Re: Routing doesn't start automatically
Bob Nielsen <[EMAIL PROTECTED]> writes: > I'm not that well versed on calculating netmasks and broadcast address, > but these look a bit strange. You might try the standard settings: a netmask of 255.255.255.0 could only be consider "standard" in a LAN environment. the mask in the original message (255.255.255.252) is typical of a DSL environment. it allows 4 addresses on the local network: the user's computer, the router, the broadcast address, and a "network address" (which is never actually used, AFAIK). I just checked on my machine. I've got a file called "/etc/init.d/network" which brings up the interfaces and updates the route tables. it says: # In new Debian installations, this file is deprecated in favour of # the ifup/ifdown commands (invoked from /etc/init.d/networking), which # can be configured from the file /etc/network/interfaces. so I guess I'm using an out-of-date method. -- joe -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: AFS over ipmasq - does it work?
Andrew Perrin <[EMAIL PROTECTED]> writes: > So, two questions: > 1.) Is it possible to run an openafs client via NAT? and no. > 2.) If not, is it possible somehow to re-export an AFS filesystem? I'm > thinking then of mounting AFS on the firwall machine and exporting it to > the other machine via NFS. I don't know about this. note: I'm not an AFS expert. I toyed with it some time ago, and ran into exactly the NAT problem you seem to have run into. I did some googling and other research and found that AFS and NAT are incompatible. that doesn't mean that it's impossible to get working, and if you ask on AFS specific mailing lists (are there any?) then maybe you'll get more information. -- joe -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Alternatives To NFS?
Jason Healy <[EMAIL PROTECTED]> writes: > Yes. AFS allows for NFS-like connectivity and mounting (some large > universities use it for home directories). AFS servers and clients > are available in Debian (though I've never gotten around to setting > one up). I have. it was not simple, and not at all for the faint-hearted. however, some kind soul has produced a transcript of an AFS configuration session and included it in the openafs-client Debian package. with the help of the transcript I was able to get things moving fairly quickly. without it it would have taken me, well, a really long time to even begin to understand what to do. when you consider that AFS is meant to be secure, and that the largest part of security in any system is the human administration of the system, it's clear that an administrator fumbling his way through an installation, blindly following a script he doesn't understand is not a recipe for good security. which isn't meant to be a knock on AFS. just to illustrate that an AFS is probably more suited to a department level organization than a single user "convenience" system. -- joe -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Ethereal rights
> "Jeff" == Jeff Johnson <[EMAIL PROTECTED]> writes: Jeff> are there any more permanent solutions so this doesn't have to Jeff> be done every time i wish to run ethereal? one way would be to use the XAUTHORITY enviroment variable, which names the file in which authorization info is kept. if the X server was always run by the same user (typical in a workstation), then you could have something like this in one of root's shell init scripts: export XAUTHORITY=/home/notroot/.Xauthority export DISPLAY=:0.0 I think that this should work, but you should be aware that this sort of thing can also cause weird problems. say you login to the machine via an ssh connection with X forwarding enabled, su to root, and run an X program. not only will it not display on the initial host, but it might even end up displaying on the workstation display, assuming you're logged in there also. maybe the best thing to do is to make a script that root can use to run ethereal that has the commands above in it, then runs ethereal. -- joe -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Ethereal rights
> "Jeff" == Jeff Johnson <[EMAIL PROTECTED]> writes: Jeff> I try to capture traffic with ethereal and I dont have the Jeff> permissions for the device. sou i try to run it from a root Jeff> console and X cannot open display. one way of doing this is * as non-root owner of the X session, do xauth extract /tmp/auth :0.0 * as root, do xauth merge /tmp/auth export DISPLAY=:0.0 ethereal note that if the /tmp/auth file is publicly readable, then anyone on the local system can hijack your X server. -- joe -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: this post is not off-topic
> "David" == David Wright <[EMAIL PROTECTED]> writes: >> s/not\s+//; David> I appreciate the good-natured jibe. I didn't think the analogy David> to the Debian release process was so far-fetched, but it David> appears that it is. I'll admit to being one (of many, probably) who read the first sentence of your original post and decided that "this post *is* off-topic", having mysteriously landed here after being misdirected from an epidemiology list. David> I never understood people who claim that to relase Woody for David> mainstream architectures (essentially i386 and PPC) before David> releasing it for non-mainstream architectures would make the David> non-mainstream architectures "second class citizens." has it really been demonstrated that HPPA (the designated whipping boy, judging from other posts I've read) significantly holds up releases? I personally wouldn't know, not being at all involved in the process. David> I was hoping someone who takes this position would either David> explain why my ananlogy fails or explain why we really should David> spend on all 11 diseases equally, even though this does not David> help the most people that we can. it fails because it appears to be based on a false premise: that port specific bugs are significantly holding up the release process. then again, whenever someone starts moaning about the outdatedness of debian, someone usually counters with "Don't you know that Debian has to support X platforms? That's not easy!" it seems common-sensical that supporting more platforms is going to be harder than supporting fewer platforms. if it could be shown that the increasing number of ports has made Debian *overall* less useful to people, then your analogy would make more sense. -- joe -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: wrapping [was: Re: disable paragraph flows in mozilla?]
