Re: [leaf-devel] Command recall Oxygen(yes) Bering(no) DF(no)
Nathan Angelacos wrote: Well, command recall is my wish for Bering rc4. >>Matt Sigh. Me too! It pains me to realize that I've been using command recall in Ox, DF, and BF for a long time now. The up arrow works great. My apologies for losing my mind. The reason is Oxygen uses BusyBox shell, and Bering uses "real" ash. Last time I tried to get bb shell to work with bering, there was a problem with initrc (can't remember specifically where in the script, but bering initrc seemed to want something that busybox couldn't understand, and it caused the startup to fail.) Thanks for the explanation. I guess what I'm really missing in filename/directoryname completion. Best, matthew --- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf ___ leaf-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/leaf-devel
[leaf-devel] Command recall Oxygen(yes) Bering(no) DF(no)
One reason I used to use Oxygen a lot is that it has command recall and filename/directoryname completion as part of it's shell, which I think is ash. But I don't get what's different on the other LEAFs. Don't people use the same shell? Also it has a a working $PWD. Well, command recall is my wish for Bering rc4. Not enabled by default of course :) Matt --- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf ___ leaf-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/leaf-devel
[leaf-devel] Help floppy and man.lrp
Has a help floppy ever been proposed? How about it? Put all our docs and guides and whatnot on it. Then if the user is stuck he just puts in the help floppy and types #@$#$% into the console :> Well, maybe just a docs.lrp package. And how about a man.lrp package? I think most of us have enough ram to load those as part of our supplemental diskette 2. At first I thought it'd be a good idea to load man pages over the internet from our home base at sourceforge, but then I realized that would exclude people with routing problems. regards, matt --- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf ___ leaf-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/leaf-devel
[leaf-devel] SF LEAF docs
Heya can we do a little something with the way the documentation is layed out on the web site? Things are odd somehow, but I can't pin it down. Here's a few examples. 1. On the FAQs page held in SF Doc Manager: http://sourceforge.net/docman/?group_id=13751 the document titles are the same size font as the Descriptions. It makes me go hmmm that I can see a doggy poop in my yard easier than I can read those titles. Could you plz a. increase the font on the FAQ subsection Titles to like an H3 b. increase the font on the document titles themselves to something larger than the descriptions c. indent the Description lines ??? I hope this problem isn't my browser, Moz 1.1 :) 2. Why can't we put all the docs on one page? When someone clicks Documentation, can they go to a page that shows the Guides and HowTos on the top of the page and the SF Docmanger FAQs below them? Let me guess... no :) ___ |D O C U M E N T A T I O N | | | | | | 1. Guides | | 2. HowTos | | 3. Faqs | | | | | | -| | | | 1. G U I D E S | |1.1 Blah | |1.2 Bar | | -| | 2. H O W T O S | |2.1 Foo | |2.2 Bang | | -| | | | 3. F A Qs | |...| Or why can't we submit the Guides and Howtos to the Doc manager and just get everything in one place, perhaps there be a subsection of the FAQs for Guides and another for HowTos? 3. When you click documentation in the left hand pane of the home page, then click FAQs, the page that comes up http://leaf.sourceforge.net/mod.php?mod=userpage&menu=1300&page_id=9 is confusing because the link to the docs doesn't stand out some how. I don't know why this intermediate page exists, but maybe just change the background to yellow around the link to emphasize it, or make it larger, or make it say something more apparent such as: The FAQs are stored here in the docmanager I dunno. Been liking the way weblet uses bg colors. Matt --- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf ___ leaf-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/leaf-devel
[leaf-devel] bering users manual command anomaly
http://leaf.sourceforge.net/devel/jnilo/buwireless.html There's a typo below. The broadcast addy didn't turn out to be what you inteneded :) 10.6 Step 5: configure your interfaces file Trough the LEAF configuration menu type 1 to access to the network configuration menu and 1 again to edit your /etc/network/interfaces file. Enter the following information: auto lo eth0 iface lo inet loopback iface eth0 inet dhcp iface eth1 inet static address 192.168.1.254 masklen 24 broadcast 255.255.255.0 == regards, matthew --- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf ___ leaf-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/leaf-devel
Re: [leaf-devel] RFC: Oxygen Network Boot HOWTO
H. D. Lee wrote: > David, List, > > I have written a howto on booting Oxygen from network. Only with NIC > and boot ROM on the router, it should boot from tftp and dhcpd. The > packages also loaded by tftp. The whole process made easy with David's > Oxygen scripts that have been made network boot friendly. Thanks David. > > http://www.leaf-project.org/devel/hdlee/oxygen/doc/netboot/network-boot-howto > http://www.leaf-project.org/devel/hdlee/oxygen/doc/netboot/oxygen.cfg > http://www.leaf-project.org/devel/hdlee/oxygen/doc/netboot/std.conf > > I hope it would of use to someone. The document is still in a very > "alpha" stage. Comments and suggestions are welcomed. > > Thank you in advance. > > Regards, Neat. It'd be nice if you added a .txt extension to the file so that clicking on the link displayed the HOWTO in Mozilla rather than defaulting to downloading a binary/octet stream mime type. Also, when you do look at it in Mozilla, you can see the first few paragraphs weren't formatted with any carriage returns so that the lines are about 400 characters wide and we have to scroll sideways. Beyond asthetics, it'd be helpful if you said a few words about mknbi-linux which is where a lot of the meat an potatos happens. I'd not heard of it until seeing you use it. And if you could comment somewhere on the changes you made to oxygen.cfg that makes it load pacakges over the network. And finally, as far as useability goes, how do you back stuff up? I liked it, matthew --- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf ___ leaf-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/leaf-devel
Re: [leaf-devel] Zebra package development status.
