Re: pointers to material for using netbook's wireless as access point

[Joel Rees]

I now have connection for both the wireless and the netbook that is acting
as the AP. I took out the bridge entirely, quit trying to play with
port forwarding,
just used dead simple setup. dnsmasq was the only missing piece, if I had
not been focusing on bridging.  Bridging is probably for the other direction.

But the wireless is pretty slow, so I'm not sure I'm finished.

I have to go take care of some family business, when I'm done I'll
post the details.

But it's really pretty simply. I was just working too hard.

--
Joel Rees

Randomly ranting:
http://reiisi.blogspot.com

Re: Downgrading specific packages with apt

[Jimmy Johnson]

On 06/09/2017 09:44 PM, solitone wrote:

I am on Debian 9 (scratch), and I have a MacBook Pro 12,1 with retina display.

Few days ago I upgraded Google Chrome from version 58 to 59:
google-chrome-stable:amd64 (58.0.3029.110-1, 59.0.3071.86-1)
This new version no longer supports HiDPI. As a result everything in Chrome is
so small that I would need a magnifying glass!

Not a great issue though, since I usually work with Chromium, not Chrome. But
today apt has proposed the very same upgrade for Chromium too:

chromium/testing 59.0.3071.86-1 amd64 [upgradable from: 58.0.3029.96-1]
chromium-driver/testing 59.0.3071.86-1 amd64 [upgradable from: 58.0.3029.96-1]
chromium-l10n/testing 59.0.3071.86-1 all [upgradable from: 58.0.3029.96-1]
chromium-shell/testing 59.0.3071.86-1 amd64 [upgradable from: 58.0.3029.96-1]

I suspect if I upgrade I would end up having the same HiDPI issue with
Chromium. So for the time being I have put on hold those four packages. But
I'd like to test whether this is true. What should I do to be able to
downgrade to version 58.0.3029.96-1 in case 59.0.3071.86-1 does in fact break
HiDPI? I always have some difficulties when I need to downgrade with apt. This
time I want to be prepared.

Thanks!



I've never downgrade using apt, but with synaptic it's not too hard, 
just go to the menu, package, force version and then to edit, fix 
packages.  The fix package part really only checks to see no package is 
broken, but you must do it or risk losing all your current settings and 
have to start over.


Hope that helps.
--
Jimmy Johnson

Ubuntu 16.04 - KDE Plasma 5.8.7 - Intel G3220  - EXT4 at sda19
Registered Linux User #380263

Re: ifconfig network resolution

[Pascal Hambourg]

Le 10/06/2017 à 03:57, Joel Rees a écrit :

Experimenting from the command line, I find myself puzzled about the
arguments for ifconfig.

Reading the manual, it would appear that the arguments for ifconfig
should be something like this:

ifconfig eth0 netmask 255.255.255.224 netmask 255.255.255.224
broadcast 10.19.23.223 10.19.23.94

But the command returns with

SIOCSIFNETMASK: Can't allocate this address.
SIOCSIFBRDADDR: Can't allocate this address.


You set the mask twice.
The address must be set before the mask and broadcast address.

Re: ifconfig network resolution (Re: pointers to material for using netbook's wireless as access point)

[Joel Rees]

Sorry my typing is so lame.

On Sat, Jun 10, 2017 at 10:57 AM, Joel Rees  wrote:
> Experimenting from the command line, I find myself puzzled about the
> arguments for ifconfig.
>
> Reading the manual, it would appear that the arguments for ifconfig
> should be something like this:
>
> ifconfig eth0 netmask 255.255.255.224 netmask 255.255.255.224
> broadcast 10.19.23.223 10.19.23.94

   sudo ifconfig eth0 netmask 255.255.255.224 broadcast 10.19.23.223
10.19.23.194

> But the command returns with
>
> SIOCSIFNETMASK: Can't allocate this address.
> SIOCSIFBRDADDR: Can't allocate this address.
>
> If I repeat the command, it gives no errors, but the netmask and
> broadcast address end up full class A (255.0.0.0 and 10.255.255.255).

But if I repeat it often enough (with legitimate parameters), it seems to
decide to believe me, and sets the nic to the requested netmask and
broadcast address.

Once or twice may have been bad parameters, but I have just hit the
up arrow and it goes ahead and sets the parameters to what I said on
the second try.

> Anyone have an idea what's happening?

-- 
Joel Rees

One of these days I'll get someone to pay me
to design a language that combines the best of Forth and C.
Then I'll be able to leap wide instruction sets with a single #ifdef,
run faster than a speeding infinite loop with a #define,
and stop all integer size bugs with my bare cast.

More of my delusions:
http://reiisi.blogspot.com/2017/05/do-not-pay-modern-danegeld-ransomware.html
http://reiisi.blogspot.jp/p/novels-i-am-writing.html

Downgrading specific packages with apt

[solitone]

I am on Debian 9 (scratch), and I have a MacBook Pro 12,1 with retina display.

Few days ago I upgraded Google Chrome from version 58 to 59:
google-chrome-stable:amd64 (58.0.3029.110-1, 59.0.3071.86-1)
This new version no longer supports HiDPI. As a result everything in Chrome is 
so small that I would need a magnifying glass!

Not a great issue though, since I usually work with Chromium, not Chrome. But 
today apt has proposed the very same upgrade for Chromium too:

chromium/testing 59.0.3071.86-1 amd64 [upgradable from: 58.0.3029.96-1]
chromium-driver/testing 59.0.3071.86-1 amd64 [upgradable from: 58.0.3029.96-1]
chromium-l10n/testing 59.0.3071.86-1 all [upgradable from: 58.0.3029.96-1]
chromium-shell/testing 59.0.3071.86-1 amd64 [upgradable from: 58.0.3029.96-1]

I suspect if I upgrade I would end up having the same HiDPI issue with 
Chromium. So for the time being I have put on hold those four packages. But 
I'd like to test whether this is true. What should I do to be able to 
downgrade to version 58.0.3029.96-1 in case 59.0.3071.86-1 does in fact break 
HiDPI? I always have some difficulties when I need to downgrade with apt. This 
time I want to be prepared.

Thanks!

Re: pointers to material for using netbook's wireless as access point

[Joel Rees]

I posted the following to Randy, yesterday, intending it to go to the list.
I'll post it back to the list (with Randy's permission), with a bit of further
comment:

> On Friday, June 09, 2017 02:14:17 AM Joel Rees wrote:
>> (With aplogies for html mail)
>>
>> 2017/06/08 23:46 :
>> > On Thursday, June 08, 2017 10:00:17 AM Joel Rees wrote:
>> > > Maybe I'm misunderstanding that wiki, but it seems to be describing
>> > > this
>> > >
>> > > kind of setup:
>> > >WAN <--> foreign AP <-wireless-> debian box <-> more devices
>> > >
>> > > But what I'm wanting is
>> > >
>> > >WAN <-wired-> router/modem <-wired-> debian netbook AP <-> more
>>
>> devices
>>
>> > Thank you for including the above "sketch" which finally let me
>>
>> understand what
>>
>> > you are looking for.
>> >
>> >  I doubt that I can help, but I'll think about it--my setup is somewhat
>> >
>> > similar to what you describe except that the "debian notebook AP" is
>>
>> replaced
>>
>> > by a commericial wireless AP (but, as your sketch shows, wired to my
>>
>> router /
>>
>> > modem).
>> >
>> > Aside: I wouldn't think it should be very difficult, since it seems to be
>>
>> off-
>>
>> > the-shelf functionality you can buy, but ...
>>
>> Thanks for looking at it.
>>
>> Just for the record, I am presently typing on my tablet, connected wireless
>> through the
>> software AP in the netbook (which is the reason for the html). It's
>> transparent, so the
>> modem at the wall sees it as if the wireless is an extension of the modem's
>> (wired)
>> network.

And this is kind of tricky, because it seems to be a bit dependent on the
weather, whether it works or no.

Or, rather, I have since installed dnsmasq and removed and re-installed
network-manager (stupid thing gets in the way), and I'm not connecting
any more.

>> What I'm trying to get is connection on the netbook itself. I can't even
>> ping the modem
>> from the netbook, because the bridge owns the netbook's only ethernet port.
>> So I can't
>> work on anything that requires the network while the kids are playing.
>>
>> I read hints here and there about how to do it, but I keep hitting walls.
>> (And learning
>> things. :)
>>
>> I had it serving dhcp and dns over the wireless a bit back, but I couldn't
>> get outside the
>> modem with either the wireless or the netbook. That was not, of course,
>> bridged. I think.

I was able to confirm that it was bridged.

