Re: yabasic problem

[Thomas Schmitt]

Hi,

i wrote:
> > (Found the booklet. It's HP BASIC 3.0, not 2.0. Newest techology of 1985.)

David Wright wrote:
> I thought we were up to version 4.0¹ by 1985,

Indeed, the booklet says "June 1984 ... First Edition".

I think i did not get to BASIC 4.0 because in 1986 i wrote a BASIC program
which translated our other BASIC programs to C (with some handwork being
left to do).


> Would you agree, though, that "BASIC" is the language that must
> have the biggest contrast between its well-endowed versions and
> the most dire cr*p.

Well, back then i perceived HP BASIC as the best language of all. It made
me boss on all those expensive HP machines (from 9845B to 9000/320).
But C ran on all Unix workstations. And as soon as i became ambidextrous
enough, i fell in love with the display manager of the Apollo Domain DN3000.

Microsoft's Visual Basic is said to have surpassed HP BASIC in the years
later.


> Where would yabasic fit?

It seems to be inspired by C (see the syntax of "open"). But why use such
a C-BASIC when there is gcc, gdb and valgrind ?


> If you've not come across the 9845 machine, it was in my experience
> unique, in that you could edit the program while it was executing,
> not even having to pause it.

Yep. I have seen users countering Error 17 by hitting the EDIT key and
throwing out the offending line. (One can imagine the effect on the
collection of emerged CAD files. We got much less repair work after the
users were re-seated to Apollo, DEC, and Sun where they had to submit
a bug report when the program crashed.)


Have a nice day :)

Thomas

Re: yabasic problem

[Jim Popovitch]

On August 20, 2018 7:35:35 AM UTC, Thomas Schmitt  wrote:
>Hi,
>
>i wrote:
>> > (Found the booklet. It's HP BASIC 3.0, not 2.0. Newest techology of
>1985.)
>
>David Wright wrote:
>> I thought we were up to version 4.0¹ by 1985,
>
>Indeed, the booklet says "June 1984 ... First Edition".
>
>I think i did not get to BASIC 4.0 because in 1986 i wrote a BASIC
>program
>which translated our other BASIC programs to C (with some handwork
>being
>left to do).
>
>
>> Would you agree, though, that "BASIC" is the language that must
>> have the biggest contrast between its well-endowed versions and
>> the most dire cr*p.
>
>Well, back then i perceived HP BASIC as the best language of all. It
>made
>me boss on all those expensive HP machines (from 9845B to 9000/320).
>But C ran on all Unix workstations. And as soon as i became
>ambidextrous
>enough, i fell in love with the display manager of the Apollo Domain
>DN3000.
>
>Microsoft's Visual Basic is said to have surpassed HP BASIC in the years later.
>

Before VB (early 1990s) there was  Microsoft Basic (Professional Development 
System) which iirc became quite popular in the late 80s.  

-Jim P.

Aw: Re: Re: does btrfs have a future? (was: feature)

[Stefan K]

> Your English is fine.
thanks for that. :)

> Ignoring Raid5/6 and similar, I don't know what features btrfs is
> lacking that make ZFS more attractive. 
I guess it is the very poor performance, so you cant use it as backend 
fileserver, amilserver etc, and you also can't SSDs for caching like ZFS do. 
The only thing where you can use btrfs is for / (root) nothing more because for 
that you don't need much performance. If they fixed that performance thing (up 
to ext4 level), then btrfs will have chance.

> Btrfs's killer feature, imo, is its Copy-On-Write features
yep you're right, specially the Snapshot thing, you can use it before you make 
upgrades (I think the new Linux Mint version does that).


best regards
Stefan

> Gesendet: Mittwoch, 15. August 2018 um 16:50 Uhr
> Von: "Matthew Crews" 
> An: "Stefan K" , debian-user@lists.debian.org
> Betreff: Re: Aw: Re: does btrfs have a future? (was: feature)
>
> On 8/15/18 2:25 AM, Stefan K wrote:
> > Did you think that "only" the RAID5/6 problem is the reason why btrfs is 
> > not so common? what is with the performance? and some (important) featrures 
> > (not futures ;) ) are missing to catch up ZFS.
> > 
> > best regards
> > Stefan
> > (sorry for my bad english)
> > 
> 
> Your English is fine. Not perfect (no one ever is), but I know plenty of
> native speakers who speak it worse than you.
> 
> In my opinion btrfs has a bad rap partially because of the RAID5/6
> situation, but also because for a long time it was marked as
> experimental, and there are some situations where data loss has occured
> (I'm guessing because of RAID5/6). But as long as you avoid RAID5/6 and
> stick to RAID1/10, you should be fine.
> 
> Ignoring Raid5/6 and similar, I don't know what features btrfs is
> lacking that make ZFS more attractive. Btrfs does have *nice* features
> that ZFS currently lacks, like adding and removing disks to the array
> on-the-fly and intelligent data balancing while the array is mounted.
> 
> Btrfs's killer feature, imo, is its Copy-On-Write features, which you
> can read about on the Arch Wiki:
> 
> https://wiki.archlinux.org/index.php/Btrfs#Copy-on-Write_.28CoW.29
> 
> Btrfs also corrects read errors on-the-fly, something ZFS doesn't do,
> but only if you are using a RAID with some level of redundancy.
> 
> 
> 

Thunderbird 60 ignores LC_TIME environment variable

[Stefan Pietsch]

Dear list,

after the upgrade from Thunderbird 52.9.1 to 60.0 (Debian sid) it seems
for me that the environment variable LC_TIME is ignored.

I used the following command to set 24h time format in Thunderbird.

$ LC_TIME=de_DE.utf8 thunderbird

With Thunderbird 60 setting the LC_TIME or LANG variables will not help.
Can someone confirm this?

Regards,
Stefan

Sid: rpcbind not starting (was NFSv3 mounting problem)

[Grzegorz Sójka]

Hi there,

When I try to start rpc.bimd I get the following error in the log:

Aug 20 11:20:56 Zeus daemon.err rpcbind[2924]: cannot get local address 
for udp: Servname not supported for ai_socktype
Aug 20 11:20:56 Zeus daemon.err rpcbind[2924]: cannot get local address 
for tcp: Servname not supported for ai_socktype
Aug 20 11:20:56 Zeus daemon.err rpcbind[2924]: cannot get local address 
for udp6: Servname not supported for ai_socktype
Aug 20 11:20:56 Zeus daemon.err rpcbind[2924]: cannot get local address 
for tcp6: Servname not supported for ai_socktype


It's surprising because DNS is working fine:

$ host Zeus
Zeus.olimp has address 192.168.0.130

# host www.debian.org
www.debian.org has address 5.153.231.4
www.debian.org has address 130.89.148.14
www.debian.org has IPv6 address 2001:41c8:1000:21::21:4
www.debian.org has IPv6 address 2001:67c:2564:a119::148:14

# cat /etc/hosts
127.0.0.1   localhost
::1 localhost ip6-localhost ip6-loopback
ff02::1 ip6-allnodes
ff02::2 ip6-allrouters
192.168.0.130   Zeus.olimp Zeus

PS. System was installed using debootstrap with variant=minbase, missing 
packages or configuration files??


--
Pozdrawiam
Grzesiek

Wysłane z kompa wolnego od wirusów Billa Gatesa.

Kernel hang in Xen (4.9)

[Bastien Durel]

Hello,

I have a xen guest that hangs after a while if running 4.9 kernel
(since upgrade to stretch)
I get theses messages in console, then I cannot do anything. I get a
login prompt on console, but do not get shell prompt; SSH connexion is
established, I get my motd (even with updated last login time), but no
shell.

[260514.780118] INFO: task jbd2/xvda5-8:192 blocked for more than 120 seconds.
[260514.780132]   Not tainted 4.9.0-7-amd64 #1 Debian 4.9.110-3+deb9u2
[260514.780137] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables 
this message.
[260514.780145] jbd2/xvda5-8D0   192  2 0x
[260514.780156]  8800f1caec00  8800f0beb140 
8800f5c18980
[260514.780171]  81c11500 c900410bba70 8160fed9 
81090d5a
[260514.780183]  00ffe8c0b000 8800f5c18980 8130f759 
8800f0beb140
[260514.780196] Call Trace:
[260514.780239]  [] ? __schedule+0x239/0x6f0
[260514.780249]  [] ? queue_work_on+0x2a/0x40
[260514.780263]  [] ? blk_mq_flush_plug_list+0x139/0x160
[260514.780271]  [] ? bit_wait+0x50/0x50
[260514.780278]  [] ? schedule+0x32/0x80
[260514.780286]  [] ? schedule_timeout+0x1dd/0x380
[260514.780295]  [] ? xen_clocksource_get_cycles+0x11/0x20
[260514.780303]  [] ? bit_wait+0x50/0x50
[260514.780310]  [] ? io_schedule_timeout+0x9d/0x100
[260514.780319]  [] ? prepare_to_wait+0x57/0x80
[260514.780326]  [] ? bit_wait_io+0x17/0x60
[260514.780333]  [] ? __wait_on_bit+0x55/0x80
[260514.780344]  [] ? find_get_pages_tag+0x158/0x2e0
[260514.780353]  [] ? wait_on_page_bit+0x7f/0xa0
[260514.780360]  [] ? wake_atomic_t_function+0x60/0x60
[260514.780370]  [] ? __filemap_fdatawait_range+0xe0/0x140
[260514.780378]  [] ? filemap_fdatawait_range+0xf/0x30
[260514.780397]  [] ? 
jbd2_journal_commit_transaction+0x73d/0x17b0 [jbd2]
[260514.780408]  [] ? __switch_to_asm+0x34/0x70
[260514.780416]  [] ? __switch_to_asm+0x40/0x70
[260514.780424]  [] ? xen_load_sp0+0x84/0x170
[260514.780432]  [] ? finish_task_switch+0x7d/0x200
[260514.780440]  [] ? _raw_spin_unlock_irqrestore+0x16/0x20
[260514.780452]  [] ? kjournald2+0xc2/0x260 [jbd2]
[260514.780463]  [] ? prepare_to_wait_event+0xf0/0xf0
[260514.780474]  [] ? commit_timeout+0x10/0x10 [jbd2]
[260514.780482]  [] ? kthread+0xd9/0xf0
[260514.780489]  [] ? kthread_park+0x60/0x60
[260514.780496]  [] ? ret_from_fork+0x57/0x70

When running on jessie's 3.16 kernel, the guest is running fine.
I have other VMs on this host running 4.9 kernel without problem.

Does anyone have an idea about this bug ?

Thanks,

-- 
Bastien

Re: Thunderbird 60 ignores LC_TIME environment variable

[Curt]

On 2018-08-20, Stefan Pietsch  wrote:
> Dear list,
>
> after the upgrade from Thunderbird 52.9.1 to 60.0 (Debian sid) it seems
> for me that the environment variable LC_TIME is ignored.
>
> I used the following command to set 24h time format in Thunderbird.
>
> $ LC_TIME=de_DE.utf8 thunderbird
>
> With Thunderbird 60 setting the LC_TIME or LANG variables will not help.
> Can someone confirm this?

I can't confirm anything but this bug report

https://bugzilla.mozilla.org/show_bug.cgi?id=1426907

looks both pertinent, and, after a cursory read (recent posts re 60 toward the
bottom), discouraging concerning your plight (at least for the moment).

> Regards,
> Stefan
>
>


-- 
"She understands everything, recognizes everyone, but through her half sleep,
she simply cannot understand what power binds her hand and foot, oppresses her,
and keeps her from living."  -- Anton Chekhov, “Sleepy”

Re: Kernel hang in Xen (4.9)

[Steve Kemp]

> I have a xen guest that hangs after a while if running 4.9 kernel
> (since upgrade to stretch)

  You might try the 4.9.110-3 kernel from stretch-proposed-updates,
 which I noticed is required to fix a reboot-loop on stretch:

https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=903767

  That doesn't seem to match your situation 100% as you do
 seem to gain access, albeit briefly, but it might be worth a go.

> When running on jessie's 3.16 kernel, the guest is running fine.
> I have other VMs on this host running 4.9 kernel without problem.

  It is interesting that not all guests fail ..

Steve
-- 
https://steve.fi/

Re: Sid: NFSv3 mounting problem (again!)

[Grzegorz Sójka]

I added /etc/services /etc/rpc and now rpcbind is starting but:

#rpcinfo -p localhist
   program vers proto   port  service
104   tcp111  portmapper
103   tcp111  portmapper
102   tcp111  portmapper
104   udp111  portmapper
103   udp111  portmapper
102   udp111  portmapper
1000241   udp  49035  status
1000241   tcp  58571  status

there is no nlockmgr.

