The following packages have been kept back
Fellows, it's time for a dist-upgrade, but The following packages have been kept back python-gtk2 python2.3-gtk2 Could it be a fault of mine? # apt-get -o Debug::pkgProblemResolver=true dist-upgrade Package perl-modules has broken dep on libcgi-pm-perl Considering libcgi-pm-perl 0 as a solution to perl-modules 152 Fixing perl-modules via remove of libcgi-pm-perl Package python2.3-gtk2 has broken dep on python-gdk-imlib Considering python-gdk-imlib 1 as a solution to python2.3-gtk2 2 Package python2.3-gtk2 has broken dep on python-glade Considering python-glade 0 as a solution to python2.3-gtk2 2 Package python2.3-gtk2 has broken dep on python-gnome Considering python-gnome 0 as a solution to python2.3-gtk2 2 Package python2.3-gtk2 has broken dep on python-gtk Considering python-gtk 4 as a solution to python2.3-gtk2 2 Holding Back python2.3-gtk2 rather than change python-gtk Fixing python2.3-gtk2 via remove of python-gdk-imlib Fixing python2.3-gtk2 via remove of python-glade Fixing python2.3-gtk2 via remove of python-gnome Investigating python-gtk2 Package python-gtk2 has broken dep on python-gtk Considering python-gtk 4 as a solution to python-gtk2 1 Holding Back python-gtk2 rather than change python-gtk The following packages will be REMOVED: libcgi-pm-perl python-gdk-imlib python-glade python-gnome -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
how could Packages mention files that don't exist yet?
How do the mirrors work? Are the Packages files always copied last, to ensure all the versions mentioned in them already exist on the disk when you try to get them? In what cases could an hours old Packages file mention packages that apparently haven't been copied to the mirror yet? E.g. http://debian.linux.org.tw/debian/pool/main/f/file/file_4.06-1_i386.deb: ERROR 404: Not Found. Plenty of space there: $ df Filesystem 1K-blocks Used Available Use% Mounted on /dev/hdc1118729904 103213752 14309912 88% /home/ftp /dev/hdd1118729904 104041660 13482004 89% /home/ftp/pub -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: how could Packages mention files that don't exist yet?
P> Possible. Could be a bad copy of files. Could be anything. Try manually P> browsing the mirror using your webbrowser or lftp etc. to see if the P> files are there. Yes, http://debian.linux.org.tw/debian/pool/main/f/file/ shows no Oct 03 additions, whereas http://debian.linux.org.tw/debian/dists/sid/main/binary-i386/Packages.gz already lists them. This must reveal a fundamental flaw in the debian mirrors system, where some mishap could cause Packages to be updated, even if not all the packages it lists were successfully updated. $ lynx -dump http://linux.cdpa.nsysu.edu.tw/debian/pool/main/f/file/|grep Oct-2003 a nearby machine, shows plenty though. Sure, maybe the next day, the hole will get filled automatically, but still, that such a condition (Packages file updated before confirming what it lists are confirmed to be successfully updated) is bad. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
what tool to view CDROM containing TV show?
Gentlemen, I have been send a CDROM with a television show on it. What might be the right tool to view it? The CDROM's structure is: $ tree -s |-- [ 27] autorun.inf |-- [ 2048] cdda |-- [ 2048] cdi |-- [ 2048] ext |-- [ 155648] mpeg2dmx.ax |-- [ 2048] mpegav | `-- [445904896] avseq01.dat |-- [ 129024] mpgdec.ax |-- [36864] playwnd.exe |-- [ 2048] segment `-- [ 2048] vcd |-- [ 2048] entries.vcd |-- [ 2048] info.vcd |-- [ 2048] lot.vcd `-- [ 2048] psd.vcd However even with the right tool, it seems there will be additional problems, seen with similar CDs, and both my CD drives: $ wc mpegav/avseq01.dat wc: mpegav/avseq01.dat: Input/output error 0 0 12288 mpegav/avseq01.dat In /var/log/kern.log: I/O error: dev 0b:00, sector 2436 Of course trying mount -o sync doesn't help either. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
only one new debian-user thread a day according to google
After Oct 21st, this group has suddenly dwindled to only about one new thread a day: http://groups.google.com/groups?group=linux.debian.user Some of us low bandwidth users read via google, as gmane.org's search isn't as good yet. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
how to apt-get the smaller Packages.bz2s?
I notice there are Packages.bz2 and Packages.gz on the mirrors. How to tell 'apt-get update' to get the smaller .bz2s? -- http://jidanni.org/ Taiwan(04)25854780 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
give cdrom source more priority in apt-get
Assuming $ apt-cache policy twlog twlog: Installed: (none) Candidate: 1.3-2 Version Table: 1.3-2 0 500 http://linux.cdpa.nsysu.edu.tw sid/main Packages 500 cdrom://[Debian GNU/Linux SID _Sid_ - fsn.hu unofficial i386 Binary-4 (20021010)] unstable/main Packages How can I tell apt-get install that it should try the cdrom first? Maybe the apt_preferences man page mentions it, The final selection method is by origin. This is simply the site name of the originating package files. The empty string is used for file URIs. Is that what I want to use? If so, how? Shall I make the cdrom priority 501? -- http://jidanni.org/ Taiwan(04)25854780 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
apt 1000 different levels of priority mere eye candy?
Is it true that the apparently 1000+ different levels of priority mentioned on the apt_preferences(5) man page are actually just a handful? If one manages to make 'apt-cache policy' report one item with priority 504 and another with priority 506, there are still other factors that override this and end one up apt-getting the 504 version, e.g. if it appears first in the 'apt-cache policy' output? If one really wants a higher number to mean something one must cross one of the boundaries mentioned on the apt_preferences man page? Is it like "all students with B's are candidates if there are no A's. We don't look at the percent grade."? -- ...because "Barksdale" is quite a mouthful. --Bdale Garbee http://jidanni.org/ Taiwan(04)25854780 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
XIM vs. mozilla
Even though /etc/mozilla-*/prefs.js has
// XIM style
pref("xim.input_style", "over-the-spot");
but I still can't use it. I.e. forget about typing Chinese into
mozilla-snapshot, etc.
Apparently that how life is for the last few months, bug reports or not.
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
add devfs=mount as a kernel command-line boot parameter?
LILO version 22.5.7.2 says Warning: '/proc/partitions' does not match '/dev' directory structure. The kernel was compiled with DEVFS_FS, but 'devfs=mount' was omitted as a kernel command-line boot parameter; hence, the '/dev' directory structure does not reflect DEVFS_FS device names. So should I add devfs=mount as a kernel command-line boot parameter? http://jidanni.org/comp/lilo.conf is my lilo.conf. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
just get the changelog
What is the URL these days to just get a changelog? http://cgi.debian.org/cgi-bin/get-changelog?package=somepackage doesn't work. Aptitude's "C" won't work. Is http://packages.qa.debian.org/s/somepackage.html and then browsing news items there the best one can do without downloading any extra files in addition? -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
apt-mature: take the edge off the bleeding edge of sid
Announcing apt-mature, which should supposedly tell you how long upgradable packages have been on the mirror so you can pick older ones etc., all offline. http://jidanni.org/comp/apt-mature -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: obsolete Suggests field seen on apt-get installs
Regarding seeing obsolete Suggests fields when doing apt-get installs, it turns out that just doing a daily apt-get update isn't enough, one should instead do a dselect update, it turns out. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
ps axu|grep grep doesn't output always
Just curious, why doesn't this make output every run: $ ps axu|grep grep jidanni 1163 0.0 0.2 2376 608 pts/1R02:26 0:00 grep grep $ ps axu|grep grep $ ps axu|grep grep $ ps axu|grep grep $ ps axu|grep grep jidanni 1171 0.0 0.2 2376 608 pts/1R02:26 0:00 grep grep Is it some kind of race condition? -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
rescue mail off secondary server contingency plan
Let's assume it's Friday and one's primary mx server cannot be fixed until Monday. Maybe I can write a standard procedure for other low bandwidth users to use. Mail is accruing on the secondary, but not yet cleaned of spam, nor in a mailbox fetchable. We copy our latest .procmailrc, and .spamassassin/user_prefs (hopefully ok for the spamassassin version on the secondary) to the secondary, then trigger some special command the dumps the mail queued for us that is waiting for the down primary mx server, into our mailbox here on the secondary, as per our .procmailrc. We then use fetchmail to fetch it. We disconnect. New mail still gets queued waiting for the primary, until if we repeat the procedure. I wonder if the above is somewhat on track. More flexible would be: I make the mx record for the primary now point to a good machine where all is ready to do my spam filtering. But that would take time to propagate, I suspect. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: rescue mail off secondary server contingency plan
>> Let's assume it's Friday and one's primary mx server cannot be fixed >> until Monday. Maybe I can write a standard procedure for other low >> bandwidth users to use. S> It has been around for years, and is called uucp. and how does one use uucp to manipulate the exim queue on a remote machine? is uucp ssh-ready? It seems manipulating the mx record to make the primary point to a backup machine would be the right way. Then have the secondary machine refresh its DNS, so it then sends all queued mail to the backup, no? -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
exim.conf for the man with two ISPs
Dear fellas, outgoing mail should be sent thru my ISP's SMTP server,
so my exim.conf has
smarthost:
driver = domainlist
transport = remote_smtp
route_list = "* ms46.myisp.com bydns_a"
end
But let's say now I have and an additional ISP that I sometimes
connect to, with its different SMTP server. What would be a good way
to tell exim [3] that that rule should be different short of editing
exim.conf for those calls? I can make a script: do_something;pon isp2
but what should that something be, short of an ed(1) script to edit
exim.conf?
Question two: actually that second ISP has no SMTP server, so I
thought of a neat rule that would cover both ISPs: first try to
deliver an email directly, then if rejected ("we don't accept mail
from known dialups", etc.), try thru myisp.com... how does one write
that?
BTW, my setup is fine but sometimes I turn off the computer with
unsent messages and turn it on a few days later and exim sends notices
that messages have not been delivered for more than a day, but it send
them quite understandably thru my ISP as it was warning the sender
which was my full address. So in /etc/aliases I put
[EMAIL PROTECTED]:jidanni
but I don't think that worked as /etc/aliases is for local addresses
or something, so how do I tell exim that if it ever sees a
[EMAIL PROTECTED] it should send it to jidanni and not out to the
net and back when sending its 'day old warning' messages?
