Re: Any interest in SNMP web based network monitor?
>>>>> "Joey" == Joey Hess <[EMAIL PROTECTED]> writes: Joey> R. Chris Ross wrote: >> For the past couple of weeks I have been using a package called >> MRTG. Joey> It's already a debian package: (snip) It might still be a good candidate for some work; the package is version 2.2, and MRTG 2.3 has been out for a while now. I'm not sure Dermot Bradley's packages are getting worked on right now; I got the impression recently that he'd become too busy (always understandable). -- Ed Donovan [EMAIL PROTECTED] -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: group permission & nethack
Hi - >>>>> "Rick" == Rick Hawkins <[EMAIL PROTECTED]> writes: Rick> I"m trying to run nethack. however, i get repsonses to hte Rick> effect of Rick> No write permission to lock /var/lib/games/nethack/perm! Rick> checking the permissions, Rick> 0 -rw-rw-r-- 1 root games 0 May 27 02:02 logfile Rick> 0 -rw-rw-r-- 1 root games 0 May 27 02:02 perm Rick> 0 -rw-rw-r-- 1 root games 0 May 27 02:02 record Rick> 1 drwxrwxr-x 2 root games1024 May 27 02:02 save The problem is the directory listed there itself. If "." was listed above, I think it'd be mode 755, with no permission for the group to write. chmod 775 /var/lib/games/nethack/ should do it (it did for me with version 3.2.2-7). Rick> And I did add myself to group games (though i'm pretty sure i Rick> didn't have to do this last time). Yeah, nethack is setuid games, you don't have to be in the group. I submitted a bug report on this, since I didn't see a fixed version on master or anywhere yet, but the developer said, > Nethack should either be up to -8 or -9, depending on whether I've > uploaded the one I'm looking at right now. :) > > Anyway, the bug was fixed a while ago. (in -8, according to the changelog) so it's coming down the pike somewhere. Hth, -- Ed Donovan [EMAIL PROTECTED] -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: libc5 and libc6
>>>>> "Paul" == Paul Miller <[EMAIL PROTECTED]> writes: Paul> How can I install *BOTH* libc5 and libc6?? Can't we all just Paul> get along? -Paul As people have said, you can install them both--the packages named 'libc5' and 'libc6'. The complications are in the '-dev' packages and the compilation environment. libc5-dev and libc6-dev cannot coexist. The current gcc package conflicts with libc5-dev, so if you want to keep up with unstable, you'll have to install libc6-dev to compile any C programs. If you want to do that, *and* also be able to compile in a libc5 environment, there are a few packages meant to allow that by having an alternative gcc and C library on the system: libc5-altdev, altgcc, and libdl1-altdev (the only ones I've seen). I haven't tried these. Hope this clarifies things a little, anyway - -- Ed Donovan [EMAIL PROTECTED] -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: bash won't see existing binaries
>>>>> "JT" == Joachim Trinkwitz <[EMAIL PROTECTED]> writes: JT> After installing the 'unstable' package wn_1.17.11-1.deb JT> the installation failed with a (for me) mysterious error: JT> bash says "/usr/sbin/wn: No such file or directory", and JT> the same thing with all other executables which belongs to JT> WN, but all files _are_ correctly installed (in my JT> opinion). JT> The binaries really are in /usr/sbin, they have the same JT> file permissions as all other files in the same dir (ls -l JT> says "-rwxr-xr-x 1 root root 86232 Jun 9 21:28 JT> /usr/sbin/wn"), which wn says "/usr/sbin/wn", but the JT> server won't start at all -- neither through inetd nor as JT> standalone (swn). JT> Is there an explanation (and better: a solution) to this JT> strange behaviour? Hi Joachim - This is something I always find *very* confusing. It's hard to distinguish what program is saying it can't find what other program in that error message. Although it looks like bash is saying it can't find wn, the message is actually bash reporting for wn, saying that wn can't find something it's looking for. (Leastaways that's how it looks from here; it really is confusing.) I'm not sure what you are missing in this case; in general it could be a library, or an executable. I couldn't test-install wn on my system, since it would conflict with apache, but I downloaded the wn_1.17.11-1.deb file, and extracted the wn binary. It only depended on libc5, and I don't believe you can get away without that on a debian system yet: $ldd /usr/local/src/wn libc.so.5 => /lib/libc.so.5 (0x4000e000) Still, since you have all the wn binaries in place, it'd seem like some kind of library problem. I agree with Emilio that some dependency seems to be broken. Since you can't execute wn, I think that 'ldd' on it won't work either. You can use 'strings' to try and see what else that binary might be looking for. Let's see, from the 'wn' binary in the above package, I only get: /lib/ld-linux.so.2 libc.so.5 as things that look like references to external files. Do you not have ld-linux.so.2? I can't remember; that isn't on any Debian system older than 'hamm', the current unstable, right? Mm, no, I don't think it is: $dpkg -S ld-linux.so.2 libc6: /lib/ld-linux.so.2 So, it looks to me like wn_1.17.11-1.deb needs libc6 but doesn't list that among its dependencies. You don't have libc6 installed on your system, do you? That's my guess. So, you could install wn_1.17.0-5.deb from stable, or start upgrading to libc6 and the packages in unstable. You could submit a bug report on wn, too. Does that fit with what you see on your system? Hope it helps, -- Ed Donovan [EMAIL PROTECTED] -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: compiling with libc5
Hi Rick - I am *not* an expert on this, but I've asked about this stuff a little before. :-) The altdev packages are intended to coexist with libc6-dev, in the way the regular libc5-dev packages cannot. So do keep that libc6-dev. 'Altgcc' is the compiler package intended to compile with libc5 under that setup, and I think you'll probably need libdl1-altdev as well. Those are the only altdevs in my list of packages from unstable. I haven't tried this setup myself, since I haven't yet needed to; I just went from using libc5-dev to using libc6-dev. Until recently, the other way to go would have been to stick with libc5-dev and not install libc6-dev. I did that until a gcc package came out that only wanted lib6-dev, and I played along. I couldn't say what .deb version that was, but if the 'altdev' setup doesn't work for you, downgrading gcc might allow you to back out of the libc6-dev environment and go back to libc5-dev, straight up. Hope that's somewhat clear, -- Ed Donovan [EMAIL PROTECTED] >>>>> "Rick" == Rick Hawkins <[EMAIL PROTECTED]> writes: Rick> I know we've gone to libc6, but I'm not quite clear on how to Rick> compile with libc5 instead. I know I need libc5-altdev. Do I Rick> need to remove libc6-dev, which also requires removal of Rick> libg++272-dev? And do I need to install anything else? Rick> I'm getting some strange errors compiling development versions Rick> of lyx, and one of them suggested trying the older libraries. -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: rxvt ignores x resource settings?
