Re: [Emc-users] Some odd things

[paul_c]

On Wednesday 18 February 2009, Jon Elson wrote:
> I recently wanted to check the latency numbers, and ran into this when I
> tried to run latency-test :
>
> ./latency-test: line 136: halrun: command not found

Has been commented on several months ago. A fix requires the setting of PATH 
to include current working directory.

> After some poking around, it seem the last line of latency-test is
> halrun lat.hal

This file is created in a /tmp/$random directory - The full path/name should 
be passed to halrun. Fix that, and halrun will then go tits up over hal.xml..

> And the file lat.hal dosn't exist.  So, for some reason, latency-test
> doesn't seem to be creating this file.

Both issues can be resolved with a trivial fix (works on my install).


--
Open Source Business Conference (OSBC), March 24-25, 2009, San Francisco, CA
-OSBC tackles the biggest issue in open source: Open Sourcing the Enterprise
-Strategies to boost innovation and cut costs with open source participation
-Receive a $600 discount off the registration fee with the source code: SFAD
http://p.sf.net/sfu/XcvMzF8H
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] CAN-Bus drivers

[paul_c]

On Saturday 07 February 2009, Thomas Kaiser wrote:
> you can find a CAN PCI card for your PC. The question is if there is a
> drive available which works with Linux

There are several drivers available for use with Linux (and a few with 
realtime Linux) - Most of the code I have relates to SJA1000 variants.

> and EMC2. 

Short answer, No.


--
Create and Deploy Rich Internet Apps outside the browser with Adobe(R)AIR(TM)
software. With Adobe AIR, Ajax developers can use existing skills and code to
build responsive, highly engaging applications that combine the power of local
resources and data with the reach of the web. Download the Adobe AIR SDK and
Ajax docs to start building applications today-http://p.sf.net/sfu/adobe-com
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] pagesides

[paul_c]

On Friday 06 February 2009, John Thornton wrote:
> I didn't see \papersides in any of the preambles nor could I find
> Documents/Settings/Pagelayout...
> We are still using Lyx 1.3.7 are you refering to a newer version?

I have LyX 1.5.5 installed here.

> I tried adding \papersides to the preambles but got an error. Do you have a
> link that explains \papersides?

I will send you a diff - It will have to be against an old cvs checkout 
because I can not currently do a `cvs up`[1]







[1] Error:

~/emc2# cvs up -dPA
ssh: connect to host cvs.linuxcnc.org port 22: Connection refused
cvs [update aborted]: end of file from server (consult above messages if any)

--
Create and Deploy Rich Internet Apps outside the browser with Adobe(R)AIR(TM)
software. With Adobe AIR, Ajax developers can use existing skills and code to
build responsive, highly engaging applications that combine the power of local
resources and data with the reach of the web. Download the Adobe AIR SDK and
Ajax docs to start building applications today-http://p.sf.net/sfu/adobe-com
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] How to turn the spindle / coolant on to resumeafter program stop ? (Alex Joni might know about this)

[paul_c]

On Thursday 05 February 2009, John Thornton wrote:
> It looks like the page numbering is some random mix... we use lyx to create
> the docs as well as some custom things to create the html's. I took a quick
> look and didn't find a way to specifiy where the page numbers should be...

Change \papersides from 1 to 2 in the preamble of each LyX file - This can be 
done quite quickly with sed and friends. Or Documents->Settings->PageLayout 
and tick the Two-sided document box (be warned that it may well overwrite 
other settings).

Alternatively, use the fancyhdr package for finegrained control over header & 
footer placement.


--
Create and Deploy Rich Internet Apps outside the browser with Adobe(R)AIR(TM)
software. With Adobe AIR, Ajax developers can use existing skills and code to
build responsive, highly engaging applications that combine the power of local
resources and data with the reach of the web. Download the Adobe AIR SDK and
Ajax docs to start building applications today-http://p.sf.net/sfu/adobe-com
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] EMC Board Election Results

[paul_c]

On Thursday 05 February 2009, Michael Cornelius wrote:
> With 1047 ballots distributed and 97 returned, the result of the voting
> is as follows:

Thank you Michael for your efforts in conducting a survey in fair and neutral 
manner. However, you have been provided with an incomplete list for the 
second year (possibly third), so the results should be declared null and 
void.



--
Create and Deploy Rich Internet Apps outside the browser with Adobe(R)AIR(TM)
software. With Adobe AIR, Ajax developers can use existing skills and code to
build responsive, highly engaging applications that combine the power of local
resources and data with the reach of the web. Download the Adobe AIR SDK and
Ajax docs to start building applications today-http://p.sf.net/sfu/adobe-com
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Telnet Interface Questions

[paul_c]

On Sunday 11 January 2009, Leslie Newell wrote:
> A much more likely scenario is a malicious user trying to mess around
> with a CNC machine that is on the same local network.

Or it could be a curious login user - For anyone interested in system security 
and stability, try poing a few hal variables, look in the /proc directory 
tree, echo a random string to any/all of the rtapi or hal entries.

Try also, with the aid of emc_module_helper, unloading some of the multitude 
of "components" followed by the main modules, finishing with rtai_rtapi - Let 
us know what happens when you shut down.

For anyone with a Mesa card (or any other fpga card), using the provided 
utility, upoad a new bit file, then pick a bit file at random and try that.



--
Check out the new SourceForge.net Marketplace.
It is the best place to buy or sell services for
just about anything Open Source.
http://p.sf.net/sfu/Xq1LFB
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] FW: Telnet Interface questions

[paul_c]

On Sunday 11 January 2009, AKSYS Tech Pty Ltd wrote:
> Just a quick update.  I did as paul_c suggested and removed the queue
> type from the *.nml file. This seems to have fixed my issue, and I now
> seem to have error reporting coming back into my micro.  I can see where
> Alex is coming from and understand the possible problems of doing this.

Ideally, you should only ever have one process writing to a buffer, and one 
process reading (as per the NIST RCS handbook). For multiple processes 
reading or writing to a single buffer, it would be better if a 
secondary "channel" was spawned for each additional connection - This would 
require some small changes to libnml along with the addition of some 
arbitration policy.



--
Check out the new SourceForge.net Marketplace.
It is the best place to buy or sell services for
just about anything Open Source.
http://p.sf.net/sfu/Xq1LFB
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] New cutter compensation algorithm in TRUNK

[paul_c]

On Sunday 11 January 2009, Chris Radek wrote:
>  Please send a patch that makes it build without warnings
>  You can send it to emc-developers or to me personally.

Restore my CVS access, and you won't need to mess with "patches".

--
Check out the new SourceForge.net Marketplace.
It is the best place to buy or sell services for
just about anything Open Source.
http://p.sf.net/sfu/Xq1LFB
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] New cutter compensation algorithm in TRUNK

[paul_c]

On Thursday 08 January 2009, Chris Radek wrote:
> Today I merged my cutter compensation work.  The new algorithm
> handles concave corners and is not picky about entry moves.

You need to include cstdlib & cstring headers in interp_queue.cc and then it 
will at least compile with gcc-4.3.

--
Check out the new SourceForge.net Marketplace.
It is the best place to buy or sell services for
just about anything Open Source.
http://p.sf.net/sfu/Xq1LFB
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Telnet Interface questions

[paul_c]

On Saturday 10 January 2009, Alex Joni wrote:
> > Remove the 'queue' type from the emcError buffer line in your *.nml
> > config.
>
> Removing the 'queue' type might lead to problems, as more than one error
> messages won't get stored.
> Any subsequent error messages will overwrite existing ones in the channel.
> Currently task and iocontrol are 2 examples of error message sources, and
> if errors happen from both, one of the two could get lost. (at least that's
> my interpretetion of it from the RCS handbook...)

Yes, which is why I said:
>> You will then be able to read most recent error message passed from lower
> > levels.



--
Check out the new SourceForge.net Marketplace.
It is the best place to buy or sell services for
just about anything Open Source.
http://p.sf.net/sfu/Xq1LFB
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Telnet Interface questions

[paul_c]

On Saturday 10 January 2009, Leslie Newell wrote:
> The problem is that the messages are deleted once they have been read.
> In most cases the GUI will get the messages because it is polling them
> quite frequently. By the time emcrsh gets there they have been swallowed
> by the gui.
>
> Unfortunately to change this would require a major redesign of emc's
> messaging system.

Remove the 'queue' type from the emcError buffer line in your *.nml config. 
You can also locate the updateError() function in shcom.cc, and change 
emcErrorBuffer->read() to emcErrorBuffer->peek(), then recompile. You will 
then be able to read most recent error message passed from lower levels.






--
Check out the new SourceForge.net Marketplace.
It is the best place to buy or sell services for
just about anything Open Source.
http://p.sf.net/sfu/Xq1LFB
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Software question

[paul_c]

On Thursday 08 January 2009, Alex Joni wrote:
> NML is the mechanism which is used for defining the dataset and
> communication paths between the various userspace components.
> If you look at the (slightly outdated) overview of EMC:

> * there were (every couple of years) talks about replacing it with
> something else, but:
>     * while there are a couple alternatives, I'm not sure anyone studies
> them deeply enough to really understand which would fit best

So what do *you* consider to be the "alternatives" capable of providing 
similar functionality ?


--
Check out the new SourceForge.net Marketplace.
It is the best place to buy or sell services for
just about anything Open Source.
http://p.sf.net/sfu/Xq1LFB
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Software question

[paul_c]

On Thursday 08 January 2009, Kenneth Lerman wrote:
> For my use and for (as far as I can tell), almost everyone else, it is
> unnecessary.

NML describes the data structures as well as the methods for transmitting the 
PODs to the various parts of EMC - *Everyone* uses it. Even within the task 
planner, the NML PODs are central to how canonical commands are queued and 
then dispatched.

> It is also complex. 

So you remove the data structures and defines in emc.hh (also emc_nml.hh along 
with several other files in src/emc/nml_intf) and delete libnml, what viable 
alternative do you suggest ?

> For my use and for (as far as I can tell), almost everyone else, it is
> unnecessary. It is also complex.

That also describes several other parts of emc2.

--
Check out the new SourceForge.net Marketplace.
It is the best place to buy or sell services for
just about anything Open Source.
http://p.sf.net/sfu/Xq1LFB
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Software question

[paul_c]

On Wednesday 07 January 2009, Kenneth Lerman wrote:
> NML is used because it supports an environment where functions are split
> across multiple machines.
>
> I don't know how widely that functionality is used, though. As far as
> I'm concerned, it adds an unnecessary level of complexity.

Please define "unnecessary ... complexity" within the context of emc2.





--
Check out the new SourceForge.net Marketplace.
It is the best place to buy or sell services for
just about anything Open Source.
http://p.sf.net/sfu/Xq1LFB
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] info on emcsvr

[paul_c]

On Saturday 03 January 2009, Jeff Epler wrote:
> The source code to 'emcsvr' is ./src/emc/task/emcsvr.cc

Correct - It also uses a number of classes from libnml (or rcslib for EMC1).

> Its function is to manage the lifetime of the nml shared memory segments
> used by various parts of emc (gui, milltask and iotask).

WRONG. It's purpose is to arbitrate communications with remote processes via a 
network connection. It should only ever be run as an optional binary when 
remote connections are expected.


--
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] EMC1 ?

[paul_c]

On Friday 02 January 2009, Richard F. Amaral wrote:
> I'd like to know where I can get the public domain version of EMC? Is  
> it available on a cvs server like cvs.linuxcnc.org? If so, please  
> provide name of branch/tag etc.

cvs -d:pserver:anonym...@emc.cvs.sourceforge.net:/cvsroot/emc login
(Press return when asked for a password)
cvs -d:pserver:anonym...@emc.cvs.sourceforge.net:/cvsroot/emc co -rnist emc
cvs -d:pserver:anonym...@emc.cvs.sourceforge.net:/cvsroot/emc co -rnist rcslib

Or: 
ftp://ftp.isd.mel.nist.gov/pub/emc/emcsoft
ftp://ftp.isd.mel.nist.gov/pub/emc/rcslib

You will need both rcslib and emc tarballs/checkouts.

--
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Ethernet, rtai, and rtnet

[paul_c]

On Monday 29 December 2008, Sebastian Kuzminsky wrote:
> The rtnet front page  claims they provide a POSIX
> socket API, but I was unable to find out if they support SOCK_RAW.

ChangeLog:
0.9.6
- added support for SOCK_RAW packet sockets

Grepping the rtnet sources, I see two example sources that implement raw 
socket services (one a listener, the other, a server).



--
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] end of program

[paul_c]

On Wednesday 17 December 2008, Ray Henry wrote:
> (1) What you are describing was addressed when an attempt was made to
> allow integrators the ability to set these behaviors by parameter
> changes.  I've argued for such things for years and a first stab at it
> was made by Keith Rumley about three+ years ago on the BDI versions.
> You could see his revisions in CVS.

Minor correction - The changes were made on a branch of the original EMC code. 
The bug fixes were merged in to the BDI-4 branch of emc2, but possibly not in 
to the main trunk.

> IMO the proper way to handle these things will require a major rework
> and expansion of the EMC2 task planning code.

Offsets are predominantly in the realm of the interpreter and have little to 
do with the task planner.

> (2) A related ability, suspending a program using abort or some such
> then switching to manual and moving to allow the replacement of a broken
> or worn tool, has been discussed at some length.  A similar issue is
> raised with auto tool wear sensing.

Now this does require changes at the task planner level, and also in the 
trajectory planner - Manual moves & unplanned tool changes opens up a number 
of problems associated with tool offsets.

> IMO there is a solution to this and several other "issues" but it would
> require a significant revision of the interpreter code.