By the way, in the official examples we see:
route_list = dict.ref.book mail-1.ref.book:mail-2.ref.book byname
Would the average person write .com or .org where he writes .book or
is he talking about subdomains or something?
--
http://jidanni.org/ Taiwan(04)25854780
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
sid or sarge for wimp user
In December a friend will burn 9 debian CDs for me to upgrade my poorly networked 'woody of may 02' system. Shall I have him burn sid or sarge? I'm basically a wimp user but need fresh software so I don't sound too silly in my many bug reports. What are the chances my system will come to a screeching halt with sid vs. sarge? After I do the upgrade I don't suppose it is easy to revert if not happy. -- http://jidanni.org/ Taiwan(04)25854780 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
debian user vs. browser plug-ins
Just curious what the debian user does if he wants to browse like his microsoft counterpart when encountering flash, java, etc. 'enhanced' web pages. These cause mozilla 0.99 to ask me to download some 'plug-ins' in some wacky fail prone way on my frail costly modem connection. Why can't we 'plug-in' all the plug-ins before hand from our debian CD set? I'm tired of complaining that various sites shouldn't be using all this 'advanced technology'. Instead I was supposing that the highly advanced debian system with 3 times as many packages as other distributions certainly somehow can meet and exceed the convenience of a microsoft user when encountering .swf etc. Yes I noted some packages to deal with .swf, but what is the 'integrated' way? -- http://jidanni.org/ Taiwan(04)25854780 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
10MB lost per CD multi-session?
http://www.cliffshade.com/dpfwiw/cd-r.htm#multi says each time you add another session (more data) to a disc you loose 10MB of space to the needed header files for the new session. Is it that serious on GNU/Linux with cdrecord, and if so, I don't think I saw it in the docs. -- http://jidanni.org/ Taiwan(04)25854780 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
wget PASV no route to host
I updated to sid and now have a wget failure: --06:52:39-- ftp://ftpsv.cwb.gov.tw/pub/satellite/s1q2002-12-10_06.jpg (try:20) => `s1q2002-12-10_06.jpg' Connecting to ftpsv.cwb.gov.tw[163.29.179.241]:21... connected. Logging in as anonymous ... Logged in! ==> SYST ... done.==> PWD ... done. ==> TYPE I ... done. ==> CWD /pub/satellite ... done. ==> PASV ... couldn't connect to 163.29.179.241:58773: No route to host Giving up. I don't see any iptables blocking activity by me at that time in syslog. P.S. I see portmap is running. Was that one of the things we were supposed to turn off for security? -- http://jidanni.org/ Taiwan(04)25854780 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: 10MB lost per CD multi-session?
> "E" == Eduard Bloch <[EMAIL PROTECTED]> writes: E> Blind? E> # zgrep overhead /usr/share/doc/cdrecord/README.multi.gz E>overhead + 150 sectors for the pre-gap of the first track after the E>overhead + 150 sectors for the pre-gap of the first track after the E> The total amount of the overhead is 11250 + 150 sectors for the first additional Oh, OK, but it should be emphasised again, and in bytes, otherwise dummies like me will keep on making 3MB incremental backups thinking that we are thrifty. It also says "So in theory, the can be 44 sessions on s 74 minute disk" ^? OK, I see, if we fill it up with separators and no content, we can have 44 separators. -- http://jidanni.org/ Taiwan(04)25854780 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
looking at deborphan's output first
Is part of deborphan's job to clean up for sloppy package maintainers' bad Depends: etc. fields? Or during normal usage does a debian system accrue useless packages? In /usr/share/doc/deborphan/README.gz : REMOVING ALL SUGGESTED PACKAGES --- deborphan -p4 | xargs apt-get -u remove Never run this command without looking at deborphan's output first! $ deborphan|xargs|fold -s|sed 's/ $//' libgal19 gstreamer-vorbis libaiksaurusgtk0 libcapplet0 libcapplet1 libident libmetakit1 zlibc libcdk4 libmimelib1 libmagick5 libole2-0 libzipios++0 libid3-3.7-13 libgtop1 libwraster2 libqt2-mt libfltk1 libgmime0 libc-client2001 gstreamer-audiofile tix41 libghttp1 tclx8.3 libarts1-audiofile libjack0.38.0-0 libflash0 libsigc++0 libarts1-xine libpgsql2 libnautilus2-2 libgdbmg1 OK, I looked at the output as instructed. Now I run the dangerous command... Wait, by looking at the output I think he means look for names of loved ones that are to be spared... etc. Any sure tests? Would I have linked to one of them if I never compile anything on my own? -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
removed kde but remnants remain
I recall I at one point "removed kde from my system", however I still see # dpkg -l kde\*|grep ^i ii kdeaddons-doc- 3.1.2-1KDE add-ons documentation in HTML format ii kdeartwork-the 3.1.2-2icon themes released with KDE ii kdebase-data 3.1.2-1KDE Base (shared data) ii kdeedu-data3.1.2-1shared data for KDE educational applications ii kdegames-card- 3.1.2-2Card decks for KDE games ii kdelibs3-doc 2.2.2-14 KDE core library documentation ii kdetoys-doc-ht 3.1.2-1KDE toys documentation in HTML format This must be because their cross-dependencies are inadequately intertwined, so they didn't go away with the rest, and are left sitting there mostly useless on their own[?], so I must remove them by hand if I want to really clean things up, eh? -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
let's remove fortran from my system
I never use fortran, so # apt-cache show libg2c0 Description: Runtime library for GNU Fortran 77 applications # apt-get --purge remove libg2c0 The following packages will be REMOVED: atlas2-3dnow* libg2c0* pdl* r-base* r-base-core* r-base-html* r-base-latex* After unpacking 56.1MB disk space will be freed. Do you want to continue? [Y/n] n Uh oh doesn't look that simple. I don't recall using the R language either... But removing pdl, # apt-cache show pdl Description: The perl data language. Perl extensions for numerics. looks threatening to the system. Ok, better quit while I'm ahead. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
can see almost all languages in mozilla
Looking at the bottom of http://www.debian.org/Bugs/ I see the names of various languages in their native fashion. Everything is hunky dory, except the Korean (Hangul), I see squares with hex in them for the 한국어 Did I forget to apt-get something, or is Mozilla Debian Package 1.3.1-3 deficient? (I would upgrade and check again, but I'm on a modem.) -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
/var/lib/dpkg/available a day behind apt
Gentlemen, my /var/lib/dpkg/available file always seems a day behind apt:
$ grep-available -s version -XP zsh
version: 4.0.7-6
$ apt-cache policy zsh
Installed: 4.0.7-3
Candidate: 4.0.7-6
$ grep-available -s version -XP base-config
version: 1.67
$ apt-cache policy base-config
Installed: 1.67
Candidate: 1.68
So it's not if a package has been installed or not. It just takes an
extra day for dpkg to know about it after apt, for me here.
Furthermore,
$ cmp /var/lib/dpkg/available*
always shows that the -old backup is the same as what it is supposed
to backup.
P.S. Is it normal to have a /var/lib/dpkg/lock always present 24 hours a day?
Do real apt-get update users see this too? (Or is it because I'm, erm,
using a rsync hack for modem users?) I do apt-cache gencaches. Must I
also everyday do something like
dpkg --clear-avail
find /var/lib/apt/lists -name \*Packages -newer /var/lib/apt/lists/partial \
-exec dpkg --merge-avail {} \;
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
APT::Cache-Limit 16777216 still needed?
How can one tell if I still need to set APT::Cache-Limit 16777216; for the current version of apt? $ apt-config dump doesn't say if the default value has been increased. $ which apt-get|xargs strings -a|grep -i limit doesn't help. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
can't post via gmane.linux.debian.user
I notice nntp postings via gmane.org's gmane.linux.debian.user to this group never get posted, even though NNTP returns 240 OK and http://lists.debian.org/debian-user/ says "This list is not moderated, posting is allowed to anyone." On gmane.discuss they have no further solutions. I don't know if the following is related: >> Which newsgroups are you gatewaying to, and which groups are >> read-only? I think if we isolate that, we might be able to move >> forward. Jan> I'm reading the ML as 'gmane.linux.debian.user' and I think I also Jan> have 'linux.debian.user' in my list. Jan> The first handles postings to the Newsgroup quite well, the latter is Jan> AFAIK not as good. -> news.gmane.org -> www.gmane.org -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
apt-get install c-compiler
# apt-get install c-compiler
Package c-compiler is a virtual package provided by:
gcc-3.0 1:3.0.4-13
tcc 0.9.19-1
gcc272 2.7.2.3-18
gcc-3.3 1:3.3.1-1
gcc-3.2 1:3.2.3-7
gcc-2.95 1:2.95.4-17
gcc 4:3.3.1-1
bcc 0.16.3-2
altgcc 1:2.7.2.3-2
You should explicitly select one to install.
# apt-get install c-compiler 2>&-|awk '/^ /{print $1}'|xargs -r dpkg -l|grep ^i
ii gcc-3.03.0.4-13 The GNU C compiler.
ii gcc-3.33.3.1-0pre0The GNU C compiler
ii gcc-3.23.2.3-6The GNU C compiler
ii gcc-2.95 2.95.4-17 The GNU C compiler.
ii gcc3.3-2 The GNU C compiler.
Indeed I see I have a "subscription" to every new version, with no
automatic mechanism to get rid of the old. Is this the ideal situation?
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: can see almost all languages in mozilla
Ladislav> apt-get install ttf-baekmuk should solve your problem.
OK, now looking at Japanese, I occasionally notice boxes with "9A28",
"6CA2", etc. I must be missing a few chars. What would the
mozilla user of today be sure to have apt-gotten in order to be
prepared for almost any language?
Do I just do
COLUMNS=222 dpkg -l ttf\*|awk '/^pn/{print $2}'|xargs apt-get install
I.e. get all the ttf- packages?
Wait, there are problems with that and some conflicts, and even some
free/non-free decisions. Hmmm, then I suppose I must ask for a list
of what to apt-get still.
One moment, now looking at Japanese under emacs, it turns out some
very common characters are missing, so something must be maladjusted
in mozilla.
