Re: [Emc-users] Those $20 80mm hand dials at MPJA.com

2017-03-12 Thread Gene Heskett 
On Sunday 12 March 2017 23:54:49 Jon Elson wrote:

> Got some rotary switches in, and built up a palm-sized MPG
> pendant for my Bridgeport today.
> I built my first pendant using a big Fanuc MPG that John
> Kasunich gave me at a codefest/CNC meeting in about 2007.
> When I built a second one, I used an encoder very similar to
> the ones from MP Jones and found I could do touching-off
> with a paper feeler if I held the pendant in one hand and
> operated the MPG with my thumb.  My first version was not
> set up so that was very easy to do.  So, I just made a new
> one that allows me to do that.
>
> See http://pico-systems.com/pendant.html
> for pictures of my first and 2nd versions.  The one I just
> built with the MP Jones encoder is VERY similar to the 2nd
> picture.
> You can hold it in one hand, press the enable button with
> your index finger and jog with your thumb.
>
> Jon
>
The switches I wound up using for the enables are much higher force 
switches than I would have liked. At least a pound to make them click. 

So I will have a timedelay in the code as a time to disable the wheels. 
When the switch is depressed to the click point, it will steer the 
wheels A=B=on signal to an updown, input controlled by the wheels turn 
direction to drive the count.up and count.down inputs.  The updown is 
restricted to a 0-7 range, and decoded by a bitslicer whose 3 output 
lines are the sel's for a mux8 which has the jog size setp'd to its 8 
inputs. When the button is released the timedelay eventually times out 
and disables the wheel. Currently set for a 10 second timeout. More time 
is obtained by pressing the button again w/o moving the wheel.  With the 
range clamped, for largest jog, just press button & spin cw, back to 
minimum, press button and spin ccw.

I think each of us likely has his own take on how a wheel should work, 
and my back, already tired from the housework (but the microwave is 
fixed horaayyy!) is limiting my time at the machine as that forces 
me to stand and lots of walking around it, so my code isn't working yet 
because I tested it with halshow without either the wheels or the 
switches hooked up.  So I have some logic that needs inverted and the 
timedelay re-routed. I suspect my code is 3x the complexity of that for 
a single wheel pendent. And I also suspect that the wheel reading parts 
may have to be moved to the faster servo-thread else counts might be 
missed at a 100hz sample rate. The wheel can be turned faster than one 
rev/second. That, or speed up my "jog-thread" to 500hz, as cpu loading 
doesn't seem to be growing that much with a 100hz thread.

Full opengl video would help also as the halmeter response is noticeably 
slow. But my questions on that to the debian-arm list have been ignored, 
twice now.  Does anyone here have a better video drive set up and 
working on a pi with the jessie install as base, kept uptodate daily?

Thanks everybody

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

--
Announcing the Oxford Dictionaries API! The API offers world-renowned
dictionary content that is easy and intuitive to access. Sign up for an
account today to start using our lexical data to power your apps and
projects. Get started today and enter our developer competition.
http://sdm.link/oxford
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Those $20 80mm hand dials at MPJA.com

2017-03-12 Thread Jon Elson 
Got some rotary switches in, and built up a palm-sized MPG 
pendant for my Bridgeport today.
I built my first pendant using a big Fanuc MPG that John 
Kasunich gave me at a codefest/CNC meeting in about 2007.
When I built a second one, I used an encoder very similar to 
the ones from MP Jones and found I could do touching-off 
with a paper feeler if I held the pendant in one hand and 
operated the MPG with my thumb.  My first version was not 
set up so that was very easy to do.  So, I just made a new 
one that allows me to do that.

See http://pico-systems.com/pendant.html
for pictures of my first and 2nd versions.  The one I just 
built with the MP Jones encoder is VERY similar to the 2nd 
picture.
You can hold it in one hand, press the enable button with 
your index finger and jog with your thumb.

Jon

--
Announcing the Oxford Dictionaries API! The API offers world-renowned
dictionary content that is easy and intuitive to access. Sign up for an
account today to start using our lexical data to power your apps and
projects. Get started today and enter our developer competition.
http://sdm.link/oxford
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Thoughts on extending the life of the microsd's.

