Re: How do I set different JAVA_HOME for different users?
From: Thomas Schmitt To: debian-user@lists.debian.org Sent: Wednesday, 6 April 2016, 16:50 Subject: Re: How do I set different JAVA_HOME for different users? Hi, Jaimz Fairfax wrote: > The internet seemed to be convinced that it had to be done via systemd > user.conf. That opinion does not have to be wrong. It depends on the level of software which you want to influence. Environment variables are a very fundamental feature. Programs can see them even if they have not been started by a shell. But of course, the settings in the shell cannot influence what those programs see. So it is sometimes necessary to use the init software or the desktop to define the variable names and values for programs. On the other side ot the spectrum, you may write wrapper programs around the programs which you want to influence. Those wrappers would set up the desired environment variables before starting the desired program. In the most simple case this would be a shell script like #!/bin/sh export JAVA_HOME="/usr/lib/jvm/java-8-openjdk-amd64" android-emulator ...options... So you can set up different environments per program rather than per user. Have a nice day :) Thomas Thomas,A shell script exactly like that was what what I started out with!It worked ONLY for that terminal. I have seen a lot references to wrapper scripts but no examples for me to follow.I never thought to try 'man bash'.I did 'man environment' which returned nothing...The trouble is, often you need to know where to look, and if you know where to look you probably know the answer anyway. Jaimz
Re: How do I set different JAVA_HOME for different users? SOLVED
From: Thomas Schmitt To: debian-user@lists.debian.org Sent: Wednesday, 6 April 2016, 9:59 Subject: Re: How do I set different JAVA_HOME for different users? I have most of my personal preferences in ~/.bashrc .Stuff like: export LC_COLLATE="C" export HISTCONTROL="ignoredups" export PS1='\u@\h:\W> ' So i guess it would be a good place to set export JAVA_HOME="/usr/lib/jvm/java-8-openjdk-amd64" Have a nice day :) Thomas Thanks for that! Adding export JAVA_HOME="/usr/lib/jvm/java-8-openjdk-amd64"to the end of ~/.bashrc fixed it.The internet seemed to be convinced that it had to be done via systemd user.conf.
How do I set different JAVA_HOME for different users?
I am running debian unstable, systemd version 229.I am learning Android development. Android Studio uses JAVA_HOME = /usr/java/jdk1.8.0_66.Some of the code I want to investigate uses Apache maven and that wants JAVA_HOME = /usr/lib/jvm/java-8-openjdk-amd64 I setup user2 so that when logging in JAVA_HOME is automatically set for maven.Except nothing I try will make it work.export JAVA_HOME="/usr/lib/jvm/java-8-openjdk-amd64" works in a terminal, but only for that terminal. The internet thinks that /home/user2/.config/systemd/user.conf withDefaultEnvironment=JAVA_HOME=/usr/lib/jvm/java-8-openjdk-amd64will work. It doesn't. Thanks, Jaimz
Link Transfer Fail message on remote printing
Since upgrading to Potato, I cannot print from remote boxes on my network, I get a message "Link Transfer Fail." I also note that in syslog, I see tcplog: printer request from brubaker.martinet ippl: ICMP message type destination unreachable - bad port from localhost [127.0.0.1] Also, auth.log shows PAM_unix[16418]: authentication failure; (uid=0) -> scm\artin for samba service M Does anyone know what these messages mean? Thanks, Steve Martin [EMAIL PROTECTED]
remote printing broken due to PAM?
