Re: [gentoo-user] Command Line Mail Clients?
On Wed, Jan 28, 2004 at 04:20:31PM -0800, Anupam Kapoor wrote: > > emacs -nw -f gnus > but that's not the same as mail now is it ? gnus will do both news and mail. I assume it was originally meant to be just a news client, but there's plenty of documentation describing how to use it as your mail client as well. mh (and nmh) are other command-line mail programs. I'm not sure if they support gnupg or not. Personally, I use mutt. Matt -- Matt Garman email at: http://raw-sewage.net/index.php?file=email -- [EMAIL PROTECTED] mailing list
Re: [gentoo-user] office (or at least excel)
On Fri, Jan 30, 2004 at 07:20:00PM +0100, Meka[ni] wrote: > Is there some kind of office that is not so huge like Open Office? I > actually need Excel, but if there is an office that is small enough > (source not above 20mb) I would like to test it. Please anyone. I need > it for some test on the university. Thanx. :o) Check out gnumeric, it's a standalone spreadsheet program. I'm not sure if it's under 20 MB or not, since it requires the GNOME library. Hope that helps, Matt -- Matt Garman email at: http://raw-sewage.net/index.php?file=email -- [EMAIL PROTECTED] mailing list
[gentoo-user] i2c device creation?
I emerged i2c and lm-sensors hoping to get some temperature readings on my hardware. However, when I run "sensors-detect", it says the following: No i2c device files found. Use prog/mkdev/mkdev.sh to create them. I presumed that the ebuild would automatically create the i2c devices, but apparently not. So I extracted the archive /usr/portage/distfiles/i2c-2.8.1.tar.gz. The prog directory is empty, though (even after doing a make). Does anyone happen to know how I can get these i2c devices created? Thanks, Matt -- Matt Garman email at: http://raw-sewage.net/index.php?file=email -- [EMAIL PROTECTED] mailing list
[gentoo-user] openbsd partitions, grub
I have two hard drives in my system, a SCSI drive from which the OS runs, and a big IDE drive that I use for backups. I recently converted the system from OpenBSD to Gentoo. I didn't "touch" the IDE drive during the install, except during grub configuration. Since it can be somewhat confusing to distinguish between SCSI and IDE drives in grub, I accidently ran the grub "setup" command on my IDE drive. Now it appears as though there are *no* partitions on the IDE drive (it used to have four FFS partitions). No partitions are reported at boot time, and none are shown when I print the partition table from fdisk. Plus, I have no /dev/hda[1-4] devices. What are the chances I ruined the partition table of the IDE drive while I was playing with grub? Is there anything else I can do to further diagnose the problem and/or get at the data on that drive? Thanks! Matt -- Matt Garman email at: http://raw-sewage.net/index.php?file=email -- [EMAIL PROTECTED] mailing list
Re: [gentoo-user] i2c device creation?
On Wed, Feb 04, 2004 at 12:10:30PM -0500, Alex Nelson wrote: > Matt Garman wrote: > >However, when I run "sensors-detect", it says the following: > > > >No i2c device files found. Use prog/mkdev/mkdev.sh to create them. > > If you are running a 2.6.x kernel, make sure you have the i2c-dev > module loaded. That one caught me for a while too. I'm running kernel version 2.4.22-gentoo-r5. But, for what it's worth, I did load the i2c-dev module (to no avail). Thanks! Matt -- Matt Garman email at: http://raw-sewage.net/index.php?file=email -- [EMAIL PROTECTED] mailing list
[gentoo-user] sensors -s: can't access /proc file for writing
I'm using the gentoo-sources kernel version 2.4.22-gentoo-r5. I believe it has a patch that breaks the "sensors -s" functionality. See support ticket #1564 on the lm_sensors page: http://www2.lm-sensors.nu/~lm78/readticket.cgi?ticket=1564 Essentially, when I try to run "sensors -s", even as root, I get the following message: it87-isa-0290: Can´t access /proc file for writing; Run as root? On the suggestion of the lm_sensors support team, I compiled a vanilla 2.4.24 kernel. Under this kernel, I was able to run sensors -s. The lm_sensors folks are guessing that the problem is caused by one of the gentoo kernel patches. For what it's worth, the problem also seems to plague users using a Debian kernel. Is there anywhere I can get a list of patches applied to the gentoo-sources kernel? Has anyone else experienced similar problems, and perhaps figured out what is specifically causing the problem? Thanks! Matt -- Matt Garman email at: http://raw-sewage.net/index.php?file=email -- [EMAIL PROTECTED] mailing list
Re: [gentoo-user] bulloney 2.6.X benchmarks
On Thu, Feb 12, 2004 at 02:23:37AM +0600, Grendel wrote: > Well one possibility is that if you have ACPI turned on it can > drastically slow down your linux box. so disable acpi and rerun your > ... > disable ACPI the 3d performance under 2.4 was much faster than 2.6. I've seen a few references to the ACPI situation now, but haven't bothered to get the full story. Does anyone happen to have a good link discussing the ACPI? When you suggest disabling ACPI, do you mean in the Linux kernel or in the BIOS? Thanks, Matt -- Matt Garman email at: http://raw-sewage.net/index.php?file=email -- [EMAIL PROTECTED] mailing list
Re: [gentoo-user] How can I open my port 80?
On Wed, Feb 11, 2004 at 11:36:49PM +0200, Nickolay Savchenko wrote: > I've just installed mini_httpd web-server on my home computer, and it > works, but there is a problem. It doesn't listen either to the 80 or > to the 8080 port. (Remote portscanner shows "stealth status"). I've > flushed all my firewall rules, stopped unneeded network services, > tried to change webserver (tried boa and apache), but my port 80 > still isn't opened. Are you using DSL or a cablemodem? There's a chance that your ISP is blocking some or all of your inbound ports. You might try starting other services (sshd, for example) to see if you can connect to that. I don't know about mini_httpd, but many applications can be compiled with tcpd (tcp wrappers), in which case they'll look at /etc/hosts.allow, /etc/hosts.deny (I think) to see who can and cannot connect to that service. Hope that helps! Matt -- Matt Garman email at: http://raw-sewage.net/index.php?file=email -- [EMAIL PROTECTED] mailing list
Re: [gentoo-user] OT: Win NT & 2k
On Fri, Feb 13, 2004 at 12:34:32AM +0100, Spider wrote: > other technologies that can be seen as "infringing" on some of MS > turf. Its an easy, blunt lega weapon to weild against whole projects, > stating that all progress past this point is only because you copied > MS sources, weilding a large batch of patents, tradesecrets and > copyright infringement claims and slamming large and wide. What scares me, truly, is that it only takes one bonehead to do a copy and paste job from the MS code to a (previously) legitimate OSS application. And, as you say, that gives Microsoft, with their vast financial and legal resources, an opportunity to start doing a lot of damage to the OSS world. I'm probably overly paranoid, but part of me is thinking about a conspiracy. Even if it's not a conspiracy, you've got to admit that Microsoft is pretty good at taking advantage of circumstances, and working things to their favor. I'm sure someone there will figure out a way to use this "opportunity" to kill Linux---as many have suggested, Microsoft's (and SCO's) only way to beat OSS is through manipulation of the legal system. I hope this isn't the first step towards the criminalization of Linux. I've often thought that the best way to get a lot of converts to Linux is for OSS to offer the next "killer app". My vision of the next killer app is highly integrated, "ubiquitous security"---systems that are by default completely secure, but the security implementation and infrastructure is transparent to the users; systems that are *by design* inherantly secure, so they can be open as well (kinda like "open DRM"). I think a component of such a system would be unmistakeable, impossible-to-forge digital audit trails and authentication systems. This would serve a huge purpose for OSS: accountability, and and easy means to verify source code (who made it, where it came from, etc, etc). The intent is to help OSS "prove" that it is legitimate, to avoid SCO-like fiascos. Much of that infrastructure or capability probably already exists, but it's extra work, and not necessarily trivial to use. For an OSS project manager to keep explicit track of every code submission (who, when, where, etc) would (1) take a lot of fun out of the development process, and (2) slow down the development process. But if such a system were fully automated, easy to use---ubiquitous---it would be easy to add the kind of accountability OSS needs to combat the naysayers (and anti-OSS FUD). Sorry, I'm starting to rant/ramble/daydream :) Matt -- Matt Garman email at: http://raw-sewage.net/index.php?file=email -- [EMAIL PROTECTED] mailing list
