Re: [LUAU] How long should it take to put a system together?
Scott E Foulk wrote: I suppose I missed this part of the thread, but how does one pick a motherboard? I have a LTSP server down, and I want to replace the MB. I see all kinds of options and haven't a clue. This server needs minimally 4 Gb of ram. A decent bus speed, which CPU, etc. Its for Linux, obviously. Is there a difference in MB's that I would choose between a Linux standalone and a Windows WS? Sorry if this is too late to be relevant (I don't check this list much). If you buy the wrong motherboard, you're going to have to buy a new one, or spend a long time making it work. I bought an Asus A7V, and ever since, I do a ton of research before buying a motherboard. The A7V had ridiculously common hard-lock issues, as described at http://episteme.arstechnica.com/eve/forums/a/tpc/f/77909585/m/9880993003/inc/1 . So features are nice, but they aren't my main concern. If I were to buy a new motherboard, I would look at reviews. I wouldn't buy anything with less than 5 stars on newegg. I would even go so far as to go to their support forum and read about problems people have with the motherboard. If all the questions are How do I overclock it more, or When I do this, and this, and that, and stand on my head, this one function is 10% slower than I'd expect; then it's probably a good motherboard. If the questions are about avoiding hardlocks and how to get it to work with all of the RAM slots, or anything about needing a beta version of the BIOS, run. Anyway, as far as features go, I would look for future-proof type of technologies. Make sure it has firewire and USB. Make sure that if you're going to get a quality raid card that you have enough of that type of slot (PCI express 16, maybe?). Usually you can get 2 Gigabit ports on the motherboard, but if you're buying more ethernet cards or even 1GBit/s cards or fibre channel for storage, then you'll need whatever port those use. The bus speed isn't as important as allowing DDR3 RAM and a modern processor. DDR3 isn't important in and of itself, but it's probably the future of RAM. If you need at least 4GB RAM, then you obviously need a 64-bit processor, since memory-mapped I/O will max 32-bit processors at probably 3.0-3.5GB or so. ASUS now has a complete Linux-based diagnostic boot environment entirely contained on a PROM on most of their server motherboards. I've never played with it, but it sounds neat. I think ultimately the processor socket(s) is the limiting factor in future-proofing. Nowadays it's hard to upgrade one component of a machine. They go through RAM standards and processor sockets almost every year now. Intel and AMD both went a little bonkers with their number of product lines, so you'll probably need to find some good benchmarks (is a X3 Phenom better than a high-end opteron? How do core2quads compare to xeon 7 series?). -Eric Hattemer ___ LUAU@lists.freesoftwarehawaii.org mailing list http://lists.freesoftwarehawaii.org/listinfo.cgi/luau-freesoftwarehawaii.org
Re: [LUAU] Wine 1.0 can be amazing
Wine development has grown by leaps and bounds, but unfortunately, that's been from utter garbage to a sometimes passable API. You'll almost surely need to execute some or all of the suggestions on http://appdb.winehq.org/objectManager.php?sClass=versioniId=9194 for Guild Wars, including reading up on all the bugs. Aside from Guild Wars, you may benefit from some of the modifications from http://wiki.winehq.org/winetricks . I've never really gotten anything to work without importing a dozen or so .dll files from Windows XP and turning on native mode for them. There used to be problems with the installer that came with wine, which caused it to skip certain registry keys. So I had to take a Windows install of the program, export its registry keys, then import them into my wine environment. Wine has gotten a lot better, but until they add features like working anti-aliasing, back-buffering, and predictable sound output, my games are running on Windows. I used to run EVE on Wine at 0.5fps, whereas on Windows, it was close to 300fps. I played with all sorts of options and dll files, read every forum entry and every workaround, and never found an answer besides Wine fundamentally doesn't support back-buffering, which is used extensively by EVE. So I gave up. Maybe Wine will finally implement all of DirectX 9 by the time all games require DirectX 10... -Eric Hattemer Jason Axelson wrote: I haven't tried it on World of Warcraft but I've been having trouble getting to run Guild Wars. Which is odd because Guild Wars is on the Platinum list so it should work without any configuration. On Fri, Jul 4, 2008 at 11:29 AM, Peter Besenbruch [EMAIL PROTECTED] wrote: I recently experimented with Wine 1.0 under Ubuntu 8.04. I was utterly amazed at running World of Warcraft perfectly and at a stunning speed under an older P4 with a mediocre AGP card. Everything works. Even the sound is flawless. There are quite a few other apps listed in the Wine DB. It's really come a long way. I'm still waiting for it to run Word Perfect 10, or Microsoft Publisher 98. It tends to work on graphics programs, and rather imperfectly on a Windows 3 era dictionary. I still run the dictionary, because it's good, and I put up with the display imperfections. -- Hawaiian Astronomical Society: http://www.hawastsoc.org HAS Deepsky Atlas: http://www.hawastsoc.org/deepsky ___ LUAU@lists.freesoftwarehawaii.org mailing list http://lists.freesoftwarehawaii.org/listinfo.cgi/luau-freesoftwarehawaii.org ___ LUAU@lists.freesoftwarehawaii.org mailing list http://lists.freesoftwarehawaii.org/listinfo.cgi/luau-freesoftwarehawaii.org ___ LUAU@lists.freesoftwarehawaii.org mailing list http://lists.freesoftwarehawaii.org/listinfo.cgi/luau-freesoftwarehawaii.org
Re: [LUAU] non-distro firewall recommendations?
Jeff Mings wrote: Is there a good up-to-date firewall solution like MonMotha's? It is _really_ easy to configure, and make work. The only thing missing is the ability to deny packets from certain hosts. The MonMotha script was supposed to do this, but the office girls were still able to suck away the bandwidth and their productivity at myspace.com, forever21.com, etc. Shorewall unequivocally blocks the crap sites. I also need to accomodate Gizmo and the SIP / RTP functionality of the Talkswitch mini-PBXes. The servers are primarily Fedora Core (8,7,6) boxes that primarily serve OpenVPN, Samba, HTTP and DHCP. Suggestions? I don't know if you can hook it into dynamic lists, but http://www.fwbuilder.org/ is my favorite. It's a GUI program where you create objects like loopback interface, my subnet, ICMP, UDP Port 1, and you drag them around and make a layout of your rules. Once you're done, you can export a firewall script for a dozen different operating systems. -Eric Hattemer ___ LUAU@lists.hosef.org mailing list http://lists.hosef.org/listinfo.cgi/luau-hosef.org
Re: [LUAU] high performance SCP/SSH
So I know this is pretty off-topic and all, but has anyone ever found an scp client for Windows that doesn't perform like ass? I can pick any two random linux/solaris/osx machines, and get 20-30MB/s transfers. But between windows and linux/solaris/osx, the most I've ever gotten was 3MB/s. I've tried cygwin ssh, winscp, pscp, coreftp, and probably others, never with any decent speed. Regular FTP on the same machine is 30MB/s+, FTP over SSL was about the same, I think, but scp and suddenly the speed goes down by a factor of 10. Using various -c and -C options hasn't helped. Anyone know what it is that caps this speed? -Eric Hattemer ___ LUAU@lists.hosef.org mailing list http://lists.hosef.org/cgi-bin/mailman/listinfo/luau
Re: [LUAU] help on port 25
Clifton Royston wrote: On Fri, Nov 09, 2007 at 12:27:12AM -0800, goku ball z wrote: hay guys. I have a funny problem and I just can't figure out. I just installed open suse 10.2 here is my problem on the suse box to test to see if port 25 is working i did the following telnet localhost 25 and it worked BUT. when I did the following telnet domain.com 25 I got connection refused. I check the services, dns and everything I can think of can someone point me in the right direction? A few more notes. Try `iptables -L` and see if there are any iptables rules that might specify port 25. I still recommend a program called fwbuilder if you're not familiar with iptables. Or if you want to be defenseless, on FC you can do `service iptables stop`. I'm sure there's something similar on suse. Depending on how your routing works, it is possible that you are going out to your ISP before you go back to your own IP address. Many intelligent home ISP's block and filter port 25 on their network to help oppress the massive spam bots. Unless you have a commercial server-type agreement, it is unlikely that you will be able to have port 25 open, deliver mail, and not have your outgoing mail marked as spam. Often to use another ISP's smtp server, you need to configure your mail client to use port 465 or 587. Many domains block all mail from dynamic or home user IP addresses. You might try configuring your MTA to bind only to port 587 and only accept authenticated smtp if you want to use it as a local SMTP server. If you actually want to collect mail on that box, but your ISP doesn't want you to, you're pretty much screwed. But Clifton is probably right that if you didn't do enough configuring on your MTA, you probably just missed the part where you open it up to the public. If you're using sendmail, don't. The only reason to use sendmail is to put it on your resume for those companies that are stuck with it. The configuration belongs back in the early 70's. Even then, they probably could have written a better configuration file, but didn't. Postfix is excellent and easy to setup. I haven't looked into qmail, but anything is better than sendmail. -Eric Hattemer ___ LUAU@lists.hosef.org mailing list http://lists.hosef.org/cgi-bin/mailman/listinfo/luau
Re: [LUAU] all your GPU are belong to us
Dave Burns wrote: You're assuming that they can't get in and read /etc/shadow. If they can, then either you've got a broken configuration and they will own you in 5 minutes, or they have root already and ordinary user-level passwords aren't really stopping them from doing much. I suppose this situation deserves some contemplation, but I'd prefer to spend a lot more effort preventing them from getting to that point in the first place. I agree. I think my post was a little vague, but the idea is that there used to be vulnerabilities in Windows where you could use a null session to download the password hash anonymously. I suppose it's possible that you could find a network vulnerability for any OS that lets you read files but no execute arbitrary code. This would make cracking a password hash like /etc/shadow worthwhile. But I think this situation is pretty uncommon. -Eric Hattemer ___ LUAU@lists.hosef.org mailing list http://lists.hosef.org/cgi-bin/mailman/listinfo/luau
Re: [LUAU] all your GPU are belong to us
Vince Hoang wrote: On 10/25/07, Jim Thompson [EMAIL PROTECTED] wrote: If passwords weren't dead already, this (or having the botnet do it on the CPUs) finished them. In a world where bank PINs are 4 numeric digits can you suggest practical alternatives? Biometrics are not mature enough. Two-factor authentication has existed for a long time but is not cost effective for the average consumer. The article talks about ntlm and pgp. The answer is not passwords that are more complicated, it is passwords that can't be anonymously downloaded and cracked offsite. It doesn't matter how crappy your shadow password is if someone can only try an ssh attempt every 2 seconds or so. NTLM passwords are freely available to any decent cracker with a network connection to the windows machine. If your PGP secrets are important, and you expect someone to get at them, you'd better have a ridiculously large key. -Eric Hattemer ___ LUAU@lists.hosef.org mailing list http://lists.hosef.org/cgi-bin/mailman/listinfo/luau
Re: [LUAU] headless fedora
Peter Besenbruch wrote: Dave Burns wrote: Maybe I should go in the opposite order - I can log in over the serial connection when it is still booted runlevel 5 and has a monitor attached, can't I? You can, but you can also use a remote desktop connection. Yeah, the thing to keep in mind is that having a monitor and a graphical desktop doesn't hinder you from doing anything. I'll give you a chance to configure it and make sure everything is working before disconnecting it. You can test ssh, look into vnc, and make sure it all works from off your local network before disconnecting anything. The other thing to think about is that as long as you can guarantee something like sshd will be running, a second ethernet card with no firewall rules might be easier than a serial console. If you have multiple machines that will be on most of the time, I suggest you setup netconsole so that you can get kernel crash reports, and maybe something like nagios so that you can get emailed if something goes down. -Eric Hattemer ___ LUAU@lists.hosef.org mailing list http://lists.hosef.org/cgi-bin/mailman/listinfo/luau
Re: [LUAU] Re: [buug] nano help!!!!
Vince Hoang wrote: The solution is the same in both: fix the PATH. For tcsh, the OP will want to add something like the following to your .tcshrc file: set PATH = ($PATH /usr/local/bin) Using a symlink will work too, but using a custom PATH will continue to work for other local package installations. In my experience with tcsh, $PATH is set by setenv, and is : separated, whereas $path is handled by set, and is space separated. So you'd want setenv PATH $PATH:/usr/local/bin or set path=($path /usr/local/bin) My point that a lot of people missed was someone wrote in saying basically I would like to use nano. I went out of my way to install it, because I am familiar with it and enjoy it. Sean wrote back saying basically, I recognize that you've probably seen vi and hate it enough to drive you to go out of your way to install nano, but you're wrong, and you need to use vi. Do you ever wonder why vi was written in 1976, but in 1984, 1999, and 2000 respectively, people had the audacity to publicly release emacs, nano, and gedit, when vi is clearly better? The answer is that not everybody likes vi. Some people will never like vi. Some people have used it extensively, either when it was necessary, or to give it a chance, are familiar with it, proficient, and still hate it. If you grabbed 100 people off the street, gave them a copy of nano and a copy of vi, and told them go, I think you'd see my point. The point is not which editor is the best. The point is that if someone says I'm having trouble with A, the answer of A sucks, use B is usually inappropriate. I have yet to find this hypothetical system that has vi, but doesn't have emacs, noxemacs, nano, pico, nedit, gedit, kwrite, jed, or joe; the ability to install one of these easier to use editors, or a file transfer utility that would allow you to do your editing remotely. Maybe I'm fortunate that I work for a company whose servers are all from the last 2 decades. -Eric Hattemer ___ LUAU@lists.hosef.org mailing list http://lists.hosef.org/cgi-bin/mailman/listinfo/luau
Re: [LUAU] Re: [buug] nano help!!!!
