gqclient-discuss Mailing List for GQ LDAP client (Page 2)

Thanks allot for letting me know, I never did get around to working on
the other changes I was looking at.  I'm still using the svn checkout
I had made so definitely looking forward to the release.

Thanks
Brian

On Tue, Jun 10, 2008 at 7:59 AM, Sven Herzberg <her...@gn...> wrote:
> Hi Brian,
>
> sorry for getting back to you that late. I really like your patch, and I
> plan to commit it to the upcoming version (as we're currently trying to
> fix the remaining issues on 1.4 and try to get this released ASAP).
>
> Regards,
>  Sven
>
>

Hi Davor,

Am Dienstag, den 10.06.2008, 18:54 +0200 schrieb Davor Ocelic:
> On Tue, 10 Jun 2008 16:57:36 +0200
> Sven Herzberg <her...@gn...> wrote:
> > > ** (gq:24693): CRITICAL **: gq_formfill_get_values: assertion
> > > `GQ_IS_FORMFILL(self)' failed **
> > > ** ERROR:(gq-input-form.c:1466):add_entry_from_formlist_no_close:
> > > assertion failed: (formlist)
> > 
> > The warnings are really important to get fixed. Can you please run gq
> > with --g-fatal-warnings and provide a backtrace then?
> 
> I'll do it later and report back, but I think in this case we should 
> focus on the primary problem -- why it's not possible to add an entry,
> instead of on the segfault that happens on the second click.

I'm not hunting the segfault; the warnings should absolutely not happen
and something goes really wrong if they do. So fixing those should be
the first step to solving the problem.

Regards,
  Sven

Seem to remember this happened before but can't remember how I worked 
around it.

make fails when working through the po directory:

Making all in po
make[2]: Entering directory `/local/gmatt/gq-1.3.4/po'
file=`echo cs | sed 's,.*/,,'`.gmo \
	  && rm -f $file &&  -o $file cs.po
/bin/sh: -o: not found
make[2]: *** [cs.gmo] Error 127
make[2]: Leaving directory `/local/gmatt/gq-1.3.4/po'
make[1]: *** [all-recursive] Error 1
make[1]: Leaving directory `/local/gmatt/gq-1.3.4'
make: *** [all] Error 2

I can run the subsequent ./src/gq but can't browse even anonymously. 
keep getting the error window: "This dialogue could not be completely 
constructed" See attached screenshot.

This is on fairly vanilla ubuntu (had to install all build deps before 
getting this far):

gcc v4.2.3
gq 1.3.4

GREG
-- 
Greg Matthews           01491 692445
Head of UNIX/Linux, iTSS Wallingford

-- 
This message (and any attachments) is for the recipient only. NERC
is subject to the Freedom of Information Act 2000 and the contents
of this email and any reply you make may be disclosed by NERC unless
it is exempt from release under the Act. Any material supplied to
NERC may be stored in an electronic records management system.

On Tue, 10 Jun 2008 16:57:36 +0200
Sven Herzberg <her...@gn...> wrote:

> Hey Davor,
> 
> sorry for getting back to you that late; but unfortunately I have been
> quite busy during the last 2 months. To properly fix the bugs you list
> here, I need some more information:

No probs.

> > ** (gq:24693): CRITICAL **: gq_formfill_get_values: assertion
> > `GQ_IS_FORMFILL(self)' failed **
> > ** ERROR:(gq-input-form.c:1466):add_entry_from_formlist_no_close:
> > assertion failed: (formlist)
> 
> The warnings are really important to get fixed. Can you please run gq
> with --g-fatal-warnings and provide a backtrace then?

I'll do it later and report back, but I think in this case we should 
focus on the primary problem -- why it's not possible to add an entry,
instead of on the segfault that happens on the second click.

When I reported that the error was "no attributes provided", it was
gq's problem, not mine -- I mean, the attributes were filled in, but
gq didn't pass them on properly.

Here's a copy of what I wrote in the followup to my initial report;
hope it helps till I try with --g-fatal-warnings:


the  update_formlist(iform);  function does not work correctly
after the changes introduced in upcoming 1.4.0 source.

With stable gq, after filling in the form and calling
update_formlist(iform);, g_list_length(formlist) returns a correct
number, such as 10.

With svn trunk, after filling in the form and calling
update_formlist(iform);,
g_list_length(formlist) always returns 1. (I suspect this one entry
found is the entry's DN, without any attributes).

So in update_formlist(), I suspect the problem is either in
gq_formfill_add_value
which doesn't work correctly (i.e. doesn't add attributed) or in
gq_input_form_set_entry_full called at the end which erases the work
done by formfill_add_values.