Eric B Kiser wrote: > Howdy all, ... > Then I had a revelation. I should be able to compile the individual protocol > as independent *.lrp's that way you can go in and just load the protocols > that you need. NOTE: I have not tested this it is merely conjecture at this > point. For now I am going to proceed forward with a basic proof of concept > that will include zebrad, bgpd, ospfd, and ripd. When I have this working > then I am going to put my efforts toward creating and testing the individual > packages. > > Regards, > Eric Kiser When I'm working on something like that, I find that ls -ltu is invaluable. regards, matthew --- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf ___ leaf-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/leaf-devel
Re: [Leaf-devel] Updated Oxygen Development Image - and lrp/ sourcedirectory
"Angelacos, Nathan" wrote: > > >If people could bang away on the Oxygen development image, I'd > >appreciate it. I'd like to hear from some testers before I release... > > Finally got a chance to download (and test), so far looks good. The > config.lrp handling works nicely now. Thanks! > > As an observation, /dev/boot seems to be a permanent symlink to > /dev/fd0u1680, even when booting from /dev/hda1 (I had globally replaced > fd0u1680 /w hda1 in oxygen.cnf) But I probably missed something somewhere > on changing that (RTFM, right?) > > So far the new image looks good. I tried it too, and the dhcpcd on eth0 didn't work for connecting to the new attbi.net replacement for @Home. Works on Dachstein rc2, but the system hangs during boot on Oxygen 12/8. Matthew ___ Leaf-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/leaf-devel
[Leaf-devel] Need Steinkuehler localtime for PST8PDT
Apparently, the time control files for Oxygen are causing problems, and my localtime file is a UTC one. I'd like to try the PST8PDT timezone file from CS, but he's taken them off his website. Anybody know where to get one of these? Thanks, Matthew ___ Leaf-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/leaf-devel
[Leaf-devel] Oxygen boot problems with fd1
In testing the latest development Oxygen-090601, I've found that I can't get packages to load from fd1 when I put the correct statements into large.cnf, for instance. I've found that the boot program won't recognize some of the commands listed in Oxygen.cnf. mountpoint /mnt/fd1 mount -t msdos /dev/fd1u1680 /mnt Both return an error that they are unknown commands. Even when I give it the correct device /dev/fd1u1680 filesystem msdos it can't stat the filesystem and returns a lot of errors reading the diskette. This diskette loads fine if I tell it to use /dev/fd0u1680. Any ideas David? Matt ___ Leaf-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/leaf-devel
Re: [Leaf-devel] Echowall on Oxygen needs ipmasqadm and ipfwd
Matthew Schalit wrote: > > I guess I need ipmasqadm and ipfwd for Echowall > to work on Oxygen, but I can't find them in > .lrp format. Are they compiled, David? > > Thanks, > Matt I found ipmasqadm, but not ipfwd. I couldn't even find ipfwadm as a .lrp package any more. If I had that, I could continue running rc.pf, and not have to deploy echowall or seawall. Thanks, Matt ___ Leaf-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/leaf-devel
[Leaf-devel] Echowall on Oxygen needs ipmasqadm and ipfwd
I guess I need ipmasqadm and ipfwd for Echowall to work on Oxygen, but I can't find them in .lrp format. Are they compiled, David? Thanks, Matt ___ Leaf-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/leaf-devel
Re: [Leaf-devel] Oxygen default gateway problems
Matthew Schalit wrote: > > David, > On oxygen-090401, the default gateway is not > getting entered into routing table, even though > the data is correctly entered into /etc/network.conf. > > Thanks, > Matt Strike that. The default gateway loads fine. Thanks, Matt ___ Leaf-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/leaf-devel
[Leaf-devel] Oxygen /etc/init.d/sysklogd broken
David, In the Oxygen-090601 image, the script /etc/init.d/sysklogd has something in the RCDLINKS line that breaks RCS, and the script doesn't get copied to rcS.d, rc1.d, or any other rc.? directory. I removed all the #,K16 statements from the RCDLINKS line, saved, rebooted, and now the script works. What I did was to set: RCDLINKS="2,S00 3,S00 4,S00 5,S00" Thanks, Matt ___ Leaf-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/leaf-devel
[Leaf-devel] Oxygen default gateway problems
David, On oxygen-090401, the default gateway is not getting entered into routing table, even though the data is correctly entered into /etc/network.conf. Thanks, Matt ___ Leaf-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/leaf-devel
[Leaf-devel] Oxygen timezone problems
David, When using Oxygen-090601, I've been setting my /etc/timezone file to make TZ=PST8PDT for myself in California. That worked, I thought. Now what I see when I run the date command is: Thu Oct 25 16:18:22 ??? 2001 instead of Thu Oct 25 16:18:22 PDT 2001 In addition to that, the local clock doesn't seem to be repsecting the timezone, and the time I get back from my local time server on my network appears shifted by 8 hours. I know we worked on this, and I looked at the /etc/init.d/hwclock.sh and /etc/rc.config.d/hwclock.sh files, but I think you've changed what we had working in the oxygen-051401. Thanks, Matt ___ Leaf-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/leaf-devel
[Leaf-devel] Oxygen ipmask and /etc/init.d/network broken
David, The recent updates to /etc/init.d/network that you made include a function call to mask() which calls your new ipmask program. That program gets the CIDR wrong, and consequently the network comes up incorrectly. We originally used my masktocidr() function. I'm using oxygen-090601. Thanks, Matt ___ Leaf-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/leaf-devel
Re: [Leaf-devel] gatping segfault in Oxygen
David Douthitt wrote: > > "Scott C. Best" wrote: > > > Heya. Actually, yea, gatping is a play on Gatling. :) > > It essentially sends out 256 minimum-sized icmp echo requests, > > ignoring the responses, so as to get a freshened-up ARP > > table. > > In compiling gatping (0.2), I get lots of interesting things :-/ > [snip] Thanks for compiling that David, but you left out the debugging symbols, so it didn't help when it segfaulted just now. I couldn't debug it. It fails in the fourth function call, whatever thatis. There's other bigger problems, though in the other thread. Thanks, Matt ___ Leaf-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/leaf-devel
Re: [Leaf-devel] gatping segfault in Oxygen
David Douthitt wrote: > > Matthew Schalit wrote: > > > Not so big a deal as that, David, but still, > > gatping was compiled against 2.0.7. > > Shouldn't matter. glibc 2.0.7 binaries run against glibc 2.1.3 all the > time. Ok, I just guessed, because it segfaulted and didn't have any debugging symbols, so I'm asking someone to compile it with symbols, and I'll run inside gdb. > > Could you compile if for me agains 2.1.3? > > Sure. Where do I find it? > > Is it related to Gatling Guns? :-) Scott Best posted gatping.c here in this thread, and I've put it on my ftp site, in case you want: ftp://ftp.schalit.net/pub/Lrp/Oxygen/Packages/gatping.c Thanks. Matt ___ Leaf-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/leaf-devel
Re: [Leaf-devel] gatping segfault in Oxygen
David Douthitt wrote: > > Matthew Schalit wrote: > > > Gatping, Dave... gatping. > > I HAVE to tell this story - the first time I saw this message thread, I > about had a heart attack :) > > I initially misread it as "gaping segfault in Oxygen" :) Not so big a deal as that, David, but still, gatping was compiled against 2.0.7. Could you compile if for me agains 2.1.3? I don't have any dev platform around to do that. Thanks, Matt ___ Leaf-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/leaf-devel
Re: [Leaf-devel] gatping segfault in Oxygen
"Scott C. Best" wrote: > > Matt: > > Heya. Source code attached. If you could send > back to me a compiled working versionI'd be thankin > ya. :) > BTW, yes, echoWall has gatping built for ES2B > only, glibc-2.0 (I believe). > > -Scott I think I see that I'll have to get Dave's help to compile gatping for a libc-2.1.3 based Oxygen. Can you just see him hunched over his little PDA in a dark corner of the study? The back's probably off the thing, and he's looking for a way to dock it into his router. He's tryin to do some sort of Forth-Perl-DynamicKernelModule-viaWavelan-ssh-pbm2gif nob. Gatping, Dave... gatping. ___ Leaf-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/leaf-devel
Re: [Leaf-devel] gatping segfault in Oxygen
"Scott C. Best" wrote: > > Matt: > > Heya. Source code attached. If you could send > back to me a compiled working versionI'd be thankin > ya. :) > BTW, yes, echoWall has gatping built for ES2B > only, glibc-2.0 (I believe). > > -Scott Hi Scott, Thanks I got the archive. It has gatping gatping.c in it. I don't have the setup to compile it, though. Do you? How was the gatping in the archive you sent compiled? Thanks, Matt ___ Leaf-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/leaf-devel
[Leaf-devel] gatping segfault in Oxygen
Is there a way to code around gatping? It's stopping me from deploying echowall. If someone has a few minutes, make me debug version of gatping I can run through gdb. As usual, mine has no symbols. It's probably a library mismatch. I did an ldd and it it found the three libs. I forget which. Sorry it's not booted. I think this latest Oxygen uses glibc-2.1.3 or something. Thanks, Matt ___ Leaf-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/leaf-devel
[Leaf-devel] Oxygen-090601 issues
David, I tried to install the 090601 image choosing the 'setup' option at the boot: prompt. I used the easy-072601 image as the setup disk. It hangs during boot right after I see this on the screen: running rcS I found that if I hit ^C a few times, that it will finally finish loading. But it didn't run any setup, just put me into acfg. I didn't know there was a setup program. So I just used the acfg stuff. Then I happend to type setup and went through that. The acfg option to setup the Network settings works fine. The acfg option to modify the System Settings just takes me right back to the Network Settings. There's no ipchains. What should I use? Or should I just now create my own second disk out of the packages that you have at http://leaf.sourceforge.net/pub/oxygen/ Thanks, Matt ___ Leaf-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/leaf-devel
Re: [Leaf-devel] Current Development Image available
David Douthitt wrote: > > I've created a directory for Oxygen development images. The URL is > > http://leaf.sourceforge.net/pub/oxygen/development > > I would very much like people to try it out. The current status of > development goals and desires is: > > PRIOR TO RELEASE > (with priorities: 1-high 9-low) > > * (1) Fix setup scripts to work completely, and perform live tests David, I just saw your post and tried the two images. The same thing happened with those two that happened with the 071201 series, it hangs on my system right after it says it's going to run the rcS. I had not other errors show up before that point. No tcpd(err!) this time. I have no CD, no IDE, no serial. Just the two floppies, and three nics (pro100, pro100, 3c509b), p150, 64MB. I let it sit there for at least 10 minutes, so I don't think it's dns related. Perhaps one of the first init scripts is hanging. Regards, Matthew ___ Leaf-devel mailing list [EMAIL PROTECTED] http://lists.sourceforge.net/lists/listinfo/leaf-devel
Re: [Leaf-devel] New User Interface Tool: Breakthrough!