Joey - Seems like you may've got your answer already, but I'd like to toss this in: at some point, I don't now remember when, my rxvt stopped paying attention to X resources written as "RXvt*", but would recognize them written as "rxvt*", all lowercase. I'm pretty sure it was a Debian rxvt before and after, but I can't swear to it; I've built ones for /usr/local/ once or twice. I'd be curious to see if this made any difference on your system. Thanks - (And, hey, while I'm at it, thanks for all your packages and pdmenu. :-) I have, let's see, 9 packages installed on my machine, and also pdmenu being used by many people on some not-yet-Debian machines.) -- Ed Donovan [EMAIL PROTECTED] >>>>> "Joey" == Joey Hess <[EMAIL PROTECTED]> writes: Joey> Peter S Galbraith: >> > Gee, sounds exactly like the debian version, which mentions >> just these > features in its changelog. >> >> rxvt latest is v2.21 Joey> Debian version is 2.20. I don't see anything to fix my problem Joey> in the changelog for 2.21. (Another poster on this thread Joey> pointed ouyt the problem is that rxvt doesn't use standard X Joey> resources and parses the file directly.) >> My rule of bug fixing is to not bother fixing old versions, and >> to start with the latest version. If you find and fix a bug, >> where does it go? Debian only? XFree? Or at the real source >> where it belongs? Joey> If the debian maintainer fixed it, it would be merged upstream Joey> into the real source, of course (assumming the author Joey> accepeted the fix). -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: netatalk error: atalkd socket: Invalid argument
>>>>> "Scott" == Scott Hanson <[EMAIL PROTECTED]> writes: Scott> Hi, After upgrading the Linux server in our Macintosh network Scott> to Debian 1.3, the atalkd daemon in netatalk won't load, but Scott> returns the error: Scott> socket: Invalid argument Scott> AppleTalk worked fine before. Anyone else seen/fixed this? Hi Scott - (I'm a bit behind on debian-user, so you've probably fixed this, but...) If your kernel Appletalk support is compiled as a module, you'll get that error message when /etc/conf.modules file has the line: alias net-pf-5 off instead of: alias net-pf-5 appletalk I think my current conf.modules is the 1.3 default, and it has the 'off' alias, disabling appletalk. HTH, -- Ed Donovan [EMAIL PROTECTED] -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: SU Problems
>>>>> "Paul" == Paul McDermott <[EMAIL PROTECTED]> writes: Paul> Hi folks, I just installed X version 3.2 on my system this Paul> past few weeks. I am having problems do things as su. ie for Paul> the first error message is one that I receive whenever I try Paul> to run an X program as su. This error message is when I was Paul> trying to use xclock. I have check xclocks permissions, Paul> ownership both seem to be in order. Ownership is root.root and Paul> the Permissions are 755. This error happens when I try to run Paul> all x programs. Paul> Xlib: connection to ":0.0" refused by server Xlib: Client is Paul> not authorized to connect to Server open display failed Paul> InitXWindows failed, exiting Paul> This error is when I try to run emacs as su. Paul> Xlib: connection to ":0.0" refused by server Xlib: Client is Paul> not authorized to connect to Server emacs: Cannot connect to X Paul> server :0.0. Check the DISPLAY environment variable or use Paul> `-d'. Also use the `xhost' program to verify that it is set Paul> to permit connections from your machine. A little while ago this started happening to me and others under Debian's X setup. I've added localhost to my 'xhost +' list in my .xinitrc, and now programs run under other uid's in my X display work fine. It seems to me as if localhost used to be allowed by default, and is no longer, but maybe something about userids allowed to use the display changed instead. I don't know if the change would be XFree86's or Debian's; I can't find a changelog for xbase. You could just add xhost +localhost to your .xinitrc file, if you don't have any xhost commands there already. Jens has suggested "export XAUTHORITY=~paul/.Xauthority"; I don't immediately know which method would be preferable. The xhost method would solve this problem for any userid, and wouldn't require changes to any other config files to automate it. On the other hand, there could be valid security reasons not to do so--most of the machines I've run X on have been single-user workstations, so I haven't had to worry about inter-user issues like that. Hope this helps, -- Ed Donovan [EMAIL PROTECTED] -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: how to safely install libc6
Thanks, Rick - >>>>> "Rick" == Rick Jones <[EMAIL PROTECTED]> writes: Rick> ld.so has nothing to do with compiling. I suspect the problem Rick> here is either the "-L /usr/lib" type entries in the Makefile Rick> are wrong for library search path for the compilers "ld" Rick> linker (not ld.so) or the software you are trying to compile Rick> has been developed with libc6 in mind or you have the new Rick> compiler which may have default changes to libc6 linking Rick> instead of libc5. As I was just saying in reply to Craig's message, I didn't figure out well enough just what was going wrong. My compile attempts were kernels--I thought it was ld.so when, running 'make menuconfig,' the lxdialog binary it built died instantly with 'No such file or directory'. But I don't know exactly why that was. And when I tested it by just running 'make config,' that crapped out, too--but I appear not to have saved the error message, ugh. Usually I really want to know what happened, but after the dselect downgrade killed my ld.so, I got more 'Dammit, let's get this show on the road.' :-) Rick> Since libc6 is going to be the standard very shortly, it's Rick> very possible that you are compilling libc6 dependant code. Rick> Correct me if I'm wrong, but isn't hamm supposed to be mostly, Rick> if not all, libc6 software? >> it's safe to have the shared libraries for libc6 and libc5 >> installed at the same time, just as it was safe to have both >> libc4 and libc5 installed. It only gets complicated when you want >> to do development for both libc5 and libc6 on the same machine. Rick> I believe that installing hamm (unstable) would insure that Rick> the 2 development environments are seperate but equally Rick> accessable, since it is the intermediate release between libc5 Rick> and libc6 standards. And the 'altdev' packages are the means of maintaining this, then? (I think...) rick> I also suspect that libc5 and libc6, like a.out and ELF, will Rick> be co-existing on systems for some time before libc5 is Rick> finally completely phased out. Rick> Bottom line: Rick> Check your Makfiles and gcc/ld man pages for references to Rick> libc6 changes. I guess I may have had an unusual problem with the current releases; I'm not sure. Usually my systems are so standard Debian that I feel like Everyman (or, person), deb-style. :) I will plan to go with both libraries, then--but for now, a little gun-shy, I'll hold off on the libc6-dev packages. :) I guess libc6 is stabler than I thought--at first I thought glibc 2.0 released meant readiness, but then I saw a lot of warnings that it was still being worked into production on each platform. I'll keep an eye on the '-L' paths, etc. I hadn't thought that through, thanks. Appreciate the time & thought, Ed -- Ed Donovan [EMAIL PROTECTED] -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: how to safely install libc6
Craig -
Thanks, for the second time this week, I think. :-) I'm not sure which
change was decisive on my system, but after following some of your
suggestions I could compile a kernel again.