2017-03-12 Thread Phillip Carter 
On 13/3/17 1:45 pm, Gregg Eshelman wrote:
> Then why hasn't its issue with noatime been fixed?
>
>
>
>From: Bertho Stultiens 
>   To: Enhanced Machine Controller (EMC) 
>   Sent: Sunday, March 12, 2017 4:22 PM
>   Subject: Re: [Emc-users] Thoughts on extending the life of the microsd's.
> 
> Mutt is a command-line email reader; see
> https://en.wikipedia.org/wiki/Mutt_(email_client)
>
> This has been a long time standard email client on many systems and is
> still in use today. It is *still* in development (last release ~16 days
> ago).
>
> 
> --
> Announcing the Oxford Dictionaries API! The API offers world-renowned
> dictionary content that is easy and intuitive to access. Sign up for an
> account today to start using our lexical data to power your apps and
> projects. Get started today and enter our developer competition.
> http://sdm.link/oxford
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users

http://homepages.inf.ed.ac.uk/imurray2/compnotes/noatime_mutt.html



--
Announcing the Oxford Dictionaries API! The API offers world-renowned
dictionary content that is easy and intuitive to access. Sign up for an
account today to start using our lexical data to power your apps and
projects. Get started today and enter our developer competition.
http://sdm.link/oxford
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Thoughts on extending the life of the microsd's.

2017-03-12 Thread Gregg Eshelman 
Then why hasn't its issue with noatime been fixed?



  From: Bertho Stultiens 
 To: Enhanced Machine Controller (EMC)  
 Sent: Sunday, March 12, 2017 4:22 PM
 Subject: Re: [Emc-users] Thoughts on extending the life of the microsd's.
   
Mutt is a command-line email reader; see
https://en.wikipedia.org/wiki/Mutt_(email_client)

This has been a long time standard email client on many systems and is
still in use today. It is *still* in development (last release ~16 days
ago).

   
--
Announcing the Oxford Dictionaries API! The API offers world-renowned
dictionary content that is easy and intuitive to access. Sign up for an
account today to start using our lexical data to power your apps and
projects. Get started today and enter our developer competition.
http://sdm.link/oxford
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Example configuration for THC + stepper

2017-03-12 Thread John Thornton 
Are you using the thcud component?

JT


On 3/12/2017 10:52 AM, Alexander Brock wrote:
> Hi,
>
> I built a XYZ router and attached a plasma cutter to it with a THC
> module which gives me "Arc ok", "up" and "down" signals.
> I figured out how to use Stepconf to move the stepper motors but I can't
> get the THC part to work.
>
> Does someone have a similar / example configuration? I saw that there is
> a example for simulating plasma operation in
> "configs/sim/gmoccapy/gmoccapy_plasma", but I don't understand how I
> need to modify it so it moves my steppers and actually uses my plasma
> cutter.
>
> Is it better / easier to modify the gmoccapy_plasma so it actually does
> something or should I try to modify the mill configuration which does
> already move the steppers?
>
> Best Regards,
> Alexander
>
> --
> Announcing the Oxford Dictionaries API! The API offers world-renowned
> dictionary content that is easy and intuitive to access. Sign up for an
> account today to start using our lexical data to power your apps and
> projects. Get started today and enter our developer competition.
> http://sdm.link/oxford
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users


--
Announcing the Oxford Dictionaries API! The API offers world-renowned
dictionary content that is easy and intuitive to access. Sign up for an
account today to start using our lexical data to power your apps and
projects. Get started today and enter our developer competition.
http://sdm.link/oxford
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Thoughts on extending the life of the microsd's.

2017-03-12 Thread Dave Cole 
noatime
http://en.tldp.org/LDP/solrhe/Securing-Optimizing-Linux-RH-Edition-v1.3/chap6sec73.html

. Mutt
http://www.tecmint.com/send-mail-from-command-line-using-mutt-command/

Dave

On 3/12/2017 6:04 PM, Peter Blodow wrote:
> Please, could someone explain to a poor physicist what noatime and Mutt are?
> Peter
>
> Am 12.03.2017 16:34, schrieb dragon:
>> For about three years now, I know of no other applications that has
>> issues with noatime other than Mutt. Everyone always says 'but it breaks
>> programs like Mutt' but in reality these days, my experience has been
>> that it only breaks Mutt. There used to be more applications that it
>> caused issues with but mostly they have either been patched or
>> deprecated. I have been using noatime for about 5 years now and have
>> experienced no issues with it on everything from desktops to servers to
>> dedicated appliance type setups for realtime audio and radio automation.
>>
>>   From what I have read, SSHFS has a bit more overhead in comparison to
>> NFS. In fairness though, I have never personally tested for it. Perhaps
>> someone else on the list has more info. On a desktop or server I
>> wouldn't sweat it too much but on the Pi it could make a difference. You
>> could just try both and see what happens. I have used both of them,
>> depending on the situation, but have never compared the impact on system
>> performance. Personally I wouldn't encrypt the remote mount on a secure
>> LAN when using the Pi unless sensitive data was involved. That just
>> seems like a lot of extra clock cycles for the little Pi, but perhaps I
>> am worrying about nothing.
>>
>> Amanda... check the perms on the directory itself, in addition to what
>> is set up for the mount.
>>
>>
>>
>>
>> On 03/12/2017 08:23 AM, Gene Heskett wrote:
>>> On Sunday 12 March 2017 08:39:03 Erik Christiansen wrote:
>>>
 On 11.03.17 16:42, Gene Heskett wrote:
> On Saturday 11 March 2017 12:02:36 dragon wrote:
>> I doubt that you will ever hit the end of life on a quality SD
>> card, especially a large size one. The writes that you are doing,
>> and thus the number of blocks, are TINY compared to what the cards
>> were designed for... photos and videos. You can also run a flash
>> file system instead of ext4 if you like. Checking that noatime is
>> set for the filesystem would have a far greater effect than all of
>> the writes that you will ever do in daily use. While true at one
>> time, this whole wearing out flash storage thing is almost a non
>> issue for use cases like this nowdays.
> How do I check that, and with the rube goldberg's hired hand boot
> configs used on the pi, how would I set the noatime option?
 AIUI, as relatime has long been the default (since kernel 2.6.30), you
 don't really need to do anything. Given that relatime only updates the
 access time if the previous access time was earlier than the current
 change time, there'll be slightly more writes than with noatime, but a
 lot less than with neither - and noatime breaks applications like
 mutt.

