Re: Android Studio and AVDs

[paul . wallrabe]

Same issue here. I literally tried every suggested fix on stack
overflow and it is not possible to use hardware acclaration on Debian
Stretch. KVM is installed and emulator -accel-check is returning

accel:
0
KVM (version 12) is installed and usable.
accel

But somehow Android Studio does not start the AVD.

Re: Re: Laptop randomly reboots

[Alexander V. Makartsev]

You can check system logs after reboot with: sudo journalctl -b -1
where -1 stands for previous boot (-2 for one boot before previous boot,
etc).
Maybe it will show some hints about what happened with your system
before laptop reboot.

If there is no error messages in the logs, then it is definitely a
hardware problem. And since you bought used mobo it could be anything,
from bad soldering to faulty controller that was overheated during
repair procedures.
>From my experience, I'd suggest you swap mobos to the one you bought
laptop with( with nvidia GPU on it ). Power consumption and
hibernate\resume won't be an issue because you can use integrated
graphics adapter from your CPU and use nvidia GPU only on demand via
primus\bumblebee.

Install Debian 9.1 stretch on Pentium II system: pata_hpt3x2n required?

[David Max]

Hello

I am attempting to install Debian 9.1 stretch on an old Pentium II system.

Setup --

cpu Pentium II 400 MHz
ram 512 MB
video card ATI RAGE IIC AGP
hdd #1 6.8 GB Fujitsu (with Windows installed, also a 1GB Linux swap
partition)
hdd #2 80 GB Seagate, attached to Highpoint 133SB Rocket Raid (PCI) card

Install process goes well until the stage of searching for disks. Then the
first disk is detected fine, but 2nd doesn't show up.

If I pause the installation sequence at disk detection, driver pata_hpt37x
is shown by lsmod. lspci shows the Highpoint card as "RAID bus controller
HighPoint Technologies, Inc. HPT302/302N (rev 02)".

As some background, Knoppix 7.2.0 installs okay on the Seagate drive but I
seem to have broken it by upgrading some packages (and removing
libreoffice). On boot, Knoppix complains that graphics cannot be
initialised, I think because package xserver-xorg-video-ati has been
unpacked but not installed (i.e., on the Knoppix hdd installation).

The aim is to see if this old machine can take on a new life as a router.
Traffic rates through it are adequate for current needs. If Debian install
can be completed, I will switch to LXDE for the desktop (quite lightweight
as I understand).

If Debian can be installed Knoppix will be replaced but the old Windows
install on hdd #1 will be retained.

After a little research, I believe a driver pata_hpt3x2n might be required?

All help gratefully received.

David Max

Re: Install Debian 9.1 stretch on Pentium II system: pata_hpt3x2n required?

[tomas]

-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

On Tue, Sep 05, 2017 at 10:34:46AM +0100, David Max wrote:
> Hello
> 
> I am attempting to install Debian 9.1 stretch on an old Pentium II system.
> 
> Setup --
> 
> cpu Pentium II 400 MHz
> ram 512 MB
> video card ATI RAGE IIC AGP
> hdd #1 6.8 GB Fujitsu (with Windows installed, also a 1GB Linux swap
> partition)
> hdd #2 80 GB Seagate, attached to Highpoint 133SB Rocket Raid (PCI) card

Uh. Historical. (Don't get me wrong: if it runs... i find it commendable
when people try to wring some life from a system instead of sending it
off the dump -- unless it is a 12kW devouring monster, that is.

> Install process goes well until the stage of searching for disks. Then the
> first disk is detected fine, but 2nd doesn't show up.
> 
> If I pause the installation sequence at disk detection, driver pata_hpt37x
> is shown by lsmod. lspci shows the Highpoint card as "RAID bus controller
> HighPoint Technologies, Inc. HPT302/302N (rev 02)".

Hm. Parallel ATA support seems to still be there, but you are exercising
paths which don't get exercised these days very much. Perhaps some module
isn't being loaded which should (best case) or some driver isn't working
any more (worst case). If you don't know any better, you can play around
with

  find /lib/modules -type f -name "*pata*"

and try modprobing things which seem to make sense. Wear protective
goggles 8-)

> As some background, Knoppix 7.2.0 installs okay on the Seagate drive but I
> seem to have broken it by upgrading some packages (and removing
> libreoffice). On boot, Knoppix complains that graphics cannot be
> initialised, I think because package xserver-xorg-video-ati has been
> unpacked but not installed (i.e., on the Knoppix hdd installation).
> 
> The aim is to see if this old machine can take on a new life as a router.
> Traffic rates through it are adequate for current needs. If Debian install
> can be completed, I will switch to LXDE for the desktop (quite lightweight
> as I understand).

Consider ditching the desktop environment (and X) altogether: it's a
router after all, and you have one piece of historical hardware less
to worry about (unless it Just Works, or it is part of your challenge,
that is).

> If Debian can be installed Knoppix will be replaced but the old Windows
> install on hdd #1 will be retained.
> 
> After a little research, I believe a driver pata_hpt3x2n might be required?

On my system, there are two (very roughly) matching candidates:

  /lib/modules/4.3.0-1-amd64/kernel/drivers/ata/pata_hpt37x.ko
  /lib/modules/4.3.0-1-amd64/kernel/drivers/ata/pata_hpt366.ko

You can have a look at the sources (I'd expect those drivers to be
pretty stable, so the exact version wouldn't matter much). You can
browse sources at https://git.kernel.org/, e.g.:

  
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/drivers/ata/pata_hpt37x.c?h=v4.3&id=6a13feb9c82803e2b815eca72fa7a9f5561d7861

The source, and your PATA chipset model together might offer some
hints.

Good luck, and... keep us up to date :-)

Cheers
- -- t
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.12 (GNU/Linux)

iEYEARECAAYFAlmug7cACgkQBcgs9XrR2kbWBQCeP3JbibWWsIjL1zAunFKytHem
w8gAn2nQgu4Q91QSAfzLN5ex0Ik9Pf03
=gBu2
-END PGP SIGNATURE-

Re: Laptop recommendation

[Dekks Herton]

Jonas Hedman  writes:


Hello I hope that is not OT for this list.

Basically I'm on the hunt for a newish laptop on which I 
naturally want
to run Debian. I'm a student and I spend most of my daily 
outandabout 
computer time reading pdfs, writing LaTeX docs, surfing and 
doing some 
light coding, no heavy duty stuff in other words. I use i3vm and 
I
generally like to keep things as light and minimal as feasibly 
possible.


I really value light weight and good battery life and my upper 
bound on prize
is somewhere around 875$ (~7K SEK). Memory should be greater 
than or

equal to 2GB and an sdd would be nice.