There is/was an extensible interpreter being worked on that allowed for custom 
G/M codes via a config file - The intent was to provide 
different "personalities" to emulate various controls (i.e. Fanuc or 
Heidenhain) along with closer coupling of PLC functions..





--
SF.Net email is Sponsored by MIX09, March 18-20, 2009 in Las Vegas, Nevada.
The future of the web can't happen without you.  Join us at MIX09 to help
pave the way to the Next Web now. Learn more and register at
http://ad.doubleclick.net/clk;208669438;13503038;i?http://2009.visitmix.com/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Anyone have EMC2 running on OSX?

[paul_c]

Hi George

On Sunday 19 October 2008, George Lawrence Storm wrote:
> Before I have to jump through a number of hoops, does anyone have EMC2  
> running on any version of OSX?

Not emc2, no. But I've just compiled some code and got a GUI and supporting 
libraries running (on 10.4) - The middle layer(s) compiled without error, so 
once the low level stuff can be built, it should run.

I doubt the real time services provided by OS X will give adequate for a 
truely deterministic trajectory planner - I would envisage the core of this 
(and any pulse generator) having to run on an embedded board or fpga. The 
alternative would be to install Linux along with a suitably patched kernel 
and thus sidestepping the complexities of having to deal with the IOkit.


Regards, Paul.


P.S. Screenshot for the interested.. 
http://zathras.homelinux.org/images/OSX_cnc.jpg



-
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] No version for "nano2count" found

[paul_c]

On Thursday 02 October 2008, Евгений Александрович wrote:
> rtapi: No version for "nano2count" found: kernel tained.
>
> Is it problem or just information message?
> If it is a problem, what could be wrong?

It is a compile time problem that is easily solved - What kernel & RTAI 
versions are you using ?



Paul.

-
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Problem on EMC2 starting with rtai_hall.ko

[paul_c]

Hi Evgeny

On Thursday 25 September 2008, Евгений Александрович wrote:
> > sounds like rtai is already loaded. Either you compiled it into the
> > kernel, not as a module.
> What do you mean. Did I somesing worng in kernel?

It is impossible to "compile rtai_hal" in to the kernel - The Makefile 
associated with rtai_hal explicitly states rtai_hal.ko will be compiled as a 
module.

> # emc
> EMC2 - 2.2.5
> Machine configuration directory is
> '/usr/local/etc/emc2/sample-configs/stepper' Machine configuration file is
> 'stepper_mm.ini'
> Starting EMC2...
> insmod: error inserting '/usr/realtime/modules/rtai_hal.ko': -1 File exists
> Realtime system did not load
> Shutting down and cleaning up EMC2...
> /bin/pidof: illegal option -- x

There is the most likely root cause of your system not cleaning up after 
running emc2.

> BusyBox v1.6.1 (2007-08-26 04:52:27 MDT) multi-call binary

BusyBox has not been tested or validated with the emc2 run scripts. It is 
quite possible that more scripting errors are waiting to bite if you are not 
using a bash shell.

Question: Are you loading the RTAI modules from an init script during boot 
up ?



-
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Back on Ubuntu 8.04 + EMC2 Problems

[paul_c]

On Tuesday 23 September 2008, Alex Joni wrote:
> That is correct. 2.6.24-16-rtai is the latest available. It is usually such
> a big pain to build a working -rt version that works on as many PCs as
> possible,

Rubbish, perpeptuating the myth that compiling kernels is some black art.. 
Very few configuration options cause problems, and these are well documented. 
Patching a kernel takes less than a minute, and even after pulling a kernel 
config from an earlier version, building a kernel package takes thirty 
minutes or so (for a single flavour).

 A bash script to automate the task is only a couple of lines, maybe six if 
you are uploading to a "proper" Debian repository.

> that I don't really consider building new versions unless there 
> are really serious updates/problems.

In the last couple of weeks there have been, what, three reports ?

> If there is someone out there who wants to contribute by doing this job,
> then we'll be happy to hear so. 

Four i386 flavours along with associated header and support packages, 30 
seconds to trigger the build script. A couple of hours of CPU time, job done.



-
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Arduino with Ubuntu 8.04

[paul_c]

On Sunday 21 September 2008, Dirk wrote:
> If you are lucky you can use the debian way of compiling  
> kernels. This will solve dependency's too. Try: apt-get install kernel-
> kpkg" this will install all stuff needed to compile your own kernel in  
> a nice .deb package.

That should be kernel-package - However, Ver 11.001 currently produces 
slightly broken header packages for 2.6.24 (and up). See:
http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=475029 for a workround.

Have a patch (almost) ready to send to the package maintainer(s), so an 
updated version could be out soon.


Regards, Paul.


-
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] New error after upgrade

[paul_c]

On Saturday 13 September 2008, Alex Joni wrote:
> Sure, it's always been open for review/comment, though we should move this
> to the emc-devel list,

Still awaiting answers to other questions on the dev list, so does it really 
make a difference where the topic is raised ?

-
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] New error after upgrade

[paul_c]

Alex

On Thursday 11 September 2008, Alex Joni wrote:
> I get the error your pastebin shows when I try to load probe_parport more
> than one time.

Attempting to load a module more than once should result in -EEXIST being 
returned. If you are getting as far as triggering a call trace, then there is 
a more serious bug.

Do you wish to discuss the correct way to handle parport registration, or is 
the code not open to review/comment without inciting cries of "fud" from the 
CCM dept. ?


---

Paul.

-
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] New error after upgrade

[paul_c]

Alex

On Wednesday 10 September 2008, Alex Joni wrote:
> the culprit might be probe_parport, can you check if parport_pc shows up
> under lsmod (right after booting the machine, then after running emc2), and
> if you have loadrt probe_parport in your hal file.

Indeed, probe_parport is the offending "module" - Ripping $random code from a 
common kernel module without considering the implications opens the door to a 
whole load of problems.

Take a look at the pnp_driver struct (in probe_parport.c) and note the name 
field - A conflict is just waiting to happen.


---

Paul.


-
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] New error after upgrade

[paul_c]

Hi Steve

On Wednesday 10 September 2008, Steve Blackmore wrote:
> Thanks Paul - I do have the modules If anybody needs them,

Thanks, but no - You get to keep all the broken pieces ;-)

> what's the next course of action? Is it something I've done or an error in
> the updates somewhere?

It is nothing you have done - The error is upstream and I'm afraid you will 
have to wait for a "fix"... (unless you can speak C).


Regards, Paul.




-
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] New error after upgrade

[paul_c]

Alex

On Tuesday 09 September 2008, Alex Joni wrote:
> can you post a list of modules loaded?

You don't need a list of loaded modules to diagnose the problem (actually, 
there are two).

> From: "Steve Blackmore" <[EMAIL PROTECTED]>
> > [ 6204.322690] Pid: 5635, comm: insmod Tainted: GF       2.6.24-16-rtai

First problem - A compile time issue is resulting in the kernel being tainted.

> > [ 6204.322787] kobject_add failed for parport_pc with -EEXIST, don't try
> > to register things with the same name in the same directory.

parport_pc is already registered, which means a module of that name is loaded. 
To fix this requires a more involved fix along with an understanding of 
kernel internals.



Paul.


-
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Stepper configuration

[paul_c]

On Wednesday 13 August 2008, John Kasunich wrote:
> Unfortunately, conversion from clocks to time is non-trivial, so we
> simply report the results in clocks.

(get_cycles()*10^6)/cpu_khz

Is that non-trivial ??

> The situation is getting worse, not better.  In the name of "progress", we
> now have to deal with CPUs that change their clock frequencies on the
> fly.

You need to read the RTAI docs again - Disabling CPU_FREQ is *strongly* 
recommended because it messes up realtime response. But if you have disabled 
all power management, that option will not be available.

> I've even heard a  report that on some modern CPUs, the TSC can no longer
> even be counted on to be monotonic.

Almost correct - What happens with multiple processors is the TSC drifts 
slightly, however, recent developments in the kernel & RT patches 
periodically resyncronise the TSCs. Disregarding TSC drift, the value 
returned will be accurate enough for your needs.

---

Paul.




-
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Stepper configuration

[paul_c]

Hi Sebastian

On Tuesday 12 August 2008, Sebastian Kuzminsky wrote:
> Hm, looks like rt_spin_lock() and friends are not exported through rtapi.

Why would you want rtapi functions for spinlocks ?

> It looks like it would be easy to add a bunch of functions (or maybe
> preprocessor macros) to call the RTAI locking code through RTAPI
> front-end functions, but I dont (yet) know how to do this stuff with
> RT-Linux, and I dont currently have a way to build & test RT-Linux side
> of any RTAPI locking implementation.

RT-Linux is pretty much a dead project from what activity I have seen. The 
last patch I saw was for a very early 2.6 series kernel and possibly 
experimental. A quick look at the WindRiver overview suggests that they use a 
2.6.14 kernel, but I doubt anyone (here) uses it.

> Are there maybe some RT-Linux systems in the EMC2 build farm?

There was a 2.2.x series kernel patched with a very old RT-Linux version, but 
your driver(s) won't compile on anything other than 2.6.xx.

> Does anyone run EMC2 on RT-Linux?

RT-Linux support is a dead end, rtapi for RT-Linux does little more than add 
to the general bit rot.


---

Paul.

-
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Stepper configuration

[paul_c]

On Sunday 10 August 2008, Matt Shaver wrote:
> 1b. I can't help but think that there is (buried under the IDE) a command
> line VHDL>bit-image compiler.

Command line interface is trivial enough to allow non-interactive builds.. 
e.g.:
#! /bin/bash
xst -ifn ${project}.xst -ofn ${project}.syr
ngdbuild -dd . -uc ${project}.ucf ${project}.ngc ${project}.ngd
map -o ${project}_map.ncd ${project}.ngd ${project}.pcf
par -w -ol high -t 1 ${project}_map.ncd ${project}.ncd ${project}.pcf
bitgen -w -d ${project}.ncd ${project}.bit
promgen -w -u 0  ${project}.bit -p bin -o  ${project}.bin

> I couldn't find it, and we probably couldn't 
> redistribute it without permission, but if there is, and we could, that
> would solve the first problem with making new FPGA code if incorporated
> into the build system.

If you read the Xilinx license.txt, you will find it goes on at length 
about "copyrighted material", "trade secrets", "proprietory information", 
and "Intellectual Property". Section 2 make interesting reading along side 
the GPL text..

I dare say Altera's software would also provide conflicts in licensing 
conditions.. N.M.P.

---
Paul.

-
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Disabling Sound for Increased performance.

[paul_c]

Hi Dan

On Friday 25 July 2008, Organic Engines wrote:
>   After a whole bunch of thought it occurred to me that sound is
> essentially real time and probably a resource hog.

I'd suggest using tools like top or htop to profile the system in order to 
identify the resource hogs - Sound doesn't consume much CPU time in 
comparison to some of the other stuff. If you are running Gnome or KDE 
desktop, replace it with a lightweight window manager such as xfce4 or 
fluxbox.

>   Any reason not to unload all those alsa, oss, snd and etc modules?

None at all - Ultimately, you may well want to compile your own kernel 
optimised for the processor in question.

>   Ultimately network, printing, bluetooth and other stuff can go once
> your system is working.

Networking is very useful to have - The multimedia stuff can go, along with 
non-Linux file system support (retain ISO9660 & NFS).

>   The idea is getting toward what would be essentially an embedded
> system. No fluff to distract your pc from doing it's machine control job.

Yup - A minimal install should fit quite comfortably on a 500Mbyte partition 
and yet leave room for additional utilities.


Regards, Paul.

-
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Data Types

[paul_c]

Thanks for taking the time to reply Ken.

On Sunday 08 June 2008, Kenneth Lerman wrote:
> I've changed the subject of this to better reflect what it is about, and
> am responding on both users and developers lists so that the original
> audience will still see it.

Noted - See below as to why it is relevant for the user.

> I do seem to recall a prior discussion about data types, but since I
> don't keep files of such things, I don't have a pointer to it and I
> could easily be mistaken.

I don't recall seeing any discussion on a mailing list, so if one did take 
place, it was probably some little nook in cyberspace..

> While using doubles to get higher precision would be a solution, it is
> aesthetically nasty for us purists. Why would one use a floating point
> number for something that is inherently an integer?

Primarily, it simplifies the math and avoids having to use a fixed point math 
library. But...

> Also, one could easily imagine porting HAL to a platform where floating
> point arithmetic has significantly higher cost than integer arithmetic.

I don't think that is likely to happen any time soon - However, there is a 
strong case for moving *all* floating point math out of kernel space and 
performing the minimum of trivial integer math at the lower levels.

> For example, some of us are thinking about a split where the HAL layer would
> be moved to a remote controller using (say) an ARM processor.

Sure - If it was kept as a simple low level layer for hardware configuration, 
I would go along with that. Dump the cruft, and you might get an acceptable 
level of performance out of an ARM platform.. Take a look at the latency 
figures under load before you commit to such an undertaking though - 100mSec 
and higher is quite normal.

> The general answer to the issue of integer typing is that the C standard
> (at least the older versions I'm familiar with) does not say that an
> integer is 32 bits. I've worked on platforms where an int was 16 bits,
> and I can imagine machines where it might be eight or 64 bits.

ISO/ANSI only gives a minimum range for an int (-32767 to 32767), but it is 
generally accepted that it uses what ever size is most efficient on the host 
machine. So on a 32bit processor (i386 or PowerPC), it would be 32 bits, and 
on an x86_64, 64 bits (but never 8 bits).

> My answer to the issue of more precision for encoders would be to use a
> type hal_s64. That would be defined as a long long on the machines I
> usually use.