For instance, in a charset=EUC-JP document, mozilla shows a square
"5FDC" where emacs shows:
character: 応 (0154376, 0, 0xd8fe)
charset: japanese-jisx0208 (JISX0208.1983/1990 Japanese Kanji: ISO-IR-87)
code point: 49 126
buffer code: 0x92 0xB1 0xFE
file code: B1 FE (encoded by coding system japanese-iso-8bit-unix)
font: -JIS-Fixed-Medium-R-Normal--16-150-75-75-C-160-JISX0208.1983-0
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: APT::Cache-Limit 16777216 still needed?
Greg> It is needed. If you have a ba-zillion deb and deb-src lines and all Greg> three/four Debian tick-marks. (stable, testing, unstable, experimental). I see. Can one tell what APT::Cache-Limit value is without resorting to the source? apt-config dump doesn't tell default values. Would adding this capability be a good wishlist item? However, I was told not to report any more apt bugs. So perhaps somebody else could. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Bug#205997: /usr/bin/apt-get: package's new versions ignoredafter failed config
Just curious, is it best that apt-get, aware that there is a newer version of a program, still no questions asked, just tries again to install the old stuck version --- would any variation in that behavior have any even wishlist merit? >>>>> "M" == Matt Zimmerman <[EMAIL PROTECTED]> writes: M> On Mon, Aug 18, 2003 at 07:11:17AM +0800, Dan Jacobson wrote: >> Package: apt >> Version: 0.5.9 >> Severity: normal >> File: /usr/bin/apt-get >> >> Consider the case of netbase 4.11, a package with a broken script that >> fails configuration. >> >> Then one does apt-get update and 4.13 is now the candidate as seen in >> apt-cache policy. M> apt always tries to fix any broken packages on the system (or let you M> specify a solution) before letting you try to install new ones. This is a M> feature. It is not installing anything; it is only trying to configure the M> package which was already unpacked on your system, and failed to configure M> previously. M> Did you consider copying and pasting the output showing exactly what apt was M> doing, rather than trying to explain what you thought it was doing? That M> would have saved me the guesswork of figuring out what you were talking M> about. M> I am not going to provide technical support to you through the BTS. Your M> bug reports are not contributing to the development process. So, please do M> not file any more bug reports against apt. If you have a question about why M> apt is behaving a certain way, ask on [EMAIL PROTECTED] Do not M> file a bug report. Is this clear? M> -- M> - mdz -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
backing out from apt-get postinst failures
Just curious. If a $ apt-get install some_package gets stuck in some bug in /var/lib/dpkg/info/some_package.postinst what is the proper way to back out until a newer version is available? At this point restoring ones system to the previous state is now beyond the ability of apt-get and will take several dpkg commands, trying to install older .deb's. eh? -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: backing out from apt-get postinst failures
>> $ apt-get install some_package >> gets stuck in some bug in /var/lib/dpkg/info/some_package.postinst >> what is the proper way to back out until a newer version is available? C> Usually, edit the postinst script until it doesn't get stuck. I assume if one does not want to hack the system, the answer would be to find a earlier .deb, and use dpkg -i on it. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
debian lists have no nomail option
I was surprised to find that some debian lists require a subscription to post, but have no "nomail" option. This means e.g. gmane.org users will get a copy of each message, even though they've already read them. For modem users, for the privilege of posting one message, one must get one's POP box buried with extra copies of messages one normally reads elsewhere. >>>>> "D" == Debian Bug Tracking System <[EMAIL PROTECTED]> writes: D> This is an automatic notification regarding your Bug report D> #206959: lists.debian.org: no options control information, D> Dan Jacobson wrote: >> Package: lists.debian.org >> Version: unavailable; reported 2003-08-24 >> Severity: normal >> >> Gentlemen, I subscribed to a list and was surprised that the user is >> told no way how to control the list options. Not in the welcome D> Since there are none. Thanks. OK, then mention that there, else we would have never imagined it. D> Regards, D> Joey -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
dselect: how to jump to next section?
In dselect, hitting O gives me: - All packages - @-- Installed packages --- - Installed Required packages -... Say my cursor is on the @, how do I jump to the end of Installed packages? -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
spamdealer: smtp time filter for non-root offline users
Is my http://jidanni.org/comp/spam/spamdealer.html a breakthrough for non-root offline users? If so, maybe someone can give me an account so I can set up camp to get my email back into shape. -- http://jidanni.org/ Taiwan(04)25854780 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Packages file under version control
I am fortunate to have an account on debian.linux.org.tw, I.e. a
mirror. Maybe I can make a script to just do a diff of the Packages
file and then just download this diff over my modem?
I hate downloading megabytes just because I want to upgrade some puny
package without resorting to dpkg.
I would use ssh to run a script that on the mirror diffs the current
Packages with a copy of Packages from the last time I ran the script.
Then I would scp this diff.gz over my modem, and then apply it to the
Packages on my home computer. Does that make sense? Is there some
modem overhead due to being forced to use scp as that's all I can use
there? Don't tell me to learn some complex new software just to
retrieve this one file, though.
Before I noticed this thread in the debian weekly news, I was going to send:
Subj: Packages.gz too big for modem users
Gentlemen, I found I could reduce Packages.gz files by half if I
ripped out Descriptions. Makefile:
#see how shorter Package files are if we rip out descriptions
no_desc:
/var/lib/apt/lists/debian.linux.org.tw_debian_dists_sid_main_binary-i386_Packages
awk '/^$$/;/^Description:/,/^$$/{next};1' $?
You see, doing apt-get update is too much for us modem users,
especially when all we then want to install is a measly few Kbyte
package. Hence, I now mostly just download the .debs and use dpkg -i.
Anyways, apparently one could even strip out more unnecessary fields,
just leaving the dependency info so apt-get would still work... one
would do this stripping on a remote server, and use one's modem to
download the stripped Packages.gz file, every time one would want to
do a apt-get update, perhaps only getting the real Packages.gz file
once a month, so "apt-cache show" would still show some semi-fresh
description info.
Anyway, I'm just thinking of ways costly phoneline users could use
debian's package system better... I'm not sure if my ideas actually
work, i.e. I haven't actually fed these files into apt yet.
--
http://jidanni.org/ Taiwan(04)25854780
Hope my mail address is working by now, if not see my contact.html page
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Packages file under version control
$ wget http://home.tiscali.cz:8080/~cz210552/download/apt-rsync $ chmod +x apt-rsync Traceback (most recent call last): File "./apt-rsync", line 58, in ? f=PrivoxyWindowOpen(SOURCES) NameError: name 'PrivoxyWindowOpen' is not defined Now what? I do in fact use privoxy BTW. Of course the author has no email. -- http://jidanni.org/ Taiwan(04)25854780 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
why is the access time of dirs not chaged?
The following message is a courtesy copy of an article that has been posted to comp.unix.shell as well. If I do "cat dir/file", why is the access time of dir not changed, as seen with stat(1) or ls -ul? Only if we do "ls dir" will dir's access time be changed. Maybe it is on purpose to avoid too much disk access. In version 7 Unix did they do that too? -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
first audio hopes dashed: drain playback error -512
Dear sirs, I have connected speakers to my system (http://jidanni.org/comp/system.txt), done # modprobe via82cxxx_audio # play file.au But nothing happens and I have to hit C-c to get my prompt back. # dmesg Via 686a audio driver 1.9.1 ac97_codec: AC97 Audio codec, id: 0x414c:0x4710 (ALC200/200P) via82cxxx: board #1 at 0xE400, IRQ 12 via_audio: ignoring drain playback error -512 I consulted the various HOWTO's but like they are all for just any distribution whereas I can just apt-get... anyways, help. -- http://jidanni.org/ Taiwan(04)25854780 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
need Ph.D. for sound even with common hardware?
Apparently I can forget about sound as the Sid CD of 20021010 that I'm [rural modem user] using has Alsa versions clustered around a flurry of changes and http://www.linuxorbit.com/modules.php?op=modload&name=Sections&file=index&req=viewarticle&artid=541&page=1 says I have to compile from alsa-source anyway. Anyway, all too much of a challenge for me. Yes I also read http://www.tux.org/~tbr/sound-debian/ kernel-source-2.4.18/Documentation/sound/Introduction etc. I suppose it is too much to ask for a simple way a simple user can use sound on Linux version 2.4.18-k7 with VIA VT8233 AC97 Audio Controller. I looked at many a Debian Sound HowTo already. http://jidanni.org/comp/system.txt is what I've got. -- http://jidanni.org/ Taiwan(04)25854780 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: need Ph.D. for sound even with common hardware?
How kind of you to help me find the simplest way to make my system give even the littlest bleat. I did # modprobe ac97 # modprobe ac97_codec # modprobe via82cxxx_audio # cat bleat.au > /dev/audio # tail /var/log/syslog Via 686a audio driver 1.9.1 ac97_codec: AC97 Audio codec, id: 0x414c:0x4710 (ALC200/200P) via82cxxx: board #1 at 0xE400, IRQ 12 Via 686a audio driver 1.9.1 ac97_codec: AC97 Audio codec, id: 0x414c:0x4710 (ALC200/200P) via82cxxx: board #1 at 0xE400, IRQ 12 via_audio: ignoring drain playback error -512 isn't 0xE400 what my http://jidanni.org/comp/system.txt says it should be? I want to try this the non Alsa way to reduce complexity. I cannot find tool "via-audio-diag" mentioned in some articles. Drain playback error -512 is not explained by google searching. -- http://jidanni.org/ Taiwan(04)25854780 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: need Ph.D. for sound even with common hardware?
|It sounds like you just don't have sound.
I confirmed that if the jack is pulled out of the Epox 8KHA+ 'speaker'
jack in the manual's photo, and stuck into a radio, there is sound.
|First make sure you're in the audio group (" addgroup dan
|audio" - then log out and back in as "dan" -- type "groups" to verify
|you're in the audio group).
or even as root, no sound.
|Second, kill X (for now), and make sure you don't have any sound daemons
|(esd, artsd) running.
I thought init 3 would kill x. well, i had to do init 1. then i did
the tests as root. same deal. Anyways, a pstree shows no sound
related things running. i use icewm BTW.
|Use "splay http://jidanni.org/ Taiwan(04)25854780
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: need Ph.D. for sound even with common hardware?
