Thank you for your continued business.
I line 508 of display_finfo() of client.c is not suitable for recognizing output of file list.
The separator of d_printf() is space but filename often includes space.
If the last character of filename was space, then we couldn't recognize it as a part of filename or separator.
It might be good if the separator is \t, not space.
Thank you for your attention.
Your description of the problems (also bug 7700) is very concrete, almost a plaintext description of the patch. It would probably make it quicker to get the changes in if you uploaded them as a patch. See
for a quick description to create such a patch.
Thank you for your advice.
All right. I'll study it and create patch.
Thanks a lot.
Created attachment 6038 [details]
Patch for 3.x (include bug 7700 and 7701)
A patch for solving the problems 7700 and 7701.
Against which tree did you create this patch? It does not apply to any of the v3-x-test branches we right now have open.
I am so sorry for late reply. Please forgive
me that I have made a incorrect descriptions.
The against branch is "master", but not any
I have a question, does it need to make a
new one for v3-5-test? How about the version
Oh, that is interesting. Your patch contains the two lines
-void do_list(const char *mask,
+int do_list(const char *mask,
The - line usually means that this line is about to be deleted. When I look at current master, it contains a line
NTSTATUS do_list(const char *mask,
My implementation of the "patch" program is not able to delete the line "void do_list...", because it finds a line "NTSTATUS do_list...".
What version of "patch" are you using to cleanly apply this patch?
Thank you for reply,
That means I have an old version of the source of samba? I have
got them from last month. Until now, I have not updated the latest
version. Has someone else improved the problem already during
these days? All right, I will update the latest version for checking.
Add after that, I will offer a new patch.
We hope that the following versions will be applied for the patch.
origin/v3-3-* (presently used)
origin/v3-5-* (be used soon)
Would you like to tell us if we want to use the upgraded versions
such as "origin/v3-6-*" and "origin/v4-0-*", we must offer the new
patchs for them respectively?
Besides, we are familiar with "SVN", but "git" is new for us. I'd
like to study it. Could you tell me the following steps are correct?
$ git clone git://git.samba.org/samba.git samba
$ git checkout -b my_v3-3-test origin/v3-3-test
Then, we do our changes,
$ git add my_v3-3-test
$ git commit -m "7701-xxx-yyy-zzz"
Finally, we make the patch by,
$ git format-patch -1
Created attachment 6047 [details]
Patch for v3-6-test
The local sources have been updated today.
Found that the bug 7700 has been fixed partly
except for "cmd_mkdir()"" and "cmd_rmdir()".
Therefore, the new patch has been made for the
two remained functions and bug7701. The agaist
branch is "origin/v3-6-test".
Created attachment 17468 [details]
patch for master
The core issues in the attached patch (that is, ignoring some errors in source3/client/client.c) are still unfixed. It was a couple of these:
+ return 1;
This patch updates that, though I honestly don't know if anyone cares about this code.
(In reply to Douglas Bagnall from comment #9)
ah, just read the title of the bug, and I patched the bits that *don't* pertain to that. Oh well.