I know very little about hardware and I'm not quite sure where 
or how to 
start looking. I though the Lenovo Thinkpad 13 looked promising 
but it's a bit 
over my price target. Could chromeboosk be something to look 
into? I
know the older thinkpads has good debian support but if I 
remember

correctly they are also very heavy.

As a small reference: I made due with a Asus EEE PC 1005HA for a 
number 
of years and it generally worked well for my purposes but it 
died and

surfing was a pain so I'm looking to upgrade a bit from that.

Any suggestions?

Best regards


Check out ebay for something like a Thinkpad X230 or a Thinkpad 
T430 with an IPS HD+ screen. I know in sweden availability might 
be an issue so check out T440 or T450 prices too. 

Re: Laptop randomly reboots

[rhkramer]

On Tuesday, September 05, 2017 12:46:58 AM Sam Smith wrote:
> Well I guess I really spoke too soon Just got a random reboot a
> minute ago... Pretty much at a loss now. The only thing that hasn't been
> replaced now is the CPU, battery, and power adapter. I do have a spare
> battery and power adapter. Guess I'll try running with those for a month
> or so and see what happens. Though not really sure how either one of
> those would make the computer reboot randomly once a month. Laptop is
> mounted in a docking station. At least for the last 6 months, every
> reboot was while it was docked. But I do remember times last year
> sitting at the kitchen table undocked and getting a reboot so I don't
> think the docking station is the issue.. But have no idea now :/

I haven't paid much attention to this before today, but, if I had a problem 
with unexplained reboots, the first area I'd look at is the power supply 
"chain"--that is, the battery, the connections between the battery and the 
laptop, power "glitches" to the house,etc.

I run a laptop with a known bad battery--if I disconnect the power supply it 
crashes in seconds.  OTOH, if your battery is good, it serves as a (long 
lived) UPS--you should be able to ride out hours long power glitches.

If I had a known good battery, I might try examining the contacts between the 
battery and the laptop--are they clean / shiny, is the "springy side" still 
"springy", does the battery fit snugly in the case or is there enough clearance 
for it to possibly move and break contact with the laptop?  I might even try  
picking up the laptop and reasonably gently shaking it (while running) to see 
if it reboots.  (You don't want to shake it so hard that you damage something 
else.)

If you have a known bad battery, I'd look at the chain from the power outlet 
through the various cords and contacts on both sides of the power adapter.  
Try shaking the power adapter.

Is the power to your house reliable--is there any chance you've had brief 
power outages overnight?  (In the US, most electric utilities have automatic 
reclosers on the power line circuit breakers--if they detect a fault they 
open, then try closing again after a few seconds (in hopes that either the 
fault has cleared (maybe a shorting tree branch has either blown or burned),  
away, open again if there is still a fault, again try to reclose, after a few 
seconds.  In most cases, they try this something like 3 times, with a somewhat 
longer delay before the last retry.

Oh wait, you have (presumably known good) spare battery and power adapter--try 
those.

Strange results with an additional HD -- any idea why?

[James H. H. Lampert]

The box I've been reconfiguring over the past few weeks has a hardware 
RAID controller card, with one mirrored (RAID 1) pair on it at the time 
of installation. Over the weekend, I plugged two more drives into the 
two empty sockets, to create a second mirrored pair, which shows up in 
Linux as "sdb."


Initially, it auto-mounted much the same way the external drive 
auto-mounted. I added this line to fstab, and now it mounts at the 
mountpoint of my own choosing:



/dev/sdb /media/Auxiliary ext4 errors=remount-ro 0 2


Everything seems to work fine, except for two things:

1. Gnome produces a desktop icon for it. Before I added the fstab entry, 
that icon was a picture of a modern hard drive; now, with the fstab 
entry, it is, of all things, a generic *document* icon. (By contrast, my 
USB external drive shows up with the "Merlin disk pack" icon I gave it 
years ago, and I didn't have to do anything new for Linux that I hadn't 
done for WinDoze; I've got both .ico and .png files of that icon on the 
new mirrored pair.)


2. It can be unmounted. I'd rather it not be unmountable.

What can I do about these things?
--
JHHL

Re: Strange results with an additional HD -- any idea why?

[Dan Ritter]

On Tue, Sep 05, 2017 at 08:35:27AM -0700, James H. H. Lampert wrote:
> The box I've been reconfiguring over the past few weeks has a hardware RAID
> controller card, with one mirrored (RAID 1) pair on it at the time of
> installation. Over the weekend, I plugged two more drives into the two empty
> sockets, to create a second mirrored pair, which shows up in Linux as "sdb."
> 
> Initially, it auto-mounted much the same way the external drive
> auto-mounted. I added this line to fstab, and now it mounts at the
> mountpoint of my own choosing:
> 
> > /dev/sdb /media/Auxiliary ext4 errors=remount-ro 0 2
> 
> Everything seems to work fine, except for two things:
> 
> 1. Gnome produces a desktop icon for it. Before I added the fstab entry,
> that icon was a picture of a modern hard drive; now, with the fstab entry,
> it is, of all things, a generic *document* icon. (By contrast, my USB
> external drive shows up with the "Merlin disk pack" icon I gave it years
> ago, and I didn't have to do anything new for Linux that I hadn't done for
> WinDoze; I've got both .ico and .png files of that icon on the new mirrored
> pair.)
> 
> 2. It can be unmounted. I'd rather it not be unmountable.

I suspect both of those are the result of mounting it in /media.

Try this:

mkdir /aux
umount /dev/sdb

edit fstab to:

/dev/sdb /aux ext4 errors=remount-ro 0 2

mount /dev/sdb

and see if the same behavior repeats.

-dsr-

Re: leaving the hibernate state

[Pierre Frenkiel]

On Sat, 2 Sep 2017, Andr? N B wrote:


Have you tried any of these tests?

https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable.git/plain/Documentation/power/basic-pm-debugging.txt?h=v4.12.10



   I tried the first one: it crashed my laptop, and after reboot, I got "recovering 
journal"
   I thing that the problem acrually comes from a too small swap partition. 
I'll increase it, and post the result.

best regards,
--
Pierre Frenkiel

Re: testing, upgrade of openssl libssl1.1 ( 1.1.0f-3 => 1.1.0f-4 )

[Michael Grant]

I upgraded openssl today in my server running testing.  It installed
version 1.1.0f-5.  To my surprise, my mac clients can no longer send
and receive email!

How do I roll back to the previous version of openssl?

"apt-cache showpkg openssl" only shows version 1.1.0f-5.

apt install openssl=1.1.0f-3
Reading package lists... Done
Building dependency tree
Reading state information... Done
E: Version '1.1.0f-3' for 'openssl' was not found

I gather I need to add something to my sources.list to get at the
older versions.