> "Richard" == Richard Cobbe <[EMAIL PROTECTED]> writes: Richard> Lo, on Saturday, May 18, Hans Ekbrand did write: >> On Fri, May 17, 2002 at 03:40:47PM -0700, Vineet Kumar wrote: >> > The reason most people suggest 72 is that traditionally, >> > terminals >> >are 80 characters wide, and 72 leaves enough room to be quoted >> >with "> " four times. Richard> That's one of the reasons I like VM and Gnus. They run in Richard> (X)Emacs, and fill-paragraph-or-region (M-q) is almost Richard> always smart enough to get the quoting brackets right when Richard> it refills a paragraph. also, when reading a non-or-improperly wrapped article in Gnus, the summary command 'W w' can be used to wrap the article. -- joe -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: safe load average
> "Colin" == Colin Watson <[EMAIL PROTECTED]> writes: Colin> The load average refers to the average number of processes Colin> that are runnable or in uninterruptible sleep. The latter Colin> usually indicates I/O. I did not know that. still, processes in uninterruptible sleep are certainly waiting for short-term I/O, e.g. from a disk, correct? 50 webservers are more likely to be blocked on socket reads and/or accepts, which would take them off the run queue. -- joe -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: safe load average
> "Kirk" == Kirk Strauser <[EMAIL PROTECTED]> writes: Kirk> "Safe"? I think you should be more interested in CPU states Kirk> than load average. For example, consider running 50 webserver Kirk> processes, all of which are in an I/O wait state. Your load Kirk> average may be near 50, but your CPU may be sitting mostly Kirk> idle. As another example, pretend you're running an RC5 or no. "load average" refers to the average length of the run queue. processes blocked on I/O are not on that queue. to the OP: there's no such thing as a "safe" load average. however, higher than normal load averages might indicate something wrong, runaway processes or a misconfigured package or something. "normal" depends on the context: my development workstation rarely rises above 1, but my ISP's shell server is often >10. -- joe -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Emacs advice,what to install?
> "Elizabeth" == Elizabeth Barham <[EMAIL PROTECTED]> writes: Elizabeth> emacs is simply a great editor. A lot of people are put Elizabeth> off at first by it's complexity and memory foot-print size Elizabeth> but once you get the hang of it, vi and other editors Elizabeth> become a real drag to use. OH MY GOD RUN FOR THE HILLS -- joe -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: maintenance mode in woody
> "Marcelo" == Marcelo Chiapparini <[EMAIL PROTECTED]> writes: Marcelo> Hello! Marcelo> how can I enter the maintenance mode in woody at boot time, Marcelo> for example for run e2fsck? TIA Marcelo -- Marcelo Marcelo> Chiapparini DFT-IF/UERJ [EMAIL PROTECTED] as others have mentioned, you can use the "single" switch when booting to get to single user mode. this should be enough for you. if things are really hosed, you can try the "emergency" switch, which gets you to a prompt faster than "single", and w/o running any init scripts. -- joe -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Which window-manager for a kiosk?