--
Pozdrawiam
Grzesiek

Wysłane z kompa wolnego od wirusów Billa Gatesa.

Re: Kernel hang in Xen (4.9)

[Bastien Durel]

Le 20/08/2018 à 11:58, Steve Kemp a écrit :

I have a xen guest that hangs after a while if running 4.9 kernel
(since upgrade to stretch)


   You might try the 4.9.110-3 kernel from stretch-proposed-updates,
  which I noticed is required to fix a reboot-loop on stretch:

 https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=903767

   That doesn't seem to match your situation 100% as you do
  seem to gain access, albeit briefly, but it might be worth a go.

I've noticied that, and tried to run 4.9.110-3+deb9u2 as 4.9.110-3 is 
out from stretch-proposed-updates.
I ran 4.9.110-3 too, same problem. (and almost every 4.9 version that 
where in repo since stretch is out, got a freeze each time, had to 
fallback each time)



When running on jessie's 3.16 kernel, the guest is running fine.
I have other VMs on this host running 4.9 kernel without problem.


   It is interesting that not all guests fail ..

Indeed, it is. But I don't figure WHY this guest fails. And as it's the 
"put-it-here-if-there-is-no-obvious-choice" machine of my network, 
there's many installed services on it.


--
Bastien

Re: yabasic problem

[Richard Owlett]

On 08/19/2018 08:50 AM, Richard Owlett wrote:

On 08/19/2018 08:36 AM, to...@tuxteam.de wrote:
[snip


   - try running your program under strace: there you'll see which
 system calls are being issued, and perhaps gain some insight
 on where this "No such file..." (ENOENT) really happens and
 what argument values are being passed to the system call.

 (as a very short intro:

   strace -o  yabasic
   [do your thing]
   [exit]
   [go carefully with a pitchfork through your trace file
    looking out for your file names and for ENOENT]

 Holler here if you need help with strace. It's a handy tool
 in any hacker's toolbox.


Didn't know of strace. Just glanced at manpage. On my way out - pursue 
this afternoon.




MIXED results.

In the yabasic mode I was using {enter basic program and hit Enter twice 
to execute}, strace and yabasic do not "play well together". The basic 
program is never executed.


However if you do:
  > strace -o mytrace yabasic test.bas
it executes, but doesn't apparently doesn't give any explanation ow why 
the two files cannot be opened.


*HOWEVER* perhaps my system and strace do not "play well together".
A tutorial I found gave as a first example:  > strace -o mytrace 
yabasic ls

That gave 10 {apparently spurious} file errors before running "ls" normally.

I then tried some brute force trouble shooting and discovered that the 
problem was specifying a complete path to target files. I moved them to 
my home directory and all worked fine.


Now to solve logic flaws in my code ;/
Thank you.

Re: yabasic problem

[Richard Owlett]

On 08/20/2018 02:35 AM, Thomas Schmitt wrote:

David Wright wrote:
 [snip] 

Would you agree, though, that "BASIC" is the language that must
have the biggest contrast between its well-endowed versions and
the most dire cr*p.


Well, back then i perceived HP BASIC as the best language of all. It made
me boss on all those expensive HP machines (from 9845B to 9000/320).
But C ran on all Unix workstations. And as soon as i became ambidextrous
enough, i fell in love with the display manager of the Apollo Domain DN3000.

Microsoft's Visual Basic is said to have surpassed HP BASIC in the years
later.



Where would yabasic fit?


I think it would rate fairly well. I browsed the BASICs in the 
repository and seemed best matched to my preferences, specifically no GUI.




It seems to be inspired by C (see the syntax of "open"). But why use such
a C-BASIC when there is gcc, gdb and valgrind ?


'Cause the last time I used C was about 4 *DECADES* ago ;}
Although I have programmed, I would claim to be a *programmer*.

Re: painted into a corner

[Eike Lantzsch]

On Sunday, August 19, 2018 5:51:24 PM -04 Gene Heskett wrote:
> On Sunday 19 August 2018 17:15:43 Eike Lantzsch wrote:
> > On Sunday, August 19, 2018 10:37:05 PM -04 john doe wrote:
> > > On 8/19/2018 9:40 PM, Gene Heskett wrote:
> > > > Greetings all;
> > > > 
> > > > I just installed stretch to a fresh 2T HD. letting it
> > > > autopartition and format for separate /, swap, /var and /home
> > > > partitions. But I didn't let it overwrite the grub on the 1st
> > > > drive it was/is  booting wheezy from.
> > > > 
> > > > I figured I'd mount it to wheezy and copy over my personal stuff,
> > > > like an email corpus well over 15GB reaching back to 2002.
> > > > 
> > > > But I can't mount much of the drive, / is all that will actually
> > > > mount, because the 2 versions of ext4 are incompatible, nearly all
> > > > the mount and e2tools can't touch the installers ext4 file
> > > > systems.
> > > > 
> > > > For instance, its not mounted:
> > > > gene@coyote:~$ e2fsck /dev/sdb8
> > > > e2fsck 1.42.5 (29-Jul-2012)
> > > > /dev/sdb8 has unsupported feature(s): metadata_csum
> > > > e2fsck: Get a newer version of e2fsck!
> > > > 
> > > > And of course whats installed to wheezy is the latest available
> > > > wheezy version of e2fsck.
> > > > 
> > > > Whats the recommended way to do these mounts so I can maintain as
> > > > much continuity as possible?
> > > 
> > > Maybe using apt-pinning.
> > > 
> > > In other words, installing the version of Stretch/Jessie on wheezy.
> > > 
> > > On the wheezy host, can't you backupt on an external hardware?
> > 
> > In those cases I usually boot a recent live-CD (or USB-stick) like
> > KNOPPIX and mount and copy from there.
> 
> I just rebooted to it, and I found a desktop interface that will wear out
> a set of batteries in my mouse weekly because it takes at least 8 to 10
> clicks and some scroll wheel work just to find a #@%# terminal, and it
> can't even add tabs to a different shell either! I didn't install
> anything special for a desktop, took the default because I intended to
> replace it with TDE asap, and going from 10 workspaces, 4 of which have
> multi-tabbed (up to 7 tabs each) konsoles running on them, with a
> pulldown text menu to run half the stuff I run on the other workspaces,
> to a single window, single tasking system thats worse than the last
> windows box I was asked to configure the networking on, was very
> disheartening.  So I added the trinity stuff to /etc/apt/sources.list.d
> using mc to copy that from the old disk, changing the wheezy in the deb
> line to stretch, and that did not get me the TDE desktop I've been using
> for years, but did get me some sort of a warning window that was taller
> than my screen, fussing that some repo I hadn't added, was duff. If
> debian is trying to kill itself, it was a heck of a good start, not even
> a windows user looking for something better would be impressed.
> 
> The only thing that Just Worked was the networking, it took everything
> for a static net and Just Worked on the reboot.  That was a rather
> pleasant surprise considering the only stretch based install on my
> rock64's that works at all was armbian.  None of the other arm|hf|64
> *bians will accept a gateway assignment except as as a route command
> after a login.
> 
> I'm burned out for today, my cataracts might have to be the next thing I
> fix.
With KNOPPIX to do such simple things as copying don't boot into the X-Window.
On boot enter: knoppix 2
(look at the cheat code)
and you have text only
use mc if you must or want

Sorry but I don't have time to wade through rants ...
All the best
Eike

Re: yabasic problem

[Mark Fletcher]

Isn’t the problem that you misspelled “experimental” in your original file
paths?

Mark
On Mon, Aug 20, 2018 at 21:13 Richard Owlett  wrote:

> On 08/20/2018 02:35 AM, Thomas Schmitt wrote:
> > David Wright wrote:
> >  [snip]
> >> Would you agree, though, that "BASIC" is the language that must
> >> have the biggest contrast between its well-endowed versions and
> >> the most dire cr*p.
> >
> > Well, back then i perceived HP BASIC as the best language of all. It made
> > me boss on all those expensive HP machines (from 9845B to 9000/320).
> > But C ran on all Unix workstations. And as soon as i became ambidextrous
> > enough, i fell in love with the display manager of the Apollo Domain
> DN3000.
> >
> > Microsoft's Visual Basic is said to have surpassed HP BASIC in the years
> > later.
> >
> >
> >> Where would yabasic fit?
>
> I think it would rate fairly well. I browsed the BASICs in the
> repository and seemed best matched to my preferences, specifically no GUI.
>
> >
> > It seems to be inspired by C (see the syntax of "open"). But why use such
> > a C-BASIC when there is gcc, gdb and valgrind ?
> >
> 'Cause the last time I used C was about 4 *DECADES* ago ;}
> Although I have programmed, I would claim to be a *programmer*.
>
>
>

Re: Repository Problem

[Greg Wooledge]

On Fri, Aug 17, 2018 at 10:32:39PM +0100, Brian wrote:
> On Fri 17 Aug 2018 at 21:49:53 +0200, john doe wrote:
> 
> > As an aside, you should consider using 'apt' and not 'apt-get'.
> 
> Why? What functional difference is there between 'apt-get upgrade' and
> 'apt-get install' and the apt equivalents?

I don't agree with john doe's assertion; I'm just answering the literal
question.

"apt upgrade" will install new packages, if necessary, to satisfy the
upgrade dependencies; "apt-get upgrade" will not.  This is quite visible
every time there's a new kernel ABI bump.

"apt upgrade" or "apt install ..." will remove the *.deb files that
it downloads into /var/cache/apt/archives on this particular run, but
not any pre-existing files there.  "apt-get" will not.

"apt" uses yellow text that is virtually unreadable on a white terminal
background.  "apt-get" does not; it doesn't use ANY colors.

Re: "accept_ra 1" vs "accept_ra 2" in interfaces configuration-file

[Martin T]

On Mon, Aug 20, 2018 at 7:55 AM Andy Smith  wrote:
>
> Hi Martin,
>
> On Mon, Aug 20, 2018 at 05:12:56AM +0300, Martin T wrote:
> > According to "man interfaces" "accept_ra 1" makes interface to accept
> > IPv6 RA messages. "accept_ra 2" does the same and in addition, it also
> > enables forwarding. What does the forwarding mean in this context? One
> > could think, that it modifies the /proc/sys/net/ipv6/conf/*/forwarding
> > file(s), but this does not seem to be the case.
>
> If forwarding = 1 then by default RAs will not be accepted. Setting
> accept_ra to 2 allows RAs to be accepted even when forwarding = 1.
>
> Changing the values of either forwarding or accept_ra does not alter
> the values of the other. Only the behaviour of the system.
>
> Back in 2011 this was a hard-won battle:
>
> 
> http://strugglers.net/~andy/blog/2011/09/04/linux-ipv6-router-advertisements-and-forwarding/
>
> Cheers,
> Andy
>

Hi Andy!

Thanks for this very informative blog post! However, setting the
"net.ipv6.conf.all.forwarding" to 1 in /etc/sysctl.conf and
"accept_ra" to 2 in /etc/network/interfaces for ISP facing
interface(eth0) didn't work for me. I expected SLAAC to work, but it
didn't. I'm running kernel version 4.9.0. Settings can be seen below:

# sysctl net.ipv6.conf.all.forwarding
net.ipv6.conf.all.forwarding = 1
#
# sysctl net.ipv6.conf.eth0.accept_ra
net.ipv6.conf.eth0.accept_ra = 2
#

When I set the "net.ipv6.conf.all.forwarding" to 0 and reboot the
router, then SLAAC works. What might cause this?


thanks,
Martin

Re: Sid: NFSv3 mounting problem (again!)

[Reco]

Hi.

On Mon, Aug 20, 2018 at 12:42:14PM +0200, Grzegorz Sójka wrote:
> I added /etc/services /etc/rpc and now rpcbind is starting but:
> 
> #rpcinfo -p localhist
>program vers proto   port  service
> 104   tcp111  portmapper
> 103   tcp111  portmapper
> 102   tcp111  portmapper
> 104   udp111  portmapper
> 103   udp111  portmapper
> 102   udp111  portmapper
> 1000241   udp  49035  status
> 1000241   tcp  58571  status
> 
> there is no nlockmgr.

I always said that part of the result is better that none at all.
This one is suitable for the client, and it advertizes NFS support from
version 2 to version 4 inclusive.
For NFS server I'd expect to see 'nfs' (v3 or v4).

