and guess what....

The AuthInfoRequired line will keep re-appearing.

It is written there by cupsd

Remove the line, restart cups and you'll be able to print once.

If you want to print again, you'll need to restart cups.



I've spent more than a day solving this.



Ended up downloading the cups-1.3.8-source.tar.bz2 from the cups
website and applying the following patches and then recompiling /
installing cups:

diff --recursive cups-1.3.8/scheduler/printers.c
cups-1.3.8ae/scheduler/printers.c

1563c1563

<******** p->auth_info_required[p->num_auth_info_required] =
"negotiate";

---

>******** p->auth_info_required[p->num_auth_info_required] =
"none";

1640c1640

<****** p->auth_info_required[p->num_auth_info_required] =
"negotiate";

---

>****** p->auth_info_required[p->num_auth_info_required] =
"none";



I have no idea how this crap got into cups or even why authentication
should be necessary to print locally.

Anyway, it has nothing to do with the security deny/allow stuff in
cupsd.conf



The above effectively forces the word 'negotiate' after the
AuthInfoRequired keyword to be treated as none.









On Sun, 2008-07-20 at 18:07 +0000, g wrote:

Craig
White wrote:

<snip>

AuthInfoRequired
negotiate



got into /etc/cups/printers.conf in the printer definition and removing

the line completely solved it.



I have no idea how it got there




because you use a computer and it forgot "1st law"?



interesting anyway.



do printers set up with this install. i use it just for internet. will

check files under other installs. will be interesting to see what they

have.



glad you got it working.


----

sometimes, I haven't a clue to references that you make - i.e. - 1st

law?



I routinely set up cups to offer print queues for users whether Linux

(automatic discovery, automatic setup), Macintosh (via cups or via

netatalk) and Windows (via Samba, providing APW) and cups has been a

remarkable print server/service with very few problems.



Obviously I try to stay away from editing the cups config files by hand

because they tend to get trampled on by cups configuration tools

(http://localhost:631)
or via system-config-printer (which finally seems

to live in peace alongside cups configuration tools).



Something, somewhere, somehow entered that line in the printers.conf

file (AuthInfoRequired negotiate) as I am quite certain that I didn't

put it in there - not even from an errant selection in one of the above

mentioned configuration tools.



Anyway, c'est la vie - that's what I get for not raking through the

config files with emacs and relying upon the gui tools



Craig


--
fedora-list mailing list
fedora-list@redhat.com
To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list

Andy Eager wrote:
<snip>

I have no idea how this crap got into cups or even why authentication
should be necessary to print locally.
Anyway, it has nothing to do with the security deny/allow stuff in
cupsd.conf


this got me wondering.

craig,

have you give thought to *group*?

because problem is with *authentication*, are you
'authenticated' if you add your user name to 'cups group'?

ria, under md 9 or 10, i had problems in printing as a 'user'
and i believe that doing so enable me to print.

i do recall that i still had to enter 'root' password to make
any changes to configs.


--

tc,hago.

g
.

in a free world without fences, who needs gates.

--
fedora-list mailing list
fedora-list@redhat.com
To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list

On Thu, 2008-07-24 at 13:21 +0000, g wrote:
> Andy Eager wrote:
> <snip>
> > I have no idea how this crap got into cups or even why authentication
> > should be necessary to print locally.
> > Anyway, it has nothing to do with the security deny/allow stuff in
> > cupsd.conf
>
> this got me wondering.
>
> craig,
>
> have you give thought to *group*?
>
> because problem is with *authentication*, are you
> 'authenticated' if you add your user name to 'cups group'?
>
> ria, under md 9 or 10, i had problems in printing as a 'user'
> and i believe that doing so enable me to print.
>
> i do recall that i still had to enter 'root' password to make
> any changes to configs.
----
The problem isn't with authentication, the problem is that somehow the
printer configuration inherited an authentication requirement when there
was no desire or action for it to do so.