Re: [gentoo-user] OT: Win NT & 2k
On Fri, Feb 13, 2004 at 02:02:54PM -0800, Eric Paynter wrote: > Matt Garman said: > > This would serve a huge purpose for OSS: accountability, and and > > easy means to verify source code (who made it, where it came from, > > etc, etc). The intent is to help OSS "prove" that it is > > legitimate, to avoid SCO-like fiascos. > [...] > > But if such a system were fully automated, easy to > > use---ubiquitous---it would be easy to add the kind of > > accountability OSS needs to combat the naysayers (and anti-OSS > > FUD). > > CVS? Yeah, that's kind of what I'm thinking about, but with a much more rigorous authentication/validation/verification system. Plus, on smaller projects, it may not be worth setting up CVS. Even with CVS, administrators might play loose with permissions, allowing easy corruption of the source. I know those are effectively user problems, which can never be avoided, but that's the point I was trying to make: that this system has security so deeply imbedded and integrated that laziness, carelessness, maliciousness and human error become non-issues. The ultimate goal is for the legitimacy of all open source software to never come into question. I think some of the infrastructure is already there. If all OSS projects used CVS, plus digitally-signed all code with public key encryption, then we'd have a nice collection of "verifyably authentic" code. The legitimacy of anything that is not digitally signed is uncertain. Unfortunately, it's easier said than done (and my thoughts are certainly over simplifications). If I won a bunch of money and could quit my job, I think it would be a fun concept to research and try to develop. It's fun to think about anyway :) Have a good weekend! Matt -- Matt Garman email at: http://raw-sewage.net/index.php?file=email -- [EMAIL PROTECTED] mailing list
Re: [gentoo-user] OT: just a comment
On Mon, Feb 23, 2004 at 11:56:20PM -0800, Sami Samhuri wrote: > I haven't run into any problems that someone else didn't have already. > In my experience if you spend long enough on google, you'll find the > answer. It may come around as parts of different problems or fixes but There's a boatload of great information on the web. Proficiency in web searching is an invaluable skill for any computer user (Linux or Windows). These days, whenever I have a problem, I usually spend more time looking for the answer on the web than I do actually fixing the problem. Whenever I get stuck, my first thought is "someone has been here before". It's actually somewhat of a crutch---I often search the web before I even read the docs! I have to laugh at myself whenever I need help, search google groups or mailing list archives and find someone who's asked the same question, only to be given a bunch of "RTFM" responses. > it's always there. I bet you'll think more clearly tomorrow and trust > me... there's no going back to Windows. Yup, everybody gets frustrated once in a while. Sometimes it helps to rant and vent. The thing to keep in mind, as many have pointed out, is that most things in the Linux world are done on a volunteer basis. It's really hard to keep that in perspective when you consider how much good software has been created in the OSS world; it's easy to start thinking of it as "free beer" as opposed to "free speech". (I'm guilty of this from time to time.) Many anti-OSS folks used to (still do?) rant about lack of support. Granted, you have to know where to look (web, google groups, mailing lists), but once you do, the support for a lot of OSS is quite impressive. When you take into consideration that a lot of this is hobby/volunteer work, I personally think it's astounding how robust the support is. I've seen OSS projects that have more structured support facilities and bug tracking than I have at work. Just my thoughts (this is my "feel good" message for the day :) Matt -- Matt Garman email at: http://raw-sewage.net/index.php?file=email -- [EMAIL PROTECTED] mailing list
[gentoo-user] can't get rw permissions on nfs mount
I'm trying to export an NFS mount for my LAN. Here's the relevant line of my /etc/exports file: /video192.168.1.1/255.255.0.0(rw,sync) Now, when I mount this share from another PC, I can't create any new files on it, even though mount reports the mount as being mounted "rw": > mount 192.168.1.1:/video on /video/septictank type nfs (rw,addr=192.168.1.1) Even as root I cannot create any files. In fact, I can only cd one directory below /video. On the NFS server, I set all my directory (and sub-directory) permissions to 777, yet that doesn't change anything. I have other NFS shares exported that work as I expect. The only that is different about this one is that it's an XFS filesystem. Can XFS filesystems not be used with NFS? I was under the impression that the whole point of NFS was to be a filesystem abstraction. Any hints would be appreciated! Thank you, Matt -- Matt Garman email at: http://raw-sewage.net/index.php?file=email -- gentoo-user@gentoo.org mailing list
[gentoo-user] mysterious lockups
Since doing my last 'emerge -vuD world' a few days ago, my system has locked up two or three times. And it hasn't locked up while I was using it---I step away for a while, come back, and the screen is blank (but the monitor is *not* in power save mode), keyboard/mouse interaction doesn't do anything; also, the computer will respond to pings, but trying to ssh into the computer results in a timeout. Now, I know this is about the most vague question I can ask, but does anyone have any suggestions on likely places to start looking? I haven't changed any hardware, so I'm guessing that it's a software issue. I have no idea what is causing the lockup. If I had to make a random guess, I would guess it's the nvidia drivers (nvidia-kernel and nvidia-glx version 1.0.6629-r1). But that's just a totally wild guess :) Anyone have any other ideas? Is there a good diagnostic tool for catching something as nebulous as this? Thank you for any ideas or suggestions, Matt -- Matt Garman email at: http://raw-sewage.net/index.php?file=email -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Which Graphics Card?
Regarding graphics cards, what about the Matrox Millenium series? At least at one time, those were considered *the* best cards to have for use with Linux. Looking at Matrox's driver page, and a bit of google'ing makes it seem as though the Millenium g550 is the latest generation card from Matrox with complete open source drivers. However, I can't seem to find any information on 3D performance of the g550; I know 3D has never been the reason to buy Matrox, but as a non-gamer, I really don't care. I just want to be able to run the OpenGL xscreensaver hacks without using CPU :) So does anyone out there have a Matrox g550, who can comment on it's hardware 3D support? Thanks, Matt -- Matt Garman email at: http://raw-sewage.net/index.php?file=email -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] Filesystem Choice
On Thu, Feb 10, 2005 at 11:21:11AM -0800, David Busby wrote: > I don't want this to turn into one of those ext3 vs xfs > conversations. It kind of turned into a "filesystem versus database" conversation! > I want to know which filesystem is a better choice for having lots > (>4billion) directories. The dirs will be in tree format, so at > the root will be 256 dirs, each with 256 sub-dirs, each with 256 > ... If you have the time, you could probably create some benchmarks/tests pretty easily. For filesystems, I think you could use the bonnie++ program and create a simple shell/perl/python script that reflects the behavior/requirements of the system you intend to implement. You could probably also do the same with some basic SQL scripting in a database. I once saw a nice webpage that compared benchmark results of the various filesystems available for Linux (but I don't remember the link). But I don't think the benchmarks took your specific (and heavy-duty) application into consideration. I'm sure a lot of folks ould enjoy seeing such benchmark results as well, especially if you did a similar comparison to a database. In short: benchmarking could be a really interesting study if you have the time! Matt -- Matt Garman email at: http://raw-sewage.net/index.php?file=email -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] OT software to block IPs automatically?