Reco

Re: yabasic problem

[Thomas Schmitt]

Hi,

i wrote:
> > why use such a C-BASIC when there is gcc, gdb and valgrind ?

> 'Cause the last time I used C was about 4 *DECADES* ago ;}

This way you missed the everlasting joy of hunting bugs which trigger
SIGSEGV at some arbitrary point of program execution.
valgrind widely industrialized that hunt. But its reports can be nearly
as cryptic as the messages from gcc if you omitted a semicolon somewhere.

Nevertheless, between C and shell there is few room for BASIC.
C for what is complicated and must run fast.
shell for what is rather simple and must be programmed fast.


Back to the literal question:

I doubted that the "open" commands failed at all. A decent BASIC in dialog
mode should show the error message automatically. And the docs say that
a running BASIC program shall abort if "open" fails and is not wrapped in
an "if" command.
Instead i suspected that the messages of peek$("error") stem from other
earlier attempts to open files. You meanwhile reported that strace shows
such attempts.

Did you make experiments to verify or falsify my theory ?
E.g. by provoking an error different from "File not found" and then
checking whether this message sticks to peek().


But Mark Fletcher has a different theory, which should be verified:
> Isn’t the problem that you misspelled “experimental” in your original file
> paths?

Your directory name "expermental_copies" looks fishy indeed.

Did you really test by copy+paste to an "ls" command in a shell terminal
that your path in the BASIC commands is correct ?
Don't trust your eyes. Trust your mouse.


Have a nice day :)

Thomas

Re: Repository Problem

[john doe]

On 8/20/2018 2:54 PM, Greg Wooledge wrote:

On Fri, Aug 17, 2018 at 10:32:39PM +0100, Brian wrote:

On Fri 17 Aug 2018 at 21:49:53 +0200, john doe wrote:


As an aside, you should consider using 'apt' and not 'apt-get'.


Why? What functional difference is there between 'apt-get upgrade' and
'apt-get install' and the apt equivalents?




See below.


I don't agree with john doe's assertion; I'm just answering the literal
question.



Reading Debian wiki 'apt-get' is no longer the recommended way to manage 
pkgs:


https://wiki.debian.org/DebianPackageManagement


"apt upgrade" will install new packages, if necessary, to satisfy the
upgrade dependencies; "apt-get upgrade" will not.  This is quite visible
every time there's a new kernel ABI bump.



This behavior is expected if you read apt(8):

https://manpages.debian.org/stretch/apt/apt.8.en.html

--
John Doe

Re: yabasic problem

[Greg Wooledge]

On Mon, Aug 20, 2018 at 03:26:20PM +0200, Thomas Schmitt wrote:
> Nevertheless, between C and shell there is few room for BASIC.
> C for what is complicated and must run fast.
> shell for what is rather simple and must be programmed fast.

In between C and shell, you've got a huge number of scripting languages,
like python, perl, tcl, ruby, etc.  I haven't seen anyone use BASIC
in years, because these other languages are generally better.

But if BASIC is working for someone out there, then hey, great.

Re: yabasic problem

[Richard Owlett]

On 08/20/2018 07:29 AM, Mark Fletcher wrote:
Isn’t the problem that you misspelled “experimental” in your original 
file paths?


No 
I had misspelled it when creating the directory.
And, being a lousy typist, I copy-n-past whenever possible.
I had used Caja's Properties menu to get the path.
But there was a typo which I caught after my initial initial post
[A missing "." before the extension of the input file].



Mark
On Mon, Aug 20, 2018 at 21:13 Richard Owlett > wrote:


On 08/20/2018 02:35 AM, Thomas Schmitt wrote:
 > David Wright wrote:
 >  [snip]
 >> Would you agree, though, that "BASIC" is the language that must
 >> have the biggest contrast between its well-endowed versions and
 >> the most dire cr*p.
 >
 > Well, back then i perceived HP BASIC as the best language of all.
It made
 > me boss on all those expensive HP machines (from 9845B to 9000/320).
 > But C ran on all Unix workstations. And as soon as i became
ambidextrous
 > enough, i fell in love with the display manager of the Apollo
Domain DN3000.
 >
 > Microsoft's Visual Basic is said to have surpassed HP BASIC in
the years
 > later.
 >
 >
 >> Where would yabasic fit?

I think it would rate fairly well. I browsed the BASICs in the
repository and seemed best matched to my preferences, specifically
no GUI.

 >
 > It seems to be inspired by C (see the syntax of "open"). But why
use such
 > a C-BASIC when there is gcc, gdb and valgrind ?
 >
'Cause the last time I used C was about 4 *DECADES* ago ;}
Although I have programmed, I would claim to be a *programmer*.

Re: painted into a corner

[Michael Stone]

On Sun, Aug 19, 2018 at 03:40:55PM -0400, Gene Heskett wrote:

I just installed stretch to a fresh 2T HD. letting it autopartition and
format for separate /, swap, /var and /home partitions. But I didn't let
it overwrite the grub on the 1st drive it was/is  booting wheezy from.

I figured I'd mount it to wheezy and copy over my personal stuff, like an
email corpus well over 15GB reaching back to 2002.

But I can't mount much of the drive, / is all that will actually mount,
because the 2 versions of ext4 are incompatible, nearly all the mount
and e2tools can't touch the installers ext4 file systems.


If you unmounted cleanly you don't need fsck. Install a newer kernel 
from wheezy backports and it should mount fine. 


Mike Stone

Re: yabasic problem

[Thomas Schmitt]

Hi,

Greg Wooledge wrote:
> But if BASIC is working for someone out there, then hey, great.

Sure. I am still wondering what will be the final solution to Richard's
riddle. (The typo theory seems to have died now. Remains the sticky error
theory.)


Have a nice day :)

Thomas

Re: Sid: NFSv3 mounting problem (again!)

[Grzegorz Sójka]

On 08/20/18 15:12, Reco wrote:

Hi.

On Mon, Aug 20, 2018 at 12:42:14PM +0200, Grzegorz Sójka wrote:

I added /etc/services /etc/rpc and now rpcbind is starting but:

#rpcinfo -p localhist
program vers proto   port  service
 104   tcp111  portmapper
 103   tcp111  portmapper
 102   tcp111  portmapper
 104   udp111  portmapper
 103   udp111  portmapper
 102   udp111  portmapper
 1000241   udp  49035  status
 1000241   tcp  58571  status

there is no nlockmgr.


I always said that part of the result is better that none at all.
This one is suitable for the client, and it advertizes NFS support from
version 2 to version 4 inclusive.
For NFS server I'd expect to see 'nfs' (v3 or v4).


If I'm getting this right nlockmgr is needed only on the server side?

Anyway, I still get the following error:
# mount -v /mnt/users
mount.nfs: timeout set for Sat Aug 18 18:55:52 2018
mount.nfs: trying text-based options 'nfsvers=3,addr=192.168.0.129'
mount.nfs: prog 13, trying vers=3, prot=6
mount.nfs: trying 192.168.0.129 prog 13 vers 3 prot TCP port 2049
mount.nfs: prog 15, trying vers=3, prot=17
mount.nfs: trying 192.168.0.129 prog 15 vers 3 prot UDP port 37385
mount.nfs: Protocol not supported

On all the working clients I do have:

# rpcinfo -p
   program vers proto   port  service
104   tcp111  portmapper
103   tcp111  portmapper
102   tcp111  portmapper
104   udp111  portmapper
103   udp111  portmapper
102   udp111  portmapper
1000211   udp  44149  nlockmgr
1000213   udp  44149  nlockmgr
1000214   udp  44149  nlockmgr
1000211   tcp  37814  nlockmgr
1000213   tcp  37814  nlockmgr
1000214   tcp  37814  nlockmgr
1000241   udp  52987  status
1000241   tcp  55392  status

So, nlockmgr is missing only on the broken client.


--
Pozdrawiam
Grzesiek

Wysłane z kompa wolnego od wirusów Billa Gatesa.

Re: Sid: NFSv3 mounting problem (again!)

[Michael Stone]

On Mon, Aug 20, 2018 at 12:42:14PM +0200, Grzegorz Sójka wrote:
I added /etc/services /etc/rpc 


What on earth does this mean? How did the system get into such a state 
that these were missing?


Mike Stone

Re: painted into a corner

[Gene Heskett]

On Monday 20 August 2018 08:29:14 Eike Lantzsch wrote:

> On Sunday, August 19, 2018 5:51:24 PM -04 Gene Heskett wrote:
> > On Sunday 19 August 2018 17:15:43 Eike Lantzsch wrote:
> > > On Sunday, August 19, 2018 10:37:05 PM -04 john doe wrote:
> > > > On 8/19/2018 9:40 PM, Gene Heskett wrote:
> > > > > Greetings all;
> > > > >
> > > > > I just installed stretch to a fresh 2T HD. letting it
> > > > > autopartition and format for separate /, swap, /var and /home
> > > > > partitions. But I didn't let it overwrite the grub on the 1st
> > > > > drive it was/is  booting wheezy from.
> > > > >
> > > > > I figured I'd mount it to wheezy and copy over my personal
> > > > > stuff, like an email corpus well over 15GB reaching back to
> > > > > 2002.
> > > > >
> > > > > But I can't mount much of the drive, / is all that will
> > > > > actually mount, because the 2 versions of ext4 are
> > > > > incompatible, nearly all the mount and e2tools can't touch the
> > > > > installers ext4 file systems.
> > > > >
> > > > > For instance, its not mounted:
> > > > > gene@coyote:~$ e2fsck /dev/sdb8
> > > > > e2fsck 1.42.5 (29-Jul-2012)
> > > > > /dev/sdb8 has unsupported feature(s): metadata_csum
> > > > > e2fsck: Get a newer version of e2fsck!
> > > > >
> > > > > And of course whats installed to wheezy is the latest
> > > > > available wheezy version of e2fsck.
> > > > >
> > > > > Whats the recommended way to do these mounts so I can maintain
> > > > > as much continuity as possible?
> > > >
> > > > Maybe using apt-pinning.
> > > >
> > > > In other words, installing the version of Stretch/Jessie on
> > > > wheezy.
> > > >
> > > > On the wheezy host, can't you backupt on an external hardware?
> > >
> > > In those cases I usually boot a recent live-CD (or USB-stick) like
> > > KNOPPIX and mount and copy from there.
> >
> > I just rebooted to it, and I found a desktop interface that will
> > wear out a set of batteries in my mouse weekly because it takes at
> > least 8 to 10 clicks and some scroll wheel work just to find a #@%#
> > terminal, and it can't even add tabs to a different shell either! I
> > didn't install anything special for a desktop, took the default
> > because I intended to replace it with TDE asap, and going from 10
> > workspaces, 4 of which have multi-tabbed (up to 7 tabs each)
> > konsoles running on them, with a pulldown text menu to run half the
> > stuff I run on the other workspaces, to a single window, single
> > tasking system thats worse than the last windows box I was asked to
> > configure the networking on, was very disheartening.  So I added the
> > trinity stuff to /etc/apt/sources.list.d using mc to copy that from
> > the old disk, changing the wheezy in the deb line to stretch, and
> > that did not get me the TDE desktop I've been using for years, but
> > did get me some sort of a warning window that was taller than my
> > screen, fussing that some repo I hadn't added, was duff. If debian
> > is trying to kill itself, it was a heck of a good start, not even a
> > windows user looking for something better would be impressed.
> >
> > The only thing that Just Worked was the networking, it took
> > everything for a static net and Just Worked on the reboot.  That was
> > a rather pleasant surprise considering the only stretch based
> > install on my rock64's that works at all was armbian.  None of the
> > other arm|hf|64 *bians will accept a gateway assignment except as as
> > a route command after a login.
> >
> > I'm burned out for today, my cataracts might have to be the next
> > thing I fix.
>
> With KNOPPIX to do such simple things as copying don't boot into the
> X-Window. On boot enter: knoppix 2
> (look at the cheat code)
> and you have text only
> use mc if you must or want
>
> Sorry but I don't have time to wade through rants ...

Sometimes they can be educational, and I try to make mine so if you can 
get past my frustration, but in this case its partly my own fault, a 
naming confusion on my part, I had put the wrong dvd-1 for 9.4 in the 
reader, when I should have used the adapted stretch from the LinuxCNC 
guys. I have now burned that one, and if I don't install a desktop at 
all, then my tde install will be clean and hopefully uncontaminated by 
what is obviously a broken gnome. Trying to make it look and run like 
windows 95 is IMO a huge mistake, but one that I must say debian has 
done remarkably well. Bring back the pulldown text stuff where 2-3 
clicks gets you anything installed.