I have apparently now gotten beep(1) to beep thru the speaker, [as well as the tiny PC speaker like it used to.] This I did by aumix's Vol and Spkr controls. How then can I make play(1),cat $file > /dev/audio, splay, mpg123, mpg321,mpg123-oss etc. also make the speaker come alive? they do noting at present except just hang there waiting for a C-c. How can the lowly beep program do it but they can't? It's not a permission problem as i've long ago minded /etc/groups and my experiments are carried out as root anyway. $ beep & find /proc/$! -type f|xargs more|less is what I thought might give me a clue why, but i don't understand it... I removed the sound associated modules from the kernel and it still it beeping thru the speaker... Regarding via-audio-diag, # /home/jidanni/tmp/via82cxxx-1.9.1/via-audio-diag via-audio-diag.c:v1.00 05/06/2000 Jeff Garzik ([EMAIL PROTECTED]) Unable to find a recognized card in /proc/pci. If there is a card in the machine, explicitly set the I/O port address using '-p -t ' Use '-t -1' to see the valid chip types. # /home/jidanni/tmp/via82cxxx-1.9.1/via-audio-diag -t -l /home/jidanni/tmp/via82cxxx-1.9.1/via-audio-diag: invalid option -- t /home/jidanni/tmp/via82cxxx-1.9.1/via-audio-diag: invalid option -- l Usage: via-audio-diag [-aps] [-p ] # /home/jidanni/tmp/via82cxxx-1.9.1/via-audio-diag -aps -p 0xE400 didn't help, never mind. -- http://jidanni.org/ Taiwan(04)25854780 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
just a small package and its docs, spread over many CDs
Just to install gocr,gocr-gtk,gocr-tk,gocr-doc I had to insert three different SID CD's. Is that normal? I wonder what if I wanted the other parts of it... -- http://jidanni.org/ Taiwan(04)25854780 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
files who won't tell who their parent packages are
Let's say a file is bothering me and I want to report it to its parents or owners or anything $ locate sndstat /dev/sndstat #no similarly named brothers $ apropos sndstat sndstat: nothing appropriate. $ dpkg -S /dev/sndstat dpkg: /dev/sndstat not found. $ reportbug -f /dev/sndstat Finding package for /dev/sndstat No packages match. No package specified; stopping. Is it therefore true that there can be some files, say created by scripts, who have laundered their origins, so that there is no record of what caused them to get there, and being that they arrived via installation of a package, that some better way should be invented of keeping track of them... -- http://jidanni.org/ Taiwan(04)25854780 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Computer beeping but no sound on
> "H" == Hanasaki JiJi <[EMAIL PROTECTED]> writes: H> Any thoughts as to why my system would be beeping? Its the speaker H> inside the case. The speakers are off and the Linux sound control is H> muted. When mute is off, the beeps come from the speakers. My problem is ever since I turned up the volume with aumix, echo -e \\a>/dev/console now comes out my fancy speakers as well as the built in one. However every other program or way of playing an audio file just hangs. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
SmartMedia: buy what card reader?
I'm about to buy a smartmedia card reader for my http://jidanni.org/comp/system.txt . Can I get away with just a $15 single reader, or must I get a 5 in 1 with onboard memory at 4 times the price to avoid coming all the way back home and it won't work? I suppose USB is the best bet these days? -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
^G beeps also come out my fancy speakers
Gentlemen, what is the deal when one can make C-g beeps also come out one's speaker that is plugged into the speaker jack, with modprobe -v via82cxxx_audio aumix -v 88 -p 88 modprobe -rv via82cxxx_audio echo -e \\a > /dev/console as well as the PC board speaker, and even after the modules are removed? Of course this is the only sound I've managed to get out my Million Dollar speakers. My system is http://jidanni.org/comp/system.txt -- http://jidanni.org/ Taiwan(04)25854780 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
am I a sitting duck for fetchmail hole because can't upgrade?
Just curious, we see a security bulletin that fetchmail should be upgraded to avoid a malformed hostname threat. I could wget the new fetchmail .deb with my flimsy modem but I see it depends on libc6 that is newer than the one on the sid cd set of 2002.10.10 that I have installed. If I also get the new libc6 that probably means I got to make a lot of other changes too which means I really should have a new CD set? Which means I have to go to town and ask my friend to burn a new set which is something I didn't want to have to do more than once a year. So I am a sitting duck for a malformed mail message? Perhaps as at least my other pal runs an spamassassin and exim filter on my mail before I download it with fetchmail, perhaps I can change a rule in .exim/filter to protect myself? Or maybe something can be done in fetchmailrc? P.S. on http://packages.debian.org/unstable/mail/fetchmail-ssl.html it says it depends on "fetchmail-common (= 6.1.2-1) (NOT AVAILABLE)" Does that mean one is screwed either way? -- http://jidanni.org/ Taiwan(04)25854780 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
old /lib/modules still there after purging kernel
I long ago stopped using the older kernel, 2.2. However in /lib/modules/ I see 2.2.20/ 2.2.20-idepci/ 2.4.18-k7/ . What is the recommended way of cleaning up the older ones? Is the user supposed to just rm -r? -- http://jidanni.org/ Taiwan(04)25854780 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
is my exim filter with white word cool?
Fellas, what do you think of my exim filter file filtering spamassassin results... I first give 'em a pass-word in case they're real. [You see, I don't operate spamassassin, just have access to its results.] I do the best I can with the paltry exim user assignable variables available, numbers. add 334424 to n9 if $h_subject: contains $n9 then logwrite "**Passed with $n9: $tod_log $h_from $h_subject" finish endif #--- if $h_X-Spam-Status: contains FROM_NO_USER then logwrite "FROM_NO_USER spam: $tod_log $h_from $h_subject" fail "FROM_NO_USER spam? If not, put \"$n9\" in Subject." seen finish endif #--- etc. etc., the whole file is in http://jidanni.org/test/exim-filter -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
sound possible on Epox 8kha+ motherboard?
Did anybody ever get sound to work on your EPOX 8kha+ board? I have 10/10/02 sid cds. I tried everything. I can make the console beep out of the speaker only. Alsaconf doesn't have the card. http://jidanni.org/comp/system.txt is my sorry system. I can use kernel -18 or -19. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
console messages flit by too fast
I can't believe this is still not solved. I boot the system. I see some ominous warnings. They scroll by so fast and are gone. E.g. something about mtab. Ok, cd /var/log; grep mtab * */* Nothing. You will answer "oh, mtab, don't worry about that". But the general problem of messages that appear at boot but go off the screen is not solved for me still. Am I really supposed to hit ^S^Q like back 30 years ago? What if I am not fast enough still? You will answer: that is the fault of whatever package author for also not logging his message to syslog. But that isn't helping me: what package? the name went by too fast. Isn't there something that I can turn on to capture all these, or are we too early in the startup? Am I supposed to boot my system thru some remote terminal like the certainly must do a Linux Labs so they can scroll back? But I only have 1 equipment and am not into learning something fancy. Perhaps all the stupid questions I post about why I can't enter the audio age could be solved if I didn't miss one of those ominous messages that go by too fast. I hit ALT CTRL F1 etc. but I don't suppose those `man console(4)` tty can be scrolled backwards. -- http://jidanni.org/ Taiwan(04)25854780 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
make CD of only newest 3 months of sid
Dear Debian Sirs, a few months ago I want to town and my friend made me 10 cd's: _Sid_ - fsn.hu unofficial i386 Binary (20021010)] Now if I go to town again can I say, have him use the "jigdo" mechanism to make just a few cd's that hold only the changes to sid since 20021010, avoiding the embarrassment of having him do the whole job over? Or maybe make some kind of custom CD of anything in sid newer than 20021009? It seems any of these would me more trouble than having him burn me a whole new set. Or maybe I should just keep my mouth shut and not ask for an update until say 6 months have passed... ah, the luck of the piddly modem user. (BTW, I would use apt-cdrom on the update CD, that is no problem, I can handle that.) -- http://jidanni.org/ Taiwan(04)25854780 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
CDROM's headphone jack old-fashioned?
What are the 4 wires from "analog" prongs of my CDROM unit to the "CD1" prongs on my motherboard for? Is this an old fashioned way of listening to music? How about the headphone jack on the front of the CDROM unit? Old fashioned too? Take my "RICOH CD-R/RW MP7080A, ATAPI CD/DVD-ROM drive", which seems to have difficulty often realizing that a cd had been inserted. Anyway, say I retire it. Can it then play CDs without a computer around? -- http://jidanni.org/ Taiwan(04)25854780 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: sound possible on Epox 8kha+ motherboard?
The following message is a courtesy copy of an article that has been posted to alt.comp.periphs.mainboard.epox as well. >> Did anybody ever get sound to work on your EPOX 8kha+ board with >> debian? I have 10/10/02 sid cds. I can make the console beep out >> of the speaker only. Rico> Would you mind mailing the Output of these two Commands? Rico> lspci; lspci -n OK, I put it all in http://jidanni.org/comp/system.txt L> I have never had much joy with these embedded_audio_devices on the L> 8KHA+, the 8H5A2 (or any mobo for that matter). When I have L> managed to get them to work, they have never sounded that great L> either. L> IMHO, these embedded_audio_devices are more trouble that they are worth. L> My approach is to disable the embedded device and to install a L> relatively inexpensive (e.g. Ensoniq 5880 based) pci audio card. I just traded in my old computer with its card still inside ... I'd hate to have to make a trip to town to buy one... One sign of trouble is I tried knoppix and it made lots of complaints upon boot [which of course flew off the screen and are not in /var/log...] Seems 2.4.19 is worse than 18 here... Anyway, it must be just the IRQ's or whatever... if you have the recipe I don't think I would notice quality issues... I mean if I can make it beep out the speakers with echo -e \\a>/dev/console that means I must not be too far off? A> Alsa 0.9.0rc6 seems to support your card, look here: A> http://www.alsa-project.org/alsa-doc/doc-php/template.php3?company=VIA&card=&chip=via8233a&module=via82xx A> You might have to compile it yourself, I don't know whether 0.9.0-rc6 A> is in Debian stable. I use Debain testing, there you can get a package A> for it. since it's not easy for me to update my SID 20021010 cd set, nor do I dare to 'compile the kernel', might there be a way to do it with 'sox'? One bleat will satisfy me. P.S. wildly entering commands produces this dmesg: Via 686a audio driver 1.9.1 ac97_codec: AC97 Audio codec, id: 0x414c:0x4710 (ALC200/200P) via82cxxx: board #1 at 0xE400, IRQ 12 via_audio: ignoring drain playback error -512 ALSA ../../../alsa-kernel/core/seq/oss/seq_oss.c:225: can't register device seq VIA 8233 soundcard not found or device busy ALSA ../../alsa-kernel/drivers/mpu401/mpu401.c:70: specify snd_port MPU-401 device not found or device busy -- http://jidanni.org/ Taiwan(04)25854780 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
hard to find out what those daemons are about
I do pstree and I see "khubd". "What could that be about?," I say. I
try locate, man, apt-cache search, looking in /proc/{it's pid},
everything I can think of. After removing the "k",
find /var/tmp/kernel-source-2.4.18/|grep hubd
finds some uncommented files that may have something to do with it...
