Re: Execute shutdown as other user than root
Iago Sineiro wrote: Hi. I want to execute shutdown as other user than root. How to do it? Is it possible? Note: I want to do it in one box with Debian that doesn't have command sudo and I don't want to install it. make shutdown suid root and only executable by the shutdown group? -- Mental ([EMAIL PROTECTED]) Old aunts used to come up to me at weddings, poking me in the ribs and cackling, telling me, "You're next." They stopped after I started doing the same thing to them at funerals. CARPE NOCTEM, QUAM MINIMUM CREDULA POSTERO. GPG public key: http://www.neverlight.com/pas/Mental.asc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Single-use root account?
Ron Johnson wrote: On Fri, 2003-11-07 at 07:55, J. Bruce Fields wrote: Why not? They already have physical access to the machine, what more would you give up to them by telling them the root password? For a home computer, I don't see much reason not to just stick the root password on a post-it note on the monitor You already trust anyone that's in a position to see it. And if a not-so-trustworthy "friend" or acquaintance wanders by, he can destroy you. The all-privilege sudo is the best idea, since the actions are audited. Who cares? If a not-so-trustworthy anyone is there when you're not, the game is over. Physical access == no security. And the auding of full sudo is trivial to circumvent. sudo vim /etc/network/interfaces :!bash Anything done in this shell (which is running as root) will not be audited. or init=/bin/bash Insta-root, no password needed. No auditing. Works with just about any boot loader. Passworded boot loaders can be bypassed by bootable media. Passworded bios can be reset. And for the truly hard core, the hard drives can be removed, and remounted in another machine, modified, then reinstalled (yes, I'm wearing my tin foil hat today). If it were me I'd either just give them the root password or sudo. You could always do the low-tech method. Put the root password in a sealed evelope in your desk drawer. When needed, the seal can be broken and the password used. Post-use, change the password and put a new sealed evelope in the desk drawer. Either way its honor system based. Sudo implies trust not to abuse it as does giving the root password. -- Mental ([EMAIL PROTECTED]) CARPE NOCTEM, QUAM MINIMUM CREDULA POSTERO. GPG public key: http://www.neverlight.com/pas/Mental.asc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Crash in fork.c on everything = install hosed
Malcolm Box wrote: OK, that looks like it might fix things. But can anyone tell me how to get the new version onto my machine given that both dpkg & apt-get don't work at the moment? Is there some way to manually unpack the files from the .deb into the right places? Malcolm Last time I broke libc, I fixed it by booting off my rescue (knoppix) cd. Mounted all the filesystems under /mnt/debian. It looked like /mnt/debian /mnt/debian/boot /mnt/debian/usr /mnt/debian/var etc... I used knoppix to download the libc6 deb and copied it to the root filesystem (/mnt/debian) of the hosed installation. I then cd'd into /mnt/debian and did the following: ar x PACKAGE.deb That left 2 files, data.tar.gz and control.tar.gz. I untared data.tar.gz and rebooted. It worked, I was fine. Just to be sure, I did an apt-get --reinstall install libc6 to make sure everything was updated (package info, control scripts, etc). HTH -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Browsers that *don't* support about:blank
Tom wrote: On Mon, Nov 03, 2003 at 01:30:21PM -0500, Mental Patient wrote: Tom wrote: I filed a wishlist bug against "links" asking for it to support about:blank (highly useful in frames pages as a default for the "body" frame). Thats an abuse of side effects and its usefulness is debatable. A highly useful default page in a frameset is one with relevant CONTENT. If you really want a blank page, create a blank page. Okay, then the correct action is *not* to file a bug against links asking it to support about:blank. The correct action is to file a bug against Mozilla asking it to remove support for it in HREF and frame SRC arguments, correct? As long as we care about being standards-compliant. /me ducks The standard doesnt forbid it. Additional tags are what killed browser compatibility in the first place. We need more tags like iframe and embed. A generic csshole tag would suffice. And if at somepoint that happened (fixing href and src tags) and your site broke, where would that leave things? You're relying on a side effect subject to interpretation. Thats it. This is not a flame, just an observation. -- Mental ([EMAIL PROTECTED]) "The Torah... The Gospels... The Koran... Each claimed as the infallible word of GOD. Misquoted, misinterpreted, misunderstood, and misapplied. Maybe that's why he doesn't do any more interviews." - sinfest.com CARPE NOCTEM, QUAM MINIMUM CREDULA POSTERO. GPG public key: http://www.neverlight.com/pas/Mental.asc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Browsers that *don't* support about:blank
Tom wrote: How come y'all are being adversarial? I agree with the maintainer. I'm just curious what browsers support it :-) Jeez, grandstanding Sorry if I came off as adversarial. Neither lynx nor w3m support about:blank. Neither does wget, LWP::UserAgent or any of the Java user agents I've thrown it at. Python doesn't seem to support it either. Simply put, its not a valid url. Any language suited for doing HTTP would have to be patched/modified to deal with it. The more things you think of, the more clearly it is that the 'big' gui browsers use it as a hack to get to a blank startup page. You know there are such things as "defacto" standards as well as "dejure" standards. It is my contention that many browser authors have found it useful to include "about:blank" primarily as a way of saying "no start page." If it's not a standard, how come everybody's doing it? Links is doing that. If you start it without any arguments, you get a blank page. Thus the 'defacto' standard of a blank start page is preserved without an explicit tag. There's also a concept of 'interpretation'. :) The Romans couldn't think of the concept of zero, and thus doing long division in Roman Numerals sucked. :-) They not only couldn't concieve of zero, the very concept was discouraged. There's a cute book called the History of Zero (or something very close to that), you should read it, its fun. Did you know that in SQL the truth value of NULL = NULL is NULL. And the group by statement groups together things whose equality truth value is TRUE. Except it groups together NULLs :-) People have trouble of thinking of things that aren't there. :) Fair enough. However, the maintainer never said no. They said to show them a standard. So write an rfc. The thing to remmeber is that any changes the debian maintainer makes to links will need to be preserved from that point forward. Therefore it is in their best interrest to only make the minimum changes needed to make moving changes upstream into debain less painful. Perhaps they feel that supporting about:blank will not make it into upstream and will thus be a burden to support moving forward. Remember, debian supports multiple architectures and branches. Now, if there were a standard the likelyhood of about:blank being supported upstream would increase. So would the likelyhood of a patch to add such behavior being accepted. So really, you could take 2 aproaches. Write a standard, or ask upstream. My original question of 'what value does it provide' still stands. I can't think of a use that only about:blank can solve. In fact, I think the use of about:blank in the case you mention causes more problems than it solves. -- Mental ([EMAIL PROTECTED]) "The Torah... The Gospels... The Koran... Each claimed as the infallible word of GOD. Misquoted, misinterpreted, misunderstood, and misapplied. Maybe that's why he doesn't do any more interviews." - sinfest.com CARPE NOCTEM, QUAM MINIMUM CREDULA POSTERO. GPG public key: http://www.neverlight.com/pas/Mental.asc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Browsers that *don't* support about:blank