The original poster was requesting a way to automatically block suspicious IPs. Lots of good responses. Another idea, and I've only read about this (no actual experience), but may be worth looking into: "port knocking". The basic concept is that you would keep your ssh port closed *all* the time. You need a secret "knock" to open the port. The knocking method is achieved by pinging various ports in a specific order (and with specific timing). So basically, before you can connect to port 22, you may have to ping ports 302, 50, 17, 17, 22, 542, 1002, 98, 12. The server will recognize the sequence and open port 22. Like I said, I don't have any firsthand experience with such a tool, but I've always thought it sounds incredibly clever. Maybe someone around here has some experience with port knocking and can offer some more insight. Good luck! Matt -- Matt Garman email at: http://raw-sewage.net/index.php?file=email -- gentoo-user@gentoo.org mailing list
Re: [gentoo-user] arrow keys in dosemu
On Thu, Nov 13, 2003 at 07:36:50AM -0800, Mark Knecht wrote: > Interesting. I was running a standard xterm yesterday when I tested, > so following on I just tried dosemu under Eterm. The arrow keys work, > but the Alt key doesn't, so I cannot get back to the menu when I try > to get out of FreeDOS Edit. Like you, I am fine in the console which > is where I'll probably use this anyway. I realized there's an "xdosemu" program as well. In this, everything appears to work as expected. My main reason for using dosemu was to look at a bunch of "ANSI art" I created many years ago. xdosemu automatically picks a terminal font that supports (correct) displaying of those high-ascii characters. (In other words, another downside to running "vanilla" dosemu in an xterm is that you have to make sure you're using a font capable of displaying those high-ascii characters.) Thanks again for all the feedback! Matt -- [EMAIL PROTECTED] mailing list
Re: [gentoo-user] arrow keys in dosemu
On Thu, Nov 20, 2003 at 08:59:47AM -0800, [EMAIL PROTECTED] wrote: > Interesting to know about xdosemu. I'll try that out. > > One question - how do you cleanly exit for dosemu? I didn't find a > command to shut it down, other than just killing it which probably > isn;t the right thing to do. I used the "exitemu" command. Although, it has locked up on me a few times, and I've just killed the process. :) Matt -- [EMAIL PROTECTED] mailing list
[gentoo-user] update: c++ performance: gentoo, debian, windows
A while back I posted a message talking about the performance differences of a c++ program I wrote on gentoo and debian. In the end, I chalked up the performance degrade on gentoo to my having built gcc with too many optimizations. When I recompiled gcc with more "conservative" settings, my program ran as fast or faster on gentoo. If you recall, the processor/memory intensive part of my program is reading a 50,000+ line CSV file into memory. On my gentoo box, it takes about 1.6 seconds to load this into memory. On Windows, using MS Visual Studio 6.0, it takes five or six seconds to load into memory! I don't know how big a role computer speed plays, though. My home computer (gentoo) is an Athlon XP 2500 with 1 GB of ram. My work computer (windows) is a dual Xeon 1.5 GHz with 2 GB of ram. Judging by Windows Task Manager, only one CPU is used to load the data. So, loosely speaking, my home PC is 1.6 times faster, but the same process is 3.1 times slower (on my work PC). Again, this testing isn't scientific by any means, but I thought some folks might be interested. Looks good for gcc or AMD, in my opinion :) Matt -- [EMAIL PROTECTED] mailing list
Re: [gentoo-user] update: c++ performance: gentoo, debian, window s
On Mon, Nov 24, 2003 at 08:44:44AM -0600, Van Eps, Nathan D. (James Tower) wrote: > That is weird. A compiler should generate the same code whether it is > optimized or unoptimized. It would be interesting to hear an > explanation from the gcc folk as to what causes this. I think it had more to do with the *library* moreso than the compiler itself. For example, in the C language, all of your standard functions (fopen, strlen, malloc, etc) are part of the libc library. If an application makes a lot of calls to libc functions, different compilations of the libc library will probably result in different performance characteristics for the application. To put it another way, if you recompile (the same version of) a library, you don't have to recompile the applications that use it. In my case, my application makes heavy use of the C++ Standard Template Library (STL). The STL is built when I emerge (read: compile) gcc. (Note that libc is *not* built with gcc; it's part of it's own package, sys-libs/glibc.) Just a tip for anyone who may not be aware of this: it's easy to find out what libraries a program uses and what package those libraries belong to using ldd and qpkg: ldd Will list all the libraries that references. My application, for example, shows the following: ldd a.out libstdc++.so.5 => /usr/lib/gcc-lib/i686-pc-linux-gnu/3.2.3/libstdc++.so.5 (0x4001f000) libm.so.6 => /lib/libm.so.6 (0x400ce000) libgcc_s.so.1 => /usr/lib/gcc-lib/i686-pc-linux-gnu/3.2.3/libgcc_s.so.1 (0x400f) libc.so.6 => /lib/libc.so.6 (0x400f9000) /lib/ld-linux.so.2 => /lib/ld-linux.so.2 (0x4000) >From there you can use "qpkg -f " to tell you what package created a library (or any file for that matter). My example, continued: qpkg -f /usr/lib/gcc-lib/i686-pc-linux-gnu/3.2.3/libstdc++.so.5 sys-devel/gcc * I hope that doesn't add any confusion :) Matt -- [EMAIL PROTECTED] mailing list
[gentoo-user] per-user package administration
While daydreaming during a boring meeting, I was thinking how nice it would be to have a Linux box at work (currently I have only a Win2k machine). However, I'm so used to having root that if I got a Linux machine at work, I'd probably only have normal user access. In the past, on the various Unix systems I've used where I don't have root, but want a package installed, I usually just get the source, compile and install in my home directory. That's fine for small packages, but it's just too much work if the package has a lot of dependancies or needs a makefile tweaked, etc. Then I thought it wouldn't be too hard (in theory ) to have per-user access to the package tool (portage in gentoo's case). The user would have his own set of package accounting data; it wouldn't affect the system-wide database. For example, say I want to install package slrn, which has a dependancy on the slang package. I don't have admin access, and my sysadmin doesn't want it installed. It would be nice if I could, however, do an "emerge slrn" and have portage download, build and install slrn and slang all in my home directory. Assuming I have the disk space, I could do this manually anyway. But, obviously, portage makes it easier. Plus, say slang was already installed at the system level, it would be cool if the "per-user" version of the package tool recognized that. This would also be nice for testing new or development versions of packages. You could do a local install (to a test account, for example) and play with the software without affecting the integrity of the rest of the system. I really have a need for such a feature, but someone else might. I just thought I'd throw the idea out there... maybe someone will be inspired :) Matt -- [EMAIL PROTECTED] mailing list
[gentoo-user] LAN performance issues
I have a 100/10 Mbps home LAN whose performance is considerably less than capable. The setup is as follows: - OpenBSD server with two PCI NICs (a D-Link and a Linksys). One NIC connects to the Internet, the other connects to the LAN. (This machine provides the firewall/gateway/NAT functions.) - A Linksys "Etherfast Dual-Speed 5-Port Workgroup Switch" - My Linux workstation. It's using an onboard 3Com 3c59x NIC (Asus A7N8X Deluxe motherboard) - My roomate's Win2k box. Also has a Linksys 100/10 PCI NIC. The best LAN transfer speeds I can get are around 230 KB/s. This is between Linux and OpenBSD as well as Win2k and OpenBSD. Interestingly, between Linux and Win2k, the speed is about half that: 130 KB/s. The Linux NIC & driver appears to be functioning correctly; I checked its operation with mii-diag and vortex-diag [1]. I'm not aware of OpenBSD NIC diagnostic tools, so I don't know how to check those NICs as extensively. But all three NICs (two in BSD + one in Linux) are operating in full-duplex 100baseTX mode. All my cables are CAT5e UTP. Just to be sure, I bought new cables. Transfer speeds did not improve. I tried using a cross-over cable between OpenBSD and Linux. No improvement. I also tried switching the roles of the BSD NICs (i.e. Internet NIC became DSL NIC and vice-versa). No change in transfer speeds. Short of phsically moving NICs around, I'm not sure how to futher diagnose the problem. Anyone have any ideas, thoughts, hints, suggestions, etc? Thanks, Matt [1] http://www.scyld.com/diag/index.html -- Matt Garman email at: http://raw-sewage.net/index.php?file=email -- [EMAIL PROTECTED] mailing list
Re: [gentoo-user] gcc 3
On Fri, Dec 05, 2003 at 03:06:05PM +0100, Oliver Lange wrote:
> warning: undefined reference to `operator new(unsigned)'
> warning: undefined reference to `operator delete(void*)'
> warning: undefined reference to `vtable for
> __cxxabiv1::__si_class_type_info'
>
> These are calls from a C function which is using a C++ class. What's
> the difference between GCC2 and GCC3 ? Is it a missing software
> package ?
There are two things that may help:
1. Try compiling your code with g++ instead of gcc. If you are using
new and delete, then you are writing c++ code.
2. In your header file, try to wrap your function prototypes like the
following:
#ifdef __cplusplus
extern "C" {
#endif
int some_function(int x);
char another_function(const char* s);
#ifdef __cplusplus
}
#endif
Linking is performed a bit differently for C and C++. Using extern "C"
tells the compiler to make C++ functions visible to C.
You really can't just declare a C++ object in C. The C language alone
knows nothing of classes, objects, function overloading, the STL and all
the other stuff C++ adds. When you start using those features in your
c file, you may as well rename it .cc (or .cxx or .cpp or whatever
extension you use for C++ source files). Note that that means that you
can't just wrap your C++ classes with extern "C" and expect them to be
available to regular C.
If you have some C++ source whose functionality you wish to make
available to vanilla C, you have to write some kind of interface or
wrapper code.
Look at it this way, C++ is a *superset* of C. You can call C functions
from C++. But you can't call C++ code from C without writing some C++
that allows you to do that.
Hope that helps,
Matt
--
Matt Garman
email at: http://raw-sewage.net/index.php?file=email
--
[EMAIL PROTECTED] mailing list
Re: [gentoo-user] what will happen..