Anyway, is it really sometimes that hard to track down things one sees
in ps(1) output?
--
http://jidanni.org/ Taiwan(04)25854780
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: console messages flit by too fast
> Isn't there something that I can turn on to capture all these, or are > we too early in the startup? < Try using dmesg (or less /var/log/dmesg) i wanted to see the ones that dont go in there > when booting the machine, hook up the console to another system > if you only have 1 system, I'd suggest getting another. just to capture a few lines of ascii right there on the linux console > I hit ALT CTRL F1 etc. but I don't suppose those `man console(4)` tty can > be scrolled backwards. Have you tried scroll lock? I seem to recall it worked for me when I >had this type of issue but they go by too fast for that > Yeah, every bug report you file must be dealt with as an emergency even my mom is even worse. > Sarcastic comments like "I can't believe this is still not solved" about You mean like: "Hurmph, they expect me to use pencil and paper to copy those messages... why can't one hook some 'loopback tee' into a file from say the boot prompt... Didn't that Knoppix guy put the whole kaboodle into a ram disk or something... maybe have the console be a fifo|tee file etc..." ok, sorry :-) -- http://jidanni.org/ Taiwan(04)25854780 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
make CD of only newest month of sid
Rural modemed me goes to town every two months or so and hopes to bring back fresh sid CDs. But my pal just burned a set of 10 last month, how embarrassing to ask him to burn a whole fresh set each time. I bet the changed packages would fill up just a couple of CDs... 1. let's find just what I need for my update cd's: $ ssh my.account.on.debian.server $ touch -t 20021009 timestamp $ cd where? $ find . -newer timestamp ... #?? #anyways, generate a list of sid packages newer than timestamp Or: wget ftp://...some..list and analyze the list vs. dpkg's lists on my machine.. then somehow make CDs of just those. then use apt-cdrom to register them. >>>>> "R" == Richard Atterer <[EMAIL PROTECTED]> writes: R> Hi Dan, R> On Tue, Jan 07, 2003 at 01:58:01AM +0800, Dan Jacobson wrote: >> Now if I go to town again can I say, have him use the "jigdo" >> mechanism to make just a few cd's that hold only the changes to sid >> since 20021010, avoiding the embarrassment of having him do the whole >> job over? R> No, sorry - that's just not how jigdo works! jigdo can only recreate CDs R> that have previously been created by someone else and then made available R> for download. You're expecting it to create custom CDs which contain only R> certain packages of your choice - no-one has created such CDs, so jigdo R> can't reproduce them. R> What you /can/ do, however, is to bring all of your old CDs to your R> friend's, then he can use jigdo to update them to the latest sid snapshot but I see him at a social event where he gives me the cds he has burned the day before. i.e. no time for more interaction. R> from fsn.hu. This means that your friend will still have to burn the R> complete set of CDs, but he will not have to download the contents of all R> those CDs again, because jigdo is able to reuse those packages on the old R> CDs that are still present on the new CDs. even if i were to mail him a list of the contents of the cds and he could use that to tell jigdo, he still would need top make 10 cds i see. R> I should probably point out that this update mechanism works much better R> for stable or testing than for unstable, because unstable changes so R> rapidly. But it should still be worth the effort even for unstable. R> For details on making jigdo reuse old packages, see the HOWTO R> <http://www.dirac.org/linux/debian/jigdo/>. Anyways it seems the case where updates would only comprise <30% of the original are not ideally covered yet... Wait, maybe one of the switches to apt-get could figure out what is newer than my current CDrom set that i registered last time with apt-cdrom, and just report a list... note that it must not actually change anything, and it should see what's newer with respect to my current cds, not what packages i've installed or not on my system. then i would take that list an have cds made... but those cd's i suppose must look like those .hu ISOs, i.e. more than just .debs anyway, i believe i stumbled on a legitimate problem. assume 'mom' doesn't allow sonny to use the modem for more than 10 minutes a day. -- http://jidanni.org/ Taiwan(04)25854780 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: make CD of only newest 3 months of sid
[add a ftp mirror to sources.list and run] F> apt-get update # a few megabytes hmm, this would have permanent effects... but then I can just restore sources.list back to just my listing my current cds to make those effects go away later? Apparently all it does is add files to /var/lib/apt/lists I suppose. F> apt-get dist-upgrade --print-uris -y > file reading the man page, the -y worries me... ok don't worry. F> remove every line up to 'Need to get ...' F> cut the first field after that (the url of the file) [then have the friend who will burn the cd's do:] F> get those packages with wget -i file and put the on a cd or laptop apparently figuring out by hand how much to put on each cd... [PS, what happens one day when debian has one package so big that it wont fit on one cd :-)] [now, back at the ranch:] F> copy the packages to /var/cache/apt/archives OK, but I was thinking I would have him make a CDrom that I could use apt-cdrom on... hmmm, then one needs to do more than wget... maybe there's a tool for that stage. F> apt-get dist-upgrade -u [I suppose the -u just gives more info] F> This worked rather well for me, although I seem to remember that F> sometimes it wanted to download some more packages. [PS sorry for making two threads] -- http://jidanni.org/ Taiwan(04)25854780 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: sound possible on Epox 8kha+ motherboard?
The following message is a courtesy copy of an article that has been posted to alt.comp.periphs.mainboard.epox as well. Regarding sound on debian, indeed, I followed Colin Keefe's advice http://groups.google.com/groups?selm=20030107020013$64eb%40gated-at.bofh.it&rnum=6 with initial success after reboot (using "snd-via8233" instead of "snd-via82xx") with stock kernel 2.4.18-k7 ! -- http://jidanni.org/ Taiwan(04)25854780 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: make CD of only newest month of sid
> "R" == Richard Atterer <[EMAIL PROTECTED]> writes: R> Hi Dan, R> hmmm, there's one more possibility: apt-zip. You first do an "apt-get R> update" on your machine (will take closer to 15 than 10 minutes, I'm R> afraid!-), choose the packages you want and run apt-zip. It outputs a shell R> script which your friend can use to download the desired packages. I also noticed /usr/share/doc/apt/offline.html/index.html 's wget method. R> apt-zip doesn't create any Packages.gz file. Either your friend needs to do R> that or you could create a symlink farm on your HD to the files on the R> CD... The apt-ftparchive tool (apt-utils package) can create Packages.gz. R> Obviously, this will only output the packages you selected, and not all R> newer packages like you seem to prefer. hmmm, nobody seems to consider that. maybe I shouldn't either. R> It would be possible to write some tool which does what you describe, but R> you're on your own there... Try asking on debian-devel, maybe someone else R> has done something similar. P.S. Now I hear things like "do NOT use libc6-blah-9, use -8!" Well, after doing apt-get update, I see I am in line for -9 how can I insure the lists I am about to create "pin?" the libc version? Also for getting sid apparently there are other connoisseur issues i should be aware of? -- http://jidanni.org/ Taiwan(04)25854780 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
apt-get --print-uris: why filenames sometimes different?
$ man apt-get
--print-uris Instead of fetching the files to install their
URIs are printed. Each URI will have the path, the
destination file name, ... Note that the file name to
write to will not always match the file name on the
remote site!...
Indeed, why are some of those names different? The difference I found
was that sometimes an [0-9]%3a is inserted:
$ apt-get -qq --print-uris dist-upgrade|sed "s=.*/==;s/'//;s/.%3a//"|\
awk '$1!=$2{print $1,$2}' #eliminated the differences with
But why the need to be different? I'm following
/usr/share/doc/apt/offline.html/ch3.html .
--
http://jidanni.org/ Taiwan(04)25854780
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: make CD of only newest month of sid
One problem with the methods discussed is that we are only bringing home a CD with updates to installed packages, thus will miss out on the next kernel sub version, alsa etc. drivers for it, "emacs22" etc... >From studying man 5 apt_preferences etc., it seems to avoid the deadly libc6 2.3.1-9 in sid, one could do: # cat $PWD/preferences Package: libc6 Pin: version 2.3.1-8 Pin-Priority: 1001 But why: # apt-get -o Dir::Etc::Preferences=$PWD/preferences \ -s dist-upgrade|grep upgraded 296 packages upgraded, 21 newly installed, 87 to remove and 325 not upgraded. # apt-get -s dist-upgrade|grep upgraded 684 packages upgraded, 58 newly installed, 3 to remove and 4 not upgraded. when all the dependencies I've seen point to 2.3.1-1 only. -- http://jidanni.org/ Taiwan(04)25854780 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
I do not have Super Cow Powers
So what does Super Cow Powers mean? I see it with 'aptitude --help'. --- This is an automatic notification regarding your Bug report #175939: aptitude: "This aptitude does not have Super Cow Powers." ... has to go., which was filed against the aptitude package. It has been closed by one of the developers, namely Daniel Burrows <[EMAIL PROTECTED]>. Their explanation is attached below. If this explanation is unsatisfactory and you have not received a better one in a separate message then please contact the developer, by replying to this email. Dan Jacobson <[EMAIL PROTECTED]> was heard to say: > $ aptitude --help > This aptitude does not have Super Cow Powers. > > sorry fellas, this has to go. Only hip club members know what > you're talking about, the rest of us feel left out. *shrug* Then ignore it. Daniel -- http://jidanni.org/ Taiwan(04)25854780 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
top bug reporters (quantity, not quality) are who?
http://www.debian.org/Bugs/Reporting does not show how to find who are the top bug reporters for the past month, etc. statistics. Also top bug reported packages and top bugs themselves... how to know these? -- http://jidanni.org/ Taiwan(04)25854780 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
do I really need to be in all those /etc/groups?