After I installed Potato on my machine, I noticed that remote printing from the Windows boxes on my network was failing. Looking into the problem, I noted entries like this in auth.log when I tried to print: PAM_unix[535]: authentication failure; (uid=0) -> scmartin for samba service I have never used PAM before, but it was necessary to install it this time, several other packages are dependent on it. All I have loaded is libpam_modules, libpam_runtime and libpam0g. Do I need to install libpam_smb, or is this a password authentication problem? (That is, do I need to do something to the password database to correct this?) Thanks, Steve Martin [EMAIL PROTECTED]
Re: ppp: frame with bad fcs
"...something munging bytes on the way through," is what I assume. I can load the same pages on my Win95 laptop with no problem with the same ISP, so it appears to be Linux somehow that is doing this. Last night I recompiled the kernel (a potato system, I had been using the default 2.2.14 kernel), and this seems to have relieved the problem somewhat -- stuff downloads very slowly from the problem sites, but I don't get the timeouts that I was getting. No error messages are now evident. Maybe I need to examine the packets that are getting sent over to me? > fcs is Frame Check Sequence. PPP over an async link uses HDLC-like framing > for each packet. The FCS is used to detect errors in packets which are > received. Such an > error could happen on a link which is not error-corrected. On a link which is > error-corrected this probably indicates something is munging bytes on the way > through. > This could happen because of things like software flow-control, which inserts > XON/XOFF bytes into the stream to tell the other side to stop sending data. > Make sure your > modem is configured to only do hardware flow control. > > [EMAIL PROTECTED] wrote: > > > I am having trouble recieving files via PPP from certain sites. I checked > > the debugging log, and I am repeatedly getting a two-line error message > > from those sites: > > > > kernel: ppp: frame with bad fcs, length = 994 > > kernel: ppp: bad frame, count = 994 > > > > Can anyone tell me what this means? > > > > It seems that my PPP connection receives a bad frame, and keeps requesting > > that same frame over and over again (I get the same message repeated > > throughout the log). > > > > Thanks, > > > > Steve Martin > > -- > Jens B. Jorgensen > [EMAIL PROTECTED] > >
ppp: frame with bad fcs
I am having trouble recieving files via PPP from certain sites. I checked the debugging log, and I am repeatedly getting a two-line error message from those sites: kernel: ppp: frame with bad fcs, length = 994 kernel: ppp: bad frame, count = 994 Can anyone tell me what this means? It seems that my PPP connection receives a bad frame, and keeps requesting that same frame over and over again (I get the same message repeated throughout the log). Thanks, Steve Martin
ppp: error in VJ decompression
I have a potato system, with a 2.2.14 kernel. I keep getting a message in syslog saying ppp: error in VJ decompression. As far as I can tell, VJ decompression is handled by a utility in the PPP package. Is there any way I can debug this? Are there any bugs in PPP that might cause this? Something is slowing my ppp sessions down to a crawl, I'm wondering if this might be the cause? When I log on with my Windows 95 laptop, I have no problems downloading from the same sites. So I believe it is something in the Potato setup of PPP, but I haven't been able to determine what. Steve Martin [EMAIL PROTECTED]
Unable to load lp module
I just wiped out my system and reloaded with Potato and a 2.2.14 kernel. When I tried configuring modules, I got this error on the lp module: /lib/modules/2.2.14/misc/lp.o: invalid parameter parm_io /lib/modules/2.2.14/misc/lp.o: insmod /lib/modules/2.2.14/misc/lp.o failed /lib/modules/2.2.14/misc/lp.o: insmod lp failed Installation failed. When I try insmod lp.o, I get this: lp.o: unresolved symbol parport_enumerate_Rd3cb4efd lp.o: unresolved symbol parport_claim_or_block_R21556569 lp.o: unresolved symbol parport_release_R139844a3 lp.o: unresolved symbol parport_unregister_device_R60d23b61 lp.o: unresolved symbol parport_register_device_Rb152ec85 I have none of the parport modules loaded. lp was working fine before I did the switch. What is causing this error? Steve Martin [EMAIL PROTECTED]
