[gentoo-user] Re: RPM equivlents

[Eamon Caddigan]

Radu Filip <[EMAIL PROTECTED]> wrote:
> 
> Hi, on Gentoo, what are the equivalents of the following RPM options:

Does this mailing list have a FAQ? This should be added to it.

-Eamon


--
[EMAIL PROTECTED] mailing list

[gentoo-user] Lightweight httpd

[Eamon Caddigan]

Hey all,

I'm thinking of setting up a small httpd on my desktop computer. Since I
have a dynamic IP and no extra box to use as a "server", I just want
something really simple that will let me serve static files, e.g.
torrents, images, and the occasional MP3.

Some programs that look interesting (found 'em all in portage) are: boa,
monkeyd, cherokee, and fnord. Would anybody with experience in one or
more of these share their opinion? Are there any others I'm missing?

Thanks,
Eamon


--
[EMAIL PROTECTED] mailing list

[gentoo-user] Re: SCSI emulation under Gentoo

[Eamon Caddigan]

Mark Knecht <[EMAIL PROTECTED]> wrote:
> Hi,
>If I use a grub boot option like hdc=ide-scsi, then what /dev device do I
> mount in fstab to use the device?
> 
>Do I need to make any changes at all in /etc/devfsd.conf to make this
> work?

IIRC, the default devfsd.conf will create a symlink to hdc at
/dev/cdrom. You can use this in your fstab, or the complete path -- I'm
not sure if one method is prefered over another.

Assuming hdc is, in fact, removable media, have you considered
automounting the drive? I think there's a good need for a "Gentoo
Automounting Guide" -- I recently switched from autofs to supermount,
and I had to do a good bit of Googling to get going with each. There
are, as far as I'm aware, three ways of automounting removable media,
and I'm only aware of some of the pros and cons of the two I mentioned.

-Eamon


--
[EMAIL PROTECTED] mailing list

[gentoo-user] gtk2 (was Re: MozillaFirebird Font Issues)

[Eamon Caddigan]

Doug Gorley <[EMAIL PROTECTED]> wrote:
> Eamon Caddigan wrote:
>
>>Second, I just upgraded to 0.7-r1 from 0.6.1, and the font used for the
>>menu bar and bookmarks has changed. I don't care for this serif font,
>>and I can't figure out how to reset it in the options. I don't believe
>>my GTK2 theme is to blame -- other apps look fine, and changing themes
>>didn't help.
>>
>>Any insight into these problems would be greatly appreciated.
>>
> With regards to the last bit, are you certain you compiled Firebird with 
> the gtk2 USE flag?

Yep! The only reason I upgraded firebird was because I recently made the
jump to "+gtk2". Now Firebird's buttons, scrollbars, checkboxes, etc.
match most of my other apps. I'm looking to move completely to gtk2 as
soon as the Audacity/wxWindows issues are worked out.

On *that* note, can anyone suggest a good gtk2 terminal emulator that
doesn't rely on gnome libs? Also, are there any tools that make it easy
to "convert" a gtk2 theme to gtk1 or qt? 

-Eamon


--
[EMAIL PROTECTED] mailing list

[gentoo-user] MozillaFirebird Font Issues

[Eamon Caddigan]

Hi,

I'm having a couple issues with fonts in MozillaFirebird.

First, I've never been able to use some of the truetype fonts with
Mozilla. The fonts in '/usr/X11R6/lib/X11/fonts/truetype' are never
presented in the list of available fonts in the configuration window,
although TrueType fonts in *other* directories (e.g.,
/usr/share/fonts/ttf-bitstream-vera) are. I did notice that other
directories contain a file fonts.cache-1, while the problem directory
does not. Other programs, however, are able to use these fonts.

Second, I just upgraded to 0.7-r1 from 0.6.1, and the font used for the
menu bar and bookmarks has changed. I don't care for this serif font,
and I can't figure out how to reset it in the options. I don't believe
my GTK2 theme is to blame -- other apps look fine, and changing themes
didn't help.

Any insight into these problems would be greatly appreciated.

-Eamon


--
[EMAIL PROTECTED] mailing list

[gentoo-user] Rsync Outage

[Eamon Caddigan]

As seen on gentoo.org:

  update: Service should be largely restored now, though traffic seems
  to be high on most mirrors because of the outage. The cause of the
  problem seems to be localized to some filesystem corruption on the
  root partition.

Well?! Which filesystem was it running?

-Eamon


--
[EMAIL PROTECTED] mailing list

[gentoo-user] Re: i need to wine to run...

[Eamon Caddigan]

Eamon Caddigan <[EMAIL PROTECTED]> wrote:
> blade- <[EMAIL PROTECTED]> wrote:
>> Eamon Caddigan wrote:
>>>[EMAIL PROTECTED] <[EMAIL PROTECTED]> wrote:
>>>
>>>>If the setup program is made with install shield you need to install
>>>>dcom95.exe first into your fake_windows
>>>
>>>Hmm, where can this file be found? Microsoft has a page offering it
>>>(http://www.microsoft.com/com/dcom/dcom95/download.asp), but the link is
>>>broken. Couldn't find it on my XP partition, either.
>> 
>> Check the link again, I just downloaded it fine.
>
> D'oh! Clicking on the link in Firebird sent me to another page, but
> copying the link and wget'ting it worked fine. Thanks!
>
> Running the executable in wine, however, pops-up the following error:
>
> DCOM95 can only be installed on Windows 95.
> For Windows 98, please install latest DCOM98.
> For Windows NT, please install latest service 
> packs.
>
> I'm using wine-20031212.

Okay, I found the section of in the config where the windows version is
set, and it was set for win98. However, downloading and running dcom98
gave /this/ error:

A newer version of DCOM95 or DCOM98 had been
installed.
To override, you must uninstall the current
version first.

This leads me to believe that dcom has nothing to do with my wine
errors. Fortunately, I'm not trying to do anything important with Wine
-- just some children's games. I'll play around with this later.

Thanks again for the help,
Eamon



--
[EMAIL PROTECTED] mailing list

[gentoo-user] Re: i need to wine to run...

[Eamon Caddigan]

blade- <[EMAIL PROTECTED]> wrote:
> Eamon Caddigan wrote:
>>[EMAIL PROTECTED] <[EMAIL PROTECTED]> wrote:
>>
>>>If the setup program is made with install shield you need to install
>>>dcom95.exe first into your fake_windows
>>
>>Hmm, where can this file be found? Microsoft has a page offering it
>>(http://www.microsoft.com/com/dcom/dcom95/download.asp), but the link is
>>broken. Couldn't find it on my XP partition, either.
> 
> Check the link again, I just downloaded it fine.

D'oh! Clicking on the link in Firebird sent me to another page, but
copying the link and wget'ting it worked fine. Thanks!

Running the executable in wine, however, pops-up the following error:

DCOM95 can only be installed on Windows 95.
For Windows 98, please install latest DCOM98.
For Windows NT, please install latest service 
packs.

I'm using wine-20031212.

-Eamon


--
[EMAIL PROTECTED] mailing list

[gentoo-user] Re: i need to wine to run...

[Eamon Caddigan]

[EMAIL PROTECTED] <[EMAIL PROTECTED]> wrote:
>
> If the setup program is made with install shield you need to install
> dcom95.exe first into your fake_windows

Hmm, where can this file be found? Microsoft has a page offering it
(http://www.microsoft.com/com/dcom/dcom95/download.asp), but the link is
broken. Couldn't find it on my XP partition, either.

Thanks,
Eamon


--
[EMAIL PROTECTED] mailing list

[gentoo-user] XFS won't work with port -1

[Eamon Caddigan]

Another XFS thread -- but I've been able to get it working. I just can't
get it to work with the default port, -1.

Before, in /etc/X11/XF86Config, I had:
  FontPath   "unix/:-1"

and in /etc/conf.d/xfs, I had:
  XFS_PORT="-1"

With this configuration, XFS would log the following error:
  xfs: cannot establish any listening sockets
and X would log:
  Could not init font path element unix/:-1, removing from list!

However, when I changed the port from -1 to 7100 in both of the above
files, xfs started working fine. However, since I like to stick to
defaults, and I don't like not knowing why things don't work, I was
hoping somebody might have insight into *why* the port change worked.