Just look at me, $ id uid=1000(jidanni) gid=1000(jidanni) groups=1000(jidanni),20(dialout),24(cdrom),29(audio),1004(scanner) My latest addgroup was disk, so I wouldn't get error messages when eject(1)ing USBs. However $ find /dev |wc -l 5142 $ find /dev -group disk -perm -20|wc -l 4006 that gives me write permission to most of /dev. By the way, those error messages were eject: unable to open `/dev/sda1' #if my id(1) is not in the group "disk", or eject: unable to eject, last error: Invalid argument #if it is. Either way, it still does its job. One has to be root to not get the annoying messages. System is debian 2.4.19-k7. Another item is I can switch groups with ease, $ newgrp disk $ newgrp dialout $ newgrp jidanni Password: ** Sorry. Except for my own group, which fails when I give my login passwd... Maybe I didn't read the manual. But more exciting is when run in a emacs *shell* window, $ newgrp disk Segmentation fault $ newgrp dialout Segmentation fault $ newgrp jidanni Password: ** Sorry. $ newgrp audio Segmentation fault $ reportbug -f newgrp P.S. even after doing # deluser user group #(PPS: can use addgroup this way but not delgroup!) the processes still have those privileges until they die. But I guess that is how the system is designed. -- http://jidanni.org/ Taiwan(04)25854780 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
can exim try other route if first is rejected or down?
Dear Sirs [great spam phrase, actually. Thought I was really renaissance, but just added to my SpamAssassin score by using it, probably.] Anyway, Dear Sirs, I have an officially approved "great question", i.e. I wouldn't send a normal question to the list, only great ones > "M" == Mark Symonds <[EMAIL PROTECTED]> writes: M> hrm great question actually. I don't know offhand - M> did you figure it out? The setting: young Dan at home connecting with his modem twice a day to send his the email he has just written. >> just curious, anyway in exim.conf to make it try the third one if the >> second is down? >> >> route_list="\ >> *.tw smtp.apol.com.tw bydns_a;\ >> * tx.symonds.net bydns_a;\ #for non TW, and symonds is down: >> * smtp.apol.com.tw bydns_a;\ >> " >> #last item makes sense? no. I suppose this is not the place to do that kind of thing... not the right section of exim.conf ... My dream in fact was 1. try direct connection to the final destination, if get failure: 2. try thru my ISP, if get failure: 3. try thru my pal's USA machine which never gets rejected as spam suspect, and, if say there's a windstorm and it's down, try a backup... However all these test apparently are not a smart way to dispatch 20 mails in a 10 minutes modem call... Anyway, I was hoping exim could _for each message_ adapt to the failures and successes of what happens as it goes down the list. Maybe exim can't try another method of delivering the same message when it say encounters "looks like spam, service not available". Of course I read the manual only once. -- http://jidanni.org/ Taiwan(04)25854780 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
dist-upgrade better than "freshen my sid CDs"
Summary: "freshen my sid CDs" probably worse than "get fresh CDs" or
"dist-upgrade via apt-zip".
Recently I posted on
http://lists.debian.org/deity/2003/deity-200301/author.html
about ideas for modem users on how to freshen their sid CD set:
1. not like apt-zip, which is gets fresh versions of what you already
have installed. Therefore one never gets new kernel releases, etc.
2. not like jigdo, which is for making a whole new set of cds,
therefore one ends up making lots of CDs even if only a few programs
changed, I recall.
Instead my fantabulous idea was: have my friend burn a CD with all the
.debs in sid minus the ones I already have on my 20021010 CDs.
Well, since my sid cds are three months old, it turns out that more
than 1/2 of the .debs are older than current. Hmmm, therefore say I
wait a few more months, probably most of the programs will have new
versions. So maybe I should just 'a new sid cd set once a year, and
in the meantime, apt-zip and just plain apt-get install for small
stuff that one wants freshened badly. wwwoffle and
APT::Get::List-Cleanup false; are helpful for offline users.
Anyway here's how I computed it:
$ make -s bytes_needed
Unchaged on CDs 2612964310, changed 3496978462
or if just apt-zip, dist-upgrade: 500961328
I.e. 3GB vs. 500MB
here's the Makefile to compute it:
PV=awk '/^Package:/ {p=$$2};\
/^Filename:/{f=$$2};\
/^Version:/ {v=$$2};\
/^Size:/{s=$$2};\
/^$$/ {sub(":","%3a",v);print f,p"_"v,s}'
LD=/var/lib/apt/lists
H=debian.linux.org.tw
remote_list:
$(PV) $(LD)/$H_debian_dists_sid*-i386_Packages|sort>$@
cd_list:
$(PV) $(LD)/Debian%20GNU_Linux%20SID*i386_Packages|sort>$@
bytes_needed: remote_list cd_list /tmp/fetch-script-wget-debian
comm -2 remote_list cd_list|awk '/^\t/{h+=$$3;next};{s+=$$3};\
END{printf "Unchaged on CDs %d, changed %d\n",h,s}'
awk '{s+=$$3};END{printf "or if just apt-zip, dist-upgrade: %d\n",s}' \
/tmp/fetch-script-wget-debian
/tmp/fetch-script-wget-debian:
/usr/sbin/apt-zip-list -s -m /tmp -a dist-upgrade
--
http://jidanni.org/ Taiwan(04)25854780
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
apt-get install libc6 but want non-latest version
It seems if I want to get any new things from sid over my modem, I will have to get a new libc6. This shouldn't be too big: # apt-get install libc6 The following extra packages will be installed: libc6-dev locales Need to get 9348kB of archives. After unpacking 279kB will be used. Do you want to continue? [Y/n] n Questions: I was told to get 2.3.1-8, not 2.3.1-9. Therefore, a simple apt-get install libc6 will not be enough. Will putting Package: libc6 Pin: version 2.3.1-8 Pin-Priority: 1001 in /etc/apt/preferences be? No, # apt-get install libc6 Package libc6 has no available version, but exists in the database. This typically means that the package was mentioned in a dependency and never uploaded, has been obsoleted or is not available with the contents of sources.list Hmm, what is the procedure if one wants less than fresh versions then? Let me guess: apt-get --print-uris install libc6, then edit the uris to the version one wants. Download them and then install them with dpkg -i? P.S. having libc6 2.3.1-8 while the rest of my system is sid of 2002.10.10 won't break anything, right? I must get new versions of programs in the security bulletin so must get the new libc6. P.S. Does this "way of life" mix of units bother you? $ apt-cache show libc6|grep Size Installed-Size: 12116 Size: 3190622 -- http://jidanni.org/ Taiwan(04)25854780 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
shutdown without waking up the monitor
Problem: the electric company often cuts power here either for 10 minutes or for the whole day. One can only guess what it will be. With both my monitor and computer on, my UPS can only last 5 minutes, but with the monitor off, the UPS can last 20 minutes. (No, I have the most basic UPS with no com port hook up.) I could hit the plastic button on my monitor to turn it off, but if power is restored, then it is a shame to have turned it on and off: wear and tear. Best would to be to put it into a standby mode. Then if 15 minutes have gone by, go into shutdown via script, without reviving the monitor via keyboard actions. However I am having trouble writing this. xset dpms force standby #works, but monitor gets turned on again when X #exits during shutdown! How to stop that? #Ok, try from plain tty1 or after "init 1", on /dev/console: setterm -powersave powerdown #no effect, and no error message echo -e "\033[9;0]\033[14;0]" >/dev/console #again, no effect sleep 900 #interrupt with ^C if I notice power is restored shutdown -h now #I have a Lemel 15+ monitor, OS: debian woody -- http://jidanni.org/ Taiwan(04)25854780 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
cdrecord/README.ATAPI boot camp
Just thought you might like to see how I append files to my CDs, a
multitrack backup CD. My hair is turning grey in no coincidence to
the trial and error involved. I step thru it and it actually works.
You know what's neat? How the windows users just pop in the disks
like on the Nero screen, whilst us Linux Debian users get
/usr/share/doc/cdrecord/README.ATAPI.gz boot camp, if indeed we still
have the 'cells to comprehend it.
#Makefile:
sec=/home/jidanni/tmp/first_dir/second_dir
a19copy: #must be root
-rm $(sec)/*
find /var/tmp/*bkp*.bz2 -mtime -1 -print -exec cp -p {} $(sec) \;
#cp a cpio backup into the empty 'second dir', then
insmod: #2002.7 must have disk in recorder
insmod ide-scsi #debian, else bananas. in lilo.conf I appended stuff too
debdev=0,0,0
a19: #must have disk in recorder, silly me
-mount /cdrom
set -x;Z=`cdrecord -msinfo dev=$(debdev)` && \
mkisofs -v -o ../tmp/isoimage_2.raw -r -C $$Z -M /dev/cdrom ../tmp/first_dir
a19d:
-mount /cdrom #I forgot if needed
#needed fs=4m otherwise it can't read it out of the default file
cdrecord fs=4m -dummy -v -multi ../tmp/isoimage_2.raw
##did you make room infront of the computer as the disk will get
#ejected (a male thing, else no guarantee of success) after this:
#also don't do any other stuff at the same time to slow this down!
a19w:
#Must unmount first otherwise it will think it is read only no matter
#what you do in fstab or -o rw etc. Also you will think you burned ok
#but it didn't burn at all
-umount /cdrom
#needed fs=4m otherwise it can't read it out of the default file
cdrecord fs=4m -v -multi -eject ../tmp/isoimage_2.raw
#ok, see if it worked
close_drawer:
eject -t /dev/scd0 #if already closed also ok
#deb /usr/share/doc/cdrecord/README.ATAPI.gz
a19check: close_drawer
-mount /cdrom
ls -Ul /cdrom/* $(sec)
find /cdrom/ $(sec) -mtime -1 -type f|xargs sum|sort -n
#hope all is well :-)
--
http://jidanni.org/ Taiwan(04)25854780
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: shutdown without waking up the monitor
Dan Jacobson <[EMAIL PROTECTED]> wrote: > Problem: the electric company often cuts power here either for 10 > minutes or for the whole day. One can only guess what it will be. > With both my monitor and computer on, my UPS can only last 5 minutes, > but with the monitor off, the UPS can last 20 minutes. (No, I have > the most basic UPS with no com port hook up.) > > I could hit the plastic button on my monitor to turn it off, but if > power is restored, then it is a shame to have turned it on and off: > wear and tear. Best would to be to put it into a standby mode. Then > if 15 minutes have gone by, go into shutdown via script, without > reviving the monitor via keyboard actions. However I am having > trouble writing this. csj> Same power problemo here. I found my solution a few months ago. I csj> installed emacs + emacspeak + flite + eflite (needs a few bits of csj> non-automatic configuring). Emacspeak lets you compute without a csj> monitor! In my case it lets me do basic computer administration. When a csj> power outage hits I set my terminal (window or tab) to an emacs(peak) csj> shell and promptly turn off the monitor. If the power isn't restored csj> within say 10 mins, I can su to root and issue the shutdown csj> command. Yes, but you still physically (push the button) turn off the monitor. You have not managed to shutdown without somehow waking up the monitor, so you must physically turn it off. I don't need to touch the keyboard again. My script will do the shutdown in 10 minutes. I just can't figure out to run the shutdown command without waking up the monitor. -- http://jidanni.org/ Taiwan(04)25854780 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
are we going to have a Y2.1K problem like Y2K?