> All the best
> Eike

Thanks Eike.

-- 
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: Sid: NFSv3 mounting problem (again!)

[Reco]

Hi.

On Mon, Aug 20, 2018 at 04:04:01PM +0200, Grzegorz Sójka wrote:
> On 08/20/18 15:12, Reco wrote:
> > Hi.
> > 
> > On Mon, Aug 20, 2018 at 12:42:14PM +0200, Grzegorz Sójka wrote:
> > > I added /etc/services /etc/rpc and now rpcbind is starting but:
> > > 
> > > #rpcinfo -p localhist
> > > program vers proto   port  service
> > >  104   tcp111  portmapper
> > >  103   tcp111  portmapper
> > >  102   tcp111  portmapper
> > >  104   udp111  portmapper
> > >  103   udp111  portmapper
> > >  102   udp111  portmapper
> > >  1000241   udp  49035  status
> > >  1000241   tcp  58571  status
> > > 
> > > there is no nlockmgr.
> > 
> > I always said that part of the result is better that none at all.
> > This one is suitable for the client, and it advertizes NFS support from
> > version 2 to version 4 inclusive.
> > For NFS server I'd expect to see 'nfs' (v3 or v4).
> 
> If I'm getting this right nlockmgr is needed only on the server side?

And only if you're using NFSv3 and your mount options does not include
'nolock'. The question is - what's on the server?


> Anyway, I still get the following error:
> # mount -v /mnt/users
> mount.nfs: timeout set for Sat Aug 18 18:55:52 2018
> mount.nfs: trying text-based options 'nfsvers=3,addr=192.168.0.129'
> mount.nfs: prog 13, trying vers=3, prot=6
> mount.nfs: trying 192.168.0.129 prog 13 vers 3 prot TCP port 2049

A client tries NFSv3 via tcp:2049 and presumably meets TCP RST.

> mount.nfs: prog 15, trying vers=3, prot=17
> mount.nfs: trying 192.168.0.129 prog 15 vers 3 prot UDP port 37385
> mount.nfs: Protocol not supported

A server says to the client that NFSv3 over udp is not supported here.

But without rpcinfo from the server it's impossble to tell what's really
supported.


> On all the working clients I do have:
> 
> # rpcinfo -p
>program vers proto   port  service
> 104   tcp111  portmapper
> 103   tcp111  portmapper
> 102   tcp111  portmapper
> 104   udp111  portmapper
> 103   udp111  portmapper
> 102   udp111  portmapper
> 1000211   udp  44149  nlockmgr
> 1000213   udp  44149  nlockmgr
> 1000214   udp  44149  nlockmgr
> 1000211   tcp  37814  nlockmgr
> 1000213   tcp  37814  nlockmgr
> 1000214   tcp  37814  nlockmgr
> 1000241   udp  52987  status
> 1000241   tcp  55392  status
> 
> So, nlockmgr is missing only on the broken client.

Note that your client advertizes NFSv4 support. Maybe that's what they
really use?

Reco

Re: Sid: NFSv3 mounting problem (again!)

[Michael Stone]

On Mon, Aug 20, 2018 at 05:25:07PM +0300, Reco wrote:

Note that your client advertizes NFSv4 support. Maybe that's what they
really use?


The output of `mount` or /proc/mounts will show what's actually in use.

Mike Stone

Re: painted into a corner

[Gene Heskett]

On Monday 20 August 2018 09:37:05 Michael Stone wrote:

> On Sun, Aug 19, 2018 at 03:40:55PM -0400, Gene Heskett wrote:
> >I just installed stretch to a fresh 2T HD. letting it autopartition
> > and format for separate /, swap, /var and /home partitions. But I
> > didn't let it overwrite the grub on the 1st drive it was/is  booting
> > wheezy from.
> >
> >I figured I'd mount it to wheezy and copy over my personal stuff,
> > like an email corpus well over 15GB reaching back to 2002.
> >
> >But I can't mount much of the drive, / is all that will actually
> > mount, because the 2 versions of ext4 are incompatible, nearly all
> > the mount and e2tools can't touch the installers ext4 file systems.
>
> If you unmounted cleanly you don't need fsck. Install a newer kernel
> from wheezy backports and it should mount fine.
>
> Mike Stone

sudo reboot doesn't unmount the system drive cleanly enough? Then I'd 
call it a bug.

And my wheezy kernel is
gene@coyote:/$ uname -a
Linux coyote 3.16.0-0.bpo.4-amd64 #1 SMP Debian 3.16.39-1+deb8u1~bpo70+1 
(2017-02-24) x86_64 GNU/Linux
Which is not pinned and the machine is updated about 2x a week.


-- 
Cheers Mike, 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: Group ID conflicts between different distros: how to manage them with NIS?

[Joao Roscoe]

Excellent.
Time to read about udev.
Will do my homework, try a few tests, and will get back with results and,
possibly, more questions :-D
Thank you, everybody for your valuable help (and time).

Best regards,
João

On Mon, Aug 20, 2018 at 2:54 AM Reco  wrote:

> Hi.
>
> On Mon, Aug 20, 2018 at 12:51:24AM -0300, Joao Roscoe wrote:
> > Hmmm...
> >
> > If I create a NIS group (with a high ID), called serial_ports, dhould I
> > just, as root, chgrp /dev/ttyS0 so that it's group is serial_ports ?
>
> You could, and it may even work, but it would be temporary.
> To make it truly work you should write your own udev rule for these (and
> other) devices.
>
> The reason being - udev creates everything under the /dev (system boot).
> Udev changes everything under the /dev (vt switch, user relogins).
>
> In that particular case you should override changes made by
> /lib/udev/rules.d/50-udev-default.rules.
>
> Reco
>
>

[OT] Re: yabasic problem

[David Wright]

On Mon 20 Aug 2018 at 09:35:35 (+0200), Thomas Schmitt wrote:
> Hi,
> 
> i wrote:
> > > (Found the booklet. It's HP BASIC 3.0, not 2.0. Newest techology of 1985.)
> 
> David Wright wrote:
> > I thought we were up to version 4.0¹ by 1985,
> 
> Indeed, the booklet says "June 1984 ... First Edition".
> 
> I think i did not get to BASIC 4.0 because in 1986 i wrote a BASIC program
> which translated our other BASIC programs to C (with some handwork being
> left to do).
> 
> 
> > Would you agree, though, that "BASIC" is the language that must
> > have the biggest contrast between its well-endowed versions and
> > the most dire cr*p.
> 
> Well, back then i perceived HP BASIC as the best language of all. It made
> me boss on all those expensive HP machines (from 9845B to 9000/320).

For our purposes, that was true, and consequently we spent £20,000 on
a 9845 with a £1000/year contract (1979 prices). The back was well
populated with 16-bit GPIO, HPIB and RS232 interfaces. Later we bought
a 9836, before makiing the transition to HTBasic on 486DX/DOS5.0.

If they had a weak point, it was those little tape drives; I lost
count of the number we had replaced, and memories of rethreading the
tape under that plastic drive band. But I gained much kudos with
the engineers when I managed to rebuild one of the cassettes where
the band had come completely off.

> But C ran on all Unix workstations. And as soon as i became ambidextrous
> enough, i fell in love with the display manager of the Apollo Domain DN3000.
> 
> Microsoft's Visual Basic is said to have surpassed HP BASIC in the years
> later.

DOS6.22 was an incredibly stable platform for HTBasic. We ran NFS
clients (into linux servers) on it too (they only output; no input)
as well. The idea of exposing our instruments to Windows of the time
was laughable. These machines had to run non-stop with complete
reliability.

[…]

> > If you've not come across the 9845 machine, it was in my experience
> > unique, in that you could edit the program while it was executing,
> > not even having to pause it.
> 
> Yep. I have seen users countering Error 17 by hitting the EDIT key and
> throwing out the offending line. (One can imagine the effect on the
> collection of emerged CAD files. We got much less repair work after the
> users were re-seated to Apollo, DEC, and Sun where they had to submit
> a bug report when the program crashed.)

Ah, well there's the difference. When you're developing software
interactively that's controlling a machine analysing a sample (which
represents a foreign field trip and a week's chemistry lab work by
the person sitting next to you) and said sample is burning away in a
vacuum on an incandescent filament with a lifetime of perhaps 20
minutes, a bug report is their saying "Er, Houston, we have a
problem". That's peer pressure.

It was very rare to actually need that particular editing facility
(which IIRC you didn't get with the 9836). But the ability to
interrogate and set any variable in the current context (which could
be made very large with COM statements) was invaluable, as was the
ability to PAUSE, EDIT, RUN and CONTINUE within seconds and be back
running your sample.

Cheers,
David.

Re: Sid: NFSv3 mounting problem (again!)

[Grzegorz Sójka]

The package netbase was missing.

Thanks for your help

--
Pozdrawiam
Grzesiek

Wysłane z kompa wolnego od wirusów Billa Gatesa.

Re: yabasic problem

[tomas]

-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

On Mon, Aug 20, 2018 at 06:55:19AM -0500, Richard Owlett wrote:

[...]

> However if you do:
>   > strace -o mytrace yabasic test.bas
> it executes, but doesn't apparently doesn't give any explanation ow
> why the two files cannot be opened.

Hm. You will have realized that strace's output is extremely chatty:
you'll see the open for all possible libraries your program tries
to get hold of, then typically, after that succeeded, mmaping those
libraries into the process's address space, many failed attempts
at opening config and auxiliary files (in search of them at different
possible locations) until the program is where you consider it "up
and running".

You'll have to wade through it. Ah -- and if the program is just a
wrapper calling other processes, you'll have to tell strace to follow
them (option -f). For example (I use the option '-f', although I know
that 'cat' doesn't fork):

  strace -o /tmp/trace -f cat /no/such/file

I get the (expected) error:

  cat: /no/such/file: No such file or directory

and my /tmp/trace contains:

  4157  execve("/bin/cat", ["cat", "/no/such/file"], [/* 26 vars */]) = 0
  4157  brk(NULL) = 0x55850a279000
  4157  access("/etc/ld.so.nohwcap", F_OK) = -1 ENOENT (No such file or 
directory)
  4157  mmap(NULL, 12288, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 
0) = 0x7f7c303fc000

  # here you see the loading of the binary (execve), allocating memory (brk),
  # start of the search for the "interpreter" for the binary, etc.

  # lots of lines elided (roughly 85, in your program they'll be
  # quite a few more

  4157  fstat(1, {st_mode=S_IFCHR|0620, st_rdev=makedev(136, 2), ...}) = 0
  4157  open("/no/such/file", O_RDONLY)   = -1 ENOENT (No such file or 
directory)

  # this is the failed attempt at opening "/no/such/file"

  4157  write(2, "cat: ", 5)  = 5
  4157  write(2, "/no/such/file", 13) = 13
  4157  open("/usr/share/locale/C.UTF-8/LC_MESSAGES/libc.mo", O_RDONLY) = -1 
ENOENT (No such file or directory)
  4157  open("/usr/share/locale/C.utf8/LC_MESSAGES/libc.mo", O_RDONLY) = -1 
ENOENT (No such file or directory)
  4157  open("/usr/share/locale/C/LC_MESSAGES/libc.mo", O_RDONLY) = -1 ENOENT 
(No such file or directory)
  4157  write(2, ": No such file or directory", 27) = 27
  4157  write(2, "\n", 1) = 1

  # and this is the output of the error message to stderr (file descriptor 2)

  4157  close(1)  = 0
  4157  close(2)  = 0
  4157  exit_group(1) = ?
  4157  +++ exited with 1 +++

  # cleanup and exit.

> *HOWEVER* perhaps my system and strace do not "play well together".
> A tutorial I found gave as a first example:  > strace -o mytrace
> yabasic ls
> That gave 10 {apparently spurious} file errors before running "ls" normally.

See above: since libraries and other auxiliary files can be at one of
several places, many open errors are just the normal noise due to "looking
around".

Strace's output can be at first overwhelming, but it's a good tool
to have around.

> Now to solve logic flaws in my code ;/

Ah, that's the true joy :-)

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

iEYEARECAAYFAlt62hcACgkQBcgs9XrR2kauygCfYQHujYlJno4qvZ7bqOKvSKBG
CN4An3yyNA0JFRQiyF/5uce+yIbNtuzv
=h0BJ
-END PGP SIGNATURE-

Re: painted into a corner

[Michael Stone]

On Mon, Aug 20, 2018 at 10:33:29AM -0400, Gene Heskett wrote:

sudo reboot doesn't unmount the system drive cleanly enough? Then I'd
call it a bug.