[EMAIL PROTECTED] 808.356.2913 wrote: this might not sound nice but use 'vi' not 'nano' you will never feel any editor woes again Sean http://www.kokuatraffic.com You're right, it doesn't sound nice. It sounds elitist and stupid. -Eric Hattemer ___ LUAU@lists.hosef.org mailing list http://lists.hosef.org/cgi-bin/mailman/listinfo/luau
Re: [LUAU] help with iptables
Whoever Whatever wrote: Hi, when I have a linux box with two NICs, one $EXTIF (external) and one $INTIF (subnet=192.168.176.0/20). With ip masq turn on: iptables -t nat -A POSTROUTING -o $EXTIF -j MASQUERADE how do I block $INTIF(192.168.176/20) from accessing ports(ie: 80) on the linux box(192.168.176.1), but still allow port 80 access using ip masq to the outside world? I tried the filters below, but I still can access port 80 after applying the rules: iptables -A INPUT -i $INTIF -p tcp -m tcp --dport 80 --syn -j DROP iptables -A FORWARD -i $INTIF -p tcp -m tcp --dport 80 --syn -j DROP iptables -A INPUT -p tcp -s 192.168.176.0/20 -d 192.168.176.1 --dport 80 -j DROP anyone know if it's possible? I don't know any iptables, because I've never had to. Maybe someone on the list will answer your question, but in the meantime, I strongly suggest a program called fwbuilder. You drag your rules around, and when you're done, you can auto-generate scripts for every major routing/firewall language you can think of. -Eric Hattemer ___ LUAU@lists.hosef.org mailing list http://lists.hosef.org/cgi-bin/mailman/listinfo/luau
Re: [LUAU] help with iptables
Eric Hattemer wrote: Whoever Whatever wrote: Hi, when I have a linux box with two NICs, one $EXTIF (external) and one $INTIF (subnet=192.168.176.0/20). With ip masq turn on: iptables -t nat -A POSTROUTING -o $EXTIF -j MASQUERADE how do I block $INTIF(192.168.176/20) from accessing ports(ie: 80) on the linux box(192.168.176.1), but still allow port 80 access using ip masq to the outside world? I tried the filters below, but I still can access port 80 after applying the rules: iptables -A INPUT -i $INTIF -p tcp -m tcp --dport 80 --syn -j DROP iptables -A FORWARD -i $INTIF -p tcp -m tcp --dport 80 --syn -j DROP iptables -A INPUT -p tcp -s 192.168.176.0/20 -d 192.168.176.1 --dport 80 -j DROP anyone know if it's possible? I don't know any iptables, because I've never had to. Maybe someone on the list will answer your question, but in the meantime, I strongly suggest a program called fwbuilder. You drag your rules around, and when you're done, you can auto-generate scripts for every major routing/firewall language you can think of. But if you do have a pretty good idea of what you're doing and want to debug it, Phil Dibowitz wrote a nice program called IPTState at http://www.phildev.net/iptstate/ that shows you what iptables is up to in realtime. It's of wide enough acclaim that you can probably find it in your local package management tool. -Eric Hattemer ___ LUAU@lists.hosef.org mailing list http://lists.hosef.org/cgi-bin/mailman/listinfo/luau
Re: [LUAU] Top 10 Best / Worst Cities For Software Developer Pay
Jim Thompson wrote: Where are the exports for Hawaii? I think this is the key question. You can't pay people with money you don't have. It's not so much about how much are they paying as who is there that can actually pay? I think the only way to solve this problem would be to get more companies with more cash-earning products to Hawaii. It's not direct, but there's definitely a small correlation between corporate earnings and employee salaries. How many giant (Microsoft, Adobe, IBM, Apple, etc.) software/computer companies have offices in Hawaii? The only one I can think of is IBM. -Eric Hattemer ___ LUAU@lists.hosef.org mailing list http://lists.hosef.org/cgi-bin/mailman/listinfo/luau
Re: [LUAU] MS to EOL Win98 and WinMe - July 11
patches, then make sure that your OS patches stay up to date. However, yes, it would be nice if you didn't need the hardware/software/SP2 firewall because there weren't any open ports to begin with. I have read that Vista will ship with no open ports, but considering how often they change fundamental things in between betas, who knows whether they'll botch something simple like that or not. I firmly believe that if you turn on any network server-like services, you take responsibility for understanding what they do and certify that you have the newest, safest version of the service. If you don't, then you deserve what you get. I've seen OSX machines become raging messed up hacked open mail relays because some fool ran through the Sharing Preferences and checked all the checkboxes without knowing what any of them do (turns on samba, ssh, apache, sendmail and several others with default settings). -Eric Hattemer
Re: [LUAU] MS to EOL Win98 and WinMe - July 11
Jim Thompson wrote: I told him to not connect it to the Internet, because it would be rooted in minutes. http://www.realtechnews.com/posts/1511 Its ugly out there... I haven't found a link to the original article or anything, but this sounds largely absurd. A statement like this requires many qualifications. First, Windows ME by some definitions can't be rooted, since it only has one user. Second, any worms that might do any sort of automatic rooting almost certainly do NOT apply to the dos-based versions of windows. From sarc.com about blaster worm: *Systems Affected http://securityresponse.symantec.com/avcenter/refa.html#systemsaffected:* Windows 2000, Windows NT, Windows Server 2003, Windows XP *Systems Not Affected http://securityresponse.symantec.com/avcenter/refa.html#systemsnotaffected:* Linux, Macintosh, OS/2, UNIX, Windows 95, Windows 98, Windows Me This is true of most if not all non-self-inflicted and non-browser-related worms. If you can find me a worm capable of automatically infecting a windows ME machine without user input, I will be extremely surprised. The best I've ever found is one that will propagate to shared drives on the network that have manually turned on read/write sharing on the root of the hard drive regardless of whether they set a password on the share. No sharing is enabled on windows ME by default. If you turn on sharing to the root of your hard drive with read/write without ever going to windowsupdate.microsoft.com , then you do deserve what you get. Otherwise, a default install of windows ME is relatively safe. This article is surely very specific to a certain class of computers that came out before 2004, computers running a a copy of Windows NT4 (any) Windows 2000 SP0-SP3 or Windows XP SP0-SP1 with no security patches slipstreamed onto the installation media, with no software nor hardware firewall. I don't think Windows 2000 or NT with the latest service pack and all recent security patches slipstreamed into the installation CD is common, although you used to be able to order a free CD from microsoft with all the security patches for the price of shipping at http://www.microsoft.com/security/protect/cd/order.asp . Windows NT/2000/XP all were vulnerable to several classes of network worms because they had retarded default security settings with open ports for running services normal people would never need. All of these ports are firewalled by default in XP SP2, and almost all computers that have been built since August 2004 have SP2 built into the installation/restoral CD. Until a worm comes out that hacks the firewall itself (hasn't happened yet, but isn't impossible), all of these XP SP2 machines are safe by default, and can only be hacked via self-inflicted security holes (running randomly obtained exe files, turning off the firewall, etc). Furthermore, a windows machine of any sort that is alone behind a hardware firewall won't have any hackable open ports either. It seems like almost all broadband ISPs give their customers NAT firewall routers/modems nowadays, so this mostly just leaves dialup and static IP users of windows NT-2000-XPSP1, who either haven't been to windowsupdate before the viruses came out, or who managed to find an installation CD for one of those old OSes, which is an increasingly small segment of windows users. So I contend that to say that all windows machines are hackable in 12 minutes is more absurd than to say that all linux machines are hackable in 12 minutes just because redhat 7 was ( http://xforce.iss.net/xforce/alerts/id/advise71 ). The only people who have it right are OSX and a few linux distributions. OSX has no open ports by default. Almost all Linux distributions have ssh enabled by default, which has had a few exploits. I strongly believe that ALL open ports should be an opt-in policy and not an opt-out/firewall policy. -Eric Hattemer
Re: [LUAU] Pics and news about the mythtv build
Hawaii Linux Institute wrote: It appears that there's a lot of interest in MythTV, which is spilling over to Linux. Does any one have any thoughts/comments re Hauppauge 250 vs 350? Thanx. Wayne You'd do really well to ask the mythtv-users list from http://www.mythtv.org/mailman/listinfo/mythtv-users/ . There are hundreds of them with lots of experience with those cards. -Eric Hattemer
Re: [LUAU] Myth TV Thanks!
Jim Thompson wrote: There was also talk of having a annual conference as a fundraser to highlight OSS in the community - j/k I find your humor here derogatory in tone and tenor, and sarcastic at best. Perhaps I misinterpret, and you will clarify. Interestingly the record shows that you were involved with TPOSSCON 05. Can it be true that you now deride same? You should come next build, you can ask the 'old-timers' - people over 30 about the good old days. Matt, you can choose to either help build, or help tear things apart. I'll be 44 in May, what is your point? Do you mean to insinuate that because Julian is under 30, that he can't hope to understand your politics? That life really was better before he was born, when 1MB of memory cost more than any of us made in a week? Because I'll back Julian in a Linux bar fight. I'll lay $500 that he can out-code anyone in your company (I get to pick the language and platform.) Jim _ Whoa, let's be cool, everyone. I don't think I read anything negative in Matt's comments. This was about history of the list, and I'm fairly sure that 'old-timers' are just people that can remember what the list was like back then. I'm 23, but I can remember back to 99/2000. I remember that this is exactly why we had two LUAUs. I wasn't there for it, but I remember hearing rumors that the original LUAU fell apart because everyone fought all the time. Warren kept the name, but made a point to say that it wouldn't be as ugly and mean-spirited as the end of the last group. It almost got that way anyway in the first part of this century, because a couple of people resented Warren taking a lot of leadership in the group (saying it wasn't democratic enough). Now I imagine Scott would see some liability in throwing the HOSEF name around randomly at meetings and events that don't involve him. You may be able to get the HOSEF organization to agree to and sponsor whatever events you are planning. I'm guessing that funds are chartered more for school projects, but you'd have to ask the HOSEF organizers. But since LUAU hasn't had a real meeting since Warren moved to New Hampshire, you can use that name and make meetings, and do what you like. LUAU used to have meetings at Mid-Pacific Institute with pizza and talks about kernels and virtual machines and things like that. It was at Mid-pac because Warren volunteered there. He made a linux computer lab there. Many of the MPLUG/LUAU (same group then, basically) got together to help build that lab. People were bringing in computers and stuff like that. As time went on, MPLUG/LUAU started working on other school labs. This is where HOSEF basically came from. So many of the main/active members were working on these things, that MPLUG/LUAU/HOSEF became somewhat synonymous. This doesn't mean that just because some people are working on linux labs and the newer HOSEF school projects, that everyone has to stop holding meetings to talk about kernel developments and building telecom servers. Is the space at McKinley an option? Is Scott willing to let people work on these things at his workshop area? I hope everyone can give the benefit of the doubt to what each other say. This message is not meant to be negative or anti-anyone. I just mean to try to keep the peace, and tell what I remember of LUAU. It may not be factually accurate, but please let me know that in a polite manner. -Eric Hattemer
Re: [LUAU] Re: Putting the UG back in LUG
Matt Darnell wrote: Speaking of 1080p..when should I show up for the Superbowl? Meh, I don't think it'd be worth your while to fly out to LA for it... That's where I live nowadays. -Eric Hattemer
Re: [LUAU] Re: mythtv install
Eric Jeschke wrote: I can definitely recommend using Ubuntu as the base. Just enable the universe and multiverse repositories and then synaptic is your friend-- just search for mythtv and it will install everything that you need. I think just use what you're familiar with. Almost all distros have a binary distribution of mythtv. Fedora has one from Axel Thimm http://atrpms.net/ , which has never been entered into the main tree. I've never tried it, but people seem to be happy with it. I just compiled from source, and if you follow the directions, this is a very smooth process. Oh, and streaming works great to a laptop running 802.11g wireless with just the front end installed; I watch/pause live tv or recorded shows anywhere on the property. What platform is the laptop? -Eric Hattemer
Re: [LUAU] Re: Putting the UG back in LUG
Matt Darnell wrote: On 2/2/06, Jim Thompson [EMAIL PROTECTED] wrote: Maddog wrote: Mythtv sounds like a winner too! I'm not in Hawaii anymore, but let me know if you get stuck with this. I have had a mythtv box setup for over a year now. It's not terribly hard. I'm using Fedora 4. It works great. I do have to worry about RedHat changing my udev configuration and stopping mythtv from finding the TV card, but I just have to remember to play with the udev settings after any major rpm upgrade. You may want to join the mythtv mailing list. It's pretty chatty, though. But join it for a bit, and people will have interesting suggestions (not as important now that you have your TV cards). I have a bttv card, which is somewhat low quality, doesn't have onboard processing, but does come automatically in the kernel. The ivtv based cards are significantly trickier to get working on the kernel level. I do suggest a card with an onboard mpeg2 processor, though. That will significantly lower your overhead and let you tape more things simultaneously while using the computer. The main page is at http://www.mythtv.org/ . Now that they've added plenty of functionality and worked out most of the bugs, I would stick to the point (non-svn) versions. The svn versions work fine too, but you feel like you have to keep upgrading. They are as step-by-step as possible here: http://www.mythtv.org/modules.php?name=MythInstall . I have literally never seen a linux project as refined and user-friendly as mythtv. Once it's compiled, you never really need to touch the command line again (except to start the frontend). All of the setup and customization are through the cute little menus. It's well documented, and remarkably easy, unless you have some non-standard hardware. I downloaded the dsmyth filters http://dsmyth.sourceforge.net/ , installed mythweb, and stream my shows to my windows machine. You can even do live TV over your favorite Windows media player (except Windows Media Player TM). I hear you need something just slightly faster/more reliable than 802.11b to stream them. I use gigabit ethernet, but I'm pretty sure 802.11g or 100 Mbit ethernet would do it great. I haven't played with any of the mythplugins, except for the mythweb. But it seems that they install pretty cleanly. -Eric Hattemer
Re: [LUAU] Re: Putting the UG back in LUG
just that at the same bitrate, large frame sizes give good stills, and small frame sizes give good motion. If I did [EMAIL PROTECTED]/s, I'd get insane macroblocking. But that's not relevant for talk shows. With any of the cards with the MPEG2 encoders builtin, you can tape to MPEG2 directly. This uses negligible cpu (I think someone said about 4%). Even at max quality, the shows don't take up an unreasonable amount of space. You can then run the transcoder in the background at minimum priority (in mythtv, you setup a recording profile that automatically transcodes certain shows you pick). I think the Haupauge 150/250/350 cards are the most common ones with builtin MPEG2 encoders. There are cards with 720p capabilities, but since I have SD cable, I've never looked into that. -Eric Hattemer
Re: [LUAU] new display card needed
Jim Thompson wrote: Also, are you using the drivers that come with Xorg, or the ones from ATi's website? I don't have much experience with ATi, but I would try this driver: https://support.ati.com/ics/support/default.asp?deptID=894task=knowledgefolderID=27 It should have faster acceleration, and probably has access to more features like non-standard video modes. The only question is stability. I haven't run ATi drivers, but I don't think stability is a problem. Here's a modeline I found that you can try # 1920x1200 @ 75Hz Modeline 1920x1200 246.590 1920 2064 2272 2624 1200 1201 1203 1253 +hsync +vsync Section Screen ... DefaultDepth 24 Subsection Display Depth 24 Modes 1920x1200 1024x768 800x600 640x480 ViewPort0 0 EndSubsection EndSection -Eric Hattemer
Re: [LUAU] new display card needed