>>>>> "Craig" == Craig Sanders <[EMAIL PROTECTED]> writes:
Craig> On 15 May 1997, Ed Donovan wrote:
>> Here's a slightly old thread I'm continuing; I'm very curious how
>> people are dealing with this. I usually follow 'unstable,' in
>> part because it largely is quite stable. I hold off on things
>> that look like more trouble, like libc6. Much advice recommends
>> not using it yet on important systems. I tried to avoid it, but
>> fell for ldso 1.9.2, not realizing it (apparently) commits me to
>> the libc6 development environment.
Craig> I don't think that's the case. ldso doesn't depend on either
Craig> libc5-dev or libc6-dev. According to the notes in
Craig> /usr/doc/ld.so the only major incompatibility is with a.out
Craig> (libc4) programs. My reading of the notes indicates that it
Craig> cant even run a.out binaries any more, but i might have
Craig> misread that.
I'm now running ok with ldso_1.9.2-2. I'd thought it wasn't compatible
in practice before because of the non-linking binaries I was producing.
I'm not going to try an earlier 1.9 package again to investigate if that
was the difference. :)
Craig> it's safe to have the shared libraries for libc6 and libc5
Craig> installed at the same time, just as it was safe to have both
Craig> libc4 and libc5 installed. It only gets complicated when you
Craig> want to do development for both libc5 and libc6 on the same
Craig> machine.
Craig> suggestion:
Craig> install libc6, libc5, libc5-dev (and other libc5 development
Craig> stuff(*)), and ldso-1.9.2.
Craig> don't install libc6-dev (or related stuff) or the
Craig> altdev-libc5 stuff. You'll have to either install stuff
Craig> manually with dpkg or put quite a few packages on HOLD status
Craig> with dselect.
Craig> Doing this will give you a system capabale of developing for
Craig> libc5, and also capable of running libc6 programs.
This makes sense, and I hadn't gotten it before. Made these changes at
the same time as the ldso minor version upgrade, so, science bemoans the
fact that conditions were not controlled. But I'm very happy I could
build that new kernel with IP multicast (so I could get on with failing
to make gated work. :-)
Craig> (*) 'stuff (noun): highly technical term meaning "thingies"'
Craig> :-)
Had to leave this so the casual reader could follow. ;-)
Gratefully yours,
Ed
--
Ed Donovan [EMAIL PROTECTED]
--
TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to
[EMAIL PROTECTED] .
Trouble? e-mail to [EMAIL PROTECTED] .
Re: how to safely install libc6
Here's a slightly old thread I'm continuing; I'm very curious how people are dealing with this. I usually follow 'unstable,' in part because it largely is quite stable. I hold off on things that look like more trouble, like libc6. Much advice recommends not using it yet on important systems. I tried to avoid it, but fell for ldso 1.9.2, not realizing it (apparently) commits me to the libc6 development environment. I have not installed libc6 yet, and right now my compiles fail. All the binaries generated give "No such file or directory" errors, because they can't link, I believe. I hesitate to go ahead and install the libc6 packages, though, if things might become more unstable. I'd like to build a new kernel at the moment, and wouldn't want to do so with a C library that wasn't ready for prime time. I tried downgrading ldso to ldso_1.8.10-2. That didn't work at all; when dpkg finished my linker was blown out, and I had to go to the rescue disk. I'm not sure how the new environment should look; what packages I'll need, and if they're all in place yet; the 'altdev' packages are continuing to appear. Are people compiling stable binaries & kernels with libc6, or with it installed but using the altdev-libc5 stuff? If everyone were as uncertain as I am, I think I'd be seeing more discussion on the topic, so maybe I'm up a tree somewhere :-) But I'd love some opinions on whether, with libc6, we're getting much more 'unstable' than we usually are. Thanks all, loving Debian as always, Ed Donovan [EMAIL PROTECTED] -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Adpkg (was Re: How to trick debian into thinking a package is installed)
Thanks, Craig - >>>>> "Craig" == Craig Sanders <[EMAIL PROTECTED]> writes: Craig> if you install dpgk again before removing adpkg, nothing will Craig> break: Craig> dpkg -i dpkg_1.4.0.8.deb dpkg -r --force-remove-essential Craig> adpkg Craig> I've successfully removed adpkg from at least a dozen systems Craig> like this. That's very good to know, will do. Craig> I used adpkg for a while - i really like the way it's dselect Craig> scans the binary directories first and builds a list of Craig> packages to install, and i also like the way it configures Craig> packages immediately. Unfortunately, it needs some dependancy Craig> ordering so that it doesn't try configuring a package before Craig> all packages it depends on are configured - which leads to Craig> having to run Install about a million times and also manually Craig> install some packages. There are other problems with adpkg as Craig> well. I hadn't been aware of that stuff; it's interesting. (I wasn't sure what, if, or how much adpkg was doing for the standard dpkg & dselect commands. Adpkg says it replaces & provides dpkg, but dpkg's files are still listed in dpkg.list, and its status is 'installed'; I haven't know what to consider whose.) Is there a doc source you know about, or is that from the source or debian-devel, or general smarts? :-) Maybe I could bone up on adpkg before I remove it. Craig> adpkg shows a lot of promise, but it needs more work. (Not just for Craig; he's helped enough :) Is there still work being done on adpkg, or are the ideas being moved into the deity project or some such? Well, I'm chattering a bit, those are just browsing-level interest questions, I know what I need to know practically now. Thanks again, Craig, Ed -- Ed Donovan [EMAIL PROTECTED] -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Adpkg (was Re: How to trick debian into thinking a package is installed)
>>>>> "Christoph" == Christoph Lameter <[EMAIL PROTECTED]> writes: Christoph> Those tools are in the adpkg package... But that package Christoph> also has an experimental version of dpkg in it. I will Christoph> put it back somehow. If you have access to an earlier Christoph> version of adpkg then get that. Hi - While the topic is raised--I installed adpkg a while ago, mistakenly thinking it could come out cleanly if I wanted to remove it. I haven't used deb2asc or asc2deb yet, and don't think I'm using anything else provided by adpkg. I'd like to remove it for now, but as an 'Essential' package dpkg/dselect doesn't want to let it go. I could force-remove-essential it, but with it being tied so closely to dpkg, I haven't wanted to risk that going wrong (not fully confident in my prediction of dpkg's actions). Or I could purge it out manually. I like to leave my dpkg and debian installation as clean and uninterfered-with as possible, so I'm curious to hear what the group knows before trying anything more. Thanks all, Ed Donovan [EMAIL PROTECTED] -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: perplexing procmail problem
Douglas L Stewart <[EMAIL PROTECTED]> writes: > > Anyone have any idea why mail to debian-user is bypassing my procmail > filter? I'm using the following .poprc and .procmailrc. > [snip] > :0 > * [EMAIL PROTECTED] ^ Well, usually I don't see that dot between ^TO and the string you want to match. "^TO.*debian-user" might match anything, but you don't need to put any metacharacters around your string when you use one of the special procmail expressions like "^TO". Not absolutely sure that's it, but that's the thing I don't understand in this setup. Maybe, then... (Hm, catchy subject line, I think I'm starting to hum it to the tune of "Hooray for Captain Spaulding" :-) -- Ed Donovan [EMAIL PROTECTED]
Re: Error while installing unstable/binary/base/at_3.0.deb
Hi Andreas -
This has been reported as a bug (by Dirk Eddelbuettel) and is fixed in
at_3.1.2-1. But, with the problem in the postrm script, normal
reinstallation of the new version doesn't work. :-( If your system gets
stuck in that state, you can edit the buggy postrm script pretty simply
to make it work.
