pnethood - samba share mounter

[HairyWill]

Version 0.6 ready for testers
This has involved massive internal restructuring and should be considered beta quality. Feedback especially welcome from anyone with an unusual setup. The major highlights are:
optional password saving
cifs support if you have it in your kernel (it is included by default in puppy 2.14R) may not work with samba servers
much improved handling of servers and shares with funky characters in their names
user variable network scan duration
the mountpoints have moved to /mnt/network

[Béèm]

I'll recall my installation
Laptop (LT) and desktop (DT) WiFi connected to a rooter.
Name of DT = MD8818
Also LT and DT are connected through a Firewire (1394) connection.
LT = 10.0.0.11 and DT = 10.0.0.10

I downloaded and installed PNH 0.6 in Dingo 4 beta 1

Started pnethood without arguments.

Scanning is done and I have two tabs, a 10_0_0_10 and a MD8818
User guest, password unknown to me but has 5 positions. I didn't give a user or password anywhere.

I can mount the shared resource as well via 10_0_0_10 as MD8818
They are available as expected in /mnt/network.

I don't know if cifs is installed in Dingo 4 beta 1, so I didn't try.

So apart from that user/password issue, I think PNH 0.6 is working so far.

[HairyWill]

User guest, password unknown to me but has 5 positions. I didn't give a user or password anywhere.

Glad it works, thanks for the feedback Béèm.
I expect the password is guest. You could confirm this by checking the "save passwords" box, doing a refresh or quit and then looking in /root/.pnethood/passwords. This happens when the server fails to list its shares when smbclient asks for the list without giving a username and password. The next thing that pnethood tries is the guest account, if this works then it sets the username and password accordingly. By doing this I have managed to access the memory card holder in my HP C5180 AOI printer. It used to really bug me that pnethood showed it as a samba server but didn't show the memory card.

I presume it is possible on the server to set some shares as listable by guest and some that require a password. In this scenario pnethood will find the guest shares and stop scanning so it won't find the password listed shares. All this is hypothetical as I haven't read up on how to configure my XP box this way.

The cifs module is not included with dingo beta 1. Ticking the cifs box shouldn't cause a problem, you will get an error message each time you try to mount a share and it will be done automatically with smbmount instead.

As your shares are accessible to guest, hopefully there is no-one on your network that you don't trust.

I should have added to the release notes. Thanks to Disciple for fixing all the errors in the help file.

[Béèm]

Hi Will,
Guest was a good guess then.
As a test I tried with another user, but pnethood reverted to guest.
I suppose the password is guest as well.
Confirmed when doing a save passwords and refresh.
But somehow strange anyway.
I have a Dutch XP (hate this) and the guest account is gast and isn't password protected.
And yet pnethood can use guest with a password.
But can't use another account with its password. (Or should I save password first?)

[disciple]

Testing in both Dingo (alpha 6 I think) and Puppy 1.x:
1. The first time it is run it gets into some sort of eternal loop, and spits out this sort of message

Code: Select all

/usr/local/apps/pnethood/pnethood: line 22: /root/.pnethood/pnethood.rc: No such file or directory
/usr/local/apps/pnethood/pnethood: line 24: /root/.pnethood/pnethood.rc: No such file or directory
FATAL: Module cifs not found.
rm: cannot remove `/tmp/pnethood/temp': No such file or directory
/usr/local/apps/pnethood/pnethood: line 93: *1000: syntax error: operand expected (error token is "*1000")
/usr/local/apps/pnethood/pnethood: line 376: 13011 Terminated              gtkdialog3 --program=MSG_SCANNING_NETWORK --center
/usr/local/apps/pnethood/pnethood: line 376: 13049 Terminated              gtkdialog3 --program=MSG_SCANNING_MACHINE --center
/usr/local/apps/pnethood/pnethood: line 1: 13110 Terminated              gtkdialog3 --program=MSG_SCANNING_MACHINE --center

** ERROR **: gtkdialog: Error in line 31, near token '</default>': syntax error

aborting...
/usr/local/apps/pnethood/pnethood: line 93: *1000: syntax error: operand expected (error token is "*1000")
/usr/local/apps/pnethood/pnethood: line 376: 13137 Terminated              gtkdialog3 --program=MSG_SCANNING_NETWORK --center
/usr/local/apps/pnethood/pnethood: line 376: 13175 Terminated              gtkdialog3 --program=MSG_SCANNING_MACHINE --center
/usr/local/apps/pnethood/pnethood: line 1: 13236 Terminated              gtkdialog3 --program=MSG_SCANNING_MACHINE --center