Thanks,
Eamon


--
[EMAIL PROTECTED] mailing list

[gentoo-user] Re: joystick and Linux games

[Eamon Caddigan]

Andrew Gaffney <[EMAIL PROTECTED]> wrote:
> Now that I've finally got my joystick working, I can't remember what
> game I wanted it for from about 6 months ago when I originally tried
> to get it working. So now, in order to justify the time I spent
> getting the joystick to work, I'm looking for a good Linux game that
> supports the joystick ;) Any recommendations?

xmame and roms (provided you own the original arcade boards, of course).

-Eamon


--
[EMAIL PROTECTED] mailing list

[gentoo-user] Re: server use-flags

[Eamon Caddigan]

Ben Munat <[EMAIL PROTECTED]> wrote:
> First, apologies for (sort of) re-sending a thread, but I realized that 
> asking about the best packages to install for a web server was jumping 
> the gun a bit. I need to set my USE flags and update my system first.
>
> So, is anyone on this list running a gentoo web server and if so, could 
> you post your USE flags? Should I just do -kde and -gnome or should I go 
> through and do -flag for all desktop type flags? 

USE="-*" turns off all default flags, then you can enable only those USE
flags you know you want. Definitely the way to go for a server.

-Eamon


--
[EMAIL PROTECTED] mailing list

[gentoo-user] Re: problem installing audacity

[Eamon Caddigan]

[EMAIL PROTECTED] <[EMAIL PROTECTED]> wrote:
> On Sat, Jan 10, 2004 at 04:36:34AM -0600, Chris wrote:
>> >>> md5 src_uri ;-) audacity-src-1.1.3.tgz
>> >>> Unpacking source...
>>  *
>>  * Audacity will not build with wxGTK compiled
>>  * against gtk2.  Make sure you have set
>>  * -gtk2 in use for this program to compile
>>  *
>> 
>> !!! ERROR: media-sound/audacity-1.1.3 failed.
>> !!! Function src_unpack, Line 37, Exitcode 0
>> !!! Make sure -gtk2 is in USE
>> 
>> Which file do i need to edit and who to do as it requests?
>
> It seems that audacity depends on wxGTK. 
> But it need that GTK is compiled without using GTK2.
>
> So i think you have already installed and compiled GTK 
> and it has been compiled by
> using gtk2 that audacity doesn't want.
>
> So try to rebuild wxGTK and dont't make it depend on 
> GTK2 by using right flags.
> Then you have the prob to find out the others stuff that
> was compiled and installed by using wxGTK based on gtk2

Wow, that's a pretty bad bug. I recently changed to "+gtk2" in my USE
flags after seeing some pretty nice themes -- I was *about* to recompile
all the apps previously compiled with "-gtk2" -- so thanks for the heads
up on this.

If the comments on Bug 27804 are to believed, 1.2.0 will fix this mess. 

-Eamon


--
[EMAIL PROTECTED] mailing list

[gentoo-user] Re: nmap/iptables

[Eamon Caddigan]

Eamon Caddigan <[EMAIL PROTECTED]> wrote:
> SN <[EMAIL PROTECTED]> wrote:
>>
>> - Original Message - 
>> From: "Jorge Almeida" <[EMAIL PROTECTED]>
>>> (The 1597 ports scanned but not shown below are in state: closed)
>>> Port   State   Service
>>> 6/tcp  filteredunknown
>>> 25/tcp filteredsmtp
>>> 80/tcp openhttp
>>> 135/tcpfilteredloc-srv
>>
>> Okay the output here means, the firewall is blocking 6, 25,135, since they
>> show up here you didn't completely drop all packages, but only block them,
>> this is usually safe.
>
> What exactly is the difference?
>
> I ask because I'm also running shorewall, and although I've closed all
> but a couple ports, I get the following results when running nmap from
> an outside machine:
>
> (The 1527 ports scanned but not shown below are in state: filtered)
> Port   State   Service
> 113/tcpclosed  auth
> 139/tcpclosed  netbios-ssn 
> 445/tcpclosed  microsoft-ds
>
> Interestingly, I need to run 'nmap -PT ', where  is one
> of the ports I've opened, to make nmap realize the host isn't down.
> Presumably, this is because port 80 is closed -- but why would it, and
> all others, be reported as "filtered"?

After following the recent... "discussion" on gentoo-security, I
realized that the real difference between "closed" and "filtered" ports
is whether iptables is REJECTing or DROPping the packets. A quick change
in /etc/shorewall/policy has nearly all of my ports turn up "closed" in
an nmap scan, with a couple still being reported as filter.

My only remaining concern is a handful of ports (135, , -6668)
that aren't specifed anywhere in my shorewall configuration, but are
still dropping packets. Anyone know why these aren't closed?

-Eamon


--
[EMAIL PROTECTED] mailing list

[gentoo-user] Re: Memory Usage

[Eamon Caddigan]

Pooh Sun Tzu <[EMAIL PROTECTED]> wrote:
> Maybe it is just me, but I've never in my entire experience had swap 
> touched. And that is at 400 megs of RAM, in gnome, XFCE4, and fluxbox. 
> Cached, sure. But he is running into swap, and that isn't supposed to 
> happen.

Well, I wrote a script to check my memory usage and which processes I
was running, with the intention of tracking down what was to blame. I
found that I only hit swap when running bittorrent or ftp, which I
routinely use to download reasonably large files (500 MB). 

What suprised me before I started this thread wasn't that these
processes used a lot of memory, but the fact that my memory usage didn't
drop when they terminated. Now I have a slightly better understanding of
why this is, and I'm no longer worried about some renegade process
leaking memory.

Thanks again,
Eamon


--
[EMAIL PROTECTED] mailing list

[gentoo-user] Re: Memory Usage

[Eamon Caddigan]

David Hollister <[EMAIL PROTECTED]> wrote:
> On Wed, 2004-01-07 at 21:29, Eamon Caddigan wrote:
>> Sorry if this is a trivial question...
>> 
>> But I recently upgraded my RAM (from 245 MB to 768 MB), and because of
>> this, I've been paying close attention to my memory usage. I was pretty
>> surprised to see that I'm still hitting swap after running a couple
>> simple apps in X. 
>> 
>> ps shows that none of the currently running processes are using a
>> significant percentage of my RAM. Logging out and back in doesn't help.
>> I can only guess that the kernel is using all this memory -- can anyone
>> recommend a tool to watch kernel memory usage, or any possible fixes for
>> this? I can't believe I'm using more physical memory now than I
>> previously had in physical and swap together!
>
> It's probably not the kernel per-se, but the filesystem's buffer cache. 
> Run "free" and look in the "cached" column.  The fs buffer cache will
> use as much memory as you have.  You will still hit swap occasionally,
> but generally, buffers should be purged from the cache to allow other
> processes access to physical memory when needed.

Ah ha, there's almost 500 MB cached! I figured it was something like
this going on -- my system wasn't slowing down, but I was surprised to
see all the memory tied up. Now that I know this is normal, my mind is
set at ease.

Thanks for the replies!

-Eamon


--
[EMAIL PROTECTED] mailing list

[gentoo-user] Memory Usage

[Eamon Caddigan]

Sorry if this is a trivial question...

But I recently upgraded my RAM (from 245 MB to 768 MB), and because of
this, I've been paying close attention to my memory usage. I was pretty
surprised to see that I'm still hitting swap after running a couple
simple apps in X. 

ps shows that none of the currently running processes are using a
significant percentage of my RAM. Logging out and back in doesn't help.
I can only guess that the kernel is using all this memory -- can anyone
recommend a tool to watch kernel memory usage, or any possible fixes for
this? I can't believe I'm using more physical memory now than I
previously had in physical and swap together!

-Eamon


--
[EMAIL PROTECTED] mailing list

[gentoo-user] Re: need help unsubscribing

[Eamon Caddigan]

Gerhard W Gruber <[EMAIL PROTECTED]> wrote:
> On Tue, 6 Jan 2004 20:53:59 +0000 (UTC), Eamon Caddigan <[EMAIL PROTECTED]>
> wrote:
>
>>That's the mail-to-news gateway I'm using to read this list (right
>>now!).
>
>>Yes, that's it. If you remain subscribed, you can read and post to the
>>group through the news gateway -- but as I mentioned previously, I
>>haven't figured out how to remain subscribed but get no mail.
>
> But I don't use this software. The linux.gentoo.user/dev shows up on my
> regular newsserver, so is this true only because you are using that gmane or
> is this true in general?

I'm not sure! I wasn't aware of the linux.* hierarchy in general -- my
news admin must be badgered accordingly.