The file is `/var/lib/dpkg/info/at.postrm'. You have to edit it as
root, or at least be root to put the edited version in place. The
problems are on lines 5 and 6:
update-rc.d atd remove >/dev/null; fi
rm -rf /var/spool/at{jobs|spool}
should be changed to:
update-rc.d atd remove >/dev/null
rm -rf /var/spool/at{jobs,spool}
That is, take the premature "; fi" off the end of line five, and change
the pipe ("|") into a comma on line 6. After this, I could install the
new version cleanly. Hope this helps,
--
Ed Donovan [EMAIL PROTECTED]
Andreas Tille <[EMAIL PROTECTED]> writes:
>
> While doing
>
>dpkg -i at_3.0.deb
>
> I got the following errormessage.
> The at_2.9b??.deb in stable works.
>
> (Reading database ... 21163 files and directories currently installed.)
> Preparing to replace at 3.0 (using at_3.0.deb) ...
> Unpacking replacement at ...
> /var/lib/dpkg/info/at.postrm: spool}: command not found
> dpkg: warning - old post-removal script returned error exit status 127
> dpkg - trying script from the new package instead ...
> /var/lib/dpkg/tmp.ci/postrm: spool}: command not found
> dpkg: error processing at_3.0.deb (--install):
> subprocess new post-removal script returned error exit status 127
> /var/lib/dpkg/tmp.ci/postrm: spool}: command not found
> dpkg: error while cleaning up:
> subprocess post-removal script returned error exit status 127
> Errors were encountered while processing:
> at_3.0.deb
>
> Is there anybody who observed the same or should I send a bug report
> (or are there any developers reading this and take it as a bug report
> and I havn't to send this twice)
>
Re: samba
Hi Peter, It's on master.debian.org and hasn't made it to ftp.debian.org yet. I got through and snagged a copy if you wouldn't mind it as a mail attachment or some such... -- Ed Donovan [EMAIL PROTECTED]
Re: Any Linux books that focus on Debian?
CoB SysAdmin (Joe Emenaker) <[EMAIL PROTECTED]> writes: > > > I'm interested in getting a book on adminstering Linux systems, but just > about *every* book I look at focuses on Slackware. *Blech!*. > > Are there any books that focus on Debian? Well, perhaps partially, and soon :-) This was on c.o.l.announce not long ago; I meant to bring it up then. I was really pleased to hear it: At Matt Welsh's request, Specialized Systems Consultants has taken over the maintenance of Matt's book "Linux Installation & Getting Started". This book can be found on the web at http://sunsite.unc.edu/mdw/LDP/gs/gs.html It can also be accessed through SSC's Linux Resources Page, http://www.ssc.com/linux/ under The Linux Documentation Project. "Linux Installation and Getting Started is an installation and new-user guide for the Linux system, meant for Unix novices and Gurus alike. It contains information on how to obtain Linux, installation of the software, a beginning tutorial for new Unix users, and an introduction to system administration. It is meant to be general enough to be applicable to any distribution of the Linux software." Currently, SSC is working on a new installation chapter just for the Debian Linux release, and has plans to do chapters related to each of the major distributions. If you have any suggestions or corrections pertaining to this book, please send them to [EMAIL PROTECTED] (There's nothing Debian-specific in those versions on the web yet.) Does anyone on the list happen to know more about this; anyone working with them yet? -- Ed Donovan [EMAIL PROTECTED] -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Re: xman
Rick Macdonald <[EMAIL PROTECTED]> writes: > Ever tried TkMan? If not, you're in for a treat (IMHO) if you don't > mind loading tcl, tk and rman. > Oh, btw, anyone think tkman has to depend on lpr? I don't imagine so; I don't have a printer. Not that I've mailed the maintainer or anything, just popped into my head, sorry. -- Ed Donovan [EMAIL PROTECTED] -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Re: Archive, out of date web pages, Navigator 3.0, etc
Jeff Blaine <[EMAIL PROTECTED]> writes: > > 6. Naming packages '-dev' and putting them in the 'stable' area is > really confusing. I specifically avoided downloading anything with Nay, nay, '-dev' means a package that has files you need to _compile_ with a certain set of libraries, as opposed to the runtime files. I ran the packaged Mosaic, I needed zlib, libjpeg, libpng, etc. I wanted to build the Mosaic alpha sources, I had to get zlib-dev libjpeg-dev, libpng-dev (well, names approximated). libc5 and the many other runtime packages just have object files; libc5-dev and the rest are where the C header files (among other things) go. It's late and I'm not explaining this well, but you probably knew what I mean immediately. If not, someone else'll do better. Oh, yeah, info-browser is provided by emacs, which you probably already do or don't like, or the lowly 'info', which many admit is not so hot a way to see them, if you don't use emacs or xemacs. Hope this is of some use, -- Ed Donovan [EMAIL PROTECTED] -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word "unsubscribe" to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED]
Re: A couple of package spec glitches
Randy Gobbel <[EMAIL PROTECTED]> writes: > > I tried to install xemacs 19.14, and dselect complains that it conflicts with > FSF emacs. This is simply not true. The two emacsen keep their support files > in different places--as a matter of fact, I have xemacs 19.14 installed, and > both it and FSF emacs work fine. This is a bug in the package spec for xemacs > 19.14. I don't have the file around anymore, but in the last debian-changes announcement for xemacs, the maintainter noted that he had a problem with the installation of info files that both emacs and xemacs provide. He couldn't get xemacs to not overwrite the emacs-provided ones, and so had to list the packages as conflicting for now, though he was very unhappy about that. I haven't downloaded the xemacs package, so I don't know if this is mentioned anywhere within. -- Ed Donovan [EMAIL PROTECTED]
Re: -lX11 problem
"Brian C. White" <[EMAIL PROTECTED]> writes: > Don't do this!!! I tried this fix for netscape and it causes problems > with 'man'. It search X11, X11R6 and X386 directories for man pages so > if you create a link like this, it will find all of you X manpages twice. > > The better solution is to add the -L flag to the compile line. Agreed. A little earlier in the thread (article's already gone) if was mentioned you could add the -L/... flag to CFLAGS in a Makefile. I'd like to amend that: if the Makefile keeps LDFLAGS separately, as some do, it belongs there instead, as a flag specifically to the linker. Messed this up just today :-) -- Ed Donovan [EMAIL PROTECTED]
Re: how to sign up to debian-devel?