> "Kent" == Kent West <[EMAIL PROTECTED]> writes: Kent> Any suggestions? sawfish. it's small, doesn't have a "start menu" type thing, and I think will place windows for you. it does pop up a menu when a certain key is pressed ("mouse-2"??) on the root window, which could allow for the launching of new programs. but it shouldn't be too hard to disable that, or strip the menus of any dangerous entries. you can usually configure the size and geometry of X programs by passing a "-geometry" argument on the command line. galleon or mozilla or whatever may not honor this, though. you might also want to think about disabling the normal control key sequence which can shutdown X. maybe you can change it to a secret key combination which shuts X down. I don't know how this is done. -- joe
Re: NEWBIE TIP #110 [was Re: suggestion[data in .sig file]]
> "Hans" == Hans Ekbrand <[EMAIL PROTECTED]> writes: Hans> Since no one else has disputed this post yet, I think it is Hans> time to do so. I have used X-forwarding over SSH enough to know Hans> that you need not and you should not set $DISPLAY manually. no, you don't need to set DISPLAY. but you said: This tip is bad. It does not work. The first line makes the following fail (or, I think, in case of bad security on client succeed but by-pass the ssh-tunnel). which is incorrect. the tip in question was this: > DEBIAN NEWBIE TIP #110 from Dimitri Maziuk <[EMAIL PROTECTED]> > : > Here's how to TUNNEL SECURE X11 CONNECTIONS THROUGH SSH: on the > client, do this: > client# export DISPLAY=client:0.0 > client# ssh -X server > then once you're logged in at the server, do: > server# netscape & > The environment created at the server will include the DISPLAY > variable, so netscape (or whatever) will dialogue with the > client machine. (See "man ssh" for more.) you claimed that the tip above was wrong. I think you were suggesting that setting the DISPLAY variable would bypass ssh. this would be correct, *if* you were setting DISPLAY on the remote side of the connection. but that's not what the tip above is doing. it's setting the DISPLAY variable on the local side, which is usually not necessary, but is probably in the tip as a misguided attempt at clarity. the tip is *not* setting it on the server side. get it? -- joe
Re: question about /bin utils
> "Michael" == Michael Marziani <[EMAIL PROTECTED]> writes: Michael> While I was replacing my vi with vim, I noticed that there Michael> is no /bin/vi at all, it's in /usr/bin. I've never seen a Michael> distro without a /bin/vi; how do I edit my files when my Michael> /usr partition crashes? ed! it's the standard! text editor! -- joe
Re: NEWBIE TIP #110 [was Re: suggestion[data in .sig file]]
> "Hans" == Hans Ekbrand <[EMAIL PROTECTED]> writes: Hans> This tip is bad. It does not work. The first line makes the Hans> following fail (or, I think, in case of bad security on client Hans> succeed but by-pass the ssh-tunnel). no, it works as expected. if the tip had been client> ssh -X server server> export DISPLAY=client:0.0# DON'T DO THIS!!! server> netscape& then that would have been bad. but that's not what was in the post. -- joe
Re: fork: Resource temporarily unavailable
> "Pete" == Pete Harlan <[EMAIL PROTECTED]> writes: Pete> Does anyone know how to increase the number of allowed Pete> processes? if you are running a 2.2 series kernel, you have to recompile the kernel after tweaking a header file. it's been a while, but I think you only have to change one line in include/linux/tasks.h. mine now looks like this: /* * This is the maximum nr of tasks - change it if you need to */ #ifdef __SMP__ #define NR_CPUS 32 /* Max processors that can be running in SMP */ #else #define NR_CPUS 1 #endif #define NR_TASKS2048/* On x86 Max 4092, or 4090 w/APM configured. */ #define MAX_TASKS_PER_USER (NR_TASKS/2) #define MIN_TASKS_LEFT_FOR_ROOT 4 the NR_TASKS symbol was oringally set to 512, I think, meaning a max of 256 processes per user. the 2.4 kernels work differently. I have no experience with these. -- joe
Re: Sharing network from VMWare client