I wouldn't give up on the current graphics card just yet. If you tell Xorg to try 1920, does it give an error? You might have to go old-school and use a modeline, but if ATi says it works, then it's Xorg that's being the problem. Maybe you could send or post your xorg.conf and Xorg.log. -Eric Hattemer Jim Thompson wrote: My company found itself with a bit of cash at the end of the year and decided to buy me a Dell 2405 display. This has a native resolution of 1920x1200. It works great on the Mac, but I'd like to also use it on the linux box. 'lspci -v' reports: :01:00.0 VGA compatible controller: ATI Technologies Inc Radeon R300 ND [Radeon 9700 Pro] (prog-if 00 [VGA]) Subsystem: ATI Technologies Inc: Unknown device 0002 Flags: bus master, stepping, 66MHz, medium devsel, latency 255, IRQ 16 Memory at e000 (32-bit, prefetchable) [size=128M] I/O ports at 9000 [size=256] Memory at f500 (32-bit, non-prefetchable) [size=64K] Capabilities: [58] AGP version 3.0 Capabilities: [50] Power Management version 2 :01:00.1 Display controller: ATI Technologies Inc Radeon R300 [Radeon 9700 Pro] (Secondary) Subsystem: ATI Technologies Inc: Unknown device 0003 Flags: stepping, 66MHz, medium devsel Memory at e800 (32-bit, prefetchable) [disabled] [size=128M] Memory at f501 (32-bit, non-prefetchable) [disabled] [size=64K] Capabilities: [50] Power Management version 2 Relavent parts of the Xorg.log file state: (II) fglrx(0): Connected Display1: DFP on internal TMDS (II) fglrx(0): Display1 EDID data --- (II) fglrx(0): Manufacturer: DEL Model: a010 Serial#: 810826067 (II) fglrx(0): Year: 2005 Week: 41 (II) fglrx(0): EDID Version: 1.3 (II) fglrx(0): Digital Display Input (II) fglrx(0): Max H-Image Size [cm]: horiz.: 52 vert.: 33 (II) fglrx(0): Gamma: 2.20 (II) fglrx(0): DPMS capabilities: StandBy Suspend Off; RGB/Color Display (II) fglrx(0): Default color space is primary color space (II) fglrx(0): First detailed timing is preferred mode (II) fglrx(0): redX: 0.640 redY: 0.330 greenX: 0.300 greenY: 0.607 (II) fglrx(0): blueX: 0.149 blueY: 0.060 whiteX: 0.312 whiteY: 0.328 (II) fglrx(0): Supported VESA Video Modes: (II) fglrx(0): [EMAIL PROTECTED] (II) fglrx(0): [EMAIL PROTECTED] (II) fglrx(0): [EMAIL PROTECTED] (II) fglrx(0): [EMAIL PROTECTED] (II) fglrx(0): [EMAIL PROTECTED] (II) fglrx(0): [EMAIL PROTECTED] (II) fglrx(0): [EMAIL PROTECTED] (II) fglrx(0): [EMAIL PROTECTED] (II) fglrx(0): Manufacturer's mask: 0 (II) fglrx(0): Supported Future Video Modes: (II) fglrx(0): #0: hsize: 1280 vsize 1024 refresh: 60 vid: 32897 (II) fglrx(0): #1: hsize: 1600 vsize 1200 refresh: 60 vid: 16553 (II) fglrx(0): #2: hsize: 1152 vsize 864 refresh: 75 vid: 20337 (II) fglrx(0): #3: hsize: 1680 vsize 1050 refresh: 60 vid: 179 (II) fglrx(0): Supported additional Video Mode: (II) fglrx(0): clock: 154.0 MHz Image Size: 519 x 324 mm (II) fglrx(0): h_active: 1920 h_sync: 1968 h_sync_end 2000 h_blank_end 2080 h_border: 0 (II) fglrx(0): v_active: 1200 v_sync: 1203 v_sync_end 1209 v_blanking: 1235 v_border: 0 (II) fglrx(0): Serial No: T61335A40T9S (II) fglrx(0): Monitor name: DELL 2405FPW (II) fglrx(0): Ranges: V min: 56 V max: 76 Hz, H min: 30 H max: 81 kHz, PixClock max 170 MHz (II) fglrx(0): End of Display1 EDID data http://www.ati.com/products/radeon9700/radeon9700pro/specs.html says that it should run 1920x1200 @ 75Hz. For some reason X.org doesn't see it that way. So I guess I'm looking for a new 'graphics' card. I don't care (much) about 3D, but I do like my text to draw fast. I prefer DVI ports to VGA ports. I run Ubuntu (but will soon reload with kbuntu) on this machine. (3GHz P4, 1GB ram, 2x200GB SATA drives, all in a Shuttle SFF case.) I've had this card in various machines since early 2001 (my days at Musenki). Anyone got recommendations? jim ___ LUAU@lists.hosef.org mailing list http://lists.hosef.org/cgi-bin/mailman/listinfo/luau
Re: [LUAU] new display card needed
Eric Hattemer wrote: I wouldn't give up on the current graphics card just yet. If you tell Xorg to try 1920, does it give an error? You might have to go old-school and use a modeline, but if ATi says it works, then it's Xorg that's being the problem. Maybe you could send or post your xorg.conf and Xorg.log. Also, are you using the drivers that come with Xorg, or the ones from ATi's website? -Eric Hattemer
Re: [LUAU] Re: UH Manoa
I was disappointed that the UH web site student login pages only provided help instructions to Windows IE/Netscape and Mac Safari users... Write to the webmaster to update the documentation. It's probably more appropriate to write the webmaster (or department) WITH the updated documentation. It's unlikely that anyone there is prejudiced against any particular platform or browser, but more likely that they haven't allocated the time and funding to produce documentation for every web platform on the planet. -Eric Hattemer
Re: [LUAU] How do you delete over 500,000 files in a directory
Jim Thompson wrote: Hawaii Linux Institute wrote: for i in m*; do rm m*; done still requires bash to glob all 500,000 files. Fails for the same reaons. Ok, I wonder if I'm the only one who immediately read this as what I thought it should have been, then didn't get the criticism. The typo should have read for i in m*; do rm $i; done; -Eric Hattemer
Re: [LUAU] on-topic, finally!
Jim Thompson wrote: The Land Cruiser (which will eventually be a hybrid) has a mini-itx computer mounted in it. Its the box on the right in this (warning: quite large) photo http://www.smallworks.com/~jim/LandCruiser/P8280034.JPG Two questions: First, do you mean you are buying a new hybrid to replace this car, or you're actually hybridizing the car yourself? Second, what is the computer currently used for in the car? -Eric Hattemer
Re: [LUAU] goofy RAM question
Tim Newsham wrote: But it does look like 4GB minus about 680M. The machine is probably not capable of PAE or the kernel being used is not built for PAE support. Does the CentOS have /proc/config.gz or /boot/config-`uname -r` ? If so, grep for MEM in there. If HIGHMEM4G is already set, you could try setting HIGHMEM64G instead. It looks like 64G sets the PAE parameter. Don't know if it will slow things down. You can also try booting with kernel parameter mem= and then something like 4G or more likely mem=4096M. It looks like you need a kernel with HIGHMEM64G. -Eric Hattemer .
Re: [LUAU] goofy RAM question
Eric Hattemer wrote: If HIGHMEM4G is already set, you could try setting HIGHMEM64G instead. Oops, if this isn't clear, I mean grab a new kernel source, and configure it with this option, then install the new kernel. Editing a /boot/config-* or /proc/config.gz doesn't do anything. -Eric Hattemer
Re: [LUAU] Intel Doubles Down on Linux
Jim Thompson wrote: This was why I asserted that your complaints were against X applications and not X in general, or more specifically the X protocol. Though I do recognize that the X protocol isn't perfect. My point was, when people say 'X sucks', they usually mean 'X applications' suck or the 'X server configuration' sucks. These things do suck, but they can be (some would say 'are being') fixed. What I'd say sucks more than applications, more than vague issues with low level protocol stuff and how XYZ is missing, is that X11 seems slow. Really slow. On Windows, you don't need a fancy graphics card with hardcore graphics acceleration to drag windows across the screen. On mac, drivers aren't really an issue. On Linux, just dragging a local window around can feel laggy and jerky. You need to super accelerated proprietary graphics driver to make it feel better. Things are better more recently with the newer versions of Xorg and all that. What is worse, though, is when you actually try to use X over the network. It feels very synchronous. I use vnc, RDP, and X11 remotely. Let's say you click a fancy fade-in menu in each protocol. X11 wants to show you all 50 frames of the fade-in menu, and will not let you click on anything or do anything until you've seen them all. On a distant network, this might mean tens or seconds of waiting around doing nothing. In vnc or RDP, they will show you the first frame, the last frame, and don't appologize for not showing you the frames in the middle. This makes them seem so much more responsive. When I sit in front of my mythTV machine (mythTV runs QT), it has all of these really fancy qt buttons, and shading, and everything cute like that. If I sit in front of a vnc copy from my other machine, it looks almost identical. Yet the remote machine running vncviewer over the network actually responds MUCH faster than the local machine with X11. This is baffling. I believe this is because of the synchronous nature of X11, and the fact that VNC just tells the program that its drawn the 18 stages of button press, and whatnot, and gets on with it. That machine has a Geforce2 with NVIDIA drivers, DRI, GLX, etc., so I don't think its an issue that the video hardware can't handle the picture. So maybe its just mythTV, but I feel like you know your protocol has problems when a VNC emulation of it over the network seems to respond 10X faster than the protocol running natively on all local hardware. On the plus side, openGL games, such as America's Army actually run much faster on Linux than their direct3d counterparts on Windows. So its clear that X11 can be fast when it needs to be. Has anyone else noticed the 50 frames of button pushing and menu drawing in X11? I don't know anything about how the protocol works, but it just doesn't feel right. Maybe its the QT toolkit that's bugged, maybe its something that can be fixed with lots of spaghetti patches on top of the old code, or maybe it helps to show somehow that X11 is fatally flawed in terms of speed. Most modern desktop systems set X11 at a nice of negative 10. X11 tends to take up 25%+ of the processor when doing things like moving windows, even when you have accelerated drivers. Now maybe Microsoft is hiding something from us, but I've never seen Explorer.exe take up more than 1% of the processor when dragging windows around like mad. And I know that an amazing amount of desktop things you'd never expect are handled within the video card on OSX. -Eric Hattemer
Re: [LUAU] Re: Fedora mythtv cpu load
Whoever Whatever wrote: I will nice things at different level next time to see how they changes. I just tried to watch tv at 2200, I can see blocks, I can tell the most when switching to a channel with some noise/snow, so putting it back to 4500 made me happy. Oops, I should have been more specific. I do 2200 with 400x300. At the same bitrate, a smaller picture will give you better motion, and a larger picture will give you better stills. You can see if lowering the bitrate and the picture size still looks ok for you. But if you have the space and the power, may as well go for whatever looks best for you. -Eric Hattemer
Re: [LUAU] Fedora mythtv cpu load
Whoever Whatever wrote: Any thing else which I am over looking? I am recompiling the kernel today to see if that make a different. I did some search on google, there are alot of complain on fedora running slower compair to some other dist, I want to nail down the cause first else I will be switching to debian. Someone mentioned about KDE; I think what may be more likely is the threads priority. If you check in top for instance, X in one of the systems was set to nice 0, and one was set to nice -10. I don't remember which of redhat/mandrake used which. Try changing the nice values for X mythbackend mythfrontend And who knows what to change them to? I don't. If the stuttering happens inside the video file itself if you recorded it durring compiling, then you should start mythbackend with nice (-10,-17). If it happens durring live TV, then you should start mythbackend with (-5,-10) and mythfrontend (-10,-17). I always do my compilations with nice of 19. You might want to see if you can get one of the newer kernels with the ionice patch working. Make sure that selinux is set to the one that does nothing, not to the one that generates warnings. You could try a new kernel, even if you don't use ionice on it. My mythtv on fedora works fine on my system, which is comparable to yours. I use a bttv card, which is very processor intensive. I try not to run big processes while recording/watching, though. You should start those processe with nice -19, though. ionice would be really really cool if you had it working, but I never got around to it. Did you try compiling mythtv yourself, or are you using the atrpms version still? I've never touched that version. Try the CVS version, and make sure you set all the optimal configuration options. What bitrate are your recordings? -Eric Hattemer
Re: [LUAU] Re: Fedora mythtv cpu load
Whoever Whatever wrote: No, it wasn't KDE, I recompiled the kernel with athlon, then ended up recompling ivtv, lirc and other needed modules, reinstalled nvidia driver, system came up fast but with no sound. It was pretty good with very low load average, video not jerky while compiling. I have to recompile alsa 1.09 to get the sound working again on via82xx, looks like kernel/OSS driver not good for my board. I was back to that heavy load average with sound working again, I guess I nailed down the cause, most likely the sound modules, I will try the bleeding version later see if it's better. It sounds like it has a lot to do with the kernel and ivtv versions then. The alsa that comes with 2.6.11.11 will probably work, but may require a little bit of .asoundrc configuration (ie. black magic). I already have frontend running nice -15, do you mean options nice -15 or nice level of -15 ? let's call it nice -n -15 to be perfectly clear. But honestly, I don't know which one is best. Its kind of a question of whether X, mythfrontend, or mythbackend has too high of a priority, the others might lag and skip. If you record, then watch later, certainly the mythbackend should have a very important priority. If you watch in realtime, the mythbackend might need a bad priority and the frontend and X an important priority. The should probably all have a negative nice value so that default processes don't mess them up. Do I need to bleed for ionice? I remember back in years ago, I was chasing kernel release on 1.3.x tree to get ip masq working on my 33.6k dial up, stop doing that since the lap over of glib, remember new lib on kernel 2.x.x ? Like I said, I never got around to ionice, but it looks like a great (and well overdue) concept. It looked like 2.6.11.11 had an option for ionice, and I compiled it in, but I don't see it now, so maybe not. I don't have an ionice binary, so I assume that this version is built into the nice syscall. Maybe I just forgot what I saw. It might be related to cfq. Anyway, look at http://kerneltrap.org/node/4406/print . You might consider changing the io scheduler http://kerneltrap.org/node/3851 . I was compiling since .11 running on the good old Mandrake, I am just tired of compiling and try to find a easy way out, well.. there's another reason, .18 starting to use qt3.3 on some plugins, so I am too lazy to upgrade from qt3.1 to qt3.3. I will recompile it again to compare with the atrpms, but I will try to solve the high load average cause by sound modules first. Should be able to upgrade qt using yum, right? Should only take a few minutes, I'd figure. Then you can configure the mythtv sound the way you want it and feel comfortable that nothing has to do with some weird way that the atrpms stuff was compiled. What bitrate are your recordings? 720x480 bitrate 4.5k max 6k, I did try 3k.. same problem, but shouldn't matter much to the cpu on hardware encoder right? That's pretty hardcore. You're right that it won't matter on the cpu, but it will matter on the disk. All my normal recordings are 2200, and my cartoons and talk shows are 800 or so. If you check top, or gkrellm, is your cpu usage iowait or user? You might just be overloading the hard drive. The main problem I always had with the compiling was that even at nice -19, it still brutalizes the hard drive, and things start to skip. That's why I was always excited about ionice. However, he specifically says that its a proof of concept, so I suspect its not production ready yet. -Eric Hattemer
Re: [LUAU] Fedora mythtv cpu load
Whoever Whatever wrote: Hi, I was running Mandrake 9.2 for mythtv up to .18.1 on a shuttle sk41g using PVR350 for about 2 years, nvidia tv-out, no problem The PVR350 is hardware accelerated via ivtv, so that's unlikely to use more than 5% cpu. HD is enabled with DMA, using KDE as WM, the only difference would be xorg server on fedora, xwindow on Mandrake. Thanks for the help. I would blame the HD stuff. Make sure that it is running in ATA whatever mode. I have in my /etc/sysconfig/harddisks EXTRA_PARAMS=-X udma5 MULTIPLE_IO=16 EIDE_32BIT=1 You might need to use a different udma level for your needs. Also, if the nvidia TV out is in use regularly, make sure that you're using the nvidia.com driver. -Eric Hattemer
Re: [LUAU] Become a CACert Assurer and Create valid SSL Certificates for free
Michael Bishop wrote: HOSEF is sponsoring a cert party. For more information about this free service, visit http://www.cacert.org and click on the About Us link at the bottom. Are these certs in the common root cert directories, or do they give those This SSL certificate is not from a trusted source warnings? USC gives out free valid certs, but you have to click to get the SSL warning to go away. Not very professional. They pay for the certs that they use on their own websites. -Eric Hattemer
Re: [LUAU] MythTV on Fedora
Michael Matsuzaki wrote: j Hello all, I've been out of the house this memorial day weekend. It is wonderful to watch tv programs you when you want to watch them. I've been running mythtv for 14 months without a glitch. I'm using FC 1 because I've figured, if it ain't broke... don't fix it. One more thing I should mention about mythTV. Its by far the most pollished linux software I've ever used. They have the type of attention to detail that linux software never has. The frontend looks just like a Tivo or whatever, it typically installs without trouble, doesn't crash, and has tons and tons of features and addons. More linux programs need to work more like this. Firefox is probably the only one that comes anywhere close. When I first saw it, I said Wow, it runs under linux? Then why didn't I have fo fix or debug anything, or spend days looking at documentation and mailing lists to try to fix some weird dependency problem? Without networking it to other machines, its comparable to Tivo or any other mainstream DVR, but maybe more expensive. If you network it with other machines, you get a lot more flexibility. If you needed that machine anyway to do other jobs (mythtv won't use 100% or a good desktop's resources by any means), then its not expensive at all. -Eric Hattemer
Re: [LUAU] MythTV on Fedora
Hawaii Linux Institute wrote: Anyway, I am only interested in the most rudimentary part of MythTV (or any other similar package) that will allow me to, as a first step, watch TV or rental DVD movies via a coaxial-to-USB converter, from a Linux machine. Wayne Watching TV is much easier through tvtime or xawtv, which are easier to get up and running than MythTV. MythTV gives full PVR functionality, which is much better. There's really no need to install MythTV via rpms. It'll install onto just about any linux platform (or OSX); just follow http://www.mythtv.org/modules.php?name=MythInstall . I installed several different versions over the years and its really not too hard. I'm not in Hawaii anymore, though. Its not testing by any means. Its very stable and works fine. You can use one of the point versions, or you can use the CVS version. Both work well. The CVS is not unstable, but it does develop rapidly. -Eric Hattemer