I thought you were just supposed to mail Bruce, directly. I haven't ever tried to get on, though... -- Ed Donovan [EMAIL PROTECTED]
Re: gcc can't find termcap library
Hi Susan, "Susan G. Kleinmann" <[EMAIL PROTECTED]> writes: > It would probably be a good idea to remind folks (in the FAQ or wherever) > about the rationale for using curses vs. termcap. I tried to find one > using AltaVista and DejaNews, but I'll bet you can guess what a snowstorm > of "information" that yielded. I understand that ncurses provides more > support for shared libraries (than curses), but that's all I could find > so far. Any additional pointers would be helpful. I could certainly use some more education on the ncurses/terminfo vs. curses/termcap usage issues, but you could draw on the reasons in the ncurses documentation--in /usr/doc/ncurses-dev, and on the web at http://www.ccil.org/~esr/ncurses.html and http://ltpwww.gsfc.nasa.gov:81/ltpcf/about/unix/Depotdoc/ncurses/ (I don't know if the second one's canonical.) These are the parts that say the whole combination's more powerful, and curses and termcap are supposed to be on the way out. They don't say that the binary terminfo files are much faster than the ascii termcap definitions; I don't know where I've been reading that, sorry. >From the announce: "The ncurses library is a freeware emulation of System V Release 4.0 curses. It uses terminfo format, supports color and multiple highlights and forms characters, and function-key mapping, and has all the other SYSV-curses enhancements over BSD curses. In mid-June 1995, the maintainer of 4.4BSD curses declared that he considered 4.4BSD curses obsolete, and is encouraging the keepers of Unix releases such as BSD/OS, freeBSD and netBSD to switch over to ncurses." And from the ncurses intro: "System III UNIX from Bell Labs featured a rewritten and much-improved curses library. It introduced the terminfo format. Terminfo is based on Berkeley's termcap database, but contains a number of improvements and extensions. Parameterized capabilities strings were introduced, making it possible to describe multiple video attributes, and colors and to handle far more unusual terminals than possible with termcap. In the later AT&T System V releases, curses evolved to use more facilities and offer more capabilities, going far beyond BSD curses in power and flexibility." hth? I think I'll be posting my own latest ncurses compilation question soon :-) -- Ed Donovan [EMAIL PROTECTED]
Re: Make errors
[EMAIL PROTECTED] (joost witteveen) writes: (snip) > You just don't have the command "makeinfo" installed. > It's in the texinfo package. > > Installl texinfo, and you should be OK. > Or, emacs can do the job, if you have it. (And what are you making? I've always run into this building emacs-related packages, so I'm guessing you have it, but perhaps it shows up far afield.) The makefile you're using should have an entry like MAKEINFO = makeinfo and perhaps also EMACS = emacs which you could add, if not. Change the first to MAKEINFO = $(EMACS) -batch -q -f batch-texinfo-format and it will run emacs in batch (non-interactive) mode to do the job. Does that fit? -- Ed Donovan [EMAIL PROTECTED]
Re: UPDATE: deb-view.el 1.2: emacs tool for browsing deb files!
Well, how many times can I follow myself up? Let me know if I get near a record. Ed Donovan <[EMAIL PROTECTED]> writes: > Well, wait, I didn't add: there is also a 2.2.26 release version, which > I'm pretty_ includes the fast new parser, though not this and that > spiffiness, blah, blah... Uh, I may or may not be pretty, but that should have said "...I'm pretty sure..." with little underlines which I won't try again. Thank you for your support. -- Ed Donovan [EMAIL PROTECTED]
Re: UPDATE: deb-view.el 1.2: emacs tool for browsing deb files!
Ed Donovan <[EMAIL PROTECTED]> (me) writes: >>Mark Eichin <[EMAIL PROTECTED]> writes: > >> More usefully, w3 can be found in the binary-i386/net/w3-el_2.2.25-4.deb >> on any debian mirror... > Yeah, but those 2.2.x w3's are way slower than the 2.3 & now 3.0 > betas, which are very usable. There's a new html parser. It's a > dramatic difference, & I really recommend it, well worth what betaness > there is. In the 'betas' subdirectory of the above path; there may be a > symlink to newest version in the above directory also, don't recall. Well, wait, I didn't add: there is also a 2.2.26 release version, which I'm pretty_ includes the fast new parser, though not this and that spiffiness, blah, blah... Raul, if you want to, great...if you don't have the time, maybe I'll give it a shot. Thanks, -- Ed Donovan [EMAIL PROTECTED]
Re: UPDATE: deb-view.el 1.2: emacs tool for browsing deb files!