Tom wrote: I filed a wishlist bug against "links" asking for it to support about:blank (highly useful in frames pages as a default for the "body" frame). Thats an abuse of side effects and its usefulness is debatable. A highly useful default page in a frameset is one with relevant CONTENT. If you really want a blank page, create a blank page. Maintainer closed it as a nonstandard feature, but asked me I could point to a standard. Do you know of any significant graphical browsers that don't support "about:blank" by returning a blank page? I know they all handle other "about:xxx" commands differently. I think what he meant was as standard as in some sort of accepted and somewhat followed document like say... an RFC or w3c standard. In other words some sort of guarentee that is less subject to interpretation than my-browser-does-it-this-way. In addition to browsers, how are spiders that make an attempt at doing frames supposed to support about:blank? In other words, what does it offer to the world at large as a standard that can be relied on and is different from a blank page? If its just a way to avoid 3 html tags, is it really even worth creating a patch and supporting? Further, if you did use about:blank for a page and at some point MS decides to make about:blank an msn page, opera decides to sell advertising space on about:blank and konqeror points to kde.org/news, who is right? Beides. When I start links with no arguments, I get a blank page. Whats the problem? You should file a bug against whatever site uses about:blank as content. -- Mental ([EMAIL PROTECTED]) "The Torah... The Gospels... The Koran... Each claimed as the infallible word of GOD. Misquoted, misinterpreted, misunderstood, and misapplied. Maybe that's why he doesn't do any more interviews." - sinfest.com CARPE NOCTEM, QUAM MINIMUM CREDULA POSTERO. GPG public key: http://www.neverlight.com/pas/Mental.asc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: no sound in galeon/mozilla flash
Nori Heikkinen wrote: short version: i can't get sound with flash in mozilla or galeon. longer version: i've installed flashplugin-nonfree and libflash0. i can see flash movies in my two browsers (haven't tested it on anything else), but i can't hear anything. sometimes xmms is running when this happens; sometimes not. sound in general works. can anyone help me out here? all i want at work is a little homestarrunner! thanks a lot, Some sound cards (sblive and... i forget) support multiple opens in hardware. Meaning, more than one program can open and use /dev/dsp at once. Other cards do not support this, so if one program is using /dev/dsp, nobody else can. For desktop environments, this is largely worked around by using a dsp wrapper (esound for gnome, artsd for kde) and using that daemon to get to sound instead. xmms has plugins for both, or it can use alsa/oss natively. The downside is that this adds a little latency to the sound, but personally I've never noticed it. Then again, at home my primary sound card supports multiple opens so not everything is using artsd. You didn't mention what the underlying sound system is, but it doesnt matter much as I do belive flash only supports oss. If you're using alsa, load the oss compatibility modules for playing pcm data and using the mixer. If none of this helps, do an 'fuser -v /dev/dsp' and see what processes are using it. Kill them and retry using flash (you might have to restart the brower completely as I'm not sure that flash ever retries the open on the soundcard). -- Mental ([EMAIL PROTECTED]) "The Torah... The Gospels... The Koran... Each claimed as the infallible word of GOD. Misquoted, misinterpreted, misunderstood, and misapplied. Maybe that's why he doesn't do any more interviews." - sinfest.com CARPE NOCTEM, QUAM MINIMUM CREDULA POSTERO. GPG public key: http://www.neverlight.com/pas/Mental.asc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Time is runnign too fast
Nori Heikkinen wrote: oh my god, i can't decide if that link is hilarious or what ... "Barbie Wizards guide girls through the process of partitioning their disks, formatting volumes, mounting Samba shares, and installing packages. This kind of attention to detail and thorough understanding of female limitations also shows in the step by step Barbie Wizards that guide girls through the process of partitioning their disks, formatting volumes, mounting Samba shares, and installing packages. During the installation, girls are allowed to play a fashion-plate game or view a slideshow of rainbows, kittens, and Mattel products." !! is this a joke?? Its something, certainly not serious. It being a joke is subject to your sense of humor. -- Mental -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: classic deficiancy in both windows and linux ?
Jean-Marc V. Liotier wrote: On Fri, 2003-10-24 at 17:11, Mental Patient wrote: I found myself using imagemagic often to manipulate photos taken with a digital camera. I use nautilus/gnome as my desktop environment. After a while it got annoying to have to keep dropping into a shell to rotate, scale or montage the picture(s). So I wrote a couple pygtk scripts and put them in the scripts dir for nautilus. So now I can select a bunch of pictures, right click and send them to the wrapper. Up pops a gtk2 interface that I can use to set options like rotational direction, or filetype to output as a montage. It fits in with the rest of the desktop and I dont have to keep opening/closing terminals. Very interesting. Are your scripts distributed somewhere ? Nah, not yet. The pygtk stuff is still new to me. It was an excuse to learn some python. My TODO list includes cleaning up the code and releasing it. I still have to write the 'scale image' script. Its just some ugly duct tape and bubble gum code at the core of it. The one thing perl has on python is that there's an imagemagic module for perl. My python scripts are doing lots of os.system() calls :) However, I do believe there's a plugin site for this sort of stuff and lots of other people have done this. But if ya want a tarball, I could put a cvs snapshot somewhere. Its nothing too terribly exciting. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: classic deficiancy in both windows and linux ?
Daniel B. wrote: When you're looking at a directory using a graphical interface, a lot: - You have to open or at least switch to a different window. - You have to type or copy and paste the directory name. - You won't get the same data displayed by the graphical interface (which you'd expected it to print if it supported printing). Not that I dont see your point, but why should a graphical interface emit text? If you want to print a graphical representation of the filesystem, print a screen shot. If you want a text file, use the text inteface. I think that to some degree you're confusing 'want' with 'need'. I'm sure different people have different ideas on what you're talking about is supposed to do, or would want slightly different functionality, etc. If or when one wants to use a command line, that's fine, but when one it using a graphic interface, it should still support printing directory listings. I can relate to this. I found myself using imagemagic often to manipulate photos taken with a digital camera. I use nautilus/gnome as my desktop environment. After a while it got annoying to have to keep dropping into a shell to rotate, scale or montage the picture(s). So I wrote a couple pygtk scripts and put them in the scripts dir for nautilus. So now I can select a bunch of pictures, right click and send them to the wrapper. Up pops a gtk2 interface that I can use to set options like rotational direction, or filetype to output as a montage. It fits in with the rest of the desktop and I dont have to keep opening/closing terminals. They're guified shell scripts really. Might be one way to scratch your itch. They serve a specialized need and aren't flexible. But if its a common task you're doing, thats one way to solve the problem. Not sure, but I do believe kde has a similar functionality. -- Mental -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: [OT] CVS diff: hard vs. soft tabs
Nori Heikkinen wrote: right, that would have been nice ... but, as you say, i just inherited this one. not much i could have done about setting conventions first. For things like indenting, etc, you could always adjust what you have your tabstop set to. what i have my tabstop set to doesn't matter -- that's how my editor interprets hard tabs on disk. what i have is _no_ hard tabs on disk, and i want to put them there. that's more complex, right? Indeed it is. What you're essentially saying is along the lines of "I've made sweeping format changes to the code before working on it, and now I only want to commit the actual code changes. How do I do that?", right? I'm not sure how well this would work, but you could try something like this: Checkout the original, unaltered file you started with. Do a diff -uNrw originalfile newfile > changes.diff patch -p0 < changes.diff This should (I've only done some basic, rudimentary tests) only change the orignial file in the way that you want. In other words, the original file should be formated in its original state, but your new code should now be in it. Please make backups/check the output. Keep in mind that you may want to go back and reformat your bits of code so that the styles match, but this should get you most of the way there. My general rule of thumb is that unless I'm taking over the code for an extended period, I simply preserve the style it is written in. Its easier to deal with that way, especially if you know you're either not going to work on it again, or infrequently. -- Mental ([EMAIL PROTECTED]) "The Torah... The Gospels... The Koran... Each claimed as the infallible word of GOD. Misquoted, misinterpreted, misunderstood, and misapplied. Maybe that's why he doesn't do any more interviews." - sinfest.net CARPE NOCTEM, QUAM MINIMUM CREDULA POSTERO. GPG public key: http://www.neverlight.com/pas/Mental.asc pgp0.pgp Description: PGP signature
