One handy approach for dealing with multiple OSs is to use Unicode
internally in your code and factor all OS references out into an OS
"object". This object then deals with the anomalies of the various OSs you
want to support. That way you restrict any #ifdefs to your OS object and
you make it easier to support yet another OS, should you need to do so. You
can also use different OS objects for different OSs and delegate the choice
to your make file. In the RichEdit project, we use a combination of these
techniques to support a variety of OSs.
Murray
-----Original Message-----
From: Smita Desai [SMTP:sdesai@inconcert.com]
Sent: Thursday, November 12, 1998 6:36 AM
To: Unicode List
Subject: RE: OS 99
Thanks Asmus for the clarifications. For someone struggling with the
decision of porting a Solaris 2..6 app to Unicode or wait out for
Solaris 7.0, it is quite misleading to say that the former supports
Unicode. Solaris 2.6 has only two native Unicode locales, namely US
English and Korean. And in all other cases, wide character does not
translate to a Unicode character.
My source code has more #ifdefs for Win 9x, Solaris and other Unix
flavors than the actual lines of code.
In any case, it would be a good idea if the Consortium were to
define a
guideline in this respect.
Smita
Smita Desai
Software Engineer, Internationalization
InConcert, Inc.
Tel - 617.499.4427
Fax - 617.499.4409
smita@inconcert.com
http://www.inconcert.com
-----Original Message-----
From: Asmus Freytag [mailto:asmusf@ix.netcom.com]
Sent: Wednesday, November 11, 1998 3:38 PM
To: Unicode List
Subject: RE: OS 99
Smita,
your comment is well taken. The mail in question was
probably not meant to
go to go out on this wide an alias in this form.
Last I
saw it, it was being
circulated among knowledgeable people to try to
gather
information on Unicode
support in OS's, not in order to 'publish' a list
of
compliant systems.
"Unicode compliant" is a term that is fraught with
problems, because it is
in fact not defined anywhere in the standard.
"Unicode
conformant" is the
legal term, however, we should not officially use it
as
header on the table
below, since the consortium are not certifying
conformance.
"Supports Unicode" is the informal term for
attempting
to be Unicode
conformant, but not implying the same guarantees
that
"Unicode conformant"
entails.
Unicode 'compliant' gets mistaken for Unicode
'conformant' too easily,
which is bad. The chart should be entitled
"Operating systems which support Unicode.
Finally, I believe this is true for the Windows
family.
>Windows NT all yes yes
>Windows CE yes yes
>Windows 95 no some *)
>Windows 98 no some *)
>Windows 2000 yes yes
*) code set conversions and limited, low level text
output
If anybody else has information on what to add here,
please send it
to info@unicode.org (not the whole list please). We
are
considering
posting the table on our web site.
BTW, if someone wants to sumamrize the browsers for
us
in the same manner
please send info to info@unicode.org
A./
At 07:42 AM 11/10/98 -0800, you wrote:
>What are the criteria for making "Unicode
compliant"?
Is it the ability
>to convert to and from or actual support in all
functions?
>In the list below, I see a lot of instances where
the
latter would not
>work, e.g. Windows 9x, Solaris 2.6 etc.. Win 9x
does
not have Unicode
>support for "all" APIs and Solaris 2.6 has no
support
except for to/from
>conversion, at least as far as I know.
>
>Just curious,
>Smita
>
> -----Original Message-----
> From: Julia Oesterle (Unicode)
>[mailto:v-juliao@microsoft.com]
> Sent: Monday, November 09, 1998
9:54
PM
> To: Unicode List
> Subject: FW: OS 99
>
> Hi -- update help requested:
>
> > A year ago, the question of which
OSs
were Unicode
>compliant was answered
> > and
> > we constructed the following
chart:
> >
> > \OS Version
UI API
> > Sun Solaris 2.6
yes yes
> > IBM AIX 4.1x
yes no
> > IBM OS/2 4.1
no yes
> > IBM OS/400 3.7
no yes
> > SGI IRIX* 6.4
no no
> > Next Open Step
yes yes
> > Apple Rhapsody
yes yes
> > NetWare 4.x
no yes
> > Windows NT 4.x yes
yes
> > Windows 95
no yes
> > Windows 98
no yes
> > Linux no
response
> > *(IRIX supports conversions
between
UCS4, UCS2, UTF8 &
>other character
> > sets)
> >
> > Has this changed?
> >
> > Thanks,
> >
> > Julia Oesterle
>
>
This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:20:42 EDT