 is not at all sufficient for "all scripts based on Latin".
>... if you exclude French, which ISO/IEC 8859-1 can't handle integrally
>(and France is Western-European, isn't it ?) (; That said, your conclusion
>that Latin 1 is not sufficient converges with mine, it is not
>I only disagree with the remark that Latin 1 is Western-Eurocentric. It is
>not even at this height. (;
>>Most Americans and Brits use Windows or Macintosh for most documents. Win
>>and Mac have used 8-bit encodings from day one. This discussion is all
>>"legacy documents". Well, outside of some email and unix configuration
>>files, you simply can't assume that a US or UK document is encoded in
>>"US-ASCII". It almost never is (even if some header claims it is).
>>We use curly quotes and apostrophes, we use diacritics on many common
>>English words and to write the names of a large number of our citizens.
>Ah! Those nice words are dear to my heart, thanks for the intention and the
>idealism (sincerely appreciated). However in practice I count on the
>fingers of a single slightly mutilated hand the number of Angles and
>Americans who dare to write my name correctly even if they indeed have the
>tools (except for the keyboard (; ). Yesterday again, an appreciated
>American colleague, who asked me to correct a snail-mailing list, said when
>I did make fixes, that he could not enter the final E ACUTE of my family
>name and the ones in my coordinates. I showed him (and wrote him!) two
>methods to do that on Windows (use the character table of Windows
>Accessories, or use the quick-and-dirty "Alt-Numkeypad-130" method). He
>then answered honestly at last: "Well, even if I now know how to do that,
>I'm not sure I want to do that".
>I guess it is the average feeling as the results are not there. I
>experience this in practice... every day... with my American and British
>friends (some make real efforts and they are very nice [I would say they
>are heroic], but they are very few). Good tools are not there yet, one must
>admit, in English-speaking-only environments (and even when they are there
>if one wants, there is no big will!)
>Helsingør, nær København, Danmark
>(I translate for Angles and Americans : "Elsinore, near Copenhagen,
>[even if ø and æ are not French characters, I have them directly on my
> keyboard enabled to Québec government specs, which require *all* Latin 1
> répertoire characters, répertoire which is itself short of 3 French
> characters; our specs are a superset of level-2a-conformance to the
> CAN/CSA Z243.200 standard]
> The revision this year of the Canadian keyboard standard is going to
> French characters <OE>, <oe> and <Y:>, i.e. "", "" and "" in the
> Windows 'ANSI' code, or "¼", "½" and "¾" if one uses Latin 9
> (ISO/IEC 8859-15) encoding, or the equivalent in Unicode even if I can't
> unfortunately reproduce these yet with any of my pieces of software, in
> particular email, nor exchange these data with our EBCDIC systems yet.
> The Canadian keyboard standard is code-independent, it is a user-interface
> standard. It can be implemented under Windows (3, 95, 98), under DOS,
> on Macs, under UNIX, Windows NT or IBM 327x terminals, etc. (and it is
> indeed implemented on all these platforms, even with the <oe>s on Windows
> 3.1, 95, 98 and NT platforms, unfortunately with a nonstandard, albeit
> blessingly useful, external code, not exchangeable with our other 8-bit
> platforms... whence Latin 9 as the missing link between current 8-bit
> platforms and software and even with Unicode platforms... this for
> and Finnish-speakers, and Europeans in general for the EURO SIGN).
This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:20:41 EDT