Looking forward to one less slrn instance,
Eamon


--
[EMAIL PROTECTED] mailing list

[gentoo-user] Re: need help unsubscribing

[Eamon Caddigan]

Gerhard W Gruber <[EMAIL PROTECTED]> wrote:
> On Tue, 6 Jan 2004 19:31:13 +0000 (UTC), Eamon Caddigan <[EMAIL PROTECTED]>
> wrote:
> 
>>You should be aware that you can only post to the group through gmane if
>>you're subscribed to the list. 
> 
> Whats gmane? 

That's the mail-to-news gateway I'm using to read this list (right
now!).

> I just wanted to unsubscribe because I prefer newsgroups over
> mailinglists. Do you mean that, when I now unsubscribe, I can use the
> newsgroup only for reading?

Yes, that's it. If you remain subscribed, you can read and post to the
group through the news gateway -- but as I mentioned previously, I
haven't figured out how to remain subscribed but get no mail.

-Eamon


--
[EMAIL PROTECTED] mailing list

[gentoo-user] Re: need help unsubscribing

[Eamon Caddigan]

Brian Hall <[EMAIL PROTECTED]> wrote:
> 
> P.S. I am trying to unsubscribe because I recently discovered there
> is a Usenet gateway for this list. How come the email addresses of
> posters to this list aren't obscured to prevent harvesting by bots?

You should be aware that you can only post to the group through gmane if
you're subscribed to the list. 

I've only put minimal effort (a mail to the help address) into finding
how to subscribe in no-mail mode. It may be inelegant, but it was much
easier to just have procmail send list messages to /dev/null.

-Eamon


--
[EMAIL PROTECTED] mailing list

[gentoo-user] Re: Tip Winamp 2 Skins for XMMS

[Eamon Caddigan]

Norberto Bensa <[EMAIL PROTECTED]> wrote:
> SN wrote:
>> you can use winamp
>> 2 skins *.wsz with xmms
>
> XMMS must die. It's ugly. Hard to use.

"Stop using a WinAmp clone."

> If you're using KDE take a look at JuK. Or Rhythmbox if you're on Gnome.

"Start using an iTunes clone."

-Eamon


--
[EMAIL PROTECTED] mailing list

[gentoo-user] Re: System Load Monitor

[Eamon Caddigan]

[EMAIL PROTECTED] <[EMAIL PROTECTED]> wrote:
> On Sun, 04 Jan 2004 01:39:49 -0500
> KamaolaKid <[EMAIL PROTECTED]> wrote:
>
> ok, below is the message and reply is it possible your (and the
> other 3 people before you who replied) mail reader is broken? WHY QUOTE
> THE ENTIRE FREAKING MESSAGE???! Are you just trying to annoy everyone?
> Is it some kind of meow meow plot? What is the problem here?
>
> How difficult is it to leave in the part that applies to your reply and
> delete the rest? Is it the concept of convenience for others? Ignorance
> of common courtesy? A desire to be annoying? 
>
>
>
> # note # you do not have to read the below quoted message as I did what
> the previous writer did which was to quote everything even though it
> didn't have a damn thing to do with the message I wrote.
>
>> gkrellm
>> 
>> -- Kyle S.
>> 
>> Steve B. wrote:
>> 
>> >-BEGIN PGP SIGNED MESSAGE-
>> >Hash: SHA1
>> >
>> >I would like to graphicaly be able to monitor various aspects of my
>> >gentoo system. For example, cpu load, mem load, eth load.  Are there
>> >any sys load monitors one would recommended (preferably one that will
>> >work with fluxbox).
>> >
>> >Thanks,
>> >Steve
>> >- -- 
>> >
>> >Perhaps human's greatest vanity is the belief
>> >that there is only one way to the divine. 
>> >(Scott Cunningham)
>> >
>> >-BEGIN PGP SIGNATURE-
>> >Version: GnuPG v1.2.3 (GNU/Linux)
>> >
>> >iD8DBQE/+CUxCY3CQTSteGwRAvoZAJ9muoISHsF4BHLxYe57b84/nBomxQCfal9s
>> >ZIJdPGa35yiYtZfB6cxYAYs=
>> >=RUmg
>> >-END PGP SIGNATURE-
>> >
>> >
>> >--
>> >[EMAIL PROTECTED] mailing list
>> >
>> >
>> >  
>> >
>> 
>> 
>> --
>> [EMAIL PROTECTED] mailing list
>> 
>
>

Don't top post.


--
[EMAIL PROTECTED] mailing list

[gentoo-user] Re: how long does it take to commit a patch to portage

[Eamon Caddigan]

Gerhard W Gruber <[EMAIL PROTECTED]> wrote:
> On Fri, 2 Jan 2004 20:31:30 +, Ciaran McCreesh <[EMAIL PROTECTED]> wrote:
> 
>>Not really. Contributed ebuilds will generally be accepted pretty
>>quickly so long as they meet the following criteria:
> 
>>* The package is maintained upstream
> 
> What exactly does that mean?

I read this to mean that Gentoo won't accept packages that aren't
actively being developed. Which, in my opinion, isn't necessarily a
great policy.

Sometimes, a program gets to the point where it does everything the
developer wants, and it's *done*. I submitted an ebuild for one program
I use *all the time*, but I guess it won't make it into portage because
it's a couple years old. I'm not taking this personally, but it's a
shame that a lot of Gentoo users will never use this package because
it's "unmaintained".

> Actually the only good experience with patch policy, I had so far, is
> with the wine folks. They accept patches from anybody and they are
> commited in a reasonable time. Of course I don' t expect each and
> every patch to show up immediately, but in other projects I also had
> the experience that I submitted code and you get no good answers. One
> time I submitted an entire new control for a gui lib, which the author
> said would be great to have and I never got a response. Not if the
> author accepts or not and if not for what reason whatsoever. This is
> quite frustrating.

That's the cool thing about (free and) open source -- if you don't like
where the project is headed you can just fork it. Maybe you don't have
time to maintain your fork, but if somebody else wants the feature you
added, they can have it.

-Eamon


--
[EMAIL PROTECTED] mailing list

[gentoo-user] Re: GUI install

[Eamon Caddigan]

Hemmann, Volker Armin <[EMAIL PROTECTED]> wrote:
> On Friday 02 January 2004 16:37, Mac Intyre, Steven wrote:
>> Hi all,
>>
>> Is there possible a GUI install interface for Gentoo - similar to
>> Mandrake.
> 
> no, and I hope, that there will be never one.
> genflags/genkernel and the rest of this crap is newbie-friendly and 
> non-working enough. A graphic interface would inhibit me, you and everybody 
> to install gentoo the way, I, you and everybody wants it.

I don't buy into the rhetoric about GUIs inhibiting power users. First
of all, the command line *is* a graphical user interface, albeit a
one-dimensional one. While it may be true that most other GUIs currently
*do* place unnecessary limitations on the user, it doesn't follow that
they must.

Right now just about every new Gentoo user installs using the command
line on one VC, with the complete install documentation on another VC,
computer, or pile of dead trees on their desk. Heaven help the n00bs if
they're trying to cross reference this with one of the alternate install
docs.

A good GUI could *integrate* Gentoo's (excellent) documentation with the
complete install process. There are plenty of sections of the
installation that could *easily* be automated, and there's nothing about
a GUI that would force us to try automating those that can't.

And if you don't like this theoretical GUI, you should be free to ignore
it. Power users trying something new and different won't have to bother
with it, and wouldn't want to anyway. 

-Eamon


--
[EMAIL PROTECTED] mailing list

[gentoo-user] Re: worm games

[Eamon Caddigan]

William Kenworthy <[EMAIL PROTECTED]> wrote:
> My youngster likes the "worm" type games on the console.  He used to
> play with "nil" back in the bad old (mandrake) days, but it seems this
> game is very out of date (mid 2000) and the tarball doesnt want to
> compile under latest gentoo.
>
> Does anyone know of a more modern version, or a similar game?

The CVS sources for NiL compile and run (I'm using gcc 3.2.3-r3), but
the program spews warnings and sound won't work. Looks like a fun game
though, it might be worth tracking down the sources of these errors and
making an ebuild (after submitting a patch upstream, of course).