Unable to load lp module
I just wiped out my system and reloaded with Potato. When I tried configuring modules, I got this error on the lp module: /lib/modules/2.2.14/misc/lp.o: invalid parameter parm_io /lib/modules/2.2.14/misc/lp.o: insmod /lib/modules/2.2.14/misc/lp.o failed /lib/modules/2.2.14/misc/lp.o: insmod lp failed Installation failed. lp was working fine before I did the switch. What is causing this error? Steve Martin [EMAIL PROTECTED]
Unable to load lp module
I just wiped out my system and reloaded with Potato. When I tried configuring modules, I got this error on the lp module: /lib/modules/2.2.14/misc/lp.o: invalid parameter parm_io /lib/modules/2.2.14/misc/lp.o: insmod /lib/modules/2.2.14/misc/lp.o failed /lib/modules/2.2.14/misc/lp.o: insmod lp failed Installation failed. lp was working fine before I did the switch. What is causing this error? Steve Martin [EMAIL PROTECTED]
Re: Modem dials without reason
> I configured demand dialing with diald on my machine. Now my modem dials > every>five minutes without reason. I increased logging but could not find a > hint why > I apologize if there is a duplicate message here, I hit the wrong key just a few minutes ago, and seem to have deleted the message I was working on. I am having the same problem as Klaus, but the hints in this thread have not helped me. The only port activity I see in the server logs which I cannot explain is on port 1024 and on ports higher than 61000. In the period during which the modem dials out every few minutes, there are several of these requests on these ports. One more note about my case, I have diald installed, but demand dialing doesn't work unless I type the command "ipmasq" at the server prompt. If I do not, there is no dialout, the workstation browser times out, and the only thing that shows in the logs is named[221]: ns_forw: sendto([204.72.134.10].53): Operation not permitted Does anyone have any ideas on this one? Steve Martin [EMAIL PROTECTED]
printing from Windows to Linux printer
I am trying to print from Windows to a printer connected to a Linux machine. What I have discovered, is that unless the spool directory has permissions 777, Windows cannot print to the Linux machine. However, if I set the directory permissions to 777, it reverts back to 700 after I reboot. (Can I presume LPRNG is doing this?) I don't necessarily want to change the permissions on the spool directory. What do I need to do to get this to work? Steve Martin [EMAIL PROTECTED]
printing from Windows to Linux printer
I am trying to print from Windows to a printer connected to a Linux machine. What I have discovered, is that unless the spool directory has permissions 777, Windows cannot print to the Linux machine. However, if I set the directory permissions to 777, it reverts back to 700 after I reboot. (Can I presume LPRNG is doing this?) I don't necessarily want to change the permissions on the spool directory. What do I need to do to get this to work? Steve Martin [EMAIL PROTECTED]
IP Masq not running?
I have a 2.2.12 kernel with IP Chains, IP Masq, diald, etc. I have determined that the server never dials out, when one of the workstations tries to access the internet, unless I type in the command "ipmasq" at the server prompt. I do have the file ipmasq in /etc/init.d (and in rcS.d, etc. as S41ipmasq), I have included the first portion of the file below. I tried adding the line "/sbin/ipmasq" to this file, right below the line that starts out "test", but this had no effect. What do I need to do to make the dialout process automatic? IPMASQ- #!/bin/bash # # ipmasq.init Set up IP Masquerading for Debian systems # # v3.0 19 July 1998 test -x /sbin/ipmasq || exit 1 case $1 in start|restart|force-reload) ... - END --- Thanks, Steve Martin
Re: diald failing with no messages
> > But diald isn't working, and no log messages are being produced. > > If you type 'diald' at a prompt, does it segfault? If so, try making > lines in /etc/services that look like: > No segfault :( Steve Martin
Re: diald failing with no messages
> I have had similar problems with diald though not in the past few days. BTW > your e-mail to the list (just this one) keeps crashing my XFMail for some odd > reason. Very odd. > It's not that much different from any other mail I've sent, i.e., same editor, same mail program, etc., so I can't imagine why it's doing that. If you find a reason, let me know. Steve Martin