Re: Install Debian 9.1 stretch on Pentium II system: pata_hpt3x2n required?

[deloptes]

David Max wrote:

> cpu Pentium II 400 MHz
> ram 512 MB
> video card ATI RAGE IIC AGP
> hdd #1 6.8 GB Fujitsu (with Windows installed, also a 1GB Linux swap
> partition)
> hdd #2 80 GB Seagate, attached to Highpoint 133SB Rocket Raid (PCI) card

the problem here might be the CPU support in the stock i386 kernel. You'll
probably need to build own kernel. I am not sure how generic the i386
kernel is. If so do

make ARCH=i386 menuconfig

and select Pentium II.

I recently build 4.X for Geode X2 with IDE support. It took time to find the
magic combination of options, but it works great.

regards

Re: testing, upgrade of openssl libssl1.1 ( 1.1.0f-3 => 1.1.0f-4 )

[Gene Heskett]

On Tuesday 05 September 2017 13:40:00 Michael Grant wrote:

> I upgraded openssl today in my server running testing.  It installed
> version 1.1.0f-5.  To my surprise, my mac clients can no longer send
> and receive email!
>
As that is a security related upgrade, I would next push the Mac people 
to match it, or if possible, configure the Macs to use the more secure 
protocol.

> How do I roll back to the previous version of openssl?
>
> "apt-cache showpkg openssl" only shows version 1.1.0f-5.
>
> apt install openssl=1.1.0f-3
> Reading package lists... Done
> Building dependency tree
> Reading state information... Done
> E: Version '1.1.0f-3' for 'openssl' was not found
>
> I gather I need to add something to my sources.list to get at the
> older versions.


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 

Re: Install Debian 9.1 stretch on Pentium II system: pata_hpt3x2n required?

[didier gaumet]

Le 05/09/2017 à 11:34, David Max a écrit :
> Hello
> 
> I am attempting to install Debian 9.1 stretch on an old Pentium II system.
> 
> Setup --
> 
> cpu Pentium II 400 MHz
> ram 512 MB
> video card ATI RAGE IIC AGP
> hdd #1 6.8 GB Fujitsu (with Windows installed, also a 1GB Linux swap
> partition)
> hdd #2 80 GB Seagate, attached to Highpoint 133SB Rocket Raid (PCI) card
> 
> Install process goes well until the stage of searching for disks. Then
> the first disk is detected fine, but 2nd doesn't show up.
> 
> If I pause the installation sequence at disk detection, driver
> pata_hpt37x is shown by lsmod. lspci shows the Highpoint card as "RAID
> bus controller HighPoint Technologies, Inc. HPT302/302N (rev 02)".
> 
> As some background, Knoppix 7.2.0 installs okay on the Seagate drive but
> I seem to have broken it by upgrading some packages (and removing
> libreoffice). On boot, Knoppix complains that graphics cannot be
> initialised, I think because package xserver-xorg-video-ati has been
> unpacked but not installed (i.e., on the Knoppix hdd installation). 
> 
> The aim is to see if this old machine can take on a new life as a
> router. Traffic rates through it are adequate for current needs. If
> Debian install can be completed, I will switch to LXDE for the desktop
> (quite lightweight as I understand).
> 
> If Debian can be installed Knoppix will be replaced but the old Windows
> install on hdd #1 will be retained.
> 
> After a little research, I believe a driver pata_hpt3x2n might be required? 
> 
> All help gratefully received.
> 
> David Max

didier@hp-dm1:~$ grep -i hpt /boot/config*
CONFIG_SCSI_HPTIOP=m
CONFIG_PATA_HPT366=m
CONFIG_PATA_HPT37X=m
# CONFIG_PATA_HPT3X2N is not set
# CONFIG_PATA_HPT3X3 is not set

It seems you have at least to declare the driver either built-in (=y) or
as a module (=m) and rebuild the kernel
(cf
https://kernel-handbook.alioth.debian.org/ch-common-tasks.html#s-common-building)

Re: Strange results with an additional HD -- any idea why?

[deloptes]

James H. H. Lampert wrote:

> What can I do about these things?

use a different mount point

/media is used for auto mounting

Re: testing, upgrade of openssl libssl1.1 ( 1.1.0f-3 => 1.1.0f-4 )

[Michael Grant]

On 5 September 2017 at 19:15, Gene Heskett  wrote:
> On Tuesday 05 September 2017 13:40:00 Michael Grant wrote:
>
>> I upgraded openssl today in my server running testing.  It installed
>> version 1.1.0f-5.  To my surprise, my mac clients can no longer send
>> and receive email!
>>
> As that is a security related upgrade, I would next push the Mac people
> to match it, or if possible, configure the Macs to use the more secure
> protocol.

Any clues how to configure the Mac to use the more secure protocol?
All the software is up-to-date on the Mac side.  I don't see any
obvious option in any of the mail settings on the Mac side.

This is the error I see in the mail logs for both dovecot and sendmail:

dovecot:
TLS handshaking: SSL_accept() failed: error:1417D102:SSL
routines:tls_process_client_hello:unsupported protocol, session=<...>

sendmail:
STARTTLS=server: 0:error:1417D102:SSL
routines:tls_process_client_hello:unsupported
protocol:../ssl/statem/statem_srvr.c:974:

I realize this isn't a MacOS forum but the error message here is on
the Debian side.  Other mail clients like Windows Mail connect fine.

Is there something I can set on Debian side to force this newer
openssl to accept older 1.x connections?

Re: Laptop randomly reboots

[Fungi4All]