> "C-Cose" == C-Cose Masters <[EMAIL PROTECTED]> writes: C-Cose> My questions are then: C-Cose> 1. Would I be able to change the vmnet IP to something in the C-Cose> 10.x.x.x range? Assuming that would also involve netmask C-Cose> changes, what would they be? you can do this, although I can't recall the terminology vmware uses. "bridged networking" springs to mind, but I could be wrong. at any rate, it's the other of the two options from the one you are using. but, ... C-Cose> 2. Would it be easier to us either forwarding or masquing to C-Cose> allow traffic from vmnet1 access to the net through eth0? IMHO, yes. this is what I do, and it works great. it means you don't have to worry about getting a "real" address from your network people. C-Cose> 3. I have none of the "advanced" net packages included: C-Cose> ie. ipmasq, ipchains / tables. Which of these would I need. I think with 2.2 kernels you can use ipchains, which appears to be in it's own "ipchains" package. the downside is that you have to have a kernel that has a few options turned on which are not I think normally on. there is a very well written HOWTO on firewalls and masquerading around somewhere that explain how to do this. once the kernel is setup, and you've got ipchains installed, you just do # ipchains -A forward -s 192.168.46.0/24 -j MASQ you will get a warning about enabling forwarding by twiddling a file in /proc somewhere. do that, and you're on the net. -- joe
Re: C Integrated Development Environment
> "johnpf" == johnpf writes: johnpf> Actually, as much as it shames me to admit it, there is one johnpf> feature in the VC6 M$ bloat thing IDE I really want to see on johnpf> a Unix platform, and that's the incredibly powerful way it johnpf> can back reference callers, classes etc. Perhaps it's me, but johnpf> I just can't get ctags etc to work the same way, and my LISP johnpf> is too weak to make emacs do it. take a look at LXR. works great for C, and also (I think) for java. -- joe
Re: Mozilla fonts huge--'File' menu item takes up most of screen
> "Daniel" == Daniel Farnsworth Teichert <[EMAIL PROTECTED]> writes: Daniel> When I start up Mozilla, the fonts are *huge*. I'm running at Daniel> 1600x1200, and the 'File' menu item takes up the majority of Daniel> the screen (nice scaling, by the way--not a bit blocky : ). no help here, but I'm getting the same behavior. this started after a recent unstable dist-upgrade. mozilla is at 0.9.7-4. it was pleasant at first to see well scaled fonts taking up half my screen. it became annoying quickly, though. -- joe
Re: Mozilla & Address Books
> "Ron" == Ron Johnson <[EMAIL PROTECTED]> writes: Ron> Thanks. Are there any network-based address books that I (using Ron> kmail, but maybe switching to evo1 or mozilla-mail) and my wife Ron> (using Outlook Express, but soon to be going to Linux) can use Ron> to share addresses? Ron> I don't care how had it is to set up, just so long as she can Ron> easily add new names to the address book. I spent some time yesterday getting openldap going for exactly this purpose. I want to share address books between mozilla and pine. while in the middle of the process a possible show-stopper occured to me: can mozilla (and pine) *create* new entries in an LDAP directory? I suspect they cannot, and therefore LDAP is nearly useless to me. anyhow, I'm going to keep going, and see what happens. -- joe
Re: UNSTABLE USERS: dpkg, doc-linux-html, and MD5sum fields
> "Colin" == Colin Watson <[EMAIL PROTECTED]> writes: Colin> I'm getting three or four reports a day about errors like Colin> this: dpkg: parse error, in file `/var/lib/dpkg/available' Colin> near line 88004 package doc-linux-html': empty file details Colin> field `MD5sum' E: Sub-process /usr/bin/dpkg returned an error Colin> code (2) I was bit by this. my 'solution' was to delete the offending MD5Sum line from /var/lib/dpkg/availble. this seemed to work OK for me, but YMMV. -- joe
Re: vmware express
> "Jeffrey" == Jeffrey W Baker <[EMAIL PROTECTED]> writes: Jeffrey> You can also do it the other way: give your vmware machine Jeffrey> two interfaces and a real IP. Then give your debian machine Jeffrey> a private subnet. NAT all the traffic through the VMware Jeffrey> machine. wow. my head hurts just thinking about this. -- joe
Re: vmware express
> "Royce" == Royce Bell <[EMAIL PROTECTED]> writes: Royce> Jeffrey... Have you tried connecting to DSL ~through~ vmware? I do this, sorta. there is a vmware networking option ('host only', I think it's called) whereby the host OS is given an interface on a private network (192.168.x.x). I masquerade the traffic from that network onto my DSL line. works great. There is another networking setup where you give the guest OS a real address. this would also probably work with DSL, as long as you had >1 address you could use. -- joe
Re: What process started/owns another process?