Ray Olszewski wrote: > > At 08:37 AM 7/13/01 -0700, [EMAIL PROTECTED] wrote: > >On Fri, 13 Jul 2001, David Douthitt wrote: > > > >> Hilton Travis wrote: > >> > >> > Are those the 8" floppy disks? > > > >I think the 8" floppy disks never held more than 250KB at most, anyway. > > Actually, Jeff, us old timers remember them growing to the *huge* (at the > time, anyway) capacity of 1.2 MB. I ran my first real CP/M compiler (Digital > Research Pascal, I think) on a dual-8 system. > > -- Ah yes, the joy of loading Wordstar on a CP/M off of an 8" diskette and fighting through that revolutionary user interface of theirs. Matt ___ Leaf-devel mailing list [EMAIL PROTECTED] http://lists.sourceforge.net/lists/listinfo/leaf-devel
Re: [Leaf-devel] seg faults in busybox routines in most recent oxygen
David Douthitt wrote: > > Mark Seiden wrote: > > > as an experiment, i went back to the /lib/lib* from nov 2000 > > oxygen. tar doesn't seg fault, ssh mkhostkey doesn't seg fault, > > but more still does, as before. it's mystifyin'. > > There was briefly some talk about whether libraries get corrupted when > root is backed up; can you verify or disclaim this? David, I just found that problem. 1) I burned a fresh 1.68 3/26 diskette. I modified the default boot down to 8192 MB ram I boot single disk and configure + setup + passwds + /var/boot/modules/* I loaded ldd, strace, ltracc. I ran a couple of ldd's on more and whatnot. I > /var/log/syslog I ran apkg -e (Everything) I rebooted. Nothing odd in dmesg that I see. System boots, I log in, /proc is empty. Nothing really works. ifconfig, route, etc. > Only thing I can think of is that the libraries are probably open when > they are "copied" into the tar file, but that should not matter - should > it? The other thing would be if they are self-modifying, but they > shouldn't be that either - should they? Shared libraries don't modify themselves, if that's what you're wondering, but then again, you know that. No idea about the open file issue and tar issue. Matt ___ Leaf-devel mailing list [EMAIL PROTECTED] http://lists.sourceforge.net/lists/listinfo/leaf-devel
Re: [Leaf-devel] New release of Oxygen (March 2001)
George Metz wrote: > > (Lets face it gang, ain't no UNIX browser out there > that matches up to IE5, no matter how much I want Mozilla to succeed) IE5 is p@@p. I happily run Netscape Communicator 4.61 w/128 bit strong encryption on UnixWare 7.1.1. You gone to the Dark Side, huh? Boo. George="$DarkSide" Boo. George works for M$... I never would've guessed. Sigh. ___ Leaf-devel mailing list [EMAIL PROTECTED] http://lists.sourceforge.net/lists/listinfo/leaf-devel
Re: [Leaf-devel] seg faults in busybox routines in most recent oxygen
Mike Noyes wrote: > > Mike Noyes, 2001-03-27 12:59 -0800 > >Matthew Schalit, 2001-03-27 12:47 -0800 > >>Found it, thanks. Either you touched the dates funny > >>or sourceforce is having a 2000/ 2001 problem. > >>gdb.lrp is listed as > >> > >> gdb.lrp 573 KbMon Mar 27 13:27:00 2000 > >> > >>Same date on develop.ima > >> > >>I guess that's a cc: Mike Noyes > > > >Matt, > >SourceForge isn't having a problem with dates. see. > > > >http://leaf.sourceforge.net/pub/oxygen/ > >gdb.lrp 27-Mar-2001 05:27 573k > > Matt, > There is a slight problem. David uploaded the file with the permissions set > to 600. No one can download it until this is corrected. :( Mike, Upon further investigation: The one I was looking at was at the ftp site leaf.sourceforge.net in /pub/leaf/oxygen and when I checked it at 13:13 PST it said, -rw-r--r-- 1 37002 users1720320 Mar 27 13:27 develop.ima and the problem was probably that their xntpd is not working because you notice that their time stamp is 13:27, but the real time right now is only 13:19. So I think I'm right, rather than it's my client like I just thought. Matt ___ Leaf-devel mailing list [EMAIL PROTECTED] http://lists.sourceforge.net/lists/listinfo/leaf-devel
Re: [Leaf-devel] seg faults in busybox routines in most recent oxygen
Mike Noyes wrote: > > Matt, > SourceForge isn't having a problem with dates. see. > > http://leaf.sourceforge.net/pub/oxygen/ > gdb.lrp 27-Mar-2001 05:27 573k > You're right, it was a problem with my ftp client. Thanks, Matt ___ Leaf-devel mailing list [EMAIL PROTECTED] http://lists.sourceforge.net/lists/listinfo/leaf-devel
Re: [Leaf-devel] Re: Oxygen Problems
David Douthitt wrote: > > Matthew Schalit wrote: > The new March 2001 release uses the WORKING format, as you described it. Thanks for that. I found another issue, and I thought I'd tack it one here, though, because it's serious, but not critical. Apparently, in 3/26 you changed /etc/fstab so that the /boot --> /mnt specification is now commented out and we can no longer type mount /mnt from a prompt and mount the default diskette devicel, while the directions still say that's possible. Of course, removing the comment and backing up etc is enough to fix it, but I thought I'd mention it. Matt ___ Leaf-devel mailing list [EMAIL PROTECTED] http://lists.sourceforge.net/lists/listinfo/leaf-devel
Re: [Leaf-devel] seg faults in busybox routines in most recent oxygen
David Douthitt wrote: > > I updated develop.ima -- just in case it's been updated; it almost NEVER > changes -- and put gdb.lrp there as well. You'll have to use the URL to > get there (there's no links): > > ftp://shell1.sourceforge.net/pub/leaf/oxygen/gdb.lrp Found it, thanks. Either you touched the dates funny or sourceforce is having a 2000/ 2001 problem. gdb.lrp is listed as gdb.lrp 573 KbMon Mar 27 13:27:00 2000 Same date on develop.ima I guess that's a cc: Mike Noyes > > "more" appears to segfault in termios support > > > > tar in malloc. > > Hm I wonder if this would be worth sending to the busybox list? I'd just like to know how Mark found where it was segfaulting. I tried strace and ltrace, but I don't know how to use them. I got somewhere, but it didn't jive with what Mark found. Cu, Matt ___ Leaf-devel mailing list [EMAIL PROTECTED] http://lists.sourceforge.net/lists/listinfo/leaf-devel
Re: [Leaf-devel] seg faults in busybox routines in most recent oxygen
David Douthitt wrote: > > Matthew Schalit wrote: > > I'm asking, and I'll give it a try. I'm a bit newish > > to gdb, but I know enough to load a core and do a backtrace > > and such. > > Thanks very much. The "Developer's Disk" I think is available for > download; it includes at least the following: > > strip, ldd, ltrace, strace, ascii, patch, make, cpio and more... David the developers disk is not on leaf as far as I can find. Do you mean the rescue disk? I need to know where those apps and where gdb.lrp are so that I can load them selectively. I don't have enough ram to load two disks at this point. > > Recently I started to wonder where you ftp site is. > > Do you need an account with a few hundred megs of space? > > Well, right now, everything is building up at Sourceforge. I was asking because I sort of knew this was going to happen where I'd want some stuff that you mention is around, but I hadn't seen yet. > Here is the "SegFault ToDo List" - > > * file - segfaults > * tcpdump 3.6 with libpcap 0.6 - segfaults, esp. with -X > * busybox more/tar (0.50) - segfaults > * fdisk / cfdisk / sfdisk - segfault after initial display or text > > I've tried gdb with most, but it immediately segfaults even there. > > Most of these work just perfectly fine on a full distribution... > > I would REALLY REALLY like to get these fixed. It looks like libpcap, libtcap, libm, and file have the same date on the rescue.ima disk. Not sure if that means anything. I don't recall hearing of any segfaults in 2/14. Do you? What's new since then? Matt ___ Leaf-devel mailing list [EMAIL PROTECTED] http://lists.sourceforge.net/lists/listinfo/leaf-devel
