Re: [CAD_CAM_EDM_DRO] Re: CNC Control software recommendations
Posted by
Matt Shaver
on 2001-12-21 01:33:20 UTC
Alex wrote:
them is pretty big. I just don't see this as a very likely scenario in a
machine shop environment. Even a little effort at system security would
be enough to deter the casual tinkerer.
image into memory and runs out of a ramdisk from then on, so it's not
bad performance wise after you get past the boot sequence. If this still
isn't satisfactory, see Brian Pitt's comments on solid state disks
(which are currently used with Linux).
accessing their devices...
You could accomplish something similar at run time by doing an md5sum of
the system files (or the whole partition, or just the .ini file) and
then checking it against what it should be for a stock system. This
would make it extra hard for someone to make unauthorised changes. You
could also log any failures or strange conditions to
/var/log/alexslogfile and then e-mail it to yourself daily (if the
machine was net connected). You could run the EMC as a regular user and
not give out the root password! That would slow 'em down...
controls have a "setup screen" or some other place where you configure
the control to fit the machine. I haven't seen one that was even
password protected, and it was possible to really goof up things by
making changes in there.
the people who would be your typical users. Anyone who can do what I
described in the paragraph above is probably not pressing the green
button for a living ;) . I'm also certain that a Linux based system can
be made as secure as needed to satisfy even the most paranoid insurance
underwriter. As a machine maker you have some liability exposure no
matter what. I've said it before, and I'll say it again, "There's no CNC
for Dummies...".
Matt
>> 1. Don't install development tools on the machines you ship.They are readily available, but the amount of knowledge required to use
>
>
> Ok, I understand but how available are these tools on the internet for a free download?
them is pretty big. I just don't see this as a very likely scenario in a
machine shop environment. Even a little effort at system security would
be enough to deter the casual tinkerer.
>> 2. Password protect configuration files, or put them (or the wholeActually, the system boots from the CD and then uncompresses a disk
>> system for that matter) on read only media. An example would be that
>> your system booted from and ran from a CDROM (can be done today). You'd
>> probably need a hard drive for program storage, but there would be NO
>> user alterable files that affect system operation.
>
>
> I agree but due to access speed and other factors I would prefer to have the system on the hard drive.
image into memory and runs out of a ramdisk from then on, so it's not
bad performance wise after you get past the boot sequence. If this still
isn't satisfactory, see Brian Pitt's comments on solid state disks
(which are currently used with Linux).
> Is it possible to use a parallel port type key with Linux?Sure, but... gack! The dongle makers all provide c source code for
accessing their devices...
You could accomplish something similar at run time by doing an md5sum of
the system files (or the whole partition, or just the .ini file) and
then checking it against what it should be for a stock system. This
would make it extra hard for someone to make unauthorised changes. You
could also log any failures or strange conditions to
/var/log/alexslogfile and then e-mail it to yourself daily (if the
machine was net connected). You could run the EMC as a regular user and
not give out the root password! That would slow 'em down...
>> As an example, some yahoo could make changes to theMaybe Fanuc parameters was a bad example... What I mean is that most all
>> parameters of a Fanuc control and cause the same type of problems that
>> you listed, but the manufacturer would disclaim liability because the
>> changes were unauthorized and they couldn't be made inadvertently.
>
>
> Fanuc parameter setting aren't readily available on the internet for download like Linux and its tools.
> Fanuc charges big money for their parameters so they can fall back on - No proof of payment for parameters, no liability.
controls have a "setup screen" or some other place where you configure
the control to fit the machine. I haven't seen one that was even
password protected, and it was possible to really goof up things by
making changes in there.
>> Whoever made the changes had to go to considerable trouble. If someoneI think you greatly overestimate the computer programming abilities of
>> went to the trouble of downloading the source code, making changes,
>> recompiling, and burning a new boot CD, I think you'd be in the clear.
>> Or at least as "in the clear" as you would be with any other system
>> since they're all at least somewhat user configurable.