Re: [OT] CVS diff: hard vs. soft tabs
Nori Heikkinen wrote: hey all, this is kind of off-topic, but i figured this is the community most likley to have dealt with this sort of thing in the past, and be opinionated about it. i've been editing a lot of code over the past few months that was originally saved to disk with hard tabs for indenting. i can't work with hard tabs, and so managed to reformat the entire thing to use spaces (basically a "s,^I, ," iirc) before i began my massive overhaul of this file. now it's time to check it into CVS. i don't want every single line to show up as different just because of tab characters, so i need to find a good solution on how to transform my indents back into tab characters. clearly the reverse -- "s, ,^I," -- won't just work, as there are places where two spaces exist that i wouldn't want a tab. is there some way to open the file in emacs (in which i assumer it was originally written; i use vim) and run it through a re-indentder with hard tabs on? or could i do this in vim? suggestions & opinions welcome. thanks a lot, I've done this with mixed results. In general if you're going to work on projects, its a good idea to come up with your format conventions first. :) However, sometimes you just inherit code and really there isnt much you can do about it. Its right up there with cuddled elses, some people do it, others dont. CVS is going to have huge deltas due to indent formatting. If I'm going to change the format of the code, I will usually commit ONLY format changes with a comment to that effect, then commit code changes. It makes it easier to see what was actually changed that way. When you do huge reformats, you're going to wind up in merge hell if you have lots of developers or multiple branches of the code. It'll be difficult to tell what actually changed, vs what was merely moved around. For things like indenting, etc, you could always adjust what you have your tabstop set to. -- Mental ([EMAIL PROTECTED]) "The Torah... The Gospels... The Koran... Each claimed as the infallible word of GOD. Misquoted, misinterpreted, misunderstood, and misapplied. Maybe that's why he doesn't do any more interviews." - sinfest.net CARPE NOCTEM, QUAM MINIMUM CREDULA POSTERO. GPG public key: http://www.neverlight.com/pas/Mental.asc pgp0.pgp Description: PGP signature
Re: simply mutt (Re: Mutt with evolution)
duck wrote: Alternatively you could bind the f12 key to fetchmail as I have: macro index "!fetchmail\n" fetchmailconf is working fine in sid (used it yesterday). For running fetchmail periodically I am using gnome's inbox monitor applet (execute fetchmail before each update). I prefer not to use the daemon method (/etc/fetchmailrc) as this method allows running fetchmail manually without becoming root. crontab is good too (crontab -u). You dont need to be root to run fetchmail as a daemon. Use the -d parameter when you start fetchmail and supply a time interval. It'll run in the background as you and poll mail every N seconds. When I used fetchmail, I would have it start in the background when I logged in. -- Mental ([EMAIL PROTECTED]) "The Torah... The Gospels... The Koran... Each claimed as the infallible word of GOD. Misquoted, misinterpreted, misunderstood, and misapplied. Maybe that's why he doesn't do any more interviews." - sinfest.net CARPE NOCTEM, QUAM MINIMUM CREDULA POSTERO. GPG public key: http://www.neverlight.com/pas/Mental.asc pgp0.pgp Description: PGP signature
Re: Debian-based distros ??
Ralph F. De Witt wrote: On Monday 06 October 2003 08:49 pm, Mental Patient wrote: I also removed the apt-pinning that was going on. Mental: I do not understand what you mean by apt-pinning. The term is unfamilliar to me. How do I remove it? Its in the apt howto. For the sake of brevity, here's a url: http://www.debian.org/doc/manuals/apt-howto/ch-apt-get.en.html#s-default-version Also see section 3.10, it explains in detail what pinning is and how it works. Short answer: /etc/apt/preferences -- Mental ([EMAIL PROTECTED]) "The Torah... The Gospels... The Koran... Each claimed as the infallible word of GOD. Misquoted, misinterpreted, misunderstood, and misapplied. Maybe that's why he doesn't do any more interviews." - sinfest.net CARPE NOCTEM, QUAM MINIMUM CREDULA POSTERO. GPG public key: http://www.neverlight.com/pas/Mental.asc pgp0.pgp Description: PGP signature
Re: Debian-based distros ??
Ralph F. De Witt wrote: Mental: Please forgive me jumping in here, I have just finished a Knoppix hard disk install, and need guideance in how to convert to unstable. The last time I tried I ended up with several version of hotplug and several essential files went missing, It was a horrible mess. But I need to move to unstable as there are several apps I need to run and they are only avalible there. So I was wondering if you could help me convert to unstable? Or point me to some good documentation/how-to's. Being brand-new to debian is not helping. I would have installed woody, but I was unable to get x working. Thanks for any help you can offer. Ok, in no particular order here's the essentials of what I did. Keep in mind I've been using debian for.. a while. Several years, and I'm still learning. I also tend to brute force things, so I'm sure there are more elegant solutions. I'm by no means an expert an package management yet. I abuse having broadband and have no issues with purging all of kde/gnome/X11 and reinstalling if it fixes a weird dependency problem. The first thing I did was backup my sources.list and my list of installed packages. Just make a backup copy of the sources.list. To store the list of installed packages on your system, just do: dpkg --get-selections > selections.txt Next, I pruned down the actual sources.list to only the unstable entries as well as an entry for a couple things I wanted fetched from unofficial apt repositories. I also removed the apt-pinning that was going on. Next I did an apt-get update followed by an apt-get dist-upgrade. Read whats on hold. See what is going to be removed. Proceeding WILL break things in the sort term, so read this whole email before giving this a try. :) If, during the dist-upgrade dpkg bails out saying package X conflicts with package Y, I'll apt-get remove both (one could be testing, the other unstable). Same with dependency problems. The list of removed software in my case was minimal. After I can run apt-get dist-upgrade and I'm as up to date as possible, I'll tackle the missing packges. This usually takes several itterations of apt-get dist-upgrade ; apt-get -f install. Read the error messages, they pretty much tell you what to type. During the upgrade some pieces probably broke. Some app I use all the time was purged (by me) because it was getting in the way. This is where the selections.txt comes in. I do: dpkg --get-selections < selections.txt apt-get dselect-upgrade That _should_ install all packages that were installed when you saved off the list in the first place. The only thing that wont be replaced are packages that aren't findable in your sources.list. I think libranet had one or 2 admin gui things I never used, so I didn't mind loosing them. Just to be clear, I have a very destructive method. The upside is that apt keeps your system config files around, and your personal config files are safely in $HOME, so your settings for say.. mozilla will remain even if you apt-get remove it. The downside is that if I were on a dialup, I'd have to rethink things. So far I've done this twice. Besides the usual package FOO is temporarily uninstallable (sid gets like that sometimes) for a few days, no inherent problems with it. Dont bother runnig apt-get clean til you're done (or your /var size is too small for apt to cache all the debs), you'll probably need to uninstall/reinstall a couple apps several times. All in all its not difficult so much as time consuming. Before I bother attempting this I usually wait for a sunday afternoon with crappy weather and no plans. Things take a while. Excuse the incoherency, I'm trying to work and write this at the same time. :) -- Mental ([EMAIL PROTECTED]) "The Torah... The Gospels... The Koran... Each claimed as the infallible word of GOD. Misquoted, misinterpreted, misunderstood, and misapplied. Maybe that's why he doesn't do any more interviews." - sinfest.net CARPE NOCTEM, QUAM MINIMUM CREDULA POSTERO. GPG public key: http://www.neverlight.com/pas/Mental.asc pgp0.pgp Description: PGP signature
Re: Debian-based distros ??