On Fri, Dec 05, 2003 at 07:20:03PM +0100, Marius Mauch wrote: > Either others will join and continue the work or it will die, it's the > same as with any other OSS project. Even in the commercial world there > is no guarantee that a product line will live forever (see Redhat). > But considering that we currently have 200-300 people working on > Gentoo it would take a long time before that would happen. Agreed. And looking at the "big picture"---considering all open source and free software---I'd have to say that it's only getting better. More users, more developers, more support, more exposure, etc, etc. Of course there will be a few gems that get lost, but overall, there's more for everyone. And even if a good project does die, being open source means that there's not all this political cruft attached to it. Anyone can pick up a dead project and bring it back to life. MG -- Matt Garman email at: http://raw-sewage.net/index.php?file=email -- [EMAIL PROTECTED] mailing list
Re: [gentoo-user] LAN performance issues
On Sat, Dec 06, 2003 at 03:52:44PM -0500, Jerry McBride wrote: > How's the data throughput of your hard drives? If it's low, having a > fast network won't make getting data off or onto the hard drives any > faster. Nope, I've done a bit of hard drive benchmarking (using bonnie++), and the drives certainly aren't a bottleneck. Thanks for the idea though! Matt -- Matt Garman email at: http://raw-sewage.net/index.php?file=email -- [EMAIL PROTECTED] mailing list
Re: [gentoo-user] LAN performance issues
On Sat, Dec 06, 2003 at 04:36:03PM -0500, Jerry McBride wrote: > On Saturday 06 December 2003 04:14 pm, Mike Arrison wrote: > One other thing I'm waiting to hear... how's the network configed? You > using NFS, SAMBA, AFS, etc? Sorry, I meant to mention that in my original email. But, the transfer speeds are all essentially the same using various protocols. Thus far I've used the following methods to transfer files: - http - sftp/scp (encryption overhead doesn't seem to impact performance) - rsync - samba Thanks again! Matt -- Matt Garman email at: http://raw-sewage.net/index.php?file=email -- [EMAIL PROTECTED] mailing list
[gentoo-user] partition restructuring
I used to run my system from a 120 GB drive. The drive is partitioned as follows: partition size typefilesystem - -- /dev/hda116M primary ext2 /dev/hda2 3.7G primary ext3 /dev/hda3 1.5G primary Linux swap /dev/hda4 (extended) /dev/hda5 3.7G logical ext3 /dev/hda6 3.7G logical ext3 /dev/hda715G logical ext3 /dev/hda828G logical ext3 /dev/hda928G logical ext3 /dev/hda10 28G logical ext3 Now I no longer need partitions 1 through 7. I'd like to roll them up into one partition, or even better, add them to partition 7 (so I'd have four, roughtly equally-sized partitions). That in mind, I do NOT want to lose the data on partitions 8, 9, and 10. Will the available partioning tools (e.g. GNU parted) allow me to do this safely? If so, would it take a long time? I have another 120 GB drive in a second computer; I could just use rsync to make a backup of everything, then go to work with fdisk if that would be safer and/or faster. Thanks for any thoughts! Matt -- Matt Garman email at: http://raw-sewage.net/index.php?file=email -- [EMAIL PROTECTED] mailing list
[gentoo-user] font availability in gvim
Whenever I start gvim, I have only one font available. In particular, I like the "10x20" font. I can do something like xterm -fn 10x20 And I'll get an xterm that uses that font. So in my ~/.vimrc, I have the following line: set guifont=10x20 However, gvim loads with a tiny, fixed-width font that I can't identify. Even if I manually do a :set guifont=10x20 (within gvim), the font does not change. In fact, if I do an Edit -> Select Font, I get a font list that is quite abbreviated---there is certainly nothing useable by me, anyway. All my other X apps (gkrellm, MozillaFirebird, konqueror, etc) do not have any apparent font problems. Any thoughts? Thanks, Matt -- Matt Garman email at: http://raw-sewage.net/index.php?file=email -- [EMAIL PROTECTED] mailing list
Re: [gentoo-user] font availability in gvim
On Thu, Jan 08, 2004 at 11:54:22PM -0500, Chris Bare wrote: > > Whenever I start gvim, I have only one font available. > > > Any thoughts? > > I had a problem when I built gvim with gtk2. Gtk2 uses the new anti-aliased > font stuff and I was never able to configure it to use the good old-fashoned > font that matches my xterm which I have been staring at every day for at least > 10 years. > > My solution was to set USE=-gtk2 and rebuild gvim. Yup, that fixed it! Thanks, Matt -- Matt Garman email at: http://raw-sewage.net/index.php?file=email -- [EMAIL PROTECTED] mailing list
[gentoo-user] arts-1.1.5 emerge failure
I don't know if this warrants a bug report or not... but I'm trying to do an "emerge -vUuD world" and it is failing on the arts-1.1.5 package: ... checking if STL implementation is SGI like... no checking if STL implementation is HP like... no configure: error: "no known STL type found - did you forget to install libstdc++-devel ?" !!! ERROR: kde-base/arts-1.1.5 failed. !!! Function kde_src_compile, Line 117, Exitcode 1 !!! died running ./configure, kde_src_compile:configure Any thoughts or ideas? Thanks, Matt -- Matt Garman email at: http://raw-sewage.net/index.php?file=email -- [EMAIL PROTECTED] mailing list
Re: [gentoo-user] AMD64 or Dual Athlon or Other? Motherboard suggestions?
On Wed, Jan 21, 2004 at 12:42:22AM -0500, Thomas Achtemichuk wrote: > 1 x Tyan TigerMPX (S2466N-4M Rev B or 2 or something) [1] > 2 x Athlon MP 1900+ > 2 x 512 Corsair PC2100 REG ECC > 2 x Alpha PAL6035 Heatsinks > ... How quiet is that system? I seem to recall reading somewhere that the Alpha PAL HSFs aren't known for their quietness. Just curious :) Matt -- Matt Garman email at: http://raw-sewage.net/index.php?file=email -- [EMAIL PROTECTED] mailing list
[gentoo-user] ipchains: firewall, nat, gateway...
For those of you using Gentoo (or Linux in general) as your firewall, gateway and/or NAT box, I was curious as to how you went about setting up Netfilter/iptables. I'm currently using OpenBSD for this role, in part because I wanted another system to tinker with, but also because I find it's firewall system a bit easier to setup and configure (than iptables). Now I barely have time to tinker with my Gentoo box, let alone OpenBSD. So I'd kind of like to migrate the OpenBSD box to Gentoo. The most important thing, though, is to make sure I can get a good firewall up. I've played with iptables in the past, but felt a bit overwhelmed. To me it seems that there should be a "standard" iptables script for the scenario where the Linux box connects do the Internet via DSL or cablemodem and performs NAT, firewall and gateway duties to an internal (private) LAN---I'm assuming this is an extremely common situation. That in mind, I'm just curious as to how folks have gone about implementing their iptables-based firewall. I did a bit snooping around the Gentoo forums, and came up with a few promising links: Projectfiles.com [1] has a pre-written iptables script. It has a handful of endorsements on the forums, but it's length (and assumed complexity) makes me sheepish. The Gentoo Linux Security Guide [2] also has some information on setting up iptables as well as a sample script [3]. Finally, many folks recommend Shorewall [4] which apparently is a higher-level Netfilter configuration tool. I haven't studied any of the above links in any detail yet, but thought they'd make a good starting point for discussion. Thanks for your thoughts and feedback! Matt [1] http://projectfiles.com/firewall/ [2] http://www.gentoo.org/doc/en/gentoo-security.xml [3] http://www.gentoo.org/doc/en/gentoo-security.xml#doc_chap12_pre5 [4] http://shorewall.sourceforge.net/ -- Matt Garman email at: http://raw-sewage.net/index.php?file=email -- [EMAIL PROTECTED] mailing list
[gentoo-user] strange install situation
Hello: I used the Athlon XP v1.4 (dated 9/11) Live CD disk 1 to do a Stage 3 install of Gentoo. Everything *seemed* to go okay, until reboot, when Grub couldn't find the kernel. I used the Live CD as a rescue disk, looked at my install drive, and, sure enough, no kernel was installed! Also, the /lib/modules directory tree was empty. Now, I'm not (completely) crazy; I specifically remember (during the initial install) doing an "emerge -k sys-kernel/gentoo-sources" then a "emerge -k genkernel" and finally running "genkernel --config". I made a few minor configuration changes to the kernel (mostly removing unneeded modules), then waited while it built bzImage, modules, did a make modules_install, etc. It took a while, so *something* was going on :) Unfortunately, I don't remember if it exited with an error code (I'm pretty sure I would have noticed that though). As I said, however, for some reason, neither the kernel nor the modules (or even the initrd) got installed. So I repeated the above steps after rescueing from the Live CD. My kernel config options had persisted, and this time the kernel, initrd, modules, etc all actually got installed. But now I'm wondering if other things aren't quite right with my installation. For example, I installed from my cdrom, which is typically /dev/hdc, but there's no hdc showing up in the /dev directory. (I'm using ide-scsi on that drive, but there also aren't any /dev/scd* files.) I'm trying to build the "nvidia-kernel" modules, but when I do this, I get the following error: emerge nvidia-kernel Calculating dependencies ...done! >>> emerge (1 of 1) media-video/nvidia-kernel-1.0.4496-r3 to / >>> md5 src_uri ;-) NVIDIA-Linux-x86-1.0-4496-pkg0.run >>> Unpacking source... Creating directory NVIDIA-Linux-x86-1.0-4496-pkg0 Verifying archive integrity... OK Uncompressing NVIDIA Accelerated Graphics Driver for Linux-x86 1.0-4496. * Linux kernel 2.4.20 >>> Source unpacked. rm -f nv.o os-agp.o os-interface.o os-registry.o nv-linux.o nv_compiler.h *.d NVdriver nvidia.o echo \#define NV_COMPILER \"`gcc -v 2>&1 | tail -n 1`\" > nv_compiler.h gcc -c -Wall -Wimplicit -Wreturn-type -Wswitch -Wformat -Wchar-subscripts -Wparentheses -Wpointer-arith -Wcast-qual -Wno-multichar -O -MD -D__KERNEL__ -DMODULE -D_LOOSE_KERNEL_NAMES -DNTRM -D_GNU_SOURCE -D_LOOSE_KERNEL_NAMES -D__KERNEL__ -DMODULE -DNV_MAJOR_VERSION=1 -DNV_MINOR_VERSION=0 -DNV_PATCHLEVEL=4496 -DNV_UNIX -DNV_LINUX -DNV_INT64_OK -DNVCPU_X86 -DREMAP_PAGE_RANGE_4 -I. -I/usr/src/linux/include -Wno-cast-qual nv.c In file included from nv.c:14: nv-linux.h:24:31: linux/modversions.h: No such file or directory nv.c: In function `cleanup_module': nv.c:861: warning: unused variable `i' make: *** [nv.o] Error 1 !!! ERROR: media-video/nvidia-kernel-1.0.4496-r3 failed. !!! Function src_compile, Line 117, Exitcode 2 !!! (no error message) And it's true, I don't have a /usr/src/linux/include/linux/modversions.h file. xmms also fails: >>> emerge (40 of 54) media-sound/xmms-1.2.7-r20 to / >>> md5 src_uri ;-) xmms-1.2.7.tar.gz >>> Unpacking source... >>> Unpacking xmms-1.2.7.tar.gz to >>> /var/tmp/portage/xmms-1.2.7-r20/work * Applying xmms-jump.patch... [ ok ] * Applying xmms-sigterm.patch... [ ok ] * Cannot find $EPATCH_SOURCE! Value for $EPATCH_SOURCE is: * * /usr/portage/distfiles/xmms-1.2.7-mmx.patch.gz !!! ERROR: media-sound/xmms-1.2.7-r20 failed. !!! Function epatch, Line 182, Exitcode 0 !!! Cannot find $EPATCH_SOURCE! Now, I'm not sure if these things are related to the kernel problem or not... I'm tempted to just re-do the whole install, but that takes a fair amount of time. Any suggestions or ideas? Thanks! Matt -- [EMAIL PROTECTED] mailing list
Re: [gentoo-user] strange install situation
On Wed, Oct 15, 2003 at 02:19:11PM +0100, Jon Dye wrote: > Did you create a seperate /boot partition? If so did you look in > there for your kernel rather than the / partition (which would of > had an empty /boot in that case). You also need to tell grub to > look in your /root partition rather than the / one for the kernel > and then as a kernel option pass the correct root. e.g. > > kernel (hd0,0)/kernel-2.4.22-gentoo-r7 root=/dev/hda2 > > (assuming /boot as first parition and / as second). Yup, I created the boot partition, although, I guess there is a chance I didn't mount it when I installed the kernel the first time. But, keep in mind, that the modules were also not installed the first time (and they live on the root partition in /lib/modules/). However, once I booted the LiveCD (2nd time), mounted filesystems, chroot'ed, and re-did the genkernel, everything seemed to work (I didn't have to change GRUB's settings). Thanks! Matt -- [EMAIL PROTECTED] mailing list
Re: [gentoo-user] strange install situation
On Wed, Oct 15, 2003 at 11:17:28AM -0400, Chris I wrote: > > In file included from nv.c:14: > > nv-linux.h:24:31: linux/modversions.h: No such file or directory > > nv.c: In function `cleanup_module': > > nv.c:861: warning: unused variable `i' > > make: *** [nv.o] Error 1 > > > And it's true, I don't have a > > /usr/src/linux/include/linux/modversions.h file. > > Basically, it looks like you didnt do the symlink from your kernel > source to /usr/src/linux . Shouldn't emerging the kernel source or using genkernel do that for me? At any rate: bash-2.05b# ls -la /usr/src/linux lrwxrwxrwx1 root root 22 Oct 14 19:43 /usr/src/linux -> linux-2.4.20-gentoo-r7 bash-2.05b# updatedb bash-2.05b# locate modversions.h bash-2.05b# So maybe it's just not part of the gentoo-sources 2.4.20-gentoo-r7 kernel? On my old system, I have a modversions.h file in both include/linux and include/config in my "vanilla" kernel source trees, for versions 2.4.20 through 2.4.22. I think I'm just going to re-install anyway (if only for peace of mind). I must have made some subtle mistake or overlooking something. > > xmms also fails: > > This i dont know. try emerge sync'ing again and see if the problem > still exists. Yup, xmms is no longer problematic. I'll write that one off to random strangeness :) Thanks for all the feedback! Matt -- [EMAIL PROTECTED] mailing list
[gentoo-user] multiple instances of emerge and package naming
Can I run two (or more) simultaneous instances of emerge? Would this create any consistency problems or is the Portage system smart enough to synchronize its accounting? Also, is there a difference in specifying a package name by its fully-qualified name as opposed to just the package name? In other words, is there any difference between the following: emerge net-www/mozilla-firebird emerge mozilla-firebird If not, why would you ever use one over the other (unless two or more packages had the same name)? Thanks! Matt -- [EMAIL PROTECTED] mailing list
Re: [gentoo-user] nVidia or ATI
On Thu, Oct 16, 2003 at 10:30:24AM -0700, Ian Truelsen wrote: > I am looking to get a new video card for my desktop (the old Voodoo > 3 is starting to show its age). What I would like to know is which > of the big two nVidia or ATI are better supported under Linux for > framebuffer stuff and for DRI. nVidia gets my vote. I've been using one for a few months now without any problems. I just switched to Gentoo, so before I was installing the nvidia drivers manually---it was trivial. I'm not a gamer, but I do use Pro/Engineer on my Linux box; that application is pretty picky with video cards. Our standard workstations at work *used* to come with ATI cards, but they had too many problems and we switched to nvidia. Based entirely on hearsay, the nvidia drivers are generally of much better quality and more stable (though maybe not as well performing). Although, (again based on hearsay) the ATI drivers seem to have improved reliability as of late. Hope that helps! Matt -- [EMAIL PROTECTED] mailing list
[gentoo-user] reiserfs corruption likelihood?
Purusing the Gentoo FAQ, I ran across the following: "ReiserFS and filesystem corruption issues -- how to fix'em, etc If your ReiserFS partition is corrupt, try booting the Gentoo Linux boot CD and run reiserfsck --rebuild-tree on the corrupted filesystem. This should make the filesystem consistent again, although you may have lost some files or directories due to the corruption." Granted I should have read this before doing the install, but, I made all my partitions Reiser (except for boot). So now I'm wondering about the likelihood of data corruption. Has anyone experienced it? Are there things I can do to ensure it doesn't happen to me (or at least reduce the possibility)? (Other than "backup early, backup often" of course :) Thank you! Matt -- [EMAIL PROTECTED] mailing list
[gentoo-user] ext3 fs corruption likelihood? was: Re: [gentoo-user] reiserfs corruption likelihood?
On Thu, Oct 16, 2003 at 04:24:37PM -0600, Nate Duehr wrote: > On Thu, Oct 16, 2003 at 04:24:04PM -0600, Collins Richey wrote: > > 2) Re both resierfs and xfs, I've seen about equal numbers of > > responses on any number of mailing lists: > > > > "pick-a-fs" ate my lunch, hard drive, and system. > > "pick-a-fs" is the greatest thing since sliced bread. > > > > You pays you money (or time), you takes you chances. > > Pretty wise viewpoint, I'd say. And basically what I was trying to > get across. If you value your data, don't use filesystems that have > regular reports of eating other people's data for lunch. :-) > Boring, I know... but safer. (GRIN) I'm now leaning more towards ext3. My previous system (Debian unstable) ran ext3 without any problems, and was plenty fast. At this point I value reliability/stability over performance. As was said above, reiserfs looks like it's either feast or famine. So now I'm going to ask the converse question: has anyone experienced unexplained data corruption with ext3? By the way, I'm not trying to start a "my fs vs your fs war". I just saw that blurb about reiserfs corruption in the Gentoo FAQ and (1) wondered if it's due to user error or flaky software; and (2) if the latter, what is the likelihood of corruption? Thanks to everyone for the interesting feeback! Matt -- [EMAIL PROTECTED] mailing list
[gentoo-user] changing reiserfs partitions to ext3
I'm a bit worried about data corruption with reiserfs, and more concerned with stability and reliability than performance. So I think I'd like to convert my reiserfs partitions to ext3. What would be the best way to accomplish this (without doing a re-install)? I was thinking of the following procedure: - boot from the LiveCD - use tar to backup a partition to some temporary location (making sure tar preserves links, permissions, ownership, etc) - re-format the partition as ext3 - use tar to move the data from the temp location to the newly formatted ext3 partition I would repeat this procedure for every one of my reiserfs partitions. Finally, I would update /etc/fstab to reflect the changes. Assuming I didn't make any mistakes in the above procedure, would this preserve my system integrity? Thanks again! Matt -- [EMAIL PROTECTED] mailing list
Re: [gentoo-user] Gentoo's preference, Gnome or KDE ?