>> 8-/
>>
>> It would be cheaper, time-wise, to buy a portable access point, of course,
>> and I will probably do so.
>> But I'm finally getting an idea of how a bridge really works, so it's also
>> worth the
>> education.
>>
>> Joel Rees

Randy replied,

On Fri, Jun 9, 2017 at 9:43 PM,   wrote:
> This provides me with even more understanding of what you're looking for and
> the problems you're having--up until I read this, I didn't understand that (1)
> the netbook is already working "properly" as a wireless access point (in that
> your tablet wirelessly connects via it), but (2) the problem is that the
> netbook apparently can't be used simultaneously as a WAP and to provide its
> own connection to the Internet.
>
> I would suggest that you post this to the list--I think others may have the
> same misunderstanding as to your goal and the current problem, and once
> understanding this, may be able to help you.
>
> If you want, I can post it to the list as well.
>
> I would tend to say that your problem is not a common problem, so I suspect
> most documentation won't be helpful.  I also tend to doubt that you want the
> netbook to be in bridge mode, but I don't know that for sure.  IIRC (and I
> can't easily check at the moment), I believe my (commercial) WAP is working as
> a full fledged router, in the sense that it creates a new network (with a
> different IP address and provides DHCP functionality).  I'll have to turn my
> WAP on later and see what IP address my phone gets to see if it is on my "main
> LAN" (i.e., wired) or if it is an IP on a different network.
>
> Good luck,
> Randy Kramer
>

Indeed, that's what I want to do. But I'm not smart enough to set the wireless
up myself, so I was hoping hostapd would do that.

On Fri, Jun 9, 2017 at 10:30 PM,   wrote:
> Update: I just turned on my WAP and my cell phone, and found that the cell
> phone gets an IP address in the 10. network (specifically, in this case
> 10.0.0.2), while my "main" (wired) LAN is on the 192.168.1 network (the server
> is 192.168.1.1), thus confirming my suspicion that the WAP is not in bridge
> mode, but is a "full" network server with services such as DHCP.

That is the usual case for portable access points in Japan. (And, unless you
ask, they don't tell you that you can log into the admin page and set the
number of devices that can connect wirelessly to the AP.)

> I guess if you're in bridge mode, the IP address of your tablet is coming from
> your modem / router.

That is correct. It was getting addresses in the range the wall modem was
providing. A coupl

Re: Reply-to-all or reply-to-list again

[Joel Rees]

On Sat, Jun 10, 2017 at 12:57 AM, Nicolas George  wrote:
> Le primidi 21 prairial, an CCXXV, The Wanderer a écrit :
>> Disagreed. This results in sending extra copies to people who are
>> subscribed to the list, which is incorrect.
>
> Not if the list is properly configured.
>
> Debian's lists are badly configured, it results in burden to all users,
> but the users should reject that burden. I do.
>
> I will not answer Gene's message, he obviously missed the whole point of
> my mail.
>
> Regards,
>
> --
>   Nicolas George

There was a time when mailing lists were e-mail, and newsgroups were
newsgroups. Newsgroups have unfortunately mostly gone the way of the
dodo, and mailing list participants want the mailing lists to behave like
newsgroups.

And Now Google Doesn't Think Users Should Ever Set Headers.

Progress is progress.

-- 
Joel Rees

One of these days I'll get someone to pay me
to design a language that combines the best of Forth and C.
Then I'll be able to leap wide instruction sets with a single #ifdef,
run faster than a speeding infinite loop with a #define,
and stop all integer size bugs with a bare cast.

More of my delusions:
http://reiisi.blogspot.com/2017/05/do-not-pay-modern-danegeld-ransomware.html
http://reiisi.blogspot.jp/p/novels-i-am-writing.html

ifconfig network resolution (Re: pointers to material for using netbook's wireless as access point)

[Joel Rees]

Experimenting from the command line, I find myself puzzled about the
arguments for ifconfig.

Reading the manual, it would appear that the arguments for ifconfig
should be something like this:

ifconfig eth0 netmask 255.255.255.224 netmask 255.255.255.224
broadcast 10.19.23.223 10.19.23.94

But the command returns with

SIOCSIFNETMASK: Can't allocate this address.
SIOCSIFBRDADDR: Can't allocate this address.

If I repeat the command, it gives no errors, but the netmask and
broadcast address end up full class A (255.0.0.0 and 10.255.255.255).

Anyone have an idea what's happening?

-- 
Joel Rees

One of these days I'll get someone to pay me
to design a language that combines the best of Forth and C.
Then I'll be able to leap wide instruction sets with a single #ifdef,
run faster than a speeding infinite loop with a #define,
and stop all integer size bugs with a bare cast.

More of my delusions:
http://reiisi.blogspot.com/2017/05/do-not-pay-modern-danegeld-ransomware.html
http://reiisi.blogspot.jp/p/novels-i-am-writing.html

Re: Predictable Network Interface Names prevents WiFi connections.

[Marcos Raúl Carot]

Hi Miguel,

Did you ever get an answer about this?

I can't get Network Manager (from KDE) to connect if the predictable
network interface names are enabled.

Cheers,

Marcos

Re: Debian installation issues

[Michael Fothergill]

On 9 June 2017 at 20:59, Fungi4All  wrote:

> Here is some relevant reading of installing linux system besides Win8 and
> in some cases the same problem exists on Win 10.
> https://askubuntu.com/questions/221835/installing-ubuntu-alongside-a-pre-
> installed-windows-with-uefi
>


​I read through some of this.  If I understood it correctly, if you buy a
machine that comes with e.g. Windows 10 installed for you then this secure
boot feature would make it difficult to boot and install certain Linux
distributions - but some versions of Ubuntu might be OK apparently.

But if you would buy such a machine, do you not also get the Windows key
codes for the OS...?

If you do, then could you not just back up the work files on the
installation and then uninstall Windows and reinstall it with the secure
boot feature turned off and then install the Linux distro of your choice?

When I get a new PC I specifically request that it has no operating system
on it and then install everything from scratch.

I have not encountered this problem as yet.

Cheers

MF









>
> Ok, MS did what they did, but manufacturers accepted this and incorporated
> it into their product which you pay.
> It is like paying for an MS only system without knowing it.
>

Re: Debian installation issues

[Fungi4All]

Here is some relevant reading of installing linux system besides Win8 and in 
some cases the same problem exists on Win 10.
https://askubuntu.com/questions/221835/installing-ubuntu-alongside-a-pre-installed-windows-with-uefi

Ok, MS did what they did, but manufacturers accepted this and incorporated it 
into their product which you pay.
It is like paying for an MS only system without knowing it.

Re: Reply-to-all or reply-to-list again

[The Wanderer]

On 2017-06-09 at 11:57, Nicolas George wrote:

> Le primidi 21 prairial, an CCXXV, The Wanderer a écrit :
>
>> Disagreed. This results in sending extra copies to people who are
>> subscribed to the list, which is incorrect.
> 
> Not if the list is properly configured.

What configuration would avoid this result, when using "Reply to All"?

When using "Reply", a mailing-list configuration which sets Reply-To to
point back to the list (presumably in addition to any Reply-To set by
the user, or only when the user did not set such) would avoid that
result, but I've seen arguments presented in the past that it also
results in undesirable results of other types in some cases; although I
didn't retain those arguments in full, I was also not able to come up
with effective counterarguments at the time.

When using "Reply to All", at least with some mail clients, Reply-To
will be ignored - and in fact, I think I might argue that any mail
client which does not ignore Reply-To in the case of "Reply to All" is
broken, since that is failing to do what the user explicitly requested.

Even if using a mail client which does treat Reply-To as overriding
"Reply to All", that just means that the user is required to keep track
of which mailing lists set Reply-To (and use "Reply to All" in those
cases), which don't (and use some other reply method in those cases),
and also which specific messages on some lists have changed the Reply-To
configuration (and then have to decide whether to override that
manually, including potentially copying some or all of the addresses
from the original mail by hand since the mail client won't do it for you).

*Maybe* that's better than any of the known alternatives as a practical,
real-world matter, but it hardly seems anything remotely close to ideal,
in my view.

> Debian's lists are badly configured, it results in burden to all users,
> but the users should reject that burden. I do.

And your chosen method of avoiding this burden seems to result both in
burden to you (in the form of having to modify message headers every
time you compose a reply) and in burden to others (in this case, in the
form of me receiving two copies of your reply, rather than only the copy
sent through the mailing list).

I'm not sure I'd agree that that's better.

