Folks,
ISO-8859-8-i exists only in the realm of the "charset" directive of MIME
(and HTTP).
It is *not* a character set. It's a "charset" value. Yes, I know it's confusing.
When either ISO-8859-8 or ISO-8859-8-i is specified in "charset", the character
set is that of ISO 8859-8. As added semantics, the former implies 8859-8 with
"visual" direction control, while the latter implies implicit (aka "logical")
direction control (Unicode algorithm).
The reason for registering these names this way is historic. At first, "ISO-8859-8"
was registered in IANA as a charset value string. In terms of direction control,
"Visual" was the only one in existence in the Internet realm, so no indication
was required.
A bit later, implicit (aka "logical") direction control gained some momentum.
Now, there was no verb in MIME to specify direction control. It was just not
there.
The choices were:
(a) Go for spec'ing a new verb ("dir" ?), which would have taken *years* of
IETF process, or
(b) overload a charset spec to "contain" an indication of direction control,
which would have taken only a registration in IANA.
The second option was patently un-elegant and far from ideal; however a way to
standardize Hebrew content was required, and waiting a few years for the proper
verb was just not an option. So, ISO-8859-8-i was registetered, with the semantics
of "charset" being overloaded. Now, for Hebrew, the "charset" value strings
indicate both character set and direction control.
In any other context except MIME and HTTP (and hence HTML META tag), ISO-8859-8-i
does not exist, and ISO-8859-8 should strictly map to a character set, carrying
no direction control semantics.
Doron
Shachar Shemesh wrote:
Nadav Har'El wrote:
On Sat, Aug 21, 2004, Shachar Shemesh wrote about "Re: charset iso-8859-8-i":I don't need to. You agree with it yourself. Let me try to rephrase in a simpler way. ISO-8859-8 mean visual Hebrew ONLY when used in a MIME context. With any other context, it does not relate to ordering at all, leading, as you suggested, to an assumption of "logical order".
Both the last sentences are not exactly accurate.
Care to back up this claim with a reference?
It is therefor, broadly, correct to say that ISO-8859-8 refers to visual ordering when in MIME context, and logical ordering otherwise.
True, but I don't see how an RFC titled "Handling of Bi-directional Texts in MIME" can contradict anything I said :-)As far as I know, ISO-8859-8-I does not exist outside the HTML/Mime world, and ISO-8859-8 does mean logical Hebrew there.
Iso-8859-8-i was defined as early as 1993 in RFC 1556.
Care to show an example of that? So long as all MIME clients treat ISO-8859-8 as "visual ordering", I don't see how anyone can bend the standard. Like I said above, as far as OS, file, and other (non-MIME) encoding goes, ISO-8859-8 is usually logical ordering anyways.I already mentioned that while this RFC is "de jure" law, the "de facto" law is that the old visual representation is all but obsolete, so people are thinking more and more that it's silly to tack on the extra "-i" for the only useful encoding. So they don't.
Shachar
================================================================= To unsubscribe, send mail to [EMAIL PROTECTED] with the word "unsubscribe" in the message body, e.g., run the command echo unsubscribe | mail [EMAIL PROTECTED]