stan wrote: A general wor of wisdom here, for what it's worth. I once made the mistake of allowing myself to be sucked in by the siren song of these offshot distors. As a result I am stuck with 2 machines (one at home, and one at work), bith of which I still actively use, but can't, in any reasonable way, update. Both machines were, at one time, lot's nicer than the Debian mainstream at that tiem. In mycase it was Progey, who sure looked like they would be around :-( So why can't you update? I ask because I've installed via libranet and knoppix (as well as the usual debian installer), and I haven't had this problem. Sure, knoppix jumped the gun on packageing X 4.3 and I had a couple minor problems with it, but it was easily fixed. Further the dist-upgrade to sid was a little bumpy, but thats to be expected anyhow. Are you saying that you cant upgrade to stable? Just wondering what the problem was. I've known people running testing who've backed out to stable, so there must be a way you can 'get back on track' so to speak. Then again, your definition of reasonable may differ from mine. -- Mental ([EMAIL PROTECTED]) "The Torah... The Gospels... The Koran... Each claimed as the infallible word of GOD. Misquoted, misinterpreted, misunderstood, and misapplied. Maybe that's why he doesn't do any more interviews." - sinfest.net CARPE NOCTEM, QUAM MINIMUM CREDULA POSTERO. GPG public key: http://www.neverlight.com/pas/Mental.asc pgp0.pgp Description: PGP signature
Re: /etc/group changes don't take effect immediately
Alphonse Ogulla wrote: Hi all, In an effort to run ppp as non root, I had to include my normal user id in the group 'dip' by directly editing /etc/group using vi. However, on saving and exiting /etc/group, I still could *not* access files owned by user root and available to users in group dip whilst using my normal uid. Just as I was about to give up, I decided to exit the shell and log in once again and to my amazement, I now could read files in /etc/ppp/peers directory whose permission is as follows:- [EMAIL PROTECTED]:/etc/ppp$ ls -ld peers drwxr-s--- 2 root dip 4096 Oct2 12:06 peers How does one explain this? Is the file /etc/group cached somewhere and updated only so often? How can the changes be forced to take effect immediately without exiting and logging in again? Simply put, thats how unix works. On login, your groups are pulled from wherever nsswitch.conf points. Not everyone uses files (I use ldap), so caching can be desired. If you dont want to logout and back in, you can always start a new shell with the newgrp command. See man newgrp and the nscd docs (about caching) HTH -- Mental ([EMAIL PROTECTED]) "The Torah... The Gospels... The Koran... Each claimed as the infallible word of GOD. Misquoted, misinterpreted, misunderstood, and misapplied. Maybe that's why he doesn't do any more interviews." - sinfest.net CARPE NOCTEM, QUAM MINIMUM CREDULA POSTERO. GPG public key: http://www.neverlight.com/pas/Mental.asc pgp0.pgp Description: PGP signature
Re: cannot find libncurses
Reaz Baksh wrote: Hello I don't know if this has been asked before, as I searched the list and found nothing close to this problem but I'm hoping this is the right forum to ask. Problem: I'm trying to upgrade my Kernel from 4.2.20 to 4.2.22. I am doing this the long way not through Dpkg or apt-get. When I type 'make menuconfig' I get the following: /usr/bin/ld: cannot find -lncurses collect2: ld returned 1 exit status Install the -dev package for curses. libcurses5-dev Or use make xconfig or make config -- Mental ([EMAIL PROTECTED]) "The Torah... The Gospels... The Koran... Each claimed as the infallible word of GOD. Misquoted, misinterpreted, misunderstood, and misapplied. Maybe that's why he doesn't do any more interviews." - sinfest.net CARPE NOCTEM, QUAM MINIMUM CREDULA POSTERO. GPG public key: http://www.neverlight.com/pas/Mental.asc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: debian Sid & Alsa & kernel 2.6.0-test6
Seppe Van Sande wrote: Hi all,Well, I'm trying to setup ALSA with my Debian Sid with kernel2.6.0-test6 for my Audigy (emu10k1), but I get nothing but errors :-( I downloaded the driver, libs and utils package, but the driver failswhen I do a make. First, I did: You just need libs and utils. Use the alsa drivers in the kernel. I did, and my audigy works just fine with the alsa drivers. Remember to unmute the volume when you insmod the drivers. -- Mental ([EMAIL PROTECTED]) "The Torah... The Gospels... The Koran... Each claimed as the infallible word of GOD. Misquoted, misinterpreted, misunderstood, and misapplied. Maybe that's why he doesn't do any more interviews." - sinfest.net CARPE NOCTEM, QUAM MINIMUM CREDULA POSTERO. GPG public key: http://www.neverlight.com/pas/Mental.asc pgp0.pgp Description: PGP signature
Re: X is slow on 2.6-test6 kernel
Nathan Weston wrote: I just installed 2.6.0-test6, and X is incredibly slow -- if I minimize a window which covers most of the screen, I can watch the redraw creep down the screen, taking about 1 second to finish. I'm using a PCI GeForce2 MX with the 4496 driver from NVidia. Under 2.4.21, the same driver produces good results. I haven't been able to test other cards or drivers. Nathan You wont need to run X with a negative nice value if you have a preemptable kernel. Doing so will have a negative impact on performance. I had the same problem until I changed my /etc/X11/Xwrapper.config. Just comment out or remove the nice_level line. -- Mental ([EMAIL PROTECTED]) "The Torah... The Gospels... The Koran... Each claimed as the infallible word of GOD. Misquoted, misinterpreted, misunderstood, and misapplied. Maybe that's why he doesn't do any more interviews." - sinfest.net CARPE NOCTEM, QUAM MINIMUM CREDULA POSTERO. GPG public key: http://www.neverlight.com/pas/Mental.asc pgp0.pgp Description: PGP signature
Re: radeon opengl help
Mark Roach wrote: Hi folks, I am trying to understand how to get opengl working on my compaq evo n1000v laptop with radeon M7 LW (7500) and it seems pretty confusing. Can someone break it down for me a little bit? Here's what my current understanding is: Essentially you have 2 choices. 1. Go with all free/oss drivers (dri). 2. Go with the commercial ati linux drivers (some proprietary code). I've opted for the commercial driver for performance reasons. See http://www2.ati.com/drivers/firegl/readme0325.txt ( I dont see M7 listed, so you may only have one option) If commercial drivers are undesirable, see http://dri.sf.net/ and check the documentation. -- Mental ([EMAIL PROTECTED]) "The Torah... The Gospels... The Koran... Each claimed as the infallible word of GOD. Misquoted, misinterpreted, misunderstood, and misapplied. Maybe that's why he doesn't do any more interviews." - sinfest.net CARPE NOCTEM, QUAM MINIMUM CREDULA POSTERO. GPG public key: http://www.neverlight.com/pas/Mental.asc pgp0.pgp Description: PGP signature
Re: quake 2 mouse problem
Sebastian Kapfer wrote: On Fri, 19 Sep 2003 12:50:12 +0200, Andrew Ingram wrote: Hi List, I just found out about "apt-get install quake2" :) Everything is fine apart from my mouse. It works only horizontally, Enable the mouse look feature. I don't know if there is a menu setting for that. If not, try to enter +mlook at the game console. but far too quickly No idea I think there's a setting called something like in_dgamouse. Try toggling it from whatever its set to. -- Mental ([EMAIL PROTECTED]) "The Torah... The Gospels... The Koran... Each claimed as the infallible word of GOD. Misquoted, misinterpreted, misunderstood, and misapplied. Maybe that's why he doesn't do any more interviews." - sinfest.net CARPE NOCTEM, QUAM MINIMUM CREDULA POSTERO. GPG public key: http://www.neverlight.com/pas/Mental.asc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: ATI Radeon 9600G