$ touch -t 2000 a $ touch -t 2011 a $ touch -t 2111 a touch: invalid date format `2111' Oh great, can't deal with dates in the next century. Hope this isn't a deep routed problem for all of Unix or something. -- http://jidanni.org/ Taiwan(04)25854780 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
how long does stuff in /var/tmp stick around?
I know that stuff in /tmp sticks around until the next of my daily power-ups, but what about /var/tmp for the default woody configuration? Seems like months, but I'd like to know if there's some cleaner program that is going to come along every 1/2 a year etc. when I least expect it. -- http://jidanni.org/ Taiwan(04)25854780 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
how to unwrap M$ self extracting .exe?
Just curious, here we have a MS executable http://members.fortunecity.com/d140738/M3-E.exe that I remember contains a excel file that I can deal with with gnumeric, etc. But now that I have removed M$ from my system, how to deal with the MS self extracting exe, as I believe it is. -- http://jidanni.org/ Taiwan(04)25854780 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
memory not pro-actively marked free?
Instead of doing shutdown, I did init 1, doing away with X windows, etc. and in maintenance mode, I did top(1). However it reported total memory usage still as high as ever. How might that be? P.S. is that the usual way to get rid of X windows if in case one wants just to use the humble console? I know there is a startx program, but no stopx. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
CTRL+ALT+Backspace will kill the X-server
>> is that the usual way to get rid of X windows if in case one >> wants just to use the humble console? I know there is a startx >> program, but no stopx. Preston> "CTRL+ALT+Backspace" will kill the X-server. Indeed it does, with no questions asked. But then it just springs back to life again, with a login prompt (at least with xdm here). Let's pretend my monitor is being borrowed for a few days and being replaced by a VT100, and I want to properly end all X processes and revert to plain tty mode, without rebooting or editing any files. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
communication structures crumbled
To us debian users, the most notable thing during this break in or whatever episode, is how the communication structures crumbled. debian-announce had one message on the 21st, five days ago, saying for more information, see www.debian.org. Nothing special there, so I checked http://www.debian.org/security/, same problem. With the mailing lists affected, what would average user me do to learn the latest on the situation, google around? Googleing around just lead me to some stale discussion on the mailing lists before they got turned off. At least some latest news could have been posted to the main website. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
no way to remove something if the control file is damaged
M> I'm trying to get rid of a package here M> or at least put it on hold, or just get apt M> to shuttup about it already, and I'm at M> a loss. Any ideas? M> # dpkg --purge --force-all sysadmin-guide M> dpkg - warning, overriding problem because --force enabled: M> Package is in a very bad inconsistent state - you should M> reinstall it before attempting a removal. M> (Reading database ... 57729 files and directories currently installed.) M> Removing sysadmin-guide ... M> error: empty control file at /usr/sbin/install-docs line 600. M> dpkg: error processing sysadmin-guide (--purge): M> subprocess pre-removal script returned error exit status 25 odd, no way to remove something if the control file is damaged. Perhaps try wget .../sysadmin-guide_0.7-2_all.deb and variations of dpkg -i sysadmin-guide_0.7-2_all.deb which other readers of this list might mention... -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Bug#223372: apt-zip: dynamic computation of what files we need
OK, the idea is we depart our faraway Antarctic (or mere hilltop, or
Space Station) debian machine with as little information possible that
describes its state, in terms of which of sid/sarge/woody, and which
versions of packages we have installed:
COLUMNS=222 dpkg -l|awk 'NR==1,/===/{next};{print $1,$2,$3}
Also we would have a list of packages we want to add, but didn't have
the bandwidth do download.
When we arrive at the well connected site, oh weeks later, we furnish
the above information and create a, oh, CDROM or two, of the same
files we would get if we had done a 'apt-get dist-upgrade' back in
Antarctica, however with current Packages files instead.
This gives us weeks fresher packages than if we had computed what we
needed before we left the faraway site.
We return to the faraway site and install the packages.
When we were at the well connected site, we told apt: given these
packages installed, and these available, etc. etc., please compute
what I need for a dist-upgrade etc.
Apt programs like to have a "status" file. This we can construct from
the minimal
COLUMNS=222 dpkg -l|awk 'NR==1,/===/{next};{print $1,$2,$3}
that we left home with.
But what happens when we arrive at the well connected site only to
find that they all use some other leading brand operating system... no
apt anywhere to compute what we need.
No problem, we just slide our pitiful floppy with the dpkg -l,
sid/sarge/woody choice, and additional wanted packages list, into
their computer and tell the apt-computations website to compute what
we need for download, (and ask about problem resolution too.)
Of course one might ask, why not take the whole computer or the hard
disk to the well connected site... well, as we take a kayak thru a bad
neighborhood to get there we could likely loose all along the way,
plus mom needs to use it while we are gone.
>>>>> "G" == Giacomo A Catenazzi <[EMAIL PROTECTED]> writes:
G> Dan Jacobson wrote:
>> Package: apt-zip
>> Version: 0.13.2
>> Severity: wishlist
>> Let's say I leave home with fetch-script-wget-jidanni.org for a two
>> week trip to a place with fast connections, I'll return home with
>> files no newer than two weeks.
>> If however, I left home with an extract of the dpkg-status file, and
>> some other vital snippets, I could be able to get the freshest
>> versions available for my system... but how... probably thru various
>> apt config stuff...
G> Hello.
G> I used a script that download the Packages file and from this it
G> extracts the newer version of package to download. So I could
G> update unstable with few problems. It is no perfect, but it can
G> help. Maybe I can implement an other script to put the complete
G> package list on zip and than remotely to see what new packages are
G> needed. This methos would have nearly no more dependence
G> problem. (but in case of packages split, packages that depends on
G> new created packages. To solve this it would to complicates, but if
G> we use the next methods)
G> apt-get documentation give some hint about use a remote debian machine to
G> calculate the new list of packages of download.
G> These methods are more complex as the actual methods, and the
G> remote machine (fast connected machine) maybe doesn't have the
G> needed tools (shell scripts are not so portable in the various unix
G> because of "creative" implementation of some features). I'm
G> working on doing some scripts so that user can use powerfull script
G> on GNU/Linux machines and simple scripts on some old unix machines
G> (and maybe also a .bat for dos/windoze).
G> ciao
G> giacomo
G> PS: sorry about the big delay of fixing apt-zip bugs :-( . Really I
G> should/want to correct shortly the apt-zip, and in a second step
G> upload the newfeatured apt-zip
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
HTML debloater for reading on PDA
Any package for converting the average bloated webpage into slim and trim HTML for a turn of the century browser on a weakling black and white PDA? "tidy" even with all options turned on isn't aggressive enough. Maybe I will just end up doing lynx -dump|txt2html. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Debian's at(1) is so bad, what does GNU use?
Constructive me says: Debian's at(1) is so bad, one can't even do $ at next year parse error. Last token seen: year Garbled time $ at now + 12 months #Sigh. Must use instead. No wonder. Not genuine GNU parts. Say, Stallman to the rescue. What do those cool GNU people use? Don't tell me they don't use at(1)? -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Debian's at(1) is so bad, what does GNU use?
A> nohup sh -c "(sleep 100h; move-my-toes)" & Yeah, never rebooted where you live. Out with it, what at(1) command do they use at GNU? -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
encountered new file extension (.dwg)
> "Jack" == Jack Varga <[EMAIL PROTECTED]> writes: Jack> Now, a question. Does anyone know of efforts to develop a DWG Jack> library that is compatible with GPL, LGPL or a similar license? Jack> The Open Design Group's OpenDWG library (a misnomer in its own Jack> right), is seriously problematic in regards to licensing. Say, I just encountered a .dwg file and don't know which Debian tool to use to do anything with it. All I know how to do on Debian when encountering a new file extension is apt-cache search dwg, aptitude search .dwg, grep-available -i dwg. OK, I suppose I should search Google (proprietary solution) for "linux dwg". Hmmm, so probably Debian packages should be sure to list uncommon file extensions that they can deal with in the Package Description. Wait, according to what Jack says, there won't be any Debian package to deal with .dwg's. Oh. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Outlook to GNU/Linux mailbox conversion is possible with existing tools
Hello Fellas, unfortunately when we read and reply to this list via
netnews, the mailing list folks won't see it [smart]. So here's a
repost via mail. My real address is jidanni @t kimo.com.tw
Message-ID: <[EMAIL PROTECTED]>
Subject: Outlook to GNU/Linux mailbox conversion is possible with existing tools
Date: 02 Mar 2001 00:01:03 +0800
How I got my mail out of Outlook 2000 and Outlook Express 5 into GNU/Linux
without special tools By Dan Jacobson 3/2001
---
Below is me commenting on my earlier [D>] discussion. Below I've left
the steps a little loose... maybe somebody will polish this up. I
didn't try the IMAP thingy. I did try lots of windows programs
mentioned on webpages, that didn't work...
---
They say one needs fancy tools to extract ones Outlook 5 mail files
into say a Unix mbox format or whatever. However, I have found a
obscure passageway out.