> "Peter" == Peter S Galbraith <[EMAIL PROTECTED]> writes: Peter> I have a lot of these processes on my system: ... Peter> How do I determine what process started them? pstree. it's in the psmisc package. or you could figure it out the hard way by asking 'ps' for the parent PID. the '-f' option to ps will do this, or you can use the '-o' option to ask for specific fields. -- joe
Re: OT: tail -f | while read
>>>>> "Martin" == Martin F Krafft <[EMAIL PROTECTED]> writes: Martin> also sprach Joseph Dane (on Thu, 06 Sep 2001 01:26:37PM Martin> -1000): >> why not just use xargs? >> >> find . -name \*.c | xargs wc -l >> >> I think you can pass '-n 1' to xargs to cause it to execute the >> command a separate time for each argument. Martin> no: xargs(1) --max-args=max-args, -n max-args Use at most Martin> max-args arguments per command line. Fewer than max-args Martin> arguments will be used if the size (see the -s option) is Martin> exceeded, unless the -x option is given, in which case xargs Martin> will exit. right. meaning that if you specify '-n 1', then at most one argument will be used for each invocation. is that not what I said? -- joe
Re: OT: tail -f | while read
> "Martin" == Martin F Krafft <[EMAIL PROTECTED]> writes: Martin> and could someone give me a perl one-liner that takes each Martin> such line fed into its STDIN, and for each line, calls an Martin> external shell script with the entire line as argument? why not just use xargs? find . -name \*.c | xargs wc -l I think you can pass '-n 1' to xargs to cause it to execute the command a separate time for each argument. -- joe
Re: Adding a user to a group
> "Dave" == Dave Sherohman <[EMAIL PROTECTED]> writes: Dave> On Tue, Jul 17, 2001 at 07:50:19AM -0400, Nathan Weston wrote: >> Oops, I meant to post back to the list... is there a reason that >> the list address isn't in the reply-to by default, like it is on >> most lists? Dave> 1) Yes, there is a reason. Do a search on "reply-to considered Dave> harmful" for more information. do a search for 'reply-to munging considered useful', for what I feel is a much more convincing argument. the link where I found it originally, http://www.metasystema.org/essays/reply-to-useful.html appears to be down, but you can probably find it elsewhere. Dave> 2) I dispute your assertion that most lists use reply-to-list. Dave> I am currently subscribed to... um... about 15? lists (not Dave> including the ones that I own) and exactly one of them sets a Dave> reply-to header. really. I'm on about the same number, and the split is more like 40/60 for me, in favor of not setting reply-to. also, most of the lists I admin set reply-to, at the request of the people on the lists. in the end, the 'right' policy is the one chosen by the list admins, since they have the right to set whatever policy they want. but this does seem like an issue that just won't go away, ever. -- joe
Re: MUAs that compare with Outlook (your chance to show how much better Linux is than MS!!)
