password change does not work: LDAP, sssd, nss or pam error?
by Volker Potworowski
Hi everyone,
last week I switched my small home network from NIS to OpenLDAP (I am running
Fedora 13 on all machines). Everything went fine except the ability to change
LDAP user passords.
When I try to change the password as a (LADP-) user on a client I get
#> passwd
Changing password for user vp.
Current Password:
New password:
Retype new password:
passwd: Authentication token manipulation error
On the client side I get in /var/log/secure:
Oct 6 12:18:06 thal passwd: pam_unix(passwd:chauthtok): user "vp" does not
exist in /etc/passwd
Oct 6 12:18:43 thal passwd: last message repeated 2 times
Oct 6 12:18:43 thal passwd: pam_sss(passwd:chauthtok): Password change failed
for user vp: 28 (Module is unknown)
Oct 6 12:18:43 thal passwd: gkr-pam: couldn't update the login keyring
password: no old password was entered
I am obviously missing something in the pam-configuration (or in the slapd-
conf or wherever...). I do not have a clue where to search for the error.
Could somebody help, please?
Cheers
Volker
13 years, 7 months
... why top-posting should be avoided... (was Re: Broken mail readers (was Re: Properly wiping a hard drive ?))
by JD
On 10/10/2010 01:22 PM, Sam Sharpe wrote:
> You forgot to mention why top-posting should be avoided...
>
> --
> Sam
>
> On 10 October 2010 21:01, Tim<ignored_mailbox(a)yahoo.com.au> wrote:
>> On Sun, 2010-10-10 at 09:02 -0500, Matthew J. Roth wrote:
>>> I'd really appreciate it if anyone could explain why these things are
>>> happening and how I could configure the Zimbra Web Client to fix them.
>> You probably can't fix the threading problems, those things sound more
>> like faults, or badly designed software which simply doesn't do
>> everything that it should do with email. But the long line issue may be
>> a configuration option. All the things you brought up are mentioned
>> below.
>>
>> Threading - Mail clients insert a header, as they reply to a message,
>> that indicate which message you're replying to (the in-reply-to header
>> will be followed by its message id). And add the same message id to a
>> references header, which lists message ids of all the messages that
>> belong in the same thread. Then, when they show a list of messages, the
>> references headers are used to group them all together, and the
>> in-reply-to headers to put them in the right sequence.
>>
>> Headers - These are data put ahead of the message, not where you can
>> usually see them, but the pre-amble before the message. Use a "view
>> message source" option in your client to see them all.
>>
>> Chances are that if your mail client is destroying threading, then it
>> doesn't do anything with those headers (doesn't insert the in-reply-to,
>> and doesn't add to the references). View the source of one of your
>> replies to see if it does (send an email to your own address to test
>> without bothering anybody else with all your tests). You're unlikely to
>> be able to add the function to inadequate software. It's a basic part
>> of email clients, should already be there, there shouldn't be any
>> options to remove it, with the exception of anonymous mail services used
>> to protect people in certain countries and situations.
>>
>> Line length - Customary practice was to set software to wrap lines at
>> about 72 characters. That ensured that the message would fit into 80
>> column displays, even when there were a few generations of quotes
>> prefixed with ">" marks. Messages longer than the screen size would get
>> badly mangled, being hard line-broken, not re-wrapped. And often still
>> are, even with modern software, when someone replies to those messages
>> (just about everyone's seen hideously ugly and annoying to read messages
>> with alternating 79 and 6 character length lines). For those who think
>> that 72 is too short consider that (a) reading longer lines than that is
>> more difficult, and (b) it's still about the right size when printing
>> email to A4 or US letter paper, no matter what your printer resolution
>> is.
>>
>> As we all know, software authors, and users, ignore custom as they see
>> fit. And options creep in to change this, or the recommendation is just
>> plain ignored (no wrapping occurs, at all).