D> Preliminary results on getting one's sent netnews messages [and
D> perhaps any other stuff in a .dbx file] out of the clutches of
D> Outlook and Outlook Express 5:
D> Goal: stop using M$ and get on GNU/Linux. Problem: I've got
D> several hundred of my brilliant netnews posts that I must take with
D> me if I want to keep my record of never having lost a byte of my
D> precious golden keystrokes intact. Email too, sent & received.
Can you believe were about to move our stuff from Outlook into Outlook
Express then into GNU/Linux.
Indeed, seemed like they sealed up plain Outlook2000 pretty good, but
wait, in Outlook Express I noticed I could "import" my plain Outlook
folders! Ah ha! Be sure to rename any Outlook Express folder of the
same name before you do this, else plain Outlook messages will be
copied into it alongside your original messages. Yuck.
Import steps were: [using the letter symbols, because I don't know
what the original English names are because I use the Chinese
version.] File (F)>I>E, select 'Microsoft outlook' (S), Messages,
select the plain Outlook folder to integrate or all, note that you will get
you messages intermingled with Outlook Express messages unless you first
renamed you Outlook Express folders. Yuck.
[brief interlude of me banging around inside the maze that they thought
they had sealed up:]
D> you control-click them [to select more than one, better yet,
D> there's a select all option on a menu] and then you got to the file
D> menu choice and discover that the 'save to file' choice is no
D> longer highlighted, ha ha they only allow you to save one file, no
D> problem: give up and instead now go to 'search' thru that folder [I
D> don't know the exact English menu names because I'm using the
D> Chinese version, plus too lazy to check], you search match all your
D> items, and again the save to file option's lights are out as you
D> have chosen more than one file, but not the forward option !!! ha
D> ha the chumps at M$ forgot to seal that off [Win98SE]! Conspiracy
D> failed! OK, so you 'send' all your news items in that folder as a
D> multi file attachment and save this unsent mail to disk.
D> Indeed, it has a long attachment with your files uuencoded! we are
D> almost home!
To reiterate: Still in Outlook Express, select the folder of interest,
and select find: Edit (E)>F>M, now search for everything, e.g., by
checking all items 'earlier' than 'now', now all items are added to
the found set, then select them all: Edit (E)>A.
Now pick Mail(M)>A to 'send them all as attachments' [devious or
what?!], now choose File (F)>A to save this letter with its many
attachments as a file. [Don't pick "save the attachments" (V) as that
will create a mess, especially with same Subject lines...]
Save this file as whatever.eml . [Inside this file one can see that the
messages are also uuencoded .eml files.]
D> OK, now boot up GNU/Linux and go to that file and run my brilliant
D> shell script below. when I did it my file with the long attachment
D> was called *.eml . "col" I think cured the ^M's at end of
D> lines. [try dos2unix next time]
: #middle part of Outlook 5 -> Outlook Express 5 -> gnus travail
set -ex
#my personal setup for this run
source=mysentmail.eml #file forced out of OE 5 by extreme methods
target=outlook_sent_mail #I plan to call a new gnus nnml folder
#or for another run of mine:
source=mymailbox.eml
target=outlook_inbox
maildir=~/Mail/mail/$target #gnus's'
grep -c begin\ 666 $source #informational and quit if error [0 found]] [set -e]
above]
mkdir $maildir #fail if already there... we would be overwriting!
cat $source|col| (cd $maildir
awk '
BEGIN {xxx="uudecode"}
#uud
debian mailing lists now available via NNTP at news2.sinica.edu.tw
It is with extra large pleasure that [EMAIL PROTECTED] has given me permission to announce the uni-directional read-only, no posting allowed availability of lots and lots of debian, free-bsd etc. mailing lists, via NNTP server news2.sinica.edu.tw . Any mail regarding this should be sent to him and not me. The debian groups are called os.debian.* -- http://www.geocities.com/jidanni Tel886-4-25854780
floppy in kernel 2.6: is not a block device
How does one use one's floppy under kernel 2.6? $ mount /floppy mount: /dev/fd0 is not a valid block device I tired /usr/share/doc/fdutils/Fdutils.html#SEC7 's idea: for i in /dev/fd0*; do dd if=$i of=/tmp/foo count=1 && break; done I tried modprobe floppy ide-floppy. I didn't change fstab from 2.4 days. They must have changed the kernel. Same lilo, append="pci=biosirq" or not. dmesg shows: inserting floppy driver for 2.6.7-1-k7 FDC 0 is a post-1991 82077 Yes unit is plugged in. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: floppy in kernel 2.6: is not a block device
Can't use my floppy drive in Linux 2.6. G> As root in /dev type: MAKEDEV floppy ./MAKEDEV: don't know how to make device "floppy" G> Another piece, is "udev" installed? OK, I installed it and rebooted. G> If it is, it should create those auto-magically. No it didn't. ls /dev/f* shows none. Now trying ./MAKEDEV fd0 shows lots of supposed action, but ls doesn't show any changes. On a second run there is no output, so supposedly its changes are registered somewhere. Maybe I have to reboot to see them too... G> Or you can manually use "mknod" to create them. G> Here is my listing for /dev/fd0: G> brw-rw 1 root floppy 2, 0 2003-11-12 09:24 /dev/fd0 G> That would be: G> mknod --mode=0660 /dev/fd0 b 2 0 Done. However still, mount: /dev/fd0 is not a valid block device OK, that's better than the earlier mount: special device /dev/fd0 does not exist no help from modprobe ide-floppy and floppy either. I was able to use the floppy in Linux 2.4. Not in 2.6. In fact, I now notice using reboot(8) no longer reboots, just hangs ttys. pstree shows it is stuck doing "shutdown -r 0 w". I will remove udev forthwith. A second try hung my last tty and I had to walk over to press a hard reset. Geez. I notice purging udev leaves files behind. Their name flew off my screen and now I won't ever be able to find them without repeating an install/purge. Ok, now rebooting again, luckily with no filesystem damage after the hard reboot, I see all the /dev/fd* stuff is indeed there. However, as much as I try, /dev/fd0 is not a valid block device. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
mystery file spotted in /
Say if we spot a file, # ls -l /root-n -rwxr-xr-x 1 root root 0 2004-03-24 00:31 /root-n and we can't tell what package(s) made it, # dpkg -S /root-n dpkg: /root-n not found. nor is it mentioned in debian-policy/fhs, how do we know it is safe to remove it, or must we just let it sit there for eternity out of fear, even though all along it might have just been the product of our fumble fingers, or maybe it indeed is the linchpin that holds the system together. Why is there no Department of Homeland Security, whom we can ask if this file really belongs on our system, and not is some wetback, enjoying the good times right on "/" itself, immune from meaningful stat(1) checks, as "I am a 0 bytes file, access times don't necessarily reflect my importance, you still dare to delete me?" Anyway, I don't even know if the available security packages will investigate randomly named files in this free for all file system of ours where any traces of what did what are only in each .deb itself. Can I delete /root-n? -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
ide: Assuming 33MHz
Upon boot, one sees "ide: Assuming 33MHz system bus speed for PIO modes; override with idebus=xx". But upon reading e.g., http://storagereview.com/guide2000/ref/hdd/if/ide/modesUDMA.html one feels their computer sounds more like the 100Mhz kind. How can one tell if one is taking full advantage of the hardware? Maybe I am, as I see "IDE UDMA100 controller" in dmesg etc. output (that I put in http://jidanni.org/comp/system.zip) -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: mobile disk racks
I slide in a mobile disk. I then do $ mount /mnt/backup/ mount: /dev/hda5 is not a valid block device Does this mean that mobile IDE disks need a full reboot to be recognized? -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
any last words before I install kernel 2.6?
Any last words before I # apt-get install kernel-image-2.6-k7 on my home (sid) PC? Will things break that used to work in 2.4? -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
it was supposed to be my first DVD
It's my first DVD, and # mplayer dvd://1 libdvdread: Using libdvdcss version 1.2.5 for DVD access libdvdread: Could not open /dev/dvd with libdvdcss. libdvdread: Can't open /dev/dvd for reading Couldn't open DVD device: /dev/dvd Does this mean the physical device, # cat /proc/ide/ide1/hdd/m* cdrom DVD-ROM DDU1621 is bad? or is my lilo.conf append="pci=biosirq hdc=ide-scsi hdd=ide-scsi max_scsi_luns=1" bad? and why above does it say cdrom? I assume it is these issues and we haven't even got to the encryption business. Yes I linked hdd->dvd. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
env|grep HZ=100
Where is the mysterious HZ=100 documented that I see in the environment of some accounts after login, and always after doing su? # env|grep HZ # su - nobody No directory, logging in with HOME=/ [EMAIL PROTECTED]:/$ env|grep HZ HZ=100 -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
many marillat users don't realize their Packages.gz is old
Dan> Because of the broken Last-Modified, I am unable to get fresh Dan> copies of Dan> http://marillat.free.fr/dists/unstable/main/binary-i386/Packages.gz Dan> from my ISP's cache!: Cristian> Please read my home page. http://marillat.free.fr: Since yesterday all packages uploaded got the nice 17-Apr-1971 date, thus apt-get is unable to see these files as news. The only solution is to do (as root or with sudo) a "rm /var/lib/apt/lists/marillat*" and do another "apt-get update". However, this still will not solve the problem with many ISP caches! Thus I suppose many marillat.free.fr users do not realize that they are not getting new Packages.gz files, rm or not! And, there is nothing they can do about it. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
no response from the most generic 3-button mouses
In XF86Config-4, these work fine for my Trackman Marble+: Option "Device" "/dev/ttyS1" Option "Protocol" "IntelliMouse" But when I instead attach the most common simplest 3 button mouses, I get no response from the mouses at all. mdetect doesn't detect them, moving around the mouse or not, yes with xdm stopped, using just the console. mdetect detects the Trackman Marble+ fine. No, I don't have gpm or imwheel installed. Saying # Option "Protocol" "MouseSystems" # Option "Protocol" "Microsoft doesn't help. I don't think these matter Option "Emulate3Buttons" "true" Option "ZAxisMapping" "4 5" when trying to get the least response from the mouse, so I didn't remove them. For instance, I'm talking about mouses like this: http://home.att.net/~nalutz-lgi/List_of_Product_Table_Contents/index/Product__5/Roltek_PS2_3-Button-Mouse__New/roltek_ps2_3-button-mouse__new.html Krone Whale Mouse, and Mouse GM 203P. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