Sebastian Canagaratna wrote: Hi: I am hoping to buy Gateway 700CX computer which comes with ATI Radeon 9600G card. I list archives don't seem to have anything about it; they do talk about the Radeon 9000. Has anyone any experience with it with Debian? I will probably be installing unstable on it. From what I have read the ATI Radeon cards do not seem to be well supported. Any suggestions will be gratefully appreciated I just upgraded my old nvidia card to an ATI Radeon 9800. It works great. ATI has binary drivers on their website for their radeon cards. There is also an OSS dri project for them that works, but based on what I've seen do not perform as well. ATI cards are very well supported on linux. I too thought they weren't. The binary drivers come with utilities to configure X for you, acceleration is fast, common games like quake3 run great, etc. The only thing to be mindful of is that unlike nvidia, the ATI drivers REQUIRE that your kernel recognize you agp chipset. Even if the ati drivers override agpgart, agpgart must still recognize your chipset. This caused much agravation for me. I'm using a kt400 chipset on the motherboard that isn't well supported in 2.4. I wound up installing 2.6.0 and installed a port of ati's kernel module. There is a backport for 2.4 that helps fix kt400 support. I just went with 2.6 as I wanted to see how it was shaping up anyways. That all said, the ATI card works well for me. 2.6 kernel driver port of ATI's latest driver: http://zinx.xmms.org/misc/tmp/fglrx-module-3.2.5-for-2.6.0-test.tar.gz Here's how a friend of mine setup the open source ATI drivers. http://lists.netisland.net/archives/plug/plug-2003-07/msg00462.html hope that helps. I've found that the ATI drivers work well. The nvidia drivers would cause lockups in X once in a while. So far things seem more stable. -- Mental ([EMAIL PROTECTED]) "The Torah... The Gospels... The Koran... Each claimed as the infallible word of GOD. Misquoted, misinterpreted, misunderstood, and misapplied. Maybe that's why he doesn't do any more interviews." - sinfest.net CARPE NOCTEM, QUAM MINIMUM CREDULA POSTERO. GPG public key: http://www.neverlight.com/pas/Mental.asc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: ATI drivers 3.2.5, kernel 2.6-test4 and modules
Nick Lidakis wrote: I have been trying my hardest to get ATI's 3.2.5 driver (the 3.2.5 are supposed to be kernel 2.6 compatible) module to compile with kernel 2.6-test4 on my Debian unstable box. I have read the kernel doc's concerning modules, but I can't seem to figure out what I am doing wrong. I don't have any trouble with the 2.4 series. The error I keep getting when I run ./make.sh is: marvin:/lib/modules/fglrx/build_mod# ./make.sh ATI module generator V 2.0 == kernel includes at /lib/modules/2.6.0-test4/build/include not found or incomplete file: /lib/modules/2.6.0-test4/build/include/linux/version.h This happens if make a symlink to linux in /usr/source or edit the make.sh to point to /usr/src/kernel-source-2.6-test4 Any suggestions? The link to the complete ATI installation tips is http://www2.ati.com/drivers/firegl/readme0325.txt Lots of stuff changed from 2.4 to 2.6 and the ati drivers arent' totally prepared for it. Zinx on #ati (freenode) has a port that I'm using with my ATI 9800 + 2.6.0-test5. You can get it here: http://zinx.xmms.org/misc/tmp/fglrx-module-3.2.5-for-2.6.0-test.tar.gz Nothing changes as far as the setup goes, just use the tarball to build the kernel module and put it somewhere in /lib/modules/`uname -r`/ Hope that helps. -- Mental ([EMAIL PROTECTED]) "The Torah... The Gospels... The Koran... Each claimed as the infallible word of GOD. Misquoted, misinterpreted, misunderstood, and misapplied. Maybe that's why he doesn't do any more interviews." - sinfest.net CARPE NOCTEM, QUAM MINIMUM CREDULA POSTERO. GPG public key: http://www.neverlight.com/pas/Mental.asc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: directfb w ATI Radeon
[EMAIL PROTECTED] wrote: On Mon, Sep 15, 2003 at 08:39:01AM -0400, Mental Patient wrote: Turns out it was a kernel config error. Recompiled with frame buffer console enabled, it goes. Is the framebuffer accelerated with a Radeon [9700]? I got it to work but it was very slow compared to not using it. This would make things that spew lots of text (like compiles) take much longer, so even though it was beautiful I preferred not to use it. Did I misconfigure it? Not sure, I'm running 2.6.0-test5. I haven't really had time to make any kind of judgement. I got the console to come up in frame buffer mode, logged in and thats about it. I'm using a similar card as yours. The console didn't seem overy slow, but plain text mode is probably still faster. As far as directfb goes, I saw on the web site that initial radeon acceleration support is working, however I still have a good bit of reading to do before I get gtk+ apps,ect working on the console. -- Mental ([EMAIL PROTECTED]) "The Torah... The Gospels... The Koran... Each claimed as the infallible word of GOD. Misquoted, misinterpreted, misunderstood, and misapplied. Maybe that's why he doesn't do any more interviews." - sinfest.net CARPE NOCTEM, QUAM MINIMUM CREDULA POSTERO. GPG public key: http://www.neverlight.com/pas/Mental.asc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
sata support in installer
Hello, I'm researching sata support. Thinking about buying a new system and going all SATA. The only docs I've seen on SATA support are in new (2.4.21 and 2.6.0) kernels. Additionally, the links I've seen seem to indicate that people are installing debian on an EIDE drive, then upgrading the kernel, then using their SATA devices (see http://www.e-aiyama.com/~toshi/Computer/Linux/SATA.html). Is the simplest solution for this to use a live cd like http://www.e-aiyama.com/~toshi/Computer/Linux/LiveCD.html? Are there debian installers that support SATA? Anyone have any advice/recomendations on brands of motherboards (asus prefered, but not required)? -- Mental ([EMAIL PROTECTED]) "The Torah... The Gospels... The Koran... Each claimed as the infallible word of GOD. Misquoted, misinterpreted, misunderstood, and misapplied. Maybe that's why he doesn't do any more interviews." - sinfest.net CARPE NOCTEM, QUAM MINIMUM CREDULA POSTERO. GPG public key: http://www.neverlight.com/pas/Mental.asc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: directfb w ATI Radeon
Arnt Karlsen wrote: On Sun, 14 Sep 2003 00:31:55 -0400, Mental Patient <[EMAIL PROTECTED]> wrote in message <[EMAIL PROTECTED]>: Anyone using this setup? I recently switched my Geforce3 for a Radeon. I have glx/X11 stuff working fine, but was wondering about getting directfb working. I've tried passing vga=791 at the grub prompt, as well ..at the grub prompt, hit "c" then tab twice. Try 'vbe-detect(?)' to see what video modes it can do. Turns out it was a kernel config error. Recompiled with frame buffer console enabled, it goes. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
directfb w ATI Radeon
Anyone using this setup? I recently switched my Geforce3 for a Radeon. I have glx/X11 stuff working fine, but was wondering about getting directfb working. I've tried passing vga=791 at the grub prompt, as well as video=radeonfb, but the console stays black until X starts (gdm). Anyone have this stuff working? -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: problems with Via AC97 controller