>>
>> No wrapping has its problems, especially with older client or server
>> software, as many had a 999 character limit, and would have some form of
>> error if the line was any longer. Each line in a message is parsed by
>> the software, headers and message content, looking for information that
>> it needed to handle the mail (addresses, content headers, etc.). And
>> such software may have only been able to handle 999 characters per line.
>> Not to mention the difficulty in reading messages with huge line
>> lengths, especially when users don't break their writing up into any
>> paragraphs.
>>
>> Options came to be that would allow you to send apparently unwrapped
>> text, so that the reader could resize their window to suit themselves,
>> wrapping your message within their window, but with the message source
>> still using short lines, that would keep all other software happy.
>> Various content-encoding schemes use some method that virtually says,
>> ignore the line breaks actually in the source, only wrap at a special
>> character sequence. And most mail clients would decode that content,
>> though some older ones cannot. They'll show those /codes/, or simply
>> show the text as 72 character wrapped lines.
>>
>> And that leads to a couple of things you can try to change how your
>> message does line length. Look for user configuration options regarding
>> line length or line wrapping. And/or try out different content-encoding
>> schemes (plain, quoted-printable, base64), your client may wrap
>> differently for some of them.
>>
>> Plain will just send out what you typed, as if you were still working
>> with ASCII-only computing. Email was originally only 7-bit, and often
>> still is, requiring encoding to get through many services.
>>
>> Quoted-printable, inserts control codes that may look odd (an equals
>> sign, followed by a number), but are a sequence of ordinary characters,
>> that you could just skip over if you happened to see them in a client
>> that can't handle them.
>>
>> Base64 will encode 8-bit textual data into sequences of characters that
>> only use 7-bit ASCII characters. It looks like gibberish to the naked
>> eye, hence why users still using *ancient* mail client software don't
>> like it.
>>
>> Don't, however, send HTML content-encoded mail to the list, it will not
>> be appreciated, and you will get flamed for it. You can Google why HTML
>> mail is bad, for yourself.
>>
>> See:
>> http://en.wikipedia.org/wiki/Quoted-printable
>> http://en.wikipedia.org/wiki/Email#Content_encoding
>>
>> --
>> [tim@localhost ~]$ uname -r
>> 2.6.27.25-78.2.56.fc9.i686
>>
>> Don't send private replies to my address, the mailbox is ignored. I
>> read messages from the public lists.
>>
>>
>>
>> --
>> users mailing list
>> users(a)lists.fedoraproject.org
>> To unsubscribe or change subscription options:
>> https://admin.fedoraproject.org/mailman/listinfo/users
>> Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
Perhaps the list server, when sending a final confirmation of subscription,
should include this reminder (I am sure there are other reminders that
everyone can think of). But this particular one has definitely caused many
reply posts asking the OP to not top post.
13 years, 7 months
Sound, Fedora13, and VMware
by Dean S. Messing
I just installed VMware Workstation 7.1.2 on my i686 machine. I
attached my old Windows XP client image, and things came up
nicely---except for sound ("no /dev/dsp found" was the sad
announcement).
Has anyone successfully fixed this on F13?
I've run across one solution (uncommenting the magic line in
/etc/modprobe.d/dist-oss.conf to enable legacy OSS) but the person who
did it had first ripped out the PulseAudio stuff and was running
straight ALSA.
Dean
13 years, 7 months
Broken mail readers (was Re: Properly wiping a hard drive ?)
by JD
On 10/09/2010 07:24 PM, James McKenzie wrote:
> On 10/9/10 10:07 AM, mike cloaked wrote:
>> On Sat, Oct 9, 2010 at 6:04 PM, Patrick O'Callaghan
>> <pocallaghan(a)gmail.com> wrote:
>>> On Sun, 2010-10-10 at 01:25 +1030, Tim wrote:
>>>> I see this quite a lot. There's a few clients around that are just
>>>> plain broken, and shouldn't be used with mailing lists.
>>> +1
>>>
>>> IMHO the most egregious offender is Yahoo Mail. Friends don't let
>>> friends use this POS on mailing lists.
>>>
>>> poc
>> This discussion is moving somewhat off topic!