-Eamon


--
[EMAIL PROTECTED] mailing list

[gentoo-user] Re: devfs problem

[Eamon Caddigan]

Jerry McBride <[EMAIL PROTECTED]> wrote:
>
> Call me OLD FASHION, but devfs is the first thing I dump after installing 
> gentoo on a new box. The steps I take are; recompile the kernel without devfs 
> support, add "gentoo-nodevfs" to the append line in lilo.conf, correct any 
> defvs associations iin lilo.con and fstab, run lilo, rc-update del devfs and 
> then reboot. Once booted, move to /dev and run MAKEDEV. If you are running 
> alsa, also run snddevices anad add in any symlinks you might need for various 
> applicaitons...
>
> Once done, no more devfs MADNESS.

Wow, that *does* sound easier than dealing with devfs.

-Eamon


--
[EMAIL PROTECTED] mailing list

[gentoo-user] Re: dvdwriter package

[Eamon Caddigan]

Gerhard W Gruber <[EMAIL PROTECTED]> wrote:
> I did an esearch and found the following two packages:
>
> *  app-cdr/dvdrtools
>   Latest version available: 0.1.4
>   Latest version installed: [ Not Installed ]
>   Size of downloaded files: 1,407 kB
>   Homepage:http://www.nongnu.org/dvdrtools/
>   Description: This is the dvdrtools package a fork from cdrecord which
> permits dvdwriting
>
> *  app-cdr/dvd+rw-tools [ Masked ]
>   Latest version available: 5.13.4.7.4
>   Latest version installed: [ Not Installed ]
>   Size of downloaded files: 71 kB
>   Homepage:http://fy.chalmers.se/~appro/linux/DVD+RW/
>   Description: A set of tools for DVD+RW/-RW drives.

Hey, good timing. I suppose I'm not the only one who got a DVD-burner
for Christmas () -- does anyone have a preference between the above
packages? What about burning DVD movies?

-Eamon


--
[EMAIL PROTECTED] mailing list

[gentoo-user] Re: File list from package?

[Eamon Caddigan]

Patrick Börjesson <[EMAIL PROTECTED]> wrote:
> AFAIK there's no way to see what files would be installed by an emerge
> before hand.

Well, it's pretty convoluted, but the following method will tell you
what files will be installed by a package if you NEED to know:

ebuild  install
cd /var/tmp/portage//image
find . -printf '/%P\n'

This will display the files that will be installed on your filesystem,
along with directories that will be created if they do not exist. Note
that using ebuild instead of emerge will cost you automatic dependancy
fetching, and that /var/tmp should be replaced with your PORTAGE_TMPDIR. 

If everything checks out, you can save yourself some time by installing
with 'ebuild  qmerge', but be forewarned that the package
won't be added to your world favorites file.

-Eamon


--
[EMAIL PROTECTED] mailing list

[gentoo-user] Re: Iso : To be or not to be bootable ...

[Eamon Caddigan]

Ciortea Cristian <[EMAIL PROTECTED]> wrote:
> I was wondering if there is a way to check and see if a .iso is bootable ..
> Anybody has a clue ?

Just a guess... 
od [-t x1] filename.iso | less

I'm going from memory here -- so please excuse any innacuracies -- but
the first several 2048-byte sectors of a non-bootable iso are filled
with zeros. Therefore, if you have data in the first sector, you've
probably got a bootable iso. Whether the boot sector works, however, is
probably more difficult to determine.

-Eamon


--
[EMAIL PROTECTED] mailing list

[gentoo-user] Re: [gentoo-users] self-compiling external programs

[Eamon Caddigan]

Primero.Franz <[EMAIL PROTECTED]> wrote:
> I'm tryng to compile by myself the CVS version of mplayer.
>
> when i try the normal
> ./configure
> make
> make install
>
> i receive this error message :
>
> bash: ./configure: /bin/sh: bad interpreter: Permission denied

This error is usually the result of attempting to run a shell script
located on a partition that has been mounted 'noexec'. Note that some
mount options (e.g. 'user') imply noexec.

HTH,
Eamon


--
[EMAIL PROTECTED] mailing list

[gentoo-user] Re: emerge failes

[Eamon Caddigan]

Michael Spohn <[EMAIL PROTECTED]> wrote:
> 
>> Are you able to run /bin/sh by itself?  My hunch is a hard drive problem
>> because everything else looks OK.
> 
> No problem at all.
> 
> Even a script "hello"
> 
> #!/bin/sh
> echo "Hello World"
> 
> produces "Hello World"

Out of curiosity, is this working script on a different partition than
the one giving you errors? Might the latter partition be mounted noexec?
Your error matches the one I get every time I try to execute a script in
/tmp, as I often forget that I have that partition mounted noexec.

-Eamon


--
[EMAIL PROTECTED] mailing list

[gentoo-user] Re: update: c++ performance: gentoo, debian, wi ndow s

[Eamon Caddigan]

Van Eps, Nathan D. (James Tower) <[EMAIL PROTECTED]> wrote:
>>From: Eamon Caddigan
>>
>>>Van Eps, Nathan D. (James Tower) <[EMAIL PROTECTED]> wrote:
>>> That is weird. A compiler should generate the same code whether it is
>>> optimized or unoptimized. It would be interesting to hear an explanation
>>> from the gcc folk as to what causes this.
>>
>>Plenty of programs break when compiled with aggressive optimizations --
>>why should a compiler be any different?
> 
> If the compiler broke (because of aggressive optimizations), the compiler
> wouldn't make an executable. And if you compiled the compiler with good
> optimizations, the compiler would simply run faster (not the executable it
> generates). Its like deterministic.
> 
> You expect your programs to run the same regardless of compiler
> optimization. You just want them to run faster...

That's a fair expectation, but not quite how it works. When optimizing
code, compilers rearrange instructions (among other things) to achieve
faster (or smaller, etc.) code. It's not too difficult to imagine an
incorrectly-optimized compiler creating executables that themselves
fail, rather then simply refusing to compile any code. Especially when
you consider how subtle the problems introduced by incorrect
"optimizations" can be.

-Eamon


--
[EMAIL PROTECTED] mailing list

[gentoo-user] Shorewall filtering (was Re: nmap/iptables)

[Eamon Caddigan]

Molnar Peter <[EMAIL PROTECTED]> wrote:
> On Sun, 2003-11-23 at 22:32, Eamon Caddigan wrote:
>> Interestingly, I need to run 'nmap -PT ', where  is one
>> of the ports I've opened, to make nmap realize the host isn't down.
>> Presumably, this is because port 80 is closed -- but why would it, and
>> all others, be reported as "filtered"?
>> 
> Because the packets are fragmented, and the fragments get through your
> firewall. To see them being completely closed, you need to use
> connection tracking. This will cause the packets being reassembled
> before they are filtered.

Interesting, thanks for the tip. Are any other Shorewall users doing
this? Is it worth the effort -- do packet fragments pose a reasonable
risk? 

-Eamon


--
[EMAIL PROTECTED] mailing list

[gentoo-user] Re: update: c++ performance: gentoo, debian, window s

[Eamon Caddigan]

Van Eps, Nathan D. (James Tower) <[EMAIL PROTECTED]> wrote:
> That is weird. A compiler should generate the same code whether it is
> optimized or unoptimized. It would be interesting to hear an explanation
> from the gcc folk as to what causes this.

Plenty of programs break when compiled with aggressive optimizations --
why should a compiler be any different?

-Eamon



--
[EMAIL PROTECTED] mailing list

[gentoo-user] Re: safe gnome/kde testing

[Eamon Caddigan]

Christian Schäfer <[EMAIL PROTECTED]> wrote:
> hi gentoo-user,
>
> I want to test gnome and kde and maybe some other desktops.
> but as soon as I decided which one to use, I want to get rid of the
> others. can I safly unmerge kde and everything is fine?
> or are there traces of the install left on my harddisk?

One very safe way to test KDE, xfce, fluxbox, WindowMaker, and a couple
others is with a Knoppix CD. This way, there's no risk of anything
getting on your hard drive, and you can rule out several contenders
without downloading or compiling a thing.

-Eamon


--
[EMAIL PROTECTED] mailing list

[gentoo-user] Re: nmap/iptables

[Eamon Caddigan]

SN <[EMAIL PROTECTED]> wrote:
>
> - Original Message - 
> From: "Jorge Almeida" <[EMAIL PROTECTED]>
>> (The 1597 ports scanned but not shown below are in state: closed)
>> Port   State   Service
>> 6/tcp  filteredunknown
>> 25/tcp filteredsmtp
>> 80/tcp openhttp
>> 135/tcpfilteredloc-srv
>
> Okay the output here means, the firewall is blocking 6, 25,135, since they
> show up here you didn't completely drop all packages, but only block them,
> this is usually safe.