-- 
   The Wanderer

The reasonable man adapts himself to the world; the unreasonable one
persists in trying to adapt the world to himself. Therefore all
progress depends on the unreasonable man. -- George Bernard Shaw



signature.asc
Description: OpenPGP digital signature

Re: Debian installation issues

[Celejar]

On Fri, 09 Jun 2017 10:47:25 +0200
"Thomas Schmitt"  wrote:

> Hi,
> 
> Fungi4All wrote:
> > try from linux to burn an
> > ms-win-installation image, chances are that you will fail despite of what
> > way you may try to do so.
> 
> Are there any such images available for free and legally safe to have
> and to talk about ?

Well, Microsoft will let you download ISOs to do a Windows install,
although you'll need a product key to activate it:

https://www.microsoft.com/en-us/software-download/windows8ISO

I went through this procedure a while back when I bought a ThinkPad
with Windows installed. I installed Debian on the bare metal, and then
used the ISO (burned onto a USB, with a simple dd, IIRC) to reinstall
Windows in a qemu-kvm VM.

Celejar

Re: Debian installation issues

[Thomas Schmitt]

Hi,

Dan Ritter wrote:
> https://www.microsoft.com/en-us/software-download/windows10ISO/

The site creates a download link which is valid for one day.
4+ GB. Only good that my phone provider forced me on a 50 Mbit/s
line last year. 

  $ xorriso -indev Win10_1607_English_x64.iso -report_el_torito plain 
-report_system_area plain
  ...
  Volume id: 'CCSA_X64FRE_EN-US_DV5'
  El Torito catalog  : 22  1
  El Torito images   :   N  Pltf  B   Emul  Ld_seg  Hdpt  Ldsiz LBA
  El Torito boot img :   1  BIOS  y   none  0x  0x00  8 539
  El Torito boot img :   2  UEFI  y   none  0x  0x00  1 541
  El Torito img blks :   1  2
  El Torito img blks :   2  2138320
  xorriso : NOTE : No System Area was loaded

No MBR and no partition tables to see.
So the ISO is only prepared for booting from CD, DVD, or BD media.
Not for hard disk or USB stick.


Since we are at it, let's have a look at the filesytems:

There is only one data file in the ISO 9660. Its content says
  This disc contains a "UDF" file system and requires an operating system
  that supports the ISO-13346 "UDF" file system specification.

So i mount it
  # mount -t udf -o loop Win10_1607_English_x64.iso /mnt/iso
and then can see lots of files underneath /mnt/iso.
  $ du -s /mnt/iso
  4273306 /mnt/iso

The content of the EFI System partition is similarly meager as in
a Debian ISO.
  # mount -t vfat -o loop,offset=1107968 Win10_1607_English_x64.iso /mnt/fat
The offset number 1107968 was computed from UEFI boot image LBA 541
multiplied by DVD block size 2048 bytes.
  $ find /mnt/fat -exec ls -ld '{}' ';'
  drwxr-xr-x 3 root root 7168 Jan  1  1970 /mnt/fat
  drwxr-xr-x 3 root root 512 Jul 15  2016 /mnt/fat/EFI
  drwxr-xr-x 2 root root 512 Jul 15  2016 /mnt/fat/EFI/BOOT
  -rwxr-xr-x 1 root root 869216 Jul 15  2016 /mnt/fat/EFI/BOOT/BOOTX64.EFI


Have a nice day :)

Thomas

Re: Debian installation issues

[Joe]

On Fri, 09 Jun 2017 16:24:20 +0200
"Thomas Schmitt"  wrote:

> Hi,
> 
> i wrote:
> > > Are there any [MS-Windows] images available for free and legally
> > > safe  
> 
> Joe wrote:
> > Oddly enough, there are:
> > https://www.microsoft.com/en-GB/evalcenter/evaluate-windows-10-enterprise  
> 
> The registration gives me creeps. I came up to the question whether
> i want to use a Microsoft or LinkedIn account. The Microsoft account
> creation asks me ifor info which i'd rather not want to give.
> 

I have a very old (@hotmail.com) emergency Hotmail account, which is
still acceptable as a Microsoft ID, though MS truncated the password
to 16 characters some years ago. I bought MS Access for a client using
it about a year ago. I've no idea what information I gave to create it,
but that must have been around 20 years ago.

> The description speaks of "Windows 10 Enterprise, version 1703 |
> 64-bit ISO" which lets me expect that one really would get a ISO 9660
> filesystem image.

I think so, my only concern was how big it is, it may only fit on a
Blu-Ray disc, or not even that.
> 
> I cannot find a legal statement whether it is permissible to give the
> ISO image to others. So i have to assume that there is no such
> permission. All in all i am not yet curious enough to start a legal
> relation with Microsoft Inc.

Me neither, but it *is* a free and legally installable MS OS, fulfilling
the conditions... there are also Win 8 versions. I may possibly still
have NT4 180-day evaluation CDs somewhere in my loft, but probably the
installation is date-limited and won't work. Though they would still
work as a test of iso-burning and booting.

-- 
Joe

Re: Debian installation issues

[Dan Ritter]

On Fri, Jun 09, 2017 at 04:24:20PM +0200, Thomas Schmitt wrote:
> Hi,
> 
> i wrote:
> > > Are there any [MS-Windows] images available for free and legally safe
> 
> Joe wrote:
> > Oddly enough, there are:
> > https://www.microsoft.com/en-GB/evalcenter/evaluate-windows-10-enterprise
> 
> The registration gives me creeps. I came up to the question whether
> i want to use a Microsoft or LinkedIn account. The Microsoft account
> creation asks me ifor info which i'd rather not want to give.
> 
> The description speaks of "Windows 10 Enterprise, version 1703 | 64-bit ISO"
> which lets me expect that one really would get a ISO 9660 filesystem image.
> 
> I cannot find a legal statement whether it is permissible to give the
> ISO image to others. So i have to assume that there is no such permission.
> All in all i am not yet curious enough to start a legal relation with
> Microsoft Inc.

Microsoft requires you to pay for their software.

They do not enforce this in the software itself, however. I
would consider it ethical to use their software once to
determine compatibility, after which you should pay for a
license or delete it.

https://www.microsoft.com/en-us/software-download/windows10ISO/

is where one downloads ISOs. There is no account creation or
registration required. The installation process will ask you to
agree to their EULA; if you do not, the install halts. It will
also ask for a license key, but you do not have to fill that in
immediately.

-dsr-

Re: Debian installation issues

[Michael Lange]

Hi,

On Fri, 9 Jun 2017 14:34:47 +0100
Michael Fothergill  wrote:

(...)
> > I am willing to bet the image written through rufus is fine, your
> > problem is booting up from usb as people have mentioned before.  I
> > like to assume that you did not install win10 in an older pc but
> > bought a recent pc with win10 in it. The following will not help you
> > "fix your problem" but it is more like talking about THE problem so
> > your frustration is not misdirected to the wrong direction.
> > MS with win10, using the excuse of a protection of your system,
> > managed to effectively block other systems of getting installed next
> > to win10.
> >
> 
> ​Is this really the case?  I use windows 10 and it does not stop me
> running a triple boot system on a single disk.
> 
> Try using grub2win: https://sourceforge.net/projects/grub2win/  -
> install it in windows 10 and play around with it and see if you can get
> it to see your usb stick.

A while ago I installed linux (not debian, but I don't think that matters
here) onto a laptop with a pre-installed win10, there was not any problem
with the EFI/BIOS finding the usb thumb drive with the installer iso and
booting into it. Hence I doubt that the OP's problem is that windows
prevents him from booting the install media. I did not follow this thread
closely from the beginning, but I believe it is more likely that either
the installer drive was not set up correctly or that some Efi/Bios
setting is not correct. Or maybe it is one of these odd devices that have
been discussed here a couple of days ago in a different thread, which are
64bit machines with a 32bit EFI.

Regards

Michael



.-.. .. ...- .   .-.. --- -. --.   .- -. -..   .--. .-. --- ... .--. . .-.

No one can guarantee the actions of another.
-- Spock, "Day of the Dove", stardate unknown

Re: Reply-to-all or reply-to-list again

[Nicolas George]

Le primidi 21 prairial, an CCXXV, The Wanderer a écrit :
> Disagreed. This results in sending extra copies to people who are
> subscribed to the list, which is incorrect.

Not if the list is properly configured.

Debian's lists are badly configured, it results in burden to all users,
but the users should reject that burden. I do.

I will not answer Gene's message, he obviously missed the whole point of
my mail.

Regards,

-- 
  Nicolas George


signature.asc
Description: Digital signature

Re: Grow an ext4 filesystem

[David Parker]

Indeed it has!  I upgraded to Jessie and all is well  Thank you!