> From: rhkra...@gmail.com
> To: debian-user@lists.debian.org
>
> On Tuesday, September 05, 2017 12:46:58 AM Sam Smith wrote:
>> Well I guess I really spoke too soon Just got a random reboot a
>> minute ago... Pretty much at a loss now. The only thing that hasn"t been
>> replaced now is the CPU, battery, and power adapter. I do have a spare
>> battery and power adapter. Guess I"ll try running with those for a month
>> or so and see what happens. Though not really sure how either one of
>> those would make the computer reboot randomly once a month. Laptop is
>> mounted in a docking station. At least for the last 6 months, every
>> reboot was while it was docked. But I do remember times last year
>> sitting at the kitchen table undocked and getting a reboot so I don"t
>> think the docking station is the issue.. But have no idea now :/
>
> I haven"t paid much attention to this before today, but, if I had a problem
> with unexplained reboots, the first area I"d look at is the power supply
> "chain"--that is, the battery, the connections between the battery and the
> laptop, power "glitches" to the house,etc.
>
> I run a laptop with a known bad battery--if I disconnect the power supply it
> crashes in seconds. OTOH, if your battery is good, it serves as a (long
> lived) UPS--you should be able to ride out hours long power glitches.
>
> If I had a known good battery, I might try examining the contacts between the
> battery and the laptop--are they clean / shiny, is the "springy side" still
> "springy", does the battery fit snugly in the case or is there enough 
> clearance
> for it to possibly move and break contact with the laptop? I might even try
> picking up the laptop and reasonably gently shaking it (while running) to see
> if it reboots. (You don"t want to shake it so hard that you damage something
> else.)
>
> If you have a known bad battery, I"d look at the chain from the power outlet
> through the various cords and contacts on both sides of the power adapter.
> Try shaking the power adapter.
>
> Is the power to your house reliable--is there any chance you"ve had brief
> power outages overnight? (In the US, most electric utilities have automatic
> reclosers on the power line circuit breakers--if they detect a fault they
> open, then try closing again after a few seconds (in hopes that either the
> fault has cleared (maybe a shorting tree branch has either blown or burned),
> away, open again if there is still a fault, again try to reclose, after a few
> seconds. In most cases, they try this something like 3 times, with a somewhat
> longer delay before the last retry.
>
> Oh wait, you have (presumably known good) spare battery and power adapter--try
> those.

Power is not the only thing that will shutdown a laptop, but rebooting 
automatically
sounds a bit strange in itself.  Most I have seen take your input to power-up 
unless
a specific instruction is given to reboot.
Are you sure you have not set a hot-key for reboot?
Is it possible a fan or an overheating alarm is not shutting it off? But then 
again, you
are talking about a reboot not a shutdown.
Anything else from the manufacturer that is triggering a reboot?
If the operating system is at fault for ordering a reboot that will be in the 
logs.
If it is a mechanical shutdown it may be instant and not giving enough warnings
to the system to log anything.
So why is it rebooting?  Maybe it goes through an interruption that is not a 
shutdown
so instantly the system is trying to recuperate from the interruption.  
Possibly insufficient
or too much voltage is going through a self protected circuit which causes the 
interruption
and then allows current back through again, giving you the illusion of a reboot.
Sometimes disk drives on their way out draw too much current, or cd/dvd drives, 
which
the power supply can't keep up with which cause an interruption till enough 
juice is
built up to get the motor rolling again.  It could be as simple a small 
capacitor in one of
the drives.  Do your drive/s go to sleep in some state to preserve power?  Look 
at your
power settings and turn off the energy conserving options.
Maybe boot from usb into a live system, disable the drives from the bios and 
run a
usb stick or drive.  If the problem comes up again then it is not drive related 
or operating
system related.  Operating systems are much more likely to freeze up than to 
order
an anauthorized shut/reboot.  I am willing to bet it is a mechanical/electric 
glitch and
not an OS problem.

Unless you have discovered a new black-hole in systemd  (rust never sleeps). ;)

Re: Laptop randomly reboots

[Sam Smith]

On 09/05/2017 06:25 AM, rhkra...@gmail.com wrote:

On Tuesday, September 05, 2017 12:46:58 AM Sam Smith wrote:

Well I guess I really spoke too soon Just got a random reboot a
minute ago... Pretty much at a loss now. The only thing that hasn't been
replaced now is the CPU, battery, and power adapter. I do have a spare
battery and power adapter. Guess I'll try running with those for a month
or so and see what happens. Though not really sure how either one of
those would make the computer reboot randomly once a month. Laptop is
mounted in a docking station. At least for the last 6 months, every
reboot was while it was docked. But I do remember times last year
sitting at the kitchen table undocked and getting a reboot so I don't
think the docking station is the issue.. But have no idea now :/


I haven't paid much attention to this before today, but, if I had a problem
with unexplained reboots, the first area I'd look at is the power supply
"chain"--that is, the battery, the connections between the battery and the
laptop, power "glitches" to the house,etc.

I run a laptop with a known bad battery--if I disconnect the power supply it
crashes in seconds.  OTOH, if your battery is good, it serves as a (long
lived) UPS--you should be able to ride out hours long power glitches.

If I had a known good battery, I might try examining the contacts between the
battery and the laptop--are they clean / shiny, is the "springy side" still
"springy", does the battery fit snugly in the case or is there enough clearance
for it to possibly move and break contact with the laptop?  I might even try
picking up the laptop and reasonably gently shaking it (while running) to see
if it reboots.  (You don't want to shake it so hard that you damage something
else.)

If you have a known bad battery, I'd look at the chain from the power outlet
through the various cords and contacts on both sides of the power adapter.
Try shaking the power adapter.

Is the power to your house reliable--is there any chance you've had brief
power outages overnight?  (In the US, most electric utilities have automatic
reclosers on the power line circuit breakers--if they detect a fault they
open, then try closing again after a few seconds (in hopes that either the
fault has cleared (maybe a shorting tree branch has either blown or burned),
away, open again if there is still a fault, again try to reclose, after a few
seconds.  In most cases, they try this something like 3 times, with a somewhat
longer delay before the last retry.

Oh wait, you have (presumably known good) spare battery and power adapter--try
those.




The power "chain" imo seems good. Never heard of a power adapter going 
bad.. but I do have a spare. The battery I normally use is a smaller 6 
cell I bought off of ebay, it is pretty wore out. It reports 50% 
capacity loss but runs for about 2 hours. I only bought it because the 
stock battery is a bigger 9 cell that is bulky and heavy so around the 
house I just leave the smaller one in. I'll use my spares for now. As 
for household power, it is normally plugged into my UPS (apc smartups, 
not the cheap consumer version).


CPUs seem pretty cheap on ebay, so I could try that. But it seems that 
CPU issues would cause more problems than a random reboot once a month.


One other thing, I do run the 'thinkfan' fan control program. You can 
set fan levels by temp. It was made for thinkpads made probably 10 years 
ago, but still seems to work on the newer models. But I do know the 
embedded controller on the T520 series is different than prior models. 
So perhaps taking control of the fan causes the EC to detect this as a 
fault. And maybe, just maybe, after so many "faults" it will kill power 
to the machine. A wild idea but who knows. I've also since disabled the 
thinkfan service.


Other than these, not sure what else. Kinda strange, but definitely 
annoying.


Regards,

Re: testing, upgrade of openssl libssl1.1 ( 1.1.0f-3 => 1.1.0f-4 )

[Sven Hartge]

Michael Grant  wrote:

> Is there something I can set on Debian side to force this newer
> openssl to accept older 1.x connections?

No, you can't.

Kurt Roeckx, the DD maintaining OpenSSL, patched it in such a way that a
program needs to call a special function of OpenSSL to override the
default minimum TLS-version of TLS1.2.