Re: [Leaf-devel] seg faults in busybox routines in most recent oxygen
David Douthitt wrote: > > Matthew Schalit wrote: > > I'm afraid Oxygen 3/14 is a bit too buggy. I await the > > next version with renewed interest, though. > > I'm not sure what you base that on, but the only specified bug here is > that more segfaults. This appears to have gone away in the development > version; if you're desperate for more, you can replace it with a Debian > more instead. Ok. I'd be happy to test out your development versions in the future, before you release them, so that there's another set of eyes poking around in there. It's such a tight project already, that it'd be a bummer, imo, to let a few odd bugs show up. I can help in small ways, and maybe that'll take some of the load off you. > I'd REALLY like it if someone would be able to research some other > SegFaults I'm getting. The segfaults with more seem to have gone away. > However, fdisk (and cfdisk and sfdisk) segfault, as does file. I'd > REALLY like some help I managed to create a gdb.lrp to go with > ltrace.lrp, strace.lrp, and others of like ilk that I have on the > Developer's disk. gdb.lrp isn't there, its too big... but just ask. I'm asking, and I'll give it a try. I'm a bit newish to gdb, but I know enough to load a core and do a backtrace and such. Recently I started to wonder where you ftp site is. Do you need an account with a few hundred megs of space? Matthew ___ Leaf-devel mailing list [EMAIL PROTECTED] http://lists.sourceforge.net/lists/listinfo/leaf-devel
Re: [Leaf-devel] Re: Oxygen Problems
David Douthitt wrote: > > Matthew Schalit wrote: > > $ ifconfig eth0 63.194.213.179 netmask 255.255.255.0 > > > > The output shows the correct IP and correct MASK, but the > > BROADCAST was unchanged and still reads 63.255.255.255.0 > > This is MOST interesting. I'll put in support for IF#_BROADCAST before > I release. Thanks. It seems like ifconfig just does what you tell it to do, without derivation of the other variables. Haven't tried to do things with ip. I'll doink with it later today. Matt ___ Leaf-devel mailing list [EMAIL PROTECTED] http://lists.sourceforge.net/lists/listinfo/leaf-devel
Re: [Leaf-devel] Re: Oxygen Problems
David Douthitt wrote: > > Matthew Schalit wrote: > > > I tried this on the Oxygen machine in my own test script and > > it works my way but fails David's way. Set -x didn't show > > any difference in the command that was to be run. > > I wonder if it could be a problem in splitting of arguments? > Perhaps I have a feeling that it's either a parsing/IFS issue, or more likely it's some hard coded issue with the ${foo:+bar} format. If my day goes well, I might post this over to comp.unix.shell and see what turns up. Matt ___ Leaf-devel mailing list [EMAIL PROTECTED] http://lists.sourceforge.net/lists/listinfo/leaf-devel
Re: [Leaf-devel] Re: Oxygen Problems
Mark Seiden wrote: > > it is driving me crazy why a seemingly identical command > is not correctly executed in these two cases. does anyone > understand the problem? Yes I see it now that you quoted the offending command. The syntax is crap :) Wrong - ifconfig $IFNAME $IPADDR ${MASK:+netmask $MASK} ${ETHER:+hw $ETHER} ${BROADCAST:+broadcast $BROADCAST} $STATE Correct -- ifconfig $IFNAME $IPADDR ${MASK:+netmask} $MASK ${ETHER:+hw} $ETHER ${BROADCAST:+broadcast} $BROADCAST $STATE I tried this on the Oxygen machine in my own test script and it works my way but fails David's way. Set -x didn't show any difference in the command that was to be run. > i have no problem with the gateway setting. perhaps you inadvertantly > munged or removed the GATEWAY_IFNAME definition, which is required. > arguably it should default on a one interface machine, but it doesn't. > maybe you deleted the newline at the end of the file, e.g. Odd, it's there now, and there are more routes than before. The routes look ok, though, just a couple of host routes that took me a second to decipher. Must have been operator error. :) If this mail makes it to the list, then my feeble ipchains masq rule worked and Oxygen works, too. W00t! Thanks for the heads up, Matthew > On Sat, Mar 24, 2001 at 05:23:51PM -0800, Matthew Schalit wrote: > > Mark Seiden wrote: > > > > > > i had this problem also. > > > > > > i still don't understand why it's happening. > > > > > > you might notice that i've added an IF_BROADCAST variable as > > > well to network.conf... > > > > > > i dare say that with a class c netmask and a class a address ifconfig > > > will get your broadcast address wrong, also. > > > > > > here's a bit of code that works around in /etc/init.d/network > > > > > > echo -n "network: bringing $STATE interface $IFNAME on $IPADDR" > > > [ -n "$MASK" ] && echo -n " with netmask $MASK" > > > [ -n "$BROADCAST" ] && echo -n " with broadcast address $BROADCAST" > > > [ -n "$ETHER" ] && echo -n " (with hardware address $ETHER)" > > > echo > > > > > > # what a weird bug: ifconfig complains "netmask... unknown host". > > > # ifconfig $IFNAME $IPADDR ${MASK:+netmask $MASK} ${ETHER:+hw $ETHER} >${BROADCAST:+broadcast $BROADCAST} $STATE > > > # but this works: > > > echo ifconfig $IFNAME $IPADDR ${MASK:+netmask $MASK} ${ETHER:+hw $ETHER} >${BROADCAST:+broadcast $BROADCAST} $STATE | sh > > > } > > > > > > > > > > > > > Thanks for the workaround ideas, Mark. > > Have you noticed that the default gateway is not getting > > added to the routing table? It wasn't in mine even > > though it was entered in /etc/network.conf. I haven't > > had time to track it down though. > > > > Matthew > > > > ___ > > Leaf-devel mailing list > > [EMAIL PROTECTED] > > http://lists.sourceforge.net/lists/listinfo/leaf-devel > > -- > mark seiden, [EMAIL PROTECTED], 1-(650) 592 8559 (voice) Pacific Time Zone > > ___ > Leaf-devel mailing list > [EMAIL PROTECTED] > http://lists.sourceforge.net/lists/listinfo/leaf-devel ___ Leaf-devel mailing list [EMAIL PROTECTED] http://lists.sourceforge.net/lists/listinfo/leaf-devel
Re: [Leaf-devel] Re: Oxygen Problems
George Metz wrote: > > If you specify the netmask and the IP address, > it will go to the first valid range that includes the netmask and the > given IP, then figure the broadcast address from there. > > Are you seeing the broadcast address being set incorrectly even with the > netmask correctly identified by ifconfig? That could break all SORTS of > lovely things... > > -- > George Metz Hi George, I used a stock 3/14/01 Oxygen and booted it with a correct network.conf that should have setup eth0 with IP = 63.194.213.179 MASK = 255.255.255.0 <-- Correct data BCST = 63.194.213.255 but it doesn't set it up right. As was mentioned /etc/init.d/network ignores the IF0_NETMASK line and uses the one that you describe as the default: IP = 63.194.213.179 MASK = 255.0.0.0 <-- Oxygen creates these BCST = 63.255.255.255 So I read your followup that I quoted above, and you mention that ifconfig should get it right, "if you specify the nemask and the IP address." I tried that and is doesn't use those two to calculate the broadcast. $ ifconfig eth0 63.194.213.179 netmask 255.255.255.0 The output shows the correct IP and correct MASK, but the BROADCAST was unchanged and still reads 63.255.255.255.0 Regards, Matt ___ Leaf-devel mailing list [EMAIL PROTECTED] http://lists.sourceforge.net/lists/listinfo/leaf-devel
Re: [Leaf-devel] seg faults in busybox routines in most recent oxygen
Mark Seiden wrote: > > i'm getting a lot of random-seeming (but completely reproduceable) seg > faults that i don't remember existing in the older versions of oxygen. > > sul-lockss10.stanford.edu:/etc/init.d # more network ... > Segmentation fault > sul-lockss10.stanford.edu:/etc/init.d # I get the same segmentation violations with more on my Oxygen 3/14/01 ystem, when I try more /var/log/syslog I'm afraid Oxygen 3/14 is a bit too buggy. I await the next version with renewed interest, though. Matthew ___ Leaf-devel mailing list [EMAIL PROTECTED] http://lists.sourceforge.net/lists/listinfo/leaf-devel