I'm sure you would. But for all of your wall-of-text ranting, you failed 
to ever provide basic information like "what is the result of trying to 
mount the partition" so we're left guessing here.


Mike Stone

Re: Wanted - Debian(preferred)/Linux handheld

[Glenn English]

On Sun, Aug 19, 2018 at 1:35 PM Eric S Fraga  wrote:
>
> On Thursday, 16 Aug 2018 at 14:28, Glenn English wrote:
> > It's all over Amazon (search: planet gemini pda computer), but, as
> > best I can tell, there's no computer.
>
> I'm not sure I understand what you mean by "there's no computer."  The
> Gemini is not vapourware.  It exists.

When I looked the other day, there were lots of books and such, but no
hardware. Gemini was there, but their computer wasn't.

-- 
Glenn English

Re: painted into a corner

[Andrew McGlashan]

On 20/08/18 05:40, Gene Heskett wrote:
> Whats the recommended way to do these mounts so I can maintain as much 
> continuity as possible?

Those other areas, are they logical volumes perhaps?  lvms.

Cheers
A.

Sid: NETDEV WATCHDOG: eth1 (r8169): transmit queue 0 timed out

[Grzegorz Sójka]

Hi there,

I have following integrated network adapter:

# lspci -s 02:00
02:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd. 
RTL8111/8168/8411 PCI Express Gigabit Ethernet Controller (rev 06)


By a heavy traffic i get the following error:
NETDEV WATCHDOG: eth1 (r8169): transmit queue 0 timed out

As a workaround a turned off jumbo frames support. Is there any other 
solution making it possible to use jumbo frames??


Thanks for any help in advance.

--
Pozdrawiam
Grzesiek

Wysłane z kompa wolnego od wirusów Billa Gatesa.

Re: yabasic problem

[Richard Owlett]

On 08/20/2018 08:49 AM, Thomas Schmitt wrote:

Hi,

Greg Wooledge wrote:

But if BASIC is working for someone out there, then hey, great.


Sure. I am still wondering what will be the final solution to Richard's
riddle. (The typo theory seems to have died now. Remains the sticky error
theory.)



I'll attempt to claim at least minimal sanity and common sense.
This project started out in Tcl, which I visualize as a super-BASIC.

Project precis: SeaMonkey can save bookmarks in two formats.
JSON which is lossless but not suitable for human.
HTML which is human friendly, but loses hierarchical information 
[critical for my purposes]

Due to inexperience I got lost in how Tcl handled my JSON data.
Now that I have yabasic reading/writing files I've solved my program 
logic problems. It's losing critical data. I have a strong suspicion 
that when it reads my input file it sees some data as a line terminator.


Now that I know my program logic is valid, I'll likely translate it into 
Tcl for which there is a friendly debugging tool - the code pane of 
tcltutor30b7.


As to:
   > strace -o mytrace2 ls
I did a power off/power on sequence. I got a 9.5kb file with 10 
instances of "ENOENT (No such file or directory)". If anyone is 
interested, I'll send them a copy.


As to:> As to problem doing:> Did you make experiments to verify or 
falsify my theory ?

E.g. by provoking an error different from "File not found" and then
checking whether this message sticks to peek().


With a good filename, nothing printed.
With a bad filename, it reports file not found.

Re: Sid: NETDEV WATCHDOG: eth1 (r8169): transmit queue 0 timed out

[Dan Ritter]

On Mon, Aug 20, 2018 at 05:57:16PM +, Grzegorz Sójka wrote:
> Hi there,
> 
> I have following integrated network adapter:
> 
> # lspci -s 02:00
> 02:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd.
> RTL8111/8168/8411 PCI Express Gigabit Ethernet Controller (rev 06)
> 
> By a heavy traffic i get the following error:
> NETDEV WATCHDOG: eth1 (r8169): transmit queue 0 timed out
> 
> As a workaround a turned off jumbo frames support. Is there any other
> solution making it possible to use jumbo frames??
> 
> Thanks for any help in advance.
> 

Looks like this:

https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1752772

so at least you're neither alone nor crazy.

-dsr-

Re: Wanted - Debian(preferred)/Linux handheld

[Tony van der Hoff]

On 20/08/18 16:16, Glenn English wrote:
> On Sun, Aug 19, 2018 at 1:35 PM Eric S Fraga  wrote:
>> On Thursday, 16 Aug 2018 at 14:28, Glenn English wrote:
>>> It's all over Amazon (search: planet gemini pda computer), but, as
>>> best I can tell, there's no computer.
>> I'm not sure I understand what you mean by "there's no computer."  The
>> Gemini is not vapourware.  It exists.
> When I looked the other day, there were lots of books and such, but no
> hardware. Gemini was there, but their computer wasn't.
>
Try Here:
https://www.amazon.co.uk/Gemini-4G-WiFi-Space-Grey/dp/B07DNG5YMG/ref=sr_1_1?ie=UTF8&qid=1534779869&sr=8-1&keywords=gemini+pda


-- 
Tony van der Hoff| mailto:t...@vanderhoff.org
Buckinghamshire, England |

Re: Sid: NETDEV WATCHDOG: eth1 (r8169): transmit queue 0 timed out

[Grzegorz Sójka]

On 08/20/18 16:05, Dan Ritter wrote:

On Mon, Aug 20, 2018 at 05:57:16PM +, Grzegorz Sójka wrote:

Hi there,

I have following integrated network adapter:

# lspci -s 02:00
02:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd.
RTL8111/8168/8411 PCI Express Gigabit Ethernet Controller (rev 06)

By a heavy traffic i get the following error:
NETDEV WATCHDOG: eth1 (r8169): transmit queue 0 timed out

As a workaround a turned off jumbo frames support. Is there any other
solution making it possible to use jumbo frames??

Thanks for any help in advance.



Looks like this:

https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1752772

so at least you're neither alone nor crazy.


This is slightly different problem. I don't use suspend (this is desktop 
machine) and they do not mention jumbo frames.


--
Pozdrawiam
Grzesiek

Wysłane z kompa wolnego od wirusów Billa Gatesa.

Re: yabasic problem

[Thomas Schmitt]

Hi,

Note: this mail is purely nostalgic and not about yaboot.

David Wright wrote:
> [HP 9845B] If they had a weak point, it was those little tape drives

I once had a hidden line removal program reading a 3D model from one
tape and writing the intermediate result to the other. Then i had to
change the first tape, load the next program, insert the final result
tape, and let the program gnaw the data back from the intermediate tape.
One could hear the computer work. When it was done, it beeped and
printed a screen dump on the thermo printer between the tapes.
And of course neither model nor result were allowed to be larger than
200 KB.


> sample is burning away in a
> vacuum on an incandescent filament with a lifetime of perhaps 20
> minutes, a bug report is their saying "Er, Houston, we have a
> problem". That's peer pressure.

This gives real time programming a whole new meaning.


> But the ability to
> interrogate and set any variable in the current context (which could
> be made very large with COM statements) was invaluable, as was the
> ability to PAUSE, EDIT, RUN and CONTINUE within seconds and be back
> running your sample.

Well, back then we were stupid enough to do such things and smart enough
to get away with them. OPTION BASE 1.


Have a nice day :)

Thomas

Re: Wanted - Debian(preferred)/Linux handheld

[Anders Andersson]

On Mon, Aug 20, 2018 at 5:47 PM, Tony van der Hoff  wrote:
> On 20/08/18 16:16, Glenn English wrote:
>> On Sun, Aug 19, 2018 at 1:35 PM Eric S Fraga  wrote:
>>> On Thursday, 16 Aug 2018 at 14:28, Glenn English wrote:
 It's all over Amazon (search: planet gemini pda computer), but, as
 best I can tell, there's no computer.
>>> I'm not sure I understand what you mean by "there's no computer."  The
>>> Gemini is not vapourware.  It exists.
>> When I looked the other day, there were lots of books and such, but no
>> hardware. Gemini was there, but their computer wasn't.
>>
> Try Here:
> https://www.amazon.co.uk/Gemini-4G-WiFi-Space-Grey/dp/B07DNG5YMG/ref=sr_1_1?ie=UTF8&qid=1534779869&sr=8-1&keywords=gemini+pda

OP stated in the initial email and a follow-up to the first person who
answered, that he is looking for a US distributor. That's what people
have been talking about. If you can find it on amazon.com, please
share the link, because I can't find the actual computer.

Re: Wanted - Debian(preferred)/Linux handheld

[Richard Owlett]

On 08/20/2018 12:01 PM, Anders Andersson wrote:

On Mon, Aug 20, 2018 at 5:47 PM, Tony van der Hoff  wrote:

On 20/08/18 16:16, Glenn English wrote:

On Sun, Aug 19, 2018 at 1:35 PM Eric S Fraga  wrote:

On Thursday, 16 Aug 2018 at 14:28, Glenn English wrote:

It's all over Amazon (search: planet gemini pda computer), but, as
best I can tell, there's no computer.

I'm not sure I understand what you mean by "there's no computer."  The
Gemini is not vapourware.  It exists.

When I looked the other day, there were lots of books and such, but no
hardware. Gemini was there, but their computer wasn't.


Try Here:
https://www.amazon.co.uk/Gemini-4G-WiFi-Space-Grey/dp/B07DNG5YMG/ref=sr_1_1?ie=UTF8&qid=1534779869&sr=8-1&keywords=gemini+pda


OP stated in the initial email and a follow-up to the first person who
answered, that he is looking for a US distributor. That's what people
have been talking about. If you can find it on amazon.com, please
share the link, because I can't find the actual computer.



I also stated I wished to avoid Amazon and Ebay.
From when they both first opened I couldn't figure what there role is 
in a transaction like this.
As to requiring a US distributor, it would essentially guarantee that 
the units have FCC type acceptance. That can be a pain - decades ago 
that was part of my job.
That the hardware is hard to find on Amazon causes me to suspect they 
haven't received that yet.

Re: Sid: NETDEV WATCHDOG: eth1 (r8169): transmit queue 0 timed out

[Michael Stone]

On Mon, Aug 20, 2018 at 05:57:16PM +, Grzegorz Sójka wrote:

By a heavy traffic i get the following error:
NETDEV WATCHDOG: eth1 (r8169): transmit queue 0 timed out

As a workaround a turned off jumbo frames support. Is there any other 
solution making it possible to use jumbo frames??


Which kernel is having the problems? Does it work with a different 
kernel?


What does 
dmesg | grep r8169
show? (you can delete the MAC address from that if you'd prefer) How about 
ethtool -k eth1 | grep :\ on


Mike Stone

Re: yabasic problem

[Thomas Schmitt]

Hi,

Richard Owlett wrote:
> With a good filename, nothing printed.
> With a bad filename, it reports file not found.

But it does not abort at that point, does it ?

When the file name is bad, is the resulting filenumber 0 ?
If not: Can you read or write to that filenumber ?

I cling so massively to the theory of a sticky error indicator because
that is exactly what the indicator variable "errno" does in C and what
is described in man 3 errno.


Read The Source Luke:

peek$("error") is probably implemented in
  https://sources.debian.org/src/yabasic/1:2.79.2-1/function.c/#L1582
which hands out a copy of variable "errorstring".

A search for "errorstring" by
  https://codesearch.debian.net/search?q=package%3Ayabasic+errorstring
yields only one match where this variable's content is reset. That's
when it gets created at yabasic startup:
  https://sources.debian.org/src/yabasic/1:2.79.2-1/main.c/?hl=116#L117

All other occasions fill it with new error texts or put it out.
So it does not get cleared when command "open" succeeds.

See its implementation in
  https://sources.debian.org/src/yabasic/1:2.79.2-1/io.c/?hl=683#L683
The message
  "No such file or directory"
probably stems from the C function strerror(3) via my_strerror()
  https://sources.debian.org/src/yabasic/1:2.79.2-1/io.c/?hl=683#L804
  https://sources.debian.org/src/yabasic/1:2.79.2-1/main.c/?hl=2402#L2406

To fix this stickiness of error indicators, i would insert at the start of
myopen() around
  https://sources.debian.org/src/yabasic/1:2.79.2-1/io.c/?hl=683#L701
the statement

  errorstring[0] = 0;
  errorcode = 0;

... or somewhere more upwards, in the big interpreter loop ...


Hey, that's something HP BASIC never allowed us. Peeking into its code
and making perky proposals.

-

> This project started out in Tcl, which I visualize as a super-BASIC.

