I think that this question must be noted as an important issue. But,
discussing it right now is premature - the discussion should be resumed
when Mageia has / will imminently publish a release and has become a
serious alternative for consideration by a commercial company, which is
not the case
On 09/24/2010 02:15 PM, Gustavo Ariel Giampaoli wrote:
I mean, I have a problem, I report the bug, and I keep bothering until
someone answer me.
Again: we are talking about a community distro - you, being on this list
at this time, are community elite (smiley). You know, I know whom to
keep
I agree with those who say: give them time. I understand that the
founding group works full steam on creating a documented concept that
can be shown to the community, and that will serve as a base to set up
an organisation.
I guess that the amount of time needed for formulating the concept,
I like the idea to lean on software authors and this way increase the Y.
But it only works where these authors are somewhere within reach. In
case where the author is somewhere upstream that will be difficult.
Btw upstream bugs: this is another argument illustrating where the
reporting user
On 09/26/2010 11:41 AM, Ahmad Samir wrote:
A bigger triage team
without a higher number of devs/packagers fixing bugs isn't gonna
change the situation that much
Michael Scherrer also insisted on this - I have been somewhat naive, I
now realise that this is a/the key issue.
Let me continue
On 09/27/2010 02:19 AM, Tux99 wrote:
Agreed, if we start questioning everything we might as well all go home.
Agreed again.
The first priority of Mageia is to publish, within a relatively short
time and with excellent quality, a release - and, as a base for that, to
get organised.
On 10/01/2010 11:08 AM, Marc Paré wrote:
I don't think that the mirrors will necessarily drop mirror support for
Mandriva for Mageia because the Mandriva brand does carry some weight
and they have not, as of yet' abandoned the developing MandrivaLinux.
They are still in the server market and
There is a single high-priority item for me: release-naming should
support evaluation in a script: a single and not excessivly complicated
regular expression should allow to detect whether a release is older or
younger then a reference release. In case Mageia decides to name
releases after the
Would be nice if release naming does not need to be changed if policy on
release periodicity changes (that may happen). But I agree - this is not
high up in the priority of things to be discussed - just collecting
input for future decisions.
Another idea triggered by watching the discussions on Mageia-discuss -
immediately triggered by Install Mageia from Windows, but also by the
observation that Mageia-discuss is a very noisy mixture of highly
specialised discussions and general considerations (a similar reflection
may also
On 10/12/2010 10:54 AM, Romain d'Alverny wrote:
There needs to be some sort of formal designation (to know who is
working with whom and on what) as well as progress/final reporting and
archives of discussions or results. But that's something to iron out
of practice.
Sure - I did not want to
On 10/13/2010 08:01 PM, Margot wrote:
Our target market should be current and future Linux users
worldwide, male and female, aged 0 upwards.
Are you being excessively opportunistic or not opportunistic at all?
(and I appreciate that there is no upper limit of age) (-
On 10/16/2010 09:29 AM, Morgan Leijström wrote:
I agree with most everything said
Not quite evident to understand in this quite controversial discussion!
I cant keep my tong away from my cheek - sorry. The real point for my
reply is: I am getting more and more confused with the ultimate goal
On 10/23/2010 07:16 AM, Ahmad Samir wrote:
With the current installer you can select customize (instead of
selecting KDE or GNOME) and uncheck all the options there when you
click next the installer will even give you options to:
- install/not-install X
- install/not-install suggested packages
-
The structure of packages as they are offered in the installer tries
somehow to address this issue. The affiliation of packages to groups and
sub-groups has been changed serveral times - an evident indication that
the need to change has already been perceived.
The problem with the existing
Yes, and sorry, I used far too many words to ruminate on the question:
are there technical means to improve what, fundamentally, exists already
and make it more appropriate to the needs that have been distilled in
this discussion. Thanks for including this issue in your list.
A couple of loose ideas that come when following the discussion:
1. Have packages pass description QA reading, done by non-technical
users (if implemented, easy with packages that are new or underwent a
major new release, how also catch packages that just are around)?