Problem is: next to no program implements this as of yet.

The Dovecot developers may introduce the needed change in some of the
coming versions, with sendmail I believe you will be out of luck.

First help: Grab an older OpenSSL version from snapshots.debian.org to
get going again.

My solution (other than complaining on the debian-devel mailinglist) was
to recompile OpenSSL with the patch in question removed.

Of course in doing so I burdened myself with tracking any new release of
the OpenSSL packages and recompile them until this situation has been
resolved in some other way.

Grüße,
Sven.

-- 
Sigmentation fault. Core dumped.

Re: testing, upgrade of openssl libssl1.1 ( 1.1.0f-3 => 1.1.0f-4 )

[Michael Grant]

On 5 September 2017 at 20:29, Michael Grant  wrote:
> On 5 September 2017 at 19:15, Gene Heskett  wrote:
>> On Tuesday 05 September 2017 13:40:00 Michael Grant wrote:
>>
>>> I upgraded openssl today in my server running testing.  It installed
>>> version 1.1.0f-5.  To my surprise, my mac clients can no longer send
>>> and receive email!
>>>
>> As that is a security related upgrade, I would next push the Mac people
>> to match it, or if possible, configure the Macs to use the more secure
>> protocol.
>
> Any clues how to configure the Mac to use the more secure protocol?
> All the software is up-to-date on the Mac side.  I don't see any
> obvious option in any of the mail settings on the Mac side.
>
> This is the error I see in the mail logs for both dovecot and sendmail:
>
> dovecot:
> TLS handshaking: SSL_accept() failed: error:1417D102:SSL
> routines:tls_process_client_hello:unsupported protocol, session=<...>
>
> sendmail:
> STARTTLS=server: 0:error:1417D102:SSL
> routines:tls_process_client_hello:unsupported
> protocol:../ssl/statem/statem_srvr.c:974:
>
> I realize this isn't a MacOS forum but the error message here is on
> the Debian side.  Other mail clients like Windows Mail connect fine.
>
> Is there something I can set on Debian side to force this newer
> openssl to accept older 1.x connections?

I could not find any option I could set in the dovecot.conf or the
sendmail.mc file to make libssl accept tls 1.1.  I managed to revert
back libssl to get back to a working situation until the client's get
updated.

I downloaded libssl1.1_1.1.0f-3_amd64.deb

and did:

dpkg -i libssl1.1_1.1.0f-3_amd64.deb

restarted sendmail and dovecot and everyone can now connect.

Re: testing, upgrade of openssl libssl1.1 ( 1.1.0f-3 => 1.1.0f-4 )

[Michael Grant]

On 5 September 2017 at 22:40, Sven Hartge  wrote:
> Michael Grant  wrote:
>
>> Is there something I can set on Debian side to force this newer
>> openssl to accept older 1.x connections?
>
> No, you can't.
>
> Kurt Roeckx, the DD maintaining OpenSSL, patched it in such a way that a
> program needs to call a special function of OpenSSL to override the
> default minimum TLS-version of TLS1.2.
>
> Problem is: next to no program implements this as of yet.
>
> The Dovecot developers may introduce the needed change in some of the
> coming versions, with sendmail I believe you will be out of luck.

Ugh no!

> First help: Grab an older OpenSSL version from snapshots.debian.org to
> get going again.
>
> My solution (other than complaining on the debian-devel mailinglist) was
> to recompile OpenSSL with the patch in question removed.
>
> Of course in doing so I burdened myself with tracking any new release of
> the OpenSSL packages and recompile them until this situation has been
> resolved in some other way.

Thanks for confirming that I did the best thing I could: reinstall the
previous version of libssl.

I was surprised that this problem affected fairly recent MacOS and
Windows Outlook users.  I was also surprised that not many people had
reported this and as I continued to google around for this, I found
only this chain of posts!  And this has been in the wild now for about
10 days.

I'm sure this fix needs to be in there, forcing it on people without
making sure major mailers are going to accept it is just going to
create more problems.  It probably would have been a good idea to put
a loud warning in the log files about this.  The message given by apt
during the update:

  By default the minimum supported TLS version is 1.2. If you still need to
  talk to applications that only support TLS 1.0 you should configure the
  application to set the minimum supported version.

This is highly misleading that it is easy to do this!

Re: testing, upgrade of openssl libssl1.1 ( 1.1.0f-3 => 1.1.0f-4 )

[songbird]

Michael Grant wrote:
...
> I was surprised that this problem affected fairly recent MacOS and
> Windows Outlook users.  I was also surprised that not many people had
> reported this and as I continued to google around for this, I found
> only this chain of posts!  And this has been in the wild now for about
> 10 days.

  you are running Debian Testing/Unstable to see this
change, it won't be pushed to "stable" for a few years
(maybe, we'll see).

  my conversation with the getmail package maintainer
left it at "wontfix".  i have not yet filed a bug with
the openssl maintainer, i'm not sure i will bother.

  in the meantime i'll be stuck at compatible versions
of ssl and the lib and not able to test anything more
recent.


  songbird

Re: Macbook Air - Stretch - getting rather hot

[James Montgomery]

On Tue, Aug 8, 2017 at 7:42 AM, kelsang sherab  wrote:
> The last few days my machine seems to be getting hot more than usual -
> any suggestion of what can i do?

I didn't see any replies to this message. Are you still having this
issue? I have had similar issues with my Macbook Air mid-2011 (Ivy
Bridge). It seemed to have gone away after I set 'powertop' and
cpufreq settings to conserve power. Still, I never hear the fans kick
on an the bottom aluminium is always rather warm.


- Monty

Windows program on Debian

[黃世緯]

My computer is 11 years old, with single-channel ram, 80GB IDE hard 
drive etc.


The most important thing is the virtualisation, is it okay to install 
Windows programme on linux?

Re: Laptop recommendation

[Celejar]

On Tue, 05 Sep 2017 12:13:04 +0100
Dekks Herton  wrote:

> 
> Jonas Hedman  writes:
> 
> > Hello I hope that is not OT for this list.
> >
> > Basically I'm on the hunt for a newish laptop on which I 
> > naturally want
> > to run Debian. I'm a student and I spend most of my daily 

...

> > I really value light weight and good battery life and my upper 
> > bound on prize
> > is somewhere around 875$ (~7K SEK). Memory should be greater 
> > than or
> > equal to 2GB and an sdd would be nice.

...

> > into? I
> > know the older thinkpads has good debian support but if I 
> > remember
> > correctly they are also very heavy.

> Check out ebay for something like a Thinkpad X230 or a Thinkpad 
> T430 with an IPS HD+ screen. I know in sweden availability might 
> be an issue so check out T440 or T450 prices too. 