Re: [Leaf-devel] Re: Oxygen Problems
Mark Seiden wrote: > > i had this problem also. > > i still don't understand why it's happening. > > you might notice that i've added an IF_BROADCAST variable as > well to network.conf... > > i dare say that with a class c netmask and a class a address ifconfig > will get your broadcast address wrong, also. > > here's a bit of code that works around in /etc/init.d/network > > echo -n "network: bringing $STATE interface $IFNAME on $IPADDR" > [ -n "$MASK" ] && echo -n " with netmask $MASK" > [ -n "$BROADCAST" ] && echo -n " with broadcast address $BROADCAST" > [ -n "$ETHER" ] && echo -n " (with hardware address $ETHER)" > echo > > # what a weird bug: ifconfig complains "netmask... unknown host". > # ifconfig $IFNAME $IPADDR ${MASK:+netmask $MASK} ${ETHER:+hw $ETHER} >${BROADCAST:+broadcast $BROADCAST} $STATE > # but this works: > echo ifconfig $IFNAME $IPADDR ${MASK:+netmask $MASK} ${ETHER:+hw $ETHER} >${BROADCAST:+broadcast $BROADCAST} $STATE | sh > } > Thanks for the workaround ideas, Mark. Have you noticed that the default gateway is not getting added to the routing table? It wasn't in mine even though it was entered in /etc/network.conf. I haven't had time to track it down though. Matthew ___ Leaf-devel mailing list [EMAIL PROTECTED] http://lists.sourceforge.net/lists/listinfo/leaf-devel
Re: [Leaf-devel] Oxygen problems
David Douthitt wrote: > > > 5) Netmask won't take, network comes up wrong still. > > network: bringing up interface eth0 on 63.194.213.179 with netmask >255.255.255.0 > > netmask 255.255.255.0: Unknown host Thanks for the reply David. I'll keep my eye out for a newer version that fixes the netmask not getting transferred from network.conf to the actual interface. I guess I could hack at /etc/init.d/network too. Thanks, Matthew ___ Leaf-devel mailing list [EMAIL PROTECTED] http://lists.sourceforge.net/lists/listinfo/leaf-devel
[Leaf-devel] Oxygen problems
Hi Folks, I'm having a few problems with the latest Oxygen 3/14 image. I'm booting it on a Pentium 150 w/16 MB w/2 Pro100+'s. I used the single floppy selection after decreasing the ramdisk down to 8192. eth0: 63.194.213.179/24 eth1 10.1.1.254/24 Problems: -- 1) During configure it brought up the /etc/network.conf twice. Not really important, just cosmetic. 2) Didn't bring up /var/boot/modules/modules.conf though it did mention that one should look in /var/boot/modules. 3) /etc/hosts not brought up for editing. Is it still created dynamically like in 2.9.x? 4) I could swear I answered y to the setup (y or n) question but nothing happened and I went straight to a login prompt. I'll look at that again some time. 5) Netmask won't take, network comes up wrong still. This problem is major for me and still exists from the last version, even though there's a new variable in /etc/network.conf: IF0_NETMASK=255.255.255.0 IF1_NETMASK=255.255.255.0 The above lines are in my network.conf but the network comes up with a /8 netmask on both interfaces rather than the /24 I requested. Trying IF0_ADDR=63.194.213.179/24 didn't work either. Trying IF0_NETMASK=255.255.255.0, I see the following error during bootup (I _did_ remember to backup the ramdisk): net.ipv4.conf net.ipv4.conf.all.accept_source_route = 0 net.ipv4.ip_always_defrag = 1 error: 'kernel.sysrc' is an unknown key DHCP not requested network: bringing up interface eth0 on 63.194.213.179 with netmask 255.255.255.0 netmask 255.255.255.0: Unknown host Usage: ifconfig [-a] [-i] [-v] . Then I see the same type of messages with the eth1 interface. I guess that's enough to fix for now. The rest seems fine. Thanks for the work, Matthew ___ Leaf-devel mailing list [EMAIL PROTECTED] http://lists.sourceforge.net/lists/listinfo/leaf-devel
Re: [Leaf-devel] does the lrp project have any response to this
Steven Peck wrote: > > > ...DocBook... > I finally filtered through all the technical jargan and I have re-written > most of what I have, added a bit. Next up is validating with OpenJade. > 'Course I have to get that to work, but I have confidence that I will get > the tools to work. > > After all, it isn't following the instructions, it figuring out what the > hell the instructions mean. :) > No doubt! I'm still pissed off about trying to understand the DocBook FAQ which starts off by telling me that DocBook is an sgml dtd or something. For the love of Pete! I just wanted the thing to compile and go away. I've got no clue what dtd is, a fertilizer? I have a bad feeling it has to do with slaughtering html. What the heck is going on here? I came from a time when there wasn't any Windows Web Server, and any sane person ran CERN httpd on Unix. Mosaic was just as good as Netscape 0.94beta as long as you didn't ask it to print. Everything was aligned along the left hand margin, and you _liked_ it. So who came along and said a bunch of crack heads could make sgml dtd's? I know, it's the people who think we still run Materhorn :) Must be time for a nap :-/ Matt ___ Leaf-devel mailing list [EMAIL PROTECTED] http://lists.sourceforge.net/lists/listinfo/leaf-devel
Re: [Leaf-devel] System libraries
David Douthitt wrote: > > Matthew Schalit wrote: > > > > David Douthitt wrote: > > > > > > Does anyone have a really good understanding of what each system library > > > (found in /lib/lib* and /usr/lib/lib*) is for? > > > > > > I'm still trying to get a good idea of libraries such as libresolv, > > > libdb, libdl, libnsl, ... > > > > > > I think that: > > > > * libnsl appears to be something related to NIS (can be removed??) > > > > NIS is only part of it. RPC is in there too, and so is cs and so > > are network directory services and some xti stuff too. It's called > > the Network Sevices Library and contains many functions that are used > > by various transport protocols. > > Without portmapper, RPC and NIS are useless, right? I don't rightly know, but I recall the issue being discussed on the Lrp list last year. > What is CS? XTI? And I thought "network directory services" was > taken care of by DNS, libresolv, and libnss*..? Connection Server. X/OPEN Transport Interface. My skiming of the libnsl info I found pointed to subsections that handled various aspects, like cr1cr1 authentication library cs Connection Server library interface desData Encryption Standards library netdir Network Directory functions. This contains look-up functions and the access point to network directory libraries for various network transports. netselect Network Selection routines. These functions manipulate the /etc/netconfig file and return entries. nsl UnixWare® Transport Interface. These functions support both the older TLI semantics and the newer semantics of the X/OPEN Transport Interface (XTI). The library entry points have their traditional names for the TLI functions, such as t_open(3xti). For the XTI functions, however, the entry points have new names, such as _xti_open. When applications are compiled, the XTI function names are translated to the XTI entry points by macros in the include file . rexec REXEC library interface rpc User-level Remote Procedure Call library yp Network Information Service functions The libsocket library has two components: inet containing the Internet library routines, and socket containing the Socket Interface routines. Of course that was UnixWare stuff, but hey, you do UnixWare, right :) Isn't Linux a UnixWare derivative, anyway? Heh. > > > * libdb is what? (it's