Date: Sun, 18 Apr 1999 13:55:58 +0900 From: Motomichi Matsuzaki <mzaki@e-mail.ne.jp> To: dcs@newsguy.com Cc: luoqi@watermarkgroup.com, freebsd-gnats-submit@FreeBSD.ORG, byung@wam.umd.edu, freebsd-current@FreeBSD.ORG, joki@kuebart.stuttgart.netsurf.de Subject: Re: kern/5038: FreeBSD can't read MS Joliet CDs. Message-ID: <19990418135558W.mzaki@e-mail.ne.jp> In-Reply-To: <3718E187.54571571@newsguy.com> References: <199904171549.LAA23743@lor.watermarkgroup.com> <3718E187.54571571@newsguy.com>
next in thread | previous in thread | raw e-mail | index | archive | help
From: "Daniel C. Sobral" <dcs@newsguy.com>
> Luoqi Chen wrote:
> >
> > > I've add UNICODE support to the Joliet patch.
> > >
> > > It contains few charsets now, but to add other charsets is very easy.
> > > Currently, iso8859-1 and euc-jp is included.
> > >
> > Cool! I think NTFS and VFATFS could use this code too, is it possible to
> > move the code to place like libkern/unicode?
>
> I'm concerned about the possible size of GENERIC with this code.
> Remember, it has to fit in the install floppy. (Well, not really,
> with loader, but I'm not the one who is getting killed because of a
> three-disks install.)
The UNICODE routines consist of these files.
charset/charset.c mandatory
iso8859-1.c recommended
euc-jp.c optional <-- BIG! the object file has 53k bytes
encoding/encoding.c mandatory
euc.c optional
The 'mandatory + recommended' object size is no more than 5 kbytes.
The GENERIC kernel does not require necessarily
the euc-jp support or any other charsets.
I think the iso8859-1 support alone is sufficient for GENERIC.
The custom kernels can have the euc-jp support through
the CHARSET_EUC_JP and ENCODING_EUC kernel configure option.
(They are currently defined at the top of the source files.)
> Also, it adds a sysctl node, isn't that so? Directly under vfs,
> even, so it applies to all filesystems. Ideally, this should apply
> on a per-mount basis, and not even be in a sysctl.
Yes. sysctl is not the best idea.
I think the charset preferences should apply on per-process basis ideally.
The operator mounts some disks.
The users access the disks in their own preferred charset.
The UNICODE is a multiligual codeset,
so we shoud not suppose any specific charset on the disk.
Therefore, a per-mount basis is not enough.
If the routines can refer the users' environment 'LC_CTYPE',
it is fine idea. But it can't, I suppose.
--
Motomichi Matsuzaki <mzaki@e-mail.ne.jp>
Dept. of Biological Science, Fuculty of Sciences, Univ. of Tokyo, Japan
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-current" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?19990418135558W.mzaki>