ThinkPads, besides being generally acclaimed for their excellent build
quality, user-servicability and upgradability, also run Linux quite
well. [This is particularlyl true with regard to the "Professional /
Business" lines [e.g., the T, W, X, and P series] - the "Consumer"
lines [e.g., the E and L series] are somewhat different, and I've seen
conflicting reports as to how similar they are to the acclaimed T
series.]

New ones will likely exceed your budget, but used / refurbished units
are widely available for significantly less, both from resellers and on
auction sites. For years I used a T60, and I currently use a W550s:

https://www.notebookcheck.net/Lenovo-ThinkPad-W550s-Workstation-Review.137989.0.html

For example, T460 and T460s models (the initial letter denotes the
series, the first digit denotes the screen size (14"), the second the
model / processor generation (6th generation, Intel Core), and the 's'
suffix denotes slimmer / lighter versions) with SSDs and IPS screens
are available new / almost new / refurbished on eBay in the $600-$700
(USD) range:

http://www.ebay.com/itm/Lenovo-ThinkPad-T460s-i5-6200U-4GB-128GB-SSD-1080p-IPS-FPR-Win10-Pro-Warranty-/322709906464
http://www.ebay.com/itm/Lenovo-ThinkPad-T460-i5-2-4GHz-256GB-SSD-2020-Warranty-IPS-1080p-/272821920130

Celejar

Re: Laptop recommendation

[Eero Volotinen]

Try buying levono with student discount

3.9.2017 4.46 ip. "Jonas Hedman"  kirjoitti:

> Hello I hope that is not OT for this list.
>
> Basically I'm on the hunt for a newish laptop on which I naturally want
> to run Debian. I'm a student and I spend most of my daily outandabout
> computer time reading pdfs, writing LaTeX docs, surfing and doing some
> light coding, no heavy duty stuff in other words. I use i3vm and I
> generally like to keep things as light and minimal as feasibly possible.
>
> I really value light weight and good battery life and my upper bound on
> prize
> is somewhere around 875$ (~7K SEK). Memory should be greater than or
> equal to 2GB and an sdd would be nice.
>
> I know very little about hardware and I'm not quite sure where or how to
> start looking. I though the Lenovo Thinkpad 13 looked promising but it's a
> bit
> over my price target. Could chromeboosk be something to look into? I
> know the older thinkpads has good debian support but if I remember
> correctly they are also very heavy.
>
> As a small reference: I made due with a Asus EEE PC 1005HA for a number
> of years and it generally worked well for my purposes but it died and
> surfing was a pain so I'm looking to upgrade a bit from that.
>
> Any suggestions?
>
> Best regards
> --
> Jonas Hedman
>
> PGP: 8F72 C5BE AAFA B4BA 8F46 9185 5C39 89E0 616B B08C
>

Re: Recommended editor for novice programmers?

[kamaraju kusumanchi]

On Sat, Sep 2, 2017 at 2:21 PM, Tom Browder  wrote:
> My Linux user group is setting up one desktop computer and one laptop
> computer for lending to our local library as an educational resource for
> folks who want to explore what Linux is all about.  We are using Debian 9
> for now.
>

Great initiative. May I know the location of this library?

> I am open to any suggestions for standard packages we should add. I have
> already installed gcc and friends as well as Scilab, R, Perl 6, and some
> other stuff, including emacs.
>

Two things:
1) Please make it "very easy" to request and install new packages. No
matter what set of packages are installed at first, there will be a
few packages that a few users would like to see. So putting a big fat
notice as to how to get new software installed would go a long way.

2) Please install python3 and the relevant libraries such as numpy, pandas etc.,

> I would especially appreciate other ideas for programming editors for novice
> programmers.

I recommend vim and zim. The former is great for writing code and the
latter is great for planning and organization.

-- 
Kamaraju S Kusumanchi | http://raju.shoutwiki.com/wiki/Blog

Re: Suitable text editor [NOT word processor] or workaround?

[kamaraju kusumanchi]

On Fri, Mar 17, 2017 at 7:37 AM, Gene Heskett  wrote:
> On Friday 17 March 2017 05:49:30 Jonathan Dowland wrote:
>
>> On Thu, Mar 16, 2017 at 09:54:42AM -0400, Gene Heskett wrote:
>> > gedit has caused me to have to start over again, 3 times, on a 600+
>> > line configuration file for machine control on several occasions. As
>> > the file it trashed was 6 months worth of adding new features to
>> > that machines control, that was not appreciated. 1st time is a wtf?,
>> > 2nd time sends me a message, 3rd time I shut it down w/o the save
>> > and never ran it again.
>>
>> Whilst this is very frustrating, one other thing it tells me is you
>> need to set up a backup system (although you are quite right to
>> abandon gedit based on your experiences)
>
> I do, I am a very longtime user of amanda, since 1998 TBE. I've never
> contaminated my thinking with a windows install, going from a coco3 with
> nitros9, to amigados, to linux at RH-5.0. amanda does all 5 of these
> machines every night at about 1:20 AM and it got used all 3 times. The
> problem with that is that the status of the backup was 15 to 20 hours
> old, so I lost that days work and had to re-invent that particular batch
> of work.


FWIW, I also thought that you lost 6 months of work based on what you
wrote initially. Happy to hear that the damage is much less. But even
that can be very frustrating, right?

Along with the backups, may I suggest you to store all your work in
version control such as git so that even if the editor crashes, you
can recover everything up until the last commit.

-- 
Kamaraju S Kusumanchi | http://raju.shoutwiki.com/wiki/Blog

mailx(1) core dump, Debian 8, amd64

[John Conover]

Anytime mailx is envoked, it does a core dump:

mail: mu_wordsplit failed: missing closing quote
Segmentation fault

Any suggestions?

Thanks,

John

-- 

John Conover, cono...@rahul.net, http://www.johncon.com/

Re: Macbook Air - Stretch - getting rather hot

[Mario Castelán Castro]

On 05/09/17 20:24, James Montgomery wrote:
> On Tue, Aug 8, 2017 at 7:42 AM, kelsang sherab  wrote:
>> The last few days my machine seems to be getting hot more than usual -
>> any suggestion of what can i do?
> 
> I didn't see any replies to this message. Are you still having this
> issue? I have had similar issues with my Macbook Air mid-2011 (Ivy
> Bridge). It seemed to have gone away after I set 'powertop' and
> cpufreq settings to conserve power. Still, I never hear the fans kick
> on an the bottom aluminium is always rather warm.

Another thing that may help is to configure a CPU usage indicator widget
in a task bar so as to detect if any program is using an excessive
amount of computing power.