For users like Aram, an int32 is going to roll over at 2^31, so with a 
5,000,000 count encoder, 429 rotations would be his limit. Certainly, going 
to a 64 bit integer would improve things, but not if the counts get scaled 
and truncated to a float at the higher levels - A float is only good for six 
decimal digits, whislt a double is good for ten - When you go converting to a 
double, perform some math, convert to a float, then back to a double, do some 
more math, and finally post the result as a float, accuracy is going to be 
suspect at best.

> I'm aware that there are some "standard" definitions for such types that
> do not have "hal" in the name.  The kernel source is full of things like
>   int32, s32, _s32, __s32.  Using hal_s32 makes more sense to me than
> any of those because I can at least figure out where to go to find the
> definition if I ever want to port this.

The kernel developers are working to reduce the number of typedefs and the GCC 
team move ever closer to being "standards" compliant - int32 is portable if 
you need an int with a specific range. _s32 and it's ilk should not be relied 
upon for long term support or cross platform. Hiding behind hal_yadda 
typedefs sets a bad example, ignores good advice from the likes of Stroup and 
Torvalds, and raises questions.

[Aram] If you really want to use a 5,000,000 count encoder, you'll probably 
need to increase the size of the count registers (not difficult on a Mesa 
card) and switch to using 64 bit int in the driver - You'll still suffer from 
limited range and errors from int<->float<->double<->float<->int conversions.

---
Paul.




P.S. Tom - Can't send you any emails as sbcglobal and any other part of the 
AT&T network bounces legit messages.

-
Check out the new SourceForge.net Marketplace.
It's the best place to buy or sell services for
just about anything Open Source.
http://sourceforge.net/services/buy/index.php
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Putting Encoder on AC servomotor???

[paul_c]

On Thursday 05 June 2008, Stephen Wille Padnos wrote:
> We've discussed various methods of handling larger data types, which
> this would require.

I don't recall ever seeing any discusssion about the usage of "data types"..

> The good news is that once we have that, we can use 
> doubles anywhere.

Please tell us why doubles can not be used throughout.

> >  Do any of the other encoder interfaces use the same 32-bit integer value?
> Yes, since a hal_s32 is a 32-bit int ;)

So why not use a standard int rather than hiding behind some @$&! naming 
scheme ?



-
Check out the new SourceForge.net Marketplace.
It's the best place to buy or sell services for
just about anything Open Source.
http://sourceforge.net/services/buy/index.php
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Still Crashing

[paul_c]

On Sunday 25 May 2008, Chris Radek wrote:
> Yes, congratulations, you found a network that blocks ssh, one of the
> most commonly used protocols on the internet.  If you try, you can
> find one that blocks all web access on port 80 too.

*cough* Corporate networks *cough* secure systems *cough*. And yes, I have 
worked in places where traffic on port 80 is restricted.

> If you did, would you complain about all the websites in the world being run
> incompetently?

Don't be silly Rodney - I take the sysadmin out back and give him a quiet 
duffing.

Anyways, I now have a box set up in the lab, negotiated access to the outside 
world, and can at least pull "stuff" from cvs.linuxcnc.org... Why the *** 
pserver can't be used for anon access, I don't suppose we will get an answer. 
But, hey, ssh & cvs works, so where do I upload my public key to ?


P.S. Build goes tits up in several places, for starters:

hal/drivers/pci_8255.c:45: error: syntax error before numeric constant
hal/drivers/mesa_5i2x/hal_5i2x.h:42: warning: no semicolon at end of struct or 
union
hal/drivers/mesa_5i2x/hal_5i2x.c:327: error: request for member `base' in 
something not a structure or union
Preprocessing m7i43_hm2.comp

RuntimeError: maximum recursion limit exceeded



-
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Still Crashing

[paul_c]

On Sunday 25 May 2008, Chris Radek wrote:
> > > Can you ssh to anywhere else from this machine?
> > Yes
> No you can't.  Sourceforge disallows it.  You must not have actually
> tried.

Oh shucks - Must have been another machine I was logged in to yesterday. With 
ten or more consoles open at any one time, sometimes type a command in on the 
wrong machine.

> > Log on to Sourceforge and you poke around and see what works (or
> > not) first hand.
>
> Yes I did.  Then I quickly determined that they don't allow any
> outgoing ssh.  Then I found it in their documentation.

So you found out first hand that not everyone has anon access to 
cvs.linuxcnc.org - I dare say a similar policy applies on other (secure) 
networks..


> > > Check your firewall log after the failed trace.
> For something relevant being blocked at the time you do the trace,
> obviously!

grep -m 1 "206.222.212.218" firewall.log
16:33:07 IN=ppp0 OUT=eth0 SRC=206.222.212.218 DST=192.168.0.114 LEN=96 
TOS=0x00 PREC=0x00 TTL=46 ID=21651 PROTO=ICMP TYPE=11 CODE=0 
[SRC=192.168.0.114 DST=206.222.212.221 LEN=68 TOS=0x00 PREC=0x00 TTL=1 
ID=12525 PROTO=UDP SPT=33653 DPT=87 LEN=48 ]

About a dozen lines all pretty much the same. I don't "do" firewalls, nor do I 
pretend to understand iptables beyond basic port redirection.



-
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Still Crashing

[paul_c]

On Sunday 25 May 2008, paul_c wrote:
> The external IP address should be in the 84.9.xx.xx block, but I'm not on a
> static IP, so it will change.

Update, 87.74.xx.xx today - Might be something else after lunch.

-
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Still Crashing

[paul_c]

On Saturday 24 May 2008, Chris Radek wrote:
> > Using a remote server that I have a shell account on...
> > ssh: connect to host cvs.linuxcnc.org port 22: Connection refused
>
> Can you ssh to anywhere else from this machine?

Yes, and I can also run cvs to do anon checkouts from other systems. Log on to 
Sourceforge and you poke around and see what works (or not) first hand.

> When you try again let me know the date and time and IP.  I will
> check the logs.

The external IP address should be in the 84.9.xx.xx block, but I'm not on a 
static IP, so it will change.

> Can you traceroute more than 20 hops to anywhere else from this
> machine?  blocked UDP ports or misconfigured TTL on one of your
> machines could cause this.

First, find a server that is on some dodgy network... Picking one at random 
from the pile of fake medz... fjberiaorpodkam.com completes in 22 hops. As 
for ports - Only the essential ones are enabled for outgoing connections (22 
& 123?), ones like 25, 53, 139, 445 and anything above 1024 get dropped.

> Check your firewall log after the failed trace.

And look for what ? - The firewall logs are pretty big files, and rarely get 
looked at except to see if/when there is a problem with authentication with 
my ISP.



-
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Still Crashing

[paul_c]

On Friday 23 May 2008, John Kasunich wrote:
> You seem to be the only one who is unable to connect to the server for
> anonymous read-only access.  Of course no key is required for that.

Using a remote server that I have a shell account on...

-bash-3.00$ ssh-keygen -t dsa
Generating public/private dsa key pair.

-bash-3.00$ export CVS_RSH=ssh
-bash-3.00$ cvs -d:ext:[EMAIL PROTECTED]:/cvs co emc2
ssh: connect to host cvs.linuxcnc.org port 22: Connection refused
cvs [checkout aborted]: end of file from server (consult above messages if 
any)
-bash-3.00$

Using `cvs -d:pserver:anonymous` works fine for anon checkouts from 
Sourceforge and other repositories.

> You haven't posted your traceroute results, so there is little we can do
> to help solve whatever internet problem exists between you and the server.

~$ traceroute -n cvs.linuxcnc.org
traceroute to cvs.linuxcnc.org (206.222.212.221), 30 hops max, 40 byte packets
 1  192.168.0.1  2.702 ms  2.769 ms  2.857 ms
 2  83.146.18.123  12.269 ms  13.265 ms  15.249 ms
 3  83.146.22.30  19.222 ms  22.157 ms  22.219 ms
 4  83.146.19.82  24.120 ms  24.180 ms  27.036 ms
 5  212.187.151.5  28.071 ms  29.085 ms  29.145 ms
 6  4.68.117.94  32.037 ms  29.296 ms  31.162 ms
 7  4.69.132.122  37.110 ms  27.760 ms  27.043 ms
 8  4.69.132.146  26.945 ms  22.978 ms  22.878 ms
 9  4.69.132.118  22.860 ms  34.854 ms  35.766 ms
10  4.69.137.74  97.936 ms 4.69.137.78  96.982 ms 4.69.137.66  95.887 ms
11  4.69.134.66  98.685 ms 4.69.134.70  103.945 ms 4.69.134.74  100.734 ms
12  4.69.134.93  106.869 ms 4.69.134.81  101.543 ms 4.69.134.85  101.573 ms
13  4.69.132.65  115.446 ms  113.901 ms  113.909 ms
14  4.69.134.58  112.881 ms  115.821 ms  116.855 ms
15  4.69.132.61  133.907 ms  141.757 ms  142.722 ms
16  4.68.107.162  130.690 ms 4.68.107.34  130.954 ms 4.68.107.98  130.940 ms
17  4.79.74.14  137.921 ms  137.928 ms  139.874 ms
18  151.213.28.183  139.882 ms  141.862 ms  142.874 ms
19  151.213.28.184  143.869 ms  144.814 ms  143.813 ms
20  216.170.52.114  139.414 ms  140.825 ms  139.873 ms
21  * * *
22  * * *
23  * * *
24  * * *
25  * * *
26  * * *
27  * * *
28  * * *
29  * * *
30  * * *

Note: This is from my server, not the remote box mentioned above - That 
account has limited commands available, traceroute and ping are not on the 
list.


-
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Who is Paul_c

[paul_c]

On Friday 23 May 2008, Stephen Wille Padnos wrote:
> 1) He was instrumental in getting EMC to where it is today.

Perhaps, but then there are other people that deserve credit. Stallman, 
Torvalds, Mantegazza, Yodakin, not forgetting Proctor, Shackleford, Kramer, 
and countless others (oh, and Bill Gates).

> 2) Even through all the trolling, there are some grains of useful
> information.  (kind of like the idea that legends have some basis in fact)

Whoa... You guys need to ease up on the coffee or what ever meds you are 
on ;-)   Even although we are susposedly all speaking English, I sometimes 
wonder if them 'mericuns threw out all the best bits when they declared 
independance.

> It would be nice if he'd start contributing (again) instead of trolling,
> but that's his choice.

I say "958 compile time errors", you say "we only see a few".What should be 
happening is a close look at what the warnings say, followed by some 
workflow e.g.
 "[MOD] Duplicate symbol yadda, yadda"...
Hrmmm. those symbols are exported by modules X, Y, and P.
But we bypass the symbol tables that the kernel manages by feeding 
pointers in to Wibble.
So don't need to export those symbols, which means a few bytes are saved.
OK, that's reduced the count, how can we clean up the rest

Not everyone runs the same flavour of Linux, likewise, not everyone speaks the 
same flavour of English (hell, some of the guys round here need subtitles).


-
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Still Crashing

[paul_c]

On Friday 23 May 2008, Alex Joni wrote:
> If you care about development or development related issues use the
> emc-developers mailing list.

 I have seen some of these guys, even in the same room, huddled over a 
keyboard or engaged in some trivial discussion whilst a guest speaker is 
giving a presentation.. In cyberspace, nothing has changed...

 At least here, the poor user gets a chance to see some discussion that 
*might* improve the application that (s)he may use.



-
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Still Crashing

[paul_c]

On Wednesday 21 May 2008, Ray Henry wrote:
> I don't have ordinary traceroute installed but there is a "Network
> Tools" under the System -> Administration that lets me run quite a
> number of these sorts of tools.

Switching to a console and running diagnostics from a command line is much 
easier... Trace & stuff sent off list.

---

 Paul.

-
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Still Crashing

[paul_c]

On Thursday 22 May 2008, Stephen Wille Padnos wrote:
> I don't have the full list at the moment, and it's unnecessary.

You mean to say you have no idea who has access - As a supposed project admin 
& "board member", that is a very sad state of affairs... But thanks to Chris, 
we now know. Only 24 out of 64 Sourceforge registered names.

> There are several others who could also serve, depending on the
> subsystem you're changing.  You should ask them first, or ask on this
> list for people who are interested in reviewing a particular patch.

As if anything that has gone before has gone through a review process or 
comments & suggestions acted on. i.e.:
*) Configure goes tits up with gettext 0.15
*) Giving explicit "apt-get install" instructions in a configure error is poor 
policy.
*) Hard coding kernel dependencies and/or non-standard paths is bad.
*) Reimplementing common libraries is poor policy in the long term.
*) Ripping off $random chunks of code just stores up problems for the future.
*) Sticking fingers up at published and accepted standards...

Other issues yet to be raised or resolved:
*) Regular copyright & licencing audits.
*) General bit rot.
*) Distribution of compiled binaries, temporary & hidden files, and general 
cruft in "official" tarballs.
*) Excessive memory usage requiring raising of system limits - A result of 
sloppy coding and/or serious bugs ?
*) Bad practices resulting in fragile or unstable systems.

Most users are presented with a flashy GUI and never look behind the scenes, 
and those that do will either leave well alone or say nothing. Throwing in a 
few bells & whistles and covering it all up with reams of documentation (of 
varying quality or relevance) solves nothing, nor does a FU and attempting to 
ignore or stifle a debate.

> In no particular order, and without their endorsement, here are some
> names to choose from, and the subsystems I think they may be most
> interested in:
> Ken Lerman for interpreter changes

Most of the interpreter changes appear to be reasonably well thought out, 
although I wonder who's "standard" has been used as a reference.

> Ray Henry for tkemc or mini UI changes.

The Tcl/Tk interfaces are pretty stable and haven't changed for a long time. A 
couple of bugs still remain in the underlying code..

> Jon Elson would probably be the best person for modifications to drivers
> that support his hardware.