Re: [LUAU] Relative Newbie Questions
Jacques L. Yerby wrote: Aloha Y'all, Years ago I installed a Slackware by hand on a PC so I thought I knew what I was doing. That was back when X-Windows was relatively new and I was pretty much running command line only. Is this the same old computer? How old is it? Possibly KDE is just too intense for it. Q1: What utility can I use to find out where the bulk of my files are? I'm running out of room and have created some bigger ext2 partitions and want to map the larger stuff over there. Tim is right on, but be careful not to include ... For that you would probably want to du -sk * .[^.]*|sort -n usage or something like that. Q3: I'm having some startup problems with KDE. It's a real slow startup. I don't have this problem with Gnome. (I have multiple windowing systems installed.) You might try starting with no WM, then typing startkde. See whether its hanging somewhere and timing out eventually. You may also want to upgrade to a newer version of KDE if the system can handle it. Q4: Is an NTFS partition WRITEABLE from Linux? All of my Win partitions are automatically mounted as RO. I don't want to screw around with this until I know more. (I'd like the capability of copying back forth.) There are at least three NTFS modules for Linux. You need the module to be compiled with rw support, then you need to edit the fstab to mount the partition without the -ro. People say that it is always dangerous no matter what. The two open source modules come with kernel 2.something and 2.6. Neither of the two open source modules will write to a winXP NTFS drive, but both should write to 2000 NTFS. Supposedly the 2.6 version was rewritten or something and won't allow you to create new files, or significantly change the size of an existing file. The two open source modules have a possibility to write to the FS incorrectly and completely ruin the FS to the point where windows can't read it and scandisk can't fix it. The third module is called CaptiveNTFS, and it uses the windows ntfs.sys file to do its job. This makes some people angry, so they refuse to use it, but supposedly it can read and write to NTFS perfectly. That being said, it may be safer to go the other way around: http://uranus.it.swin.edu.au/~jn/linux/explore2fs.htm . Safest is fat32/vfat, which is supported in just about all windows and linux with no scary You may ruin your computer if you use this warnings. P.S. Once again, at the mercy of my landlord I have to move. I'd prefer to stay out here on the N. Shore but if anybody sees anything nice $1000 I'd appreciate knowing about it. I'm a retiree and can't really afford much more than that. I hear its getting a lot harder to rent on Oahu. One of the biggest house rental landlords is liquidating. Maybe look for appartments on the West coast. -Eric Hattemer
Re: [LUAU] a different open source issue, from Maui
Tim Newsham wrote: The latest in this sad saga: http://www.ht-technology.com/cherryos-pearpc/cherryos-pearpc.html There seem to be a lot more than that one page: http://www.drunkenblog.com/drunkenblog-archives/000501.html If you follow the links in any of these pages, you can find more and more. Some of this may not be proven in a legal sense, but I think its unlikely that there's no code shared between the two. I suppose some of these may be libraries automatically added by compilers for all I know, but some of it looks pretty bad. -Eric Hattemer I'm not saying that he is wrong, but I dont find his current argument or his methods very convincing. If he (and others) are wrong, I would hate to be the one facing an overly rabid open source mob. Tim N. ___ LUAU@lists.hosef.org mailing list http://lists.hosef.org/cgi-bin/mailman/listinfo/luau
[LUAU] VNC extensions for X
http://www.realvnc.com/v4/x0.html I didn't know this existed. I think someone on this list asked about this. Basically the question is Can I get vnc in linux/unix to act the way it does in windows (ie. connecting the current desktop to a vnc session, rather than creating a new desktop)? The answer seems to be yes, if you use this X11 extension. I don't have a monitor attached to my linux machine anymore, so I can't really test this, but it does seem to work. At least, I can see the gdm screen and such. These extensions are precompiled in Xorg for FC3. I hear it kind of sucks to compile XFree86/Xorg from scratch. -Eric Hattemer
[LUAU] [Fwd: Fedora depmod -a Segmentation Fault]
Thanks. It wasn't me specifically who had the problem, but hopefully this will help those who did. -Eric Hattemer
Re: [LUAU] EMail Server Limits
Maddog wrote: My Exchange Server has a 16 GB limit and it was reached today and crashed the server. I have been trying to get these guys to buy me a server I can run Debian + Exim + Squirrelmail on without any luck. Will that combination crap out when the DB swells? I don't know about exim, but squirrelmail tends to depend on cyrus for sieve filtering, which uses individual files for each email, at least if they are delivered via postfix. So as long as your filesystem doesn't have a maximum number of files/inodes, you can have as much mail as you want. Straight postfix or sendmail uses one big file for the mail, but that's per account. If exchange is using one db for the entire system, even postfix/sendmail would lighten the load a lot. I just set up my mail server a week ago, so let me know if you have some questions on it. I don't have a lot of professional experience, but I know off the top of my head the answers to the types of questions someone new to linux email would ask. Unfortunately, you need several programs in a mail system, and every combination has a different setup. I use postfix, cyrus, spamassassin, and squirrelmail. Squirrelmail has a builtin filter interface based off a sieve server that cyrus provides. -Eric Hattemer
Re: [LUAU] The file that would not die...
You don't need to own or have write permission on the file, but you do need to have write permission to the directory its in. Make sure the directory that has it is at least drwx--. If that's not your problem, try running fsck on the (unounted) partition. -Eric Hattemer kilauea wrote: Aloha to all from Kauai I have a problem deleting a file. I play quake3 online and I think someone has found an exploit in networked quake3. The first time this happened, my game configuration file was deleted. Then today, the configuration file was somehow modified to make it inaccessible even to root! The game binary was also modified in this fashion. When I try to do a chown, chgrp, ls, or rm, I get something like rm: cannot remove `pak0.pa3': Permission denied . I can move the directories but can't delete the files. I have a reiser fs on Slackware. I also have RedHat on another partition so I booted RedHat and mounted the Slackware partition but this produced the same result. I moved the inaccessible files and restored the game files with no problems. I would like to just delete these files. Has anyone experience with this sort of problem? Thanks Kilauea ___ LUAU@lists.hosef.org mailing list http://lists.hosef.org/cgi-bin/mailman/listinfo/luau
[LUAU] qt problems
I have mandrake 10-cooker, libqt3-devel 3.3.2, gcc 2.96, 3.3.2, 3.4.0, binutils -2.14.90.0.7-2mdk and previously 2.15..., and I cannot compile a very basic qt program. If you have updated cooker in the last week, or an otherwise up to date system, please download and compile http://hattenator.dyndns.org/qapp.tar.bz2 and let me know what happens. You need to qmake qapp.pro -o Makefile gmake I get : undefined reference to `QApplication::QApplication(int , char **)' qapp.o(.text+0x20): In function `main': : undefined reference to `QApplication::~QApplication(void)' collect2: ld returned 1 exit status gmake: *** [qapp] Error 1 I have no idea why. It worked a month ago. Someone elsewhere mentioned I may need gcc 3.2. Is that sensible? -Eric Hattemer
Re: [LUAU] Windows shortcut equivalent in Linux
Vikram Khurana wrote: /Parse/Linux/Linux executable /Parse/Windows/Windows executable /Parse/123.dat Now the user could save this anywhere they want. Now in Win I can tell people to create a shortcut to /some directory/Parse/Windows/Windows executable place it in any folder they want. The 123.dat file is read relative to the executables. How do I do the same in Linux? Probably the best way would be to make sure its in your path and make use of which and dirname. execdir=$(dirname `which linuxParse`) datadir=$execdir/.. This will only work well if you put it in your path and don't just call it directly without it in your path. A softlink might cause a problem too. -Eric Hattemer
Re: [LUAU] How Does this Work?
From a more technical explaination, you can refer to rfc1738 among others, if that kind of thing excites you: http://www.faqs.org/rfcs/rfc1738.html . I can't explain that particular URL. The URL RFC explains that there are several special characters including @, :, that aren't considered normal text. Also, %HEXHEX represents the character of that numerical value. @ is a simple, yet somewhat obvious method. When a site asks for a password, you can either wait for it to ask, or you can type http://user:[EMAIL PROTECTED] You can leave the password out if you want. If the site doesn't actually require a user/password, it will ignore it. So you can use anything you want in the username. [EMAIL PROTECTED] will take you to google, and microsoft has no effect. Domain names don't have to be used. http://216.239.57.104 will take you to www.google.com just as well. However, even non-technical people know what an IP is, so that's too obvious in some cases. IP's can be written in other forms with hex or octal and in some cases the .'s can be omitted. The sign depends on the browser. Old versions of IE and other browsers used to read an as ignore everything before this, so www.microsoft.com/stuff/stuff/stuffwww.ijusthackedyou.com wouldn't get you to microsoft. The is much less obvious than the @, but doesn't seem to work anymore, or at least not on mozilla. http usernames and passwords don't really work with '/' marks. So www.microsoft.com/[EMAIL PROTECTED] would fail or get you to an error page within microsoft. %HEXHEX makes any charater, printable or not. %00 is NULL or \0. NULL is used to terminate a string in most programming languages. If you fill char[40] with abc\0def and leave the other 33 chars as the default, the 'string' in that array is abc. If you print www.microsoft.com/stuff/[EMAIL PROTECTED] shows up as www.microsoft.com/stuff in some cases. Otherwise you can print entire URL's in %xx%yy%zz format. You can easily abuse javascript for some purposes. A lot of URL's are of the form athis link/a but some are of the form awww.stuff.com/a. Although the second is the same as the first, and that text could be anything, people are convinced that if the link contains a url, it must point to that url. Javascript pseudo code something like: onMouseOver: statusBar.print(url) will print the url in the status bar when you point the mouse at it. This emulates the normal behavior when you point to a link in most web browsers. There are other tricks, but I don't know all of them offhand. -Eric Hattemer
Re: [LUAU] Mainline (generic) 2.6 kernel with FC2
1. The nv driver that ships with the OS does the same software rendering other open source drivers do. 2. The ati open source driver is not made by ati, and has very little hardware rendering functionality. 3. The only thing on the nforce boards that require drivers is the network card, and I'm not even sure that's true anymore. 4. There is an nvidia installer option that allows installation of multiple modules. I believe its -k, but check NVIDIA*.sh -A; for the advanced options. 5. With this in mind, I don't think there is reason to not buy NVIDIA stuff. There aren't really any reasonable alternatives besides ATi. But on the plus side, the upper range of ATi Radeon cards do seem to outperform (outbenchmark) the upper range of NVidia cards. -Eric Hattemer Hawaii Linux Institute wrote: I installed the mainline 2.6.5 kernel over the weekend on both FC1 and FC2T1/2. Have not seem to have any problem (except the following caveat). The biggest annoyance (to put it politely) is that, for systems which have either the nForce chipset or the GeForce video cards, the drivers must be recompiled for a specific kernel. Big bad pain in the rear. This also means that you are practically prevented from dual-booting from two kernel versions which share the same root partition. Thus, unless nVidia becomes more open-sourcely correct (if ever), stay away from their products. Use VIA chipsets and ATI cards. Other than that, the mainline kernels appear to work seamlessly with Fedora Core. For desktop users, you will love the new pre-emptive feature of the 2.6 kernels. (FC2 has 2.6.5 kernel, but I don't think this feature is turned on by default). wayne ___ LUAU@lists.hosef.org mailing list http://lists.hosef.org/cgi-bin/mailman/listinfo/luau
Re: [LUAU] apt RPM
Vince Hoang wrote: Apt is just less annoying to me. FC2's yum has --download-only now, so my #1 issue is gone. Now if yum would stop downloading individual headers for _every_ package... In a sense this behavior is better. apt does seem to keep its header database relatively small, but it stil has to download the whole database every time. urpmi does not do this as well and as a result, hdlist.cz is 20MB. yum headers are each tiny, and yum only needs to download ones that were changed or haven't already been downloaded. So I think in theory yum has the best approach. In practice, setting up a separate http/ftp connection for each file may add some time, plus other overhead dealing with multiple files. But yum could download multiple headers in parallel. The main concern is the total size of all headers for each. I think urpmi loses by far, but apt and yum are pretty similar, and yum has the advantage of not needing to download the whole database if one package changes. Having said all that, I have only used yum twice and almost exclusively use apt. Make sure that with apt-get you only use 'upgrade', and not 'dist-upgrade' unless you know what you are doing and are prepared to fix some problems. That's the main problem people have with apt is that everyone tells them to dist-upgrade, then it deletes kde or something. I haven't read the docs for it, but the format of the yum sources configuration file is ugly and strange. apt just uses a URL, which is pretty straightforward to change. The mirror-select function in apt is great and does all that work for you, though. RedHat does not seem to be committed to using and supporting apt, though. At least not yet. -Eric Hattemer
Re: [luau] Fedora depmod -a Segmentation Fault
I spent a long time trying to get ndiswrapper working on my friend's laptop. It required a 2.6.1-1.43 kernel rpm. I can't quite remember what else we had to do, but I remember on that particular machine it was a lot of work. It may be simple if you use the 2.6.1-1.43. But it took us a while to figure out that it doesn't work well in 2.4.22. You need to compile the ndiswrapper package (gmake install). We got a lot of errors and hardlocks in 2.4. That one is no longer readily available in rpm form, though. the 2.6.3 and .4 weren't so good either. So the main point I'm making is that we tried several different kernels and got one to work. So you may need to compile a new kernel or try the 2.6.4 or 2.6.3 from http://mirrors.usc.edu/pub/linux/fedora/fedora/fedora/1.90/i386/RPMS.updates/ . And when I say mirrors.usc.edu I really mean the equiv directory on the local mirror, ie. download.fedora.us. -Eric Hattemer Kevin English wrote: I'm running Fedora with kernel version 2.4.22-1.2115.nptl. I'm trying to install the newest verison of ndiswrapper (0.6) so I can get my linksys WMP54G card working again. For some reason when i do a depmod -a, I get a Segmentation Fault.I posted to the Fedora-list too but since so many local folks use fedora, I thought I'd give this a shot too. Here's what i get when i do 'depmod -av' xftw_readdir /lib/modules/2.4.22-1.2115.nptl/unsupported/net/wanrouter user function /lib/modules/2.4.22-1.2115.nptl/unsupported/net/wanrouter user function /lib/modules/2.4.22-1.2115.nptl/unsupported/net/wanrouter/wanrouter.o xftw starting at /lib/modules/2.4 lstat on /lib/modules/2.4 failed xftw starting at /lib/modules/updates lstat on /lib/modules/updates failed xftw starting at /lib/modules/kernel lstat on /lib/modules/kernel failed xftw starting at /lib/modules/fs lstat on /lib/modules/fs failed xftw starting at /lib/modules/net lstat on /lib/modules/net failed xftw starting at /lib/modules/scsi lstat on /lib/modules/scsi failed xftw starting at /lib/modules/block lstat on /lib/modules/block failed xftw starting at /lib/modules/cdrom lstat on /lib/modules/cdrom failed xftw starting at /lib/modules/ipv4 lstat on /lib/modules/ipv4 failed xftw starting at /lib/modules/ipv6 lstat on /lib/modules/ipv6 failed xftw starting at /lib/modules/sound lstat on /lib/modules/sound failed xftw starting at /lib/modules/fc4 lstat on /lib/modules/fc4 failed xftw starting at /lib/modules/video lstat on /lib/modules/video failed xftw starting at /lib/modules/misc lstat on /lib/modules/misc failed xftw starting at /lib/modules/pcmcia lstat on /lib/modules/pcmcia failed xftw starting at /lib/modules/atm lstat on /lib/modules/atm failed xftw starting at /lib/modules/usb lstat on /lib/modules/usb failed xftw starting at /lib/modules/ide lstat on /lib/modules/ide failed xftw starting at /lib/modules/ieee1394 lstat on /lib/modules/ieee1394 failed xftw starting at /lib/modules/mtd lstat on /lib/modules/mtd failed xftw starting at /lib/modules/unsupported lstat on /lib/modules/unsupported failed Segmentation fault - uname -a Linux localhost.localdomain 2.4.22-1.2115.nptl #1 Wed Oct 29 15:31:21 EST 2003 i686 athlon i386 GNU/Linux Kevin W. English Software Developer, X5Software [EMAIL PROTECTED] 808.689.8161 (h) 808.271.5125 (c) ___ LUAU mailing list LUAU@lists.hosef.org http://lists.hosef.org/cgi-bin/mailman/listinfo/luau
Re: [luau] Newbie Query - Browser Plugin?