How that ?
Tcl is a strange script language with the only adavantage that together
with its sibling Tk it is the GUI toolkit with the least dependencies.
But as procedural language it is really substandard.
It was an exercise in stubbornness to finish
  
https://dev.lovelyhq.com/libburnia/libisoburn/raw/master/frontend/xorriso-tcltk
and i had to build a text parser into xorriso's C code because doing
proper unquoting in Tcl was flatly impossible to me.


> Due to inexperience I got lost in how Tcl handled my JSON data.

You perhaps used some of these ?
  https://wiki.tcl-lang.org/13419
Their number raises suspicion. Why so many of them ?


> Now that I have yabasic reading/writing files I've solved my program logic
> problems.

So you want a decent procedural language. That's a healthy wish.
But Tcl is really not such a language.


Have a nice day :)

Thomas

Re: painted into a corner

[Brian]

On Sun 19 Aug 2018 at 20:58:28 -0400, Gene Heskett wrote:

> On Sunday 19 August 2018 17:33:32 David Wright wrote:

[...]
 
> > Can you not run your stretch installation then? In a new system, I'd
> > mount the old wheezy disk(s) and pull the files across. Leaving the
> > wheezy mount there makes it easy to look back at the old system in
> > case you forget a file or just want to see how you used to do a
> > particular something.
> >
> That is what I'll likely do when I next boot to it, but the tools to make 
> that easy are often in the missing list. I did get synaptic to install 
> mc, but had all sorts of perms problems I didn't expect when I tried to 
> use it, due I think to the changes in what says is ext4 on both disks. 
> We will eventually get it sorted I hope.  So I'll be, without a doubt, 
> back with more problems but hopefully making progress over the next week 
> or so. Progress always puts me in a better mood than I was for the first 
> post in this thread.

My suggestion is along the same lines as David Wright's.

Boot the installer and stop when you get to partitioning.
Switch to a console (ALT-F2) and mount the wheezy and stretch
partitions. Use cp to copy files between the two partitions.

-- 
Brian.

Re: painted into a corner

[Michael Stone]

On Mon, Aug 20, 2018 at 06:53:14PM +0100, Brian wrote:

My suggestion is along the same lines as David Wright's.

Boot the installer and stop when you get to partitioning.
Switch to a console (ALT-F2) and mount the wheezy and stretch
partitions. Use cp to copy files between the two partitions.


Except that none of this should be necessary.

Re: painted into a corner

[Brian]

On Mon 20 Aug 2018 at 13:55:55 -0400, Michael Stone wrote:

> On Mon, Aug 20, 2018 at 06:53:14PM +0100, Brian wrote:
> > My suggestion is along the same lines as David Wright's.
> > 
> > Boot the installer and stop when you get to partitioning.
> > Switch to a console (ALT-F2) and mount the wheezy and stretch
> > partitions. Use cp to copy files between the two partitions.
> 
> Except that none of this should be necessary.

More than likely. But it keeps him off the streets.

-- 
Brian.

Re: hammerfall help request

[Thomas Amm]

On Thu, 16 Aug 2018 19:00:06 +
Glenn English  wrote:

> On Wed, Aug 15, 2018 at 8:40 PM, Nicholas Geovanis
>  wrote:
> 
> > I've been totally unsystematic about this problem in my machines
> > (granted with wheezy and jessie), but what I've found is that simply
> > starting the alsamixer app and re-raising the volume "fixes" the
> > issue.  
> 
> Doesn't seem to work in Colorado :-)
> 
> > Now if you don't have alsa installed or active, you
> > shouldI don't know.  
> 
> Alsa seems to be installed (alsamixer exists and sees all the sound
> hardware).
> 
> But I'm not sure it's 'active'. I'd thing the fact that alsamixer is
> live would indicate that it's active. I guess I don't know what
> 'active' means. Do you know how to find out?
> 

Hammerfalls use to be somewhat difficult with their firmware. I had to
build my own from fairly old ALSA sources. Also, for cardbus based cards
the firmware must be loaded and the card initialized at boot time. 
Have you tried using hdsploader and HDSPConf? 

-- 
If I had a letter, sealed it in a locked vault and hid the vault
somewhere in New York. Then told you to read the letter, thats not
security, thats obscurity. If I made a letter, sealed it in a vault,
gave you the blueprints of the vault, the combinations of 1000 other
vaults, access to the best lock smiths in the world, then told you to
read the letter, and you still can't, thats security.
 -- Bruce Schneier

Re: Thunderbird 60 ignores LC_TIME environment variable

[Stefan Pietsch]

On 20.08.18 11:57, Curt wrote:
> On 2018-08-20, Stefan Pietsch  wrote:
>> Dear list,
>>
>> after the upgrade from Thunderbird 52.9.1 to 60.0 (Debian sid) it seems
>> for me that the environment variable LC_TIME is ignored.
>>
>> I used the following command to set 24h time format in Thunderbird.
>>
>> $ LC_TIME=de_DE.utf8 thunderbird
>>
>> With Thunderbird 60 setting the LC_TIME or LANG variables will not help.
>> Can someone confirm this?
> 
> I can't confirm anything but this bug report
> 
> https://bugzilla.mozilla.org/show_bug.cgi?id=1426907
> 
> looks both pertinent, and, after a cursory read (recent posts re 60 toward the
> bottom), discouraging concerning your plight (at least for the moment).

Thanks for the hint to that bug id.

I had to set the "Date and Time Formatting" option (exists since TB
version 56) in the Advanced Preferences AND the variable LC_TIME when
starting Thunderbird to get it working.


Regards,
Stefan

Re: yabasic problem

[Richard Owlett]

I'll not pursue strace further at this time.
As to programming language, I have the choice of either BASIC or Tcl.
I've not used BASIC in years.
I'm learning Tcl using tcltutor30b7. I've verified my program logic with 
yabasic. Tcltutor gives handy diagnostics. I'm translating my code to 
Tcl. It is a learning experience ;/


On 08/20/2018 10:11 AM, to...@tuxteam.de wrote:

-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

On Mon, Aug 20, 2018 at 06:55:19AM -0500, Richard Owlett wrote:

[...]


However if you do:
   > strace -o mytrace yabasic test.bas
it executes, but doesn't apparently doesn't give any explanation ow
why the two files cannot be opened.


Hm. You will have realized that strace's output is extremely chatty:
you'll see the open for all possible libraries your program tries
to get hold of, then typically, after that succeeded, mmaping those
libraries into the process's address space, many failed attempts
at opening config and auxiliary files (in search of them at different
possible locations) until the program is where you consider it "up
and running".

You'll have to wade through it. Ah -- and if the program is just a
wrapper calling other processes, you'll have to tell strace to follow
them (option -f). For example (I use the option '-f', although I know
that 'cat' doesn't fork):

   strace -o /tmp/trace -f cat /no/such/file

I get the (expected) error:

   cat: /no/such/file: No such file or directory

and my /tmp/trace contains:

   4157  execve("/bin/cat", ["cat", "/no/such/file"], [/* 26 vars */]) = 0
   4157  brk(NULL) = 0x55850a279000
   4157  access("/etc/ld.so.nohwcap", F_OK) = -1 ENOENT (No such file or 
directory)
   4157  mmap(NULL, 12288, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 
0) = 0x7f7c303fc000

   # here you see the loading of the binary (execve), allocating memory (brk),
   # start of the search for the "interpreter" for the binary, etc.

   # lots of lines elided (roughly 85, in your program they'll be
   # quite a few more

   4157  fstat(1, {st_mode=S_IFCHR|0620, st_rdev=makedev(136, 2), ...}) = 0
   4157  open("/no/such/file", O_RDONLY)   = -1 ENOENT (No such file or 
directory)

   # this is the failed attempt at opening "/no/such/file"

   4157  write(2, "cat: ", 5)  = 5
   4157  write(2, "/no/such/file", 13) = 13
   4157  open("/usr/share/locale/C.UTF-8/LC_MESSAGES/libc.mo", O_RDONLY) = -1 
ENOENT (No such file or directory)
   4157  open("/usr/share/locale/C.utf8/LC_MESSAGES/libc.mo", O_RDONLY) = -1 
ENOENT (No such file or directory)
   4157  open("/usr/share/locale/C/LC_MESSAGES/libc.mo", O_RDONLY) = -1 ENOENT 
(No such file or directory)
   4157  write(2, ": No such file or directory", 27) = 27
   4157  write(2, "\n", 1) = 1

   # and this is the output of the error message to stderr (file descriptor 2)

   4157  close(1)  = 0
   4157  close(2)  = 0
   4157  exit_group(1) = ?
   4157  +++ exited with 1 +++

   # cleanup and exit.


*HOWEVER* perhaps my system and strace do not "play well together".
A tutorial I found gave as a first example:  > strace -o mytrace
yabasic ls
That gave 10 {apparently spurious} file errors before running "ls" normally.


See above: since libraries and other auxiliary files can be at one of
several places, many open errors are just the normal noise due to "looking
around".

Strace's output can be at first overwhelming, but it's a good tool
to have around.


Now to solve logic flaws in my code ;/


Ah, that's the true joy :-)

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

iEYEARECAAYFAlt62hcACgkQBcgs9XrR2kauygCfYQHujYlJno4qvZ7bqOKvSKBG
CN4An3yyNA0JFRQiyF/5uce+yIbNtuzv
=h0BJ
-END PGP SIGNATURE-

Re: Sid: NETDEV WATCHDOG: eth1 (r8169): transmit queue 0 timed out

[Gilles Mocellin]

Le lundi 20 ao?t 2018, 19:57:16 CEST Grzegorz Sójka a écrit :
> Hi there,
> 
> I have following integrated network adapter:
> 
> # lspci -s 02:00
> 02:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd.
> RTL8111/8168/8411 PCI Express Gigabit Ethernet Controller (rev 06)
> 
> By a heavy traffic i get the following error:
> NETDEV WATCHDOG: eth1 (r8169): transmit queue 0 timed out
> 
> As a workaround a turned off jumbo frames support. Is there any other
> solution making it possible to use jumbo frames??
> 
> Thanks for any help in advance.

Hello,

Same problem for me since... a long time.
I use bonding and jumbo frames.

As my chipset is a 8168, I finally use the proprietary driver, vie the r8168-
dkms package.
Sometimes, compatibility with the kernel breaks, and I have to switch back to 
in tree kernel r8169 driver, until the problem is solved.

Re: Problem Mounting New Drives

[Felipe Salvador]

On Sun, Aug 19, 2018 at 09:10:10PM +0200, john doe wrote:
> On 8/19/2018 8:59 PM, Stephen P. Molnar wrote:
> > 
> > 
> > On 08/19/2018 02:28 PM, Reco wrote:
> > > Hi.
> > > 
> > > On Sun, Aug 19, 2018 at 02:04:39PM -0400, Stephen P. Molnar wrote:
> > > 
> > > Note that second field is a mountpoint in fstab(5), as one of the stock
> > > records show:
> > > 
> > > > UUID=8f4eeaae-a055-4262-bebb-cf99abe982a5 /var    ext4 defaults
> > > > 0   2
> > > Yours have a block device name instead of mountpoint:
> > > 
> > > > UUID=900b5f0b-4f3d-4a64-8c91-29aee4c6fd07 /dev/sdb1 ext4
> > > > errors=remount-ro 0
> > > > 1
> > > > UUID=d65867da-c658-4e35-928c-9dd2d6dd5742 /dev/sdc1 ext4
> > > > errors=remount-ro 0
> > > > 1
> > > > UUID=007c1f16-34a4-438c-9d15-e3df601649ba /dev/sdc2 ext4
> > > > errors=remount-ro 0
> > > > 1
> > > So even a stock mount(1) should refuse to mount these.
> > > 
> > > 
> > > > When I rebooted the computer the OS didn't like the new
> > > > fstab and gave me a
> > > > number of, at least to me, obscure messages.
> > > systemd-mount can be cryptic, I agree.
> > > 
> > > 
> > > > Obviously, I missed something important.  My question is what?
> > > It's a kind of dumb question, but where do you need your sdb1, sdc1 and
> > > sdc2 mounted? Your fstab(5) does not mention that.
> > > 
> > > Reco
> > > 
> > > 
> > Well, that's another red faced forehead slapper!
> > 
> > It merely reinforces that I don't know what I'm doing as far as
> > hardware is concerned.
> > 
> > It's a good thing that I'm an Organic Chemist and not an IT person.
> > 
> 
> Instead of trying your new changes in '/etc/fstab' by rebooting, you
> could also try 'mount -a'.