-- 
Do not eat animals; respect them as you respect people.
https://duckduckgo.com/?q=how+to+(become+OR+eat)+vegan



signature.asc
Description: OpenPGP digital signature

failure of cheese with a Microsoft LifeCam NX-6000

[peter]

Any insights about these messages when cheese fails?

guest@imager:~$ cheese

(cheese:1572): Gtk-WARNING **: Theme parsing error: cheese.css:7:35: 
  The style property GtkScrollbar:min-slider-length is deprecated and 
  shouldn't be used anymore. It will be removed in a future version

(cheese:1572): cheese-WARNING **: Device '/dev/video0' cannot capture 
  in the specified format: gstv4l2object.c(3583): 
  gst_v4l2_object_set_format_full (): 
  
/GstCameraBin:camerabin/GstWrapperCameraBinSrc:camera_source/GstBin:bin28/GstV4l2Src:v4l2src1:
Tried to capture in BGR3, but device returned format MPEG

(cheese:1572): Cogl-WARNING **: driver/gl/cogl-pipeline-opengl.c:932: 
GL error (1280): Invalid enumeration value

What is a "Theme parsing error"?
Is there a simple fix for the format problem?  
What is meant by 'Invalid enumeration value"?

Thanks, ... Peter E.

-- 

123456789 123456789 123456789 123456789 123456789 123456789 123456789
Tel: +1 360 639 0202  Pender Is.: +1 250 629 3757
http://easthope.ca/Peter.html  Bcc: peter at easthope. ca

Re: mailx(1) core dump, Debian 8, amd64

[kamaraju kusumanchi]

On Wed, Sep 6, 2017 at 12:10 AM, John Conover  wrote:
>
> Anytime mailx is envoked, it does a core dump:
>
> mail: mu_wordsplit failed: missing closing quote
> Segmentation fault
>
> Any suggestions?
>

It would be nice to document the problem first with full information
in order to identify what is happening and provide you some
suggestions. What is the output of the following commands.

which mailx
readlink -f /usr/bin/mailx
dpkg -l bsd-mailx
inxi -r

Note:- If inxi is not already installed on your system, you can
install it by running 'apt-get install inxi' with root privileges.

-- 
Kamaraju S Kusumanchi | http://raju.shoutwiki.com/wiki/Blog

Re: Recommended editor for novice programmers?

[Nick Boyce]

On Sat, 02 Sep 2017 18:21:07 +
Tom Browder  wrote:

> I would especially appreciate other ideas for programming editors for
> novice programmers.

If you really want to have a *simple* non-GUI (i.e. terminal) screen-mode 
editor available that novice programmers who are refugee users from Windows 
might find comfortable, you could consider 'joe' ("Joe's Own Editor" in classic 
Un*x-world recursive acronym naming style), which provides basically a Wordstar 
control-key driven modeless editing experience.  Yes, I did say "Wordstar" ... 
you may need to get off my lawn.

https://packages.debian.org/stretch/joe
http://joe-editor.sourceforge.net/

I don't like to confess to my august and more sophisticated colleagues here how 
much code I've written using joe - albeit in the simpler languages (a variety 
of Bash scripts, Perl, C, HTML and similar). There is some syntax highlighting, 
but no code-completion or compiler integration or the other trinkets that come 
with proper IDEs.  It is however, small, fast and reliable - it hasn't had a 
new feature in *years* because for it's intended use-cases it's 
*feature-complete* !

It's one of the first things I install on any Linux or *BSD system.

I don't want to provoke any religious war here, and sorry if I offend anybody, 
but: emacs is ridiculously heavy-weight, and I can never remember whether exit 
is Ctrl_C Ctrl_X or Ctrl_X Ctrl_C (yes practice would help) - and vi's power 
makes light work of many tasks but it's as user-friendly as a cornered rat ... 
novices usually remember their first time trying to find out how to exit with a 
genuine shudder.  A frequent vi moment for those familiar with modeless editors 
is to enter 'insert mode' (when you figure out how), type some text, and then 
try to use the arrow keys to move to a different line without remembering to 
first exit insert mode - depending on vi version, terminal emulation, Un*x 
flavour, phase of the Moon etc., the effect of this is that a whole bunch of 
weird character sequences get entered instead of cursor control, which you then 
spend the next 10 minutes removing again.  Ugh.

(Yes, I do use more sophisticated GUI IDEs for anything serious, but that's not 
what OP asked for.  Also, I do realise vim is much better than vi.)

Cheers,
Nick
-- 
Never FDISK after midnight.

Editor survival [Was: Recommended editor for novice programmers?]

[Erik Christiansen]

On 06.09.17 05:31, Nick Boyce wrote:
> I don't like to confess to my august and more sophisticated colleagues
> here how much code I've written using joe - albeit in the simpler
> languages (a variety of Bash scripts, Perl, C, HTML and similar).
> There is some syntax highlighting, but no code-completion or compiler
> integration or the other trinkets that come with proper IDEs.  It is
> however, small, fast and reliable - it hasn't had a new feature in
> *years* because for it's intended use-cases it's *feature-complete* !
> 
> It's one of the first things I install on any Linux or *BSD system.

In my decades of leading software teams, one thing I did not do is ask
"What editor do you use?", even in employment interviews. In my
experience, a programmer is most productive using the editor with which
he's most proficient. End of story.

> ... and vi's power makes light work of many tasks but it's
> as user-friendly as a cornered rat ... novices usually remember their
> first time trying to find out how to exit with a genuine shudder.

On the three occasions I've had to extract a marsupial possum from our
chimney (they're like a cat on steroids), I've armed myself with thick
leather gloves and grim determination. For vim, a cheat-sheet suffices,
and :help " or google do explain.

> A frequent vi moment for those familiar with modeless editors is to
> enter 'insert mode' (when you figure out how), type some text, and
> then try to use the arrow keys to move to a different line without
> remembering to first exit insert mode - depending on vi version,
> terminal emulation, Un*x flavour, phase of the Moon etc., the effect
> of this is that a whole bunch of weird character sequences get entered
> instead of cursor control, which you then spend the next 10 minutes
> removing again.  Ugh.

That's an xterm error, as the arrows simply produce motion even in
Insert-mode, if that's properly set up.

But Vim's Insert-mode/Normal-mode modality still requires user
awareness, either through memory or looking at the status line, with a
"set showmode" in ~/.vimrc - unless you also add something like:

" These days I expect to be out of insert mode, after a vertical move:
inoremap  ^[
inoremap  ^[

(The ^[ is , entered as  in insert mode.)

Now the Insert-mode exit is automated, so long as you use the arrow keys
for moves. (If you use hjkl for speed, then you're not a novice, by
definition.) I've left  and , so I stay in Insert-mode
while on the line. It works for me, but you could remap them too.