Mucking in groups to solve this issue doesn't make much sense.

As for Andy...I have not seen it reappear by itself even after printing
but I don't print all that much. Thus I don't see much need in patching
source and recompiling (seems to be overkill).

Craig

--
fedora-list mailing list
fedora-list@redhat.com
To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list

Craig White wrote:

On Thu, 2008-07-24 at 13:21 +0000, g wrote:


Andy Eager wrote:
<snip>


I have no idea how this crap got into cups or even why authentication
should be necessary to print locally.
Anyway, it has nothing to do with the security deny/allow stuff in
cupsd.conf


this got me wondering.

craig,

have you give thought to *group*?

because problem is with *authentication*, are you
'authenticated' if you add your user name to 'cups group'?

ria, under md 9 or 10, i had problems in printing as a 'user'
and i believe that doing so enable me to print.

i do recall that i still had to enter 'root' password to make
any changes to configs.


----
The problem isn't with authentication, the problem is that somehow the
printer configuration inherited an authentication requirement when there
was no desire or action for it to do so.


Absolutely correct.



Mucking in groups to solve this issue doesn't make much sense.

As for Andy...I have not seen it reappear by itself even after printing
but I don't print all that much. Thus I don't see much need in patching
source and recompiling (seems to be overkill).


If it's not reappearing then you're lucky in which case recompiling
would be overkill.

For me, (under F9 at least) it did reappear and I had no real choice.

(I got sick of removing the line & restarting cups every time I
wanted to print)







Craig







--
fedora-list mailing list
fedora-list@redhat.com
To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list

On Thu, 2008-07-24 at 23:48 +1000, Andy Eager wrote:

> > As for Andy...I have not seen it reappear by itself even after printing
> > but I don't print all that much. Thus I don't see much need in patching
> > source and recompiling (seems to be overkill).
> >
> If it's not reappearing then you're lucky in which case recompiling
> would be overkill.
> For me, (under F9 at least) it did reappear and I had no real choice.
> (I got sick of removing the line & restarting cups every time I wanted
> to print)
----
one of the first things that I did when I ran into the problem was to
move /etc/cups/cupsd.conf and replace it with the rpmsave version since
this was an upgrade from earlier versions of Fedora.

I suspect that the earlier version of my cupsd.conf contributed to the
problem.

Was yours an upgrade from an earlier version of Fedora or was this a
clean F9 install?

Craig

--
fedora-list mailing list
fedora-list@redhat.com
To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list

Craig White wrote:

On Thu, 2008-07-24 at 23:48 +1000, Andy Eager wrote:




As for Andy...I have not seen it reappear by itself even after printing
but I don't print all that much. Thus I don't see much need in patching
source and recompiling (seems to be overkill).



If it's not reappearing then you're lucky in which case recompiling
would be overkill.
For me, (under F9 at least) it did reappear and I had no real choice.
(I got sick of removing the line & restarting cups every time I wanted
to print)


----
one of the first things that I did when I ran into the problem was to
move /etc/cups/cupsd.conf and replace it with the rpmsave version since
this was an upgrade from earlier versions of Fedora.

I suspect that the earlier version of my cupsd.conf contributed to the
problem.

Was yours an upgrade from an earlier version of Fedora or was this a
clean F9 install?


Sorry, actually F8. (though I think that's not relevant. It's the cups
version that's relevant and seems to happen from 1.3.4 onwards.

It was a clean F8 install.

Even stranger, another box I have at work (also a clean F8 install) did
what you described above, where I only had to remove that line once.

If you look at the timestamp on your printers.conf, you'll see its
updated whenever you print (after the job is submitted)

Looking at the code, I think it might be a structure member that's not
initialised and therefore whether or not this happens can be 'random'.