Mark Eichin <[EMAIL PROTECTED]> writes: > >> w3 can be found via anonymous ftp to ftp.cs.indiana.edu:/pub/elisp/w3. > > More usefully, w3 can be found in the binary-i386/net/w3-el_2.2.25-4.deb > on any debian mirror... Yeah, but those 2.2.x w3's are way slower than the 2.3 & now 3.0 betas, which are very usable. There's a new html parser. It's a dramatic difference, & I really recommend it, well worth what betaness there is. In the 'betas' subdirectory of the above path; there may be a symlink to newest version in the above directory also, don't recall. -- Ed Donovan [EMAIL PROTECTED]
Re: using filter
Shaya - I hate to badmouth, (can that be used intransitively?) but, since the mail filtering faq's been mentioned, I'll quote it on the topic of filter: "IMPORTANT NOTE == If your system has both procmail and filter installed then you should use procmail which is *much* more robust and powerful than filter. This recommendation is almost universal; even the developers of Elm and Filter recommend procmail over filter. IT IS POSSIBLE TO LOSE MAIL MESSAGES WHEN USING FILTER; this is rare but it has happened." I'm not sure why they leave filter out there, then, but... If you've already assessed this situation, then excuse me. -- Ed Donovan [EMAIL PROTECTED]
Re: Q's about modifying a debian package install
Hi Al, I can't answer everything, but: Al Youngwerth <[EMAIL PROTECTED]> writes: > I'm still trying to get a grip on the whole theory of debian package = > installations and I have a few questions about how one should go about = > modifying an installation. > > I just installed the apache package. The debian installation sets up = > apache to run out of the rc.d stuff but I want it to run under inetd = > (it's for "intranet" access and security is more important than speed). = > What's the "right" way to make this modification? To take apache out of = > the start-up process should I: > > Change the apache script in /etc/init.d to just exit whenever its = > called? Not as clean as: > Run "update-rc.d apache remove" (this generates the error: "update-rc.d: = > error: /etc/init.d/apache exists during rc.d purge)? I believe you're supposed to delete the script in /etc/init.d/. Then run update-rc.d, and it will remove all the symlinks. > Just delete apache from /etc/init.d and its symlinks in all the = > /etc/rc.d directories? Yeah, this is the same as above, manually. (snip) > What about installing non-debianized programs into my system. I've seen = > people refer to installing all non-debian stuff in /usr/local. Why? I'm = > writing a lot of little perl script utilities for my system. Should they = > all go in /usr/local? What about the configuration files for the perl = > script utilities I write. I've been putting them in /etc, should they go = > someplace? That should be clarified: all unpackaged stuff _that belongs under /usr_ should go in /usr/local. /etc is still /etc, and is yours; follow the general layout, but it's not in a sort of 'distribution freeze' like /usr. Add special system-wide config files as you need, and Debian packages whose existing configuration files have been modified should respect that, offering you options like leaving your versions in place and copying the new distribution versions in as /etc/*.dpkg-dist. Nb: this is no kind of official statement. If corrections are needed, I'm sure they'll follow. -- Ed Donovan [EMAIL PROTECTED]
Re: "man" crashes
Hamish Moffatt <[EMAIL PROTECTED]> writes: > When I try to view a manual page that's located in /usr/local/man, > man says it's updating a local index, then gives a segmentation fault. > Any ideas on how to fix this? Christoph Lameter <[EMAIL PROTECTED]> writes: > Any way to fix those problems? This is going back a step: I waited, thinking I might be duplicating Dale's answer--but I'd suggest one more thing. I had this problem a little while ago. (I wasn't using an experimental version of dpkg, though I didn't ever find out just what caused it.) What worked for me was running mandb with the -c option to completely rebuild the database: -c, --create By default, mandb will try to update any previously created databases. If a database does not exist, it will create it. This option forces mandb to delete previous databases and re-create them from scratch. This may be necessary if a database becomes corrupt or if a new database storage scheme is introduced in the future. I ran it from the command line as root, and without specifying any path, letting it read /etc/manpath.config. I don't recall what directory I ran it from, but don't think it mattered. Segfaults went away as silently as they'd come. I didn't get the impression you've tried this yet, and if you haven't, do, because it sounds pretty much like my experience. Hope it helps, -- Ed Donovan [EMAIL PROTECTED]
Re: using procmail with emacs RMAIL
Hi Steve, I use Gnus & sometimes VM, so I can't tell you a thing about Rmail, except that you should point at the separate spool file that procmail will create. How to do this so that it doesn't just chuck it back in with the rest of your mail, I don't know, though it may be easy. My procmail settings are mostly cribbed from the "Mail Filtering FAQ", which I found to be very useful. Among other places: http://www.smartpages.com/faqs/mail/filtering-faq/faq.html I just copied the setup it outlines, and use these lines to catch Debian mail: :0: * ^TOdebian-user debian-user.spool :0: * ^TOdebian-changes debian-changes.spool Procmail uses one of those stanzas for each filter. The '.spool' suffix is a Gnus default that I went along with. Following the FAQ, I have these in "~/.procmail/rc.maillists"--in a more vanilla setup they'd be in "~/.procmailrc". Hth, -- Ed Donovan [EMAIL PROTECTED]
Re: perl
Andy Dougherty <[EMAIL PROTECTED]> writes: > Perl uses 'csh' for filename globbing (that is, for <*.*> sort of > constructs.) Yes, that's strange behavior, and the dependency probably > will eventually be removed, but not anytime soon. But, but...I thought it was noted, here or in the bug reports, that the csh-usage was an optional include, and was being taken out. Maybe this is a Debian stable/unstable issue?, because my current perl, and the current rex/binary-i386 Packages file say: Package: perl Version: 5.003-2 Pre-Depends: libc5, libgdbm1, libdb1 I've never had csh on this system and this version went in fine, works fine. Darren, are you around to confirm that this is behind us? -- Ed Donovan [EMAIL PROTECTED]
Re: new documentation on how to develop Debian packages available
Rick Macdonald <[EMAIL PROTECTED]> writes: > > For emacs users, my deb-view feature is just what you need to look at > doc files in a .deb file without installing _or_ unpacking the archive! > > The file is about 12KB. > > http://www.cuug.ab.ca:8001/~macdonal/deb-view.el > Ditto. Some of these dpkg docs are html; if you have w3, when you deb-view one of these docs you can 'w3-preview-this buffer' to display it right. It can't follow the links (hmm...that'd have to be pretty abstract) so view each file separately, if that's comfy for you. (Unsolicited deb-view ramble: I find it particularly valuable, since I used to 'dpkg -x' deb-files all the time to check stuff out, and often I was root. Cleaning up from that, I'd always 'rm -r usr/' in the unpacking workspace, and then gasp, realizing I'd been one character away from the apocalypse :-) Thanks again, Rick. -- Ed Donovan [EMAIL PROTECTED]
Re: printing and .profile problems
Hi Chris - Is there an existing .bash_profile in that directory? I believe Debian by default gives you one. If so, .profile will not be read: Login shells: On login (subject to the -noprofile option): if /etc/profile exists, source it. if ~/.bash_profile exists, source it, else if ~/.bash_login exists, source it, else if ~/.profile exists, source it. (from the man page) So you could rework it one way or another to get your command sourced. (You can also instead just put it in .bashrc, which practically always gets read by interactive shells. But why leave this .profile ambiguity hanging anyway? :-) The other possibility is less likely: that you are somehow running a sub-shell, not a login shell. If you were using an xterm this might be the case. But any login from a console prompt should be a login shell, and read the startup files as specified above. Specifying login behavior, or the .bashrc solution, would work if it was a sub. Hth, -- Ed Donovan [EMAIL PROTECTED]
X font packages in very limited disk space
Hi - Seeking yr advice for a fairly slim X installation. Right now I have only the xfntbase fonts installed, and I get plenty of "can't get font courier-..." or "adobe" or "helvetica" messages when starting apps. The Contents file tells me several such fonts are hanging around in xfnt75 and xfnt100, and some in xfntscl. Which of those package(s) would be most useful under 800x600 on the SVGA server? What situation calls for 75dpi, 100dpi, or scalable fonts? The common apps I run right now include rxvt, emacs, & xmosaic. It'd be great to just add one of those font packages. I can certainly try installing them in turn and see what lights up my life, but maybe I can get a little more background than I have from docs or FAQs on the roles of those different fonts. Many thanks as always, -- Ed Donovan [EMAIL PROTECTED]
X server error output (Re: X11 problems)
[EMAIL PROTECTED] (Jens Decker) writes: > > The monitor gets black and after CTRL-ALT-F4 I get: > > TRANS(SocketUNIXConnect) () can't connect: errno = 111 > giving up. > xinit: Connection refused (errno 111): unable to connect to X server > xinit: No such process (errno 3): Server error. > Re: this output - It'd help to see more of the stderr output from X. This is the tail end of what scrolled by. The X server probably found no valid modes, according to its assessment of the hardware, in your XF86Config. If this was run by startx, try 'startx 2> X_Startup &', just sending that output to a file. You'll see a lot of stuff in there; lines beginning (**) report info specified in your XF86Config, those beginning (--) show what the server is determining on its own. These will include the lines where it excludes possible configurations found in that file, like these: (--) SVGA: Mode "640x480" needs vert refresh rate of 69.98 Hz. Deleted. (--) SVGA: Mode "800x600" needs vert refresh rate of 60.32 Hz. Deleted. (--) SVGA: Mode "800x600" needs vert refresh rate of 72.19 Hz. Deleted. and later these: (--) SVGA: There is no defined dot-clock matching mode "800x600" (--) SVGA: Removing mode "800x600" from list of valid modes. (not that I'm having trouble getting my Cirrus 5420 to do 800x600 ;-) This could give you some start into what's not working in your XF86Config settings. I don't know your hardware & am not enough of a general expert, so I should let others take it from there. They'll want (parts of) your XF86Config, and perhaps the results from running Superprobe. Right now, either the settings don't match the hardware, or you have to specify more to force the server to realize that they do. HTH a little, -- Ed Donovan [EMAIL PROTECTED]
Re: mail-delivery-agents?