-doc

Hi Brian,

sorry for getting back to you that late. I really like your patch, and I
plan to commit it to the upcoming version (as we're currently trying to
fix the remaining issues on 1.4 and try to get this released ASAP).

Regards,
  Sven

You can rebuild the package for hardy. I upload a working 1.2.3  
version to launchpad:

https://launchpad.net/~herbert/+archive

Best regards
Herbert

Quoting Greg Matthews <gm...@ne...>:
> this arrives the day I finally resurrect my desktop but as of yesterday
> it is no longer running Debian. I'll test the build on Ubuntu (heron)
> which ships with a broken version of gq anyway.

Hey Davor,

sorry for getting back to you that late; but unfortunately I have been
quite busy during the last 2 months. To properly fix the bugs you list
here, I need some more information:

Am Freitag, den 11.04.2008, 15:57 +0200 schrieb Davor Ocelic:
> Starting program: /usr/local/inst/bin/gq -d 1
> [Thread debugging using libthread_db enabled]
> [New Thread 0xf7fc0000 (LWP 24693)]
> ** Message: FIXME: implement with GtkTreeView
> ** Message: FIXME: implement with GtkTreeView
> 
> (gq:24693): GLib-GObject-WARNING **: invalid uninstantiatable type
> `(null)' in cast to `GqFormfill'
> 
> ** (gq:24693): CRITICAL **: gq_formfill_get_flags: assertion
> `GQ_IS_FORMFILL(self)' failed
> 
> ** (gq:24693): CRITICAL **: gq_formfill_get_values: assertion
> `GQ_IS_FORMFILL(self)' failed **
> ** ERROR:(gq-input-form.c:1466):add_entry_from_formlist_no_close:
> assertion failed: (formlist)

The warnings are really important to get fixed. Can you please run gq
with --g-fatal-warnings and provide a backtrace then? Maybe we need to
play a ping-pong game with a new warning and a new backtrace quite a few
times, so feel free to ping me on IRC, nickname is herzi (sometimes
herz1) on GIMPnet and FreeNode.

Regards,
  Sven

Sven Herzberg wrote:
> ahja, I just increased the GTK+ dependency to GTK+ 2.10 because the
> color-handling functions only have been added recently in GTK+. Thanks
> for catching this one and sorry for not getting back to you earlier.

this arrives the day I finally resurrect my desktop but as of yesterday 
it is no longer running Debian. I'll test the build on Ubuntu (heron) 
which ships with a broken version of gq anyway.

GREG

> 
> Regards,
>   Sven


-- 
Greg Matthews           01491 692445
Head of UNIX/Linux, iTSS Wallingford

-- 
This message (and any attachments) is for the recipient only. NERC
is subject to the Freedom of Information Act 2000 and the contents
of this email and any reply you make may be disclosed by NERC unless
it is exempt from release under the Act. Any material supplied to
NERC may be stored in an electronic records management system.

Hi Greg,

Am Dienstag, den 08.04.2008, 14:18 +0100 schrieb Greg Matthews:
> Sven Herzberg wrote:
> > Thanks a lot. But now, please switch over to testing the latest 1.3.x
> > release (to make sure 1.4.0 rocks as much from the beginning).
> 
> ok... not such good news. I couldnt get it to compile. This is the last 
> few lines:
[...]
> gq-input-form.c: In function ‘gq_input_form_set_editable’:
[...]
> gq-input-form.c:1053: error: implicit declaration of function 
> ‘gtk_style_lookup_color’
> gq-input-form.c:1053: warning: nested extern declaration of 
> ‘gtk_style_lookup_color’

ahja, I just increased the GTK+ dependency to GTK+ 2.10 because the
color-handling functions only have been added recently in GTK+. Thanks
for catching this one and sorry for not getting back to you earlier.

Regards,
  Sven

On Fri, 11 Apr 2008 15:57:40 +0200
Davor Ocelic <do...@sp...> wrote:

> 
> Hey folks,
> 
> So here's the situation: It seems it's not possible to add a new
> entry in Gq using old one as template.

Looking further here's the thing:

the   update_formlist(iform);   function does not work correctly
after the changes introduced in upcoming 1.4.0 source.

With stable gq, after filling in the form and calling
update_formlist(iform);,
g_list_length(formlist) returns a correct number, such as 10.

With svn trunk, after filling in the form and calling
update_formlist(iform);,
g_list_length(formlist) always returns 1. (I suspect this one entry
found is the entry's DN).