(Didn't go into that deeply, just wanted it to work and was getting
quite annoyed that I was blowing away a day just to be able to print)







Craig







--
fedora-list mailing list
fedora-list@redhat.com
To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list

Craig White wrote:
<snip>

The problem isn't with authentication, the problem is that somehow the
printer configuration inherited an authentication requirement when there
was no desire or action for it to do so.


how ever, what ever, did put it in there. in one form of reasoning,
should every user be allowed to print? i would think so, but maybe
authors of 'cups' felt otherwise. i can see a reasons to limit use of
printing, but are they extreme reasons?.

like to keep someone from making printouts of material that they should
not copy. or someone who like to waste paper. or if a color printer,
printing nudes. extreme, yes. but reasons to limit access to printing.


Mucking in groups to solve this issue doesn't make much sense.


if you are a member of a *group*, are you not considered
'authenticated' and able to use what is in that *group*?

like 2 users who want to share files and not allow others to have
access. a third user group can be created and the 2 users added to
that group. then only the 2 have access.


from 'man group';

++++
DESCRIPTION
/etc/group is an ASCII file which defines the groups to which users
belong.
++++

anyway, need to go offline. back later.


--

tc,hago.

g
.

in a free world without fences, who needs gates.

--
fedora-list mailing list
fedora-list@redhat.com
To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list

On Fri, 2008-07-25 at 00:05 +1000, Andy Eager wrote:
> Craig White wrote:
> > On Thu, 2008-07-24 at 23:48 +1000, Andy Eager wrote:
> >
> >
> > > > As for Andy...I have not seen it reappear by itself even after printing
> > > > but I don't print all that much. Thus I don't see much need in patching
> > > > source and recompiling (seems to be overkill).
> > > >
> > > >
> > > If it's not reappearing then you're lucky in which case recompiling
> > > would be overkill.
> > > For me, (under F9 at least) it did reappear and I had no real choice.
> > > (I got sick of removing the line & restarting cups every time I wanted
> > > to print)
> > >
> > ----
> > one of the first things that I did when I ran into the problem was to
> > move /etc/cups/cupsd.conf and replace it with the rpmsave version since
> > this was an upgrade from earlier versions of Fedora.
> >
> > I suspect that the earlier version of my cupsd.conf contributed to the
> > problem.
> >
> > Was yours an upgrade from an earlier version of Fedora or was this a
> > clean F9 install?
> >
> Sorry, actually F8. (though I think that's not relevant. It's the cups
> version that's relevant and seems to happen from 1.3.4 onwards.
> It was a clean F8 install.
> Even stranger, another box I have at work (also a clean F8 install)
> did what you described above, where I only had to remove that line
> once.
> If you look at the timestamp on your printers.conf, you'll see its
> updated whenever you print (after the job is submitted)
> Looking at the code, I think it might be a structure member that's not
> initialised and therefore whether or not this happens can be 'random'.
> (Didn't go into that deeply, just wanted it to work and was getting
> quite annoyed that I was blowing away a day just to be able to print)
----
came back when I tried to print from Firefox...

see https://bugzilla.redhat.com/show_bug.cgi?id=456555

by all means add your insights/comments/frustrations

Craig

--
fedora-list mailing list
fedora-list@redhat.com
To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list

Craig White wrote:
<snip>

came back when I tried to print from Firefox...


so maybe you *and* firefox need to be in 'lp' group. <gbwg>


--

tc,hago.

g
.

in a free world without fences, who needs gates.

--
fedora-list mailing list
fedora-list@redhat.com
To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list

On Thu, 2008-07-24 at 20:51 +0000, g wrote:
> Craig White wrote:
> <snip>
> > came back when I tried to print from Firefox...
>
> so maybe you *and* firefox need to be in 'lp' group. <gbwg>
----
I hope that you won't mind that I will continue to ignore your
suggestions thinking that they are not useful and even if they worked,
that isn't the intended behavior and thus no solution at all.

Craig

--
fedora-list mailing list
fedora-list@redhat.com
To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list