2. As an alternative to
I did not want to steal your ideas - precisely the opposite: I tried
to repeat some salient points which have come up several times - I was
afraid that they might get lost in the wealth of input to this list :)
On 10/29/2010 05:44 PM, Dj Marian wrote:
So instead of guessing, or saving on the internet maybe a quick question
at the beginning of the install asking ´How do you rate your experience
with linux?´ beginner medium advanced expert... new ones should get a
small list to choose programs from
An alternative for the practical information of this help information:
Personally, I have made a habit of putting this kind of help into popup
windows that pop up when you hover over the question.
- that avoids cluttering the display with information which half of the
users dont need
-
On 10/31/2010 06:26 AM, Marc Paré wrote:
I like the Community approach that Wobo suggested.
Yes - and that could be done also for less well-organized
sub-communities: there might simply be a list of neighbours who are
willing to help out (unless slow dial-up lines are due to a general lack
On 10/31/2010 04:24 PM, Marc Paré wrote:
I agree. This is where a great Mageia Community would delegate the CD
burning to a member or neighbour. Maybe even offer some help with
installation and trouble-shooting too.
Yes, that is the spirit of my suggestion (and that does not go far
beyond
On 11/05/2010 06:55 AM, andre999 wrote:
I would make all delta updates relative to the distro release, i.e.
- main = foo-1
- upd_full = foo-2, foo-3, foo-4, etc
I have a feeling that just distributing snapshots of the update
repositories (they are relative the release) is a wise thing to
I perfectly agree with what you say, and I agree that this list is
suffering from an excessive amount of messages.
Formally: I forwarded an urgent warning of common interest to colleagues
and used the only channel available - no discussion about Mandriva at
all. What - quote - Mandriva
Well, in fact, given the way tcp work, I do not see how it could
suffer from a transmission problem.
Tcp corrects errors with a certain likelyhood. I do not know what
CRC-something is used by ftp - for CRC-16, depending on the type of
error and the package length, probablilities are smaller
It seems like this is quite more complicated then I thought.
Not really so complicated: with the technical specs, Mageia decides and
has decided.
As for the hosting, I will talk to people at Switch and ask how much
they can help: minimum = stable in summer; maximum = cauldron now with 2
On 02/19/2011 09:28 PM, James Kerr wrote:
Open a bug at: ...
As the other replies point out, opinions on this issue are quite
different. Before filing a bug, I suggest to see whether some kind of
aggreement can be reached.
My opinion:
1. I have lived for several years with having several
It might be more emotional than stupid - the quoted website is in Brasil
where Mandriva certainly does a lot of promotion.
I had a difficulty with making a Perl-Tk application work in Mageia -
the build failed with the message missing tcl.h. To make it work, I
had to install tcllib-devel.
That was not necessary under Mandriva, and does not make sense: if you
use Perl-Tk rather than Tcl-Tk you precisely want to
Talking to others helps thinking oneself - I agree, the perl Inline
module has nothing whatever to do with tcl - Inline is an infrastructure
for including sources from several languages.
That made me re-check my application once more: found a perfectly
un-needed #include tcl.h ... (a leftover
I have been wanting such a feature for a long time and have quite a bit
played around with possible approaches.
I agree with many things said
- Such a tool can substantially shorten the time needed for customizing
an out-of-the-box system.
- A tool is better implemented as an pure
You need to click on the up arrow to see more icons, and you can use
the shorcut ALT+D,S to configure it.
Thank you - that works
Looks weird; picking up this discussion I verified on my laptop
- the snd_hda_intel driver is loaded
- task-pulseaudio is not installed
- but a lot of ...pulse... packages are installed (see list at the
end of this message (all installed without asking for it)
- sound works
- I remember (about
I delayed this issue for quiet times - this may be a good moment:
When configuring / (dis)connecting a wlan interface, the user interface
to the network control center is not as bright as it could be: the
user does not get enough feedback on the success of what he wants to be
done.