** ERROR **: gtkdialog: Error in line 31, near token '</default>': syntax error

aborting...
/usr/local/apps/pnethood/pnethood: line 93: *1000: syntax error: operand expected (error token is "*1000")
/usr/local/apps/pnethood/pnethood: line 376: 13263 Terminated              gtkdialog3 --program=MSG_SCANNING_NETWORK --center
/usr/local/apps/pnethood/pnethood: line 376: 13301 Terminated              gtkdialog3 --program=MSG_SCANNING_MACHINE --center
/usr/local/apps/pnethood/pnethood: line 1: 14273 Terminated              gtkdialog3 --program=MSG_SCANNING_MACHINE --center

** ERROR **: gtkdialog: Error in line 31, near token '</default>': syntax error

aborting...
/usr/local/apps/pnethood/pnethood: line 93: *1000: syntax error: operand expected (error token is "*1000")
/usr/local/apps/pnethood/pnethood: line 376: 14300 Terminated              gtkdialog3 --program=MSG_SCANNING_NETWORK --center
/usr/local/apps/pnethood/pnethood: line 376: 14338 Terminated              gtkdialog3 --program=MSG_SCANNING_MACHINE --center
/usr/local/apps/pnethood/pnethood: line 1: 14399 Terminated              gtkdialog3 --program=MSG_SCANNING_MACHINE --center

** ERROR **: gtkdialog: Error in line 31, near token '</default>': syntax error

aborting...
/usr/local/apps/pnethood/pnethood: line 93: *1000: syntax error: operand expected (error token is "*1000")
/usr/local/apps/pnethood/pnethood: line 376: 14426 Terminated              gtkdialog3 --program=MSG_SCANNING_NETWORK --center
/usr/local/apps/pnethood/pnethood: line 376: 14464 Terminated              gtkdialog3 --program=MSG_SCANNING_MACHINE --center
/usr/local/apps/pnethood/pnethood: line 1: 14525 Terminated              gtkdialog3 --program=MSG_SCANNING_MACHINE --center

** ERROR **: gtkdialog: Error in line 31, near token '</default>': syntax error

aborting...
/usr/local/apps/pnethood/pnethood: line 93: *1000: syntax error: operand expected (error token is "*1000")
/usr/local/apps/pnethood/pnethood: line 376: 14552 Terminated              gtkdialog3 --program=MSG_SCANNING_NETWORK --center
/usr/local/apps/pnethood/pnethood: line 376: 14597 Terminated              gtkdialog3 --program=MSG_SCANNING_MACHINE --center
/usr/local/apps/pnethood/pnethood: line 1: 14658 Terminated              gtkdialog3 --program=MSG_SCANNING_MACHINE --center

** ERROR **: gtkdialog: Error in line 31, near token '</default>': syntax error

aborting...
/usr/local/apps/pnethood/pnethood: line 93: *1000: syntax error: operand expected (error token is "*1000")
/usr/local/apps/pnethood/pnethood: line 376: 14685 Terminated              gtkdialog3 --program=MSG_SCANNING_NETWORK --center
/usr/local/apps/pnethood/pnethood: line 376: 14723 Terminated              gtkdialog3 --program=MSG_SCANNING_MACHINE --center
/usr/local/apps/pnethood/pnethood: line 1: 14784 Terminated              gtkdialog3 --program=MSG_SCANNING_MACHINE --center

** ERROR **: gtkdialog: Error in line 31, near token '</default>': syntax error

aborting...
/usr/local/apps/pnethood/pnethood: line 93: *1000: syntax error: operand expected (error token is "*1000")
The application 'gtkdialog3' lost its connection to the display :0.0;
most likely the X server was shut down or you killed/destroyed
the application.

The last bit is when I killed it obviously

2. It fails to mount Windows 98 shares, for no apparent reason. It can mount shares on the same machine I run it on, but I haven't had a chance to try mounting shares on another linux machine.