part of glibc...) > > > > Libdb is the Berkeley DB database functions that are now maintained > > by Sleepycat Software at http://www.sleepycat.com/ > > dbopen(), dbhash(), weird stuff like that, often used with mysql. > > Do LRP-based systems really need this one? What would break if it was > gone? I'm not familiar enough with the code to know how much Linux uses it. It sounds like it's a rare one. I didn't realize it came stock with Linux until you mentioned it? Why not yank it and see :) > > > * libuuid (a part of e2fsprogs) seems to be some sort of utilities for > > > e2fs; can this be removed? > > > > Utilities for managing ext2 filesystem. > > So if libuuid was missing it wouldn't matter? At least, if there is no > e2fs filesystems to be had... I thought it was for ext2, but Lrp doesn't use either, right? So it sounds optional. > I finally started figuring out nm and objdump - but I'm still trying to > figure out how to read their output. The help files (man page, info > pages, Linux howtos) all talk in terms of "header files" when what I > really wanted was an answer to this question: "What functions does this > library provide to other binary programs?" > > It looks like a proper reading of nm and objdump will answer this... There are two things you can do. First, for applications, you use ldd. Let's say you want to know what scp depends on, like does it use libnsl.so? Well you can check that with ldd scp Secondly, for shared libraries you use nm. On UnixWare, you can use nm libnsl.so | grep FUNC | less and that will list all the functions, as SCO's nm has a field that lists what it finds by type. FILE, and FUNC, are some of the types. On Linux, I figure you'd be using GNU-binutils which includes nm. Then I'd do something like nm libnsl.so | grep " T " > /tmp/pout nm libnsl.so | grep " t " >> /tmp/pout It appears that all functions are listed with a ' T ' or a ' t ' in the second column of the GNU-nm output. The space I used between the quotations and the letter will keep your grep from grabbing functions with the letter T or t in their
Re: [Leaf-devel] System libraries
David Douthitt wrote: > > Does anyone have a really good understanding of what each system library > (found in /lib/lib* and /usr/lib/lib*) is for? > > I'm still trying to get a good idea of libraries such as libresolv, > libdb, libdl, libnsl, ... > > I think that: > > * libresolv is used for "getbyhostname" or whatever else to do name > lookups Roger. It's called the Internet Domain Name Server library. > * libnss* is the "name service switch" which is on the "back end" of > name lookups Yup. > * libnsl appears to be something related to NIS (can be removed??) NIS is only part of it. RPC is in there too, and so it cs and so are network directory services and some xti stuff too. It's called the Network Sevices Library and contains many functions that are used by various transport protocols. > * libdl appears to be something to do with using dynamic libraries If you say so :) I don't know that one. > * libdb is what? (it's part of glibc...) Libdb is the Berkeley DB database functions that are now maintained by Sleepycat Software at http://www.sleepycat.com/ dbopen(), dbhash(), weird stuff like that, often used with mysql. > * libcrypt is for passwords and other use of crypt() Roger that. > * libutil (a part of glibc) seems to be generic misc. functions and > utilities Login, logout, openpty, blah blah blah. > * libuuid (a part of e2fsprogs) seems to be some sort of utilities for > e2fs; can this be removed? Utilities for managing ext2 filesystem. > Can someone help out here? I usually just nm the library and look at the functions or check out the header file. I apologize for not being able to provide details about how needed these are to Linux, though. > The ultimate goal is to remove unnecessary libraries from the system, > and put them into separate packages - which could be anywhere, including > on a data disk, boot disk, or CDROM I figured at some point, developers would be stripping the necessary functions out of the libraries to make them smaller. Take it easy, Matt ___ Leaf-devel mailing list [EMAIL PROTECTED] http://lists.sourceforge.net/lists/listinfo/leaf-devel
Re: [Leaf-devel] Oxygen Release
David Douthitt wrote: > > The new version can be had at http://leaf.sourceforge.net/pub/oxygen/ Hi David, Thanks for hammering out a new release. I took a look at the page you cited, and I was wondering why 2.2.17 is listed as the section header, but 2.2.18 is in all the filenames: Thanks > 2.2.17 Kernel:< Here's >the 2.2.17 part > Intel 486: >Oxygen+Openwall+FrameBuffers-2.2.18.tar.gz <=== but >Oxygen+Openwall+IDE+ISO9660-2.2.18.tar.gz <=== all >Oxygen+Openwall+IDE+noPCI-2.2.18.tar.gz <=== these >Oxygen+Openwall+IDE-2.2.18.tar.gz <=== refer >Oxygen+Openwall+MathEmu-2.2.18.tar.gz <=== to >Oxygen+Openwall+VPN+Masq+Framebuffers-2.2.18.tar.gz <=== 2.2.18 >Oxygen+Openwall+VPN-486-2.2.18.tar.gz <=== >Oxygen+VPN+Masq-2.2.18.tar.gz <=== > Intel Pentium: >Oxygen+Openwall+Pentium-2.2.18.tar.gz <=== >Oxygen+Openwall+VPN-Pentium-2.2.18.tar.gz <=== ___ Leaf-devel mailing list [EMAIL PROTECTED] http://lists.sourceforge.net/lists/listinfo/leaf-devel
Re: [Leaf-devel] Oxygen network.conf problem
Hi Mark, Mark Seiden wrote: > > in the absence of netmask and broadcast parameters, i believe ifconfig > ends up with the "natural" masks and broadcast addresses that would > correspond with the network address. (i think getnetbyname or > getnetbyaddr actually are the routinges doing this). Thanks for the suggestion. I went ahead and set, in network.conf, the following variables that work in 2.9.4 and 2.9.8: IF1_NETMASK=255.255.255.0 IF2_NETMASK-255.255.255.0 but it didn't make any difference. I guess those variables are not used and that IF1_IPADDR is not compared to IF1_NETWORK to derive the NETMASK anyway. Matt ___ Leaf-devel mailing list [EMAIL PROTECTED] http://lists.sourceforge.net/lists/listinfo/leaf-devel
[Leaf-devel] Oxygen network.conf problem
Well I'm a bit confused today trying to understand how Oxygen-021401 calculates the netmasks of my nics. I'm guessing that network.conf is where the nics are setup, so I filled out the fields in there: --- IF1_IFNAME=eth0 IF1_IPADDR=63.194.213.179 IF1_NETWORK=63.194.213.0 IF1_STATE=up IF2_IFNAME=eth1 IF2_IPADDR=10.1.1.254 IF2_NETWORK=10.1.1.0 IF2_STATE=up (clearly my netmask is 255.255.255.0 for both IF's) Both the following commands concur that eth0 and eth1 get incorrect netmasks and broadcasts: ifconfig -a ip address show -- ifconfig -a --- eth0 Link encap:Ethernet HWaddr 00:90:27:50:50:50 inet addr:63.194.213.179 Bcast:63.255.255.255 Mask:255.0.0.0 blah blah blah eth1 Link encap:Ethernet HWaddr 00:90:27:30:32:34 inet addr:10.1.1.254 Bcast:10.255.255.255 Mask:255.0.0.0 blah blah blah I left out ip address show because I didn't want to type it all in, but trust me, it had the same erroneous data. So. What'd I do? What'd I do? I didn't do it. . . I'm new. Matthew ___ Leaf-devel mailing list [EMAIL PROTECTED] http://lists.sourceforge.net/lists/listinfo/leaf-devel
Re: [Leaf-devel] Oxygen Dec2000 comments