# e2fsck -f /dev/sdb1
e2fsck 1.42.12 (29-Aug-2014)
Pass 1: Checking inodes, blocks, and sizes
Pass 2: Checking directory structure
Pass 3: Checking directory connectivity
Pass 4: Checking reference counts
Pass 5: Checking group summary information
/dev/sdb1: 45983/183140352 files (1.0% non-contiguous), 668815777/732530432
blocks

# resize2fs /dev/sdb1
resize2fs 1.42.12 (29-Aug-2014)
Resizing the filesystem on /dev/sdb1 to 1220884224 (4k) blocks.
The filesystem on /dev/sdb1 is now 1220884224 (4k) blocks long.

And now...

root@tapesrv:~# df -h /dev/sdb1
Filesystem  Size  Used Avail Use% Mounted on
/dev/sdb1   4.5T  2.5T  1.9T  58% /export

Thanks again!

On Fri, Jun 9, 2017 at 10:34 AM, Sven Joachim  wrote:

> On 2017-06-09 10:15 -0400, David Parker wrote:
>
> > I have a storage server running Debian 7.6 x64.  It's an HP server with
> 24
> > HDDs and a hardware RAID controller.  It has a 2.9 TB ext4 filesystem
> which
> > resides on a RAID 5 volume, and I recently needed to grow this filesystem
> > so I added more disks to the volume and then used parted to grow the
> > partition to the new size of 5 TB, so the space is now available in that
> > partition:
> >
> > # parted print /dev/sdb
> > Model: HP LOGICAL VOLUME (scsi)
> > Disk /dev/sdb: 5001GB
> > Sector size (logical/physical): 512B/512B
> > Partition Table: gpt
> >
> > Number  Start   End SizeFile system  Name Flags
> >  1  1049kB  5001GB  5001GB  ext4 primary
> >
> > However, the filesystem is still stuck at the old size:
> >
> > # df -h /dev/sdb1
> > Filesystem  Size  Used Avail Use% Mounted on
> > /dev/sdb1   2.7T  2.5T  104G  97% /export
> >
> > I cannot figure out how to resize the actual ext4 filesystem to use the
> > added space.  I have tried both resize2fs without any luck.  When I try
> > resize2fs, I get this error:
> >
> > # resize2fs /dev/sdb1
> > resize2fs 1.42.5 (29-Jul-2012)
> > resize2fs: /dev/sdb1: The combination of flex_bg and
> > !resize_inode features is not supported by resize2fs.
> >
> > In a potentially stupid move, I did indeed remove the resize_inode
> feature
> > from this filesystem in order to get parted to work with it, which
> > ultimately proved unnecessary, but now I cannot add it back:
> >
> > # tune2fs -O resize_inode /dev/sdb1
> > tune2fs 1.42.5 (29-Jul-2012)
> > Setting filesystem feature 'resize_inode' not supported.
> >
> > Is there a way I can resize this filesystem to use the additional 2 TB
> > available to it?
>
> By using a newer e2fsprogs version, this particular problem has been
> fixed in e2fsprogs 1.42.8[1].
>
> Cheers,
>Sven
>
>
> 1. https://bugs.debian.org/696746
>
>


-- 
Dave Parker
Database & Systems Administrator
Utica College
Integrated Information Technology Services
(315) 792-3229
Registered Linux User #408177

Re: Reply-to-all or reply-to-list again (was: https_port)

[Gene Heskett]

On Friday 09 June 2017 10:47:29 Nicolas George wrote:

> Le primidi 21 prairial, an CCXXV, Charlie Kravetz a écrit :
> > When replying to the mailing list, hit reply. Do not use "Reply to
> > All", since that sends individual emails to the person you are
> > answering.
>
> This recommendation is unsustainable and should be eliminated from the
> guidelines. It only exists because the mailing-list server is not
> configured correctly.
>
> The reply-to-list feature is flawed because it requires the user to
> give special attention each time "am I replying to a personal mail or
> to a mailing-list"? The correct behaviour should be the default,
> always, because that is the only way to minimize mistakes. This is
> what happens with mailing-list servers properly configured.
>
> When replying to a mail, any mail, use reply-to-all, unless you
> actively know you want to reply to an unusual subset of recipients.
> (But heed the reply-to headers, of course.)
>
> If somebody complain, tell them to set up their mail headers, just as
> I did mine.
>
> Regards,

Or use an email agent that does properly support a "reply to mailing 
list" function.  Such as this old kmail still included as part of the 
tde (trinity) desktop.

I do not use the reply-all format unless I am aware the OP is not 
subscribed. Those mailing lists I am subscribed to that do not require 
the poster to be subscribed are also flooded with spam.  And debian-user 
is one such list, and because of that, and my ISP's reject policy, I get 
threats from bendel of being unsubscribed because of bounces. It will 
never happen because when I get the bounce msg from bendel, its obvious 
why mail.shentel.net rejected it.  IMO debian needs to fix that, but its 
been made very clear that they will not go to a subscription required to 
post model.  Ever.  Sigh  It is what it is.

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: Reply-to-all or reply-to-list again (was: https_port)

[Fungi4All]

UTC Time: June 9, 2017 2:47 PM
From: geo...@nsup.org
To: Charlie Kravetz 
debian-user@lists.debian.org

Le primidi 21 prairial, an CCXXV, Charlie Kravetz a écrit :
> When replying to the mailing list, hit reply. Do not use "Reply to
> All", since that sends individual emails to the person you are
> answering.

This recommendation is unsustainable and should be eliminated from the
guidelines. It only exists because the mailing-list server is not
configured correctly.

The reply-to-list feature is flawed because it requires the user to give
special attention each time "am I replying to a personal mail or to a
mailing-list"? The correct behaviour should be the default, always,
because that is the only way to minimize mistakes. This is what happens
with mailing-list servers properly configured.

right before I replied to xorriso Thomas and the reply went personal. I hit 
reply
to this message and it is going to the list. So I agree with you, it is always
easier to delete the unwanted than to edit an addition and/or both.

When replying to a mail, any mail, use reply-to-all, unless you actively
know you want to reply to an unusual subset of recipients. (But heed the
reply-to headers, of course.)

I think it seems affected by the post of each member and transmits some
but not all headers for threading. I can't explain it otherwise.
Even if the list was misconfigured it should have been consistent for all
messages and all recipients. Here we have supernatural evidence of ghost
in the machine.
Whatever happened to bounce instead of forward? Is it illegal now? Is this
list utilizing this action, bouncing instead of trimming and forwarding?

If somebody complain, tell them to set up their mail headers, just as I
did mine.

and some laxatives for side dish.

Regards,
Nicolas George

Re: Reply-to-all or reply-to-list again

[The Wanderer]

On 2017-06-09 at 10:47, Nicolas George wrote:

> Le primidi 21 prairial, an CCXXV, Charlie Kravetz a écrit :
> 
>> When replying to the mailing list, hit reply. Do not use "Reply to
>> All", since that sends individual emails to the person you are
>> answering.
> 
> This recommendation is unsustainable and should be eliminated from
> the guidelines. It only exists because the mailing-list server is
> not configured correctly.
> 
> The reply-to-list feature is flawed because it requires the user to
> give special attention each time "am I replying to a personal mail or
> to a mailing-list"? The correct behaviour should be the default,
> always, because that is the only way to minimize mistakes. This is
> what happens with mailing-list servers properly configured.

Agreed.

> When replying to a mail, any mail, use reply-to-all, unless you
> actively know you want to reply to an unusual subset of recipients.
> (But heed the reply-to headers, of course.)

Disagreed. This results in sending extra copies to people who are
subscribed to the list, which is incorrect. The only time you should
send a copy of a message both to the list and to someone who is
subscribed to the list is when you specifically want to draw that
person's attention to that particular message, e.g. if you think they
might otherwise miss it among the rest of the list traffic, or if you
have the mistaken impression that they are not subscribed to the list.

Just as it's bad to require the user to check "am I replying to a
personal mail or to the mailing list?" every time, it's equally bad to
require the user to check "did this reply include addressees which it
shouldn't, or omit ones which it should?" every time. Using "Reply to
All" as your default action leads to the latter situation.

Ideally, things would be configured so that simple Reply would work
correctly in all cases (and I have indistinct memories of mailing-list
discussions in years long past where this seemed to in fact be the
case); however, I'm not certain of how to achieve that in practice.
Regardless, Reply-to-All as the "use as baseline default" is simply not
a good approach.

> If somebody complain, tell them to set up their mail headers, just as
> I did mine.

Are there pages out there documenting how to configure various mail
clients (much less Webmail!) to do this automatically?

Because tweaking mail headers by hand on every reply is unwieldy and
impractical, and people are simply not going to do that.