So in update_formlist(), I suspect the problem is either in
gq_formfill_add_value
which doesn't work correctly (i.e. doesn't add attributed) or in
gq_input_form_set_entry_full called at the end which erases the work
done by formfill_add_values.

-doc

Hey folks,

So here's the situation: It seems it's not possible to add a new
entry in Gq using old one as template:

1) Open gq (svn head)
2) Connect to server as root
3) Browse tree, in my case entry ou=Group,dc=test,dc=hr
4) Right click New -> use current entry
5) In the window that pops up, replace "Group" with "Hosts" in two
   places and click "Add"

6) This should work (works correctly in 1.2.3), but in 1.4.0 (svn
head) an error shows up:

gq-input-form.c:1537 Error adding new entry 'ou=Hosts,dc=test,dc=hr':
  'Protocol error' errochain.c:218 Additional error:
  no attributes provided.

7) Click ok, which brings you back to the previous new entry window.
   Click "Add" again, and Gq crashes.


On LDAP server side, this is printed relevant to the first click on
the "Add" button:

<<< dnPrettyNormal: <ou=Hosts,dc=test,dc=hr>, <ou=hosts,dc=test,dc=hr>
conn=0 op=20 ADD dn="ou=Hosts,dc=test,dc=hr"
send_ldap_result: conn=0 op=20 p=3
send_ldap_result: err=2 matched="" text="no attributes provided"
send_ldap_response: msgid=21 tag=105 err=2

On the second click, no data is received by LDAP server and the
connection terminates.


Here's what's printed on the console during run (probably unrelated to
the specific bug):

Starting program: /usr/local/inst/bin/gq -d 1
[Thread debugging using libthread_db enabled]
[New Thread 0xf7fc0000 (LWP 24693)]
** Message: FIXME: implement with GtkTreeView
** Message: FIXME: implement with GtkTreeView

(gq:24693): GLib-GObject-WARNING **: invalid uninstantiatable type
`(null)' in cast to `GqFormfill'

** (gq:24693): CRITICAL **: gq_formfill_get_flags: assertion
`GQ_IS_FORMFILL(self)' failed

** (gq:24693): CRITICAL **: gq_formfill_get_values: assertion
`GQ_IS_FORMFILL(self)' failed **
** ERROR:(gq-input-form.c:1466):add_entry_from_formlist_no_close:
assertion failed: (formlist)


And here's the second "Add" click when gq crashes and the backtrace:

Program received signal SIGABRT, Aborted.
[Switching to Thread 0xf7fc0000 (LWP 24693)]
0x0f1b3eec in raise () from /lib/libc.so.6

(gdb) bt
#0  0x0f1b3eec in raise () from /lib/libc.so.6
#1  0x0f1b5b5c in abort () from /lib/libc.so.6
#2  0x0f6ba880 in g_assertion_message () from /usr/lib/libglib-2.0.so.0
#3  0x0f6bae00 in g_assertion_message_expr ()
from /usr/lib/libglib-2.0.so.0 #4  0x100332fc in
add_entry_from_formlist_no_close (add_context=121, iform=0x1012a268) at
gq-input-form.c:1466 #5  0x100334a0 in add_entry_from_formlist
(iform=0x1012a268) at gq-input-form.c:1410
#6  0x0f79f65c in g_cclosure_marshal_VOID__VOID ()
   from /usr/lib/libgobject-2.0.so.0
#7  0x0f78e1fc in g_closure_invoke () from /usr/lib/libgobject-2.0.so.0
#8  0x0f7a7d80 in ?? () from /usr/lib/libgobject-2.0.so.0
#9  0x0f7a95d8 in g_signal_emit_valist ()
from /usr/lib/libgobject-2.0.so.0 #10 0x0f7a99fc in g_signal_emit ()
from /usr/lib/libgobject-2.0.so.0 #11 0x0fbc86d8 in gtk_button_clicked
() from /usr/lib/libgtk-x11-2.0.so.0 #12 0x0fbca574 in ?? ()
from /usr/lib/libgtk-x11-2.0.so.0 #13 0x0f79f65c in
g_cclosure_marshal_VOID__VOID () from /usr/lib/libgobject-2.0.so.0
#14 0x0f78c438 in ?? () from /usr/lib/libgobject-2.0.so.0
#15 0x0f78e1fc in g_closure_invoke () from /usr/lib/libgobject-2.0.so.0
#16 0x0f7a7510 in ?? () from /usr/lib/libgobject-2.0.so.0
#17 0x0f7a95d8 in g_signal_emit_valist ()
from /usr/lib/libgobject-2.0.so.0 #18 0x0f7a99fc in g_signal_emit ()
from /usr/lib/libgobject-2.0.so.0 #19 0x0fbc8788 in gtk_button_released
() from /usr/lib/libgtk-x11-2.0.so.0 #20 0x0fbc87c4 in ?? ()
from /usr/lib/libgtk-x11-2.0.so.0 #21 0x0fcca654 in ?? ()
from /usr/lib/libgtk-x11-2.0.so.0 #22 0x0f78c438 in ?? ()
from /usr/lib/libgobject-2.0.so.0 #23 0x0f78e1fc in g_closure_invoke ()
from /usr/lib/libgobject-2.0.so.0 #24 0x0f7a78b8 in ?? ()
from /usr/lib/libgobject-2.0.so.0 #25 0x0f7a92d8 in
g_signal_emit_valist () from /usr/lib/libgobject-2.0.so.0 #26
0x0f7a99fc in g_signal_emit () from /usr/lib/libgobject-2.0.so.0 #27
0x0fe21594 in ?? () from /usr/lib/libgtk-x11-2.0.so.0 #28 0x0fcc1ae0 in
gtk_propagate_event () from /usr/lib/libgtk-x11-2.0.so.0 #29 0x0fcc34a4
in gtk_main_do_event () from /usr/lib/libgtk-x11-2.0.so.0 #30
0x0fab79e8 in ?? () from /usr/lib/libgdk-x11-2.0.so.0 #31 0x0f68c988 in
g_main_context_dispatch () from /usr/lib/libglib-2.0.so.0 #32
0x0f690bfc in ?? () from /usr/lib/libglib-2.0.so.0 #33 0x0f691084 in
g_main_loop_run () from /usr/lib/libglib-2.0.so.0 #34 0x0fcc39f4 in
gtk_main () from /usr/lib/libgtk-x11-2.0.so.0 #35 0x1001d00c in main
(argc=1, argv=0xff90c6b4) at gq.c:162