David Douthitt wrote: > > Matthew Schalit wrote: > > > > David Douthitt wrote: > > > > > > Matthew Schalit wrote: > > Just lengthen the timeout to be the same as the previous timeouts, > > 30 sec's I think, and make the timeout a variable somewhere > > so that people can set it as short as 2 sec. > > That timeout only happens during setup; once you're set up it should go > away. Ok the timeout only happends during the first setup, and after that goes away. That's was my problem. The first time through, I had no idea that setup was going to run, and it sounded very important. However, due to the relatively short timeout, I never got a chance to run setup and was left at a login prompt. I suggest a 30 sec timeout so that a new user has time to read that far and decide. > > It wouldn't have been an issue, I'm pretty sure, if the README mentioned > > that acfg, apkg, setup.sh, and configure.sh were the keys to building > > the router. I pieced that info together with a few 'find's to get passed > > my confusion, but that's where the string of clarity broke down. > > setup.sh and configure.sh are not "used" in the normal way by a general > user; a user should never have to know where these scripts are or what > they are. > Why do you need to know about configure.sh and setup.sh? I'm sorry if I've been confusing. Based on experience with 2.9.4, I could have gotten by without configure.sh and setup.sh. The fact that there was the option to configure and later an option to setup conferred the idea to me that Oxygen was either customized to work differently than 2.9.4 or was built so that success would be achieved easiest using configure and setup. Either way, it was my best guess that I'd find glory using configure and setup. I had no problem with configure. It asked me if I wanted to do it and I had time to consider that and decide yes. From that point on I suppose I never would even need to know that configure.sh was responsible or used. Then after configure, a whole mess of screen data went by and I was reading it while the setup (y or n) timeout expired. I had already made the decision to attempt to set this up, so I was distraught because I was not given enough time to realize that decision. I missed my chance to setup. So at a root prompt, I had no idea what had been missed, and I could only draw the conclusion that I had a nonfunctional system because I missed setup. Therefore, I needed to find and run setup. I hope that's a better explanation. > > I'll give ya 10 bucks to tell me some more kernel tricks :) > > Heh heh. I was serious :) > > > > 7) acfg usage. > > > > I'm confused. > > > It's no biggy. I just booted off one diskette so > > 1) acfg didn't run after logging in. > > 2) instead I saw acfg usage. > > 3) -i option looked nice. > > 4) I couldn't run acfg -i without loading dialog.lrp and libm.lrp. > > 5) How do I load those? > > 6) I didn't know there was also apkg at that point. So I kept looking > > for acfg options to load packages. It wasn't clear, because I was > > new and didn't know about the names of the 4 apps you use to setup, > > apkg, acfg, configure.sh, setup.sh > > Ahhh this makes more sense I think. However, doesn't the message > for acfg -i detail that you must load dialog and libm from another disk? Yes there's that message, but I didn't know about apkg or was getting it mixed up with acfg. When I log in I see the motd. The last part of the motd is this: > Use one of them as your second disk to use the menu based > configuration utility, or load the two packages dialog.lrp > and libm.lrp. > > usage: acfg -e > acfg -a > acfg -h > acfg -i > acfg -l ... If you simply add "using apkg" as in: Use one of them as your second disk to use the menu based configuration utility, or load the two packages dialog.lrp and libm.lrp using apkg. Then the new user would be faced with both crucial commands right there. I sort of wonder if I'm the only one who got a little confused. > The readme goes into a lot of detail about apkg; did you read that? Many times. It didn't help the fact that somehow I got apkg and acfg intertwined into one command. That confusion was short lived, though. I'd expect most people would figure it out like I did. > I don't know which version of Dec. 2000 you used; the recent versions > force you to read the readme... I used the Dec 2000 from lea
Re: [Leaf-devel] Oxygen Dec2000 comments
David Douthitt wrote: > > * Do you want to remove fdformat? [...] > * Do you want to remove syslinux? [...] > * Do you want to remove mkdosfs? [...] > > And so forth. It'll be a question and answer session, with the system > removing pieces as it goes... or perhaps all at once at the end... David, On that note, it'd be easiest to have a summary page that lists --- 1) Y=uhttpd2) Y=fdformat 3) N=mkdosfs 4) N=snarf 5) Y=syslinux 6) ssmtp Y ---> Yes I want this application available to use. N ---> No I don't want this application available. Enter the number you want to change or hit q to quit? [q] _ --- This design enables somebody to make changes and see the setup all at once. Regards, Matthew ___ Leaf-devel mailing list [EMAIL PROTECTED] http://lists.sourceforge.net/lists/listinfo/leaf-devel
Re: [Leaf-devel] Oxygen Dec2000 comments
David Douthitt wrote: > > Matthew Schalit wrote: > > 1) Requirements. I guess I couldn't find them... > Web page has them; the first page has memory requirements listed. You > don't need 32M of memory; 16M is tight though. I figure 20-24 is about > right. I hate being ignorant, but what web page? I went to leaf.sourceforge.net and looked at http://leaf.sourceforge.net/pub/oxygen/index.html and http://sourceforge.net/docman/display_doc.php?docid=1446&group_id=13751 but I didn't see what you mean. Sorry about that. > > 4) Setup (y or n). > > How about no timeout for the first boot? > > Don't know how I'd do that. Just lengthen the timeout to be the same as the previous timeouts, 30 sec's I think, and make the timeout a variable somewhere so that people can set it as short as 2 sec. It wouldn't have been an issue, I'm pretty sure, if the README mentioned that acfg, apkg, setup.sh, and configure.sh were the keys to building the router. I pieced that info together with a few 'find's to get passed my confusion, but that's where the string of clarity broke down. > You can always go back and re-read things using > "Shift-PgUp" and "Shift-PgDown" - more Linux kernel goodies. No #@$%^&* way! Do you mean that I can scroll a terminal screen? Oh Jesus. That's some sick $#|^. That's sick. I'll give ya 10 bucks to tell me some more kernel tricks :) > > 7) acfg usage. > > Secondly, I realized that I wanted to see apkg usage, but I kept seeing > > acfg usage. I guess I didn't see the logic of seeing the acfg usage right > > after the message was telling me how I'd want to load and backup packages > > very soon - acfg can't help that. > > I'm confused. Actually, acfg DOES help with packages - acfg is a > one-stop configuration menu - one menu choice takes you into apkg. If > you have everything loaded, when you log in as root you'll get the acfg > menu (with apkg as one menu option). It's no biggy. I just booted off one diskette so 1) acfg didn't run after logging in. 2) instead I saw acfg usage. 3) -i option looked nice. 4) I couldn't run acfg -i without loading dialog.lrp and libm.lrp. 5) How do I load those? 6) I didn't know there was also apkg at that point. So I kept looking for acfg options to load packages. It wasn't clear, because I was new and didn't know about the names of the 4 apps you use to setup, apkg, acfg, configure.sh, setup.sh > > module stuff. I'll work on all that later. Thanks for the heads up. Matthew ___ Leaf-devel mailing list [EMAIL PROTECTED] http://lists.sourceforge.net/lists/listinfo/leaf-devel
Re: [Leaf-devel] two port redirection questions