Richard Shepherd wrote: I am running 3.0r1+g2.2 Woody on a DFI ad77 with KT400 chipset. On startup, I get a sound device not detected error. 'sndconfig' detects a Via Technologies VT8233 AC97 Audio controller But when I try the test I hear no sound and get: modprobe provokes the following error /lib/modules/2.4.20/kernel.drivers/sound/via82cxxx_audio.o: init_module: No such device /lib/modules/2.4.20/kernel.drivers/sound/via82cxxx_audio.o: insmod: /lib/modules/2.4.20/kernel.drivers/sound/via82cxxx_audio.o: failed /lib/modules/2.4.20/kernel.drivers/sound/via82cxxx_audio.o: insmod sound-slot-0 failed Can you help? The via82cxx_audio module is failing to intialize. This could be because the chipset on your mobo is a newer version than what the driver you have supports. I have this audio chipset on my mobo at work. I had problems with older kernels. At one point (2.4.20 or 2.4.21) the modules would load but I'd STILL get no sound. I'm currently running 2.4.22 and its working. Modules loaded are: via82cxxx_audio20988 0 ac97_codec 13428 0 [via82cxxx_audio] soundcore 3844 2 [via82cxxx_audio] chipset info: 00:11.0 ISA bridge: VIA Technologies, Inc. VT8235 ISA Bridge 00:11.1 IDE interface: VIA Technologies, Inc. VT82C586/B/686A/B PIPC Bus Master IDE (rev 06) 00:11.5 Multimedia audio controller: VIA Technologies, Inc. VT8233 AC97 Audio Controller (rev 50) I'd suggest seeing what modules the stock debian kernels support and if possible upgrading your kernel via apt-get or compiling your own. -- Mental ([EMAIL PROTECTED]) "The Torah... The Gospels... The Koran... Each claimed as the infallible word of GOD. Misquoted, misinterpreted, misunderstood, and misapplied. Maybe that's why he doesn't do any more interviews." - sinfest.net CARPE NOCTEM, QUAM MINIMUM CREDULA POSTERO. GPG public key: http://www.neverlight.com/pas/Mental.asc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Mousey broken
[EMAIL PROTECTED] wrote: Only one question - why the hell does my mouse work through /dev/input/mice on 2.6.0 ??? Because perhaps the kernel has static support for USB? Different modules compiled into kernel instead? Another thing - I installed gpm last night and ran gpmconfig - it found nada on /dev/psaux. Added to that - no matter which device file I use: If the mouse driver initialized properly my optical mouse's light would've come on. It does on 2.6.0 not on 2.4.22. gpm is not a mouse driver. Until you get the mouse working, gpm is useless. BTW: USB mice work fine with Linux 2.4. I know and that's what pees me off so much. On RedHat I had no problems whatsoever. Redhat is different :) Arguably their hardware detection is much better. Fortunately you can use kudzu if you want. You could also apt-get install discover and see if that loads up some missing modules. I really think i'm missing something very stupid here, but i can almost guarantee that I've got the right kernel config. I've modprobed mousedev like Bob suggested and at first it wasn't there , cause i compiled the module into the kernel at first. If you checked my dmesg files you would've seen that the module was in fact there and did start up. Anyway's I ripped it out of the kernel and compiled as a module to no avail. Same problem... Forget about the USB part - as long as i can get the mouse to work through the PS/2 port - i'll worry about USB later. The stock debian kernel supports ps2/usb mice if you installed the bf2.4 kernel. I think you're flailing about needlessly with recompiling the kernel. The trick with usb is the order in which the modules are loaded. The usb controler needs to be initialized, then I think its input, hid, followed by mousedev. If I were you, I'd drop back to the default 2.4 kernel if you have it installed, then give discover or kudzu a shot. Baring that, check out linux-usb.org. I use usb mice at home, they worked fine in stable. If you want better hardware detection at install time (and I know many people who do), check out one of the debian derivatives that has better hw detection. It saves tons of agravation later. I'll really appreciate it if anyone can have a look at these to kernel configuration files. The one is for 2.6.0 on which mousey works the other for 2.4.22 on which it doesn't. I can't seem to find any differences in the mouse config parts... off topic question: Does sid come with xfree86 4.3 ? Thanx a bunch guys 4.3 isnt in sid yet, I dont think. I'm running an install based on libranet, so I already have 4.3. -- Mental ([EMAIL PROTECTED]) "The Torah... The Gospels... The Koran... Each claimed as the infallible word of GOD. Misquoted, misinterpreted, misunderstood, and misapplied. Maybe that's why he doesn't do any more interviews." - sinfest.net CARPE NOCTEM, QUAM MINIMUM CREDULA POSTERO. GPG public key: http://www.neverlight.com/pas/Mental.asc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Voodoo Graphics still supported?
Russell Shaw wrote: markus koller wrote: Hi, I've got an old Voodoo1 (Diamond Monster 3D) and I'd like to use it with X4.2. It worked without problems with X3.3 and Mesa, but I can't figure out how to get it to work again. I googled around a bit, and got only more confused, so I thought maybe I'd find help here. XFree86 -configure will generate a config file with available options. Actually, no it wont. It'll generate a config file for the 2d card. The voodoo1/2 cards were 3d only passthrough pci cards completely unrelated to drawing the desktop. In fact IIRC, there was no IRQ or interrupt or anything other than a memory address that showed up indicating the card's presence. Its been AGES since I used one, but when I did I think I used it on console with glide/mesa compiled for svgalib. I'm not even sure the voodoo1/2 were ever properly supported in X. Then again at the time the only real thing to run on them was quake/quake2 which (at the time) were written using svgalib for mouse/keyboard input. The only thing I can really say is look for an older howto. You may very well need to do some digging. If I had to hazard a guess I'd imagine that maybe when X was upgraded, mesa also was upgraded and is no longer compiled with glide suppport? -- Mental ([EMAIL PROTECTED]) "The Torah... The Gospels... The Koran... Each claimed as the infallible word of GOD. Misquoted, misinterpreted, misunderstood, and misapplied. Maybe that's why he doesn't do any more interviews." - sinfest.net CARPE NOCTEM, QUAM MINIMUM CREDULA POSTERO. GPG public key: http://www.neverlight.com/pas/Mental.asc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: libc.so.6
Philip Clark wrote: Hi there, I have managed to delete /lib/libc.so.6 and since the whole system depends on it eg. I can't even create a new link or new file. Also booting from a boot disk doesn't work. Does anyone have any ideas? I am presuming a rescue disk is the way to go. How do I make one? Last time I did that I wound up booting off a knoppix cd, mounting the root partition and unpacking a deb of libc by hand. You can unpack a deb file with ar. Just run "ar x ". You'll wind up with 3 files. control.tar.gz (control/install scripts) data.tar.gz (what you want to unpack in your root partition to get libc back) and debian-binary (text file). After rebooting I did an apt-get --reinstall install libc6 just to make sure everything was back the way it was supposed to be. It worked for me, good luck. -- Mental ([EMAIL PROTECTED]) "The Torah... The Gospels... The Koran... Each claimed as the infallible word of GOD. Misquoted, misinterpreted, misunderstood, and misapplied. Maybe that's why he doesn't do any more interviews." - sinfest.net CARPE NOCTEM, QUAM MINIMUM CREDULA POSTERO. GPG public key: http://www.neverlight.com/pas/Mental.asc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Sound Blaster Audigy and ati radeon 9800 not seen by kernel 2.4.18