Ryan Kawailani Ozawa wrote: Great answers, everyone. Thanks! So! I tried urpmi, and used easyurpmi linked at urpmi.org, but although the man entry for urpmi is present on my Mandrake 9.2 install, the utility/executable itself couldn't be found (bash: urpmi.addmedia: command not found). You may need to install/reinstall urpmi. I decided to go the RPM route, and found both the firefox version linked here (cooker/beta) and the Macromedia Flash RPM at UCSC. Unfortunately, I got stuck there, too. Attempting to run rpm on both of them reported Warning: filename V3 DSA Signature: nokey yada yada yada error: failed dependencies: libXinerama.so.1 is needed by filename Ah, those pesky dependencies. Where to from here? g One place to go is www.rpmfind.net (which has been swamped lately). A search on libXinerama.so.1 leads to XFree86-libs, which I would have figured you would need to run X11 in general. But anyway, go to one of those http/ftp rpm directories and find XFree86-libs*.rpm and install that one first. Urpmi will automatically handle dependencies. addmedia may have been dropped in favor of the gui stuff. Check in the mandrake control center for an rpm tool. It may be called something like mandrake-control-packages or something. Otherwise grab urpmi*.rpm http://mirrors.usc.edu/pub/linux/distributions/mandrake/9.2/i586/Mandrake/RPMS/urpmi-4.4-37mdk.noarch.rpm http://mirrors.usc.edu/pub/linux/distributions/mandrake/9.2/i586/Mandrake/RPMS/urpmi.setup-0.4.4-4mdk.noarch.rpm http://mirrors.usc.edu/pub/linux/distributions/mandrake/9.2/i586/Mandrake/RPMS/rpmdrake-2.1-35mdk.i586.rpm http://mirrors.usc.edu/pub/linux/distributions/mandrake/9.2/i586/Mandrake/RPMS/XFree86-compat-libs-4.1.0-3mdk.i586.rpm Try installing those rpms. Unfortunately I don't have a mandrake machine anymore, so that's why this stuff is all slightly vague. I may move back to mandrake, though, since I'm not so far impressed by Fedora. -Eric Hattemer
Re: [luau] Newbie Query - Browser Plugin?
Ryan Kawailani Ozawa wrote:
So! I tried urpmi, and used easyurpmi linked at urpmi.org, but although the
man entry for urpmi is present on my Mandrake 9.2 install, the
utility/executable itself couldn't be found (bash: urpmi.addmedia: command
not found).
One more thing... If you have urpmi installed, it probably comes with
one source pre-configured. It may be a slow one in france or something,
but you could probably start using it immediately. Try urpmi.update -a
and see if that works, then just run urpmi rpmdrake, or something like
that. To get the firefox and all that, you'd need to add a cooker
source, though. There is a file to manually edit. You can manually
edit /etc/urpmi/urpmi.cfg with a new source. I believe you just add
the lines from urpmi.addmedia directly into there, such as
USC9.2_1
ftp://mirrors.usc.edu/pub/linux/distributions/mandrake/9.2/i586/Mandrake/RPMS
with ../base/hdlist.cz
USCcooker_1
http://mirrors.usc.edu/pub/linux/distributions/mandrake-devel/cooker/i586/Mandrake/RPMS/
with ../base/hdlist.cz
USCcooker_2
http://mirrors.usc.edu/pub/linux/distributions/mandrake-devel/cooker/i586/Mandrake/RPMS2/
with ../base/hdlist2.cz
and so on. With USCcooker_2, you can just do urpmi.update -a ; urpmi
firefox, and it should handle all the dependencies and everything.
However, you may want to comment out all cooker sources after you have
firefox installed. In my opinion, Mandrake cooker is extremely well
built, and 5 times more stable than redhat rawhide. I updated from
cooker daily for a year and never had any serious problems. But,
especially for servers, try only to use the 9.2 stuff. If it doesn't
take that urpmi.cfg format, try this one:
USCcooker_2
http://mirrors.usc.edu/pub/linux/distributions/mandrake-devel/cooker/i586/Mandrake/RPMS2/
{
hdlist: hdlist.usccooker2.cz
with_hdlist: ../base/hdlist2.cz
}
-Eric Hattemer
Re: [luau] Newbie Query - Browser Plugin?
Ryan Kawailani Ozawa wrote: I'd love to use urpmi and RPM... if you can bear with me and explain exactly what that is! Is it automated versioning or whatever like CVS? I take it maybe you're used to some unixy system? RPMs are similar to windows installers. Somebody takes the sourcecode, compiles it for a given linux distro, zips it up, then puts it into an rpm. It is like a zip file with some extra scripting to handle upgrades and stuff. An rpm will not work on a system drastically different from the one it was compiled for. So you would look for mandrake 9 rpms (redhat 6.2 rpms probably won't work, as they rely on older libraries etc.). To deal with rpms, you must first find one. Writers of certain programs make their own rpms, and mandrake keeps a big database of their own. A mirror here at USC is http://mirrors.usc.edu/pub/linux/distributions/mandrake/9.2/i586/Mandrake/ . Out there, videl.ics.hawaii.edu doesn't seem to mirror mandrake anymore. http://sluglug.ucsc.edu/macromedia/site_ucsc.html has the flash rpm. just grab the download for mandrake. Warren was talking about a thunderbird rpm from mandrake, but I can't find a firefox one. You can get firefox from Mandrake cooker (beta) at http://mirrors.usc.edu/pub/linux/distributions/mandrake-devel/cooker/i586/Mandrake/RPMS2/mozilla-firefox-0.8-14mdk.i586.rpm . So go ahead and download the mozilla-firefox*.rpm and the flash plugin rpm. Then you need to install those. To do that, just open a terminal and type rpm -Uvh mozilla-firefox*.rpm rpm -Uvh flash*.rpm And that's it in a nutshell. For more information on rpm, type 'man rpm' or 'rpm --help'. As for urpmi, it automatically finds, downloads, and installs rpms for you. Refer to http://urpmi.org for information. Use their easyurpmi tool to add mirrors in your urpmi configuration. You may want to add plf, since it has all the cool stuff with questionable licenses, such as gameboy emulators, windows media player codecs, cable descramblers, etci. -Eric Hattemer
Re: [luau] Newbie Query - Browser Plugin?
Ryan Kawailani Ozawa wrote: The thing is, clearly I've forgotten much of the basic concepts of Linux, and specifically, I just can't get the Macromedia Flash plugin to work in either Firefox or Konqueror. The installer plops it in /home/user/.mozilla, and neither browser sees it. I tried manually moving the files from .mozilla to /firefox/plugins, but that doesn't do the trick. Any ideas? One thing I think I did wrong was installing Firefox under a user directory (/home/user/firefox) rather than in a... main directory. Could that be the problem? If so, though, I'd think at least the default Konqueror install would see the plugin files... Please explain a little more about where you got both. It sounds like you got the zip file for firefox from mozilla.org? And the macromedia plugin, where did that come from? You may want to see if there is an rpm for each of those available from urpmi. Let us know if you need help with urpmi. Firefox should check two places for its plugins. One is the directory its installed to (usually /usr/lib/firefox, but in your case something like ~/firefox). Installing the plugins to ~/firefox/pluginsshould usually fix that, I would think. The other place it looks is ~/.phoenix/plugins. That was the original original name. See if copying the files or making symlinks to there fixes the problem. I expect it would, but I'm not in front of a linux machine to see how redhat did it for their fedora rpms. Many thanks in advance for your patience and insights! And if this Mickey Mouse stuff is inappropriate for this list, let me know... and point me at a local Linux for Dummies group instead (if there is one!). This stuff is absolutely appropriate for this list. You happened upon the list in a couple of weeks when the discussion was all fiber channel and such, but usually there are plenty of questions like this one. Some linux groups are nasty about this kind of stuff (I'm not naming names, though), but LUAU has always been supportive. Generally all advanced linux types do appreciate it if you show that you've done some work or research beforehand. But this list especially understands that knowing where to do research may be the trickiest part for someone new to this stuff. If I find a good doc on mozilla.org I'll post that in a sec. -Eric Hattemer
Re: [luau] Newbie Query - Browser Plugin?
http://plugindoc.mozdev.org/faqs/firefox-linux.html So it is wherever you installed it/plugins, and ~/.phoenix/plugins, like I said. Make sure you have both the .so and the .xpt. You may need whatever the mandrake version of compat-glibc is (it may very well have the same name). It may have a menu item, like Help/About Plugins, which will tell you what plugins it detects. -Eric Hattemer
Re: [luau] 64-bit Computing
You might want to talk to Warren about it, but I think with the Fedora, you'd do better just to install the 64, and use 32-bit rpms when necessary. There shouldn't be reason to need to install both cores. -Eric Hattemer Hawaii Linux Institute wrote: I am expecting to receive an HP Athlon64 notebook in about a week. I plan to do a triple-boot install, XP/Fedora32/Fedora64. Anyone who has good info (or bad), comments, etc., please let me know. Thanx a whole bunch. wayne ___ LUAU mailing list [EMAIL PROTECTED] http://videl.ics.hawaii.edu/mailman/listinfo/luau
Re: [luau] Linux + NTFS re-partitioning (was Linux)
From all the direct and list replies it seems that no one has had the same problems I had long ago. I guess I can stop discouraging parition resizing to friends who ask about it. Personally I don't know if I'd ever feel right about doing it myself, though, since windows needs to be reinstalled every year or so anyway to keep it running properly. Thanks to everyone for all of the good success stories. -Eric Hattemer
Re: [luau] Linux + NTFS re-partitioning (was Linux)
I spoke too soon. One very discouraging experience similar to mine. I think the link is that both my experience and the one in question is that it was pre-2000. Perhaps the software has become much more reliable lately. Maybe I will always worry about resizing, but I'm glad it works for the rest of you. -Eric Hattemer Eric Hattemer wrote: From all the direct and list replies it seems that no one has had the same problems I had long ago. I guess I can stop discouraging parition resizing to friends who ask about it. Personally I don't know if I'd ever feel right about doing it myself, though, since windows needs to be reinstalled every year or so anyway to keep it running properly. Thanks to everyone for all of the good success stories. -Eric Hattemer ___ LUAU mailing list [EMAIL PROTECTED] http://videl.ics.hawaii.edu/mailman/listinfo/luau
Re: [luau] Linux
To partition your hard drive you will need a separate program. I have used PartitionMagic and recommend it, but it is commercial and the status is a bit unclear now that they have been purchased by Symantec (http://www.powerquest.com/). Other possibilities here: http://directory.google.com/Top/Computers/Software/Disk_Management/Partition/ I've always wanted to take a poll. How many of you have used a partition resizing tool such as partition magic and seen massive filesystem corruption immediately or up to a month later? That's why I shopped using or recommending those products, and was wondering if its at all common. If you want to keep from flooding the list you can send It worked great for me responses directly to me, but mainly I'm just looking to see if there are any negative responses. -Eric Hattemer
Re: [luau] Trying to upgrade from RedHat8 to Fedora
I haven't taken a look at the actual installer for fedora, but see if there is info in the (ctrl)-alt-F1, F2, F3, F4 terminals. You may be surposed if its something like downloading fedora, FTP: too many users or your class or something like that. It may also tell you where its trying to save the data. Do you have a complex mount point/partition system, or just a /boot and a /? Although some would recommend against it, you can probably update the whole system using apt-get/yum. It works for 9.x for me without any real concerns. -Eric Hattemer Brian Chee wrote: Hi folks: We're getting an error while trying to upgrade (sidegrade?) and the system is giving us a message of something like: ...uploading install image to hard disk, error, possibly out of hard drive space...is there a way to tell Fedora to use a particular partition for the upgrade stuff? /boot has 38mb free, but the other partitions got gobs of space to play with. Any wisdom? /brian chee University of Hawaii ICS Dept Advanced Network Computing Lab 1680 East West Road, POST rm 311 Honolulu, HI 96822 808-956-5797 voice, 877-284-1934 fax ___ LUAU mailing list [EMAIL PROTECTED] http://videl.ics.hawaii.edu/mailman/listinfo/luau
Re: [luau] Which Distribution To Go With
I firmlly believe that no one should comment on distributions they haven't tried and know little about. I've alternated between Mandrake and redhat every year for the last several years. I use Mandrake because its better, and Redhat every time Warren claims its better because of some kernel improvement or fedora, etc. I firmly believe average people would like Mandrake a lot better. But then again, it depends a ton on what you're using it for. If you want a very unixy system reminiscent of 1965, there are better solutions (ie. solaris, openBSD (and yes, I've used both of these)), But if you want a home user system with movie programs, tv programs, and office programs that actually work together, you might want to look at Mandrake. It had urpmi way before redhat had apt-rpm. The installer has always been better. It decides what changes you make to packages will effect dependencies for other packages in realtime. It detects your video card properly and will use vesafb if its unsupported. Redhat8 will hardlock in the installer on certain geForce cards (mobile and nforce). Mandrake is nearly 100% compatible with redhat rpms. In the worst case, you may need to find a .src.rpm and rebuild it. Because of this, it has plenty of programs available. Probably not as many as debian, but if you can think of something its missing, let me know. Now if you're upset about the webadmin system of administration, that's one of many optional packages to do system administration. I'm doubtful that many people use it. Linuxconf is available, along with the gnome and kde control panels. All of these get their settings from files in /etc and dont' really cache these settings, so you can seemlessly use both these and edit /etc files. I don't know what someone might want from an administration tool, but the mandrake control center is top notch. It has gui menus for so much that redhat does not. The printer control panel detects your printer, then automatically can install rpms for relevant things like cups or lpr. Then it updates programs like mozilla to use that new printer. I was amazed that I still haven't found a graphical default runlevel editor for redhat (although I do know how to do it manually). Once you get used to single boot diskette based network installs for redhat, mandrake, and debian, its hard to go back to downloading gigabytes of CDs unless you already own the CDs or are installing many machines. -Eric Hattemer Tom_Gordon/RISE/[EMAIL PROTECTED] wrote: Mandrake is just a shittier version of Red Hat. I've not used the drake-fly distro because of it's insistence that web-based GUI are acceptable means for system administration. Did you know Mandrake was once just a modified version of Red Hat Linux?