His hardware/driver(s), NMP - A few SMP bugs in there that are easily fixed, 
and some minor changes would reduce the resulting binary by 25-30%.

> Seb Kuzminsky for mesa driver changes.

Not seen that yet (have some cards though) - The code will be posted for 
review before it is released ?

> Chris Morley for classicladder.

Patches for CL are sent to Marc Le Douarain where they benefit a wider 
audience.

> You also have the option of providing your public SSH key to Chris, so
> you can get developer access yourself.

If anonymous access doesn't work, please explain how sending a "key" will fix 
it.

---

 Paul.

-
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Still Crashing

[paul_c]

On Wednesday 21 May 2008, Stephen Wille Padnos wrote:
>  email them to a developer for review and possible inclusion.  

Please provide a full list of "developers" who have access to the repository.

---

 Paul.



-
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Still Crashing

[paul_c]

On Wednesday 21 May 2008, John Kasunich wrote:
> Pinging it _with_ the http is NOT going to work.  You ping hosts, not
> services.  "http://"; is a URL, not a hostname.  Paul has been around
> long enough to be well aware of that.

> Pinging _with_ the http:// is doomed to fail, which I'm sure Paul
> already knew.

It was a keyboard error when typing a reply - So ping, traceroute, and cvs all 
fail. The relevant ports are all open my end, ssh is set up and fully 
functional..

--

 Paul.

-
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Still Crashing

[paul_c]

On Wednesday 21 May 2008, John Kasunich wrote:
> EVERYONE has anonymous read access:
> http://cvs.linuxcnc.org/cvs/

`ping http://cvs.linuxcnc.org` - Fail
`traceroute cvs.linuxcnc.org` - Stops somewhere in the alltel.net system.
`cvs -z5 -d:ext:[EMAIL PROTECTED]:/cvs co emc2` - Connection times out.

 Your assertion is incorrect, likewise, I was in error in claiming you had 
labelled me an ex-developer - It was some minor nobody on IRC a year or two 
ago..

---

 Paul.



-
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Still Crashing

[paul_c]

To add to the history lesson...

On Saturday 17 May 2008, Ray Henry wrote:
> The BDI evolved out of an early discussion that we came to call a
> Brain Dead Install when the EMC was still a part of NIST.  It's
> difficult to imagine now how much trouble it was to get a running
> version of the EMC software back then.  Paul Corner took up the
> challenge and produced what he officially named the BDI. 

The early releases were based on Red Hat (6.x used the RTLinux patch). Later 
releases switched to using RTAI and Debian as a base. At one point, a LiveCD 
was produced, but it proved to be tiresome to debug and update as well as 
being an inflexible medium for those wanting to perform a minimal install.

Aside from a couple of releases, effort was taken to ensure the tools and 
libraries were included so that EMC could be compiled without having to 
download extra packages. It was fortunate that EMC did not require any 
obscure or specific library versions aside from a suitable kernel & realtime 
support.

 A seperate forum was set up to provide help for users WRT general Linux usage 
as it was felt that poor advice was being given by one or two - Something 
that has continued, and now spills over to other lists/forums. Despite that, 
space has always been allocated for emc2 (since it started), and when it has 
been demonstrated to be a viable alternative, it would be included.

> When we -- fondly called EMC2 -- and He -- fondly called tuxcnc --
> parted ways,

It was JMK & sidekicks that moved CVS away from Sourceforge without discussion 
(and overstepped his mandate) - I no longer have access to the repository 
(not even anonymous), and JMK decided a long while ago that I wouldn't be 
contributing any further... Apart from the usual suspects, just who has write 
access to CVS amongst the registered "developers" listed on Sourceforge ?


Re: tuxcnc - What I do outside the remit of Sourceforge is of little concern 
here, and it is not a name I would wish to associate with what you might 
call "emc2". IF I release any code, be assured, it will have little in common 
with emc2 apart from a requirement for a realtime kernel.


emc2 - 958 compile time warnings/errors and counting.

---

Paul.

-
This SF.net email is sponsored by: Microsoft 
Defy all challenges. Microsoft(R) Visual Studio 2008. 
http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] EMC with Galil Ethernet motion controller

[paul_c]

On Saturday 22 March 2008 14:31, Thomas J Powderly wrote:
> EMC is a controller
> DMC40xx is a controller

and emc2 is ??

> there was code, a long time ago that allowed a similar controller
> ( PMAC) to control a machine, with EMC as only the 'user interface'
> but that code would need major modification to work with Galil
> and major modification to work with EMC's popular 'front ends'

I did a "proof of concept" project a few years back with a Galil DMC17xx card 
and the EMC code - A couple of files needed reworking to convert canonical 
commands to DMC instructions and handle returning data. As a demonstration 
project, it worked within expectations, but some of the more advanced DMC 
features required an overhaul of code elsewhere in the tree.

 Might try to borrow the hardware again later on in the year and update the 
driver for the latest 2.6.xx kernel (if I can find the original code)...


Regards, Paul.


-
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

[Emc-users] EMC history

[paul_c]

On Wednesday 13 February 2008 15:40, Stephen Wille Padnos wrote:
> >I'm starting to suspect that EMC is a project that started out, not to
> > emulate the commercial equivalents, but built bit by bit to do various
> > things on the cheap, I'm starting to suspect that EMC is not a realtime
> > machine control system, but rather an offline (non realtime) simulator
> > that relies on assumption (I sent signals to move X 1.01 mm, therefore I
> > shall assume it has moved 1.01 mm)
>
> Actually, you have it backwards.  Many of the commercial controls are
> based on EMC2 code.  EMC was originally developed at NIST and was public
> domain.  It was part of a standardization project for machining
> languages, and was written as the reference implementation of the
> newly-developed RS274NGC standard.

Really 

http://www.isd.mel.nist.gov/personnel/kramer/pubs/RS274NGC_3.web/RS274NGC_3TOC.html

 Section 1.2.2
"RS274 is a programming language for numerically controlled (NC) machine 
tools, which has been used for many years. The most recent standard version 
of RS274 is RS274-D, which was completed in 1979."

 Section 1.2.3
"The NGC architecture has many independent parts, one of which is a 
specification for the RS274/NGC language, a numerical control code language 
for machining and turning centers. The specification was originally given in 
an August 24, 1992 report "RS274/NGC for the LOW END CONTROLLER - First 
Draft" [Allen-Bradley] prepared by the Allen-Bradley company."

Section 1.2.4

"In 1995 the EMC project collaborated with several industrial partners in an 
open-architecture machine tool controller project known as VGER (a name, not 
an acronym). This project retrofitted an SNK 5-axis machining center with an 
open architecture controller. NIST provided the RS274 interpreter for this 
project [Kramer4]. It was intended to be able to interpret some existing 
programs for the SNK machine which were written for its former Fanuc 
controller [Fanuc]. Thus, the RS274/VGER Interpreter took Fanuc flavored 
RS274/NGC code as input."

>From that, and other references (both in assorted documentation and original 
source code), I would conclude that the EMC interpreter does NOT conform to 
any ratified "stansard" - It *may* be based on documented behaviour of a 
common (in 1990-4) control system that *might* be compliant in part with a 
recognised "standard".

 The current "extended" interpreter certainly does NOT conform to any 
published standard (i.e. as issued by DIN, ISO, EIA, or any other official 
body) - A subset of instructions may be compliant, but that is all.

As for the claim that "Many of the commercial controls are based on EMC2 
code", that is stretching things a bit - Mach1/2/3/4 may have it's roots in 
the original EMC interpreter, and there *may* be a few other hobby level 
controls out there that appear to be similar.

> Although there is a simulator mode (something you won't find on any
> commercial control

You really should look at the current offerings from the likes of Heidenhain 
and Seimens - I'm sure even the top end Fanuc controls would offer simulation 
and/or preview modes.

> - why bother, nobody would buy one just to stick it 
> on their desk), EMC2 is very much realtime - to the limits of the PC
> it's installed on.

Parts of the core code run in a realtime environment - For example, trajectory 
planning, PID, and IO. Parsing of assorted files is non-realtime, done in 
user space (unless someone has been screwing around with the code).

> It is used on machines from tabletop lathes to 20-ton machining centers.

I have EMC running on a desktop mill, but not emc2.

> There is no "hobby-class" controller that uses feedback.  None.

I fear you will lose a bet - Last I heard Mach3 can use feedback with a 
suitable IO card.

> The least  expensive commercial controller I know of with feedback is in the
> $5000 range.

With the value of the US dollar being so low, $5K would be close. Certainly, 
there are a number of commercial controls available on the market under $10K.

> Heh - "the emc coder".  EMC was developed by a few PhDs at NIST, then
> added to by about 20-30 developers over the last 10 years.

Odd that... There are over sixty names listed on the Sourceforge site. Perhaps 
you would be so kind as to tell us who currently has access to the repository 
(including anyone not listed at Sourceforge).



-
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Running on a G4 MAc (new world)

[paul_c]

Hi Jason

On Monday 11 February 2008 07:45, Jason Cox wrote:
> Paul,
>   I am about ready to start on the mac port :)
> I am trying to decide onwhich way to go.
>
>   1. install xcode and try to get rtai to work on a 10.4 10.5 kernel. I
> have done a bit of digging into this and havent had much luck in finding
> anyone who has don it. downfall as i dont know anything about howto
> program or run in OS X. realy i am a n00b here

RTAI over the OS X kernel will not work, period !

 The OS X system, whilst similar in many respects to Linux/GNU (it is, I 
believe, a BSD derived build), Linux kernel code will not work.

>   2.  install ubuntu 6.06 as that is the lattest for a PPC class cpu. i
> think i still remomber how to manualy do a patch from the rtlinux ad
> NIST days. at least i know linux :).

If you *really* want to go with Linux, might I suggest going for Debian - You 
can always upgrade to Lenny (or Sid) if you want 
testing/unstable/experimental packages without waiting for them to appear in 
ubuntu. To be honest though, if you already have a distro installed and 
running, why change.

>   3. unknown to me at this stage.
>
> either way, it should be this week that i start to give it a go

Keep the original OS X install, then you have a chance to compile and test for 
both environments (and we get to see were things go tits up).


Regards, Paul.

-
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Running on a G4 MAc (new world)

[paul_c]

Hi Jason

On Monday 11 February 2008 10:19, Jason Cox wrote:
> Ok i will move over to the Dev list.

Keep it here. Dev list is ...

> Alex, was this a "roll you own" linux distro that you were using? What i
> have is a G4 (gray case) UP with 1G ram so it should do ok. I seem to be
> seeing a lot about slow compiling, is it realy bad or just takes time? I
> could setup a cross compiler under my gentoo box, but then libs will be
> the issue.

The only thing that should be time consuming is a kernel build - This can be 
built with a cross compiler without any side effects. No user space libraries 
are used in kernel space (if they are, someone needs a kick up the..).

 If you are comfortable with Gentoo, install it on a second hard drive on the 
Mac - Just because these guys are developing for ubuntu, it doesn't mean you 
have to.

> On Mon, 2008-02-11 at 10:33 +0200, Alex Joni wrote:
> > just some snippets from the RTAI mailing list:
> > (Thread called "[Rtai] rtai on a ppc405 (Xilinx XUP Boad)" from the 1st
> > February 2008)

ppc405 != G4

> > "> - Is it possible to use a newer Kernel than 2.6.14, are there any more
> > patches?

Since 2.6.18 (as I recall), the powerpc and ppc architectures have been merged 
in to one (as have i386 & x86_64 in 2.6.24). The latest ipipe patch from the 
ADEOS team works reasonably well with 2.6.23 but I don't think it is in use 
by RTAI yet.

Before you ask, yes, I do have kernels for i386, x86_64, AND PowerPC installed 
& running - All patched with the latest ipipe patches (plus a couple more). 
The PowerPC kernel was also cross-compiled on an AMD64 box without any 
problem.



-
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Running on a G4 MAc (new world)

[paul_c]

Hi Jason

On Wednesday 06 February 2008 02:59, Jason Cox wrote:
> As i started this thread, i should put my hand up, but there are a few
> problems at the moment. I have just started uni (after 10 years) and I
> would need a guide for setting up a development environment on OSX.

All you need is xcode which should be on your installation CD (or DVD), if 
not, you can easily download it from the Apple Developer's site at 
http://developer.apple.com/ - You'll probably need to create a user account 
first..

 The xcode package provides all the tools and libraries needed to get started, 
even an IDE if you don't like the command line.


Regards, Paul.

-
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Running on a G4 MAc (new world)

[paul_c]

On Sunday 03 February 2008 05:35, Jason Cox wrote:
> as luck would have it, I was given recently an old G4 Mac. is it posible
> to run EMC on a ppc32?

After proding a few people I know, some feedback.. The core library, libnml 
compiles and appears to be functioning when linked to the test suite. The 
Tcl/Tk bindings are pretty trivial, and as Tcl is part of a native OSX 
install, shouldn't present any problems (nor should the ncurses interface).

 Gtk frontends will need supporting libraries from Fink (or elsewhere), so 
they can be omitted for the timebeing..

This all allows tkemc or similar to run as a "native" OSX application, either 
G3/4 or Intel based. Looking at the realtime code, it is unlikely it would 
run within the time constraints required, so it would have to be offloaded to 
either an external Linux box or an internal "intelligent" I/O card. Some 
experimentation would be required to determine how much of the lower level 
code would run within the OSX IO Framework and what needs to move outboard. 
There are a few DSP cards available that could provide the realtime 
environment along with I/O capabilities for high speed stepper and/or servo 
systems.