Hi,

Am Dienstag, den 08.04.2008, 08:26 +1000 schrieb Dave Horsfall:
> On Mon, 7 Apr 2008, Greg Matthews wrote:
> 
> > Excellent! been using 1.0 for years as nothing newer would work with 
> > Debian and I've just got round to trying it again and 1.2.3 works great!
> 
> Works perfectly on FreeBSD 6/7 as well.  Good work!

Thanks a lot. But now, please switch over to testing the latest 1.3.x
release (to make sure 1.4.0 rocks as much from the beginning).

Regards,
  Sven

Excellent! been using 1.0 for years as nothing newer would work with 
Debian and I've just got round to trying it again and 1.2.3 works great!

Just wanted to say thanks - this is by far the best LDAP browser and 
editor around, even the buggy 1.0 which crashes a lot is better than 
anything else I've tried. The ability to open and manage more than one 
session at a time is invaluable to me as I administer around 25 LDAP 
servers...

thanks again.

GREG
-- 
Greg Matthews           01491 692445
Head of UNIX/Linux, iTSS Wallingford

-- 
This message (and any attachments) is for the recipient only. NERC
is subject to the Freedom of Information Act 2000 and the contents
of this email and any reply you make may be disclosed by NERC unless
it is exempt from release under the Act. Any material supplied to
NERC may be stored in an electronic records management system.

On Fri, Apr 4, 2008 at 5:30 PM, Herbert Straub <he...@li...> wrote:
>  Please create a bug in Debian and Ubuntu (or confirm the open bug - I think
> #110278).

Spot on, just added a comment to it, thanks.
Steph

Hallo,

this is a problem with Debian and Ubuntu. I created a new package with  
version 1.2.3, but there is no reaction. This is sad, because I'm  
using Ubuntu on the Desktop and need gq.

Please create a bug in Debian and Ubuntu (or confirm the open bug - I  
think #110278).

Best regards
Herbert Straub

Quoting FRLinux <fr...@gm...>:

> Hello,
>
> Been following the list silently in the past few months. I confirm
> that compiling a 1.3.x on Ubuntu fixes my issue (just upgraded to
> hardy beta today and sad to report their version is still
> 1.2.1-2build1).
>
> Steph

Hello,

Been following the list silently in the past few months. I confirm
that compiling a 1.3.x on Ubuntu fixes my issue (just upgraded to
hardy beta today and sad to report their version is still
1.2.1-2build1).

Steph

I like to use gq to look at an active directory. Anyone else doing that?

I have a setup with which I can search and get results. So far so good.