diald failing with no messages
I have server with kernel 2.2.12 and a mostly slink Debian system (I have upgraded to packages from potato as necessary). I have diald set up on the server, but it is not working. I am also running IP Masquerading with IP Chains, etc. I have everything configured correctly, as far as I can tell from the documentation, but nothing happens. If I use wvdial or pon/poff, dialout is successful, and my clients can then successfully access the internet. But diald isn't working, and no log messages are being produced. If I try to access the internet from either the server or the workstations, without first dialing out on the server, nothing happens. How do I begin to debug this? Steve Martin
ipchains and DNS problem
I have ipchains working, but my workstations cannot get out to the internet because of a DNS problem. (When I try to ping any host from the workstations I get the message Host not found.) I have kernel 2.2.12, running IP MASQ and ipchains. /etc/init.d/network on the server looks like this: -- ifconfig lo 127.0.0.1 route add -net 127.0.0.0 netmask 255.0.0.0 dev lo ifconfig eth0 192.168.1.1 netmask 255.255.255.0 broadcast 192.168.1.255 up route add -net 192.168.1.0 dev eth0 echo "1" > /proc/sys/net/ipv4/ip_forward ifchains -P forward DENY ipchains -A forward -s 192.168.1.0/24 -j MASQ -- on the workstation: -- ifconfig lo 127.0.0.1 route add -net 127.0.0.0 netmask 255.0.0.0 dev lo ifconfig eth0 192.168.1.10 netmask 255.255.255.0 broadcast 192.168.1.255 up route add -net 192.168.1.0 dev eth0 route add default gw 192.168.1.1 -- resolv.conf on both machines just has two lines, one for each of the nameservers my ISP uses. Issuing the command # ipmasq -d from the server gives the following output: -- /sbin/ipchains -P input DENY /sbin/ipchains -P output DENY /sbin/ipchains -P forward DENY /sbin/ipchains -F input /sbin/ipchains -F output /sbin/ipchains -F forward /sbin/ipchains -A input -j ACCEPT -i lo /sbin/ipchains -A input -j ACCEPT -i eth0 -s 192.168.1.1/255.255.255.0 /sbin/ipchains -A input -j ACCEPT -i ppp0 -d 206.11.2.183/32 /sbin/ipchains -A input -j DENY -i ppp0 -s 192.168.1.1/255.255.255.0 -l /sbin/ipchains -A forward -j MASQ -i ppp0 -s 192.168.1.1/255.255.255.0 /sbin/ipchains -A output -j ACCEPT -i lo /sbin/ipchains -A output -j ACCEPT -i eth0 -d 192.168.1.1/255.255.255.0 /sbin/ipchains -A output -j ACCEPT -i ppp0 -s 206.11.2.183/255.255.255.255 /sbin/ipchains -A output -j DENY -i ppp0 -d 192.168.1.1/255.255.255.0 -l -- I can ping anything from the server, but not from the workstations. What am I doing wrong? Steven C. Martin
fax/efax 6100 segmentation fault
I am trying to get the fax/efax program working. I used the command "fax test
> debug.out" suggested by the man page, and note that at the bottom of the
resulting file is a note: "/usr/bin/fax: line 1129: 6100 Segmentation fault"
(this seems to correspond to the penultimate line of the script, which is
"done").
What does this mean?
For reference, I am pasting the contents of debug.out below
Thanks,
Steve Martin
debug.out---
--- /usr/bin/fax ---
FAX=fax
EFAX=efax
EFIX=efix
DEV=ttyS1
CLASS=2
FROM="+1 800 555 "
NAME="Put Your Name Here"
PAGE=letter
PRTYPE=pcl # HP-PCL (e.g. HP LaserJet)
PRCMD="lpr"
VIEWCMD="xv -"
GS=gs
DIALPREFIX="T"
DIALSUFFIX=""
LOCK='-x /usr/spool/uucp/LCK..$DEV' # "normal" systems
CONFIGFILES="/etc/efax.rc ${HOME:-~}/.efaxrc ./.efaxrc"
TSTAMP='date +%m%d%H%M%S'
LOOKUP='eval sed -n -e "/^fax($1):/{" -e "s/^[^:]*://p" -eq -e"}" $f'
DIRFILES="${HOME:-.}/.faxdir"
TELCVT='sed -e s/+1/1/ -e s/+/011/' # North America
RES=204x196 # default "Fine" resolution (196 lpi)
PRINT='$EFIX -ve -p$PAGEDIM -r300 -s0.98 -d0,0.125 -o$PRTYPE $f'
VIEW='$EFIX -ve -p$PAGEDIM -r200 -opgm $f' # 50dpi: fast, whole-page view
INIT="-iZ -i&FE&D2S7=120 -i&C0"
RESET="-kZ"
SPKR="-iM1L0"
TXCAP="1,3,0,0,0,0,0,0"
RXCAP="$TXCAP"
RXINIT=""
TXINIT=""
DATECMD='date "+%y/%m/%d %H:%M"'# YY/MM/DD HH:MM (24hour)
HDR='$DATE $NAME ($FROM) --> $TO p. %d/%d'
BUSYRETRIES="30 60 120 300 60 600 60 60 1200 60 60"
FAILRETRIES="300 300" # try two more times at 5 minute intervals
NICE="nice -n -10"
RM="/bin/rm -f"