On Fri, Oct 17, 2003 at 05:32:55PM +0100, Rick [Kitty5] wrote: > I have never really been a KDE fan, although I have been using it on > my gentoo workstation as I had problems getting Gnome to work > properly. I have to say its far to 'busy' for my tastes, would love > to purge all the bundled 'everything and the kitchen sink' stuff I > am just never going to use. You might want to give xfce a try. The latest xfce is version 4, and, although I don't personally use it, I thought it was pretty slick. It's a desktop environment like GNOME and KDE, but certainly more lightweight (not as many bells and whistles). It's somewhere between a standalone window manager and a heavyweight desktop environment. :) MG -- [EMAIL PROTECTED] mailing list
[gentoo-user] fluxbox doesn't see artwiz fonts
I merged the development version of fluxbox (v0.9.6) via /usr/portage/x11-wm/fluxbox/fluxbox-0.9.6_pre9.ebuild. I also installed the package artwiz-fonts. At least xfontsel can see the artwiz fonts. However, fluxbox is not using them. So I'm curious if I need to indicate the location of the artwiz-fonts in some config file. gvim also cannot use the artwiz fonts. In general, is there a "Gentoo X Fonts HOWTO" or similar documentation? My default font is a fairly small and somewhat blury sans-type font. It was used by MozillaFirebird (until I tweaked the settings), it's used by the gkrellm configuration screen, it's used by xchat-2, etc. So I'd like to make the default something "prettier". I'm assuming there's a fairly standard set of steps that most Gentoo folks take to setup "nice" fonts after a fresh install. Is this documented somewhere? Thanks, Matt -- [EMAIL PROTECTED] mailing list
Re: [gentoo-user] Grub setup fails
On Mon, Oct 20, 2003 at 09:10:13PM +0200, Patrick Marquetecken wrote: > Is this the same for the grub.config ? > Not the use of sda but only hda Yup. I discovered (by trial and error) that there are some subtleties to how grub sees disks. I have both IDE and SCSI hard drives in my computer. I installed to the SCSI drive. However, when I booted the Gentoo LiveCD (for installation), I had to do a "modprobe sym53c8xxx" so that Linux could see my SCSI card and disk (my CD-ROM is IDE). Keep in mind, the IDE drive had already been recognized by Linux. So during the install, (hd1,0) was my SCSI drive, and (hd0,0) was my IDE drive. My boot order (as configured by the BIOS), is Floppy, IDE CD-ROM, SCSI. So when I actually boot the *installed* version of Gentoo, my SCSI hard drive becomes (hd0,0) and the IDE hard drive becomes (hd1,0). In other words, I used (hd0,0) in my grub.conf file. Fun stuff :) Matt -- [EMAIL PROTECTED] mailing list
Re: [gentoo-user] MultiThreaded music rippers (was: iTunes on linux?)
On Tue, Oct 21, 2003 at 01:50:44PM +0200, Sigurd Stordal wrote: > > now theres a question. is there any app that i can use to rip 2 cds > > at once? i have 2 CD rom drives and 2 CPUs and enough memory (1gb) > > so i think > > Well you could use two instances of grip, and set the cdrom device in > the second one tho the second cdrom device. Another "ripper" you may want to check out is "cdparanoia". It'll eat up your CPU pretty good, but if your CDs aren't in pristine condition, it'll help make sure errors aren't transferred to your magnetic copy. I had a CD that skipped pretty badly because it had so many scratches. cdparanoia actually fixed a lot of the disc's problems (on the copied version of the music, obviously). At the same time, however, it took a LONG time to copy that CD. Hope that helps, Matt -- [EMAIL PROTECTED] mailing list
[gentoo-user] slightly OT: c++ performance: gentoo vs. debian
I just got a new hard disk and installed gentoo on it. I've got a c++ development project that I was working on that runs noticeably slower on my gentoo box than it did on under my debian unstable installation. I basically wrote a CSV file reader in C++. The implementation uses a vector of vectors (one vector per row for each field, and a vector for every row). In other words: std::vector< std::vector< std::string > > data; Anyway, my data file has roughly 50,000 records. Under the Debian system, it took, on average, two or three seconds to load this data into memory. However, under the Gentoo system, it takes eight or nine seconds on average to load the data. (Run times don't change appreciably between subsequent runs, so I don't think caching is the issue here.) That's a pretty dramatic change, in my opinion. No hardware on my system has changed (except the new disk, which I have verified is not the source of the slowdown (it's a 10k SCSI drive, should be faster if anything)). I haven't changed any compile settings from one system to the other. Just for kicks, I tried using the optimising options I'm using for Gentoo in my /etc/make.conf: "-O3 -march=athlon-xp -pipe". That knocked a couple seconds off the load time, but the Debian-compiled version is still much faster. For what it's worth, the Debian system used: g++ (GCC) 3.3.2 20030908 (Debian prerelease) And the Gentoo system is using: g++ (GCC) 3.2.3 20030422 (Gentoo Linux 1.4 3.2.3-r2, propolice) Looks like Debian's is a bit newer. I don't follow gcc development really---have their been drastic improvements in a minor revision? Tonight I plan to setup grub to boot the old system to do some more investigating. Thanks for any comments or thoughts! Matt -- [EMAIL PROTECTED] mailing list
Re: [gentoo-user] OT: How many of you are 100% Linux?
On Tue, Oct 21, 2003 at 09:55:35AM -0400, Chris Bare wrote:
> BS. What you already know is easier to use. I think Linux has a
> different learning curve than windows. Initially it's tough, but once
> you get some basic concepts (editing a file, find, grep) the curve
> flattens out. I think with windows the curve is constant because
> everything has config windows hidden in different places that work
> slightly differently.
I wish more people felt this way. I think "what you already know is
easier to use" is a profound statement. Folks that argue about Windows'
"ease of use" say all I have to do is open this, click here, click this
tab, click okay, click... And I say, in Linux, all I have to do is open
this file, add these lines...
Unix sys-admining just seems easier to me because I know where every
program option can be found, where every config file may be located.
It's easy to document as well ("Program X should have its config file
installed in /some/path/config_file_name"). It's really asking a lot of
someone to try and remember where every config option is buried in a
dialog. If you miss a tab or the "Advanced..." button or whatever
inconsistent scheme your GUI-based config utility uses, you're probably
missing have of the goodies. Plus, it's pretty hard to document (and
keep that documentation up to date) when you have to decribe a "click
trail" for every config option.
Of course, not all GUI configs are bad (I like 'em sometimes myself). I
mainly feel that Unix has an undeserved repuation of being "difficult to
configure". I think if you can know and understand Unix configuration
basics, you've really learned *concepts* rather than wrote procedures.
But understanding such concepts gives a more in-depth knowledge of how
the computer works, which ultimately (in theory) should empower you with
getting more out of your system (more productivity, etc).
I'll stop ranting now :)
Matt
--
[EMAIL PROTECTED] mailing list
Re: [gentoo-user] OT: How many of you are 100% Linux?