 If you do want to change it, then in /etc/fstab, add noatime to the
 comma-separated options in column 4. Before the relatime kernel
 default change, that was said to give a 10% performance boost. It
 shouldn't do anything noticeable now, I figure.

 If there's more recent info than that, it'd be interesting to hear.

>>> I was concerned on the pi, but its there for /, but not for /boot.  And I
>>> was never a fan of mutt, nor is email handled on that machine. It's
>>> running 3/4 ton of metal lathe.  But I use what I describe as mounts to
>>> put the high traffic dir as a mount over the top of an existing dir in
>>> order to put, when its fully operational, that directory effectively
>>> remoted to rotating media on this machine, useing sshfs.  This does of
>>> course show up in /etc/mtab.  That in turn gives amanda a tummy ache
>>> because although its wide open as far as perms go, its reported as no
>>> permission despite there being an entry for that directory on the pi's
>>> exclude list.  So amanda yowl's about it in the emailed backup report,
>>> but it does get backed up with the rest of this machine so its no
>>> biggie.
>>>
 Erik
>>> Thanks for the clarification, Erik.
 --
  Announcing the Oxford Dictionaries API! The API offers
 world-renowned dictionary content that is easy and intuitive to
 access. Sign up for an account today to start using our lexical data
 to power your apps and projects. Get started today and enter our
 developer competition. http://sdm.link/oxford
 ___
 Emc-users mailing list
 Emc-users@lists.sourceforge.net
 https://lists.sourceforge.net/lists/listinfo/emc-users
>>> Cheers, Gene Heskett
>>>
>>
>>

Re: [Emc-users] Thoughts on extending the life of the microsd's.

2017-03-12 Thread Bertho Stultiens 
On 03/12/2017 11:04 PM, Peter Blodow wrote:
> Please, could someone explain to a poor physicist what noatime and Mutt are?

Noatime refers to a flag that can be set on the filesystem. By default,
the access timestamp is recorded and saved in a unix filesystem (the
last time you access a file, any file). This requires a write to the
disk for /every/ thing you do with files, and on a unix system,
everything is a file. So it will add up to many many writes.

The flag noatime disables this behaviour and no access timestamps are
recorded/saved.


Mutt is a command-line email reader; see
https://en.wikipedia.org/wiki/Mutt_(email_client)

This has been a long time standard email client on many systems and is
still in use today. It is *still* in development (last release ~16 days
ago).


-- 
Greetings Bertho

(disclaimers are disclaimed)

--
Announcing the Oxford Dictionaries API! The API offers world-renowned
dictionary content that is easy and intuitive to access. Sign up for an
account today to start using our lexical data to power your apps and
projects. Get started today and enter our developer competition.
http://sdm.link/oxford
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Thoughts on extending the life of the microsd's.

2017-03-12 Thread Peter Blodow 
Please, could someone explain to a poor physicist what noatime and Mutt are?
Peter

