>> Yes, I'll take a look at this; callers to
>> "asn1_octet_string_value_decode()" must not assume that what
>> they get back is a '\0'-terminated string (it's a counted string,
>> and it's not even necessarily a character string - the ASN.1
>> dissector doesn't know whether it is, although its caller
>> probably does), *and* must not assume that it gets back a
>> non-null pointer. The LDAP dissector is making both of those
>> assumptions in some places....
> Here's a patch to "packet-ldap.c" to fix that; users with
> source and "patch" and compiler tools/necessary libraries can
> patch and rebuild, users without any of those will probably
> have to wait for the next release and the binary packages for
> that release.
That was a speedy fix: about 6 hours from time of original bug report!
Thanks.
I'm going to have to wait for the next Windows binary package. However, it
looks like I've got plenty to keep me going...Ethereal seems to let me
contantly find bugs and wrinkles in all sorts of things.
Gilbert said:
# About the LDAPv3... if no one on ethereal-dev stands up and
# says "I'm working on it", then there's currently no plans to
# do it; most developers add protocols when they have the need
# to analyze them.
# There's no over-all plan for protocol support. But perhaps
# someone will stand up and say "I'll work on that."
I understand that the development of the dissector does not happen on its
own and is dependent on someone wanting to do the work. I am hoping that
will be sooner rather than later, especially now that MS ADS is in use, and
OpenLDAP v2.0 is about to be released.
Keeping my fingers crossed.
Mark Symons