Re: [luau] Num Lock
BIOS is an option on any computer, as long as its an actual computer. The problem is, linux resets that setting once it boots, so it doesn't matter. Mandrake has a service called numlock that will turn on the numlock. Redhat does not have such a service. You could try installing the mandrake rpm... http://rpmfind.net/linux/rpm2html/search.php?query=numlocksubmit=Search+... I make no guarantees, though. Try rpmbuild --rebuild *.src.rpm on the src.rpm version. It'll appear in /usr/src/RedHat/RPMS/*. do a find /usr/src/RedHat -name *.rpm to find it. Redhat is missing so many of these cute things that make Mandrake so much more useable as a regular home user. As much as Warren tells me how cool redhat will be when he's done with it and how great their kernel is, I always get depressed sitting here on the RH machine when I remember how well things worked in mandrake. But try rebuliding that rpm and we'll see how it turns out. -Eric Hattemer
Re: [luau] Counterpoint: Linux vs. Windows Viruses
There are two important parts that the OS vendors play in this, though. First is making sure that unneeded services are trimmed down. In my mind, RedHat is bad at this. Just because I have sendmail installed doesn't mean I'd like it to start up the first time I boot. Services that should start on first boot should be the few that are needed for the computer to function and allow advanced users to turn on the extra services. SSH could be an exception, since it allows secure remote access, but with the latest SSH exploits, maybe not even that. However, at least linux has their services modularized. RPC in windows is the biggest security hole there is, and it can't be turned off. Kernel level exploits in linux are rare if ever, and usually require iptables or something. The second part they play is making updates obvious. They cannot overstate these. I deal with virus/worm calls at work, and they seem to be all I do anymore. We ask people if they click install on the windows updates that automatically pop up in windows XP and they say something like I didn't know what it was so I canceled it. OS vendors need to either have security updates automatically install themselves unless an advanced user turns off that feature. The other alternative is to make the popups so frightening that no one could refuse. That would be something like Install this update or your computer will likely be hacked into and your data erased. Most linux distributions don't make security updates too apparent to a novice user. But then again, I've become convinced over the last two months that only 10% of the American population should be allowed to use a computer at all. And really that's 10% of the college educated population. -Eric Hattemer
Re: [luau] Slashdot: Multiple Monitors Increase Productivity
Xinerama allows that? I thought that was required to have the 0.0/0.1 setup. I guess I'll look into that. Thanks. -Eric Hattemer Hawaii Linux Institute wrote: I was talking about a different dual monitor setup, Eric. The one operated with the Xinerama option turned on (fully functional under KDE, but only partially implemented under GNOME 2.4). This option allows your window screen to become double-wide, very useful for working on large spreadsheet or databases. Of course, if you don't need the other monitor, you can always turn if off. ___ LUAU mailing list [EMAIL PROTECTED] http://videl.ics.hawaii.edu/mailman/listinfo/luau
Re: [luau] Slashdot: Multiple Monitors Increase Productivity
I have to say I don't agree. Depending on where you get them (outside of Hawaii) you can easily get a 17-19 CRT for under $100. The obvious problem with desktop switching, or alt-tab switching is that you can only view one document at a time. There are so many applications for which there are two related document windows. For instance, a webpage and a text editor for doing web development. I like to play UO on one monitor and have the other monitor show the map. In windows, windows can be spread across multiple desktops, which allows you to save on that 35 monitor. Any sort of switching is a pain, and its very nice for me to be able to modify an entire electronic circuit without needing to zoom around. I'm glad I have two monitors at home. At work I always work on a sun and a pc at the same time. I seriously doubt I'd be more productive if you gave me a $100 raise over 5 years. But in those 5 years I'd get a lot more done if I could see all my programs at the same time. Unfortunately the savage driver doesn't work properly with my cheapo graphics card, so I only get one and a half monitors in linux. Its great in windows, though. I agree with you that flat panels/LCDs are uttlerly silly and useless, though. -Eric Hattemer [EMAIL PROTECTED] wrote: Or you can cut your chances of getting cancer in half by sticking to your trusty desktop switcher :) The only employers I know that bust the bling for dual display are game developers and graphics artists. I always have less monitors than I do computers because extra monitors are wasteful and expensive. It's like buying an SUV; my old crt will always look better than your new dual flat-panel. Besides, I bet the employee satisfaction from having two monitors alone probably attributes to the the 10%, yes, only 10% increased productivity. I can get more productivity by increasing wages, benefits or attending to cultural and social agendas and needs! I don't need a study to tell me that either, although i'm sure they've been done many times. Usually studies that show such a small variation in performance and have commercial contributors [NEC] are found to be false. Tom
Re: [luau] Single User Mode (was: ultra ata 100 okay?)
Plus in init s you can get more accurate results with hdparm -tT /dev/hdx. So you know whether your changes actually matter. -Eric Hattemer Tom_Gordon/RISE/[EMAIL PROTECTED] wrote: i was reffering to making changes with hdparm. if you set something that may cause corruption you don't want all those other processes/levels making reads/writes in the background so you boot to single-user mode first. Tom
Re: [luau] ultra ata 100 okay?
If you do need help with lilo and booting, just ask. I finally figured out how to boot the promise card properly, and therefore feel qualified to answer questions. I ran into a problem where the kernel picks drive letters according to where they are located on the motherboard, wheras lilo can't match those letters to the actual numbers the bios assigns. But I fixed it. More explanation if necessary. -Eric Hattemer
Re: [luau] ultra ata 100 okay?
I bought a WD 250GB (230Gib) drive last week and it works fine in RH9. It comes with a promise card. Windows cannot read the drive on the nvidia controller, so I had to use the promise card. The only issue is unless your bios is pretty hardcore, it will boot as hde or higher. I don't remember why, but my lilo is on my onboard controller hda. I think it was just easier that way. I know the installer told me my system wouldn't boot, I ignored it, then I had to repair it :). So the lilo part might be tricky. Windows is twice as bitchy, though. I now realize the steps I need to take to get windows onto hde, but it shouldn't be as hard as it will be. As for hdparm, you want to make sure you have it installed, then edit /etc/sysconfig/harddisks. Set USE_DMA=1 and at the bottom where it says extra_parms or something, set -X 69. I like to set -S 0 -M254, but that's just me. USE_DMA is necessary if you don't want HD reads to lock up the whole computer. -X 69 sets the drive to udma5 instead of dma33. -S 0 keeps it from going to sleep, and -M 254 means to run as fast and loudly as possible (which may shorten the life of the drive, so beware). And make sure you put the extra_parms line in quotes (ie. EXTRA_PARMS=-X 69 -S 0). But like I said, with other controlers and drives on that computer, lilo may be hard, and grub may be impossible. Windows may involve disabling the other dma channels (or unplugging the other HD's) during instalation, then getting lilo to boot the WD drive, then installing windows with the F6 driver disk. -Eric Hattemer [EMAIL PROTECTED] wrote: knowing the make model of your current controller will be plenty enough to answer that question. that can be found out by knowing the make model of the motherboard, if it is integrated. and that can be found out by writing down the numbers on the boot init /mem test screen and googling it. you also won't be getting the same kind of performance as you would in windows becuase RH9 uses concervative settings (compared to windows) so tweaking hdparms in single-user mode is a must. google is your friend there too. i hope this helps tom Please respond to [EMAIL PROTECTED] Sent by:[EMAIL PROTECTED] To: luau [EMAIL PROTECTED] cc: Subject:[luau] ultra ata 100 okay? I'm trying to figure out if this western digital 175GB 7200rpm ultra ata/100 hard drive will work in my linux system. I went to redhat and looked for their HCLs, since I am running RH9, but I don't find anything about western digital when I search. This is a bit suspicious that none of their drives would be listed. Anyhow, it has a special pci controller card warns all the windows people to load the drivers. What should I do to find out? TDB ___ LUAU mailing list [EMAIL PROTECTED] http://videl.ics.hawaii.edu/mailman/listinfo/luau ___ LUAU mailing list [EMAIL PROTECTED] http://videl.ics.hawaii.edu/mailman/listinfo/luau
Re: [luau] I need help
Something got mixed up in your directions, I think. Here's what you should do. Boot into windows, figure out what drive letter your fat32 is. (This isn't necessary, but will help you keep track of some things). Then boot into linux and create a similarly named directory. You can call it fat32, or you can call it D or E, but just don't put a : in there. You can put that in / or in /mnt. So you will type a command like: su -; mkdir /mnt/D; The next step is to mount the drive. Make sure it is /dev/hda5. You can verify this with the fdisk command. fdisk /dev/hda; p q It will print the partition table. Then for the mounting (which you will need to do every reboot unless you want to put it in your fstab or similar). mount -t auto /dev/hda5 /mnt/D; Then if that goes through without error, you should have your files in /mnt/D. Like I said, you can substitute any name you want for this dir, but carry that substitution through all of these commands. You can check for that pdf by typing ls /mnt/D/ Or look though there in some file browser. -Eric Hattemer Felipe Vegan wrote: OK He's the deal.. I'm a new linux user, and I'm trying to share files between Windows and Linux, So that's why I did a extra partion FAT32. In Windows I put some files PDF in the FAT32 partition so I could see them on Linux. So thanks to the emails the some of you people send me (Wish I appreciated very much :-) ). I put on Linux Terminal the commands mount -t vfat/dev/hda5/mnt/fat32.. I put /dev/hda5 because I saw in the hardware browser that was the direction of the FAT32 partion..But now what I have to do to see on Linuxs the files PDF that I put in the FAT32 partition when I was in windows?? Some friend told me that I could see the PDF files in /mnt but I only can see the floppy and the cd folders when I'm there.. So what I need to do?? Felipe
Re: [luau] nForce2, KM266 Business Linux Desktops
I have the nforce 1, and its not particularly tricky on linux. The graphics chip boots RH7.2+, I believe. If you want to use the onboard pci devices, I would suggest the .src.rpm. rpm -ivh *.src.rpm, then go into the directory it dumps it to. I think that's /usr/src/RedHat/SOURCES. Go into the Nvidia directory in there, and read the file in there called something like README. It says something like to do a ./configure then a gmake; gmake install. You'll need to redo this after every kernel upgrade. -Eric Hattemer Hawaii Linux Institute wrote: Both nVidia's nForce2 and VIA's KM266 provide excellent platforms for use as business Linux desktops--they are very powerful, flexible, stable and cheap (especially the KM266 based systems). The nForce2 also has a duel-monitor option, which can be very neat for certain types of applications (e.g., CPAs, financial planners, estate planning lawyers). Both systems, as far as I know, still require some tweaking. RH9/MDK9.x won't even install on KM266. Some Debian based systems (e.g., Libranet) will install on KM266, but it installs the generic VESA driver, and not the more powerful Savage driver. nVidia's instructions on installing nForce driver are so full of errors, it took me a while to figure out what's going on (particularly in trying to install the NIC driver). Would appreciate if someone would also share their experience. ___ LUAU mailing list [EMAIL PROTECTED] http://videl.ics.hawaii.edu/mailman/listinfo/luau
Re: [luau] Linux in Hawaiian Industry
Welcome and so forth. I'm not currently in Hawaii, and am probably not comming back (at least not very often), so I may be kind of biased. But if you can change your move to Oahu rather than Maui, you'd probably do a lot better. Maui and commercial industry don't seem compatible in my mind. Anyway, even Oahu isn't for me... But it does have an IBM, and IBM is pretty into linux nowadays. I doubt they're hiring, though. There's a lot of people on this list who do something with linux, although not all in a professional capacity. Wayne does some linux work in law (along with law work), Ho'ala does his own business with contracting and such, Warren has some projects and works with mic-pacific institute (high-school) sometimes. Mainly I hear about volunteer projects like Hosef and the McKinley sessions. -Eric Hattemer
Re: [luau] Powering Off a Linux Machine
I can't remember a time when you couldn't shut down linux with a button. Maybe there was a time... But I used redhat since 6.2 or 5.0 or something, and I could swear that KDE/gnome always had a way to turn it off. But then again, now that I think about it, it may have been a recent release of kdm/gdm that put shutdown on the dm menu, and it seems kde/gnome can only shut down when their own dm is running... I don't think showing people that they can put a shortcut to shutdown on their desktop is going to impress anyone. Win=2K has a shutdown command, same format as unix. You could make a shortcut to that. If linux didn't used to have a built-in shutdown button, then that really doesn't show how great it is now, it just shows how long it took it to catch up to windows and mac. I know we all like to think linux is great for everyone and that everyone who doesn't love command prompts is probably stupid or something, but seriously, the lack of a shutdown button in the past just shows how far behind linux was in the eyes of a regular user. Really its these little details that really matter. Most of linux catching up to windows (besides program/industry deveolper support) are these little things. Things that you could program in python in like 5 minutes if you stopped and thought about what was really missing. Its kind of too bad a lot of linux developers are hardcore hackers and not average shmoes. Regarding traffic, have you ever tried to drive I-465 North on the west side in the airport expressway, I-70, 10th st. area? It's been really bad recently since 65 and 70 are closed downtown right now! :) --MonMotha I'm glad LA, much like NY tends to do all its road construction between 11PM and 5AM. People complain about LA traffic, but in the areas I travel, the traffic is so much better than Hawaii. I cruise at 75 miles per hour every morning on my way to work without any wories of tickets. -Eric Hattemer
[luau] Electronics design program for linux
Anyone know of a good electronics design program for linux? Preferably free and preferably able to netlist and export to an FPGA? Berkeley had a thing called ptolemy, but I'm not sure what it is and can't compile it... -Eric Hattemer
Re: [luau] ls order changed?
That sure did fix it, thanks. Vince Hoang wrote: export LANG=C -Vince Phil Wrote: Its your LANG and LC_* env settings. I think setting LANG=C fixes it... but I don't remember. A google search will answer the question.