Am 12.03.2017 16:34, schrieb dragon:
> For about three years now, I know of no other applications that has
> issues with noatime other than Mutt. Everyone always says 'but it breaks
> programs like Mutt' but in reality these days, my experience has been
> that it only breaks Mutt. There used to be more applications that it
> caused issues with but mostly they have either been patched or
> deprecated. I have been using noatime for about 5 years now and have
> experienced no issues with it on everything from desktops to servers to
> dedicated appliance type setups for realtime audio and radio automation.
>
>  From what I have read, SSHFS has a bit more overhead in comparison to
> NFS. In fairness though, I have never personally tested for it. Perhaps
> someone else on the list has more info. On a desktop or server I
> wouldn't sweat it too much but on the Pi it could make a difference. You
> could just try both and see what happens. I have used both of them,
> depending on the situation, but have never compared the impact on system
> performance. Personally I wouldn't encrypt the remote mount on a secure
> LAN when using the Pi unless sensitive data was involved. That just
> seems like a lot of extra clock cycles for the little Pi, but perhaps I
> am worrying about nothing.
>
> Amanda... check the perms on the directory itself, in addition to what
> is set up for the mount.
>
>
>
>
> On 03/12/2017 08:23 AM, Gene Heskett wrote:
>> On Sunday 12 March 2017 08:39:03 Erik Christiansen wrote:
>>
>>> On 11.03.17 16:42, Gene Heskett wrote:
 On Saturday 11 March 2017 12:02:36 dragon wrote:
> I doubt that you will ever hit the end of life on a quality SD
> card, especially a large size one. The writes that you are doing,
> and thus the number of blocks, are TINY compared to what the cards
> were designed for... photos and videos. You can also run a flash
> file system instead of ext4 if you like. Checking that noatime is
> set for the filesystem would have a far greater effect than all of
> the writes that you will ever do in daily use. While true at one
> time, this whole wearing out flash storage thing is almost a non
> issue for use cases like this nowdays.
 How do I check that, and with the rube goldberg's hired hand boot
 configs used on the pi, how would I set the noatime option?
>>> AIUI, as relatime has long been the default (since kernel 2.6.30), you
>>> don't really need to do anything. Given that relatime only updates the
>>> access time if the previous access time was earlier than the current
>>> change time, there'll be slightly more writes than with noatime, but a
>>> lot less than with neither - and noatime breaks applications like
>>> mutt.
>>>
>>> If you do want to change it, then in /etc/fstab, add noatime to the
>>> comma-separated options in column 4. Before the relatime kernel
>>> default change, that was said to give a 10% performance boost. It
>>> shouldn't do anything noticeable now, I figure.
>>>
>>> If there's more recent info than that, it'd be interesting to hear.
>>>
>> I was concerned on the pi, but its there for /, but not for /boot.  And I
>> was never a fan of mutt, nor is email handled on that machine. It's
>> running 3/4 ton of metal lathe.  But I use what I describe as mounts to
>> put the high traffic dir as a mount over the top of an existing dir in
>> order to put, when its fully operational, that directory effectively
>> remoted to rotating media on this machine, useing sshfs.  This does of
>> course show up in /etc/mtab.  That in turn gives amanda a tummy ache
>> because although its wide open as far as perms go, its reported as no
>> permission despite there being an entry for that directory on the pi's
>> exclude list.  So amanda yowl's about it in the emailed backup report,
>> but it does get backed up with the rest of this machine so its no
>> biggie.
>>
>>> Erik
>> Thanks for the clarification, Erik.
>>> --
>>>  Announcing the Oxford Dictionaries API! The API offers
>>> world-renowned dictionary content that is easy and intuitive to
>>> access. Sign up for an account today to start using our lexical data
>>> to power your apps and projects. Get started today and enter our
>>> developer competition. http://sdm.link/oxford
>>> ___
>>> Emc-users mailing list
>>> Emc-users@lists.sourceforge.net
>>> https://lists.sourceforge.net/lists/listinfo/emc-users
>>
>> Cheers, Gene Heskett
>>
>
>
> --
> Announcing the Oxford Dictionaries API! The API offers world-renowned
> dictionary content that is easy and intuitive to access. Sign up for an
> account today to start using our lexical data to power your apps and
> projects. Get started today and enter our

Re: [Emc-users] Example configuration for THC + stepper

2017-03-12 Thread Alexander Brock 
Hi,

I forgot to mention that I use the parport interface.

Best Regards,
Alexander



signature.asc
Description: OpenPGP digital signature
--
Announcing the Oxford Dictionaries API! The API offers world-renowned
dictionary content that is easy and intuitive to access. Sign up for an
account today to start using our lexical data to power your apps and
projects. Get started today and enter our developer competition.
http://sdm.link/oxford___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Behaviour of halui bit pins, halui.program.pause

2017-03-12 Thread Sebastian Kuzminsky 
On 03/12/2017 11:37 AM, Gene Heskett wrote:
> On Sunday 12 March 2017 13:05:58 Sebastian Kuzminsky wrote:
>>>   1544 static int check_bit_changed(bool halpin, bool )
>>>   1545 {
>>>   1546 if (halpin != newpin) {
>>>   1547 newpin = halpin;
>>>   1548 return halpin;
>>>   1549 }
>>>   1550 return 0;
>>>   1551 }
>>>
>>> ...
>>>
>>>   1654 if (check_bit_changed(new_halui_data.program_pause,
>>> old_halui_data.program_pause) != 0)
>>>   1655 sendProgramPause();
>>>
>>> That line 1654 looks to me as if both edges would trigger the msg
>>> emc_task_plan_pause_msg to be send.
>>>
>>> Is that correct ?
>>
>> Almost.  Look at line 1548, if the pin has switched state it turns the
>> new state.  So that function returns True on a positive edge, only.
>>
> I must be getting dense Seb, but I read that snippet around 1548 as
> changing the return if they are not equal. Its not checking for a rising
> edge, its checking for in-equality. Or so it looks to me.