Or # findmnt -x

> https://linux.die.net/man/8/mount
> 
> -- 
> John Doe

Regards
-- 
Felipe Salvador

Re: Sid: NFSv3 mounting problem (again!)

[deloptes]

Michael Stone wrote:

> What on earth does this mean? How did the system get into such a state
> that these were missing?

he wrote net-base was missing, but why was net-base missing?

finally I think he solved the problem

regards

Re: painted into a corner

[Gene Heskett]

On Monday 20 August 2018 11:13:08 Michael Stone wrote:

> On Mon, Aug 20, 2018 at 10:33:29AM -0400, Gene Heskett wrote:
> >sudo reboot doesn't unmount the system drive cleanly enough? Then I'd
> >call it a bug.
>
> I'm sure you would. But for all of your wall-of-text ranting, you
> failed to ever provide basic information like "what is the result of
> trying to mount the partition" so we're left guessing here.
>
> Mike Stone

If you had read that wall of text, the error was that the new version of 
ext4 supported stuff the wheezy version didn't and it recommended 
getting an updated e2fs-utils, which of course is not available in a 
wheezy repo.

So I reboot to a different installer supplied by the linuxcnc people, and 
installed it. Weird, I had access to the network, and could download 
updates, but when I looked in /dev/ no sign of either eth0 or enp0s8.  
Installed some more stuff with aptitude,logged out and rebooted to that 
install, and now my network is dead.
So, working from a terminal, I've tried to configure it manually, and 
that fails, so I am now back on wheezy, which Just Works. And I've no 
quick and dirty way to copy/paste those errors after a reboot, isolating 
that filesystem from wheezy. 

If you want to help, give me a link to a printable tut on how to make a 
working static, host based for local lookups, but uses my router, which 
in turn will forward the dns requests it cannot answer from dnsmasq, to 
real servers on the outside network, like my isp's network. On stretch.

I've read the gibberish man page for ip & friends, since ifconfig is 
gone, and apparently so it route, but gibberish  is the correct term 
when there is not a working example line in the whole man page just to 
prove it works.

So I've zero network troubleshooting tools when booted to stretch.  And 
you are giving me what for, but zero help.

So show me a tut which I can use to make it work, Mike.

-- 
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: painted into a corner

[Gene Heskett]

On Monday 20 August 2018 11:23:00 Andrew McGlashan wrote:

> On 20/08/18 05:40, Gene Heskett wrote:
> > Whats the recommended way to do these mounts so I can maintain as
> > much continuity as possible?
>
> Those other areas, are they logical volumes perhaps?  lvms.
>
No, straight partitions according to gparted.

> Cheers
> A.



-- 
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: painted into a corner

[Gene Heskett]

On Monday 20 August 2018 13:53:14 Brian wrote:

> On Sun 19 Aug 2018 at 20:58:28 -0400, Gene Heskett wrote:
> > On Sunday 19 August 2018 17:33:32 David Wright wrote:
>
> [...]
>
> > > Can you not run your stretch installation then? In a new system,
> > > I'd mount the old wheezy disk(s) and pull the files across.
> > > Leaving the wheezy mount there makes it easy to look back at the
> > > old system in case you forget a file or just want to see how you
> > > used to do a particular something.
> >
> > That is what I'll likely do when I next boot to it, but the tools to
> > make that easy are often in the missing list. I did get synaptic to
> > install mc, but had all sorts of perms problems I didn't expect when
> > I tried to use it, due I think to the changes in what says is ext4
> > on both disks. We will eventually get it sorted I hope.  So I'll be,
> > without a doubt, back with more problems but hopefully making
> > progress over the next week or so. Progress always puts me in a
> > better mood than I was for the first post in this thread.
>
> My suggestion is along the same lines as David Wright's.
>
> Boot the installer and stop when you get to partitioning.
> Switch to a console (ALT-F2) and mount the wheezy and stretch
> partitions. Use cp to copy files between the two partitions.

If push comes to shove, I'll see if a wheezy disk can be mounted to a 
stretch mount point. Someplace in the last 2 days ISTR trying that after 
booting stretch, and mount worked but navigating in the disk failed, the 
root was all I could see, and when I rebooted to wheezy, I had a 15+ 
minute pause while it did an e2fsck to most of that 1 terabyte disk, 
without reporting any errors found that it told me about on the boot 
screen. That would spook most anybody I think.

But now, for no known reason, stretch's network has gone away.  And all 
the tools I am familiar with have been excised from stretch.  Sigh...

Thanks Brian.

-- 
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: painted into a corner

[Michael Stone]

On Mon, Aug 20, 2018 at 04:27:18PM -0400, Gene Heskett wrote:

If you had read that wall of text, the error was that the new version of
ext4 supported stuff the wheezy version didn't and it recommended
getting an updated e2fs-utils, which of course is not available in a
wheezy repo.


That was the output of fsck, not the output of mount. But at least 
you've provided another wall of text.

Re: Sid: NFSv3 mounting problem (again!)

[Grzegorz Sójka]

On 08/20/18 22:23, deloptes wrote:

Michael Stone wrote:


What on earth does this mean? How did the system get into such a state
that these were missing?


he wrote net-base was missing, but why was net-base missing?
As I wrote in one of my posts, the system was installed using 
debootstrap  --variant=minbase. So this is really minimal set of 
packages. Moreover, the netbase package is not in dependencies of the 
nfs-common.



finally I think he solved the problem
Yes, the problem is solved and it looks that nlockmgr is needed on both 
server and client sides.


--
Pozdrawiam
Grzesiek

Wysłane z kompa wolnego od wirusów Billa Gatesa.

Sid: please fix dependences of nfs-common

[Grzegorz Sójka]

Hi there,

Please add netbase to the deps of the nfs-common package. Without it 
mounting is impossible, more details may be found in the "Sid: NFSv3 
mounting problem" thread.


--
Pozdrawiam
Grzesiek

Wysłane z kompa wolnego od wirusów Billa Gatesa.

Re: Sid: please fix dependences of nfs-common

[Greg Wooledge]

On Mon, Aug 20, 2018 at 11:07:13PM +0200, Grzegorz Sójka wrote:
> Hi there,
> 
> Please add netbase to the deps of the nfs-common package. Without it
> mounting is impossible, more details may be found in the "Sid: NFSv3
> mounting problem" thread.

http://bugs.debian.org/

A sid user should know this stuff.

Re: Sid: please fix dependences of nfs-common

[Grzegorz Sójka]

On 08/20/18 23:12, Greg Wooledge wrote:

On Mon, Aug 20, 2018 at 11:07:13PM +0200, Grzegorz Sójka wrote:

Hi there,

Please add netbase to the deps of the nfs-common package. Without it
mounting is impossible, more details may be found in the "Sid: NFSv3
mounting problem" thread.


http://bugs.debian.org/


Thanks for the tip.


A sid user should know this stuff

I'm sid user since ... yesterday ;)

--
Pozdrawiam
Grzesiek

Wysłane z kompa wolnego od wirusów Billa Gatesa.

Re: Sid: please fix dependences of nfs-common

[Michael Stone]

(cc'ing package maintainers for insight)

On Mon, Aug 20, 2018 at 11:07:13PM +0200, Grzegorz Sójka wrote:
Please add netbase to the deps of the nfs-common package. Without it 
mounting is impossible, more details may be found in the "Sid: NFSv3 
mounting problem" thread.


This should probably actually be in the rpcbind package rather than 
nfs-common. Once upon a time, when rpcbind was still portmap, it was 
actually part of netbase so a dependency was unnecessary. But rpcinfo is 
expected to convert rpc numbers to names, and it can't generally do that 
without /etc/rpc. nfs-common may *also* want a dependency, since neither 
the kernel client nor the kernel server work properly without it, and in 
fact older versions *did* depend on netbase. There's a note in the 
changelog about removing obsolete versioned dependencies, but that 
probably should have been a conversion to a non-versioned depenedency 
instead of outright removal. But, that might be unnecessary if rpcbind 
adds the appropriate dependency.


That said, you may be the first person to not install using the 
installer, and install nfs-common, and not install nfs-kernel-server, 
and not not install any of the other packages which depend on netbase. 
You just got lucky. :)


Mike Stone

Re: "accept_ra 1" vs "accept_ra 2" in interfaces configuration-file

[Andy Smith]

Hi Martin,

On Mon, Aug 20, 2018 at 03:58:36PM +0300, Martin T wrote:
> On Mon, Aug 20, 2018 at 7:55 AM Andy Smith  wrote:
> > Back in 2011 this was a hard-won battle:
> >
> > 
> > http://strugglers.net/~andy/blog/2011/09/04/linux-ipv6-router-advertisements-and-forwarding/

[…]

> Thanks for this very informative blog post! However, setting the
> "net.ipv6.conf.all.forwarding" to 1 in /etc/sysctl.conf and
> "accept_ra" to 2 in /etc/network/interfaces for ISP facing
> interface(eth0) didn't work for me. I expected SLAAC to work, but it
> didn't. I'm running kernel version 4.9.0.

Strange. I've had a look and it seems I continued to use the
workaround mentioned in the blog post even though supposedly I no
longer need to. If you use that workaround, does it (SLAAC) start
working for you?

Also, is it just address assignment that doesn't work or is it also
default router assignment that doesn't work? On my servers that
forward v6 I don't use dynamic assignment of addresses, I statically
assign them, but I do use dynamic assignment of default route.

Cheers,
Andy

Re: painted into a corner

[Zenaan Harkness]

On Mon, Aug 20, 2018 at 04:42:00PM -0400, Gene Heskett wrote:
> On Monday 20 August 2018 13:53:14 Brian wrote:
> > On Sun 19 Aug 2018 at 20:58:28 -0400, Gene Heskett wrote:
> > > On Sunday 19 August 2018 17:33:32 David Wright wrote:
> > > > Can you not run your stretch installation then? In a new system,
> > > > I'd mount the old wheezy disk(s) and pull the files across.
> > > > Leaving the wheezy mount there makes it easy to look back at the
> > > > old system in case you forget a file or just want to see how you
> > > > used to do a particular something.
> > >
> > > That is what I'll likely do when I next boot to it, but the tools to
> > > make that easy are often in the missing list. I did get synaptic to
> > > install mc, but had all sorts of perms problems I didn't expect when
> > > I tried to use it, due I think to the changes in what says is ext4
> > > on both disks. We will eventually get it sorted I hope.  So I'll be,
> > > without a doubt, back with more problems but hopefully making
> > > progress over the next week or so. Progress always puts me in a
> > > better mood than I was for the first post in this thread.
> >
> > My suggestion is along the same lines as David Wright's.
> >
> > Boot the installer and stop when you get to partitioning.
> > Switch to a console (ALT-F2) and mount the wheezy and stretch
> > partitions. Use cp to copy files between the two partitions.
> 
> If push comes to shove, I'll see if a wheezy disk can be mounted to a 
> stretch mount point.

That's what I'd do - just boot stretch and mount your wheezy
partition(s).  Newer OSes are designed in general to be backwards
compatible - people scream at the sky when a previous disk cannot be
mounted.


> Someplace in the last 2 days ISTR trying that after 
> booting stretch, and mount worked but navigating in the disk failed,

That was probably "navigating to my old /home/ folder" perhaps?

You need to mount your original partition for /home I suspect, not
just the root partition, that's all...


> the root was all I could see, and when I rebooted to wheezy, I had
> a 15+ minute pause while it did an e2fsck to most of that 1
> terabyte disk, without reporting any errors found that it told me
> about on the boot screen. That would spook most anybody I think.

It's possible you shut down too quickly?  Or could the drive have
been disconnected before unmounting the old wheezy /root ?


> But now, for no known reason, stretch's network has gone away.  And all 
> the tools I am familiar with have been excised from stretch.  Sigh...

Hm, frustrating.

Networking should work roughly the same in both wheezy and stretch.

If you want it auto shiny, just apt install network-manager

Good luck,

Re: painted into a corner

[Zenaan Harkness]

On Mon, Aug 20, 2018 at 05:00:10PM -0400, Michael Stone wrote:
> On Mon, Aug 20, 2018 at 04:27:18PM -0400, Gene Heskett wrote:
> > If you had read that wall of text, the error was that the new version of
> > ext4 supported stuff the wheezy version didn't and it recommended
> > getting an updated e2fs-utils, which of course is not available in a
> > wheezy repo.
> 
> That was the output of fsck, not the output of mount. But at least you've
> provided another wall of text.

I thought I was the only one allowed to be snarky at others?