-- 
   The Wanderer

The reasonable man adapts himself to the world; the unreasonable one
persists in trying to adapt the world to himself. Therefore all
progress depends on the unreasonable man. -- George Bernard Shaw



signature.asc
Description: OpenPGP digital signature

Reply-to-all or reply-to-list again (was: https_port)

[Nicolas George]

Le primidi 21 prairial, an CCXXV, Charlie Kravetz a écrit :
> When replying to the mailing list, hit reply. Do not use "Reply to
> All", since that sends individual emails to the person you are
> answering.

This recommendation is unsustainable and should be eliminated from the
guidelines. It only exists because the mailing-list server is not
configured correctly.

The reply-to-list feature is flawed because it requires the user to give
special attention each time "am I replying to a personal mail or to a
mailing-list"? The correct behaviour should be the default, always,
because that is the only way to minimize mistakes. This is what happens
with mailing-list servers properly configured.

When replying to a mail, any mail, use reply-to-all, unless you actively
know you want to reply to an unusual subset of recipients. (But heed the
reply-to headers, of course.)

If somebody complain, tell them to set up their mail headers, just as I
did mine.

Regards,

-- 
  Nicolas George


signature.asc
Description: Digital signature

Re: https_port

[Charlie Kravetz]

-BEGIN PGP SIGNED MESSAGE-
Hash: SHA256

On Thu, 08 Jun 2017 11:51:35 -0700
"Adiel Plasencia Herrera"  wrote:

>I  just do not understand how to 
>respond in the list, maybe I did wrong  and that's why they are confused, I 
>beg my apologies if it is so.
>
>I'm not talking about NTP, in fact I do not know what it is.
>
>Yesterday  I sent my question hbaia with the subject https_port but 
>when I  answered I did not know how it was to respond again and what I did 
>was  create an email and send it to the list with the subject Re: 
>https_port.
>
>This was the one that wanted to respond
>
>Http://lists.squid-cache.org/pipermail/squid-users/2017-June/015565.html 
>[http://lists.squid-cache.org/pipermail/squid-users/2017-June/015565.html]
>
>
>-Original Message-
>
>From: Darac Marjal 
>
>To: debian-user@lists.debian.org
>
>Date: Thu, 8 Jun 2017 16:43:57 +0100
>
>Subject: Re: https_port
>
>
>
>
>On Thu, Jun 08, 2017 at 11:34:20AM -0400, Greg Wooledge wrote:
>
>>On Thu, Jun 08, 2017 at 04:25:11PM +0100, Darac Marjal wrote:  
>
>>> On Thu, Jun 08, 2017 at 11:18:16AM -0700, Adiel Plasencia Herrera wrote:  
>
>>> >  Hello,  
>
>>> >  I do not look for security, is that having no real internet ip in my  
>
>>> >  company I need certain programs to go to the internet and for that I  
>
>>> >  use proxycap (http://www.proxycap.com/ [http://www.proxycap.com/])   
>that makes me this function
>
>>> >  perfectly through the proxy . What happens is that with HTTP does not  
>
>>> >  work and I need to pass my squid to use HTTPS authentication for the  
>
>>> >  program (proxycap) to work well.  
>
>>>  
>
>>> I don't think squid works with NTP at all, but it's been a few years  
>
>>> since I played with Squid, so maybe someone else will be able to give  
>
>>> better advice.  
>
>>  
>
>>I don't think he's *asking* about NTP at all.  
>
>>  
>
>
>
>Ah. You mean he's a politician (replying to a topic by introducing one's
>
>own, unrelated, topic)?
>
>
>
>
>

When replying to the mailing list, hit reply. Do not use "Reply to
All", since that sends individual emails to the person you are
answering. Most people are members of the mailing list, and do not want
individual emails. If there is a "Reply to mailing list", it is okay to
use that. The subject should fill itself in.

When asking a question on the mailing list, hit "New Message" or
"Compose". Use a statement in the Subject to describe what you are
asking, and tell us in the message body what you need. The more
information you give, the better the answers you will get.

- -- 
Charlie Kravetz
Linux Registered User Number 425914
[http://linuxcounter.net/user/425914.html]
Never let anyone steal your DREAM.   [http://keepingdreams.com]
-BEGIN PGP SIGNATURE-

iQEzBAEBCAAdFiEEG5QK93YKrQMH22ZTiq6LjqbJ0IAFAlk6rGoACgkQiq6LjqbJ
0IB6+Qf/T5wybVTiO8ufwMJcuh6lRBahGFpA3RhUzUjkZC2EAqtqlgGC6JUcSbug
SCab9tZ2PU8AuFvXn43tUOXxcS3pFQY7Nw8hqekBam0n1xwe0dBe7pJATvhAu9WG
IEBlf5DaFit0lffbsyWcR9EozmU9QMr3NT/HpUoifNIlroiXE/twNiu+GQiY4fVJ
LSpfharxwYvGjBbvZ1y6DjOrZ0JfzWE2B0BJNcirwLHqVhKOiQtN2duhyBZD6Aox
Kkeg4AfR1zUmhat2llhr1vJN/jbA4BVKBuU2Lgh8351ec1ISy81A9hizZz4+irdI
geGVopHkYSXXSMG464UGuc6PTeATKg==
=VTXc
-END PGP SIGNATURE-

Re: Grow an ext4 filesystem

[Sven Joachim]

On 2017-06-09 10:15 -0400, David Parker wrote:

> I have a storage server running Debian 7.6 x64.  It's an HP server with 24
> HDDs and a hardware RAID controller.  It has a 2.9 TB ext4 filesystem which
> resides on a RAID 5 volume, and I recently needed to grow this filesystem
> so I added more disks to the volume and then used parted to grow the
> partition to the new size of 5 TB, so the space is now available in that
> partition:
>
> # parted print /dev/sdb
> Model: HP LOGICAL VOLUME (scsi)
> Disk /dev/sdb: 5001GB
> Sector size (logical/physical): 512B/512B
> Partition Table: gpt
>
> Number  Start   End SizeFile system  Name Flags
>  1  1049kB  5001GB  5001GB  ext4 primary
>
> However, the filesystem is still stuck at the old size:
>
> # df -h /dev/sdb1
> Filesystem  Size  Used Avail Use% Mounted on
> /dev/sdb1   2.7T  2.5T  104G  97% /export
>
> I cannot figure out how to resize the actual ext4 filesystem to use the
> added space.  I have tried both resize2fs without any luck.  When I try
> resize2fs, I get this error:
>
> # resize2fs /dev/sdb1
> resize2fs 1.42.5 (29-Jul-2012)
> resize2fs: /dev/sdb1: The combination of flex_bg and
> !resize_inode features is not supported by resize2fs.
>
> In a potentially stupid move, I did indeed remove the resize_inode feature
> from this filesystem in order to get parted to work with it, which
> ultimately proved unnecessary, but now I cannot add it back:
>
> # tune2fs -O resize_inode /dev/sdb1
> tune2fs 1.42.5 (29-Jul-2012)
> Setting filesystem feature 'resize_inode' not supported.
>
> Is there a way I can resize this filesystem to use the additional 2 TB
> available to it?

By using a newer e2fsprogs version, this particular problem has been
fixed in e2fsprogs 1.42.8[1].

Cheers,
   Sven


1. https://bugs.debian.org/696746

Re: Debian installation issues

[Thomas Schmitt]

Hi,

i wrote:
> > Are there any [MS-Windows] images available for free and legally safe

Joe wrote:
> Oddly enough, there are:
> https://www.microsoft.com/en-GB/evalcenter/evaluate-windows-10-enterprise

The registration gives me creeps. I came up to the question whether
i want to use a Microsoft or LinkedIn account. The Microsoft account
creation asks me ifor info which i'd rather not want to give.

The description speaks of "Windows 10 Enterprise, version 1703 | 64-bit ISO"
which lets me expect that one really would get a ISO 9660 filesystem image.

I cannot find a legal statement whether it is permissible to give the
ISO image to others. So i have to assume that there is no such permission.
All in all i am not yet curious enough to start a legal relation with
Microsoft Inc.


I wrote:
> > All known Linux
> > installation ISOs are to be put on the whole storage device.

Fungi4All wrote:
> Then you can format and partition the left over disk and even install linux
> in it with grub and stuff.  I think!

Adding partitions to claim the remaining device space is possible,
although the often crammed partition table combinations make many
partition editors scream. fdisk should be ok for Debian images,
because it ignores the Apple Partition Map and the invalid GPT in
those ISOs.

But replacing the ISO booting ISOLINUX by GRUB, or coordinating the
ISO booting GRUB with a local GRUB installation might become tricky.
At least you will have to learn about both boot loaders.
(A Debian ISOs boot by ISOLINUX on BIOS and by GRUB on EFI.)