When you
I have filed but #278
Typing in a console dolphin /home/harms I obtain the display of an
empty directory.
The display remains empty if I hit the bold Home label in the location
bar (to switch the contents of the location bar to /home/harms).
Displaying the directory /home works OK, themn selecting the icon
harms
Did you try with a fresh user (just in case) ?
I did - for the clean slate: same problem (I also checked re-booting -
also, just in case)
I have made a small script that prints a list of packages recently added
to the mirror (it does lftp -ls on the .../core/release directory of a
mirror and displays the names of all rpms that have a creation date
later than a given reference date). That helps to rapidly get informed
on what is
That could be complemented with something like the well-known periodic
ML membership reminders in order to avoid developing an account
museum: remind account owners - say every year - that the account still
exists and ask whether he wants it to be kept for another year. That
could be used for
Well, there is the rss feed of sophie for that
Sure, and I do not doubt the quality of the servers. But, with the time
needed for the first web page to come up, I already have the result of
the script on my screen (and I would still need to browse to the result
I want). The script simply
I have now arranged for a mirror to be set up at CUI - my old lab at the
university of Geneva - I hope it will come up in time for alpha-2.
CUI is glad to help, but that is not a permanent solution - CUI normally
does not provide this kind of service - permanent mirroring services
should be
I agree: Mageia can perfectly live without a mirror at Switch. I do not
agree that this kind of input from a service provision environment (I
was tempted to say run in a business environment) can just be shrugged
away - that was my point (along with the shock I perceived getting
confronted
His masters voice (header in the welcome screen of GNU/Emacs):
Welcome to GNU Emacs, one component of the GNU/Linux operating system.
On 03/22/2011 09:34 AM, Wolfgang Bornath wrote:
Wow! I'm in a time warp, thinking back 10-12 years when I spent my
time with emacs-lisp to make Emacs/Gnus do what I want it to do!
Worse than that - my fidelity to xemacs goes back more than 25 years,
when xemacs was about the only nice
Sorry, my message went away before I checked it - as a top-level message
rather than a reply - and I had not re-read Romains text: he does
clarify between web and marketing teams. That makes my example a poor
example, but does not change what I meant to say.
On 04/07/2011 06:03 PM, Oliver Burger wrote:
does contain all those clarifications Juergen asked for.
Kindly said. I had - some months back - looked at this page, re-visited
it the a few days ago but did not realise how much this page has
evolved, and that the bullets are now links with much
On 04/13/2011 07:03 AM, Dimitrios Glentadakis wrote:
I dont know, i think i read somewhere that kfind is dropped to keep the find
toolbar in dolphin that works with nepomuk.
If kfind goes away, life will become more difficult - back to
command-line find? ploughing through all the options?
This is the first time I remember seeing an article that overdoes
positive arguments: it describes Mageia as the Nachfolger - the
successor - of Mandriva.
That is correct for me, at the level of my personal usage. But as a
general observation it is neither kind nor fair to Mandriva.
Juergen
I joined the club and did an upgrade from an up-to-date Mandriva 2010.2
system with a customized KDE user (Laptop Dell Latitude D820).
The update went smoothly, but took ages (an install on my laptop takes
about 15 minutes, the update took about 45 minutes).
Some minor glitches during
On 05/22/2011 01:59 AM, Michael scherer wrote:
1 installation transaction failed:file /usr/bin/rcp from install of
krb5-appl-clients-1.0.1-1.mga1.i586 conflicts with file from package
krb5-appl-servers-1.0-4mdv2010.1.i586
Can you fill a bug report for this one ?
Done: bugzilla #1381
There appears to be a large number of users who are not satisfied with
the Mandriva way of mirror selection (count me in). But this should not
automatically be interpreted to Mageia needs easyurpm. Why not submit
the standard mirror selection offered by Mageia to an evolution that
takes the
Franklin - the X resolution problem might be the same I had on my Dell
Laptop - see Bugzilla #9. The suggested workaround corresponds to what
you have done to fix the problem.