The check_bit_changed() function returns 0 if the old and the new bits 
are the same, but returns the *new* bit if they are different.

If the old bit is 0 and the new bit is 1 (rising edge) it returns 1.

If the old bit is 1 and the new bit is 0 (falling edge) it returns 0.


-- 
Sebastian Kuzminsky

--
Announcing the Oxford Dictionaries API! The API offers world-renowned
dictionary content that is easy and intuitive to access. Sign up for an
account today to start using our lexical data to power your apps and
projects. Get started today and enter our developer competition.
http://sdm.link/oxford
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Behaviour of halui bit pins, halui.program.pause

2017-03-12 Thread Gene Heskett 
On Sunday 12 March 2017 13:05:58 Sebastian Kuzminsky wrote:

> On March 12, 2017 5:31:42 AM MDT, MH  
wrote:
> >Hello,
> >
> >I have a question about the behaviour of the halui bit pins and would
> >like to confirm that I understand the code correctly. I could not
> >answer
> >my question from the halui documentation.
> >
> >Do these pins activate on both a rising and a falling edge ?
> >
> >It like to use the input pin halui.program.pause to pause a running
> >program if I get a signal that might indicate that the material sheet
> >*might* have moved and the operator needs to check.
> >
> >Having a cursory look at the linuxcnc source code I suspect the halui
> >pins come from linuxcnc-dev/src/emc/usr_intf/halui.cc
> >
> >And in halui.cc these line (with lines numbers):
> >
> >   1114 static int sendProgramPause()
> >   1115 {
> >   1116 EMC_TASK_PLAN_PAUSE emc_task_plan_pause_msg;
> >   1117
> >   1118 return emcCommandSend(emc_task_plan_pause_msg);
> >
> >...
> >
> >   1544 static int check_bit_changed(bool halpin, bool )
> >   1545 {
> >   1546 if (halpin != newpin) {
> >   1547 newpin = halpin;
> >   1548 return halpin;
> >   1549 }
> >   1550 return 0;
> >   1551 }
> >
> >...
> >
> >   1654 if (check_bit_changed(new_halui_data.program_pause,
> >old_halui_data.program_pause) != 0)
> >   1655 sendProgramPause();
> >
> >That line 1654 looks to me as if both edges would trigger the msg
> >emc_task_plan_pause_msg to be send.
> >
> >Is that correct ?
>
> Almost.  Look at line 1548, if the pin has switched state it turns the
> new state.  So that function returns True on a positive edge, only.
>
I must be getting dense Seb, but I read that snippet around 1548 as 
changing the return if they are not equal. Its not checking for a rising 
edge, its checking for in-equality. Or so it looks to me.

line 1654 OTOH, would send it only on the rising edge with its !=0 test. 
I think...  I haven't carved any C to speak of since K #2 was out + 
maybe 5 years.

> I notice now that the names of the variables inside the function are
> the opposite of what it seems they should be, that's confusing...

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

--
Announcing the Oxford Dictionaries API! The API offers world-renowned
dictionary content that is easy and intuitive to access. Sign up for an
account today to start using our lexical data to power your apps and
projects. Get started today and enter our developer competition.
http://sdm.link/oxford
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Thoughts on extending the life of the microsd's.

2017-03-12 Thread Gene Heskett 
On Sunday 12 March 2017 11:34:27 dragon wrote:

> For about three years now, I know of no other applications that has
> issues with noatime other than Mutt. Everyone always says 'but it
> breaks programs like Mutt' but in reality these days, my experience
> has been that it only breaks Mutt. There used to be more applications
> that it caused issues with but mostly they have either been patched or
> deprecated. I have been using noatime for about 5 years now and have
> experienced no issues with it on everything from desktops to servers
> to dedicated appliance type setups for realtime audio and radio
> automation.
>
> From what I have read, SSHFS has a bit more overhead in comparison to
> NFS. In fairness though, I have never personally tested for it.
> Perhaps someone else on the list has more info. On a desktop or server
> I wouldn't sweat it too much but on the Pi it could make a difference.
> You could just try both and see what happens. I have used both of
> them, depending on the situation, but have never compared the impact
> on system performance. Personally I wouldn't encrypt the remote mount
> on a secure LAN when using the Pi unless sensitive data was involved.
> That just seems like a lot of extra clock cycles for the little Pi,
> but perhaps I am worrying about nothing.