And Dale - If you're thinking of going with emacs clients, there's an smtpmail.el package. Not to stunt your budding mail-admin-ship, but maybe it'd do the job. I don't use it but a decent number of people seem to. http://www.cs.washington.edu/homes/voelker/ntemacs/contrib/smtp> is the most recent locale I've seen (that was off the Gnus mailing list, I have no relations with NT, thank you please! :-) -- Ed Donovan [EMAIL PROTECTED] (reader should note unhealthy emacs-junkie gleam in poster's eye)
kernel-package and perl, without perl package
I'd like to continue to have a working perl binary without installing the full package; disk space is a big issue. The base set of 0.93 gave me perl 5.001m as /usr/bin/perl (and /bin/perl symlinked to it). To dpkg it's orphaned; and I can't dig up an old Contents file now to see if any old package 'fesses up. All of the 'base' packages on my system are current, but I still have that a.out perl and nothing more recent. This has become a foreground issue since kernel-package requires perl 5.002. Do I have to install the whole perl package, or can I get a no-frills, dpkg-recognized 5.002 binary? (and then force-depends kernel-package? Or can I just force it now? I'm presuming Manoj had his reasons.) Does everybody else just have the perl package installed? :-) -- Ed Donovan [EMAIL PROTECTED] PS - Repeat question, sorry; in May this was part of a long letter, but slipped through the cracks. PPS - Congratulations, Bruce and everybody, on Debian 1.1. Thanks.
ram needed to install debian
ALLAN W. BART writes: > > hello, > > i would like to hear from others out there, the question is this, is it > possible to load and install debian in 4mb of ram? > > allan Absolutely. As said, console-only. I've installed 0.93, run it, upgraded to 1.1 and run it in 4M. Standard 16M of swap, though I've never filled that. As of not long ago, Bruce's installation testbed was a 386 with 4M, and he repeatedly tweaked the install to deal with possible issues people encountered on 4M machines. -- Ed Donovan [EMAIL PROTECTED]
Re: DEBIAN Linux on floppy disks
Kevin - That'd work; I had this grim problem. If fact (beware: tale of hardship :-) to install most of my Debian system, I had to sneakernet the .deb files via msdos floppies from a _Mac_ with net access, and which could only fit ~850K on the disks. So, I ran 'split' on them on my unix ISP first, and ultimately catted them back together under Linux. This reminds me of what Dr. McCoy accused the transporter of doing to his molecules. Second topic: I've been swamped lately, and haven't thanked you for some of your help messages; I've really appreciated them. -- Ed Donovan [EMAIL PROTECTED] p.s. Now that I've gone online to mail this, I see other answers, but, gots to thank you anyway ;-)
elf Mtools screw pooch
Since I upgraded to 1.1, all of the ELF versions of mtools have failed
on my system, with the error:
init: set default params
Can't initialize 'A:'
(approximate transcription)
I'm not even sure if this is the system call init or the uber-process
init, and haven't succeeded in any attempts to fix this through
configuration. It sounds like a basic (and so common? :) error, but I
can't find any given answers in the docs, my usenet spool, the web or
dejanews. ("See, I tried, whine, whine")
I have a Compaq Presario 433, unfortunately I never saw the manuals for
this old baby and can't find anything more specific about what floppy
drive I have.
Mtools 2.0.7-15b was the last that worked; since then I've tried 2.5.4-2
and 3.0-1. That old one, reinstalled, however, still works for me.
This is like my (still unsolved) gpm/emacs/mc problem; it doesn't work
out-of-the box for me, then I'm quickly over my head with the specifics
of how this package operates. Oh, well, sorry.