If this really is the case, would be nice if you add a comment to #119
giving your specific data (and documenting
There is also the very practical problem of efficiency:
- Having to deal with an ML and a forum takes more time to deal with
than if all posts would be in a single medium
- Handling a large number of posts in a forum is much quicker than
reading such a number of ML posts.
The second point
On 06/30/2011 10:17 AM, John wrote:
Last message before yours was 14 hours ago. Some users do sleep
Or use the forum :)
I just went to http://www.mageia.org/en/contribute/. In the paragraph on
Donating there is a sentence We keep a public record of what we get
and how we use it with a link to
http://www.mageia.org/en/about/reports/
which gives Access forbidden, error 403.
That should be fixed.
Have you make a bugzilla report ?
Is this a bug, is this kind of administrative issue something that is
handled within the bugzilla reporting process?
Sure, I can easily file a bug - can I get confirmation on the
procedural question first?
Juergen
I am using BackupPC, no problems. Original reasons for my choice:
- allows to back up a mix of Linux and Windows systems
- is well supported and has a lively ML where advice can be sought
- is easy to maintain and operate
The user interface / control-facilities are nice, a good compromise
Some more suggestions of candidates to consider:
PCB
The pcb printed-circuit design tool is certainly a good candidate - a
must in my opinion. It is (one of) the best opensource tools in this
kind, stable and powerful, with excellent documentation.
The target community of this tool is
Hello - just as a dont giveup message: My production system on my Dell
Latidude D820 (also multi-boot) is Mageia 1 - no hicks whatever (all
updates installed up to yesterday).
However, I use a particular install procedure: start with a minimal KDE
install and post-install (script) all the
I see and start to understand (and I do not use MageiaUpdate).
I need password-less ssh between all Linux systems on my LAN. Worked
great until some weeks (?) ago. Still works for opening sessions from my
laptop to run on the server. But when the server opens a session on the
laptop, all at sudden a password is required.
My configuration is slightly
I forgot to reply to 2 of your questions: no known_hosts files created
in both cases (known hosts are defined in /etc/ssh) - ssh -v gives in
both cases
Host 'ltjuergen' is known and matches the RSA host key.
Found key in /etc/ssh/ssh_known_hosts:5
Yes, and the link is not removed - it
On 09/24/2011 12:06 PM, Maarten Vanraes wrote:
so you're saying that ssh doesn't accept authorized_keys file if it's a link?
That is the evidence - but it did somewhere (couple of weeks back) in
the past, I have used this approach for years.
have you tried mount --bind it?
I had not thought
Tried it now: works perfectly with mount --bind.
The questions remains:
- why does it happen on my laptop and not on my server ?
- is this a consequence of some fault I made and have to find,
or can this be reproduced elsewhere and a bug needs to be filed?
On 09/24/2011 06:30 PM, Deri James wrote:
It StrictModes is turned on in sshd_conf I assume the permissions of the
link itself is checked
StrictMode no or yes does not make any difference: password still
required if login from the laptop to the server
Another possibility is to set
On 09/16/2011 01:29 PM, philippe makowski wrote:
You may know it or not, many packages have no maintainer.
isn't time to set up teams ?
I did not keep the original message - what has happened in the meantime?
- has the goal concerning the number of maintainers been reached?
- is CERBERUS
On 10/04/2011 05:54 PM, Samuel Verschelde wrote:
we're already doing it with the mentoring program
I realise very much. But at the long term the problem might not be
mentoring, but recruiting - how?
ideas that you can apply to find more, please do
That is why I am making noise (-. But
Maybe my note was also susceptible to be misunderstood in its general
tone: I did not want to complain - quite the other way round: the
initiative to launch getting maintainers was a perfect thing done at the
perfect moment.
I just had not seen what happened after the initial exchange of mail
I am using precisely the same solution as morgan: for each bootable
file-system its own /home directory, all my permanent data live in a
separate (non-bootable) file-system (yes, and with binary uid/gid-to-
alphanumeric name mappings that match).
