Samba 4 is still feature incomplete. If you are using it for anything other
-then education you are insane.
+than education you are insane.
Please file bug reports at https://bugzilla.samba.org/, product: Samba4.
Please include as much information as possible, such as SVN revision number
Removal of nmbd and introduction of process models
==================================================
-smbd now implements several network protocols other then just CIFS and
+smbd now implements several network protocols other than just CIFS and
DCE/RPC. nmbd's functionality has been merged into smbd. smbd supports
various 'process models' that specify how concurrent connections are
handled (when to fork, use threads, etc).
Much improved SWAT
==================
-SWAT has had some rather large improvements and is now more then just a
+SWAT has had some rather large improvements and is now more than just a
direct editor for smb.conf. Its layout has been improved. SWAT can now also
be used for editing run-time data - maintaining user information, provisioning,
etc. TLS is supported out of the box.
Default: Set at compile-time
+ ntvfs handler
- Backend to the NT VFS to use (more then one can be specified). Available
+ Backend to the NT VFS to use (more than one can be specified). Available
backends include:
- posix:
+ client use spnego principal
Tells the client to use the Kerberos service principal specified by the
- server during the security protocol negotation rather then
+ server during the security protocol negotation rather than
looking up the principal itself (cifs/hostname).
Default: false
* Call the wins hook script again (metze)
- * Make sure no more then 25 records are added in the WINS database (metze)
+ * Make sure no more than 25 records are added in the WINS database (metze)
* Documentation updates (jelmer)
#define SRV_SEAL "session key to server-to-client sealing key magic constant"
/**
- * Some notes on then NTLM2 code:
+ * Some notes on the NTLM2 code:
*
* NTLM2 is a AEAD system. This means that the data encrypted is not
* all the data that is signed. In DCE-RPC case, the headers of the
r = (o->flags & (GSS_C_REPLAY_FLAG|GSS_C_SEQUENCE_FLAG))==GSS_C_REPLAY_FLAG;
- /* sequence number larger then largest sequence number
- * or smaller then the first sequence number */
+ /* sequence number larger than largest sequence number
+ * or smaller than the first sequence number */
if (seq_num > o->elem[0]
|| seq_num < o->first_seq
|| o->length == 0)
/*
* In the framework of kerberos, the length can never be shorter
- * then at least one blocksize.
+ * than at least one blocksize.
*/
if (encryptp) {
cred.enc_part.cipher.length = buf_size;
} else {
/*
- * Here older versions then 0.7.2 of Heimdal used the local or
+ * Here older versions than 0.7.2 of Heimdal used the local or
* remote subkey. That is wrong, the session key should be
* used. Heimdal 0.7.2 and newer have code to try both in the
* receiving end.
/*
* NetBSD have a thread lib that we can use that part of libc that
* works regardless if application are linked to pthreads or not.
- * NetBSD newer then 2.99.11 just use pthread.h, and the same thing
+ * NetBSD newer than 2.99.11 just use pthread.h, and the same thing
* will happen.
*/
#include <threadlib.h>
ctx->pk_nonce = ctx->nonce;
/*
- * Increase counter when we want other pre-auth types then
+ * Increase counter when we want other pre-auth types than
* KRB5_PA_ENC_TIMESTAMP.
*/
#define MAX_PA_COUNTER 3
case KRB5KDC_ERR_KEY_EXPIRED :
/* try to avoid recursion */
- /* don't try to change password where then where none */
+ /* don't try to change password where there where none */
if (prompter == NULL || ctx.password == NULL)
goto out;
if (level > 9) {
krb5_set_error_string(context, "Authorization data nested deeper "
- "then %d levels, stop searching", level);
+ "than %d levels, stop searching", level);
ret = ENOENT; /* XXX */
goto out;
}
s/\(.*\)"\n\([0-9]*\)/$ #\2 Original Message:(\1)/p
}
#
-# The .msg file contains, other then the .po file, only the translations
+# The .msg file contains, other than the .po file, only the translations
# but each given a unique ID. Starting from 1 and incrementing by 1 for
# each message we assign them to the messages.
# It is important that the .po file used to generate the cat-id-tbl.c file
/* copy specified number of elements of a dn into a new one
element are copied from top level up to the unique rdn
- num_el may be greater then dn->comp_num (see ldb_dn_make_child)
+ num_el may be greater than dn->comp_num (see ldb_dn_make_child)
*/
struct ldb_dn *ldb_dn_copy_partial(void *mem_ctx, const struct ldb_dn *dn, int num_el)
{
* that any upper layers will use.