Running a Linux system on a G3/4 is simple enough - The majority of mainstream 
distros provide a PowerPC port, although some of the packages can lag behind 
in stability or version when compared to i386 builds. RTAI support is not an 
issue - Several patches are available for a number of kernels (2.4 and 2.6). 
Failing that, grab a copy of the ipipe development tree.

Looking at the emc2 code base (under protest), it *might* be possible to build 
a native OSX simulator... But...
Building a PowerPC version running on top of RTAI is also going to be 
difficult for similar reasons...

It looks like there is sufficient interest in a native OSX application, and 
given a suitable I/O card, a practical proposition - Who else is interested ?


Regards, Paul.

-
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Running on a G4 MAc (new world)

[paul_c]

On Sunday 03 February 2008 20:31, Jon Elson wrote:
> > I was given recently an old G4 Mac. is it posible to run EMC on a ppc32?
> Does it have a parallel port?  If not, you are in pretty serious
> trouble.

The right question should be: Does it have any expansion slots that would 
allow an IO card to be plugged in ?

> Not to mention the problems of getting an rtai compatible version of
> RT-Linux running on the ppc,

RTAI has patches for various 2.4.xx and 2.6.xx kernels - Not all aimed at the 
G4 processor. Best place to ask is over on the RTAI mailing list as to which 
PowerPC processors are currently supported (and pray that it isn't 2.4.xx).

> and then getting all of EMC to compile and run properly.  It is SUPPOSED
> to, but.

Take a look at the README - The "mission statement":


2)  Reduce complexity by eliminating support for platforms
other than Linux on x86, with either RTAI or RTLinux for
realtime. (DONE)


Note: x86 only !
"or RTLinux" - nuff said.

-
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Dual Linux EMC Installation on Dual HDDs

[paul_c]

On Monday 28 January 2008 21:39, John Kasunich wrote:
> I don't know what system Paul is  getting 1000+ warnings or errors on.

Has it crossed your mind to ask ??

FYI:
  Debian: stable.
  Kernel: 2.6.23.11
  RTAI: 3.6
  Hardware: Assorted - single & dual processor, mixture of speeds.

> Breezy with stock non-RT kernel:
A snapshot of Debian's testing branch with a 2.6.12(?) kernel.

> Breezy with RT kernel:
Another testing snapshot, but what kernel version..

> BDI-4.51
Debian Sarge+backports+security-updates with a 2.6.16 series kernel.

>34 warnings during MODPOST

Attempting to suppress modpost warnings is a fragile hack at best and one that 
needs to be continually updated. But as long as it works on _your_ install...

>4 warnings related to man pages

Mainly related to "bogus files" - Hard coding kernel versions and paths in to 
executables is bad enough, so compound it with documentation.

> The compile farm webpages are intended mostly to help developers,

Just who are these "developers" ?
 Certainly not all of the 63 names listed on Sourceforge..

> The last system is Dapper, with a RT kernel

Yet another Debian snapshot with an unspecified kernel version. A pattern is 
begining to develop here..

> Errors in EMC2 builds are rare

Riiight Install gettext-0.15 and compile against a 2.4.xx kernel.


-
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Dual Linux EMC Installation on Dual HDDs

[paul_c]

On Tuesday 22 January 2008 03:35, Rafael Skodlar wrote:
> pmark wrote:
> > Good Day - Have been working on the trying to install Fresh Dual
> > Installation w/2HDD's one for Ubuntu EMC2 & one for Linux BDI EMC and the
> > being able to choose the OS on BOOT. Would like to have both EMC's
> > available in order to compliment each other.

You should be able to compile and run both EMC and emc2 on the same distro, 
however - emc2 does not install a number of files in standard locations (as 
recommended in the LFSH), and if you use that modified Ubuntu disk, some 
files need to be moved/copied. Oh, and you will need to download and install 
all the prerequisite tools and development libraries.

 The EMC code also installs much of it's runtime files in non-standard 
locations ( /usr/local/emc is typical), but this is for historical reasons.

> > expect?? My HDD's are set Master/Primary & Slave/Secondary. Must have
> > tried just about every type/way of installing, that is except the
> > right way. Cannot get startup screen to show both OS. I'm Missing
> > something???
>
> Not much likely but enough to ask for help here I guess. I'm not an
> active EMC or BDI EMC user, just learning for now. If I remember
> correctly, BDI EMC was redhat based so the process is a bit different.

The early versions (BDI-2.xx & BDI-tng) were based on Red Hat, all later 
builds use Debian. The base distro isn't that important as this emc2 stuff 
*should* compile on _any_ version of Linux as long as a suitable patched 
realtime kernel is available... But...

> Assuming your ubuntu is on drive /dev/hda and BDI on hdb I would do
> something like:

> kernel/boot/vmlinuz-2.4.xx-generic root=/dev/hdb1 ro single

Why boot to single user mode ?

> The exact kernel and initrd needs to be checked from BDI's drive and
> include it in this menu.

Not just the BDI kernel version, but all of them..

Some notes on kernel versions & older distros:

 2.2.xx kernels are no longer supported, nor is RTLinux, so forget trying to 
build emc2 on RH 6.x installs.

 2.4.xx builds are broken, and gcc-2.95 (or egcs) won't work - So if you are 
trying to assemble an embedded system (2.4.xx still boots way faster than 
2.6.xx), you have some fixing to do.

 2.6.xx builds - 1004 compile time warnings & errors.. And counting...


---
Paul.

-
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] (no subject)

[paul_c]

Hi Ray

On Thursday 29 November 2007 17:53, Ray Henry wrote:
> I find this fascinating.  Long years ago after some preliminary work by
> Will at NIST, I authored a knoppix-emc CD and worked with Paul while
> developing the Morphix (take the red pill) version that became the
> debian BDI.  A couple of things come to mind.

Minor correction - The Morphix base produced the BDI-live and laid the 
foundations for later work (Morphix also provided the original framework for 
Mr Shuttleworth's LiveCD, but that is another story)..

> One of the concerns for me, being at the far end of many miles of very
> old copper is that updating and maintaining a large install like the
> ordinary Ubuntu is a time and resource consuming thing.  At best I get
> about 3k dribbling into the basement.

One of the reasons that you always insisted on the development tools being 
available on the disk - How long does it take to download compilers, 
libraries, and assorted tools at 3K ?

> I think our current raft of developers/leaders were wise to choose
> Ubuntu.

You sure about that - Ubuntu is little more than a snapshot of Debian's 
unstable branch (maybe with a few bug fixes)... If you really want to live on 
the edge, try tracking Sid and/or Experimental.

> I didn't think so at the time but old dogs can occasionally 
> find new fire hydrants.

Now that conjures up a few interesting metaphors.. Canines, lamp posts.. ;-/

> Ubuntu has most everything to satisify folk that are in process of migrating
> from the PC crowd.

As does SuSE, Red Hat, and countless other distros. Last time I tried SuSE 
(10.something), it looked pretty solid and did a better job of configuring 
the system than ubuntu.

> It certainly fits with the release often line of thought and provides a nice
> web path to keeping current -- if keeping current is your thing.

I'll bet you quietly grumbled each time something broke, or you found 
upgrading X pulled in Y,Z, A through to F

> Now to the heart of my initial thoughts regarding reconstructor.
> Cleaning out unnecessary stuff after an install can be really time
> consuming.  Building a bootable CD can be really time consuming.

*Was* time consuming - I build bootable images every day and each one takes 
between four and eighteen minutes. (scripted job taking a partial package 
list). Testing can take considerably longer, but modern hardware & qemu 
speeds that up.

> A good CD building helper will reduce the size of the step required to get
> into the business of building these disk images.  If it works it can allow
> us to offer several levels of install along the embedded -> full blown
> continuum.

As you rightly pointed out, cleaning out the post-install cruft is time 
consuming, and installation from a LiveCD gives the end user one of two 
options - Install everything the CD builder included (hacks and all), or no 
install.

 The LiveCD provides a good way to present an application for demonstration, 
but when it comes to a vehicle for installing, it lacks flexability. The poor 
user also ends up with any number of hacks and/or redundant cruft.. Just some 
of the reasons why I moved away from the LiveCD concept as an installation 
medium several years ago.

 If you want an install CD with all the latest bleeding edge packages, or the 
option to install a bare bones system on a SSD, we can do that. Want a choice 
between a lightweight window manager without the multimedia bloat, or a full 
blown Gnome/KDE plus bells'n'whistles, sure. Or how about Samba, wine, and 
the usual assortment of compilers & tools..

It kinda sounds like you have come full circle...


Regards, Paul.

-
SF.Net email is sponsored by: The Future of Linux Business White Paper
from Novell.  From the desktop to the data center, Linux is going
mainstream.  Let it simplify your IT future.
http://altfarm.mediaplex.com/ad/ck/8857-50307-18918-4
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Is updating EMC@ impossible without internet?

[paul_c]

On Wednesday 14 November 2007 18:24, Jon Elson wrote:
> xtra209 wrote:
> > Updating emc2 seems impossible without an internet connection on the
> > machine. Is this true?

True - Either to download packages, or CD images.

> > So far the only way I have successfully updated emc2 is to wait for the
> > Live CD and reinstall. Am I missing something or is this just the way it
> > is?

Reinstalling is the Microsoft way. In the world of Linux (and in particular, 
with Debian based distros) just wrong.

> > And, can a future Live CD include a gcc compiler?
> I'm pretty sure we could fit JUST the gcc compiler on the CD,

If you look on the unmodified Ubuntu CD, you *will* find gcc. What is missing 
from this modified emc2Live is most of the development tools one would 
normally expect. Indeed, in the early days of EMC & CDs, Ray was always 
insistant on tools & libraries being included.

> but to put all the delevoper files required to compile EMC2 on
> the CD definitely would not fit.  Maybe somebody could bundle
> all these up and put them on a CD.

I have the tools, scripts, and repository mirrors to do just such a task. 
Takes minutes to edit a package list and trigger a full build (which usually 
runs for 30 minutes or so).

> I think I actually know enough to do this one.  (Gulp, have I just
> volunteered?) 

debmirror & apt-ftparchive man pages will point you in the right direction. 
But DON'T look at the crazy way that emc2 repository tries to mimic a 
Debian/pool layout..

---


-
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >> http://get.splunk.com/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] EMC on DSL

[paul_c]

Hi Roland

On Saturday 01 September 2007 14:53, Roland Jollivet wrote:
> Quick question;
> Can Emc be loaded onto  a machine with DSL (Damn Small Linux) ?

DSL is Debian based - You will find both patched kernel packages and RTAI 
available from several unofficial repositories (if you need the uri, email 
me). Once the kernel/RTAI packages are installed, you can then have fun 
compiling emc2 (strongly suggest enabling run-in-place option).


Regards, Paul.

-
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >>  http://get.splunk.com/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Just a Thought

[paul_c]

Hi Ray

On Tuesday 14 August 2007 15:45, Ray Henry wrote:
> There was a heated discussion at that long ago FEST meeting in Ann Arbor
> when we created the EMC board.  It was related to it's (us) becoming a
> legal entity.

Outside of the USA, any "entity" would have little, if any power or status.

> It was my opinion, still is, that we need such an entity 
> so that we can accept and use contributions to the central benefit of
> the project -- contributions of code, cash, or whatever.

If you remember, part of the "boards" mandate was to act as an intermediatory 
to put interested parties in touch. e.g. Party A would offer hardware in 
exchange for suitable driver code.

 As for code contributions - Who decides what is good or bad ?

> I certainly respect the opinions of the others there that disagreed.  In
> fact as a condition of the creation of a board we voted that it would
> not become a legal entity and it would not accept nor distribute funds.

Setting up a legal entity would end up being a cash cow with little benefit 
for the majority - Who would audit the accounts, fund legal representation, 
or a whole other bunch of stuff...

> What those at the meeting did not know was that I had a couple thousand
> dollars in my pocket at that moment,

Likewise, there is still quite a considerable sum of dollars from the proceeds 
of BDI disks available - Some has already been used to purchase IO cards from 
the likes of Adlink, the remainder could be used for similar purposes.

> I wonder if it isn't time to rethink the legal entity.

If certain board members overstep their mandate, or fall back on a FU/O 
attitude to sidestep any issues/arguments, I doubt a legal entity would have 
any more credibility outside a gin house.


Regards, Paul.


-
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >>  http://get.splunk.com/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] emc2 and kernel 2.6.20

[paul_c]

On Friday 03 August 2007 19:46, Jeff Epler wrote:
> http://cvs.linuxcnc.org/cvs/emc2/src/rtapi/rtai_rtapi.c.diff?r1=1.34;r2=1.35


wget 
'http://cvs.linuxcnc.org/cvs/emc2/src/rtapi/rtai_rtapi.c.diff?r1=1.34;r2=1.35'
--20:40:43--  
http://cvs.linuxcnc.org/cvs/emc2/src/rtapi/rtai_rtapi.c.diff?r1=1.34;r2=1.35
   => `rtai_rtapi.c.diff?r1=1.34;r2=1.35'
Resolving cvs.linuxcnc.org... 206.222.212.221
Connecting to cvs.linuxcnc.org|206.222.212.221|:80... connected.
HTTP request sent, awaiting response... 503 Service Unavailable
20:43:00 ERROR 503: Service Unavailable.