3. When it is closed it spits out most of this, which doesn't look particularly healthy

Code: Select all

# pnethood
FATAL: Module cifs not found.
/usr/local/apps/pnethood/pnethood: line 376: 14966 Terminated              gtkdialog3 --program=MSG_SCANNING_NETWORK --center
/usr/local/apps/pnethood/pnethood: line 376: 15005 Terminated              gtkdialog3 --program=MSG_SCANNING_MACHINE --center
/usr/local/apps/pnethood/pnethood: line 1: 15917 Terminated              gtkdialog3 --program=MSG_SCANNING_MACHINE --center
/usr/local/apps/pnethood/pnethood: eval: line 352: unexpected EOF while looking for matching `"'
/usr/local/apps/pnethood/pnethood: eval: line 353: syntax error: unexpected end of file
/usr/local/apps/pnethood/pnethood: line 352: error: command not found
/usr/local/apps/pnethood/pnethood: line 352: /mnt/network/UPSTAIRS/C: is a directory
/usr/local/apps/pnethood/pnethood: line 352: not: command not found
/usr/local/apps/pnethood/pnethood: eval: line 352: syntax error near unexpected token `in'
/usr/local/apps/pnethood/pnethood: eval: line 352: `in'
/usr/local/apps/pnethood/pnethood: eval: line 352: unexpected EOF while looking for matching `"'
/usr/local/apps/pnethood/pnethood: eval: line 353: syntax error: unexpected end of file

4. I also saw this at one stage, but i forgot the circumstances

[mntent]: line 1 in /etc/mtab is bad
[mntent]: line 2 in /etc/mtab is bad
[mntent]: line 3 in /etc/mtab is bad
[mntent]: line 4 in /etc/mtab is bad
[mntent]: line 5 in /etc/mtab is bad; rest of file ignored
[mntent]: line 1 in /etc/mtab is bad
[mntent]: line 2 in /etc/mtab is bad
[mntent]: line 3 in /etc/mtab is bad
[mntent]: line 4 in /etc/mtab is bad
[mntent]: line 5 in /etc/mtab is bad; rest of file ignored

[disciple]

Ah. It mounts Linux shares fine, so there must be something different between Win98 and Linux shares, that is causing a hiccup with this version

[HairyWill]

thanks disciple

1. Is partly caused by it writing an rc file on first run and then immediately trying to read it, inserting a sync should help. The design of the main loop causes it to continue looping unless the gtkdialog exit status is QUIT so when gtkdialog crashes due to an invalid dialog definition it keeps looping. I can adjust the logic of the loop.

2. I'm assuming you have a mount button but clicking it just leads to an error message. Maybe inserting something like
gxmessage "$@"
at the top of /usr/local/apps/pnethood/func_mount would allow you you to check that the parameters being passed are sensible.
Is this the server with a space in its name?

3. It calls eval on each of the output lines from the gui. There are a couple of fields which contain more than just variable assignments. The shell is just complaining about not being able to evaluate one of them. It is unlikely that this is actually causing a problem but it potentially could. I will stop these from being evaluated.

4 :shrug, dunno

I've just realised that the code for handling lots of servers doesn't work properly for several reasons. I think I will remove the automatic insertion of username guest password guest. I will still scan for these shares and maybe mark which shares on a host can be accessed as guest.

[disciple]

2. OK, I'll try that. Yes, I do have a mount button, and no, it is any Win98 server, not just one with spaces.

[disciple]

Code: Select all

TOMS 192.168.2.36 C /mnt/network/TOMS/C   false

I guess this looks sensible. Is the "false" about password saving?

[disciple]

It looks like there is also still a separate issue with servers with spaces in their names. I'll have to wait to test properly when I get a chance to boot "Green room" into Puppy so that I can actually mount the shares.
Check out this:

Code: Select all

/pnethood-0.6/usr/local/apps# ./pnethood/AppRun
modprobe: Can't locate module cifs
./pnethood/pnethood: line 376:  4931 Terminated              gtkdialog3 --program=MSG_SCANNING_NETWORK --center
./pnethood/pnethood: line 376:  4976 Terminated              gtkdialog3 --program=MSG_SCANNING_MACHINE --center
./pnethood/pnethood: line 376:  5065 Terminated              gtkdialog3 --program=MSG_SCANNING_MACHINE --center
./pnethood/pnethood: line 376:  5102 Terminated              gtkdialog3 --program=MSG_SCANNING_MACHINE --center
./pnethood/pnethood: line 352:  5146 Terminated              gtkdialog3 --program=MSG_SCANNING_MACHINE --center
./pnethood/pnethood: eval: line 352: unexpected EOF while looking for matching `"'
./pnethood/pnethood: eval: line 353: syntax error: unexpected end of file
./pnethood/pnethood: line 352: error: command not found
./pnethood/pnethood: line 352: /mnt/network/GREEN: No such file or directory
./pnethood/pnethood: line 352: ROOM/F: No such file or directory
./pnethood/pnethood: line 352: not: command not found
./pnethood/pnethood: eval: line 352: syntax error near unexpected token `in'
./pnethood/pnethood: eval: line 352: `in'
./pnethood/pnethood: eval: line 352: unexpected EOF while looking for matching `"'
./pnethood/pnethood: eval: line 353: syntax error: unexpected end of file

I started the program, tried to mount a share on "Green Room", it failed, and then I when I closed the program I got those errors in line 352.

