bug-glibc
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: glibc-2.2.1 iconv() prototype and eucJP/SJIS conversion tables


From: Andreas Jaeger
Subject: Re: glibc-2.2.1 iconv() prototype and eucJP/SJIS conversion tables
Date: 23 Jan 2001 15:10:26 +0100
User-agent: Gnus/5.090001 (Oort Gnus v0.01) XEmacs/21.1 (Capitol Reef)

堀川 久 <address@hidden> writes:

> Hello.
> 
> I wrote two patchs to glibc-2.2.1;
>   - correct the iconv() function prototype to be the definition of
>     SUSv2.
The iconv prototype in glibc is correct - that's the way it will be in
the next Posix standard.  Your patch will not be used.

>   - harmonize the conversion tables of eucJP/SJIS with them of the 
>     commercial systems.
>     The eucJP table is corresponding to The Open Group/JVC 
>     eucJP-open-ascii table.
>     The SJIS table is corresponding to Microsoft Windows95 table.

Where are differences between glibc and those of the above systems?
Why is glibc wrong there?  We're not following commercial systems just
because they're commercial.

Why do you disable libJIS.so and remove the copyright headers?  Please
note that code for glibc has to follow the GNU Coding guidelines, your
change do not follow.  Your patch removes a number of optimizations
(check e.g. __builtin_expect).

I don't understand these changes for EUC-JP:
-<U0000>     /x00         NULL
-<U0001>     /x01         START OF HEADING
...
+<U0000>     /x00    <control>
+<U0001>     /x01    <control>


Uli will commit on some details of the patch and has to decide whether
it's ok to apply.


> 
> I put the patch files on the following location;
>       http://www2.airnet.ne.jp/pak04955/patch.htm
> 
> I would be glad if you adopted them.
> Thank you.

Andreas
-- 
 Andreas Jaeger
  SuSE Labs address@hidden
   private address@hidden
    http://www.suse.de/~aj



reply via email to

[Prev in Thread] Current Thread [Next in Thread]