[luau] ls order changed?
RedHat and Mandrake recently changed the order in which they sort the file list in the 'ls' command. It used to be ASCII alphabetical (A-z), but now seems to be case and special character ignoring. I heard ls doesn't sort anything on its own by default, but uses the filesystem. There don't seem to be any alphabetic sorting options in 'ls', and none in tune2fs. Its different from what I'm used to, so I am bothered by it. Right now, it lists things in such a way like: .a b C d _e F .g .H Anyone know how to get it back to the old behavior? -Eric Hattemer
Re: [luau] Scrounging for parts
I once installed openBSD on a cd-less sun from my linux machine. I was working on debian, but my ram seems to be too broken to do it (my sun sucks and is broken in many ways). The install process is a bit of a pain, and takes a long time to master the approximately 6 different components you need to have running simultaneously on the server side, but it can be done, and you feel cool for doing it. Here is some info regarding the process: http://www.netbsd.org/Documentation/network/netboot/intro.sun.html http://www.tldp.org/HOWTO/Network-Install-HOWTO.html http://www.debian.org/releases/stable/sparc/ch-install-methods.en.html#s-install-tftp (refer to sections 4 and 5) -Eric Hattemer
Re: [luau] Administrivia: Reply-To header removed
I think the second article makes much more sense. There's only a few mail programs in the world that support reply to group functionality. I was looking at the archive on the internet and was impressed by how threaded it was. This will not continue if people have to go out of their way to make new messages that go just to luau. And reply to all (which is what most people are going to start using) is going to bug anyone that posted and doesn't want to go through the trouble to make some fancy duplicate mail filter. And yes, Warren, you are getting two copies of this mail in protest. The first message was written by someone who strangely assumed the whole world uses mail programs designed for mailing-lists, or would like to receive two copies of every message. I'm sure there's reply-to-group in elm and mutt and whatever, but there is no such function in OE, outlook, netscape, eudora, and I'm assuming AOL, which probably make up for 99% of American mail clients. I'm using mozilla right now, which has no such function. But if you do decide to change it, please tell someone. I'm sure a lot of people were thrown off that the reply button they had always hit no longer functions the way they expect. -Eric Hattemer Warren Togami wrote: http://www.unicom.com/pw/reply-to-harmful.html I have changed the setting as part of a test based upon the rationale within this page. http://www.metasystema.org/essays/reply-to-useful.mhtml This is some counter rationale as to why reply-to-list is good. I'll leave it this way for a few weeks and see how it goes. Basically if you want to reply to the list, use Reply All rather than Reply in your mailer. Warren Togami [EMAIL PROTECTED]
Re: [Fwd: Re: [luau] Group Policies for Linux]
And don't forget about FACLs, recently introduced into linux. You just do something like setfacl -rm user:eric:rwx /mp3; and user eric gets full access to /mp3 regardless of the user-group-all permissions. You'll need a new distro with a modern kernel, and maybe a package about facls to get setfacl and getfacl. -Eric Hattemer
Re: [Fwd: Re: [luau] Group Policies for Linux]
You are correct. It was added to the 2.5 kernels, and possibly the late 2.4.20 or so kernels. I've seen it here and there. I can't comment on where. It might be a patch redhat put in there, or a 2.5 thing, but facls were added to linux kernel in the last 6 months. You might be able to find utilities for them somewhere. -Eric Hattemer MonMotha wrote: Must have been REALLY recently, either that or I didn't hear about it... I know SGI XFS has POSIX ACL support, and grsecurity has something similar, but I never heard of FACLs... --MonMotha
Re: [luau] Hard drive clean up questions
Firstly, as far as I know, these only really matter for compiling kernel dependent sources. So you could so just as well to remove them all if you don't compile anything. But to be safe, its always nice to have some kernel headers. So go ahead and get rid of 2.2, then link linux to linux-2.4. Its becomes a crazy chain link, but it makes it easier if say you upgrade to 2.4.30 or something in the future. Of course, the most sensible way to do it is to move linux-2.2 somewhere weird like /root, then use the computer for a week and make sure nothing goes wrong. Then get rid of it entirely. Also, you can probably cd into both of those dirs and do a gmake clean;. That'll get rid of a lot of .o files that you probably don't need. You should be able to torch any files (not directories) under readhat. There are no libraries in the linux kernel directories, only header files. So once something is compiled and runable, it should theoretically not have any dependence on those anymore. But just to be safe, move 2.2, wait a while, then delete it. -Eric Hattemer Ben Beeson wrote: Aloha, I need to free up some space on my hard drive and I thought /usr/src might be a good place to start as it appears I have several versions of old kernel sources stored there. I need some advice on what is safe to delete without screwing up the system references to libraries etc.
Re: [luau] Best Linux Server?
This post kind of got lost and neglected. But in any case, I assume when you say raid-0 (striping), you really mean raid-1 (mirroring). I don't know much about servers, but I figure redhat would be a good bet. I like Mandrake much better for desktops, but redhat is concentrated on their server markets. Plus Warren is working on things for redhat right now that could make it a bunch better. As far as raid cards go, a lot of people like the 3-ware cards. I have no idea about their availability right now. Otherwise 2 hard drive channels on the motherboard and linux kernel raid does just fine. Make sure you set all the good parameters in /etc/sysconfig/harddisks, such as use DMA and maybe something like -X 68 (man hdparm). And you'll get good performance and reliability. But like has been said before, you won't get any protection from hacks or similar things, since raid backs up simultaneously, so you'll probably want something like a CD burner or if you want to go old-school, a tape drive. -Eric Hattemer message untrimmed because it seems to have been ignored earlier: Dustin Cross wrote: Aloha, What is the best distro choice for a simple web/webmail server. My SuSE 7.3 Sparc Ultra II server has been great for over two years, but I am starting to worry that I don't have any routine back-ups or disk mirroring. If anything happened to this server Sandi would be very unhappy. I am thinking of building an inexpensive x86 with cheap IDE raid 0 controller and two drives. Is there a solid distro that is easy to keep up to date, (i.e. automated)? I use SuSE's Online Update all the time now and it works okay. Do any of the other distros have anything better? What about Gentoo and its emerge? I don't have much time to spend keeping this system up-to-date and secure.Sadly my best security defense has been using sparc hardware, NO ONE writes exploits for Linux on Sparc hardware. If I switch to an inexpensive x86 systems this won't be the case anymore. So what does everyone reccomend? My currect systems is: SuSE 7.3, Apache, Postfix, Imap, Squirrel Mail, PHPGroupware, Spam Assassin, and Procmail What I am looking for is reccomendation on distro (SuSE, Redhat, Mandrake, Gentoo, etc) and recomendation for inexpensive IDE Raid 0 card that works with linux. Or maybe I should just buy a second 36GB SCA SCSI drive, rebuild my Sparc and use software raid to mirror the drives. What does everyone think about Gentoo for this? Is there any benefit to the inexpensive IDE raid cards over software raid? Mahalo, Dusty P.S. I would use my beloved OpenBSD for this but then I waste the second CPU in my Ultra II and I didn't see any inexpensive IDE raid cards supported under OpenBSD. ___ LUAU mailing list [EMAIL PROTECTED] http://videl.ics.hawaii.edu/mailman/listinfo/luau
Re: [luau] Personal Touch (was anyone know of a good computer repair place?)
You might also check to see that the cpu is attached to whatever heat sinking device its supposed to be. Laptop fans are often attached to a thermal probe so that they only spin when the system is getting too hot. So its possible that your cpu has become dislodged from its sink and is heating rapidly. Either that or there may be something wrong with the probe. Of course, considering that if the fan is louder than it used to be either implies it was never this hot, or that the fan is damaged. Personally, I've never taken apart a laptop. I'm not sure that you want to. But theoretically your machine should function without even having a hard drive attached. See if you can get into the bios, or if you can pull the hard drive out and boot off a CD. You still might want to get it professionally repaired, but I'm less liekly to think its a hard drive issue as a fan issue. -Eric Hattemer
Re: [luau] Linux Email Host-Off topic maybe
Perhaps your group should investigate other modes of file transfers such as scp, sftp, and centralized data servers. -Eric Hattemer [EMAIL PROTECTED] wrote: Hello people, Can you recommend a reliable Email Hosting services that provides 90-100 email accounts with basic HTML web hosting? My current host Interland has a limit on the size of emails per account, which takes a lot of energy to maintain. So looking for something that allows unlimited sized attachments as long as I don't break my total size limit for the account... Thanks, V
Re: [luau] what'd i miss?
Ho'ala Greevy wrote: Ocean's Night Club is a good place for these types of things, they have attractive profit-sharing for fundraiser activities. I think you're going to have to go into greater detail about this. I guess if its covered at the board meeting, that's one thing, but otherwise, holding a fundraiser at a nightclub seems like quite a strange idea to me. -Eric Hattemer
[luau] If you thought MSFT had a bad reputation, you're wrong
http://www.businesswire.com/cgi-bin/f_headline.cgi?bw.012803/230282058ticker=MDLK Yep, it turns out, the reputation institute ranks Microsoft as having the best corporate reputation in America. How this makes sense, I'm not sure, especially if you look at their criteria... Each company is scored based on how many positive and negative reputation --driving attributes are found within each story. These attributes are classified into the six dimensions of the Harris-Fombrun Reputation Quotient: emotional appeal, products and services, vision and leadership, workplace environment, social responsibility and financial performance. I think they can only qualify for financial performance, and perhaps workplace environment (although I've never been in their office, so I can't say about that one). They are doing well financially, as we all know, though. Yearly earnings per share of 1.74, price per earnings of 28.06, and of course their market cap or a mere $261 Billion. That's a thousand million for you kooky Britts. -Eric Hattemer
Re: [luau] I crashed SUSE 6.2 playing Shishensho :)
Well, unfortunately linux does crash from time to time. Its always been designed and focused on the business-unix-server types of markets. Linux will run on a big server with SCSI drives, 3com ethernet cards, and VESA VGA video under heavy load for hundreds of days without a crash or reboot. But all the newer kernel modules to support the latest in desktop home user peripherals aren't always tested so well. All the fancy new graphics drivers and openGL support and whatnot are a somewhat new advance, and they are still working on those. Nvidia cards are notorious for having problems with either the builtin drivers or the nvidia site drivers. I've had good experiences with them, but other people like to complain about them. It could just be faulty hardware on your computer, though. -Eric Hattemer On Tue, 2003-01-21 at 15:01, Karen Lofstrom wrote: Yesterday I dwaddled over my discrete math homework and played a game I found on my KDE desktop obsessively between intervals of math. I was extremely *surprised* when the computer started responding slowly, screen flickering, redraw faltering, and then finally crashed. I should have saved a copy of the error screen, I really should. Instead I just rebooted and it seems to have been working fine since. Probably something to do with the video card driver, which I didn't install, my brother did. I am bemused. I thought Linux didn't do that :) No big problem, unless it starts happening frequently.
Re: [luau] Newbie on Redhat 5.2
That's usually not true, and if it were, just leave it on overnight. I remember in the dorms at USC, I once downloaded 2 linux CDs simultaneously in 10 minutes total. Of course, that was at a good 1.3MB/s. To Warren's local sites, you can probably get nearly 300K/s, which is probably enough to download all 3 CDs (usually you only really need the first two) in under 2 hours. More likely 1 hour or less if you're not in peak time. But then I'm out of state, so I couldn't speed test. I remember getting 800K/s from a roadrunner site once, though. -Eric Hattemer On Mon, 2003-01-20 at 13:43, demon_jr808 wrote: Thanks for the links Warren! I was looking to download Redhat 8, but even with RoadRunner it would take around 5 hours + per disk. Thanks again. --- Outgoing mail is certified Virus Free. Checked by AVG anti-virus system (http://www.grisoft.com). Version: 6.0.443 / Virus Database: 248 - Release Date: 1/10/2003 ___ LUAU mailing list [EMAIL PROTECTED] http://videl.ics.hawaii.edu/mailman/listinfo/luau
Re: [luau] dhcp-146-41
I concur 100%. There's an IP standard of sorts that says that all IP addresses should have a hostname attached. There are many services that do forward and reverse DNS lookups on your IP/hostname to make sure they match. One such service is some ssh servers. If you didn't have a hostname (and some ISP's are broken enough not to assign them), then certain servers won't allow you connections into them. Now maybe you're worried that your IP doesn't change too much. Now to most people, this is wonderful. People pay lots and lots of money for static IP addresses. With any of the dynamic dns services available, you can get your own DNS name, and it will be correct over 99% of the time. You could even buy a real domain name, and just update it every time it changes. That takes a day or two, but since it only changes every couple of months, you could probably actually make a webserver on there and everything (Oceanic wouldn't be too thrilled if they found out, though). Now apparently you consider it a security risk. However, if someone were to randomly pick your IP to attack, it wouldn't really matter whether it never changed or changed everyday. They'd have an equal chance of randomly picking your address. Of course its possible that people could track you and whatnot, but considering that theoretically Oceanic is the only one who has any idea of your real identity, this should not be an issue. However, if you really want to change your IP address, you could probably do so by unplugging your cable modem for long periods of time. Hopefully it will eventually grab another IP because someone took yours. That's the way it works on many other networks, anyway. Otherwise, the oly other input I have is to make sure that your domain and your default dns suffixes search order are set to hawaii.rr.com, and that might help fix your problems. -Eric Hattemer - Original Message - From: Dustin Cross [EMAIL PROTECTED] To: [EMAIL PROTECTED] Sent: Tuesday, December 24, 2002 7:29 PM Subject: Re: [luau] dhcp-146-41 Wayne, You do not have a permenant hostname. The IP address has a permenant hostname in DNS. You have a temporary IP address that is almost never changed. Would you please go read about IP addresses, dhcp, and DNS so that you can speak inteligbly on the subject and stop trying to make something into nothing. Here is a place for you to start: http://www.linux.org/docs/ldp/howto/Net-HOWTO/index.html Dusty I believe many cable modem users will not be thrilled when they realize that their computers, instead of having a changing ip, are actually assigned a permanent hostname. Of course, our local dhcp experts are telling me that having a permanent hostname don't make no difference, and are implicitly accusing me of spreading FUD. You make your own call. ___ LUAU mailing list [EMAIL PROTECTED] http://videl.ics.hawaii.edu/mailman/listinfo/luau
Re: [luau] Help deciphering portscan entries in /var/log/messages
ethereal is really good at filtering tcpdump. Try that if you decide to look into the matter. -Eric Hattemer On Thu, 2002-12-12 at 11:05, Vince Hoang wrote: On Wed, Dec 11, 2002 at 10:44:30PM -1000, Ben Beeson wrote: I'm curious why the log shows two IP addresses that aren't mine. Is that because snort is seeing the traffic pass my external interface, or is something else up? Your external interface is probably sharing the same collision or broadcast domain. If listen with tcpdump and filter out your IP, you will probably see more foreign packets. -Vince ___ LUAU mailing list [EMAIL PROTECTED] http://videl.ics.hawaii.edu/mailman/listinfo/luau -- Eric Hattemer [EMAIL PROTECTED]
Re: [luau] Help deciphering portscan entries in /var/log/messages
Honestly, I know next to nothing about networking and try hard not to learn any more than I absolutely need in life. So I hope someone else answers this question properly. I think it could be possible that its trying to use your system as a router to get to the 12.x device. Either that or its some tricky stealth scan that routes to you, but says its routing to another address in order to trick you. But yeah, I hope someone else answers this with some expertise. -Eric Hattemer On Thu, 2002-12-12 at 21:37, Ben Beeson wrote: Eric and Vince, So essentially, snort is seeing a lot of other traffic on my external interface that is not necessarily destined for my box. Is that right??? Mahalos, Ben ___ LUAU mailing list [EMAIL PROTECTED] http://videl.ics.hawaii.edu/mailman/listinfo/luau -- Eric Hattemer [EMAIL PROTECTED]
Re: [luau] M$ thinking tux??