LS="/bin/ls"
VERB="ewin" # show errors, warnings, progress & negotiation
VERBLOG="chewmainrxtf" # log everything
CONSOLE=/dev/console
FAXDIR=/var/spool/fax
FAXLOGDIR=${FAXDIR}
ANSFNAME="%m%d%H%M%S"
UMASK=022
NOTIFY='mail -s "fax from $REMID: $FILES" $FAXMGR <$f'
FAXMGR=root
ANSRINGS=1
OWNER=root.tty # typical
MODE=666# anybody
FONE=/usr/bin/fone
NOLOG='[145]'
--- /etc/efax.rc ---
DEV=modem
FROM="+1 651 776 1364"
NAME="Steven C. Martin"
PRTYPE=ps
PRCMD=lpr
PAGE=`paperconf`
VIEWCMD="xloadimage stdin" # xloadimage is a free Debian package
LOCK='-x /var/lock/LCK..$DEV' # Debian has lockfiles in /var/lock
FAXLOGDIR=/var/log/efax
apt_0.3.11.deb CD error
I just downloaded apt 0.3.11, and it doesn't seem to recognize the CD properly. apt-cdrom works to put lines in sources.list, but when I try to install, using either dselect or apt-get, I get a series of error messages like the following: Err cdrom:Debian GNU_Linux slink (2.1) 2_4 main binary-i386 section 2, contrib SAM19990306/ debian/dists/frozen/contrib/binary-i386/ xswallow 1.0.11-3 Wrong CD The relevant portion of sources.list shows deb "cdrom:Debian GNU_Linux slink (2.1) 2_4 main binary-i386 section 2, contrib SAM19990306/" debian/dists/frozen/contrib/binary-i386/ deb "cdrom:Debian GNU_Linux slink (2.1) 2_4 main binary-i386 section 2, contrib SAM19990306/" debian/dists/frozen/main/binary-i386/ deb "cdrom:Debian GNU_Linux slink (2.1) 1_4 main binary-i386 section 1 SAM19990306/" debian/dists/frozen/contrib/binary-i386/ deb "cdrom:Debian GNU_Linux slink (2.1) 1_4 main binary-i386 section 1 SAM19990306/" debian/dists/frozen/main/binary-i386/ What am I doing wrong? Steve Martin
Apt: local URIS must not start with //
I have the Debian slink CD's, and I have a mirror of non-free on a local partition. However, I cannot get Apt to recognize the non-free mirror. Whatever line I put in sources.list, when I run apt-get update, it gives me the error message "local URIS must not start with //" I've tried using deb file:/ ... as well as deb file: ... but neither of these does anything good. Assuming that Packages.gz maybe was causing a problem, I ran dpkg-scanpackages, but this didn't help either. I tried moving Packages.gz to /Debian after that, but this also didn't help. Here is the latest message from apt-get updates: Err file:Debian/ stable/non-free Packages Invalid URI, local URIS must not start with // Ign file:Debian/ stable/non-free Release Reading Package Lists... Done Building Dependency Tree... Done W: Couldn't stat source package list 'file:Debian/ stable/non-free Packages' (/var/state/apt/lists/Debian_dists_stable_non-free_binary-i386_Packages) - stat (2 No such file or directory) W: You may want to run apt-get update to correct these missing files Can anyone point out the error of my ways? Steve Martin
lp not printing
Ever since I installed Debian slink, I have not been able to print. I had been using Magicfilter under the previous version with no problems. Now, I can print, as root, if I send directly to /dev/lp1, but anything else fails. I had uninstalled and reinstalled everything associated with the printer, so I assume the permissions on lpr and magicfilter were correct; at any rate, they show the same permissions I had before I installed slink. So, I created my own filter, and tried printing using the filter. I can still print if I use the commands in the filter and direct the output to /dev/lp1. But if I try to print using lpr, the printer doesn't print, and I get a mail message from "daemon" with the subject line "lp printer job " which states thatthe job in question did not print. The lpr.log shows the same thing. The filters I created are in /usr/local/bin which is owned by root:staff, they are set to mode 755, and they are owned by root:root (I also tried setting them to root:lp). The printcap entry for the printer is (this is as created by Magicfilter, except that I changed the filter file used): lp|hpdj722|HP Deskjet 722c:\ :lp=/dev/lp1:sd=/var/spool/lpd/hpdj722:\ :pw#80:pl#66:px#1440:mx#0:\ :if=/usr/local/bin/ps.if:\ :af=/var/log/lp-acct:lf=/var/log/lp-errs: I am not able to print without a filter, as the printer is an HP 722c, and requires a special program (pbm2ppa) to create the output the printer understands. Can anyone tell me what is going wrong here? Steve Martin
dselect/APT problems