What exactly is the difference?

I ask because I'm also running shorewall, and although I've closed all
but a couple ports, I get the following results when running nmap from
an outside machine:

(The 1527 ports scanned but not shown below are in state: filtered)
Port   State   Service
113/tcpclosed  auth
139/tcpclosed  netbios-ssn 
445/tcpclosed  microsoft-ds

Interestingly, I need to run 'nmap -PT ', where  is one
of the ports I've opened, to make nmap realize the host isn't down.
Presumably, this is because port 80 is closed -- but why would it, and
all others, be reported as "filtered"?

-Eamon



--
[EMAIL PROTECTED] mailing list

[gentoo-user] Re: building openoffice

[Eamon Caddigan]

Chris Graves <[EMAIL PROTECTED]> wrote:
> I have a request then:
> Would someone with adequate diskspace (and time, etc) be willing to 
> compile it for me (w/ my make.conf)?

Somebody already has:
emerge -pv openoffice-bin

-Eamon


--
[EMAIL PROTECTED] mailing list

[gentoo-user] Re: Gentoo internal structure

[Eamon Caddigan]

Sergey V. Spiridonov <[EMAIL PROTECTED]> wrote:
> Hall Stevenson wrote:
> 
>> Outside of Debian, I think people consider software "free" if they can 
>> download it, use it, and NOT have to pay for it. I do at least.
> 
> There are several meanings of the word 'free' in English. Here is the 
> quotation from Webster's Revised Unabridged Dictionary (1913)":
> This is the first one:
> 
> Free Free (fr=e), a. Compar. Freer (-~er); superl.
> Freest (-e^st). OE. fre, freo, AS. fre'o, fr=i;
> akin to D. vrij, OS. & OHG. fr=i, G. frei, Icel. fr=i,
> Sw. & Dan. fri, Goth. freis, and also to Skr. prija beloved,
> dear, fr. pr=i to love, Goth. frij=on. Cf. Affray,
> Belfry, Friday, Friend, Frith inclosure.
> 1. Exempt from subjection to the will of others; not under
>restraint, control, or compulsion; able to follow one's
>own impulses, desires, or inclinations; determining one's
>own course of action; not dependent; at liberty.
> 
>  That which has the power, or not the power, to
>  operate, is that alone which is or is not free.
>--Locke.
> 
> I do not quote all meanings (there are 16). Here is what you meant:
> 
> 
> 15. Not gained by importunity or purchase; gratuitous;
> spontaneous; as, free admission; a free gift.

Thank you for your insight. As Linux users, I'm sure we were all unaware
of the issues surrounding the GPL, or the political and philisophical
motivation of it's supporters. 

So inspired by your words am I, that I hereby officially announce my own
fork of Gentoo: FREEtoo. Everything will be exactly the same, except
only programs with FREE (OSI-approved) licenses will be allowed in
FREEtage (nee portage). 

Now I just need to find someone to host it.

-Eamon


--
[EMAIL PROTECTED] mailing list

[gentoo-user] Re: Masked package when upgrading world

[Eamon Caddigan]

Alexander Futasz <[EMAIL PROTECTED]> wrote:
> On Mon, 10 Nov 2003 15:46:57 +0000 (UTC), Eamon Caddigan wrote:
> 
> Currently when using a ACCEPT_KEYWORDS=arch system and mixing it with
> some ~arch ebuilds it will break 'emerge -UDvp world'. Leaving the -D
> (deep) out will fix this, but of course wont upgrade the ebuilds that
> aren't recorded in /var/cache/edb/world and have been installed as
> dependencies. What you could do, but what is pretty uggly is:
> 
> # emerge -Uvp `qpkg -I|sed s/every ebuild that is ~arch masked//`
> 
> Or instead of substituting the masked ebuilds, you could inject them.

Injecting didn't work perfectly -- `emerge -pvuD' world stopped trying
to downgrade Tcl/Tk, but `emerge -pvUD world' was still broken.

The final solution, as suggested by Jason Stubbs and Doug Weimer, was to
copy the masked ebuilds into a portage overlay and set these ebuilds to
"x86". 

> Also ugly IMHO. Hopefully portage-ng (the soon to be started new
> portage) will fix this.

Hopefully.

-Eamon


--
[EMAIL PROTECTED] mailing list

[gentoo-user] Re: Runnig an X Application without using the X-Server ?

[Eamon Caddigan]

Jason Stubbs <[EMAIL PROTECTED]> wrote:
> On Sunday 16 November 2003 07:11, Eamon Caddigan wrote:
>> I always thought this feature would be worked into Gnome or
>> KDE...
> 
> Doesn't gdm and gnome2 provide this functionality?

:)

Does it? A quick search didn't turn up anything.

Looks like Xandros does too. Too bad we'll probably never see the code
for it.

-Eamon



--
[EMAIL PROTECTED] mailing list

[gentoo-user] Re: Runnig an X Application without using the X-Server ?

[Eamon Caddigan]

David A. Bandel <[EMAIL PROTECTED]> wrote:
> On Sat, 15 Nov 2003 22:59:09 + (UTC)
> Eamon Caddigan <[EMAIL PROTECTED]> wrote:
> 
>> > Proper fast user switch would be such a killer cool app, I hope you
>> > do it.
>> 
>> I'd probably never do it alone. Instead: anyone else interested in
>> this, please drop me an e-mail. If there's enough people willing to
>> work on this, I'm sure we could have some code written soon. Mostly, I
>> need help recognizing all of the potential "got'chas" that currently
>> arise from running multiple Xsessions that we'll need to address. 
>> 
>> Work on the cool rotating rectangular-solid, textured with each
>> desktop, can come later. ;)
> 
> You know, you _can_ run multiple xdm (wdm/gdm/kdm, whatever-dm) sessions
> similar to what I mentioned.  You just need to add some lines for it in
> /etc/X11/?dm/Xservers.  You can even run logins from remote systems
> (assuming you've turned on that feature, I believe now it's off by
> default).

I'm not sure multiple *dm sessions would count as "fast user switching".
Ideally, the user would still have a choice of running a display
manager, and a single instance would handle the multiple displays.

> So the only "problem" you need to overcome is how to avoid the clunky
>++ sequence, and that should be simple enough to code up.
>  It's just a key sequence passed to the VT manager that isn't first
> intercepted by X or another program that has focus.

Well, another problem I already mentioned is handling sound. I'm sure
you can imagine what happens when you try running xmms when another user
is running it on a different display. Ideally, the fast-user-switching
stuff would take care of that, and other similar problems. Linux has
always been a multi-user OS... on the console. Now it's time to extend
this philosophy to X.

> Necessity is the mother of all hacks (and kludges, too).  I just don't
> feel the necessity.  In fact, I'd rather not my kids were changing VTs/X
> sessions all over :-).

I'll admit, I thought "fast-user-switching" was a neat feature in WinXP,
but I wasn't *really* convinced until I saw it in action on Panther.
Both implementations have several limitations, the flexibility of Linux
could really let it shine.

-Eamon


--
[EMAIL PROTECTED] mailing list

[gentoo-user] Re: Runnig an X Application without using the X-Server ?

[Eamon Caddigan]

Mike Williams <[EMAIL PROTECTED]> wrote:
> On Saturday 15 November 2003 22:11, Eamon Caddigan wrote:
> 
>> I've been using this method to allow my wife and I to share our
>> computer when booted into Linux, but it's no match for the "fast user
>> switching" features in Max OSX 10.3 and WinXP. Off the top of my
>> head, better security (see above) and handling of ALSA, as well as a
>> way to switch without using awkward key-combos[1] would be necessary
>> before this stopped feeling like a hack.
> 
> You could try qingy. A replacement for getty which can punt you
> straight into whatever WM you wish (or a text console) from which ever
> vc you happen to be at.  I can login on vc1, lock my desktop, then
> someone else with access can switch to vc2 and login to their own
> xsession.  Very cool

Hmm, that does look neat. I haven't done anything with DirectFB before,
but this might be worth playing with. Thanks for the pointer.

>> I've long toyed with the idea of created a set of scripts that made
>> it a little cleaner, or possibly integrating such features into xdm
>> (we're not using any display manager atm), but like most of my
>> projects, it would probably go unfinished. Is anyone else interested
>> in something like this? I always thought this feature would be worked
>> into Gnome or KDE, or put together by a "user friendly" distro like
>> Mandrake, before it trickled down to Linux at large. It might be a
>> real feather in Gentoo's cap if "we" did it first.
> 
> Proper fast user switch would be such a killer cool app, I hope you do
> it.