On Tue, Oct 21, 2003 at 07:47:58AM -0400, Eric Livingston wrote: > I'm curious regarding the penetration of linux as a comprehensive > solution for all computing tasks in a normal day. i.e. what percentage > of Linux users are 100% linux, or even 100% Gentoo for that matter. Well, I'll chime in with my story, in case anyone is interested. I started using Linux back in '95 or '96, when I was in high school. At the time I ran a bulletin board system (aka "BBS", does anyone remember those?). Therein lied a need for a decent multi-tasking system (run the BBS but also do work/play at the same time). I first used DESQview under MS DOS 6.22, then DESQview/X (under same DOS). Those were okay, but I wanted more... So I tried Windows 3.11. Ha! That "OS" definately did not have any usable multitasking capabilities. I finally heard about Linux, and installed Slackware. I believe the Linux kernel version was in the 2.0.36 range or so at the time (I could be mistaken). Anyway, BBSes were going out of style in favor of the Internet, so the BBS hobby got replaced by the "playing in Linux" hobby. Until my sophomore year of college, I dual-booted Windows 95 and Linux. I basically kept Windows around until I could do everything I needed under Linux. Away went Windows! I also switched to Debian about the same time. Since ditching Windows, I've also picked up OpenBSD. I use OpenBSD for my firewall/gateway/NAT box. The main reason for OpenBSD is that I find their firewall (pf) much easier to setup than Linux iptables... plus that security reputation is awfully attractive. Just last week I switched to Gentoo (although the live Debian install is still available; I'm just keeping it around for config files and such). At work, unfortunately, I have to use Windows 2000. There's certainly too much beauracracy for me to even *try* get Linux on my system. Not to mention I do development exclusively for Windows customers. I think I could get by with OpenOffice.org for the most part, but my group works with a handful of MS Access databases, complete with forms and VB script and all that. Damn vendor lock-in. Every now and then I think about looking for a job where I could actually do Linux development. :) Matt -- [EMAIL PROTECTED] mailing list
Re: [gentoo-user] slightly OT: c++ performance: gentoo vs. debian
On Wed, Oct 22, 2003 at 07:08:39AM +0800, William Kenworthy wrote: > On all the machines (athlon t-bird, p4) I have tried so far, -O3 > always decreases performance - O2 is best (dramatically so on > something like a Indeed, in this case, O2 is faster than O3. I assumed Debian's gcc package was compiled with O2, so I just re-merged my gentoo gcc using O2. I recompiled my program, and now it is back to its better running time (two or three seconds to load all the records in memory). It would be interesting to actually develop a whole suite of tests to see when O3 is faster (if ever) and when O2 is faster. Based on this one (completely un-scientific :) test, I'd have to say that the C++ standard library is best compiled with O2. For what it's worth, I did some benchmarking on my new drive as well, and it's performing as good or better than the old drive. Thanks for all the feedback! If anyone else happens to do similar testing or benchmarking of compiler optimizations, I'd be interested in reading about them. Thanks again, Matt -- [EMAIL PROTECTED] mailing list
[gentoo-user] no fullscreen with mplayer
I have the following mplayer merged on my system: mplayer --version MPlayer 0.92-3.2.3 (C) 2000-2003 MPlayer Team CPU: Advanced Micro Devices (Family: 6, Stepping: 0) Detected cache-line size is 64 bytes CPUflags: MMX: 1 MMX2: 1 3DNow: 1 3DNow2: 1 SSE: 1 SSE2: 0 Compiled for x86 CPU with extensions: MMX MMX2 3DNow 3DNowEx SSE It's got some "issues" doing fullscreen. If I simply do a "mplayer somefile.mpg" the video will play fine in a small window. If I hit "f" (to switch to fullscreen), the video will completely disappear. However, other than not being displayed, it seems that mplayer "thinks" the video is still running (as the status info in the terminal will be updating). Plus, I can hit "f" again, and return the video to the visible windowed state. On the other hand, if I start mplayer using "mplayer -fs somefile.mpg", then the video will again start in a small window. But this window is actually like a "portal" that only lets me see a smaller part of the whole video (i.e. it seems to be fullscreened, but I can only see a small square of that). If I hit "f", the video will completely disappear, but mplayer is still running (as described above). If I press "f" a third time, the video will now be shown correctly in fullscreen mode. I tried recompiling mplayer with more conservative CFLAGS (O2 instead of O3), but that didn't help. Has anyone ever seen anything like this? Any ideas on what might be wrong? Thanks, Matt -- [EMAIL PROTECTED] mailing list
Re: [gentoo-user] no fullscreen with mplayer
On Wed, Oct 22, 2003 at 06:49:05PM +0100, Ciaran McCreesh wrote: > On Wed, 22 Oct 2003 12:29:33 -0500 Matt Garman <[EMAIL PROTECTED]> > wrote: > | Has anyone ever seen anything like this? Any ideas on what might be > | wrong? > > mplayer fullscreen has issues with fluxbox. Upgrading to mplayer pre1+ > and fluxbox 0.95+ fixed these for me. Other window managers may well > have similar issues... Yup, turns out Fluxbox is the culprit. I ran a plain ol' xterm as my window manager, and mplayer worked as expected. For what it's worth, I'm running Fluxbox 0.9.6 and mplayer 0.92. Kind of strange, though. I had the same version of Fluxbox on my previous system and mplayer version 0.90. Matt -- [EMAIL PROTECTED] mailing list
[gentoo-user] More disk performance data
Inspired by the recent threads regarding disk performance, I thought I'd do a bit of (informal) benchmarking. I have two machines, an OpenBSD box and a gentoo box. Each has a SCSI drive and an IDE drive. The two IDE drives are the same make and model. All OpenBSD filesystems are FFS and all gentoo filesystems are ext3. I used "bonnie" on OpenBSD (since bonnie++ has not been packaged for 3.3). On the gentoo box I used bonnie++ and hdparm. I haven't studied the results too closely yet. I just thought I'd post them and maybe generate some interesting discussion. My first impression is that ext3 (apparently) eats up the processor, at least compared to FFS (which I think is non-journaling). Matt OpenBSD 3.3 (generic kernel) Athlon Thunderbird 1 GHz Gigabyte ga-7ixe4 motherboard 512 MB RAM Tekram dc-390u2w PCI SCSI controller Maxtor Atlas 10k 18 GB SCSI drive Filesystem: FFS ---Sequential Output ---Sequential Input-- --Random-- -Per Char- --Block--- -Rewrite-- -Per Char- --Block--- --Seeks--- MachineMB K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU /sec %CPU OpenBSD 1000 29661 53.0 29233 15.9 8478 6.0 33653 64.9 33943 16.7 165.4 0.8 On-board IDE controller Western Digital wd1200jb 120 GB IDE drive Filesystem: FFS ---Sequential Output ---Sequential Input-- --Random-- -Per Char- --Block--- -Rewrite-- -Per Char- --Block--- --Seeks--- MachineMB K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU /sec %CPU OpenBSD 1000 28199 62.2 28616 24.1 4498 4.5 23435 48.3 24982 15.2 114.7 0.7 --- Gentoo 1.4 (gentoo-sources kernel w/sym53c8xx_2 compiled in kernel (not module)) Athlon XP "Barton" 2500 Asus A7N8X (nforce 2) motherboard 1 GB RAM LSI Logic PCI SCSI controller Fujitsu MAP 10,000 RPM 36 GB SCSI drive Filesystem: ext3 hdparm -Tt /dev/sda /dev/sda: Timing buffer-cache reads: 1588 MB in 2.00 seconds = 794.00 MB/sec Timing buffered disk reads: 194 MB in 3.02 seconds = 64.24 MB/sec Version 1.03 --Sequential Output-- --Sequential Input- --Random- -Per Chr- --Block-- -Rewrite- -Per Chr- --Block-- --Seeks-- MachineSize K/sec %CP K/sec %CP K/sec %CP K/sec %CP K/sec %CP /sec %CP gentoo 2G 15761 79 38441 12 20714 6 16510 67 65751 13 267.7 1 --Sequential Create-- Random Create -Create-- --Read--- -Delete-- -Create-- --Read--- -Delete-- files /sec %CP /sec %CP /sec %CP /sec %CP /sec %CP /sec %CP 16 1901 79 + +++ + +++ 2019 78 + +++ 8410 77 gentoo,2G,15761,79,38441,12,20714,6,16510,67,65751,13,267.7,1,16,1901,79,+,+++,+,+++,2019,78,+,+++,8410,77 On-board IDE controller Western Digital wd1200jb 120 GB IDE drive Filesystem: ext3 hdparm -Tt /dev/hda /dev/hda: Timing buffer-cache reads: 1580 MB in 2.00 seconds = 790.00 MB/sec Timing buffered disk reads: 140 MB in 3.02 seconds = 46.28 MB/sec Version 1.03 --Sequential Output-- --Sequential Input- --Random- -Per Chr- --Block-- -Rewrite- -Per Chr- --Block-- --Seeks-- MachineSize K/sec %CP K/sec %CP K/sec %CP K/sec %CP K/sec %CP /sec %CP gentoo 2G 15199 79 24020 7 10881 3 16641 66 46382 9 119.7 0 --Sequential Create-- Random Create -Create-- --Read--- -Delete-- -Create-- --Read--- -Delete-- files /sec %CP /sec %CP /sec %CP /sec %CP /sec %CP /sec %CP 16 2017 79 + +++ + +++ 2052 79 + +++ 8127 78 gentoo,2G,15199,79,24020,7,10881,3,16641,66,46382,9,119.7,0,16,2017,79,+,+++,+,+++,2052,79,+,+++,8127,78 -- [EMAIL PROTECTED] mailing list
[gentoo-user] arrow keys in dosemu
Does anyone out there happen to use dosemu? If so, did you have to do anything special to get your arrow keys to work? It appears that dosemu is interpreting the arrow keys (right, left, up, down) as an escape key. If I use the "edit" program, and I invoke the menus at the top of the screen, I cannot navigate them with the arrow keys (the become "unactive" as though I pressed escape). I also have another package (PCBoard, anyone else remember BBSes?) which has a (text-based) menu for its initial screen. Again, I can't navigate the menu options with the arrow keys, as pressing an arrow key just causes the program to end. Anyone have any ideas? Thanks, Matt -- [EMAIL PROTECTED] mailing list
Re: [gentoo-user] arrow keys in dosemu
On Wed, Nov 12, 2003 at 01:07:25PM -0800, Mark Knecht wrote: > Possibly this has something to do with the keyboard driver you are > calling, or the specific keyboard you are using, and not dosemu > itself? In my initial post, I was running dosemu under aterm (within X obviously). I also tried it within cygwin rxvt (under windows) and ssh'ing to my machine. With these two cases, the arrow keys did not work. However, I also tried running dosemu under a plain old Linux virtual console (i.e. no X). This time the arrow keys worked correctly. Back in X, I also tried it under a "vanilla" xterm and it worked. Hopefully I'll get a chance to do some more investigating tonight, but right now it appears the problem has something to do with how my terminal passes keystrokes to dosemu. Thanks for the feedback! Matt -- [EMAIL PROTECTED] mailing list
[gentoo-user] any Opteron/Athlon 64 gentoo users?