I am having a bit of trouble getting some packages installed from the slink distribution. I have main and contrib on CD, and I have non-free in a separate partition on one of my hard drives (I wasn't able to find CD's for this). This was not a problem under the last version, as I could tell dselect that I was mounting from a directory that was already mounted, and it never noticed that part was on CD and part was on the hard drive. I am not using Apt, as according to the documentation, and in my experience, it has trouble with the CD's. I am using dselect with an access method of multi_mount -- this seems to be the only way to get dselect to recognize both the CD's and the partition. But when I use dselect to install, it runs through a certain portion of the packages, and then just stops, doesn't install the rest, and doesn't seem to recognize that the rest are still scheduled to be installed. In past distributions, as I recall, this indicated that there were broken packages yet to be handled. So I removed all broken packages from the install list, and re-ran install, but it still doesn't do anything with the rest of the packages slated for install. I don't want to use apt-get with HTTP or FTP, as long as I have everything downloaded locally. How can I get the rest of these packages installed? Steve Martin.
LPR/LPD error
Since I installed slink, I cannot print. lpr.log shows an error message "lp: filter 'f' terminated (termsig=13)" but when I try to print, as root, using lpr > /dev/lp1, nothing comes out of the printer. Can anyone tell me what I am doing wrong here? I suspected it might be because I didn't have an /etc/hosts.lpd file, but even adding that did not help. Steve Martin
Re: sourcing the new install
> Sounds like more trouble that it is work. Slink packages (usually) don't > change, > because they are the stable distribution. Right now, packages in unstable are > changing. I use apt-get in the apt package to do what you want to do, > keep my system up to date. > One thing I was hoping to avoid...I haven't seen slink/apt yet, but I am aware that it will install from multiple CD's...under the previous release, I had copied the binaries of Contrib/Non-Free/Non-US to my hard drive, left the i386 CD in the tray, and told dselect that I was installing from the hard drive. Then, I could tell dselect the paths to each of the sets of packages, and it installed all without my having to change CD's (which the previous release wouldn't do, anyway). Will I be able to get a Non-Free and Non-US CD, and will apt do the multiple install from them? Or, if not, how do I make apt aware of where I am keeping those sets of files? Steven C. Martin
Re: gui progs as root in x under normal user?
> On Mon, 17 May 1999 19:18:48 +0200 Peter Granroth wrote: > > How do i start graphical programs under X as root while I'm logged in as a > > normal user? > > If you put the following in root's .bashrc (assuming you use bash), > you can start X programs when su root in a users X session: > > if [ "$USER" != "root" ] ; then > export XAUTHORITY=`eval echo ~$USER/.Xauthority` > fi > Just curious...I can do su root from a user's X session, but I have nothing like that in root's .bashrc, nor in /etc/profile. (I do use bash.) Is this happening through some other mechanism, or am I missing a file? Steven C. Martin
sourcing the new install
I am about to buy a 6GB HDD to supplement the two full 1.6GB HDDs I have, and then I plan to load slink on my system. I have usually bought the CDs, but the resellers I have seen on the net don't seem to include the non-free or non-US, and besides, loading a portion of the total package (even if only the non-free) takes up valuable archive space on my system. So, I am thinking about mirroring slink onto my system from ftp.debian.org, but I was wondering (a) how much drive space I can plan on devoting to the mirror, (b) how much drive space I can plan on leaving open for future additions, and (c) how much trouble it is to install a mirror and keep it up to date. (That is the other reason I was thinking of just mirroring this stuff -- I can always count on my distribution being up to date that way.) Is there any provision for automatically installing upgraded packages, or of informing me of which packages have been upgraded? Steven C. Martin
