On 9/14/06, Matthew Burgess <[EMAIL PROTECTED]> wrote:
<para>Apply the following patch to fix an issue with using identical conversion modules in iconv:</para> I couldn't understand what the issue was by looking at that description. The description in the patch is even more vague. Looking at the test case, the problem is that one can't iconv_open() the same character encoding conversion combinations twice.
I didn't understand the issue, either. I took that description from the glibc bug from the guy who identified the problem: http://sources.redhat.com/bugzilla/show_bug.cgi?id=2569#c7 Side note: This bug should be referenced in addition to the Samba bug in the patch.
What about using the description that's in iconv_open(3C): <para>Apply the following patch to fix a bug when trying to allocate a conversion descriptor for converting byte sequences between the same character encodings twice:</para> Or is this getting too detailed? As you can tell, I'm having a hard time trying to determine the right amount of information to provide!
Sounds fine to me. I'd be joking if I said I understood how iconv works. I just didn't want the book to say "Fix an issue in iconv:". The sentence runs a bit due to the complex subject matter, but I can't think of anything better. I never mind too much information. You can always glaze over and skip it if you're not interested. -- Dan -- http://linuxfromscratch.org/mailman/listinfo/lfs-dev FAQ: http://www.linuxfromscratch.org/faq/ Unsubscribe: See the above information page