[disciple]

Yes, I see the same output with a Linux server (i.e. with a server that actually mounts successfully )
So I'm guessing there are some "" missing somewhere.
But that doesn't seem to actually cause any problems.

The

Code: Select all

[mntent]: line 1 in /etc/mtab is bad 

messages don't seem to cause any problems either, but now that I looked up mtab and know what it is, they sound hazardous

I'm wondering if the problem with mounting Win98 shares in 0.6 has anything to do with the way earlier versions took much longer to mount windows shares than linux shares...

[HairyWill]

disciple how about posting your mtab

I've done some bug fixes.
disabled the auto insertion of guest password
it now runs smbclient -L without credentials, as guest and also with password for all servers scanned
the mount function writes all its parameters to the logfile (except password) for debugging
there are now separate buttons to rescan the current servers and to rescan the entire network
cifs support is enabled by default if there is no existing rc (anticipating that this will be in the next dingo alpha with cifs included)

I haven't fixed disciples reported problems with win98 and "GREEN ROOM" as I don't really have enough info to debug. Hopefully the extra output in the logfile will help

[disciple]

1. All I see is this when trying to mount a Win98 share

Code: Select all

func mount called: with
MACHINE=GREEN ROOM
IP=192.168.2.33
SHARE=F
MOUNTPOINT=/mnt/network/GREEN ROOM/F
USERNAME=
PASSWORD=?
USECIFS=false
SMB connection failed
Mount: GREEN ROOM F failed

2. I just discovered that it is the firewall I'm running in Windows that is causing pnethood to take a long time to probe Windows machines, and to mount shares - even though it is supposed to allow anything from the local network, and ask about anything new
If I close the firewall, stuff happens immediately, as with Linux servers, and I don't see these messages (which I see after probing, or mounting)

Code: Select all

timeout connecting to 192.168.2.33:445

Not that the messages make sense, as pnethood seems to be able to do everything anyway
I think the timeout is why 0.6 takes at least 5 times as long as 0.5 took to probe a Windows machine when the firewall is enabled - I see the timeout message five times with 0.6, so I guess 0.6 tries 5 times if it thinks probing failed, but 0.5 only tried once.
So that answers the question of why pnethood was so slow with my Windows servers, but unfortunately closing the firewall doesn't help 0.6x to actually mount shares

3. With this version I don't see the messages that looked like a problem with spaces in server names, but I'm not sure if that is because the problem is fixed, or just because you've tidied up the terminal output, so I simply can't see it.

[disciple]

BTW I only see the mntent message occasionally, and afterward the mtab file is empty, so I might have to put a line in a startup script to make a copy of it or something. But I'm not sure if it would be changed by anything pnethood does before those messages.
And everything like MUT and pnethood still works with an empty mtab anyway, so I don't even understand what it is needed for

[HairyWill]

1. Is the share really called F or has pnethood dropped a dollar symbol somewhere?
Should the share should be accessible without a user name and password?
Have you checked /mnt/network/GREEN ROOM/F to see if the share really is mounted even when pnethood says it isn't?
Does this work

Code: Select all

smbmount //192.168.2.33/f "/mnt/network/GREEN ROOM/F" -o "username=,password="

in the case of passwordless shares it would probably be sensible to drop the -o parameter and do
smbmount //192.168.2.33/f "/mnt/network/GREEN ROOM/F"
though specifying password/username as blank works for me.

2. So with the firewall on smbclient -NL 192.168.2.33 returns an error but with the firewall off it works? If 0.5 saw no shares it should have called smbclient -L twice with different parameters 0.61 should do this three times. It has just occurred to me that the problem could be caused by a slightly different output format from smbclient -L after all I am making some fairly extensive use of automated string processing.

3. I think that the only output that you posted indicating problems with spaces was to do with the eval statement at line 352 in 0.6

Code: Select all

./pnethood/pnethood: line 352: /mnt/network/GREEN: No such file or directory
./pnethood/pnethood: line 352: ROOM/F: No such file or directory 

this should not have had a functional impact on the program and I have limited the gtkdialog output lines that eval is called on so this problem is fixed, not hidden.

Code: Select all

TOMS 192.168.2.36 C /mnt/network/TOMS/C   false

I guess this looks sensible. Is the "false" about password saving?