I'd probably never do it alone. Instead: anyone else interested in this,
please drop me an e-mail. If there's enough people willing to work on
this, I'm sure we could have some code written soon. Mostly, I need help
recognizing all of the potential "got'chas" that currently arise from
running multiple Xsessions that we'll need to address. 

Work on the cool rotating rectangular-solid, textured with each desktop,
can come later. ;)

-Eamon



--
[EMAIL PROTECTED] mailing list

[gentoo-user] Re: Runnig an X Application without using the X-Server ?

[Eamon Caddigan]

David A. Bandel <[EMAIL PROTECTED]> wrote:
> --=.CT)Vc9(kosSjud
> Content-Type: text/plain; charset=US-ASCII
> Content-Disposition: inline
> Content-Transfer-Encoding: 7bit
> 
> On Sat, 15 Nov 2003 12:59:45 +0100
> humbaba <[EMAIL PROTECTED]> wrote:
> 
>> Hi
>> I was wondering if it is possible to start an X program for instance
>> xMule and then logout to let somebody else login and let him do his
>> work, while xMule is still running.
> 
> Yes.
> 
> Go to a console screen (++ for example), login and type:
> startx -- :1
> note that you _must_ use two dashes.  This tells X to start on a
> different screen.  You can also tell it which vt to use if you want so
> you know where it is (man X).

I've been using this method to allow my wife and I to share our computer
when booted into Linux, but it's no match for the "fast user switching"
features in Max OSX 10.3 and WinXP. Off the top of my head, better
security (see above) and handling of ALSA, as well as a way to switch
without using awkward key-combos[1] would be necessary before this
stopped feeling like a hack.

I've long toyed with the idea of created a set of scripts that made it a
little cleaner, or possibly integrating such features into xdm (we're
not using any display manager atm), but like most of my projects, it
would probably go unfinished. Is anyone else interested in something
like this? I always thought this feature would be worked into Gnome or
KDE, or put together by a "user friendly" distro like Mandrake, before
it trickled down to Linux at large. It might be a real feather in
Gentoo's cap if "we" did it first.

-Eamon

[1] I know it's silly, but I always feel like I'm going to break my
monitor when I switch between X sessions or virtual terminals using
[]++.



--
[EMAIL PROTECTED] mailing list

[gentoo-user] Re: Masked package when upgrading world

[Eamon Caddigan]

Jason Stubbs <[EMAIL PROTECTED]> wrote:
> On Wednesday 12 November 2003 14:55, Eamon Caddigan wrote:
>> rd <[EMAIL PROTECTED]> wrote:
>> > If emerge -u world is trying to downgrade packages, you can use
>> > /etc/portage.mask and /etc/portage.umask to set your own permanent
>> > masks -- not trashed by an rsync.
>>
>> This seems like a cleaner solution (other than portage working right
>> ), but I'm having trouble.
> 
> Nice to hear that you got your problem fixed the correct way. For
> future reference, the file /etc/portage/package.{mask,unmask} are only
> for adding/ removing from/to /usr/portage/package.mask. If a package
> has KEYWORDS="~x86" and you have ACCEPT_KEYWORDS="x86", changing the
> above files wont make a bit of difference. 

Ahh, that makes sense. Now I wonder if the lack of documentation on this
feature is intentional. I saw some talk on the forums about this feature
being "for developers only".

-Eamon


--
[EMAIL PROTECTED] mailing list

[gentoo-user] Re: Masked package when upgrading world

[Eamon Caddigan]

rd <[EMAIL PROTECTED]> wrote:
> If emerge -u world is trying to downgrade packages, you can
> use /etc/portage.mask and /etc/portage.umask to set your
> own permanent masks -- not trashed by an rsync.

This seems like a cleaner solution (other than portage working right
), but I'm having trouble.

I created a file containing only the line:
>=tcl-8.4

And tried naming it:
/etc/portage.umask
/etc/portage.unmask
/etc/portage/package.umask
/etc/portage/package.unmask
/etc/portage/portage.umask
/etc/portage/portage.unmask

(those last two were clearly a shot in the dark)
But when I try `emerge -pv \>=/dev-lang/tcl-8.4', I still get the
message:

!!! all ebuilds that could satisfy ">=/dev-lang/tcl-8.4" have been
masked.

Which leads me to conclude that my unmasking isn't working.

> Somewhere on the forums there is a sticky topic about this
> relatively new emerge/portage feature.

It was looking around the forums that created the confusion over what
this file should be named. Every reference I found used something
different. I'm using portage-2.0.49-r15.

-Eamon


--
[EMAIL PROTECTED] mailing list

[gentoo-user] Re: Masked package when upgrading world

[Eamon Caddigan]

Doug Weimer <[EMAIL PROTECTED]> wrote:
> On Tue, 2003-11-11 at 07:37, Eamon Caddigan wrote:
> 
>> Well, as a temporary hack, I went ahead and 'injected' the latest
>> unmasked versions of Tcl/Tk (dev-lang/tcl-8.3.4 and
>> dev-lang/tk-8.3.4-r1). Now, `emerge -pvuD world' doesn't try to
>> downgrade Tcl/Tk, but `emerge -pvUD world' is still returning errors.
>> This will serve me well until I install another masked package (which
>> isn't likely to happen any time soon).
> 
> Have you considered copying the ebuilds for the ~x86 packages you want
> to a portage overlay? Then change the KEYWORDS in the overlay ebuilds to
> x86, so that they are no longer considered masked.

This does the trick: emerge -pvUD has no errors, and emerge -pvuD isn't
trying to downgrade Tcl/Tk. Thanks!

-Eamon


--
[EMAIL PROTECTED] mailing list

[gentoo-user] Re: X server - no screens found

[Eamon Caddigan]

downtime null <[EMAIL PROTECTED]> wrote:
> ah. i assumed incorrectly! ha!
> 
> i've been having trouble getting the drivers from nvidia to work, so i
> thought his problem was specific to the driver not working correctly.
> 
> oh well.
> 
> so, i've been having trouble getting the 'nvidia' driver to
> work. X returns '(EE) No screens found' (or something like that) when
> i have 'nvidia' loading. if i use 'nv', it works fine. i don't have
> the log showing this handy, but i can recreate it if needed.

Run lsmod. Is the nvidia kernel module running? If not, `modprobe
nvidia', and try running X with the nvidia driver again.

HTH,
Eamon


--
[EMAIL PROTECTED] mailing list

[gentoo-user] Re: Masked package when upgrading world

[Eamon Caddigan]

Eamon Caddigan <[EMAIL PROTECTED]> wrote:
> 
> Right, but since I already have the latest (unstable) versions of
> Tcl/Tk, upgrading/updating them doesn't change anything. My problem is
> that emerge is returning errors when I try to update world with the
> upgradeonly flag. 

Well, as a temporary hack, I went ahead and 'injected' the latest
unmasked versions of Tcl/Tk (dev-lang/tcl-8.3.4 and
dev-lang/tk-8.3.4-r1). Now, `emerge -pvuD world' doesn't try to
downgrade Tcl/Tk, but `emerge -pvUD world' is still returning errors.
This will serve me well until I install another masked package (which
isn't likely to happen any time soon).

I *really* wish these packages were slot aware. In addition to this
mess, I can't emerge PyOpenGL (I'm not the only one:
<http://bugs.gentoo.org/show_bug.cgi?id=32070>). I'm *this* close to
downgrading Tcl/Tk and installing the new versions "by hand", but then I
remember that I switched to Gentoo precisely because I was tired of
circumventing my old distro's "package manager".

Just another whining user,
Eamon



--
[EMAIL PROTECTED] mailing list

[gentoo-user] Re: assigning net.eth? to specific nethwork devices

[Eamon Caddigan]

Jeffrey Smelser <[EMAIL PROTECTED]> wrote:
> You can autoload the module on boot in the order you want them to be.
> 
> /etc/modules.autoload is a good start.

Out of curiosity, what's the difference between /etc/modules.autoload
and /etc/modules.autoload.d/kernel-2.X? It would make sense if you put
non-kernel-version specific modules in the former, but aren't all
modules kernel-version specific? 