> Rufus works like a champ, why and how I couldn't possibly know.  All
> attempts from linux side produced a non-bootable windows installation disk 

If you have such an ISO passing by a Linux machine, i would be interested
to see the output of

  iso=...Path.to.ISO.image.file.or.to.CD.drive...
  xorriso -indev "$iso" -report_el_torito plain -report_system_area plain

This would give indications what kind of firmware on what kind of
storage device is supposed to boot the image.
(Or maybe it just reveils a new xorriso bug. Exotic input often yields
 exotic behavior ...)


> Lightning fast was my foot kicking the machine!

Consider that they remember everything.
A future civilization of AI entities will look back and judge our behavior.
Do you want to be mentioned as human barbarian, 3rd degree, number 3210620 ?

My entry in the logs shall be "He never abandoned a piece of hardware
before it was dead" or "His hands on the keyboard were always warm".


Have a nice day :)

Thomas

Grow an ext4 filesystem

[David Parker]

Hello,

I have a storage server running Debian 7.6 x64.  It's an HP server with 24
HDDs and a hardware RAID controller.  It has a 2.9 TB ext4 filesystem which
resides on a RAID 5 volume, and I recently needed to grow this filesystem
so I added more disks to the volume and then used parted to grow the
partition to the new size of 5 TB, so the space is now available in that
partition:

# parted print /dev/sdb
Model: HP LOGICAL VOLUME (scsi)
Disk /dev/sdb: 5001GB
Sector size (logical/physical): 512B/512B
Partition Table: gpt

Number  Start   End SizeFile system  Name Flags
 1  1049kB  5001GB  5001GB  ext4 primary

However, the filesystem is still stuck at the old size:

# df -h /dev/sdb1
Filesystem  Size  Used Avail Use% Mounted on
/dev/sdb1   2.7T  2.5T  104G  97% /export

I cannot figure out how to resize the actual ext4 filesystem to use the
added space.  I have tried both resize2fs without any luck.  When I try
resize2fs, I get this error:

# resize2fs /dev/sdb1
resize2fs 1.42.5 (29-Jul-2012)
resize2fs: /dev/sdb1: The combination of flex_bg and
!resize_inode features is not supported by resize2fs.

In a potentially stupid move, I did indeed remove the resize_inode feature
from this filesystem in order to get parted to work with it, which
ultimately proved unnecessary, but now I cannot add it back:

# tune2fs -O resize_inode /dev/sdb1
tune2fs 1.42.5 (29-Jul-2012)
Setting filesystem feature 'resize_inode' not supported.

Is there a way I can resize this filesystem to use the additional 2 TB
available to it?  The output of dumpe2fs is below (not including the data
about the 22,355 groups in the filesystem):

dumpe2fs 1.42.5 (29-Jul-2012)
Filesystem volume name:   
Last mounted on:  /export
Filesystem UUID:  0306070a-6d39-48da-8235-aa99c9e48141
Filesystem magic number:  0xEF53
Filesystem revision #:1 (dynamic)
Filesystem features:  has_journal ext_attr dir_index filetype
needs_recovery extent flex_bg sparse_super large_file
Filesystem flags: signed_directory_hash
Default mount options:user_xattr acl
Filesystem state: clean
Errors behavior:  Continue
Filesystem OS type:   Linux
Inode count:  183140352
Block count:  732530432
Reserved block count: 36626521
Free blocks:  63964655
Free inodes:  183094383
First block:  0
Block size:   4096
Fragment size:4096
Blocks per group: 32768
Fragments per group:  32768
Inodes per group: 8192
Inode blocks per group:   512
Flex block group size:16
Filesystem created:   Thu Sep 18 14:40:07 2014
Last mount time:  Fri Jun  9 09:46:28 2017
Last write time:  Fri Jun  9 09:46:28 2017
Mount count:  5
Maximum mount count:  -1
Last checked: Thu Jun  8 11:51:32 2017
Check interval:   0 ()
Lifetime writes:  17 TB
Reserved blocks uid:  0 (user root)
Reserved blocks gid:  0 (group root)
First inode:  11
Inode size:   256
Required extra isize: 28
Desired extra isize:  28
Journal inode:8
Default directory hash:   half_md4
Directory Hash Seed:  81bb6e96-1eea-44c2-8dbd-f3b046f9b2a9
Journal backup:   inode blocks
Journal features: journal_incompat_revoke
Journal size: 128M
Journal length:   32768
Journal sequence: 0x020ee1a8
Journal start:1

Any help is greatly appreciated.  Thanks!

-- 
Dave Parker
Database & Systems Administrator
Utica College
Integrated Information Technology Services
(315) 792-3229
Registered Linux User #408177

Re: php5-fpm segfault error 6 in libpcre.so.3.13.1

[Lucio Crusca]

Georgi Naplatanov wrote:

Stretch (Debian 9) doesn't include PHP 5.x,


You are right, in fact it is a jessie+stretch server, mostly jessie + 
some package from stretch.

Re: Debian installation issues

[Michael Fothergill]

On 9 June 2017 at 08:01, Fungi4All  wrote:

>
> UTC Time: June 8, 2017 4:17 PM
> From: wool...@eeg.ccf.org
>
> On Thu, Jun 08, 2017 at 11:08:32AM -0500, David DLC wrote:
> > Thank you for all the replies! I haven't really used a mailing list
> before,
> > so I'm not 100% sure I'm responding to the correct location. Do I hit
> > "reply all" or just reply to the debian-user@lists.debian.org address?
>
> For this mailing list, you are expected to ..
>
>
> whaaat ever 
>
> > Anyway, I have disabled secure boot, but that didn't seem to solve the
> > problem. My computer does not have a DVD drive, so I put the ISO (
> > debian-8.8.0-amd64-netinst.iso) on a USB stick. When booting my
> computer, I
> > specifically click "use a device", then "USB Drive (UEFI)". The computer
> > runs for a minute, then pops up the error message. I don't think boot
> order
> > would change this, as I am booting specifically from the USB. I forgot to
> > mention that I am attempting to dual boot my computer, so I don't wish to
> > remove Windows completely.
>
> If you used Windows to put the ISO image onto the USB stick, it's
> quite possible it wasn't done correctly.
>
>
> Get used to Greg on this list making vague assumptions on things and
> blaming the victim for anything the victim is having problems with, never
> the system.  The system must not be challenged.
> In trying to convince others to switch from windows to a better and higher
> performing system, despite of hardware, one must learn to work with windows
> to begin with, because nobody is willing to wipe everything they have off
> and start with a clean disk.
> Rufus in my experience has been 100% reliable in burning images of all
> sorts.  It is small and fast.  If in reverse try from linux to burn an
> ms-win-installation image, chances are that you will fail despite of what
> way you may try to do so.  Propbably Thomas from xorisso fame can explain
> better the whys and why nots.  I give up trying to understand the reasons.
> Let's hope you do not live in isolation, there are others with a pc around
> you and not all run win10.  Try and boot your stick in their system, no
> harm can be done if you don't install anythin.  If by any chance you make a
> debian-live-installation image you may even get to see debian come a-live
> without an installation.
> Since the debian world is really poor in applied technical information
> location, try the ubuntu (askubuntu) for help/faq.  It is a friendlier
> environment and easier to find solutions.  99% of what you will find there
> works for debian too.  Especially on this specific issue there is a ton of
> instructions.
>
> The #debian bot currently recommends  win32diskimager/>
> if you need to write the Debian install image to a USB device from
> Windows.
>
>
> I am willing to bet the image written through rufus is fine, your problem
> is booting up from usb as people have mentioned before.  I like to assume
> that you did not install win10 in an older pc but bought a recent pc with
> win10 in it. The following will not help you "fix your problem" but it is
> more like talking about THE problem so your frustration is not misdirected
> to the wrong direction.
> MS with win10, using the excuse of a protection of your system, managed to
> effectively block other systems of getting installed next to win10.
>

​Is this really the case?  I use windows 10 and it does not stop me running
a triple boot system on a single disk.

Try using grub2win: https://sourceforge.net/projects/grub2win/  - install
it in windows 10 and play around with it and see if you can get it to see
your usb stick.

