[Date Prev][Date Next]
[Chronological]
[Thread]
[Top]
RE: Last call comments with LDAP C API #4
Isn't this a rather peculiar definition for a client-side control?
"The ldctl_value.bv_lenfield MUST always be set to 4. The
ldctl_value.bv_val field MUST point
to a 4-octet integer flags value."
Wouldn't it be more reasonable to define it in implementation-specific terms
such as impl_uint_t? For example, "The ldctl_value.bv_lenfield MUST always
be set to sizeof(impl_uint_t). The ldctl_value.bv_val field MUST point to a
impl_uint_t flags value."
This would allow the field to be used naturally by both clients and the api,
without needing to do anything special to ensure use of a 4-octet value.
Tom Salter
> -----Original Message-----
> From: Mark Wahl [mailto:M.Wahl@INNOSOFT.COM]
> Sent: Friday, November 12, 1999 8:48 PM
> To: mcs@netscape.com
> Cc: Mark Wahl; timhowes@yahoo.com; andyhe@MICROSOFT.com;
> anoopa@MICROSOFT.com; kurt@openldap.org; ietf-ldapext@netscape.com
> Subject: Last call comments with LDAP C API #4
>
>
>
> Mark, here are some comments I received, primarily from our
> engineers,
> on the C API. If they look reasonable to you, I would
> consider that most of
> them could be treated as last call clarification requests.
> Let me know if
> you have any issues with them.
>
>
> 1) In 11.3.1 it says "The ldctl_value.bv_val field MUST
> point to a 4-octet
> integer flags value."
>
> This is not 'integer' in the sense of a C int (since a C int
> may not be 4
> bytes), nor is it an ASN.1 INTEGER which is variable length.
> This sentence
> therefore needs to specify the encoding of an integer value
> in 4 octets. In
> particular, the endian of this value needs to be specified.
> Host byte order
> or network byte order?
>
> I (--mfw--) assume host, so "The ldctl_value.bv_val field
> MUST point to a 4
> byte area containing the flags integer value in host byte order."
>
[...]
> Mark Wahl, Directory Product Architect
> Innosoft International, Inc.
>