> "Brian" == Brian May <[EMAIL PROTECTED]> writes: Brian> However, Gnus seems to perform badly (IMHO) when accessing Brian> remote mail: Brian> - on startup, it tries to check for mail on every folder. This Brian> is slow and time consuming (I only have a shared 28.8kbps Brian> Internet connection...). Especially annoying if I only want to Brian> check mail on my local server. one way you might be able to accomplish this is to use 'levels'. you can set a mail (or news) group at a certain level, and when you check for new mail gnus will only fetch for groups at or below that level. the default fetch level is 3 (IIRC), so if you set groups at level 4 they would only be fetched if you explicitly asked for them. I have not tried this, so YMMV. -- joe
Re: max concurrent processes can't be > 257?
> "Nate" == Nate Amsden <[EMAIL PROTECTED]> writes: Nate> Brett wrote: >> I'm running qmail and am trying to set up linux to allow for >> hundreds of outgoing connections at once (no, I'm not a spammer >> but the new admin of some very large, dynamic mailing lists). I'm >> using Debian Linux 2.2.18pre21 and from what I read, it should be >> quite possible to adjust the maximum processes per user through >> 'limit' (or 'ulimit' depending on the shell) rather than adjusting >> and recompiling the kernel. Well, I'm root, I do the adjustments >> to limit ('limit maxproc 1000') but when I check the logs, qmail >> never gets above 257 concurrency (256 is the default limit of >> maxproc). I do a 'limit' Nate> probably because you did not restart qmail *using the shell Nate> with the modified ulimits*. probably you are correct. however, there is another possibility: the limitation on the total number of processes allowed by the kernel. check out the kernel include file include/linux/tasks.h. (this is assuming you are running some sort of 2.2.x kernel.) you will see two lines like the following: #define NR_TASKS2048/* On x86 Max 4092, or 4090 w/APM configured. */ #define MAX_TASKS_PER_USER (NR_TASKS/2) on my debian potato system, the first line was set to something much smaller than the value above. memory may be failing me, but I think it was initially set to 512. which would make MAX_TASKS_PER_USER == 256. I ran into this when I had a Java program that was unable to spawn more than 256 threads. after increasing NR_TASKS, all was well. I also initially suspected ulimit, but found that making everything unlimited did not solve the problem. a recompiled kernel with the above changes did. finally, I am not sure that the above is the "authorized" way of increasing the maximum number of processes in a linux kernel. you have been warned. also, I think that 2.4 kernels do not have this problem at all (or at least, the solution would be different) because they do not have a fixed size data structure to hold task data. -- joe
Re: combo GUI/text mail client
> "Brandon" == Brandon High <[EMAIL PROTECTED]> writes: >> * netscape and pine don't (can't?) share addressbooks, so they >> either get out of sync, or require more attention than one would >> like to offer them Brandon> Probably the ugliest point. You could over engineer it and Brandon> use OpenLDAP to store address books - most modern mail Brandon> clients can do lookups against LDAP. I'm not this Brandon> brave/foolish, and I don't need to be. I have a small Brandon> address book. interesting. I have actually set up openldap a few times, and while it's certainly not for the faint-hearted it didn't seem all that difficult. does mutt do LDAP? -- joe
combo GUI/text mail client
Here's what I'd like to have: a mail client that I can access via a nice, flashy GUI when I'm sitting at my desk, or via a simple text-mode interface when I'm connecting remotely. Actually, it's my wife I'm primarily thinking about. She uses our computer at home (Debian, natch) to read/send mail using Netscape. She also occasionally connects from her parents' house, or from work using ssh and getting a simple shell (text) interface. Right now, she uses pine for this, but this sucks because: * netscape slurps up all read mail into its own 'nsmail' directory, so she can only read new mail * netscape and pine don't (can't?) share addressbooks, so they either get out of sync, or require more attention than one would like to offer them I have been looking for such a beast, without any luck. Gnus works pretty well for me, but as I'm trying to *stay* married, I'm unwilling to inflict Gnus on my wife at this time. Any ideas? -- joe
Re: Comments VMWare?