Assume patch is:

-module_param(msg_level, int, S_IRUGO | S_IWUGO);
+module_param(msg_level, int, S_IRUGO);

Install fails (see attached). Original assertion that build of release 2.1.7 
goes tits up on 2.6.20 kernel holds true.


install -d -m 0755 -o root  /tmp/cage/lib/modules/2.6.20-11-686-rtai/rtai/emc2 \
		/tmp/cage/tmp/test/bin \
		/tmp/cage/tmp/test/etc/emc2
install -m 0644 -o root ../rtlib/*.ko /tmp/cage/lib/modules/2.6.20-11-686-rtai/rtai/emc2
install -m 4755 -o root ../bin/emc_module_helper /tmp/cage/tmp/test/bin
install -m 0644 -o root ../scripts/rtapi.conf /tmp/cage/tmp/test/etc/emc2
install -d -m 0755 -o root /tmp/cage/usr/share/locale/de/LC_MESSAGES
install -m 0644 -o root ../share/locale/de/LC_MESSAGES/rs274_err.mo /tmp/cage/usr/share/locale/de/LC_MESSAGES/rs274_err.mo
install -d -m 0755 -o root /tmp/cage/usr/share/locale/fr/LC_MESSAGES
install -m 0644 -o root ../share/locale/fr/LC_MESSAGES/rs274_err.mo /tmp/cage/usr/share/locale/fr/LC_MESSAGES/rs274_err.mo
install -d -m 0755 -o root /tmp/cage/usr/share/locale/it/LC_MESSAGES
install -m 0644 -o root ../share/locale/it/LC_MESSAGES/rs274_err.mo /tmp/cage/usr/share/locale/it/LC_MESSAGES/rs274_err.mo
install -d -m 0755 -o root /tmp/cage/etc/init.d /tmp/cage/lib/modules/2.6.20-11-686-rtai/rtai/emc2 \
		/tmp/cage/tmp/test/etc/emc2 /tmp/cage/tmp/test/bin \
		/tmp/cage/usr/local/etc/emc2/configs \
		/tmp/cage/tmp/test/lib /tmp/cage/tmp/test/include/emc2 \
		/tmp/cage/tmp/test/share/doc/emc2 /tmp/cage/tmp/test/share/emc/ncfiles \
		/tmp/cage/etc/X11/app-defaults /tmp/cage/tmp/test/share/emc/tcl/bin \
		/tmp/cage/tmp/test/share/emc/tcl/scripts \
		/tmp/cage/tmp/test/man/man1 \
		/tmp/cage/tmp/test/man/man3 \
		/tmp/cage/tmp/test/man/man9 \
		/tmp/cage/tmp/test/share/emc/tcl/msgs \
		/tmp/cage/usr/share/locale/de/LC_MESSAGES \
		/tmp/cage/usr/lib/python2.4/site-packages \
		/tmp/cage/usr/lib/python2.4/site-packages/rs274 \
		/tmp/cage/tmp/test/share/axis/images \
		/tmp/cage/tmp/test/share/axis/tcl		
install -m 0644 -o root ../docs/man/man1/*.1 /tmp/cage/tmp/test/man/man1
install -m 0644 -o root ../docs/man/man3/hal_add_funct_to_thread.3hal ../docs/man/man3/hal_bit_t.3hal ../docs/man/man3/hal_create_thread.3hal ../docs/man/man3/hal_del_funct_from_thread.3hal ../docs/man/man3/hal_exit.3hal ../docs/man/man3/hal_export_funct.3hal ../docs/man/man3/hal_float_t.3hal ../docs/man/man3/hal_get_lock.3hal ../docs/man/man3/hal_init.3hal ../docs/man/man3/hal_link.3hal ../docs/man/man3/hal_malloc.3hal ../docs/man/man3/hal_param_bit_new.3hal ../docs/man/man3/hal_param_bit_newf.3hal ../docs/man/man3/hal_param_float_new.3hal ../docs/man/man3/hal_param_float_newf.3hal ../docs/man/man3/hal_param_new.3hal ../docs/man/man3/hal_param_s32_new.3hal ../docs/man/man3/hal_param_s32_newf.3hal ../docs/man/man3/hal_param_u32_new.3hal ../docs/man/man3/hal_param_u32_newf.3hal ../docs/man/man3/hal_pin_bit_new.3hal ../docs/man/man3/hal_pin_bit_newf.3hal ../docs/man/man3/hal_pin_float_new.3hal ../docs/man/man3/hal_pin_float_newf.3hal ../docs/man/man3/hal_pin_new.3hal ../docs/man/man3/hal_pin_s32_new.3hal ../docs/man/man3/hal_pin_s32_newf.3hal ../docs/man/man3/hal_pin_u32_new.3hal ../docs/man/man3/hal_pin_u32_newf.3hal ../docs/man/man3/hal_ready.3hal ../docs/man/man3/hal_s32_t.3hal ../docs/man/man3/hal_set_constructor.3hal ../docs/man/man3/hal_set_lock.3hal ../docs/man/man3/hal_signal_delete.3hal ../docs/man/man3/hal_signal_new.3hal ../docs/man/man3/hal_start_threads.3hal ../docs/man/man3/hal_type_t.3hal ../docs/man/man3/hal_u32_t.3hal ../docs/man/man3/intro.3hal ../docs/man/man3/undocumented.3hal /tmp/cage/tmp/test/man/man3
install -m 0644 -o root ../docs/man/man3/intro.3rtapi ../docs/man/man3/rtapi_clock_set_period.3rtapi ../docs/man/man3/rtapi_delay.3rtapi ../docs/man/man3/rtapi_delay_max.3rtapi ../docs/man/man3/rtapi_exit.3rtapi ../docs/man/man3/rtapi_get_clocks.3rtapi ../docs/man/man3/rtapi_get_msg_level.3rtapi ../docs/man/man3/rtapi_get_time.3rtapi ../docs/man/man3/rtapi_inb.3rtapi ../docs/man/man3/rtapi_init.3rtapi ../docs/man/man3/rtapi_module_param.3rtapi ../docs/man/man3/rtapi_mutex.3rtapi ../docs/man/man3/rtapi_outb.3rtapi ../docs/man/man3/rtapi_print.3rtapi ../docs/man/man3/rtapi_prio.3rtapi ../docs/man/man3/rtapi_prio_highest.3rtapi ../docs/man/man3/rtapi_prio_lowest.3rtapi ../docs/man/man3/rtapi_prio_next_higher.3rtapi ../docs/man/man3/rtapi_prio_next_lower.3rtapi ../docs/man/man3/rtapi_region.3rtapi ../docs/m

Re: [Emc-users] emc2 and kernel 2.6.20

[paul_c]

Attached, configure & make logs for emc2-2.1.7 released on July 30th.




config.log.bz2
Description: BZip2 compressed data


make.log.bz2
Description: BZip2 compressed data
-
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >>  http://get.splunk.com/___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] SECONDED: Ray Henry

[paul_c]

On Wednesday 01 August 2007 14:47, John Kasunich wrote:
> John Kasunich wrote:
> > (off-list)
> Not.
> crap.

Thankyou John for nailing your colours to the mast for ... and all others to 
see. Somehow, I doubt if it was really an error.

 Despite that, my offer of help as a (co?)developer still stands - If you need 
portable code to handle tasks such as firmware, just ask.

 And just between you'n'me, don't be at all surprised if rtai_math disappears 
in the future - The RTAI team made some noises about dropping it a while 
back, and certain GCC flags are appearing in the kernel build to hinder 
things. We can migrate floating point to user space now, or wait to get 
slapped in the face with a wet kipper in a year's time. The choice is yours.



Mark: http://zathras.homelinux.org/stats/emc-stats/
 You are just one of 250 million americans (may be ?). A mere 2% of the world 
population. Not everyone uses the English language in the same way you 
might - For many, it is not even a second language.


George Bernard Shaw - "and the Americans haven't spoken it [English] for 
years."


-
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >>  http://get.splunk.com/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Modbus

[paul_c]

On Wednesday 01 August 2007 01:37, Matthew Glenn Shaver wrote:
> On Tue, 2007-07-31 at 15:36 -0700, Kirk Wallace wrote:
> > I just got a VFD with a Modbus port. Is there a plan on getting EMC to
> > use Modbus?

> Python has a good serial communication library, so it looks possible.

Matt - Before you go reinventing the wheel, look at python-modlib from 
lintouch.org.


-
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >>  http://get.splunk.com/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] SECONDED: Ray Henry

[paul_c]

On Wednesday 01 August 2007 02:43, Ray Henry wrote:
> Thank you both for the nomination and second but I must decline.

The experience, maturity, and respect that you brought to the table will be 
sorely missed. Hopefully, you will still be somewhere in the background to 
keep these guys in check when they go off the rails.

-
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >>  http://get.splunk.com/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Help sought to modify hostmotXX.vhd

[paul_c]

On Wednesday 01 August 2007 02:01, Jon Elson wrote:
> > The key you send to cradek to gain access to the CVS server is your
> > PUBLIC key! 
> Why Paul Corner is claiming this is such a hideous vulnerablility is just
> totally beyond me!

Jon - Get the attribution correct before posting. I NEVER MENTIONED SSH HACKS

 Repeat - I NEVER MENTIONED SSH HACKS. 


-
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >>  http://get.splunk.com/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] elections

[paul_c]

Stephen

On Tuesday 31 July 2007 01:24, Stephen Wille Padnos wrote:
> At the moment, I think we believe that the design of EMC2 is applicable in
> an industrial setting.

I don't even trust it to run a hobby spec mill such as a Sherline let alone 
some of the heavy iron I have. Running in an industrial environment is just 
asking for trouble...

> There may be  implementation issues (bugs) that make this untrue in some
> cases.  In either case - bugs in implementation or fundamental design
> problems 

Random use of volatile keyword produces bad (binary) code. See what Linus 
Torvalds has to say on the subject at lwn.net.

Conversion between float & double does not produce "rounding errors", it loses 
precision and range. Don't mix'n'match on a whim.

Hiding behind opaque data types - It obscures intent and if you insist on 
using _t suffix, it will lead to namespace pollution/conflicts. See 
Stroustrup, Torvalds, & Stallman for further details.

What mechanisms are in place to prevent a user from unloading any/all modules 
while a system is live & running ?

Can firmware be uploaded on the fly on a live system ?

What critical system variables can altered on a live system, and will it lead 
to a crash ?

What happened to Aunt Tillie - Did someone do the old dear in or was she 
crushed under a mountain of paper ?

---

Paul.


-
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >>  http://get.splunk.com/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Help sought to modify hostmotXX.vhd

[paul_c]

On Tuesday 31 July 2007 02:19, Ray Henry wrote:
> >  The move was made without consultation or due process.
>
> IRC meeting, April, 2006
>         "The Sourceforge CVS system has experienced troubles on
>         2006-03-30, and after a few days it still wasn't looking bright
>         for a fast fix. We decided, without a formal vote (but with 4
>         members present), to look for alternatives of CVS hosting, most
>         of the ones we found were either not-free or had serious
>         limitations, so we decided to move all CVS to cvs.linuxcnc.org

Thankyou for confirming (one) of my original assertions - No formal vote was 
taken, so due process was NOT followed. What other resolutions will be passed 
in a similar cavalier manner...

> As you well know, the SourceForge admin group was never the software
> controlling focus of the project 
> It was FredP that suggested we clean out the SourceForge
> admin list long years ago and the board made a small effort in that
> direction by putting forward a policy of yearly review.  However nothing
> has been done to clean up the now obsolete admin group.  

Thanks for making that point. The position of project admin has been rendered 
obsolete, so it is a worthless title.. Likewise, the list of "registered 
developers" is also useless.

> I guess that would be untrusted me

You personally, I would liken to an elder statesman such as ol' Ronald Reagan 
(before he lost his marbles) - Possibly the last President you guys had who 
had integrity and respect. Some of the other "board" members are more akin to 
Clinton, Bush, & Blair..

> 'cause whois reports my name addy and phone when you ask about linuxcnc.org.

And for cvs.linuxcnc.org ?

traceroute to cvs.linuxcnc.org (206.222.212.221), 30 hops max, 52 byte packets
 1  BabNet (192.168.0.1)  0.411 ms  0.289 ms  0.280 ms

 6  ae-0-54.bbr2.London2.Level3.net (4.68.117.98)  29.007 ms 
ae-0-56.bbr2.London2.Level3.net (4.68.117.162)  31.695 ms  28.705 ms
 7  ae-0-0.bbr2.NewYork1.Level3.net (64.159.1.42)  95.976 ms 
as-0-0.bbr1.NewYork1.Level3.net (4.68.128.106)  97.546 ms 
ae-0-0.bbr2.NewYork1.Level3.net (64.159.1.42)  96.675 ms
 8  ae-23-79.car3.NewYork1.Level3.net (4.68.16.69)  116.447 ms 
ae-13-69.car3.NewYork1.Level3.net (4.68.16.5)  99.677 ms 
ae-33-89.car3.NewYork1.Level3.net (4.68.16.133)  97.675 ms
 9  ggr2-p360.n54ny.ip.att.net (192.205.33.93)  100.664 ms  97.650 ms  100.955 
ms
10  tbr2.n54ny.ip.att.net (12.123.3.61)  136.088 ms  136.627 ms  137.955 ms
11  tbr2.wswdc.ip.att.net (12.122.10.54)  135.951 ms  137.771 ms  137.954 ms
12  tbr1.wswdc.ip.att.net (12.122.9.149)  135.990 ms  137.662 ms  135.973 ms
13  tbr1.sl9mo.ip.att.net (12.122.10.30)  137.985 ms  133.745 ms  135.968 ms
14  gar1.omhne.ip.att.net (12.122.82.17)  136.917 ms  134.512 ms  133.919 ms
15  12.125.72.54 (12.125.72.54)  145.162 ms  158.388 ms  143.988 ms
16  h146.33.213.151.ip.alltel.net (151.213.33.146)  160.017 ms  156.617 ms  
156.966 ms
17  h114.52.170.216.ip.alltel.net (216.170.52.114)  155.967 ms  156.666 ms  
156.965 ms
18  * * *

30  * * *

That goes tits up...

traceroute to www.linuxcnc.org (208.97.140.179), 30 hops max, 52 byte packets

 4  ge-0-0-0.dcr1.lnd.cw.net (195.10.12.1)  47.009 ms  30.612 ms  37.045 ms
 5  ge-5-0-0-dcr2.lnd.cw.net (195.2.9.90)  27.899 ms  30.602 ms  30.204 ms
 6  so-0-0-0-dcr2.nyk.cw.net (195.2.10.113)  100.783 ms  106.620 ms  99.956 ms
 7  * 66.79.152.145 (66.79.152.145)  109.914 ms  102.896 ms
 8  cr2.lax009.inappnet-9.cr2.nym009.internap.net (66.79.146.174)  175.088 ms  
174.783 ms  177.698 ms
 9  cr1-cr2.lax009.internap.net (66.79.146.205)  171.978 ms  176.534 ms  
174.920 ms
10  core1.lax.inappnet-12.cr1.lax009.internap.net (66.79.149.130)  171.960 ms  
171.577 ms  172.890 ms
11  border1.po2-bbnet2.ext1a.lax.pnap.net (216.52.255.95)  172.981 ms  177.554 
ms  171.929 ms
12  newdream-1.border1.ext1a.lax.pnap.net (216.52.220.78)  174.969 ms  172.518 
ms  172.970 ms
13  basic-ugly.dinero.dreamhost.com (208.97.140.179)  175.966 ms  169.553 ms  
174.944 ms

Totally different path that ends up at a major data centre - Hopefully with 
the support and backup that such an operation would entail.