> Mike:
>
> Should we start another thread talking about broken mail readers?
>
> I am one of the GUILTY parties and I explained that the problem in my
> case is a hosed-up mail reader provided by Earthlink. BTW, this is NOT
> the first mailing list to complain about this problem with my mail reader.
>
> Guess it is time to 'ping' Earthlink again.
>
> James McKenzie
>
Hi James,
Does earthlink allow you to use your own mail client
like thunderbird, or some other gui client?
I know yahoo does not allow that on free accounts;
i.e. they do not provide pop or imap server for your
free account.
But google does. So, why not get a free google account
and use thunderbird for reading and sending email?
Spare yourself the stress of fighting a battle with earthlink.
Cheers,
JD
13 years, 7 months
Python 3
by Zoltan Hoppar
Hi guys,
If I would like to run to learn python 3 code - is it possible to have
it installed in fedora, or is it already both 2.6 and 3.1.2 are
included?
Zoltan
--
PGP: 06853DF7
13 years, 7 months
install fedora 13 problem
by Mirko Jankovic
Hey people.
I'm jumping into fedora train right now but got some problems :)
I've already got two installs without much problems.. one on my laptop and
another fedora13 install on an older comp just to test it a bit before
installing it on my main comp.
But now I've got a problem.
Both those install went without problems but when I started installation on
my main comp first I've noticed some graphic problems
when there is installation started and there is mouse and fedora 13
welcome installation screen. there are some like glitches or something
no idea how to describe them.. like bad pixels in small upper part of the
screen.
But ignoring that everything went OK up to the part where I can choose a
standard discs or non starts like sans and stuff..
When I choose standard disc installation just freeze there. No idea what is
happening.
Only this that came across my mind is that there is a problem with raid
maybe?
There are 2x640Gb HDD in raid0, there is win7 installed,
and another 1Tb HDD, non-raid member disk, where I would like
to install Fedora and have a dual boot.
Still got programs that I need win for I'm afraid.
Sooo.. any idea what could be wrong here?
Installation went smooth on laptop and another comp but there is no raid so
I'm guessing that could be a problem.
This comp where installation froze is in short:
Intel core2duo 2.6
4Gb RAM
GeForce GTX 285
2x640 GB in raid0
1Tb non-raid separate disk.
Would be shame to stop my fedora trip right at the installation :)
Thank you!
Mirko
13 years, 7 months
repo error - anyone else seeing this?
by Dave Stevens
Error: Cannot retrieve repository metadata (repomd.xml) for repository:
fedora. Please verify its path and try again
Ideas, anyone? I've done # yum clean all
Dave
13 years, 7 months
touchpad not working in Fedora 13 x86_64
by Paul Morgan
Problem
-------
Touchpad does not work on a Sony Vaio laptop VPCEB33FM with
Fedora 13 x86_64.
Troubleshooting
---------------
I've looked at http://fedoraproject.org/wiki/Input_device_configuration
and performed some troubleshooting.
Files and command output
~~~~~~~~~~~~~~~~~~~~~~~~
I've posted several files for review at:
http://github.com/jumanjiman/sony-files
You can browse the files with a web browser; you don't
have to know git to see the files.
* dmidecode
* lspci -vvvv
* cat /proc/bus/input/devices
* /etc/X11/xorg.conf.d/00-touchpad.conf
* /var/log/Xorg.0.log
What I know
~~~~~~~~~~~
* The mouse is /dev/input/event2 (/dev/input/mouse0)
* loading synaptics driver results in error in Xorg.0.log
* loading evdev driver appears to load without error, but:
- cat /dev/input/mouse0
- cat /dev/input/event2
- move finger around touchpad fails to emit output
* I've tried custom input file /etc/X11/xorg.conf.d/00-touchpad.conf
with several iterations and can confirm that my inputclass
is recognized based on the identity (e.g., paul00 mouse0)
showing up in Xorg.0.log (see github URL above for actual files)
Any suggestions would be great.
tia,
-paul
13 years, 7 months