"Scott C. Best" wrote: > > Heyaz. First, anyone seen/used this? Hey Scott, > http://freshmeat.net/projects/nportredird > > From the blurb, it *looks* like an ipportfw > command with a -s switch. > Cool. :) And it looks like it lets you forward ports to multiple hosts so that you can probably have more than one user doing something like Gnutella. > Second...any idea if there's an autofw equivalent > of: 'ipmasqadm portfw -l -n'? Can't seem to find out if > any port range is being auto-forwarded without cat'ing > /proc/net/ip_masq/autofw directly. Weird. I had an > autofw rule sitting in my firewall for *months*, as > flushing it with 'ipmasqadm portfw -f' didn't clear > it, and there's no obvious way of checking. You refer to /proc/net/ip_masq/autofw ^^ then you mention ipmasqadm portfw ^^ What about ipmasqadm autofw? ^^ Would that do what you want? Matt > > -Scott > > PS: Yeah, EchoWall 0.51 flushes both portfw and autofw > now... > > ___ > Leaf-devel mailing list > [EMAIL PROTECTED] > http://lists.sourceforge.net/lists/listinfo/leaf-devel ___ Leaf-devel mailing list [EMAIL PROTECTED] http://lists.sourceforge.net/lists/listinfo/leaf-devel
[Leaf-devel] Oxygen Dec2000 comments
David, I had a chance to boot Oxygen for the first time the other night, and I like it. You've done a great job. A few things came to mind: 1) Requirements. I guess I couldn't find them, but I saw this: Check the requirements: verify you can run Oxygen on your PC. I guess one of the requirements is 32 MB of RAM? I only have 16. Do I need 2 diskette drives, also? 2) Tiny. I read "F4: What Should I Do Now." Very useful! Thanks for making the initial setup user friendly! But I had a problem because it suggested I load from two diskettes. I decided on the base and network-tools-1. But the tiny option, which I figured I needed, said that the second diskette drive would be /dev/fd1. I don't have an fd1. How do I run tiny with only an fd0? 3) 132x25 columns, Sweet ! Vi emulation in e3, Sweet ! 4) Setup (y or n). I was enjoying the 132 columns and the initial configuration messages so much that I was still taking it all in when the boot paused for the 10 second, Do You Want To Setup Now (y or n). I didn't even get down to reading that and my 10 seconds expired :( So I missed the setup phase and I became concerned. Luckily reseting the computer caused it to boot in the same fashion and it gave me the setup option again. Whew. How about no timeout for the first boot? 5) I ended up just hitting return at the boot prompt to use only one floppy because I knew I didn't have enough memory. So I poked around from that perpective. 6) /etc/hosts. The configure script that runs the first time didn't give me the chance to edit /etc/hosts. Would you agree that file is significant to the initial setup? 7) acfg usage. I thought it was very helpful to have the apkg and acfg usage pop up during the boot process and at other various times. I felt that there was an issue with that, though. First of all, being new to Oxygen, I kept getting apkg confused with acfg. It took me awhile to realize that there were two commands. Secondly, I realized that I wanted to see apkg usage, but I kept seeing acfg usage. I guess I didn't see the logic of seeing the acfg usage right after the message was telling me how I'd want to load and backup packages very soon - acfg can't help that. 8) acfg bug. I got a syntax error in acfg -i, and I guess there's a bug in it. I loaded dialog and libm and then ran acfg -i. I then chose one menu option, and then another and it erred. I think it was on the Network menu. I can't give more information because I can't boot Oxygen right now. But what I saw was a quick message about a syntax error in the lower left corner. That's all I could read, as it disappeared too fast. 9) eepro100 module. I run dual Pro100+'s these days and I wanted to modify the /etc/modules command, eepro100, to read: eepro100 debug=2 options=0x40,0x30 but that module is not loaded from /etc/modules. Where is it loaded from? What file do I have to edit? 10) 8390.o.This module was loaded somehow, but it's not needed for the Pro100+'s so I'd like to NAK it. Where are the default modules loaded from? Soo. Other than those cosmetic issues, I think you did a great job. Do you feel like it's nearly what you wanted it to be? It seems like it. Regards, Matthew ___ Leaf-devel mailing list [EMAIL PROTECTED] http://lists.sourceforge.net/lists/listinfo/leaf-devel
[Leaf-devel] Oxygen Dec2000 comments
David, I had a chance to boot Oxygen for the first time the other night, and I like it. You've done a great job. A few things came to mind: 1) Requirements. I guess I couldn't find them, but I saw this: Check the requirements: verify you can run Oxygen on your PC. I guess one of the requirements is 32 MB of RAM? I only have 16. Do I need 2 diskette drives, also? 2) Tiny. I read "F4: What Should I Do Now." Very useful! Thanks for making the initial setup user friendly! But I had a problem because it suggested I load from two diskettes. I decided on the base and network-tools-1. But the tiny option, which I figured I needed, said that the second diskette drive would be /dev/fd1. I don't have an fd1. How do I run tiny with only an fd0? 3) 132x25 columns, Sweet ! Vi emulation in e3, Sweet ! 4) Setup (y or n). I was enjoying the 132 columns and the initial configuration messages so much that I was still taking it all in when the boot paused for the 10 second, Do You Want To Setup Now (y or n). I didn't even get down to reading that and my 10 seconds expired :( So I missed the setup phase and I became concerned. Luckily reseting the computer caused it to boot in the same fashion and it gave me the setup option again. Whew. How about no timeout for the first boot? 5) I ended up just hitting return at the boot prompt to use only one floppy because I knew I didn't have enough memory. So I poked around from that perpective. 6) /etc/hosts. The configure script that runs the first time didn't give me the chance to edit /etc/hosts. Would you agree that file is significant to the initial setup? 7) acfg usage. I thought it was very helpful to have the apkg and acfg usage pop up during the boot process and at other various times. I felt that there was an issue with that, though. First of all, being new to Oxygen, I kept getting apkg confused with acfg. It took me awhile to realize that there were two commands. Secondly, I realized that I wanted to see apkg usage, but I kept seeing acfg usage. I guess I didn't see the logic of seeing the acfg usage right after the message was telling me how I'd want to load and backup packages very soon - acfg can't help that. 8) acfg bug. I got a syntax error in acfg -i, and I guess there's a bug in it. I loaded dialog and libm and then ran acfg -i. I then chose one menu option, and then another and it erred. I think it was on the Network menu. I can't give more information because I can't boot Oxygen right now. But what I saw was a quick message about a syntax error in the lower left corner. That's all I could read, as it disappeared too fast. 9) eepro100 module. I run dual Pro100+'s these days and I wanted to modify the /etc/modules command, eepro100, to read: eepro100 debug=2 options=0x40,0x30 but that module is not loaded from /etc/modules. Where is it loaded from? What file do I have to edit? 10) 8390.o.This module was loaded somehow, but it's not needed for the Pro100+'s so I'd like to NAK it. Where are the default modules loaded from? Soo. Other than those cosmetic issues, I think you did a great job. Do you feel like it's nearly what you wanted it to be? It seems like it. Regards, Matthew ___ Leaf-devel mailing list [EMAIL PROTECTED] http://lists.sourceforge.net/lists/listinfo/leaf-devel
Re: [Leaf-devel] FTP and Firewalls rant
"Scott C. Best" wrote: > > Heyaz. I've posted a PDF discussing the FTP > protocol and how it works/doesn't work with firewalls to > my website: > > ftp://ftp.echogent.com/docs/FTP_and_Firewalls.pdf Scott! That was sexy. Just one thought, though. Can you keep version 1.0 small enough to fit on one floppy? :) Matt ___ Leaf-devel mailing list [EMAIL PROTECTED] http://lists.sourceforge.net/lists/listinfo/leaf-devel