> "Jonathan" == Jonathan Gift <[EMAIL PROTECTED]> writes: Jonathan> Hi, I might have to run a windows app, and one code copy Jonathan> protected at that. Anyway, how is VMWare at running 32 bit Jonathan> windows apps? as I understand it you load VMWare, then Jonathan> W95/98, then your app. Does VMWare set up its own file Jonathan> system? Where would I store my apps docs? I use vmware, and love it. There are a number of ways to set it up, including allocating a chunk of a disk for it to setup a filesystem on, connecting to a samba server (possibly running on the same machine, in the 'host' linux OS), or (IIRC) using an existing NT filesystem. vmware virtualizes the entire x86 environment. prepare to be amazed as you boot your guest OS, see the familiar BIOS startup screen and POST, and have a fully functional NT desktop, all safely contained within an X Window. -- joe
Re: ssh doesn't stop prompting for password [SOLVED]
> "Ralf" == Ralf G R Bergs <[EMAIL PROTECTED]> writes: Ralf> On Thu, 22 Feb 2001 13:22:53 +0100, Andre Berger wrote: >>> '-X' option. note also that if your .bashrc (or whatever) on the >>> remote machine sets the DISPLAY variable, then this won't work. >> So I have to stop a running X server first, then do the ssh >> command? Ralf> Nope. You can always do the following: local> ssh -X remote remote> DISPLAY="" xlogo No, I don't think you can do that. When you login to the remote machine with X forwarding enabled, the DISPLAY variable gets set to something like 'remote:12.0', which points to a 'fake' X server running on the remote machine (it's actually the ssh daemon) which forwards the X connection back to your local X server. If you cleared the DISPLAY variable the client ('xlogo', in this case) won't know how to contact the X server. Getting back to the original question: no, you do not have to do anything special. If you are running X on the local machine (which I assume you are), then just login to the remote machine using 'ssh -X remote' and run your X clients. -- joe
Re: ssh doesn't stop prompting for password [SOLVED]
> "Andre" == Andre Berger <[EMAIL PROTECTED]> writes: Andre> Talking about ssh, could somebody point out the steps Andre> necessary in order to export a remote host's display to my Andre> local machine via, if possible from scratch...? Sorry if this Andre> is a stupid question. do you mean that you want to run program on the remote machine and have them display on your local machine? if so, you should be able to do local> ssh -X remote remote> xlogo there are configuration options you can set for ssh to have it automatically forward X connections, so you don't have to specify the '-X' option. note also that if your .bashrc (or whatever) on the remote machine sets the DISPLAY variable, then this won't work. -- joe
Re: About Debian documentation
Others have already mentioned how one can go about tracking more recent versions of packages. I will add that many people (well, at least one person: me) would much rather have a system which is "known good" than one with the latest versions. I personally have no need for XF4, so I'll wait until I'm reasonable certain that everything I've got now will continue to work with it. For me, this means I wait until 'stable' to upgrade. Which is not at all to say that people shouldn't be able to get more recent versions of packages. Debian wouldn't be nearly as successful as it is if you couldn't decide to live on the bleeding edge. But the general philosophy, as far as I can make it out, is conservative, and means that on a stable system you will usually be a few versions behind on most packages. -- joe
Re: exim
> "Pap" == Pap Tibor <[EMAIL PROTECTED]> writes: Pap> Try telnet to localhost port 21 to see if exim is running. Since when does exim listen on port 21? sweden [~]>grep 21 /etc/services ftp 21/tcp fsp 21/udp fspd ... SMTP (and exim) are on port 25. -- joe
Re: Relation(exim,fetchmail,mutt)=?
> "Timmy" == Timmy Douglas <[EMAIL PROTECTED]> writes: Timmy> mutt reads /var/spool/mail/blah etc. Timmy> mutt pipes sent mail to /usr/lib/sendmail etc. Timmy> /usr/lib/sendmail etc. pipes to exim exim sends mail to Timmy> another host exim is more or less a replacement for sendmail. one would not generally have both exim and sendmail installed on the same system. in fact, on my debian system, sendmail looks like this: sweden [~]>ls -l /usr/lib/sendmail lrwxrwxrwx1 root root 12 Aug 14 23:16 /usr/lib/sendmail -> ../sbin/exim* -- joe