> Ultimately the board is responsible for access to the repository and no one
> who has asked has been refused.

In the past, any project admin could hand out CVS write access, but not any 
more - Just one privilege that has been removed.

> We have had none of the down time and lost commit trouble since we have
> moved the repository.

None of the other projects I'm involved with have been impacted by downtime at 
SF - Helped in part by a migration to subversion sometime before SF had a 
hiccup. Had there been a consultation period, you would have had the benefit 
of a comparison..


Regards, Paul.


-
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configurat

Re: [Emc-users] Help sought to modify hostmotXX.vhd

[paul_c]

On Monday 30 July 2007 17:59, Mark Pictor wrote:
> For the record, Paul was given the opportunity to get access to the
> new repository -- the same way anyone else gets access.

No Mark - You miss the point entirely.

 The move was made without consultation or due process. As a result, the 
project admins no longer have any control. In addition, the repository is on 
an untrusted server reliant on a single person maintaining the hardware and 
system security.

 As I predicted when this move was foisted on us, RCSlib development has 
reverted to an in house project by NIST - Valuable contributions in both 
directions have been lost. Elsewhere, others have been making noises about an 
emc3 fork..

Sourceforge provide a web interface to manage my own ssh keys along with a 
number of other services should I wish to use them. So why the hell should I 
submit keys to an untrusted third party when I can use the SF web interface.

Bottom line - I do NOT trust the system that currently hosts the repository.

---

Paul.


-
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >>  http://get.splunk.com/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Help sought to modify hostmotXX.vhd

[paul_c]

On Wednesday 02 May 2007 17:07, Stephen Wille Padnos wrote:
> Paul - do you know of any way to use the kernel firmware infrastructure
> to load different firmware sets into multiple cards?

Stephen - If/when you get elected to the "board" and succeed in restoring the 
CVS repository at Sourceforge where we ALL can have access, I'll show you how 
to handle firmware. Even with bit flipping and assorted sanity checks, it 
would amount to less than 250 lines of code with the added bonus of being 
portable across architectures and relatively immune to on going changes to 
the PCI subsystem within the kernel.


Regards, Paul.


-
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >>  http://get.splunk.com/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Board Seats

[paul_c]

On Wednesday 25 July 2007 20:14, Stephen Wille Padnos wrote:
> The board is distinct from the developers.  The board is there to make
> decisions that affect the direction of development (do we deviate from
> the RS274NGC spec, do we add features for 23-axis machines ...),
> decisions about things like distribution and websites, that sort of
> thing.  Also, when situations arise that concern things like potential
> GPL violations or other licensing issues, the board is there as an
> authoritative body to communicate with outside parties.

The "board's" original mandate was to "guide the development of EMC and to 
oversee a move towards GPL". It was also charged with "acting as a point of 
contact between outside interests and developers" along with "resolving GPL 
and copyright infringements".

 To date, certain "board members" have taken it upon them selves to move the 
CVS repository away from Sourceforge without consultation with either 
developers or project administrators.

 The code base consists of a ragtag mix of (L)GPL, public domain, BSD/MIT, RA, 
and possibly proprietary code, but there is no desire to audit the source. As 
a consequence, it is quite likely that any claim to GPL is unenforceable so 
the "board" would rather abdicate responsibility for resolving "licensing 
issues" to the first Tom, Dick, or A*** that takes up the knobkerrie.

> In essence, the board is mostly uninvolved with day-to-day development,
> as I see it.

 Correct - Adding of features and functionality is for project administrators 
& developers under the guidance of the "board". However, the "board" members 
get elevated to project admin status and now run their own fork on an 
untrusted server.


Perhaps one day, the repository will be restored to it's rightfull home and we 
can move away from 'cargo cult' and 'crack crazed monkey' programming 
paradigms to producing a package that builds and runs on any Linux base. If 
it is any consolation, the current "official" release goes tits up with a 
2.6.20 kernel, and by the time 2.6.30 comes out, major surgery will be 
required.

---

Paul.

-
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >>  http://get.splunk.com/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Help sought to modify hostmotXX.vhd

[paul_c]

On Wednesday 02 May 2007 17:07, Stephen Wille Padnos wrote:
> Paul - do you know of any way to use the kernel firmware infrastructure
> to load different firmware sets into multiple cards?  I know the current
> method of embedding the firmware in the HAL driver doesn't support that
> either, but that's one goal for these cardds, to allow a user with
> multiple cards in the PC to use different personalities for each card.  
> (at least, that's a goal of mine)

As each card matching the devce:vendor ID is claimed by your driver, simply 
call the firmware loader with a corresponding file name. How you determine 
the file name is up to you - This could be from a default list, or an array 
passed at load time. An alternative would be:...

On Wednesday 02 May 2007 18:00, Peter C. Wallace wrote:
> If you wanted the specific configuration to be associated with a specific
> 5I20 card (not slot), there are some unused EEPROM locations on the 5I20
> that could be used as a driver hint.

Useful - Would there be sufficient space for 16 or 32 characters ?
Presumably, a trivial user space utility would be needed to "burn in" any file 
names..

On Wednesday 02 May 2007 18:11, John Kasunich wrote:
> The new system will not have an FPGA bitfile embedded in the driver.

A step in the right direction...

> bfload foo.bit 0

> Load what you want, where you want it, using a simple userspace program.
> Then load the driver.

And as always, there is a multitude of wrong ways of doing something. This 
bfload is just one of them. Presumably the "what" and "where" also extends 
to "when" so a user can dump a new firmware blob on a live system..

Yet another example of user space binary that requires root permissions, 
ignores proven, stable libraries/methods in favour of NIH programming.

-
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Help sought to modify hostmotXX.vhd

[paul_c]

Hi Seb

On Wednesday 02 May 2007 11:39, Seb James wrote:
> Mesa 5i20 board which provides 2 axes of servo/encoder control, 31
> channels of logic input and 29 channels of logic output.

Nothing too dificult there - IO is currently defined as 24 bits in at most, 
half a dozen places. Mapping the internal connections to physical pins would 
be the bulk of any work.

> I think this is a case of modifying the hostmot4 configuration, and then
> updating the hal_m5i20.c driver to suit the new fpga configuration.

Applying a couple of new constraints in the driver would expose the extra IO, 
again, nothing too taxing there. The worst part is the notion that firmware 
*has* to be embedded in the module or some untrusted user space utility is 
required to preload the fpga - The kernel already provides a mechanism for 
firmware loading without resorting to NIMBY code.

A few minutes with a text editor would see the changes you need in place. To 
rework the driver so that it does the right thing wrt firmware would take a 
little longer.


Regards, Paul.

-
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Hard disk space

[paul_c]

On Friday 30 March 2007 07:00, David A. Frantz wrote:
> Games can come or go but if the time comes that one needs to trim the
> space on the CD I would hope that the office programs stayed around for
> a bit.   There are numerous uses for "office" apps in the modern shop.

One of the problems with using a demo CD aimed at the mass market as a basis 
for "niche market" distro is that you get lumbered with the bells'n'whistles. 
Games and audio-visual applications cost space, throw in a large office 
suite, and the essential tools have to be dropped. Install from a live CD, 
and you get what ever is running without any chance of fine tuning the 
package selection - Sure, you could go through weeding out and removing stuff 
after installation, but it can be a tiresome exercise.

 For an "experimental" package such as emc2, essential tools to include are 
compilers, *-dev packages, and other sundry tools to build from source. 
Having a spreadsheet, text editor (even better if it can do highlighting for 
G code), and a calculator are advantageous. A selection of CAD tools should 
be regarded as desirable or even essential.

For building a live CD to act as a demonstration vehicle, there are a number 
of tools that can be used - Feed in a list of desired packages, point to a 
suitable repository, and as long as there are no broken dependencies, output 
an ISO image. [Have a couple of boxes online that generate Live images every 
night thanks to a cron job.]

If the primary purpose of the CD is to act as an installation medium, then 
being able to have some degree of control over what is installed is 
desirable. This would allow a user to choose a minimal install of say under 
100Megs, or go for "everything" (perhaps a Gig or two). This method requires 
a little more work from the CD builder to ensure all dependencies are 
satisfied on the CD, and if any custom packages are required, they are not 
broken just to work round a problem.
Again, there are tools available to build installation type CDs that require 
nothing more than a package list as an input. [Takes about ten minutes to 
update a BDI-4 package list, and a coffee or two to build - Could be 
automated without any real effort.]

Once either system (or even both) is set up, a new CD could be generated on 
demand or with the aid of a cron job - Most of the time involved after the 
initial setting up is minor edits to meet the requests of users.. Should 
anyone have a suggestion for package inclusion for the BDI-4, you only need 
to ask - There is some 30Megs currently unused, and there is always the 
possibility of using a second CD.. Would draw a line at a complete Debian 
archive as that would use over fourteen disks.

---

Paul.

-
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Another Look at SPI

[paul_c]

On Wednesday 28 March 2007 23:18, Kirk Wallace wrote:
> I just wanted a simple SPI device to talk to, but also have something
> useful when my lathe grows up.
> > That being said, once you got the bit-banging driver worked out it would
> > be useful for many applications.


> I did find this driver:
>
> http://www.paul.de/downloadables/#spi-driver
>
> and learned allot from the documentation and the source. Worth several
> days (for me at least)

Reasonably clean and readable code. Demonstrates several basic techniques for 
a driver without being convoluted - Adapting to compile on 2.6 series kernels 
is but a few minutes work and consists of minor changes.. Just watch out if 
you try to bit-bang more than a few bytes at a time. Each inb/outb 
instruction takes ~1uSec (on a standard parport), and the computer can do 
nothing during this period.

--

Paul.


-
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Minefields and...

[paul_c]

On Saturday 17 March 2007 03:29, Matthew Glenn Shaver wrote:
> OK, that I didn't know. The original copyright header says:
>
> # Copyright (C) 1991-2006 Altera Corporation
> # Your use of Altera Corporation's design tools, logic functions
> # and other software and tools,

Note " Altera Corporation's design tools"

> and its AMPP partner logic 
> # functions, and any output files any of the foregoing
> # (including device programming or simulation files), and any
> # associated documentation or information are expressly subject
> # to the terms and conditions of the Altera Program License
> # Subscription Agreement, Altera MegaCore Function License
> # Agreement, or other applicable license agreement, including,
> # without limitation, that your use is for the sole purpose of
> # programming logic devices manufactured by Altera and sold by
> # Altera or its authorized distributors.

Even without the MegaCore section, this would seem to suggest that any work 
using this toolchain is subject to restrictions.. Certainly sufficient 
grounds for concern.

On Saturday 17 March 2007 03:51, Jon Elson wrote:
> A VHDL file that is synthesized to program the Mesa board's FPGA isn't
> likely to contain Xilinx's IP, but may require something like a PCI macro.

It *may* be possible that a snippet of IP *might* be present, certainly not a 
PCI core. That was not the point being made - The toolchain EULA would appear 
to restrict what the user can do with the tools. The Altera header quoted 
would indicate terms that are in conflict with GPL. It could be argued that 
any HDL files edited under Quartus are "output files"..

On Saturday 17 March 2007 06:09, John Kasunich wrote:
> As Matt found, that license is here:
> https://www.altera.com/support/software/download/license/lic-weprog_lic.htm
Sec. 2b "use the Licensed Program for the sole purpose of programming logic 
devices manufactured by ALTERA"

On Saturday 17 March 2007 16:28, Matthew Glenn Shaver wrote:
> Programmable logic makers could do a better job of explaining what data
> _can_ be redistributed by their customers.

More importantly, under what terms - Can we assume HDL output to be free from 
restrictions (as long as no proprietary IP is included)...

What other files lurk in CVS with dubious terms & conditions ??

On Saturday 17 March 2007 14:55, Dale wrote:
> Just an opinion and there's no need to have a long drawnout discussion
> of the matter. consider it some thing to think about

None the less, this issue will keep slapping people in the face with a wet 
kipper until there is a policy of regular auditing from within. One would 
have hoped that lessons were learnt from the first time round.. But as can be 
seen elsewhere, nothing will get done until one of "the few" gets a kick in 
the knackers.








-
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Minefields and...

[paul_c]

On Friday 16 March 2007 18:17, Ray Henry wrote:
> I suggest that we let Paul do the audit and we keep working on
> abilities.