Several advantages to this solution:
-
The hard drive containing my system partitions is dying. I can still
read it, but will have to rapidly replace it by a new (and bigger) drive.
Does somebody have experience with - rather re-installing Mga and
Windows-7 partition - simply copying the corresponding file-system
(piped tar) and
Thank you, that sounds good, is a thumbs up (my doubts were mainly
with the making executable stuff and suspecting windows facies).
Before going to dd, I will try partitioning my new disk with gparted (or
the Mageia ISO) - do all the disk layout re-arrangements I can think of
and than copy
Thanks for all these comments. I am now mailing from my newly installed
(yes: installed) system. It is worth while to sum up my experience:
- Copying (tar) an about 15 Gb partition takes about 15 minutes ( 1
hour for the 3 partitions I needed) - much faster than any installing
- Perfect for
On 11/14/2011 03:30 PM, Morgan Leijström wrote:
... how did i end up giving support for a Microsoft product !?
Yes, shame on us - and a big thank you!
There only very few things I just cannot do without windows (wine cie
dont help):
- download updates for my GPS
- write firmware into a
On 11/16/2011 05:41 PM, Barry Jackson wrote:
sharing your thoughts and confirming that I am working along the right
lines.
Do right lines exist in linux? there are so many ways of making things
work and having fun - but, I, agree are there lots of wrong lines that
dont work. (-
it is only
Install the 9 packages with rpm -i
Great suggestion, will try.
But the other question remains: did I blunder and should throw ashes
over my head, or is there a serious problem, requiring a high-priority
bug to be filed? - this problem sounds so out of the world. Probably the
best is to wait
This could not be found because HAL was not running.
Ahh, that tells me something: after I had applied Rons recipy
(downloaded the 9 packages and installed them with rpm - followed by a
reboot) urpmi became launchable - but when the DVD was requested by
urpmi it complained about the missing
I have now filed:
Bugzilla #2476: the problem mounting a DVD medium
The urpmi / hal problem has been reported while we had this discussion:
Bugzilla #3458
a 3rd bug (the hostname problem mentioned in my original mail) had
already been filed: Bugzilla #2938 (workaround: immediately after boot:
Here is a placeholder-reminder for the release notes: they should
mention that the package apache-base does not exist any more, that the
package apache must be used instead.
Should I send this note somewhere else?
I wanted to make a small addition to the Alpha 1 wiki page, but could
not log in with my normal Mageia account (Incorrect password entered.
Please try again)
Is the new wiki already accessible for editing by community members? is
this something still suspended, or do I have a problem with my
I have seen that there is a small (apparently tested and smoothly
working) patch to backuppc that allows to have the web interface of
backuppc show graphics on pool usage.
Is there interest in having this integrated into the package being
prepared for Mageia 2? Debian has it, other users
Makes much sense - somebody who really wants it should make a case
upstream (there is some noise already - which triggered my question)
On 01/22/2012 12:16 PM, Macxi wrote:
KompoZer is an open source WYSIWYG HTML editor based on a rewritten version of
the now-discontinued Nvu editor. KompoZer is maintained as a community-driven
fork and is a project on Sourceforge
Up to some time (2 years ?) ago, kompozer was quite error-prone
I just started playing with alpha 3 and hit a somewhat weird situation:
The iso-dvd installs
urpmi-6.43-2.mga2.noarch.rpm
The result of doing urpmi --auto-select - first thing after install -
is downloading and installing
urpmi-6.43.1-1.mga2.noarch.rpm
That looks like an update where an
Dots and dashes
I suspected something like that, but got blind reading the same lines
all over again - sorry for the noise!
Lesson: only send this kind of message after a nights sleep!
How about suspending this discussion? Looking at the results, I am
surprised about the clear language they speak.
It is now time that Wobo digests the result of the poll into his report:
that is where the discussion should resume - gives time to sharpen the
knifes!