However when I try to edit an entry, the edit window comes up but 
doesn't fill (i.e. I get no more then the top window bar). When I then 
switch back to the search window I find a disturbing message at the bottom:

Searching for ....: Critical extension is unavailable

Similar thing happens in the Browse tab. I can expand the base. But when 
I try to view any branch below ... no information and a similar message 
at the bottom.

If try to expand a bustree I get a bomb with at text like "Critical 
extension is unavailable Additonal error ..."

I suspect (haven't looked at the source code though) that GQ is using 
ldap_search_ext() with server controls which are passed to the server as 
critical but which aren't supported by active directory.

Does this ring any bells?

Brian Finney wrote:
> Hello,
> 
> I decided to scratch an itch today and added a SASL Mechanism field to
> the server preferences dialog box, and I'm pretty sure I got all of
> the appropriate glue code from the config file, to the dialog, and
> finally passing it to ldap_sasl_interactive_bind_s.
> 
> Patch against trunk r1165 is attached, it's working for me, hopefully
> it will also work for other people.
> 
> Next few days I may also look at changing the field to a drop down and
> suppressing the password box for things like GSSAPI (for now just
> leave the password blank).

Hi Brian,

thanks for the patch. Could you please open a sourceforge bug ticket for
this and attach the patch to it? Otherwise it might get lost in the
mailing list.

Hello,

I decided to scratch an itch today and added a SASL Mechanism field to
the server preferences dialog box, and I'm pretty sure I got all of
the appropriate glue code from the config file, to the dialog, and
finally passing it to ldap_sasl_interactive_bind_s.

Patch against trunk r1165 is attached, it's working for me, hopefully
it will also work for other people.

Next few days I may also look at changing the field to a drop down and
suppressing the password box for things like GSSAPI (for now just
leave the password blank).

Thanks
Brian

P.S. I'm really liking the minimalist interface of GQ

Hi,

first of all: which version of GQ do you have there? And what is the
exact error message (does it appear in a dialog box, command line, log
file)?

Regards,
  Sven

Hi,



I am quite new to LDAP and GQ. I have installed LDAP and GQ in the same
system and it has worked properly for me. I am able to search for the
entries I have made using GQ.



Now I have installed GQ in a system with no LDAP. I have LDAP in another
system. I haven't faced any problems in any of the installation steps of GQ,
but when I try to run GQ, I get an error saying "Invalid option". I tried to
run GQ without any of the options, but I still get this "Invalid option"
error. I have tried some of the options also but the error persists. The
only option that doesn't give an error is "--help".



Do you guys know what is causing this error and how to prevent it?



Thanks in advance,



eashwar

Hello,

I decided to scratch an itch today and added a SASL Mechanism field to the
server preferences dialog box, and I'm pretty sure I got all of the
appropriate glue code from the config file, to the dialog, and finally
passing it to ldap_sasl_interactive_bind_s.

Patch against trunk r1165 is attached, it's working for me, hopefully it
will also work for other people. <gqc...@li...>

Next few days I may also look at changing the field to a drop down and
suppressing the password box for things like GSSAPI (for now just leave the
password blank).

Thanks
Brian

P.S. I'm really liking the minimalist interface of GQ

Brent Bice wrote:
> Sven Herzberg wrote:
>> Known issue, any reason for not using 1.2.3?
> 
>     Ah, whoops!  I grabbed 1.2.1 because it was the latest listed on the 
> News section.  Grabbing 1.2.3 now...

    Just to confirm, version 1.2.3 works MUCH better. :-) I can connect 
with authenticated sessions fine now.

Brent

Sven Herzberg wrote:
> Hi Brent,
> 
> Am Dienstag, den 19.02.2008, 11:55 -0800 schrieb Brent Bice:
>> I've just compiled gq 1.2.1 on a SLES10 x86_64 machine without 
>> errors, and had some odd problems initially adding servers, but finally 
>> got some added by editing ~/.gq myself.
> 
> Known issue, any reason for not using 1.2.3?

    Ah, whoops!  I grabbed 1.2.1 because it was the latest listed on the 
News section.  Grabbing 1.2.3 now...

    Thanks!

Brent

Hi Brent,

Am Dienstag, den 19.02.2008, 11:55 -0800 schrieb Brent Bice:
> I've just compiled gq 1.2.1 on a SLES10 x86_64 machine without 
> errors, and had some odd problems initially adding servers, but finally 
> got some added by editing ~/.gq myself.

Known issue, any reason for not using 1.2.3?

> ...
>     Any ideas?
> ...

Upgrade. 1.2.3 or 1.3.4 should both be fine.

Regards,
  Sven