>
>
> That's allot to have to prove he/she did. Might be easier for the hacker to prove that my system was bad?
the people who would be your typical users. Anyone who can do what I
described in the paragraph above is probably not pressing the green
button for a living ;) . I'm also certain that a Linux based system can
be made as secure as needed to satisfy even the most paranoid insurance
underwriter. As a machine maker you have some liability exposure no
matter what. I've said it before, and I'll say it again, "There's no CNC
for Dummies...".
Matt
Discussion Thread
mfea_2000
2001-12-20 19:31:07 UTC
CNC Control software recommendations
Richard Konnen
2001-12-20 19:43:38 UTC
Re: [CAD_CAM_EDM_DRO] CNC Control software recommendations
Matt Shaver
2001-12-20 19:45:15 UTC
Re: [CAD_CAM_EDM_DRO] CNC Control software recommendations
mfea_2000
2001-12-20 20:03:00 UTC
Re: CNC Control software recommendations
mfea_2000
2001-12-20 20:04:09 UTC
Re: CNC Control software recommendations
samnjanet
2001-12-20 20:09:11 UTC
Re: CNC Control software recommendations
mfea_2000
2001-12-20 20:23:03 UTC
Re: CNC Control software recommendations
Matt Shaver
2001-12-20 22:03:37 UTC
Re: CNC Control software recommendations
Alex
2001-12-20 22:40:18 UTC
Re: [CAD_CAM_EDM_DRO] Re: CNC Control software recommendations
Matt Shaver
2001-12-20 23:00:27 UTC
Re: [CAD_CAM_EDM_DRO] Re: CNC Control software recommendations
Alex
2001-12-20 23:33:08 UTC
Re: [CAD_CAM_EDM_DRO] Re: CNC Control software recommendations
Brian Pitt
2001-12-21 00:34:33 UTC
Re: [CAD_CAM_EDM_DRO] Re: CNC Control software recommendations
Matt Shaver
2001-12-21 01:33:20 UTC
Re: [CAD_CAM_EDM_DRO] Re: CNC Control software recommendations
Matt Shaver
2001-12-21 01:37:42 UTC
Re: [CAD_CAM_EDM_DRO] Re: CNC Control software recommendations
Alex
2001-12-21 07:49:29 UTC
Re: [CAD_CAM_EDM_DRO] Re: CNC Control software recommendations
Keith Rumley
2001-12-21 08:19:53 UTC
Re: [CAD_CAM_EDM_DRO] CNC Control software recommendations
Ray
2001-12-21 17:17:46 UTC
Re: Re: Re: CNC Control software recommendations
Jon Elson
2001-12-21 23:22:35 UTC
Re: [CAD_CAM_EDM_DRO] Re: CNC Control software recommendations
Alex
2001-12-22 00:45:14 UTC
Re: [CAD_CAM_EDM_DRO] Re: CNC Control software recommendations
ballendo
2001-12-22 02:50:47 UTC
Re: CNC Control software recommendations
Paul
2001-12-22 06:17:36 UTC
Re: [CAD_CAM_EDM_DRO] Re: CNC Control software recommendations
currinh@O...
2001-12-22 08:49:20 UTC
Re: CNC Control software recommendations
currinh@O...
2001-12-22 09:03:40 UTC
EMC without rtlinux ?
Jon Elson
2001-12-22 23:28:56 UTC
Re: [CAD_CAM_EDM_DRO] Re: CNC Control software recommendations
ballendo
2001-12-23 02:57:00 UTC
using emc commercially was Re: CNC Control software recommendations
Carol & Jerry Jankura
2001-12-23 07:21:23 UTC
RE: [CAD_CAM_EDM_DRO] using emc commercially was Re: CNC Control software recommendations
Alex
2001-12-23 10:43:06 UTC
Re: [CAD_CAM_EDM_DRO] using emc commercially was Re: CNC Control software recommendations
Matt Shaver
2001-12-23 17:16:07 UTC
Re: [CAD_CAM_EDM_DRO] using emc commercially was Re: CNC Control software recommendations