The ssh encryption lag over the wire doesn't seem to be noticeable, gcode 
files are rarely over 300k in size if machine generated, while what I 
write, because I use loop structures everyplace I can, is rarely over 
300 lines, and one routine I wrote to sharpen carbide toothed table saw 
blades, takes about a day and a half to run, is actually less than 50 
lines. Finding a bug in machine generated code is nigh impossible, but a 
bug in my own code, because its probably in a subroutine, and repeated 
possibly 100's of times, is pretty easily found and fixed.

> Amanda... check the perms on the directory itself, in addition to what
> is set up for the mount.

If the mount has not been done, amanda doesn't complain, and it skips 
that directory like the good little girl she usually is. Here on this 
machine its 777. And while I am first user=1000 as gene here, there I'm 
first user=1000 as pi. tar is 1.27.1 on both.  Same pw on both even. One 
of those things that if I scratched my head over it, the hair would be 
thinner than it already is at 82yo. :)

[...]

> > Thanks for the clarification, Erik.

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

--
Announcing the Oxford Dictionaries API! The API offers world-renowned
dictionary content that is easy and intuitive to access. Sign up for an
account today to start using our lexical data to power your apps and
projects. Get started today and enter our developer competition.
http://sdm.link/oxford
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Behaviour of halui bit pins, halui.program.pause

2017-03-12 Thread Sebastian Kuzminsky 


On March 12, 2017 5:31:42 AM MDT, MH  wrote:
>
>Hello,
>
>I have a question about the behaviour of the halui bit pins and would
>like to confirm that I understand the code correctly. I could not
>answer
>my question from the halui documentation.
>
>Do these pins activate on both a rising and a falling edge ?
>
>It like to use the input pin halui.program.pause to pause a running
>program if I get a signal that might indicate that the material sheet
>*might* have moved and the operator needs to check.
>
>Having a cursory look at the linuxcnc source code I suspect the halui
>pins come from linuxcnc-dev/src/emc/usr_intf/halui.cc
>
>And in halui.cc these line (with lines numbers):
>
>   1114 static int sendProgramPause()
>   1115 {
>   1116 EMC_TASK_PLAN_PAUSE emc_task_plan_pause_msg;
>   1117
>   1118 return emcCommandSend(emc_task_plan_pause_msg);
>
>...
>
>   1544 static int check_bit_changed(bool halpin, bool )
>   1545 {
>   1546 if (halpin != newpin) {
>   1547 newpin = halpin;
>   1548 return halpin;
>   1549 }
>   1550 return 0;
>   1551 }
>
>...
>
>   1654 if (check_bit_changed(new_halui_data.program_pause,
>old_halui_data.program_pause) != 0)
>   1655 sendProgramPause();
>
>That line 1654 looks to me as if both edges would trigger the msg
>emc_task_plan_pause_msg to be send.
>
>Is that correct ?

Almost.  Look at line 1548, if the pin has switched state it turns the new 
state.  So that function returns True on a positive edge, only.

I notice now that the names of the variables inside the function are the 
opposite of what it seems they should be, that's confusing...

-- 
Sebastian Kuzminsky

--
Announcing the Oxford Dictionaries API! The API offers world-renowned
dictionary content that is easy and intuitive to access. Sign up for an
account today to start using our lexical data to power your apps and
projects. Get started today and enter our developer competition.
http://sdm.link/oxford
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

[Emc-users] Example configuration for THC + stepper

2017-03-12 Thread Alexander Brock 
Hi,

I built a XYZ router and attached a plasma cutter to it with a THC
module which gives me "Arc ok", "up" and "down" signals.
I figured out how to use Stepconf to move the stepper motors but I can't
get the THC part to work.

Does someone have a similar / example configuration? I saw that there is
a example for simulating plasma operation in
"configs/sim/gmoccapy/gmoccapy_plasma", but I don't understand how I
need to modify it so it moves my steppers and actually uses my plasma
cutter.

Is it better / easier to modify the gmoccapy_plasma so it actually does
something or should I try to modify the mill configuration which does
already move the steppers?

Best Regards,
Alexander

--
Announcing the Oxford Dictionaries API! The API offers world-renowned
dictionary content that is easy and intuitive to access. Sign up for an
account today to start using our lexical data to power your apps and
projects. Get started today and enter our developer competition.
http://sdm.link/oxford
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Thoughts on extending the life of the microsd's.

2017-03-12 Thread dragon 
For about three years now, I know of no other applications that has
issues with noatime other than Mutt. Everyone always says 'but it breaks
programs like Mutt' but in reality these days, my experience has been
that it only breaks Mutt. There used to be more applications that it
caused issues with but mostly they have either been patched or
deprecated. I have been using noatime for about 5 years now and have
experienced no issues with it on everything from desktops to servers to
dedicated appliance type setups for realtime audio and radio automation.