Any thoughts appreciated, thanks,
Ed
Several ?'s on 1.1 upg, going all-ELF, kernel versions, orphaned a.out binaries
Here are several related questions. I'm running a 1.1 system upgraded from 0.93. All base and devel packages are the latest available in the 'unstable' tree; I'm not completely up to date with the 'Incoming' directory. So I have gcc-2.7.2-8 and libc5/libc5-dev 5.2.18-6. I'm using kernel 1.2.13, compiled myself. I wonder if I should move to a 1.3 version, and, of course, if so, which? The (hopeful) proximity of 2.0 official is another factor. I've got a lot to learn here. Issues: I'm trying to get networking going for the first time (to a dialup ppp account). Some of the docs say ppp-2.2.0 (I have 2.2.0-f) needs a 1.3 kernel to work. I'm not yet at the point of firing up to test it myself. Will a kernel upgrade be necessary? I'd like to go all-ELF, for simplicity, and because I hope it can save memory on my 4M system only having one set of shared libs. I guessed I'd have to have a kernel made with gcc-2.7.x to do this. Again, do I need to upgrade to a 1.3 kernel? I could try to apply patches from the net which claim to make 1.2.13 compilable as ELF by gcc 2.7's. I don't know if this is a good idea, or whether any of these questions will prove to require a 1.3+ kernel anyway. Beside the considering the kernel itself as an a.out binary, and I don't know if that counts; to go along with my 1.2.13 kernel I've had to keep (downgrade to, actually) the 0.93, a.out versions of the syslogd and modules packages. So, to keep using packaged versions rather than making my own (would that better at this point? I haven't figured out yet how desirable it is to have all-debianized software), I'd have to get a 1.3 kernel and the sysklogd, modules, and modconf packages that go with it. I see a 1.3.100 kernel package available but sysklog and modules for 1.3.69. Would those work together acceptably? Plus then, if I go that high, there's the new /proc system, and probably other things, to sort out. I thought I was ready to take on the Debian beta release, not ready to move beyond production kernels; do they really go together? Maybe I should have waited; excuse me if I'm just too far behind here. On the other hand, maybe I've just been warned off development kernels too hard. Also: there are a couple of orphaned a.out binaries on my system: The unstripped perl binary included in the base 0.93 system has not been replaced by an ELF version. I wasn't going to install the full perl package yet, but I know I shouldn't delete perl off the system. Is this a bug? I never installed a samba package, but something that I can't figure out, using dpkg -S or looking at the 0.93 and 1.1 Contents files, has put/left an smbclient binary on my system (I'm sorry, it's not in front of me to check where, it's probably in /usr/bin.) Are either of these last two a problem for anyone else? (Related: I see someone's already reported the a.out setterm binary as a bug.) Thanks for any education here, Ed -- Ed Donovan [EMAIL PROTECTED]
Gpm not interacting with emacs, mc
Hi, I'd really like to have mouse functionality with mc, and especially, emacs; I think I have the right stuff installed, but the applications don't hear the mouse at all. I have a mostly 1.1 system, upgraded in place from a 0.93r6. All base packages are newest versions, except for those that go along with a 1.3 kernel. Emacs is 19.30-1, mc 3.1.4-0 , gpm 1.06-1 ; also the latest versions. I have a PS/2 style Compaq mouse, on a Presario 433. My current kernel was compiled with only PS/2 support, I said no to all other mouse type options, in case that's relevant. Unfortunately, I have little idea what to fix in the interaction between the packages. Gpm works properly selection-style on the console, and mev works and seems to give the right output, including in elisp-output mode. In emacs, I 'load-library t-mouse' and execute 't-mouse-run.' That's all I can see suggested. Nothing happens, the mouse continues to function only in selection mode; it can't move point, etc. I've tried control- and shift- mouse events, too, but no different. Mc is the same; the docs for it and gpm say it should work straight off, but nada. Is selection mode getting in the way of these functions; should I disable it somehow? That doesn't seem the case but beyond that I'm in over my head. Do most people have them working correctly? Thanks, Ed -- Ed Donovan [EMAIL PROTECTED]
Re: disabling daemons from init.d
To: Ed Donovan <[EMAIL PROTECTED]> Subject: Re: disabling daemons from init.d Date: Sat, 04 May 1996 10:49:57 -0400 From: Michael Alan Dorman <[EMAIL PROTECTED]> cd to /etc/init.d, chmod -x whatever, where whatever is the script you want to turn off. If it's not executable, it won't get run. Mike. -- "Don't let me make you unhappy by failing to be contrary enough" Excellent, I'll buy that. If it throws up any error messages, are they trivial? (I'm sorry; I should just check myself, but my Debian machine isn't my network connection, so I'm bouncing back and forth, and try to look ahead.) Thanks, Ed Ed Donovan [EMAIL PROTECTED]
Re: error make-ing kernel 1.2.13
From [EMAIL PROTECTED] Sat May 4 01:40:32 1996 To: Ed Donovan <[EMAIL PROTECTED]> Cc: debian-user@lists.debian.org Subject: Re: error make-ing kernel 1.2.13 Well, you are not alone. I also have this "problem", when I recompiled this kernel the first time. What you could do is: not select PPP when you are in make config. Then just follow the steps to finish the 'roll your own kernel': make dep; make clean; make zImage. You should not have any error this time. And if you really need PPP, you can use the PPP modules. Godfrey [EMAIL PROTECTED] Thanks, it's amazing how happy you get not to be alone. And that report's excellent to hear because that's what I'm trying right now. I didn't think I could figure this out, but having sent for help, I had nothing else to do, and I got a better understanding from the makefile in that directory. So I thought screw it, I'll make it without ppp and can load it as a module. Then I was kind of red-faced running make config again and seeing that they *suggested* 'no' for ppp. More than a good suggestion in this case, it seems. Now I can go to sleep believing it'll come out right. Grazie -- (I guess your host's in Toronto, but your username has me feeling Italy :) -- Ed Donovan [EMAIL PROTECTED]
error make-ing kernel 1.2.13
Hello - I was trying my first kernel compile, making one with built-in ELF support in preparation for the 1.1 upgrade. All my tools (& whole system) were from the debian-0.93 directory at the tsx-11 mirror, as of the last few weeks; running gcc-2.6.3-5.deb on source-1.2.13-7.deb. Let me know if I should give more package version information; it seemed I had everything I needed, all dependencies satisfied to run the above. The one irregularity: I first installed the includes.deb package, then later ran 'dpkg --purge includes' before installing source.deb. Appeared to go ok. Hardware is Compaq Presario 4DX33, only 4M RAM, installed via ftp & floppy. So here's my problem: the 'make dep,' and then the 'make zImage' failed, citing the 'bsd_comp.h' library, and saying it could only be made as a module. I don't believe I answered about that package in 'make config,' and I wanted to ask here before trying all again. More detail: First, following the Kernel-HOWTO, I think, I did a 'make mrproper.' Then the make config; chose to include networking, & slip but no e-net cards, took the standard tcp/ip options but not the fancy ones, gatewaying masqerading et.al. (If I should have just done this all again to see my error, sorry.) Then a 'make dep;' it did give an error about 'bsd_comp,' and perhaps it stopped early right there, but I wasn't prompted about this by the documents I had and decided to go ahead anyway. 'Make clean' appeared to go ok; I don't have all the diagnostic output, but here's what 'make zImage' then said when it stopped: gcc -D__KERNEL__ -I/usr/src/linux-1.2.13/include -E -I../../net/inet -M Space.c <- this line wrapped for mail auto_irq.c net_init.c loopback.c ppp.c bsd_comp.c slip.c slhc.c > .depend <- ditto bsd_comp.c:56: #error This file must be compiled as a module. make[2]: *** [dep] Error 1 make[2]: Leaving directory `/usr/src/linux-1.2.13/drivers/net' make[1]: *** [dep] Error 2 make[1]: Leaving directory `/usr/src/linux-1.2.13/drivers' make: *** [depend] Error 2 I can go look in those ^^ directories, and try and pull out references in the makefile there, but wouldn't know just what I was doing. I looked at the bug-list archives & didn't see anything there, but having a small system stricly made of deb packages, I'm surprised having a non-standard problem. Maybe it's just something I did; any help appreciated. Tried to include what info I could, let me know if I need more, but this one's long enough now as it is. Thanks, -- Ed Donovan [EMAIL PROTECTED]