Antonio [EMAIL PROTECTED] wrote: I have a sound blaster audigy card and an ati radeon 9800, none of them seem to be visible to the kernel 2.4.18 (woody source) that I compiled. According to what I have seen, emu10k1 should work with this card, but insmod is not working for the reason that the card is invisible. Any ideas what could be the reason? I am attaching some info hoping for some help. Thanks. Use the latest cvs code for the emu10k chips. You can get it here http://sourceforge.net/projects/emu10k1/ Works fin with my Audigy and I believe it works with audigy2 cards as well. As for ATI, you should check their website. I dont believe any cards newer than the 9000 work in linux (as of now). -- Mental ([EMAIL PROTECTED]) "The Torah... The Gospels... The Koran... Each claimed as the infallible word of GOD. Misquoted, misinterpreted, misunderstood, and misapplied. Maybe that's why he doesn't do any more interviews." - sinfest.net CARPE NOCTEM, QUAM MINIMUM CREDULA POSTERO. GPG public key: http://www.neverlight.com/pas/Mental.asc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: libc6 optimizations
Manolis Tzanidakis wrote: [20030822] Mental Patient <[EMAIL PROTECTED]> wrote: Earlier today I was toying with the idea of recompiling libc6 with a bunch of optimization flags. I did an apt-get source libc6 and added the flags I wanted. I built binaries via "fakeroot ./debian/rules binary" and did a dpkg -i on the resulting deb files. Well, the install hung and failed. ls would segfault, ssh would still work. Things were... broken. Don't even think to compile glibc with O3 or Os. It breaks baddly ;) Debian is quite optimized and runs really fast although it's compiled for i386. If you want to speed things up more try compiling kde, xfree, gnome, mozilla etc. I noticed. :) I was using -O2, but like I said in a previous post... I believe the root of my trouble to be other CFLAGS. Thanks for the feedback tho. This was more of an educational thing than anything else. I've used optimized packages of all kinds of stuff. Honestly I dont find debians performance to be an issue in the least. It was just boredom on the rampage :) -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: libc6 optimizations
Diego Calleja García wrote: El Fri, 22 Aug 2003 16:52:12 -0400 Mental Patient <[EMAIL PROTECTED]> escribió: Its not really critical or anything, I was just wondering about it. Try apt-src/apt-build No, I know how to build the debs. The problem is I belive the CFLAGS I used. The kernel uses -fomit-frame-pointer. I do believe that was my big mistake. Like I said, everything compiled and the debs were built. Its just at install it broke. :) -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
libc6 optimizations
Earlier today I was toying with the idea of recompiling libc6 with a bunch of optimization flags. I did an apt-get source libc6 and added the flags I wanted. I built binaries via "fakeroot ./debian/rules binary" and did a dpkg -i on the resulting deb files. Well, the install hung and failed. ls would segfault, ssh would still work. Things were... broken. Not really a problem as I had a 'good' libc6 deb, so I just booted off a knoppix cd and unpacked the deb in my root partition. 5 minutes later I was back up and running. I did an apt-get --reinstall install libc6 just to make sure all the needed post-install scripts ran. It worked as expected. Back up, no data loss, things are fine. Anyways... has anyone done this sort of crazy thing? I was thinking that in addition to the kernel, libc6 was a good candidate for a recompile in terms of overall performance boosts. I know about gentoo, but I dont feel like changing distros. I jsut want to put libc6 on hold a while and play around... see if I can even notice the difference before I make any decisions like changing distros. Were my problems with the optimized deb files due the the optimization flags themselves (I used the same CFLAGS that the kernel does when you compile it optimized for athlons). Also, I used gcc 3.3.2. Should I have used a different compiler? I dont want to have to recompile all the packages, just this one. Its not really critical or anything, I was just wondering about it. -- Mental ([EMAIL PROTECTED]) "The Torah... The Gospels... The Koran... Each claimed as the infallible word of GOD. Misquoted, misinterpreted, misunderstood, and misapplied. Maybe that's why he doesn't do any more interviews." - sinfest.net CARPE NOCTEM, QUAM MINIMUM CREDULA POSTERO. GPG public key: http://www.neverlight.com/pas/Mental.asc pgp0.pgp Description: PGP signature
Re: Nvidia driver install
Tim wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 ralph bacolod wrote: | Good day! I have a problem installing the nvidia | drivers which i downloaded. Im running a Debian Woody | box on a Athlon 2100 ,Asus A7N266 vm mobo,geforce2 | mx400. Im dual booting it with WinXP. I downloaded the | driver from another machine with windows coz im not Downloading with M$ often leads to incomplete files-you should try to obtain a complete download using a more reliable OS. experienced this myself when trying to get the SiS630 driver. No it doesnt. Worst case if you down load text files you have to convert line endings. Binary downloads are binary. ;) I can see if you didn't realize that *nix and dos have an incompatible opinion on what the end of a line in a text file looks like, but its not that the file is incomplete. Its that when you download in ascii mode the line end conversion happens automagically (most times). -- Mental ([EMAIL PROTECTED]) "The Torah... The Gospels... The Koran... Each claimed as the infallible word of GOD. Misquoted, misinterpreted, misunderstood, and misapplied. Maybe that's why he doesn't do any more interviews." - sinfest.net CARPE NOCTEM, QUAM MINIMUM CREDULA POSTERO. GPG public key: http://www.neverlight.com/pas/Mental.asc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: telnet localhost slow, telnet 127.0.0.1 ok
RUPERT LEVENE wrote: Indeed. Netcat and galeon appear to do the right thing too, while telnet, lynx and links all suffer from a delay. (I first noticed the problem using lynx). My guess at the moment is that the combination of a slow machine and a slow nameserver is exposing an odd bug in these programs. Although I would be surprised if this were the case since telnet and lynx have such a long pedigree... And you're sure these aren't socks clients? -- Mental ([EMAIL PROTECTED]) "The Torah... The Gospels... The Koran... Each claimed as the infallible word of GOD. Misquoted, misinterpreted, misunderstood, and misapplied. Maybe that's why he doesn't do any more interviews." - sinfest.net CARPE NOCTEM, QUAM MINIMUM CREDULA POSTERO. GPG public key: http://www.neverlight.com/pas/Mental.asc pgp0.pgp Description: PGP signature
Re: X11 RENDER support
On Thu, 2003-07-31 at 17:39, Chris Metzler wrote: > Note the second-to-last entry, and the fact that it's "built-in". > My guess is that whoever built this X server (maybe Tal and Jon, > the Libranet folks; or maybe they got it from somewhere else) > made Xrender built-in to the server, rather than a loadable > extension . . .sorta analogous to compiling a driver into the > kernel vs. making it available as a loadable kernel module. In > such a case, the experience that the gdk library complains about > RENDER regardless of whether 'load "Xrender"' is commented out > or not would make sense. > > Yes, xdpyinfo agrees with this. And I don't care that its loaded. I kinda need it for drawing some weird stuff. I just want to fix the weird glib warning being spewed to stderr everytime I start gaim or something. Its more of an irritation than a problem. Especially since rox-session has an overwhelming need to display these messages in the root window every single time they're emitted... :/ > > Know of an easy way to do this? can i apt-get install --reinstall > > some_meta_package? > > I know of no easy way to do it. Purging (after backing up your > XF86Config-4) and re-installing the components of x-window-system > would do it, I would expect. Yeah, I've pretty much come to the same conclusion. I'll budget some time for a 'quick' dpkg --get-selections > selections.txt ; apt-get remove xlibs ; dpkg --set-selections < selections.txt ; apt-get dselect-upgrade at some point over the weekend when I'm feeling masochistic I'll probably get around to it. Thanks tho. -- Mental ([EMAIL PROTECTED]) CARPE NOCTEM, QUAM MINIMUM CREDULA POSTERO. GPG public key: http://www.neverlight.com/pas/Mental.asc You have mail in /home/mental/Maildir signature.asc Description: This is a digitally signed message part