-Eamon


--
[EMAIL PROTECTED] mailing list

[gentoo-user] Re: Masked package when upgrading world

[Eamon Caddigan]

Hall Stevenson <[EMAIL PROTECTED]> wrote:
> On Mon, 2003-11-10 at 09:47, Eamon Caddigan wrote:
>> Hall Stevenson <[EMAIL PROTECTED]> wrote:
>> 
>> Yeah, I realize I wasn't very clear: Tcl and Tk are the only masked
>> packages on my system. Trying:
>> ACCEPT_KEYWORDS="~x86" emerge -pvu tcl tk
>> Indicates that I have the latest versions installed. This won't help
>> with updating world, unfortunately. I am loathe to update world using
>> "~x86" -- really, I just want to find out which packages have newer
>> stable versions.
>> 
>> I imagine this wouldn't be a problem if the ebuilds for Tcl/Tk were
>> slot-aware. I *could* downgrade to 8.3.4, and then install 8.4 by hand
>> in /usr/local, but I really don't want to do that.. 
> 
> Try "U" instead of "u", like this:
> 
> ACCEPT_KEYWORDS="~x86" emerge -pvU tcl tk
> 
> As I understand it, mixing "unstable", i.e. x86 packages, with regular
> can confuse emerge/portage. The "u" flag will attempt to actually
> downgrade you. The "U" flag won't.

Right, but since I already have the latest (unstable) versions of
Tcl/Tk, upgrading/updating them doesn't change anything. My problem is
that emerge is returning errors when I try to update world with the
upgradeonly flag. 

-Eamon


--
[EMAIL PROTECTED] mailing list

[gentoo-user] Re: Masked package when upgrading world

[Eamon Caddigan]

Hall Stevenson <[EMAIL PROTECTED]> wrote:
> On Sun, 2003-11-09 at 23:14, Eamon Caddigan wrote:
>> I have only two masked packages installed, but they're causing me
>> trouble when updating world. I'm using Tcl and Tk 8.4.4, and the
>> following happens:
>> 
>> -- begin:
>> 
>> emerge -pvUD world
>> >>> --upgradeonly implies --update... adding --update to options.
>> 
>> These are the packages that I would merge, in order:
>> 
>> Calculating world dependencies -
>> !!! all ebuilds that could satisfy "=dev-lang/tcl-8.4.4*" have been
>> masked.
>> !!!(dependency required by "dev-lang/tk-8.4.4" [ebuild])
> 
> Try "USE="~x86" emerge -uv tk tcl" first. That should upgrade your x86
> builds by themselves. I just checked this page,
> http://www.gentoo.org/dyn/pkgs/dev-lang/index.xml, and the versions
> listed there are not as new as what you've got, so that's why I assume
> you used "~x86" to get them at some point.

Yeah, I realize I wasn't very clear: Tcl and Tk are the only masked
packages on my system. Trying:
ACCEPT_KEYWORDS="~x86" emerge -pvu tcl tk
Indicates that I have the latest versions installed. This won't help
with updating world, unfortunately. I am loathe to update world using
"~x86" -- really, I just want to find out which packages have newer
stable versions.

I imagine this wouldn't be a problem if the ebuilds for Tcl/Tk were
slot-aware. I *could* downgrade to 8.3.4, and then install 8.4 by hand
in /usr/local, but I really don't want to do that.. 

-Eamon


--
[EMAIL PROTECTED] mailing list

[gentoo-user] Masked package when upgrading world

[Eamon Caddigan]

I have only two masked packages installed, but they're causing me
trouble when updating world. I'm using Tcl and Tk 8.4.4, and the
following happens:

-- begin:

emerge -pvUD world
>>> --upgradeonly implies --update... adding --update to options.

These are the packages that I would merge, in order:

Calculating world dependencies -
!!! all ebuilds that could satisfy "=dev-lang/tcl-8.4.4*" have been
masked.
!!!(dependency required by "dev-lang/tk-8.4.4" [ebuild])

!!! Problem with ebuild sys-apps/man-pages-1.60
!!! Possibly a DEPEND/*DEPEND problem.

!!! Depgraph creation failed.

-- end

I haven't dug too deep; I wouldn't know how to address a "DEPEND/*DEPEND
problem". Is there a workaround for this?

Thanks,
Eamon


--
[EMAIL PROTECTED] mailing list

[gentoo-user] Re: iptables

[Eamon Caddigan]

Dennis Freise <[EMAIL PROTECTED]> wrote:
> --Signature=_Sun__9_Nov_2003_19_20_11_+0100_KkMeCY42_=g+UfKT
> Content-Type: text/plain; charset=US-ASCII Content-Disposition: inline
> Content-Transfer-Encoding: 7bit
> 
> On Sun, 9 Nov 2003 17:36:21 + (UTC) Eamon Caddigan
> <[EMAIL PROTECTED]> wrote:
> 
>> > root # iptables -t nat -I POSTROUTING -j MASQUERAQDE -s
>> > 192.168.1.3/16
>> > /lib/modules/2.4.22-ck1/kernel/net/ipv4/netfilter/ip_tables.o:
>> > unresolved symbol nf_unregister_sockopt
>> > /lib/modules/2.4.22-ck1/kernel/net/ipv4/netfilter/ip_tables.o:
>> > unresolved symbol nf_register_sockopt
>> > /lib/modules/2.4.22-ck1/kernel/net/ipv4/netfilter/ip_tables.o:
>> > insmod
>> > /lib/modules/2.4.22-ck1/kernel/net/ipv4/netfilter/ip_tables.o
>> > failed
>> > /lib/modules/2.4.22-ck1/kernel/net/ipv4/netfilter/ip_tables.o:
>> > insmod ip_tables failed iptables v1.2.8: can't initialize iptables
>> > table `nat': iptables who? (do you need to insmod?) Perhaps
>> > iptables or your kernel needs to be upgraded.
>> 
>> Had the same problem when I first tried to get ip_tables up and
>> running.  Some Googling revealed that ip_tables only works if
>> CONFIG_MODVERSIONS isn't set. Weird, huh?
> 
> I do have CONFIG_MODVERSIONS=y set and I do have masquerading up and
> running on ppp0.  Therefor I do think it actually _is_ working
> together :)

Wow, which kernel version are you running? CONFIG_MODVERSIONS was the
only difference between unresolved symbol errors in ip_tables.o and a
working setup for me, using gentoo-sources-2.4.20-r8. 

That kernel is a harsh mistress.

-Eamon


--
[EMAIL PROTECTED] mailing list

[gentoo-user] Re: iptables

[Eamon Caddigan]

Brian Doob <[EMAIL PROTECTED]> wrote:
>   OK, it's getting better, but it still doesn't work.  Here's what happens:
> 
> root # iptables -t nat -I POSTROUTING -j MASQUERAQDE -s 192.168.1.3/16
> /lib/modules/2.4.22-ck1/kernel/net/ipv4/netfilter/ip_tables.o: unresolved symbol 
> nf_unregister_sockopt
> /lib/modules/2.4.22-ck1/kernel/net/ipv4/netfilter/ip_tables.o: unresolved symbol 
> nf_register_sockopt
> /lib/modules/2.4.22-ck1/kernel/net/ipv4/netfilter/ip_tables.o: insmod 
> /lib/modules/2.4.22-ck1/kernel/net/ipv4/netfilter/ip_tables.o failed
> /lib/modules/2.4.22-ck1/kernel/net/ipv4/netfilter/ip_tables.o: insmod ip_tables 
> failed
> iptables v1.2.8: can't initialize iptables table `nat': iptables who? (do you need 
> to insmod?)
> Perhaps iptables or your kernel needs to be upgraded.

Had the same problem when I first tried to get ip_tables up and running.
Some Googling revealed that ip_tables only works if CONFIG_MODVERSIONS
isn't set. Weird, huh?