To avoid the need to glance to the bottom of the window to check mode, I
also add:

" Cursor Appearance and behaviour:
" (Insert_Mode == Green, Normal_Mode == Red)
if &term =~ "xterm"
   let &t_SI = "\]12;yellow\x7"
   let &t_EI = "\]12;green\x7"
endif

So the cursor itself is a reminder. (I haven't yet found a way to have
it make coffee though, so that I always have my eyes open.)
(Colours chosen for yellow text on darkslategrey background, set in the
xterm. YTMV)

> (Yes, I do use more sophisticated GUI IDEs for anything serious, but
> that's not what OP asked for.  Also, I do realise vim is much better
> than vi.)

Even without integration in an editor, in *nix you can Control-Z out to
the command line, run make, and fg back in. I too would use whatever
editor my fingers remember.

Erik

Re: Recommended editor for novice programmers?

[Ben Finney]

Nick Boyce  writes:

> I don't like to confess to my august and more sophisticated colleagues
> here how much code I've written using joe - albeit in the simpler
> languages (a variety of Bash scripts, Perl, C, HTML and similar).

Joe is a fine text editor licensed as free software. I say that as
someone who has decided I don't like to use it myself.

To paraphrase someone else paraphrasing Voltaire: I may disagree with
your choice of text editor, but I will defend to the death your right to
use it.

So, go ahead and admit how much you like Joe and how much code it has
allowed you to write. If anyone has a problem with your usage of Joe for
programming, they have me to answer to :-)


Here's where I need to object:

> I don't want to provoke any religious war here, and sorry if I offend
> anybody, but:

That doesn't alter the fact that you've disparaged programs in terms
that state an absolute problem inherent to the program. This is not
helpful, because it implies that people who choose those programs are
wrong and should be disparaged themselves.

For example:

> emacs is ridiculously heavy-weight

That's an absolute statement of objective fact. Do you think it is true
for everyone? You have expressed it as though you do.

You are, in this expression, saying that people who use Emacs deserve
ridicule because Emacs is so obviously heavy-weight they should be
embarrassed to use it. That's needlessly hostile.

You can, instead, say what *you don't like* about the program:

> and I can never remember whether exit is Ctrl_C Ctrl_X or Ctrl_X
> Ctrl_C (yes practice would help)

That's great! Thank you for stating your position here in terms of what
you are and are not able to do. This doesn't judge anyone else as
inferior.

> vi's power makes light work of many tasks but it's as user-friendly as
> a cornered rat ... novices usually remember their first time trying to
> find out how to exit with a genuine shudder.

This gets too far into stating objective fact. How do you know the
“usual” experience of novice Vi users, have you surveyed a statistically
powerful sample?


I appreciate that it can be fun to rant about difficulties using
programs, and Emacs and Vi are favourite topics of this sort. We can,
and should, do so without also dismissing other people as inferior.

When you acknowledge the possibility of provoking offense, it is on your
shoulders to either express yourself in ways that *don't* implicitly
disparage people with different preferences.

Either that, or don't provoke at all :-)

Have fun, everyone, using whatever text editors make your life better.

-- 
 \   “My business is to teach my aspirations to conform themselves |
  `\  to fact, not to try and make facts harmonise with my |
_o__)   aspirations.“ —Thomas Henry Huxley, 1860-09-23 |
Ben Finney

Re: Suitable text editor [NOT word processor] or workaround?

[Gene Heskett]

On Wednesday 06 September 2017 00:09:31 kamaraju kusumanchi wrote:

> On Fri, Mar 17, 2017 at 7:37 AM, Gene Heskett  
wrote:
> > On Friday 17 March 2017 05:49:30 Jonathan Dowland wrote:
> >> On Thu, Mar 16, 2017 at 09:54:42AM -0400, Gene Heskett wrote:
> >> > gedit has caused me to have to start over again, 3 times, on a
> >> > 600+ line configuration file for machine control on several
> >> > occasions. As the file it trashed was 6 months worth of adding
> >> > new features to that machines control, that was not appreciated.
> >> > 1st time is a wtf?, 2nd time sends me a message, 3rd time I shut
> >> > it down w/o the save and never ran it again.
> >>
> >> Whilst this is very frustrating, one other thing it tells me is you
> >> need to set up a backup system (although you are quite right to
> >> abandon gedit based on your experiences)
> >
> > I do, I am a very longtime user of amanda, since 1998 TBE. I've
> > never contaminated my thinking with a windows install, going from a
> > coco3 with nitros9, to amigados, to linux at RH-5.0. amanda does all
> > 5 of these machines every night at about 1:20 AM and it got used all
> > 3 times. The problem with that is that the status of the backup was
> > 15 to 20 hours old, so I lost that days work and had to re-invent
> > that particular batch of work.
>
> FWIW, I also thought that you lost 6 months of work based on what you
> wrote initially. Happy to hear that the damage is much less. But even
> that can be very frustrating, right?
>
> Along with the backups, may I suggest you to store all your work in
> version control such as git so that even if the editor crashes, you
> can recover everything up until the last commit.

I have considered something along those lines, but mentally I can't seem 
to make the coupling between a single file that has to be ready to go 
anytime I run linuxcnc -l (where the -l says to use the same config it 
used the last time) and a git database that usually has to be compiled 
before its capable of running.  Normally we keep a separate directory 
for each machine configuration, although there may be N parallel 
directories as the configuration is developed. A milling machine may 
start out with 3 axis's of motion, and may have up to 9 degrees of 
movement freedom as accessories such as a rotating table, which may be 
mounted to the xy table with its axis of movement aligned with any of 
the other 3 axis's.  Or the actual cutting spindle may get mounted in 
such a manner as the be tiltable in both directions so it can reach into 
the nooks and crannies.  This is generally geared together 
electronically so that the cutting tool is motionless on the workpiece 
as it tilts unless you specifically drive it otherwise. This of course 
requires very carefull tool length calibrations to pull off well, but 
the math to sub-micron accuracy is already modularized.

The total configuration generally is not a single file, usually broken up 
according to its order in the programs bootup, first being the basic 
config, then the first of what could be 2 or 3 .hal files, some of which 
can't be run until the gui is started, then once the gui is drawn, more 
gui for accessory tally's, spindle speed/direction, and dials to replace 
the cranks that no longer exist, usually written in xml or pyvcp, or 
gladevcp is done, which adds the "hal pins" that connect the machine gui 
to the machine. It can get complex.  This most recent lathe has over 
1200 lines of code just in the configuration files.  And I still do not 
have any coolant or lube facilities under control.  Stuff I have yet to 
build or buy. :)



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