*
* All local attributes will have to have a definition. Not all remote
- * attributes need a definition as LDB is a lot less stricter then LDAP
+ * attributes need a definition as LDB is a lot less strict than LDAP
* (in other words, sending unknown attributes to an LDAP server hurts us,
- * returning too much attributes in ldb_search() doesn't)
+ * while returning too many attributes in ldb_search() doesn't)
*/
struct ldb_map_context;
/*@globals fileSystem @*/
/*@modifies *str, *fp, fileSystem @*/
{
- char * s = alloca(300); /* larger then the ascii set */
+ char * s = alloca(300); /* larger than the ascii set */
s[0] = '\0';
/*@-branchstate@*/ /* FIX: W2DO? */
- support for adding/deleting keys
- support for security descriptors
-- pass parsed paths around rather then strings (i.e. just a list of strings)
+- pass parsed paths around rather than strings (i.e. just a list of strings)
- integrate various registry tools ?
- finish new patchfile code
<para>The regtree utility prints out all the contents of a
Windows registry file. Subkeys are printed with one level
- more indentation then their parents. </para>
+ more indentation than their parents. </para>
</refsect1>
}
if ((*ret)->data.length < vk->data_length) {
- DEBUG(1, ("Read data less then indicated data length!\n"));
+ DEBUG(1, ("Read data less than indicated data length!\n"));
}
return WERR_OK;
WERROR (*num_values) (const struct registry_key *, uint32_t *count);
WERROR (*get_subkey_by_index) (TALLOC_CTX *, const struct registry_key *, int idx, struct registry_key **);
- /* Can not contain more then one level */
+ /* Can not contain more than one level */
WERROR (*get_subkey_by_name) (TALLOC_CTX *, const struct registry_key *, const char *name, struct registry_key **);
WERROR (*get_value_by_index) (TALLOC_CTX *, const struct registry_key *, int idx, struct registry_value **);
- /* Can not contain more then one level */
+ /* Can not contain more than one level */
WERROR (*get_value_by_name) (TALLOC_CTX *, const struct registry_key *, const char *name, struct registry_value **);
/* Security control */
/*
This implements "Class 2 mailslots", i.e. the communication mechanism
- used for all mailslot packets smaller then 425 bytes.
+ used for all mailslot packets smaller than 425 bytes.
"Class 1 mailslots" (which use SMB) are used for messages larger
- then 426 bytes and are supported on some systems. These are not implemented
+ than 426 bytes and are supported on some systems. These are not implemented
in Samba4 yet, as there don't appear to be any core services that use
them.
/* In some cases the client maybe unsure that a particular */
/* binding will reach the server. (For example, when the oxid */
- /* bindings have more then one TCP/IP binding) This call */
+ /* bindings have more than one TCP/IP binding) This call */
/* can be used to validate the binding */
/* from the client. */
[idempotent] WERROR ServerAlive ();
- allow data structures outside of interfaces
-- mem_ctx in the interface rather then as struct ndr member.
+- mem_ctx in the interface rather than as struct ndr member.
Force the use of new_hf_name when the parser generator was going to
use old_hf_name.
-This can be used in conjunction with HF_FIELD in order to make more then
+This can be used in conjunction with HF_FIELD in order to make more than
one element use the same filter name.
=item I<STRIP_PREFIX> prefix
Force pidl to not generate a particular function but allow the user
to write a function manually. This can be used to remove the function
-for only one level for a particular element rather then all the functions and
+for only one level for a particular element rather than all the functions and
ett/hf variables for a particular element as the NOEMIT command does.
=back
{
my ($a,$b) = @_;
- # Only align if previous element was smaller then current one
+ # Only align if previous element was smaller than current one
if ($$a < $b) {
pidl "if (!prs_align_custom(ps, $b))";
pidl "\treturn False;";
{
my ($e,$l,$ndr,$var_name,$ndr_flags) = @_;
- # strings are passed by value rather then reference
+ # strings are passed by value rather than reference
if (not Parse::Pidl::Typelist::is_scalar($l->{DATA_TYPE}) or Parse::Pidl::Typelist::scalar_is_reference($l->{DATA_TYPE})) {
$var_name = get_pointer_to($var_name);
}
}
# allocate the "simple" out ref variables. FIXME: Shouldn't this have it's
- # own flag rather then be in NDR_IN ?
+ # own flag rather than be in NDR_IN ?
foreach my $e (@{$fn->{ELEMENTS}}) {
next unless (grep(/out/, @{$e->{DIRECTION}}));
=item I<--parse-idl-tree>
Read internal tree structure from input files rather
-then assuming they contain IDL.
+than assuming they contain IDL.
=item I<--dump-idl>
Pidl will read additional data from an ethereal conformance file if present.
Such a file should have the same location as the IDL file but with the
-extension I<cnf> rather th
en I<idl>. See L<Parse::Pidl::Ethereal::Conformance>
+extension I<cnf> rather th
an I<idl>. See L<Parse::Pidl::Ethereal::Conformance>
for details on the format of this file.
=item I<--diff>
/* the min_version is always 0 here, and won't be updated */
- /* if the given version is higher the then current nax_version, update */
+ /* if the given version is higher than the current max_version, update */
if (cur->owner.max_version < version) {
cur->owner.max_version = version;
/* if it's for our local db, we need to update the wins.ldb too */
git.samba.org / metze / samba / wb-ndr.git / commitdiff