trinity-users@lists.pearsoncomputing.net

Message: previous - next
Month: July 2016

Re: [trinity-users] Re: Re: tdenetworkmanager, where is the conf file

From: Gene Heskett <gheskett@...>
Date: Wed, 6 Jul 2016 13:29:18 -0400
On Monday 04 July 2016 03:46:38 deloptes wrote:

> Gene Heskett wrote:
> > But, anything that looks or acts like network manager has been
> > removed from my system. So any comments I might make about NM should
> > be taken as the best swag* I can make based on 18 years of running
> > linux.
>
> NM is useful on notebooks that change locations, interfaces etc. If
> one uses a home network with many computers I would advise setting up
> proper dhcp and perhaps name server. You don't have to take care of
> each machines own files. But this is usually an individual preference.
>
> regards
>
The last time, several years ago, that I tried using dhcp on my lappy, 
which is itself an antique, I had 3 problems.

1. NM insisted on using the bcm-4318 radio in the lappy even if it was 
powered down by the switch.  It totally ignored a radio in a usb dongle 
that actually worked where the bcm-4318 never worked for more than 1 
minute, even when useing the drivers from xp that came installed to run 
it.  Rebooting to xp disclosed that this pos radio was junk from the 
gitgo.

2. Getting it setup via dhcp with a usb radio dongle, so I could sit down 
while out in the shop, and write gcode for one of the 2 machines out 
there worked flawlessly over an ssh -Y machinename connection, until I 
wanted to ssh -Y lappy, which is its own name from the house machine.  
So the dns lookup which NM should have set up, wasn't there, forcing me 
to string a 25' hunk of cat5 plugged into its ethernet port.  Adding its 
ip address and name to all the hosts files, and fixing all the network 
related files to be immutable after edited correctly, and then its just 
one of the family.

Strangely, when NM found it couldn't rewrite those files, it made no 
complaint in any log.  And it didn't spin its wheels, burning up the cpu 
either.

3. I've left it that way in all subsequent installs.  Now I remove NM and 
other than making /etc/resolv.conf a real file, its a 10 minute job 
after a fresh install to have a working network regardless of which 
socket I plug its cat5 cable into.  With 2, 8 port switches available, 
soon to be 3 as I'll need another in the garage as I bring a bigger 
Sheldon 11x36 lathe to life with LinuxCNC.  So the end of the cat5 I 
strung thru the nether regions of the house to get to the garage, will 
eventually have an 8 port switch on the end of it.  That cable was 4 
days getting run, a right PITA.

I think by now in 2016, if the computer has a working radio, or an 
ethernet over cat5 port, that it may be able to be made to work by now 
by people willing to do whats needed, but my long term experience with 
it has caused me to shoot it on sight. Bug squawks were ignored, or told 
to buzz off, and fixed versions were way too long working their way thru 
the pipeline to the users install. Several times I pulled the src and 
built it here, and it was better but still kept tearing down a working 
link and trying to find another, so it would eventually go back to the 
working link, which took it a good 2 minutes to do, then 3 minutes later 
I'm typing in the dark to /dev/null again.  And I still have at least 
half my hair by using my lambasted by everybody methods. All I can say 
is that I can carry that lappy to anyplace on my local network, plug it 
in and boot it up, and the network Just Works(TM).  Call me an idiot, 
and quite a few have, but simpler is better when it just works 100% of 
the time here. I have no reason to take it to mcd's and use their radio 
because there I would not have dd-wrt in my router standing guard. Its 
very very good at that.  Recommended as the network guard dog by Grandpa 
Gene.

Cheers, Gene Heskett
-- 
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
Genes Web page <http://geneslinuxbox.net:6309/gene>