From what I have read, SSHFS has a bit more overhead in comparison to
NFS. In fairness though, I have never personally tested for it. Perhaps
someone else on the list has more info. On a desktop or server I
wouldn't sweat it too much but on the Pi it could make a difference. You
could just try both and see what happens. I have used both of them,
depending on the situation, but have never compared the impact on system
performance. Personally I wouldn't encrypt the remote mount on a secure
LAN when using the Pi unless sensitive data was involved. That just
seems like a lot of extra clock cycles for the little Pi, but perhaps I
am worrying about nothing.

Amanda... check the perms on the directory itself, in addition to what
is set up for the mount.




On 03/12/2017 08:23 AM, Gene Heskett wrote:
> On Sunday 12 March 2017 08:39:03 Erik Christiansen wrote:
> 
>> On 11.03.17 16:42, Gene Heskett wrote:
>>> On Saturday 11 March 2017 12:02:36 dragon wrote:
 I doubt that you will ever hit the end of life on a quality SD
 card, especially a large size one. The writes that you are doing,
 and thus the number of blocks, are TINY compared to what the cards
 were designed for... photos and videos. You can also run a flash
 file system instead of ext4 if you like. Checking that noatime is
 set for the filesystem would have a far greater effect than all of
 the writes that you will ever do in daily use. While true at one
 time, this whole wearing out flash storage thing is almost a non
 issue for use cases like this nowdays.
>>>
>>> How do I check that, and with the rube goldberg's hired hand boot
>>> configs used on the pi, how would I set the noatime option?
>>
>> AIUI, as relatime has long been the default (since kernel 2.6.30), you
>> don't really need to do anything. Given that relatime only updates the
>> access time if the previous access time was earlier than the current
>> change time, there'll be slightly more writes than with noatime, but a
>> lot less than with neither - and noatime breaks applications like
>> mutt.
>>
>> If you do want to change it, then in /etc/fstab, add noatime to the
>> comma-separated options in column 4. Before the relatime kernel
>> default change, that was said to give a 10% performance boost. It
>> shouldn't do anything noticeable now, I figure.
>>
>> If there's more recent info than that, it'd be interesting to hear.
>>
> I was concerned on the pi, but its there for /, but not for /boot.  And I 
> was never a fan of mutt, nor is email handled on that machine. It's 
> running 3/4 ton of metal lathe.  But I use what I describe as mounts to 
> put the high traffic dir as a mount over the top of an existing dir in 
> order to put, when its fully operational, that directory effectively 
> remoted to rotating media on this machine, useing sshfs.  This does of 
> course show up in /etc/mtab.  That in turn gives amanda a tummy ache 
> because although its wide open as far as perms go, its reported as no 
> permission despite there being an entry for that directory on the pi's 
> exclude list.  So amanda yowl's about it in the emailed backup report, 
> but it does get backed up with the rest of this machine so its no 
> biggie.
> 
>> Erik
> 
> Thanks for the clarification, Erik.
>> --
>>  Announcing the Oxford Dictionaries API! The API offers
>> world-renowned dictionary content that is easy and intuitive to
>> access. Sign up for an account today to start using our lexical data
>> to power your apps and projects. Get started today and enter our
>> developer competition. http://sdm.link/oxford
>> ___
>> Emc-users mailing list
>> Emc-users@lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/emc-users
> 
> 
> Cheers, Gene Heskett
> 



signature.asc
Description: OpenPGP digital signature
--
Announcing the Oxford Dictionaries API! The API offers world-renowned
dictionary content that is easy and intuitive to access. Sign up for an
account today to start using our lexical data to power your apps and
projects. Get started today and enter our developer competition.
http://sdm.link/oxford___
Emc-users mailing list
Emc-users@lists.sourceforge.net

Re: [Emc-users] Thoughts on extending the life of the microsd's.

2017-03-12 Thread Gene Heskett 
On Sunday 12 March 2017 08:39:03 Erik Christiansen wrote:

> On 11.03.17 16:42, Gene Heskett wrote:
> > On Saturday 11 March 2017 12:02:36 dragon wrote:
> > > I doubt that you will ever hit the end of life on a quality SD
> > > card, especially a large size one. The writes that you are doing,
> > > and thus the number of blocks, are TINY compared to what the cards
> > > were designed for... photos and videos. You can also run a flash
> > > file system instead of ext4 if you like. Checking that noatime is
> > > set for the filesystem would have a far greater effect than all of
> > > the writes that you will ever do in daily use. While true at one
> > > time, this whole wearing out flash storage thing is almost a non
> > > issue for use cases like this nowdays.
> >
> > How do I check that, and with the rube goldberg's hired hand boot
> > configs used on the pi, how would I set the noatime option?
>
> AIUI, as relatime has long been the default (since kernel 2.6.30), you
> don't really need to do anything. Given that relatime only updates the
> access time if the previous access time was earlier than the current
> change time, there'll be slightly more writes than with noatime, but a
> lot less than with neither - and noatime breaks applications like
> mutt.
>
> If you do want to change it, then in /etc/fstab, add noatime to the
> comma-separated options in column 4. Before the relatime kernel
> default change, that was said to give a 10% performance boost. It
> shouldn't do anything noticeable now, I figure.
>
> If there's more recent info than that, it'd be interesting to hear.
>
I was concerned on the pi, but its there for /, but not for /boot.  And I 
was never a fan of mutt, nor is email handled on that machine. It's 
running 3/4 ton of metal lathe.  But I use what I describe as mounts to 
put the high traffic dir as a mount over the top of an existing dir in 
order to put, when its fully operational, that directory effectively 
remoted to rotating media on this machine, useing sshfs.  This does of 
course show up in /etc/mtab.  That in turn gives amanda a tummy ache 
because although its wide open as far as perms go, its reported as no 
permission despite there being an entry for that directory on the pi's 
exclude list.  So amanda yowl's about it in the emailed backup report, 
but it does get backed up with the rest of this machine so its no 
biggie.