I missed this before, the false means don't use cifs

[disciple]

3 - that is cool, we can forget about spaces then

2 - that probably explains the timeout messages, although I'm not sure I should be seeing 5 timeouts. What I do know is that what took about 20 seconds before now takes about 2 minutes
I'll experiment with some of those commands when I get a chance.
But I don't think we should worry about this - at least not while there are bigger issues.

1 - That share is really called F (it is just called F: drive on the computer, so it makes sense to use the same name so people know what it is).
No, it really isn't mounting, not just thinking it was unsuccessful.
None of my shares require usernames and passwords, and that machine is the only one with spaces in the name. Apart from the speed/timeout issue with the firewalls on in Windows, pnethood 0.5 can happily access any share on any computer, from any other computer. But pnethood 0.6x can only successfully mount shares on them when they are running Puppy, not Windows (even with the firewall off). So something seems to have changed between the pnethood versions, and it doesn't relate to spaces in names. I can't test that mount command right now, but I wouldn't have thought you'd changed it?

[HairyWill]

2 - that probably explains the timeout messages, although I'm not sure I should be seeing 5 timeouts. What I do know is that what took about 20 seconds before now takes about 2 minutes

2 minutes I'm not surprised you want to avoid probing the server on startup. I found a machine at work yesterday that took about 20 seconds. I have been thinking about how to just load the last known status. I think I have a sensible method for doing it, though the win98 machine is obviously more important.

1 - That share is really called F (it is just called F: drive on the computer, so it makes sense to use the same name so people know what it is).
No, it really isn't mounting, not just thinking it was unsuccessful.
None of my shares require usernames and passwords, and that machine is the only one with spaces in the name.
...
I can't test that mount command right now, but I wouldn't have thought you'd changed it?

from what you posted, the parameters going in to func_mount look correct, before looking at the logic in the program we need to confirm that the mount command works/fails with the correct parameters

[HairyWill]

aaaarrrrgggghhhhhh
when I run 0.61 from the menu in dingo it manages to get past this code without initialising the variables

Code: Select all

if [ ! -e "$HOME/.pnethood/pnethood.rc" ]; then
echo "SAVEPASSWORDS=false
SCANSPEED=1
USECIFS=true" > "$HOME/.pnethood/pnethood.rc"
sync
fi
source "$HOME/.pnethood/pnethood.rc"

I've run it from the console a few times and it works fine.
This is a problem because the first time it runs if it doesn't initialise the variables properly from the rc file when the program exits it saves the variables as empty. It is easy to design around but I don't understand why it happens.

[HairyWill]

bug fix for above

[disciple]

If you are interested:
Without firewall

Code: Select all

# smbclient -NL 192.168.2.36
session request to 192.168.2.36 failed (Called name not present)
session request to 192 failed (Called name not present)
session request to *SMBSERVER failed (Called name not present)

With firewall

Code: Select all

# smbclient -NL 192.168.2.36
timeout connecting to 192.168.2.36:445
session request to 192.168.2.36 failed (Called name not present)
timeout connecting to 192.168.2.36:445
session request to 192 failed (Called name not present)
timeout connecting to 192.168.2.36:445
session request to *SMBSERVER failed (Called name not present)

And I found the problem with the mount command:
The server at 192.168.2.36 is called TOMS:

Code: Select all

# smbmount //192.168.2.36/C "/mnt/network/TOMS/C" -o "username=,password="
timeout connecting to 192.168.2.36:445
6888: session request to 192.168.2.36 failed (Called name not present)
timeout connecting to 192.168.2.36:445
6888: session request to 192 failed (Called name not present)                                                                                         timeout connecting to 192.168.2.36:445                                                                                                                6888: session request to *SMBSERVER failed (Called name not present)                                                                                  SMB connection failed

# smbmount //192.168.2.36/C "/mnt/network/TOMS/C"                         
timeout connecting to 192.168.2.36:445
6951: session request to 192.168.2.36 failed (Called name not present)
timeout connecting to 192.168.2.36:445
6951: session request to 192 failed (Called name not present)
timeout connecting to 192.168.2.36:445
6951: session request to *SMBSERVER failed (Called name not present)
SMB connection failed

# smbmount //TOMS/C "/mnt/network/TOMS/C" -o "username=,password="
timeout connecting to 192.168.2.36:445

That final command did successfully mount it, and I don't think any of the timeout messages wouldn't be there if the firewall was off.

So there seems to be some issue with how Windows is identifying the server names. Interestingly, if I log into the router and look at the DHCP table, it just lists the IP addresses of machines running Puppy, but for machines running Win98 it has the names as well as IP addresses.