Microsoft faced a similar situation a decade ago when its nascent server software was competing head-to-head with market leader Sun Microsystems Inc. , but Microsoft did not choose at the time to write software for Sun's proprietary version of Unix. Anyone else ever used internet explorer and windows media player in solaris, or is it just me? -Eric Hattemer On Tue, 2002-12-10 at 14:12, Beeson Benjamin Lt Col 715 AMOG/DO wrote: Aloha, I saw this today on CNN. I'm not sure it's good news yet http://www.cnn.com/2002/TECH/biztech/12/10/meta.linux.reut/index.html VR, Ben Beeson -- Eric Hattemer [EMAIL PROTECTED]
Re: [luau] Sparc 20 upgrade time
This all can be quite painful. After sucessfully doing this once, it took me several hours a few months later to get it set up again. Then NFS ate my filesystem. But if you need help with this, I can get back to you over the weekend or so. I have way too much homework to explain this all. But look at the netbsd howto for this even if you're not using netBSD. Its quite helpful. -Eric Hattemer On Wed, 2002-12-04 at 14:19, Bob wrote: server RARP boot image on server boot net It's been years, but I muddled through this at one time If you can't find a CDROM locally, let me know. I may have one in the edebris pile... please reply my [EMAIL PROTECTED] address for best response -tongaloa aka Bob Ben Beeson wrote: install? What is the best course short of buying a drive or two? Thanks, Ben ___ LUAU mailing list [EMAIL PROTECTED] http://videl.ics.hawaii.edu/mailman/listinfo/luau -- Eric Hattemer [EMAIL PROTECTED]
Re: [luau] ALSA Nvidia Drivers on Red Hat 8.0
On Tue, 2002-12-03 at 09:13, W. Wayne Liauh wrote: Has anyone ever installed ALSA Nvidia Drivers in Red Hat 8.0? Probably. I know I have both on mandrake. Do you have a specific problem? I'm a bit tired of explaining Nvidia drivers, there's about 10 messages from me about it. Look back a bit, or in the archive assuming we have one. As for ALSA, Warren wrote a little how-to deal about it. -- Eric Hattemer [EMAIL PROTECTED]
Re: [luau] Lower temperatures for Athlon in Linux
Has anyone every run into an nforce compatible version? -Eric Hattemer On Fri, 2002-11-29 at 17:03, Warren Togami wrote: http://mpet.freeservers.com/LVCool.html This program will reduce temperatures of your Athlon processor on most VIA chipsets. Read the documentation to learn how it works. -- Eric Hattemer [EMAIL PROTECTED]
Re: [luau] Videl FTP
Well, OpenBSD is usually the most solid of web servers. But its a bit archaic looking. I really do love mandrake, though. For packaging, the first thing you do is set up your urpmi database. First, find a webserver that has a mandrake mirror. Then, find its rpms. An example is ftp://videl.ics.hawaii.edu/mirrors/mandrake/Mandrake/9.0/i586/Mandrake/RPMS/ . Then find its hdlist.cz file, the file that contains the listing and info for the rpm files. Its almost always ../base/hdlist.cz. On some servers, you'll have a Mandrake/RPMS2, which should be connected to ../base/hdlist2.cz. Videl seems to be set up a little non-standard about this. CD 2 is ftp://videl.ics.hawaii.edu/mirrors/mandrake/Mandrake/9.0/contrib/RPMS/ and so the relative path to hdlist2.cz is ../../i586/Mandrake/base/hdlist2.cz , which is kind of a hassle. But basically, if you want to add videl to your list, do a urpmi.addmedia Videl1 \ ftp://videl.ics.hawaii.edu/mirrors/mandrake/Mandrake/9.0/i586/Mandrake/RPMS/\ with ../base/hdlist.cz urpmi.addmedia Videl2 \ ftp://videl.ics.hawaii.edu/mirrors/mandrake/Mandrake/9.0/contrib/RPMS/\ with ../../i586/Mandrake/base/hdlist2.cz Now you have Videl1 as an official mandrake package source, and Videl2 as a contrib package source. Every now and then (every day for cooker sources), do an urpmi.update --auto, and that'll redownload the lists. If you want to upgrade every package, do urpmi --auto-select. Otherwise just do urpmi package-name. It doesn't care whether you put the version, etc, unless you want a specific version. If you use zsh (and you should), then you get tab completion with urpmi. If you type urpmi XFtab, it gives you a list of all the XFree86 stuff along with XFDrake (Or something like that). If you accidentally add a slow mirror, you can either comment it out of the file (something like /etc/urpmi/sources.list), or go into the software sources manager and uncheck it. No need to delete it entirely, it might be useful someday. -Eric Hattemer On Tue, 2002-11-26 at 19:51, eXt wrote: 1- Is Suse 8.1 or the latest Slackware in the Videl server? 2- If not, can someone tell me a good d/l mirror? ( i tried 5 last night but the dl speed was 5-10 K) 3- Anyone knows if mandrake 9 installs iptables by default? if not, what do i need to do to get them and install them? 4- What is the best distro for a Commercial web-server (i know mandrake is newbie friendly...but i'm having more headaches trying to install stuff on it because of this) __ Join Excite! - http://www.excite.com The most personalized portal on the Web! -- Eric Hattemer [EMAIL PROTECTED]
Re: [luau] Chinese Mandrake 9.1
I'm not sure if there is some confusion about this, but the official American version number for cooker is in fact 9.1. [EMAIL PROTECTED] ~ # rpm -qi mandrake-release Name: mandrake-release Relocations: (not relocateable) Version : 9.1 Vendor: MandrakeSoft Release : 0.1mdkBuild Date: Tue 08 Oct 2002 09:47:17 AM PDT Install date: Thu 10 Oct 2002 11:14:48 PM PDT Build Host: ke.mandrakesoft.com Group : System/Configuration/OtherSource RPM: mandrake-release-9.1-0.1mdk.src.rpm Size: 45 License: GPL Packager: Warly [EMAIL PROTECTED] URL : http://www.linuxmandrake.com Summary : Mandrake Linux release file Description : Mandrake Linux release file. -Eric Hattemer On Wed, 2002-11-27 at 20:47, W. Wayne Liauh wrote: During my recent trip to Taipei and Osaka, I noticed that most if not all of the home-grown distributions have mysteriously disappeared. The only distributions I can find are localized versions of Red Hat and Mandrake. Of course, there are always players in the Debian camp. China seems to be following the same trend. No one seems to be interested in their own Red Flag Linux no more. Some Chinese developers even poked into Cooker and put together a Mandrake 9.1. http://www.linuxfans.org/nuke/modules.php?set_albumName=album17op=modloadname=galleryfile=indexinclude=view_album.phppage=1 ___ LUAU mailing list [EMAIL PROTECTED] http://videl.ics.hawaii.edu/mailman/listinfo/luau -- Eric Hattemer [EMAIL PROTECTED]
Re: [luau] New linux supercomputer
Kind of reminds me of my friend, hpc.usc.edu http://www.usc.edu/hpcc/hpc.html They haven't updated the info page, in a while, and I don't think they've benchmarked it lately, but its a pretty cool cluster (well, it runs about 20-30F hotter than the room its in). But the coolest thing about this cluster is that the nodes are connected by a thing called myrinet, which is an extraordinarily cool and probably expensive type of switch that has 1GBit bandwidth between each node to any other node through any path. Unfortunately, they make use of it pretty well, and wouldn't let me mess around with it. Stuff like computational chemistry, biology, geology (earthquakes). -Eric Hattemer On Mon, 2002-11-25 at 01:03, Brandon Jasper wrote: I saw an article on this and thought it was cool enough to post on the group. http://space-simulator.lanl.gov It is a 294 processor cluster built with Shuttle computers. They have a pic on the site, pretty cool... I want one! ___ LUAU mailing list [EMAIL PROTECTED] http://videl.ics.hawaii.edu/mailman/listinfo/luau -- Eric Hattemer [EMAIL PROTECTED]
Re: [luau] Dual Booting with WinXP Pro
On Thu, 2002-11-07 at 23:13, Brandon Jasper wrote: Windows will generally try to install its own bootloader NTLDR in the MBR, NT and 2k does so I expect XP would do so as well. Yeah, winxp almost certainly will rewrite the MBR. On 11/7/02 7:05 PM, W. Wayne Liauh [EMAIL PROTECTED] wrote: Thanks. I think you just answered the first part of my question, in that I should be able to install WinXP Pro in any of the three primary partitions. If this is not correct, please let me know. (It should be noted that although textbooks say that there are four primary partitions, since the fourth one will be used to create extended partitions, there are only three primary partitions in which to install a bootable partition. Also, since Win98 does not use MBR, I am pretty sure that you must install Win98 in the first primary partition, unless you install a multiple booting tool first, such as Norton something.) I'm not sure I agree with you here... See, first, a lot of the time you don't need extended partitions. I try to avoid them at all costs. I have two hard drives, so that helps a bit, but you could easily organize it as: Linux Swap Boot Windows Linux And have two OS on 4 partitions. Now if you want a third OS, you can replace the boot partition. As far as overwriting the MBR goes, what you can do is put LILO onto the the linux/boot partition instead of the MBR. This sets the MBR to chain load lilo. Then when XP writes over the MBR, it doesn't write over lilo. All you need to do now is go into NT fdisk (control pannel/administrative tools/computer management), and set the linux/boot partition as the only active partition. This will restore the situation with no need for a boot disk. But yes, probably the best method would be to think long and hard about how you want the partitions to be. Then, even though the NT and linux partitioning programs have similar functionallity, I prefer to use the linux one. So boot a linux CD, run fdisk, set up the partitions the way you like them. Save that and reboot with the win CD. You probably will want at this point to remove and put back the windows partition just so that windows formats it properly. Windows NT+ does not need to be installed to the first partition. Then once that's finished and happy, install the linux to the leftover partitions. -Eric Hattemer
Re: [luau] Ghost for Linux and Pattition Image
Yeah, we're similar at my workplace in a lot of ways. We only have about 6 windows, 6 macs, and 4 U10's. But we get administrative rights to the windows machines. This is basically in case we need to install something temporarily to help someone with it. Since we all have keys to the building, and we're all students, sometimes due to the irresponsibility of some people, the machines end up with a bunch of junk like kazaa on them. But rather than taking away administrative rights or anything like that, Andrew, who is in charge of the computers keeps a clean ghost image, and just wipes the machines back to their previous states pretty much every week. Then he does have some script where you need to enter in the computer name or something. The multicasting seems to help when he needs to download 3 or more of them at a time, but I have no idea of how much effort it took for him to set that up. But yeah, since we're open for walk-ins from 9-5, but open for calls until 7, 5-7 is usually game night, and 7:00 is time to download the machine again. Its a much cleaner solution than trying to uninstall the program. Our systems work great all the time. (If they ever start to act funny, we can always download them again, and they should be fine). Andrew often argues with the userlab staff because he really likes his ghosts, and they really like their odd distributed .msi system. -Eric Hattemer On Tue, 2002-11-05 at 08:52, Jeff Zidek wrote: I don't dispute what you said about Norton Ghost except one thing. The max of eight workstations. I have used Ghost to multicast 64 workstations at once over a weekend. We started them on Friday and left and came back to 64 new workstaions on Monday. This was on Windows 2000 systems though. To deal with the fact that it makes them (ununique systems) I run sysprep before I make the image shutdown without letting it reboot and reboot onto ghost multicast bootdisk. If anyone is interested I have a trick so that when the 2000 systems boot all you have to do is enter a new computer name click next then next again and your done. To do this just use the XP setup manager to make your sysprep text file and run it under 2000 sysprep. Seem it's just that the setup manager in 2000 doesn't have all the available features in the XP version. Don't use the sysprep program from XP just the setup manager. Works like a dream. Won't even have to click that pesky EULA or enter a Product Activation Number. Jeff Zidek
Re: [luau] MDK 9 NVIDIA driver installation woes
You can try other video modes like 0317, but I had another idea that you could try the riva module in the kernel. In mandrake its a module, and not built into the kernel. You can try it with modprobe rivafb. Otherwise, you can try building it into a kernel and start it with video=riva:mode:[EMAIL PROTECTED] or something like that. Personally, I can't get this to work. If someone knows about that, it'd help me out a lot. But you can experiment with it. -Eric Hattemer
Re: [luau] MDK 9 NVIDIA driver installation woes
When you say no luck, does it still lock up? I don't know anything about the GF4, but I've never had serious problems installing these drivers. The default X drivers will hardlock any of the newer GF cards, but the nvidia drivers shouldn't. Was it working properly in the framebuffer mode after the install? I'll send you my XF86Config-4 file. Make sure that the -4 file and the XF86Config file are linked together. This shouldn't be a big problem on mandrake, but since redhat is really broken in this way, just check it anyway. ie: ll /etc/X11/X* lrwxrwxrwx1 root root 27 Oct 12 16:12 /etc/X11/X - ../../usr/X11R6/bin/XFree86* -rw-r--r--1 root root 2743 Oct 12 16:13 /etc/X11/XF86Config-4 -rw-r--r--1 root root 2743 Oct 12 16:12 /etc/X11/XF86Config-4~ -rw-r--r--1 root root 2739 Oct 12 16:11 /etc/X11/XF86Config-4.old -rw-r--r--1 root root 955 Sep 25 12:40 /etc/X11/XF86Config.old -rw-r--r--1 root root13766 Oct 21 23:51 /etc/X11/XftConfig -rw-r--r--1 root root 275 Sep 10 00:15 /etc/X11/Xmodmap -rwxr-xr-x1 root root 2008 Sep 10 00:15 /etc/X11/Xresources* -rwxr-xr-x1 root root 4800 Sep 10 00:15 /etc/X11/Xsession* In my case, I don't have an XF86Config file at all. You shouldn't need one. But if you have one, back it up, and then delete it. All you need is the XF86Config-4 file. Also, be aware that the kernel module won't always show up in lsmod because its autocleaned every now and then. But right when you boot, do an lsmod and look for NVDriver. -Eric Hattemer On Sun, 2002-10-27 at 13:08, Rodney Kanno wrote: Rodney, IIRC, you must also edit you XF86Config-4 to specify the new NVdriver, instead of the default nv one. Should be in the Graphics device section, under section Device. Hope it works, and would be interested if it does, since I may attempt the same thing soon. I tried changing that as well, but no luck yet Rodney --- Outgoing mail is certified Virus Free. Checked by AVG anti-virus system (http://www.grisoft.com). Version: 6.0.408 / Virus Database: 230 - Release Date: 10/24/2002 ___ LUAU mailing list [EMAIL PROTECTED] http://videl.ics.hawaii.edu/mailman/listinfo/luau