Buck. Passing ?

On Friday 16 March 2007 18:44, Matthew Glenn Shaver wrote:
> > src/hal/drivers/pluto_servo_firmware/pluto_servo.qsf
> GPL - Header at the top of the file.

Well, I guess the Altera lawyers could have a field day with that...
CVS history says one thing for the original commit, six months later, 
something else.

On Friday 16 March 2007 19:19, Chris Radek wrote:
> It's clear that opinions about this vary. Putting non-Free code into
> our source tree would be a departure from current practice and would
> require the board of directors to discuss and vote on it.  If the time
> comes, we'll have that vote.

Um.. NO. The board had a clear mandate to oversee moves towards GPL. To change 
on a whim without consultation with ALL previous contributors is wrong. 
Segregating non-compliant code in a separate module and distributed with 
clear warnings would be a reasonable compromise.

> Until then, I think, it's not very 
> interesting to argue about it, and I agree with Ray that since Paul is
> doing this audit work for us,

So once again you seek to abdicate your responsibilities. Since you guys 
decided to run a fork outside of Sourceforge, without discussion or 
consultation, anyone choosing not to follow gets branded an "ex developer".
No. Do your own damned audits and stop shirking your obligations as a "board 
member".

On Friday 16 March 2007 20:05, Gene Heskett wrote:
> Does it not bother you that whomever does own the copyrights on the files
> Paul listed, could decide to issue a C&D letter based on a percieved
> copyright violation?

Exactly Gene, and with the US legal system being what it is, even if damages 
are not sought, the legal bills could be crippling.

> Ditto if somehow a patent gets involved.  Either way, its conceivable that
> someone could be awarded outlandish damages against a project that
> basicly has no funds from sales to pay them.

Thankfully, software patents are not enforceable over here, but copyrights may 
well be.

On Friday 16 March 2007 23:43, Steve Stallings wrote:
> > And from Section 1 of the Xilinx ISE EULA:
> > (c) Restriction.  No right is granted hereunder to use
> > the Software to program or develop designs for non-XILINX devices.
> >
> > Which imposes conflicting restrictions on any firmware blobs
> > generated from GPL HDL sources.

> There are not any non-Xilinx devices that can utilize the "firmware
> blobs" generated by the Xilinx ISE from the GPL HDL sources, so this
> is meaningless.

So Joe develops & tests on a Hollybush board using ISE, perhaps unwittingly 
incorporating some Xilinx IP, and then Bob uses the same files to compile for 
an Altera Cyclone chip (with perhaps a small change to one file) - Perhaps 
using GHDL to synthesise the design - Far from being "meaningless", it opens 
up a whole new minefield. A warning notice for the end user advising that 
there may be a conflict of interest would be prudent.


-
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Another Potential User - I Wonder If He Subscribes?

[paul_c]

On Friday 16 March 2007 14:10, John Kasunich wrote:
> Accusing EMC2 of containing non-GPL code without pointing to the
> specific files in question is a perfect example of FUD.

Fer... You want to be able to make these claims, then I suggest you start 
conducting regular audits. A small selection, not limited to, but including: 
src/emc/usr_intf/axis/extensions/togl.c 
src/hal/drivers/pluto_servo_firmware/pluto_servo.qsf

 Then I suggest you look for files containing variations on BSD/RA licences 
before dismissing suggestions as FUD. Then there is:
http://sourceforge.net/mailarchive/message.php?msg_id=38081702 which seems to 
suggest that there may be non-compliant code added at any time.

And from Section 1 of the Xilinix ISE EULA:
(c) Restriction.  No right is granted hereunder to use the Software to 
program or develop designs for non-XILINX devices.

Which imposes conflicting restrictions on any firmware blobs generated from 
GPL HDL sources.

Now go and do your audit, keep it up to date, quit crying FUD and/or throwing 
up smokescreens, then if the question of GPL comes up again, you have grounds 
for a defense.


-
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Another Potential User - I Wonder If He Subscribes?

[paul_c]

On Friday 16 March 2007 03:44, Jon Elson wrote:
> On Thursday 15 March 2007 19:01, Dale wrote:
> >You can email me directly if you care to expand on what "FUD" means.

> Fear, Uncertainty, Doubt -- something naysayers spread to disparage
> something they don't want to succeed.

A cudgel used to denounce a comment of viewpoint a developer finds unpalatable 
regardless of the grounds for the original statement. e.g.

"emc2 must be really hard to configure - It takes a whole room of developers 
to set up one machine"

"emc2 contains non-GPL code"

"There are two ways of doing things. The right way, and the emc way."



-
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] EMC2 unistall

[paul_c]

On Monday 12 March 2007 12:09, Alex Joni wrote:
> Unfortunately at the moment there is no way to uninstall automatically. If
> you think that is needed, please add a feature request

No Alex, it is not a feature, it is a bug. The `uninstall` target is a 
perfectly reasonable function of any build system using autoconf and you 
would have got one for free if automake had been adopted. It would have also 
reduced the likelyhood of the following bug:

> Linking python module emc.so
> /usr/lib/gcc/i686-pc-linux-gnu/4.1.1/../../../../i686-pc-linux-gnu/bin/ld:
> cannot find -lGL
> collect2: ld returned 1 exit status
> make: *** [../lib/python/emc.so] error 1

... which has been reported before on more than one occasion. Telling a user 
to `apt-get install ` is NOT a bug fix and smacks of nepotism.

But hey, JMK has rejected automake/libtool support in favour of a 
sick'n'twisted build system that is fragile and aimed squarely at ubuntoo 
installs.



-
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] RTAI refusing to compile

[paul_c]

On Wednesday 28 February 2007 21:50, Marc van Doornik wrote:
> Initially I just downloaded rtai-3.2 and a vanilla 2.4.27 kernel. Compile
> kernel, no problem.

Very old combination - If you *really* want 2.4.27 patched with RTAI, I may 
have an old prebuilt Debian package lurking somewhere...

> After applying the rtai patch, compilation fails on 
> compiling adeos/x86.c with a "suffix or operands invalid for 'mov' " error.

x86.c can only be found in fusion which is obsolete, deprecated and 
unsupported.

> I've repeated this process with various combinations of gcc-2.95, 3.3, 3.4,
> 4.0 and 4.1, kernel 2.4.21, 2.4.25 and 2.4.27 and rtai 3.2, 3.1 and 3.05.
> So far, no luck. Has anybody ever had these problems?

Forget gcc-4.x with 2.4 kernels - I suspect there would be a number of things 
that break. Likewise, don't bother with any version of RTAI prior to the 3.4 
release. Any bug reports will result in being instructed to upgrade. Now that 
rtai-3.5 is out, would suggest using it.

> Is it at all possible to compile a 2.4 kernel on an originally non-2.4
> system? 

Compiled a 2.4.34 deb overnight without any errors (might stick it a server 
later) - The build machine runs modified 2.6.16 kernel.. Given a 
reasonable .config as a starting point, the task of compiling a kernel is 
neither difficult nor some black art. Only when having to deal with failed 
patches or digging in to bugs does it become "interesting".

> Besides, do I want a 2.4 kernel

A 2.4 series kernel should boot a little faster, but it may lack drivers for 
some of the latest hardware. In addition, you *MUST* install modutils 
alongside module-init-tools.

--
Paul.


-
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] 5i20 step generators

[paul_c]

On Thu, 15 Feb 2007 [EMAIL PROTECTED] wrote:
> > At present, EMC has a driver (by PeteV) that supports a single
> > configuration. That one config is stored as part of the EMC source code,
> > in both VHDL and binery form - binary because most people don't have the
> > Xilinx software installed, and even if they do, the translation from VHDL
> > to bitfile is NOT something you want to do every time you build the
> > software.

Calling xst and friends from the command line is trivial enough given the 
XILINX env. variable. But few people will need/want the ISE toolchain 
installed, likewise, many will not want to recompile just to update a single 
firmware blob.

 Embedding firmware is wrong - Do a search on the LKML for the reasons why - 
The kernel team have provided a simple mechanism (for 2.6.x series) that 
avoids bloating a driver with 97652 bytes (166980 for 200K devices, and even 
more for 400k).

[Peter] Is there an easy way to probe any of the fpga cards to determine which 
chip is fitted ?
(on a related topic, any chance of a dump of the 256bytes of the PCI registers 
on both 4i6x & 5i2x cards ? ).

 Firmware loading is not as straight forward with 2.4.x kernels - For those, 
search for mod_firmware_load() to see one way of doing it. Note that the "old 
way" imposes policy with respect to where the firmware is located (and no, it 
does NOT go in /etc). For firmware handling in 2.6.x kernels, refer to 
LDD.ver3 and also the source tree - Plenty of examples to be found there.

--
Paul.


-
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Qt3 lib and dro

[paul_c]

Hi James

On Thursday 04 January 2007 00:39, James Reed wrote:
>  Now, what do I do with it?  I'm very new at EMC, Linux, and even Qt3, (I'm
>  more familiar with Qt3 than the others, though).

You may do what ever you wish with the sources as long as you (or anyone else) 
do not claim copyright and/or change the licence. A simple makefile will be 
needed to call moc on qtdro.hh and then g++ to compile/link..

For example (in src/qt_dro):

moc_qtdro.cc: qtdro.hh
moc -o $@ $<

qtdro: moc_qtdro.cc
g++ qtdro.cc moc_qtdro.cc -lqt-mt -L$(LIB_DIR) -lnml -lemc -o $@

How you integrate the above snippet in to the emc2 build system is up to you 
and is something I can not help with - The current make structure is fragile, 
ugly, and based on a false premise.

For a simple DRO display, you might find the Tcl/Tk extensions to be 
suitable..


Regards, Paul.

-
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys - and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Qt3 lib and dro

[paul_c]

Hi Ray

On Wednesday 03 January 2007 16:50, Ray Henry wrote:
> There is a way around it if we find that Paul successfully raises the
> proprietary flag with the work that I encouraged him to do with qt3.

You misjudge me - The point I was trying to get across was that the code in 
question has NOT been released, and I would expect to be consulted (as a 
courtesy) before anyone made the code public.

On Wednesday 03 January 2007 17:23, Ray Henry wrote:
> The version of Paul's qt3 stuff that I have here shows the appended GPL
> licensing.

Indeed, "GPL Ver. 2" from the start. Attached is a tarball of the code in 
question, along with a few other sources. You may add these to the fork of 
your emc2 repository with one proviso - Copyright notice and license remains 
unaltered, and any changes are clearly marked.


Regards, Paul.





QtHMI.tar.bz2
Description: application/tbz
-
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys - and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] parallel port initializing

[paul_c]

On Wednesday 03 January 2007 02:17, Stephen Wille Padnos wrote:
> There was a little bit of work done on a Qt<->NML interface a couple of
> years ago.  It never went too far, but strangely enough, the "demo app"
> is a DRO :)

The core code is being used in a couple of applications.

> I'll dig it up and either stick it on the wiki or I can email it (or
> even check it into CVS somewhere).

You will of course check with the original author and ascertain that you have 
permission to release the code in to the public domain before 
posting/committing/emailing ?


Regards, Paul.


-
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys - and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] trying to install umbuntu

[paul_c]

Hi Alex

On Thursday 30 November 2006 15:55, alex wrote:
> But I think you should find a more reliable way of disttributing emc -
> ubuntu doesn't seem to be an answer - too hard for a novice user like me.

For a long while, there has been the BDI disk (based on Debian's stable 
branch, Sarge) - The latest version includes both a stable build based on the 
original EMC code as well as emc2. Although not a Live CD, it does come with 
a simple to use point'n'click installer that takes you through the install in 
half a dozen steps. For the more experienced user, it also provides finer 
control over what packages get installed - If you don't need the development 
tools or a lightweight window manager, go with the default selection.

The installer has been tested on a wide range of machines, including a 433MHz 
Celeron with just 128M memory and an AMD64 with 1Gig RAM - SATA disks have 
not been tried...

A full list of packages on the CD can be found at 
http://bdi4emc.ourproject.org/Package_list.html and links to the downloads at 
http://bdi4emc.ourproject.org/Where_to_get_it.html - Note: You get all the 
tools needed to compile emc2 should you want to.


Regards, Paul.


-
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys - and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Using emc2 with touchscreen

[paul_c]

On Friday 17 November 2006 00:29, Kasey Matejcek wrote:
> I've been using a touchscreen for a year now
> Just set it up under the X windows setup as and input device does every
> thing the same as the mouse

Been using microtouch and magictouch screens here for years - Aside from 
having to write a driver for the magictouch, configuration was simple. Both 
work fine with te mini GUI. The only real downside is dirty finger prints all 
over the screens.

Would I trust the machines to emc2 - Certainly not. From the configure stage:
"checking for xgettext tcl bug... ./configure: line 9092: test: `)' expected, 
found 5" - Not fatal, xgettext isn't used and the output files still get 
generated.. During compilation:
"hal/halmodule.cc:430: error: invalid conversion from ‘int (*)(PyObject*)’ 
to ‘Py_ssize_t (*)(PyObject*)’
make: *** [objects/hal/halmodule.o] Error 1" - Fatal if `make -k` isn't used..

Running brings it's own set of problems - It would appear there is a nasty 
race condition that results in certain modules failing to unload. The only 
thing that can be done to recover is a hard reboot of the system. If you are 
lucky enough to log an oops message as the system finally goes tits up, it 
might be possible to fix.. To date, the only error I've managed to log is:
"Nov 17 11:27:38 localhost kernel: halvcp[16136]: segfault at 2ad868280008 
rip 2ad86531d6ca rsp 7fff458a5d80 error 6" - I suspect this is due to 
a much more serious problem elsewhere.

-
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys - and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users