Regards

MF




> So ms to promote themselves and diminish competition (what's new?) did
> what they did.  Then they manhandled manufacturers to incorporate their
> evil into what they sell to you.  And you paid for this problem to your
> ventor.  Did you know when you purchased your product and license you
> became part of the problem people are addressing to solve?
> Why not a worldwide class action suit against manufacturers, vendors, and
> retailers of all sorts for not warning you the product is nearly useless
> without win10?  Thanks to some who resist and make and sell linux based
> products off the shelf.  Let's not talk about fruit vendors here.
> If not, you deserve what you got, and I am sorry to have to tell you.  I
> think you should go back to whoever sold you a lemon for an apple and make
> them either pay you back what you paid of install Debian for you on your
> partition.  It should only take 20' not several days.
> Ohhh... you accepted some license that told you so in the fine print?  I
> am sorry!  Really   keep paying for problems you distribute around.
>
> If you're doing it from a Unix/Linux system, then follow the
> instructions at 
>
>
> I have given up reading any documentation from debian, it is all written
> by developers for eng

Compiler segfault when building the kernel

[Celejar]

Hi,

I've been building kernels (vanilla from upstream) for years with
kernel-package (typical command line: "time make-kpkg -j2 --initrd
--revision 1.custom kernel_image"; .kernel-pkg.conf contains just the
line "root_cmd = fakeroot") without problem. Recently, the builds have
begun to fail with messages like these:

*

> In file included from ./include/linux/percpu-rwsem.h:8:0,
>  from ./include/linux/fs.h:30,
>  from ./include/linux/pagemap.h:8,
>  from block/partitions/check.h:1,
>  from block/partitions/msdos.c:23:
> ./include/linux/rcu_sync.h:29:48: internal compiler error: Segmentation fault
>  enum rcu_sync_type { RCU_SYNC, RCU_SCHED_SYNC, RCU_BH_SYNC };
> ^
> Please submit a full bug report,
> with preprocessed source if appropriate.
> See  for instructions.
>   CC  fs/posix_acl.o
> The bug is not reproducible, so it is likely a hardware or OS problem.

*

> In file included from ./include/linux/linkage.h:6:0,
>  from ./include/linux/kernel.h:6,
>  from ./include/linux/list.h:8,
>  from ./include/linux/module.h:9,
>  from lib/fonts/font_8x16.c:8:
> ./include/linux/export.h:63:22: internal compiler error: Segmentation fault
>   static const struct kernel_symbol __ksymtab_##sym  \
>   ^
> ./include/linux/export.h:93:25: note: in expansion of macro ‘___EXPORT_SYMBOL’
>  #define __EXPORT_SYMBOL ___EXPORT_SYMBOL
>  ^
> ./include/linux/export.h:97:2: note: in expansion of macro ‘__EXPORT_SYMBOL’
>   __EXPORT_SYMBOL(sym, "")
>   ^
> lib/fonts/font_8x16.c:4633:1: note: in expansion of macro ‘EXPORT_SYMBOL’
>  EXPORT_SYMBOL(font_vga_8x16);
>  ^
> Please submit a full bug report,
> with preprocessed source if appropriate.

This occurred immediately following a cleaning of the source tree
("make-kpkg ... clean"), the first one I've done in quite some time, so
I'm pretty sure that that's what triggered this, whatever the
underlying problem actually is.

Googling suggests that this sort of thing can be triggered by race
conditions caused by build systems improper handling of
concurrency,e.g.:

https://askubuntu.com/questions/343490/the-bug-is-not-reproducible-so-it-is-likely-a-hardware-or-os-problem

For the last year or so, I've been building with -j2, so I tried again
without it. I still got the same error, but when I once again did a
clean and then rebuilt without -j2, the build succeeded.

Any ideas? Is this a bug I should be filing against kernel-package (or
anywhere else)?

Celejar

Re: php5-fpm segfault error 6 in libpcre.so.3.13.1

[Georgi Naplatanov]

On 06/09/2017 12:27 PM, Lucio Crusca wrote:
> Hi all,
> 
> I've a problem with a stretch/amd64 web server. Since two hours ago, I
> get the following in dmesg and syslog:
> 
> Jun  9 11:21:01 weber kernel: [  331.716831] php5-fpm[1245]: segfault at
> 7ffcb4febf70 ip 7f2e5e8d629a sp 7ffcb4febf60 error 6 in
> libpcre.so.3.13.3[7f2e5e8c2000+72000]
> 
> which in turn results in a http 503 status code. However that happens
> only for one of the two VirtualHosts that make use of php5-fpm.
> 
> I've already tried lowering  pcre values in /etc/php5/fpm/php.ini:

Stretch (Debian 9) doesn't include PHP 5.x, it includes PHP 7.0 so the
configuration file for PHP-FPM is /etc/php/7.0/fpm/php.ini. If you have
PHP 5.x packages installed then they are absolute from previous Debian
versions.

> pcre.backtrack_limit=100
> pcre.recursion_limit=1000
> 
> and restarting php5-fpm (and the whole server for that matter), but
> nothing changed. The server has 8GB of RAM memory and
> 
> # ulimit -s
> 8192
> 
> What could I do to find the cause?
> 

Re: Debian installation issues

[Joe]

On Fri, 09 Jun 2017 10:47:25 +0200
"Thomas Schmitt"  wrote:

> Hi,
> 
> Fungi4All wrote:
> > try from linux to burn an
> > ms-win-installation image, chances are that you will fail despite
> > of what way you may try to do so.  
> 
> Are there any such images available for free and legally safe to have
> and to talk about ?
> 
> 
Oddly enough, there are:

https://www.microsoft.com/en-GB/evalcenter/evaluate-windows-10-enterprise

I haven't registered and investigated, so I don't know the size
involved. They may well be netinstalls, or they may be far larger than
is easily available as a USB stick. But they are expected to be
installed, so it's doubtful that they require the user to mount
the .iso file from a hard drive location.

-- 
Joe

php5-fpm segfault error 6 in libpcre.so.3.13.1

[Lucio Crusca]

Hi all,

I've a problem with a stretch/amd64 web server. Since two hours ago, I 
get the following in dmesg and syslog:


Jun  9 11:21:01 weber kernel: [  331.716831] php5-fpm[1245]: segfault at 
7ffcb4febf70 ip 7f2e5e8d629a sp 7ffcb4febf60 error 6 in 
libpcre.so.3.13.3[7f2e5e8c2000+72000]


which in turn results in a http 503 status code. However that happens 
only for one of the two VirtualHosts that make use of php5-fpm.


I've already tried lowering  pcre values in /etc/php5/fpm/php.ini:

pcre.backtrack_limit=100
pcre.recursion_limit=1000

and restarting php5-fpm (and the whole server for that matter), but 
nothing changed. The server has 8GB of RAM memory and


# ulimit -s
8192

What could I do to find the cause?

Re: Encoding problems with latin accented characters in vfat partition

[Rodolfo Medina]

Jim Ohlstein  writes:

> On Fri, 2017-06-09 at 01:50 +0100, Rodolfo Medina wrote:
>> Hi.
>> 
>> I finally got to mount and share an USB drive plugged into my ADSL router
>> (see thread `Accessing USB storage attached to network router' on this
>> list).  The only problem now is about latin accented characters such as
>> bontà, perché, ragù, ecc...  In my home directory I have many files such
>> named, but when I want to copy them into the above vfat driver there are
>> encoding problems: those characters are messing and `Bad file decriptor'
>> error messages sometimes appear.  I've never had this problem before, in my
>> Linux home directory.  Please suggest whoever can how to face and solve the
>> problem.
>
> Have you defined iocharset=utf8 in your mount command options?

Yes.  In /etc/fstab I've put:

 //192.168.1.2/Shared /mnt/shared cifs rw,user,noauto,iocharset=utf8 0 0

, but the problem is still there.

Thanks,

Rodolfo

Re: Debian installation issues

[Thomas Schmitt]

Hi,

Fungi4All wrote:
> try from linux to burn an
> ms-win-installation image, chances are that you will fail despite of what
> way you may try to do so.

Are there any such images available for free and legally safe to have
and to talk about ?


> Propbably Thomas from xorisso fame can explain
> better the whys and why nots.

Well, Microsoft Inc. will hardly use xorriso to create its images.

If it is about ISO 9660 filesystems which shall boot on BIOS and/or EFI
from DVD and USB stick, then such a filesystem image has to be copied
plainly on the blocks of a storage device. Depending on the device type
the reader software (BIOS, EFI, Linux, MS-Windows, ...) will interpret
the first blocks as MBR boot code and/or partition table or ignore them.

An important point is whether the image is supposed to represent a whole
storage device or a single partition of such a device. All known Linux
installation ISOs are to be put on the whole storage device.

Nevertheless ISOLINUX offers an opportunity to prepare isohybrid images
which are to be put onto a primary MBR partition. This partition will be
booted by an MBR which looks for the boot flag in the partition table.
Such MBRs are often found on newly bought USB sticks. The boot flag
MBR convention seems to be usual in the Windows world.
So a Microsoft Inc. image might well be meant for a partition, not the
whole storage device.

Of course it might be that Microsoft "images" need some unpacking or
patching before they get installed at various block ranges of a device.
But actually the term "image" stems from the concept of a 1:1 copy of
a storage device in form of a data file.