> Erik

Thanks for the clarification, Erik.
> --
> Announcing the Oxford Dictionaries API! The API offers
> world-renowned dictionary content that is easy and intuitive to
> access. Sign up for an account today to start using our lexical data
> to power your apps and projects. Get started today and enter our
> developer competition. http://sdm.link/oxford
> ___
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users


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

--
Announcing the Oxford Dictionaries API! The API offers world-renowned
dictionary content that is easy and intuitive to access. Sign up for an
account today to start using our lexical data to power your apps and
projects. Get started today and enter our developer competition.
http://sdm.link/oxford
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Re: [Emc-users] Thoughts on extending the life of the microsd's.

2017-03-12 Thread Erik Christiansen 
On 11.03.17 16:42, Gene Heskett wrote:
> On Saturday 11 March 2017 12:02:36 dragon wrote:
> 
> > I doubt that you will ever hit the end of life on a quality SD card,
> > especially a large size one. The writes that you are doing, and thus
> > the number of blocks, are TINY compared to what the cards were
> > designed for... photos and videos. You can also run a flash file
> > system instead of ext4 if you like. Checking that noatime is set for
> > the filesystem would have a far greater effect than all of the writes
> > that you will ever do in daily use. While true at one time, this whole
> > wearing out flash storage thing is almost a non issue for use cases
> > like this nowdays.
> >
> How do I check that, and with the rube goldberg's hired hand boot configs 
> used on the pi, how would I set the noatime option?

AIUI, as relatime has long been the default (since kernel 2.6.30), you
don't really need to do anything. Given that relatime only updates the
access time if the previous access time was earlier than the current
change time, there'll be slightly more writes than with noatime, but a
lot less than with neither - and noatime breaks applications like mutt.

If you do want to change it, then in /etc/fstab, add noatime to the
comma-separated options in column 4. Before the relatime kernel default
change, that was said to give a 10% performance boost. It shouldn't do
anything noticeable now, I figure.

If there's more recent info than that, it'd be interesting to hear.

Erik

--
Announcing the Oxford Dictionaries API! The API offers world-renowned
dictionary content that is easy and intuitive to access. Sign up for an
account today to start using our lexical data to power your apps and
projects. Get started today and enter our developer competition.
http://sdm.link/oxford
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

[Emc-users] Behaviour of halui bit pins, halui.program.pause

2017-03-12 Thread MH 

Hello,

I have a question about the behaviour of the halui bit pins and would
like to confirm that I understand the code correctly. I could not answer
my question from the halui documentation.

Do these pins activate on both a rising and a falling edge ?

It like to use the input pin halui.program.pause to pause a running
program if I get a signal that might indicate that the material sheet
*might* have moved and the operator needs to check.

Having a cursory look at the linuxcnc source code I suspect the halui
pins come from linuxcnc-dev/src/emc/usr_intf/halui.cc

And in halui.cc these line (with lines numbers):

   1114 static int sendProgramPause()
   1115 {
   1116 EMC_TASK_PLAN_PAUSE emc_task_plan_pause_msg;
   1117
   1118 return emcCommandSend(emc_task_plan_pause_msg);

...

   1544 static int check_bit_changed(bool halpin, bool )
   1545 {
   1546 if (halpin != newpin) {
   1547 newpin = halpin;
   1548 return halpin;
   1549 }
   1550 return 0;
   1551 }

...

   1654 if (check_bit_changed(new_halui_data.program_pause,
old_halui_data.program_pause) != 0)
   1655 sendProgramPause();

That line 1654 looks to me as if both edges would trigger the msg
emc_task_plan_pause_msg to be send.

Is that correct ?

Thank you
BR
Max.


--
Announcing the Oxford Dictionaries API! The API offers world-renowned
dictionary content that is easy and intuitive to access. Sign up for an
account today to start using our lexical data to power your apps and
projects. Get started today and enter our developer competition.
http://sdm.link/oxford
___
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users