I measured the time between the moment I hit the bootstrap menu item,
and the time the kde login screen appears (fully updated cauldron Alpha
3, similar packages installed on both systems):
About 21 seconds for Mageia 1, 61 seconds for an up-to-date Mageia 2.
For an alpha, that is no issue -
On 02/08/2012 07:15 AM, Olav Vitters wrote:
Please give the output of:
systemd-analyse plot systemd-boot-plot.svg
That gives a message of 300 kbytes, might be hard on people who have a
slow connection. I am trying to put it on the server at my lab, but
presently have problems with that- do
The reason for my problem was a simple typo - needed our system engineer
to sort that out - this is my sloppy day. You can retrieve the file at
http://cui.unige.ch/~harms/Download/systemd-boot-plot.svg
Looks like substantial time is spent in network configuration: my
controller is an
On 02/09/2012 09:41 AM, Olav Vitters wrote:
Recommend filing a bug on bugs.mageia.org as I heard that the systemd
maintainer is on vacation atm. Please also attach the svg to the bug.
Sure, I will - Can you advise on what Component to file the bug -
Release? (there is no systemd rpm package)?
On 02/10/2012 01:37 AM, Morgan Leijström wrote:
Correct, filesystems on different drives should run in parallel.
What was the case with the system under observation?
That is a point I had never considered - used what sysinstall put into
my fstab.
Man fstab(5) says
This field is used by the
On 02/10/2012 10:59 AM, Marja van Waes wrote:
There is: systemd-40-2.mga2.src.rpm (please check your version, this
Yes, I replied from my Mageia 1 system and stupidly did the rpm-q there
- evidently with an empty result.
Thanks, I will do as you suggested - just waiting until the present
On 02/09/2012 12:16 AM, David W. Hodgins wrote:
Try specifying MII_NOT_SUPPORTED=yes in
/etc/sysconfig/network-scripts/ifcfg-eth0.
Regards, Dave Hodgins
Sorry, I nearly missed this one.
Now I changed MII_NOT_SUPPORTED=no to MII_NOT_SUPPORTED=yes
I measured 57 seconds - may be some small
Since a couple of days ( 2 or 3, no problem until then ) I do not manage
any more to create a clean and empty up-to-date Alpha 3 system - which I
would need to test some packages I am maintaining, and do that before
Beta 1 is created.
No problem doing the install (Alpha 3 i586 iso DVD), no
Do you have bandwidth enough to do a network install?
( = get the latest packages directly without updating )
Great, thank you, that should do the job (I get 600 kbps from the Wobo
machine, no bandwidth problem). I did a minimum install, and than from
console of the installed system a urpmi
Following the good advice from Morgan (network install) I have now been
able to make a clean install with all the latest updates and to go
through all the steps I normally do to customize and verify my
production system. A couple of interesting points, here is the lot:
1. After this minimum
Weird. Please file a bug.
I will do my best - start with buying a bunch of flowers for my wife (-
Please dont misunderstand my note: I wanted to draw a list of what a
user sees in what is going to become the Beta release - principle reason
is to allow double-checking that there are
This is a good concept - I am using something similar since about 10
years. I started off doing it with rsync. But the rsync mirrors kept
growing scarce - and the sites made noises like: plain users undesired,
rsync is mainly for mirror downloads. Recently there appears to be
movement back to
Since quite some time I have a problem with xemacs on cauldron:
My site initialisation file does a check to detect if xemacs is runs as
root. As a function of the result, the background is set to one of two
different colours.
That has worked for something like 30 years on a lots of different
Beta looks great, chapeau!
Some issues remain
- The configuration of the network interface during installation still
does not work (failure message after Testing your connection), but at
the first boot a correctly working network connection is there. Does
somebody know if a bug has already
On 02/25/2012 02:10 PM, Wolfgang Bornath wrote:
This a rather old one (may be even inherited from Mandriva):
https://bugs.mageia.org/show_bug.cgi?id=1452
Great, no need to do bug filing. I spent this afternoon with an
additional serious problem that concerns the udisks daemon. It makes
1 - 100 of 131 matches
Mail list logo