-Eamon



--
[EMAIL PROTECTED] mailing list

[gentoo-user] kernel bug on shutdown (umount segfault)

[Eamon Caddigan]

I'm getting the following error whenever I shutdown or restart:

kernel BUG at inode.c:1105!
invalid operand: 
CPU:0
EIP:0010:[]   Not tainted
EFLAGS: 00010246
eax: cd03ca20   ebx: cd03ca20   ecx: cd03cb34   edx: cd03cb34
esi: cd160600   edi:    ebp: d08fc3c0   esp: ce3e3f48
ds: 0018   es: 0018   ss: 0018
Process umount (pid: 7079, stackpage=ce3e3000)
Stack: cd160600 cd1606c4 cde95190 d08f69c3 cd03ca20 cd160640 cd160600 c01d3ebc
   cd160600 d08fc428  ce3e3f88 ce3e2000 bfff8858 c01e3e2d cd160600
   cde95190 c1abe510 ce2c44bc ce2c44b0 ce4d4210 0009 0001 bfff8798
Call Trace:[] [] [] [] []
  []

Code: 0f 0b 51 04 95 dd 30 c0 e9 17 fe ff ff 8d 76 00 8b 54 24 04
 /sbin/rc: line 141:  7079 Segmentation fault  umount -a -r -n -t 
nodevfs,noproc,nosysfs,notmpfs >&/dev/null

I suppose that means that umount is segfaulting. I've STFW'ed, and only
found a similar error on the kernel mailing list, and one reply claimed
that it was fixed in 2.4.20-pre11. Apparently not -- I'm using
gentoo-sources-2.4.20-r8

-Eamon


--
[EMAIL PROTECTED] mailing list

[gentoo-user] Re: Command Prompt question

[Eamon Caddigan]

Dennis Robertson <[EMAIL PROTECTED]> wrote:
> On Sat, 25 Oct 2003 20:00, Barry Marler wrote:
>> I may as well add that you'll have to re-source .bashrc, a la:
>>
>> source ~/.bashrc (or simply . ~/.bashrc).
>>
>> On Sat, 2003-10-25 at 05:55, Barry Marler wrote:
>> > Append
>> >
>> > PS1='\[\033[01;31m\]\h \[\033[01;34m\](\w) \$ \[\033[00m\]'
>> >
>> > to ~/.bashrc.
> 
> Hello Barry,
> Following your advice I get:  dencar (~) $  at the prompt. The aterm titlebar 
> reads: [EMAIL PROTECTED]:~ which is correct. Can you help further with the prompt 
> to have it specify user dennis, please. TIA. BTW there is a script in 
> ~/.bashrc to change the window title of X terminals, which I assume is 
> producing the titlebar entry I mentioned.

Read the Bash Prompt How-To at
, and pay close attention
to 2.4-2.6. This way, you'll be able to set your prompt to anything you
like without resorting to typing alien commands that make no sense to
you.

On my Gentoo system, I noticed that the prompt looked great for login
shells and screen sessions, but not terminals. The source of the
discrepancy is the fact that the PS1 environment variable, which
controls the appearance of your prompt, was being set in /etc/profile,
which IIRC is only sourced for login shells. The solution is to, as
Barry already pointed out, add an appropriate line to your .bashrc,
which is sourced for all shells. Others have suggested hacks to
/etc/profile, but I believe Barry's solution is "cleaner".

HTH,
Eamon


--
[EMAIL PROTECTED] mailing list

[gentoo-user] Re: How to connect broadband automatically at booting

[Eamon Caddigan]

M-A Loyer <[EMAIL PROTECTED]> wrote:
> Stephen Liu wrote:
>> 
>> Kindly advise how to configure the OS to have ISP/broadband connected 
>> automatically at booting.
> 
> You can add a start-up script for your DSL in your init.d folder.
> It should look like this : ( /etc/init.d/net.adsl )

Just out of curiosity, is there a good reason this script (or something
like it) isn't included with the rp-pppoe ebuild? It seems a lot of
folks would find it useful (mostly so other services, e.g. ntpd, can
require it) -- maybe this should go on bugzilla?

-Eamon


--
[EMAIL PROTECTED] mailing list

[gentoo-user] Re: OT: How many of you are 100% Linux?

[Eamon Caddigan]

dave willis <[EMAIL PROTECTED]> wrote:
> On Tue, 21 Oct 2003, tony Scharf wrote:
> 
>> I am NOT 100% linux, and I will tell you why: I am a musician.  
>> currently, there are NO apps that can even come close to comparing to
>> the capability and features of their windows/mac counterparts.
> 
> what apps do you use in windows?  i am an electronic musician, and all i 
> use is linux. 

Are there any apps even comparable to Cubase, Cakewalk Sonar, or even
Buzz Machines on Linux? gAlan looks interesting, but like all music
software, it appears to have a very steep learning curve, and I've only
begun to play with it. It also seems to be more of a "sound design tool"
similar to Kyma than a sequencer or tracker. 

> i do use cool edit 96 under wine for audiofile editing.

You might want to check out Audacity. 

-Eamon



--
[EMAIL PROTECTED] mailing list

[gentoo-user] Re: animated gifs

[Eamon Caddigan]

Andrew Gaffney <[EMAIL PROTECTED]> wrote:
> Eamon Caddigan wrote:
>> Andrew Gaffney <[EMAIL PROTECTED]> wrote:
>> 
>>>Is there a good program for Linux (preferably with an ebuild) that
>>>allows you to make animated GIFs. I'm interested in making 32x32
>>>animated GIFs for buddy icons for AIM.
>> 
>> 
>> ImageMagick's 'convert' tool can make animated GIFs. emerge
>> imagemagick
> 
> I also found gifsicle. I went over to my Windows computer. I whipped 3
> GIFs in Paint. I ran 'gifsicle --delay 500 --loopcount=forever
> frame1.gif frame2.gif frame3.gif > output.gif' and voila!

Glad it worked for you. I'm a big fan of ImageMagick,though -- it is to
images what SoX is to audio. I've also put the library to good use in a
couple image-manipulation programs of my own.

-Eamon


--
[EMAIL PROTECTED] mailing list

[gentoo-user] Re: animated gifs

[Eamon Caddigan]

Andrew Gaffney <[EMAIL PROTECTED]> wrote:
> Is there a good program for Linux (preferably with an ebuild) that allows you to 
> make 
> animated GIFs. I'm interested in making 32x32 animated GIFs for buddy icons for AIM.

ImageMagick's 'convert' tool can make animated GIFs. emerge imagemagick

-Eamon


--
[EMAIL PROTECTED] mailing list

[gentoo-user] Re: playing a VCD

[Eamon Caddigan]

Bruce E. Harris <[EMAIL PROTECTED]> wrote:
> I am able to mount this VCD, and I can see four directories, but it wont play 
> on anything. I tried gxine, and aKtion, yet neither can see the contents 
> either.
>
> Where can I go from there?

I like mplayer, which can play VCDs with 'mplayer -vcd '

-Eamon


--
[EMAIL PROTECTED] mailing list

[gentoo-user] Re: what is using all my processor?

[Eamon Caddigan]

Eamon Caddigan <[EMAIL PROTECTED]> wrote:
> Ernie Schroder <[EMAIL PROTECTED]> wrote:
>> 8394 root  25   0   796  796  624 R 99.0  0.2 127:50.12 dialog
>> I just noticed the box lagging a bit and from top I see the above
>> I have no VT's opened and no root consoles running, WTF is going on?
>> I have noticed that any Konsole I open dies after a few minutes and I 
>> need to open a new one to kill it. A few minutes later that one dies 
>> too. I'm going to reboot to see if that fixes it. Any ideas would be 
>> appreciated.
> 
> 'top' will tell you which processes are consuming the most CPU
> resources. 'ps' can give you more details.

Oops, I just realized that you're posting the output of top... Sorry.
Don't know what's responsible for 'dialog'.

-Eamon


--
[EMAIL PROTECTED] mailing list

[gentoo-user] Re: what is using all my processor?

[Eamon Caddigan]

Ernie Schroder <[EMAIL PROTECTED]> wrote:
> 8394 root  25   0   796  796  624 R 99.0  0.2 127:50.12 dialog
> I just noticed the box lagging a bit and from top I see the above
> I have no VT's opened and no root consoles running, WTF is going on?
> I have noticed that any Konsole I open dies after a few minutes and I 
> need to open a new one to kill it. A few minutes later that one dies 
> too. I'm going to reboot to see if that fixes it. Any ideas would be 
> appreciated.

'top' will tell you which processes are consuming the most CPU
resources. 'ps' can give you more details.

-Eamon


--
[EMAIL PROTECTED] mailing list

[gentoo-user] Re: Where the heck is Bitchx?

[Eamon Caddigan]

Jayson Garrell <[EMAIL PROTECTED]> wrote:
> This morning I needed to get some info from irc so I decided to fireup
> bitchx. I cannot find it on my system. Portage thinks it is there...

I've had a similar problem with other ebuilds, but this week's Gentoo
Weekly Newsletter has the solution: qpkg. Among other useful things,
qpkg can tell you what files were installed by a particular ebuild with
'qpkg --list '. No more sending the output of emerge to
a file!

Thank you, GWN!

-Eamon


--
[EMAIL PROTECTED] mailing list