I was just curious if anyone on this list is running gentoo on an AMD64 platform, and what the results have been. I'm interested in the stability and performance you're seeing. Are compile times affected? How does the machine feel? Any other interesting comments? Just thought that might make for interesting e-conversation :) Matt (who certainly don't *need* 64-bit computing right now, but wants it just to have it :) -- [EMAIL PROTECTED] mailing list
Re: [gentoo-user] IDE RAID suggestions
On Fri, Mar 28, 2003 at 02:19:01PM +1100, Mark Saunders wrote: > Which is the best supported IDE RAID chipset? > > I'm after a motherboard with onboard IDE RAID support. > > I gave up on an ASUS A7V8X with a Promise Fasttrack onboard IDE RAID > controller. I'm certainly no expert here, but I have been snooping around a bit for Linux and IDE RAID information. If I understand the situation correctly, there are very few "true" hardware IDE RAID controllers supported by Linux. (I believe that one company offering these is 3ware). But these are REALLY expensive. >From what I gather, most of the "commodity" IDE RAID boards (Promise, HighPoint) are not hardware RAID; they rely on drivers (read: software) to do much of the RAID work. With this in mind, it appears your time would be better spent simply using the Linux md (multi-disk) functionality for software-level RAID (without needing any special hardware). Unless you have lots of money and can't afford to give up a tiny percetage of your CPU power, Linux md is the way to go. Someone please correct me if any of this info is inaccurate! With this in mind, however, what ARE the good "RAID" motherboards (for the extra IDE channels) or good add-on IDE PCI controllers? By "good" I mean having mature, stable Linux drivers? Matt -- Matt Garman, [EMAIL PROTECTED] ``I ain't never seen no whiskey, the blues made my sloppy drunk!'' -- Sleepy John Estes, ``Leaving Trunk'' -- [EMAIL PROTECTED] mailing list
[gentoo-user] gentoo migration stragety?
I'm currently running a Debian Linux system. I'd like to move over to gentoo, but I'm not exactly sure what the easiest way would be. Mainly from the perspective of partitions: I have separate partitions for root, boot, home, var, tmp and usr. I want to wipe all those clean, except for home, and install gentoo. Should I just manually remove everything from all my partitions (except home)? Or will gentoo do that for me during installation? If the latter case is true, how do I keep /home from getting erased? One more semi-related question: to make sure I understand this conceptually, using a "stage 3" install should get me up and running the most quickly, but I can still custom-compile most (or all?) of the stage 3 programs, right? Thanks for your feedback! Matt -- Matt Garman, [EMAIL PROTECTED] ``I ain't never seen no whiskey, the blues made my sloppy drunk!'' -- Sleepy John Estes, ``Leaving Trunk'' -- [EMAIL PROTECTED] mailing list
[gentoo-user] cdparanoia, cdrdao not using DMA?
If I mount a normal iso9660 cdrom, then copy all of its contents to my hard drive, it is extremely slow, and bogs down my system if DMA is not enabled. If I use hdparm -d 1 on the CDRW, the copy then goes very smoothly and does not cause a performance hit to the computer. However, using the tools cdparanoia and cdrdao cause the non-DMA behavior, i.e. CPU usage goes through the roof, the process takes forever, typing lags, the mouse is jumpy, etc. Also, when the process is done, hdparm shows that DMA has been turned off for that drive! Does anyone know why DMA mode is forced off for these tools? For what it's worth, my system is an Athlon 1 GHz with 512 MB of RAM. There are only two IDE drives in the system, a Western Digital 1200JB drive and a Lite-On 48x CDRW. Both drives are connected to their own channel on a Promise Ultra ATA100 PCI IDE card. Thanks for any feedback! Matt -- Matt Garman, [EMAIL PROTECTED] ``I ain't never seen no whiskey, the blues made my sloppy drunk!'' -- Sleepy John Estes, ``Leaving Trunk'' -- [EMAIL PROTECTED] mailing list
Re: [gentoo-user] Package management for non-ebuild software
On Wed, Apr 02, 2003 at 09:07:09PM +0100, Jan Drugowitsch wrote: > Is there any way with standard-gentoo tools to also keep track of > software installed with ./configure && make && make install? I used > install-log with LFS, but if there's a chance to use a different > system I would be glad to hear about it. I'm not sure about the "standard gentoo" requirement, but a very good manager of compiled-from-vanilla-source packages is epkg: http://encap.cites.uiuc.edu/epkg/ It is similar in concept to GNU stow. The basic concept is, rather than letting "make install" put everything in /usr/local/[bin|lib|man|...], you do something ala "make prefix=/usr/local/encap/packagename install" to put all the installed files in another hierarchy, i.e. /usr/local/encap/packagename/[bin|lib|man|...]. You then use epkg to create symlinks from the encap/packagename directory into your actual /usr/local hierarchy. Note this program is not distribution specific. I used it on Debian for years to install the latest versions of some select packages. Good luck. Matt -- [EMAIL PROTECTED] mailing list
Re: [gentoo-user] Package management for non-ebuild software
On Wed, Apr 02, 2003 at 03:39:58PM -0800, Timothy Grant wrote: > I use stow, never heard of epkg before, can you elucidate on the > differences between them, or perchance why epkg is a better choice? Well, I'm really not familiar with stow; I've really only heard a bit about it, never used it. I "grew up" on epkg simply because it was developed where I went to school, so I received much propaganda about it. I only use it for its most basic functionality, namely, managing symlinks in /usr/local... so, for my needs any package manager will do :) Sorry I don't have more to offer! Matt -- Matt Garman, [EMAIL PROTECTED] ``I ain't never seen no whiskey, the blues made my sloppy drunk!'' -- Sleepy John Estes, ``Leaving Trunk'' -- [EMAIL PROTECTED] mailing list
Re: [gentoo-user] mysterious lockups
On Mon, Feb 07, 2005 at 08:27:42AM -0800, Bob Sanders wrote: > Typically it's the graphics drivers hanging the AGP bus. But you > didn't memtion if you were running xscreensaver. If you're > running xscreensaver, do - > Then, when it hang again, ssh in and killall xscreensaver. Well, a lockup happened again last night (of course *not* while I was using the computer). Funny thing is, I can ping the locked machine, and it responds as though everything was okay. But I cannot ssh into the computer. ssh does not timeout (like it would if the host were down); after I type "ssh host" and press enter, it just sits there---I let it sit there for about 30 minutes before I got impatient and hard-rebooted the machine. But, FWIW, I am running xscreensaver. > Look in /proc/driver/nvidia/agp/card > ... > If you have SBA and Fast Writes on, try turning them > off - /etc/modules.d/nvidia I'm hesitant to try this, only because the lockups are so random. It's been almost two weeks since it last happened; I have a feeling that I could get lucky and it won't happen again forever (or it could happen in the next five minutes). Either way, it's hard to determine the solution when I make a change and then just wait. > For general system testing - not graphics, but it will rule out > other hardware problems, there is - app-benchmarks/stress. > ... > For Gfx issues, I typically, run - x11-misc/rss-glx by launching 3 > ... Yup, I think this ought to be my next step, to see if I can consistently duplicate the problem. Wow. Those look awesome! Stress testing that is both productive AND entertaining! > The other useful tool I've found is setiathome. It doesn't do > anythign for Gfx testing, but it does help out in keeping the cpu, > cpu bus, memory bus, and ide/scsi bus active enough to remove > doubts that those might be a problem. I've had setiathome running on this machine since I built it over a year ago, so perhaps that's another indicator that the problem is graphics related. Another side note: I have another computer running on virtuall the same hardware, with which I use the same kernel, etc. X11 does seem to lock up from time to time on this machine, but I can still ssh in and kill/reset processes as needed. Thanks again! Matt -- Matt Garman email at: http://raw-sewage.net/index.php?file=email -- gentoo-user@gentoo.org mailing list