Re: X11 RENDER support
On Thu, 2003-07-31 at 16:25, Chris Metzler wrote: > > The X version is 4.3 > X 4.3 is not in sid. Did you compile it yourself? Did you install it > from one of the unofficial sources at apt-get.com? If the former, > you may wanna talk to the xfree86 folks; if the latter, then them plus > whoever put the package together. Since it's not even in sid yet, > xfree86 4.3 is not "officially" supported by Debian, so far as I know > (someone please correct me if wrong). Right you are, I forgot this box started life out as a libranet box. > So you're saying that if you comment out Xrender, thus turning off > loading of the X Rendering extension, and *restart X*, you still > get those messages? What do your XF86Config-4 and /var/log/XFree86.0.log > (including the messages from X server startup) look like? No, only that glib complains that I have a buggy render extension regardless of what I do. However, going back and looking shows that the extension gets loaded regardless of whats in the config file. I've attached XF86Config-4, and the X startup log. I'll look into downgrading X to whats in sid at some point. They're not errors, just warnings. I'll wander over to the libranet users list and see if anyone else has seen this. Know of an easy way to do this? can i apt-get install --reinstall some_meta_package? -- Mental ([EMAIL PROTECTED]) CARPE NOCTEM, QUAM MINIMUM CREDULA POSTERO. GPG public key: http://www.neverlight.com/pas/Mental.asc You have mail in /home/mental/Maildir # File generated by XFdrake. # ** # Refer to the XF86Config man page for details about the format of # this file. # ** Section "Files" #FontPath "unix/:7100" FontPath"/usr/lib/X11/fonts/misc" FontPath"/usr/lib/X11/fonts/cyrillic" FontPath"/usr/lib/X11/fonts/75dpi" FontPath"/usr/lib/X11/fonts/Type1" FontPath"/usr/lib/X11/fonts/TrueType" FontPath"/usr/lib/X11/fonts/Speedo" FontPath"/usr/share/fonts/truetype" FontPath"/usr/share/fonts/truetype/openoffice" FontPath"/var/lib/defoma/x-ttcidfont-conf.d/dirs/TrueType" FontPath"/var/lib/defoma/x-ttcidfont-conf.d/dirs/CID" EndSection Section "ServerFlags" #DontZap # disable (server abort) #DontZoom # disable / (resolution switching) AllowMouseOpenFail # allows the server to start up even if the mouse doesn't work EndSection Section "Module" Load "v4l" # Video for Linux Load"ddc" Load"GLcore" Load"dbe" Load"dri" Load"extmod" Load"glx" Load"record" Load"bitmap" Load"freetype" Load"speedo" Load"type1" Load"vbe" Load"int10" # new (experimental) Load "Xrender" Load "pex5" Load "xie" EndSection Section "InputDevice" Identifier "Keyboard1" Driver "Keyboard" Option "XkbModel" "pc105" Option "XkbLayout" "us" Option "XkbOptions" "" EndSection Section "InputDevice" Identifier "Mouse1" Driver "mouse" Option "Protocol" "IMPS/2" Option "Device" "/dev/psaux" Option "ZAxisMapping" "4 5" EndSection Section "Monitor" Identifier "monitor1" VendorName "Plug'n Play" #HorizSync 30-97 #VertRefresh 50-180 # Sony Vaio C1(X,XS,VE,VN)? # 1024x480 @ 85.6 Hz, 48 kHz hsync ModeLine "1024x480"65.00 1024 1032 1176 1344 480 488 494 563 -hsync -vsync # TV fullscreen mode or DVD fullscreen output. # 768x576 @ 79 Hz, 50 kHz hsync ModeLine "768x576" 50.00 768 832 846 1000 576 590 595 630 # 768x576 @ 100 Hz, 61.6 kHz hsync ModeLine "768x576" 63.07 768 800 960 1024 576 578 590 616 EndSection Section "Monitor" Identifier "monitor2" # Sony Vaio C1(X,XS,VE,VN)? # 1024x480 @ 85.6 Hz, 48 kHz hsync ModeLine "1024x480"65.00 1024 1032 1176 1344 480 488 494 563 -hsync -vsync # TV fullscreen mode or DVD fullscreen output. # 768x576 @ 79 Hz, 50 kHz hsync ModeLine "768x576" 50.00 768 832 846 1000 576 590 595 630 # 768x576 @ 100 Hz, 61.6 kHz hsync ModeLine "768x576" 63.07 768 800 960 1024 576 578 590 616 EndSection Section "Device" Identifier "device1" VendorName "Matrox" BoardName "Matrox Millennium G450 DualHead" Driver "mga" Screen 0 BusID "PCI:1:0:0" Option "DPMS" Option "AGPMode" "1" EndSection Section "Device" Identifier "device2" VendorName "Matrox" BoardName "Matrox Millennium G450 DualHead" Driver "mga" Screen 1 BusID "PCI:1:0:0" Option "DPMS" Option "AGPMode" "1" EndSection Section "Screen" Ident
X11 RENDER support
Hello. I'm running a current version of sid. My gtk applications all (for the most part) emit the same warnings: Gdk-WARNING **: The X server advertises that RENDER support is present, but fails to supply the necessary pixmap support. In other words, it is buggy. I've tried commenting/uncommenting the Load Xrender line out of XF86Config-4 with not changes. The X version is 4.3 running in dual head xinerama mode, vid card is a Matrox MGA G400. Has anyone else had this problem? -- Mental ([EMAIL PROTECTED]) CARPE NOCTEM, QUAM MINIMUM CREDULA POSTERO. GPG public key: http://www.neverlight.com/pas/Mental.asc You have mail in /home/mental/Maildir signature.asc Description: This is a digitally signed message part
Re: Apache Problems (please help)!
On Sun, 2003-07-27 at 17:54, Francisco Castellon wrote: SNIP > So I couldn't start Apache through the Webmin module (the samething also > happens when I reboot the server and it tries to start the apache > service, it will hang in that part of the boot up process until I enter > in the passphrase then it will continue booting), so then I went to the > command line and again I got the part that says "Enter pass phrase:" and > I did, and apache started. However, when I try to access apache through > https://localhost or https://host-ip-address or https://host.domain.com > it doesn't work it just gives me the page cannot be displayed error. It > only works if I access the server through regular "http://"; and not > through "https://"; so my two big questions are: > > 1. How can I make it so that it doesn't ask me the pass phrase every > time I start apache? And if I do disable the pass phrase feature does > that mean that SSL is not started? There's a faq for this: http://www.modssl.org/docs/2.8/ssl_faq.html#ToC31 > > 2. Why are the references to any https:// address to the web server work > even after I start apache in the command line and enter in the pass > phrase as I am prompted to do so? How come only the http:// references > work? > Not sure I follow this one. Only thing I can think of off hand is that the module hasn't been loaded before your configuration parameters are parsed. Make sure any AddModule/LoadModule directives in httpd.conf come before your IfModule directives. Other than that, tail error logs, see if apache is listening on the correct port, etc. -- Mental ([EMAIL PROTECTED]) CARPE NOCTEM, QUAM MINIMUM CREDULA POSTERO. GPG public key: http://www.neverlight.com/pas/Mental.asc signature.asc Description: This is a digitally signed message part
uw-imapd bug or feature?
I am using debian/testing on a mail server. The plan was to install uw-imapd-ssl and use that. I have a problem with this and outlook 2000. Something with the setup is causing o2k to shut down almost immediately. Since then, I've recompiled imapd from the original source with SSL disabled and I'm tunneling it through sslwrap. So far, evolution, outlook express, outlook and mozilla work just fine. I believe this is related to the -DMICROSOFT_BRAIN_DAMAGE flag that the debian package is compiled with. Has anyone else had this problem? -- Mental ([EMAIL PROTECTED]) I bought a house, on a one-way dead-end road. I don't know how I got there. --Steven Wright GPG public key: http://www.neverlight.com/Mental.asc -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