> I have given up reading any documentation from debian, it is all written by
> developers for engineers and vice-versa.  If you are not one of them it is a
> waste of time.

A better reaction would be to contact the writers and to negotiate better
representation of the facts which shall be described.
A manual must be understandable and be correct. Not easy to achieve.


> If you read the instructions on how to write a document from
> your pc to a usb disk, chances are that you will need to learn 5 more things
> to do so,

It is sincere sysadmin work to copy an image onto a storage device.
So you best first learn the ways of GNU/Linux: device model, filesystem
model, byte streams, shell, ... lots of concepts which play together.

For the desktop addicted user, there should better be a GUI tool which
keeps that user from overwriting the system disk or unpacking the ISO.


> It is easier to copy the document in marble with a cheasel than learn
> all this.

But only if the text is short and you are an experienced chiseller.

Don't forget the initial wonder of computerisation: Exactness.
It computes like a lightning and still does it correctly.
You can mistoggle your text and correct it without Tipp-Ex or rubber.
It remembers all your mistakes until you find an opportunity to correct
them.

Neither ink nor marble can do this for you. (Not to speak of the human
brain and its tendency to mess up after a few dozen steps.)


> Welcome to our world of whips and chains.

Actually it's math. The painful pursuit of exact thinking.


Have a nice day :)

Thomas

Re: Add unknown keyboard key

[Floris]

Op Thu, 08 Jun 2017 14:13:53 +0200 schreef Floris :


Op Mon, 29 May 2017 16:27:49 +0200 schreef Floris :


Hey,

I have an older Asus R2E UMPC. A long time it has Windows installed on  
it, because I only used it as a navigation device. And an attempt, long  
ago, to install and run Fedora on it, didn't work. But time has  
changed, Windows 10 is sluggish. So I give it an other try with Debian  
Sid (The same version I run on my desktop). So far so good. The  
touchscreen and fingerprint reader works. And even with Gnome3 the  
performance is good. There is only one issue so far: an extra button on  
the device.


There are 4 extra buttons and I remapped them with xmodmap. And it  
looks like they work, because when I run

xev -event keyboard
The keys are correct:
- keycode 112 XF86ScrollUp
- keycode 117 XF86ScrollDown
- keycode 234 XF86Launch0
- keycode 248 XF86Launch1  <-- This one doesn't work

But when I run:
evtest /dev/input/event9
I get
...
Event code 226 (KEY_MEDIA)
Event code 240 (KEY_UNKNOWN) <-- minus 8 from the kernel?
...

Pressing the XF86Launch1 (KEY_UNKNOWN) button dmesg gives:
asus_laptop: Unknown key 9a pressed

I added an udev hwdb rule to /etc/udev/hwdb.d/70-keyboard.hwdb
evdev:name:Asus Laptop extra buttons:dmi:bvn*:bvr*:bd*:svnASUS:pn*pvr*
  KEYBOARD_KEY_9a=prog1

But when I add a shortcut with Gnome Keyboard settings. I can add the  
key to an shortcut, but nothing happens when I press the button. When I  
add one of the other buttons to a shortcut, they work as expected.


Maybe someone has an idea to solve this problem?

Floris



Still no success, but a little progress.

I made a hwdb rule in /etc/udev/hwdb.d/99-keyboard.hwdb
evdev:name:Asus Laptop extra buttons:dmi:bvn*:bvr*:bd*:svnASUS:pn*pvr*
  KEYBOARD_KEY_95=keyboard
  KEYBOARD_KEY_9a=screen

after an udevadm update and trigger udevadm info /dev/input/event5  
reports the buttons:

...
E: ID_PATH_TAG=platform-asus_laptop
E: KEYBOARD_KEY_6b=f21  <-- This one is a default udev 
hwdb rule
E: KEYBOARD_KEY_95=keybaord
E: KEYBOARD_KEY_9a=screen
E: LIBINPUT_DEVICE_GROUP=19/0/0/0:asus_laptop
...

So far everything works as expected, but evtest doesn't remap the 9a key
...
type 4 (EV_MSC), code 4 (MSC_SCAN), value 95
type 1 (EV_KEY), code 374 (KEY_KEYBOARD), value 1		<-- This one is  
modified as expected

...
type 4 (EV_MSC), code 4 (MSC_SCAN), value 9a

[EDIT]
(I made a typo)
type 1 (EV_KEY), code 240 (KEY_UNKNOWN), value 1		<-- Why isn't this one  
modified?

...

[/EDIT]




Finally I solved this issue. I added the unknown key to asus-laptop.c and  
rebuild the kernel module. Although, I don't think this is the proper way  
of adding unknown media keys.

Re: Debian installation issues

[Fungi4All]

UTC Time: June 8, 2017 4:17 PM
From: wool...@eeg.ccf.org

On Thu, Jun 08, 2017 at 11:08:32AM -0500, David DLC wrote:
> Thank you for all the replies! I haven't really used a mailing list before,
> so I'm not 100% sure I'm responding to the correct location. Do I hit
> "reply all" or just reply to the debian-user@lists.debian.org address?

For this mailing list, you are expected to ..

whaaat ever 

> Anyway, I have disabled secure boot, but that didn't seem to solve the
> problem. My computer does not have a DVD drive, so I put the ISO (
> debian-8.8.0-amd64-netinst.iso) on a USB stick. When booting my computer, I
> specifically click "use a device", then "USB Drive (UEFI)". The computer
> runs for a minute, then pops up the error message. I don't think boot order
> would change this, as I am booting specifically from the USB. I forgot to
> mention that I am attempting to dual boot my computer, so I don't wish to
> remove Windows completely.

If you used Windows to put the ISO image onto the USB stick, it's
quite possible it wasn't done correctly.

Get used to Greg on this list making vague assumptions on things and blaming 
the victim for anything the victim is having problems with, never the system. 
The system must not be challenged.
In trying to convince others to switch from windows to a better and higher 
performing system, despite of hardware, one must learn to work with windows to 
begin with, because nobody is willing to wipe everything they have off and 
start with a clean disk.
Rufus in my experience has been 100% reliable in burning images of all sorts. 
It is small and fast. If in reverse try from linux to burn an 
ms-win-installation image, chances are that you will fail despite of what way 
you may try to do so. Propbably Thomas from xorisso fame can explain better the 
whys and why nots. I give up trying to understand the reasons.
Let's hope you do not live in isolation, there are others with a pc around you 
and not all run win10. Try and boot your stick in their system, no harm can be 
done if you don't install anythin. If by any chance you make a 
debian-live-installation image you may even get to see debian come a-live 
without an installation.
Since the debian world is really poor in applied technical information 
location, try the ubuntu (askubuntu) for help/faq. It is a friendlier 
environment and easier to find solutions. 99% of what you will find there works 
for debian too. Especially on this specific issue there is a ton of 
instructions.

The #debian bot currently recommends 
if you need to write the Debian install image to a USB device from
Windows.

I am willing to bet the image written through rufus is fine, your problem is 
booting up from usb as people have mentioned before. I like to assume that you 
did not install win10 in an older pc but bought a recent pc with win10 in it. 
The following will not help you "fix your problem" but it is more like talking 
about THE problem so your frustration is not misdirected to the wrong direction.
MS with win10, using the excuse of a protection of your system, managed to 
effectively block other systems of getting installed next to win10. So ms to 
promote themselves and diminish competition (what's new?) did what they did. 
Then they manhandled manufacturers to incorporate their evil into what they 
sell to you. And you paid for this problem to your ventor. Did you know when 
you purchased your product and license you became part of the problem people 
are addressing to solve?
Why not a worldwide class action suit against manufacturers, vendors, and 
retailers of all sorts for not warning you the product is nearly useless 
without win10? Thanks to some who resist and make and sell linux based products 
off the shelf. Let's not talk about fruit vendors here.
If not, you deserve what you got, and I am sorry to have to tell you. I think 
you should go back to whoever sold you a lemon for an apple and make them 
either pay you back what you paid of install Debian for you on your partition. 
It should only take 20' not several days.
Ohhh... you accepted some license that told you so in the fine print? I am 
sorry! Really  keep paying for problems you distribute around.

If you're doing it from a Unix/Linux system, then follow the
instructions at 

I have given up reading any documentation from debian, it is all written by 
developers for engineers and vice-versa. If you are not one of them it is a 
waste of time. If you read the instructions on how to write a document from 
your pc to a usb disk, chances are that you will need to learn 5 more things to 
do so, and each one of them will require learning 5 other new things, and so on 
and forth and back. It is easier to copy the document in marble with a cheasel 
than learn all this. Depending on your psychological profile you may actually 
like all this. Welcome to our world of whi