Feeling cheated of my snark right, time for cheesecake ;(

Re: painted into a corner

[David Wright]

On Mon 20 Aug 2018 at 16:27:18 (-0400), Gene Heskett wrote:
> On Monday 20 August 2018 11:13:08 Michael Stone wrote:
> > On Mon, Aug 20, 2018 at 10:33:29AM -0400, Gene Heskett wrote:
> > >sudo reboot doesn't unmount the system drive cleanly enough? Then I'd
> > >call it a bug.
> >
> > I'm sure you would. But for all of your wall-of-text ranting, you
> > failed to ever provide basic information like "what is the result of
> > trying to mount the partition" so we're left guessing here.
> 
> If you had read that wall of text, the error was that the new version of 
> ext4 supported stuff the wheezy version didn't and it recommended 
> getting an updated e2fs-utils, which of course is not available in a 
> wheezy repo.

Yes, I still don't understand why you install stretch but then run
wheezy. It's normal for modern systems to understand older formats,
but old systems' authors can't foresee future features. Say that fast.

> So I reboot to a different installer supplied by the linuxcnc people, and 
> installed it. Weird, I had access to the network, and could download 
> updates, but when I looked in /dev/ no sign of either eth0 or enp0s8.  

Eh? in /dev?

> Installed some more stuff with aptitude,logged out and rebooted to that 
> install, and now my network is dead.
> So, working from a terminal, I've tried to configure it manually, and 
> that fails, so I am now back on wheezy, which Just Works. And I've no 
> quick and dirty way to copy/paste those errors after a reboot, isolating 
> that filesystem from wheezy. 
> 
> If you want to help, give me a link to a printable tut on how to make a 
> working static, host based for local lookups, but uses my router, which 
> in turn will forward the dns requests it cannot answer from dnsmasq, to 
> real servers on the outside network, like my isp's network. On stretch.
> 
> I've read the gibberish man page for ip & friends, since ifconfig is 
> gone, and apparently so it route, but gibberish  is the correct term 
> when there is not a working example line in the whole man page just to 
> prove it works.

Eh?

$ cat /etc/debian_version 
9.5
$ /sbin/ifconfig 
enp1s0: flags=4099  mtu 1500
ether 08:9e:01:c8:67:7e  txqueuelen 1000  (Ethernet)
RX packets 0  bytes 0 (0.0 B)
RX errors 0  dropped 0  overruns 0  frame 0
TX packets 0  bytes 0 (0.0 B)
TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

lo: flags=73  mtu 65536
inet 127.0.0.1  netmask 255.0.0.0
inet6 ::1  prefixlen 128  scopeid 0x10
loop  txqueuelen 1  (Local Loopback)
RX packets 594  bytes 42487 (41.4 KiB)
RX errors 0  dropped 0  overruns 0  frame 0
TX packets 594  bytes 42487 (41.4 KiB)
TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

wlp2s0: flags=4163  mtu 1500
inet 192.168.1.17  netmask 255.255.255.0  broadcast 192.168.1.255
inet6 fe80::e8b:fdff:fe0b:67fb  prefixlen 64  scopeid 0x20
ether 0c:8b:fd:0b:67:fb  txqueuelen 1000  (Ethernet)
RX packets 1291291  bytes 238957455 (227.8 MiB)
RX errors 0  dropped 0  overruns 0  frame 0
TX packets 157560  bytes 65731010 (62.6 MiB)
TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0

$ /sbin/route 
Kernel IP routing table
Destination Gateway Genmask Flags Metric Ref   Use Iface
default router  0.0.0.0 UG0  0   0 wlp2s0
192.168.1.0 0.0.0.0 255.255.255.0   U 0  0   0 wlp2s0
$ 

> So I've zero network troubleshooting tools when booted to stretch.  And 
> you are giving me what for, but zero help.
> 
> So show me a tut which I can use to make it work, Mike.

O'Really? Gene, meet google. Google, meet Gene. There are whole tomes
of this stuff which are downloadable in slightly frayed editions.
I've posted references to one or two, generally under Owlett threads.

Cheers,
David.

KDE Bluedevil

[Timothy M Butterworth]

In Debian 9.5 KDE Bluedevil is unable to add devices. Is their any fix
for this other then using GNOME Blueman tool.

thanks

Tim

Re: painted into a corner

[David Wright]

On Tue 21 Aug 2018 at 09:48:37 (+1000), Zenaan Harkness wrote:
> On Mon, Aug 20, 2018 at 04:42:00PM -0400, Gene Heskett wrote:

> > But now, for no known reason, stretch's network has gone away.  And all 
> > the tools I am familiar with have been excised from stretch.  Sigh...
> 
> Hm, frustrating.
> 
> Networking should work roughly the same in both wheezy and stretch.

Except you need to install net-tools (and perhaps add net.ifnames=0
lest one comes face to face with monstrosities like enp1s0 :) )

> If you want it auto shiny, just apt install network-manager

Translation for Gene: that's ¹network-mangler :)

¹Disclaimer: I have no opinion as I don't install it.

Cheers,
David.

Re: painted into a corner

[Zenaan Harkness]

On Mon, Aug 20, 2018 at 07:19:26PM -0500, David Wright wrote:
> On Tue 21 Aug 2018 at 09:48:37 (+1000), Zenaan Harkness wrote:
> > On Mon, Aug 20, 2018 at 04:42:00PM -0400, Gene Heskett wrote:
> 
> > > But now, for no known reason, stretch's network has gone away.  And all 
> > > the tools I am familiar with have been excised from stretch.  Sigh...
> > 
> > Hm, frustrating.
> > 
> > Networking should work roughly the same in both wheezy and stretch.
> 
> Except you need to install net-tools (and perhaps add net.ifnames=0
> lest one comes face to face with monstrosities like enp1s0 :) )

Oh come on‼ Does anyone -really- think that enp1s0 is not the most
intuitive name, like ever like?  (Did I mention "like"?)

;)


> > If you want it auto shiny, just apt install network-manager
> 
> Translation for Gene: that's ¹network-mangler :)
> 
> ¹Disclaimer: I have no opinion as I don't install it.

Neither do I, ackshualley - except on computers I install for others,
since it generally works well enough that they don't thereafter need
to phone me in the thought that I am their personal help desk for
ever after. Especially in the dynamic laptop situation.

Re: painted into a corner

[Jimmy Johnson]

On 08/20/2018 01:28 PM, Gene Heskett wrote:

On Monday 20 August 2018 11:23:00 Andrew McGlashan wrote:


On 20/08/18 05:40, Gene Heskett wrote:

Whats the recommended way to do these mounts so I can maintain as
much continuity as possible?


Those other areas, are they logical volumes perhaps?  lvms.


No, straight partitions according to gparted.


Cheers
A.



Hi Gene, I've seen this before, a few times. If you run #fdisk -l while 
in stretch and get error, you need to fix that first, using gparted move 
the ailing partition '1'byte, just one digit larger or smaller, if you 
have to shrink another partition, do it '2'bytes. UUID will not change 
and it will pass fdisk -l no error.  Now in wheezy see if you can mount. 
 If not you can disable the checksums. To disable checksums on an 
existing filesystem, ensure that the filesystem will pass fsck. Then 
turn off metadata_csum via tune2fs.

 #tune2fs -O ^metadata_csum /dev/path/to/disk

The above is for ext4, for repairing partition I use this with force.
 #fsck.ext4 -pvcf /dev/sdxx
--
Jimmy Johnson

Devuan Jessie - KDE 4.14.2 - AMD A8-7600 - EXT4 at sda2
Registered Linux User #380263

Re: painted into a corner

[Gene Heskett]

On Monday 20 August 2018 19:51:36 Zenaan Harkness wrote:

> On Mon, Aug 20, 2018 at 05:00:10PM -0400, Michael Stone wrote:
> > On Mon, Aug 20, 2018 at 04:27:18PM -0400, Gene Heskett wrote:
> > > If you had read that wall of text, the error was that the new
> > > version of ext4 supported stuff the wheezy version didn't and it
> > > recommended getting an updated e2fs-utils, which of course is not
> > > available in a wheezy repo.
> >
> > That was the output of fsck, not the output of mount. But at least
> > you've provided another wall of text.
>
> I thought I was the only one allowed to be snarky at others?
>
> Feeling cheated of my snark right, time for cheesecake ;(

Ahh, but that would peak my sugar way up into the ketones range, not 
good.  Tain't fun, being an 83 yo diabetic.  Everything I like is either 
illegal, immoral, or fattening.

-- 
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: painted into a corner

[Gene Heskett]

On Monday 20 August 2018 20:08:11 David Wright wrote:

> On Mon 20 Aug 2018 at 16:27:18 (-0400), Gene Heskett wrote:
> > On Monday 20 August 2018 11:13:08 Michael Stone wrote:
> > > On Mon, Aug 20, 2018 at 10:33:29AM -0400, Gene Heskett wrote:
> > > >sudo reboot doesn't unmount the system drive cleanly enough? Then
> > > > I'd call it a bug.
> > >
> > > I'm sure you would. But for all of your wall-of-text ranting, you
> > > failed to ever provide basic information like "what is the result
> > > of trying to mount the partition" so we're left guessing here.
> >
> > If you had read that wall of text, the error was that the new
> > version of ext4 supported stuff the wheezy version didn't and it
> > recommended getting an updated e2fs-utils, which of course is not
> > available in a wheezy repo.
>
> Yes, I still don't understand why you install stretch but then run
> wheezy. It's normal for modern systems to understand older formats,
> but old systems' authors can't foresee future features. Say that fast.
>
> > So I reboot to a different installer supplied by the linuxcnc
> > people, and installed it. Weird, I had access to the network, and
> > could download updates, but when I looked in /dev/ no sign of either
> > eth0 or enp0s8.
>
> Eh? in /dev?
>
Absolutely no trace of either an eth# or an enp0s# there. At that point I 
looked to see if it was 5 o-clock yet.  Wasn't, dammit.

> > Installed some more stuff with aptitude,logged out and rebooted to
> > that install, and now my network is dead.
> > So, working from a terminal, I've tried to configure it manually,
> > and that fails, so I am now back on wheezy, which Just Works. And
> > I've no quick and dirty way to copy/paste those errors after a
> > reboot, isolating that filesystem from wheezy.
> >
> > If you want to help, give me a link to a printable tut on how to
> > make a working static, host based for local lookups, but uses my
> > router, which in turn will forward the dns requests it cannot answer
> > from dnsmasq, to real servers on the outside network, like my isp's
> > network. On stretch.
> >
> > I've read the gibberish man page for ip & friends, since ifconfig is
> > gone, and apparently so it route, but gibberish  is the correct term
> > when there is not a working example line in the whole man page just
> > to prove it works.
>
> Eh?
>
> $ cat /etc/debian_version
> 9.5
> $ /sbin/ifconfig

not found

> enp1s0: flags=4099  mtu 1500
> ether 08:9e:01:c8:67:7e  txqueuelen 1000  (Ethernet)
> RX packets 0  bytes 0 (0.0 B)
> RX errors 0  dropped 0  overruns 0  frame 0
> TX packets 0  bytes 0 (0.0 B)
> TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0
>
> lo: flags=73  mtu 65536
> inet 127.0.0.1  netmask 255.0.0.0
> inet6 ::1  prefixlen 128  scopeid 0x10
> loop  txqueuelen 1  (Local Loopback)
> RX packets 594  bytes 42487 (41.4 KiB)
> RX errors 0  dropped 0  overruns 0  frame 0
> TX packets 594  bytes 42487 (41.4 KiB)
> TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0
>
> wlp2s0: flags=4163  mtu 1500
> inet 192.168.1.17  netmask 255.255.255.0  broadcast
> 192.168.1.255 inet6 fe80::e8b:fdff:fe0b:67fb  prefixlen 64  scopeid
> 0x20 ether 0c:8b:fd:0b:67:fb  txqueuelen 1000  (Ethernet)
> RX packets 1291291  bytes 238957455 (227.8 MiB)
> RX errors 0  dropped 0  overruns 0  frame 0
> TX packets 157560  bytes 65731010 (62.6 MiB)
> TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0
>
> $ /sbin/route

not found, even with a sudo preface for both of them.

> Kernel IP routing table
> Destination Gateway Genmask Flags Metric Ref   Use
> Iface default router  0.0.0.0 UG0  0  
> 0 wlp2s0 192.168.1.0 0.0.0.0 255.255.255.0   U 0  
>0   0 wlp2s0 $
>
> > So I've zero network troubleshooting tools when booted to stretch. 
> > And you are giving me what for, but zero help.
> >
> > So show me a tut which I can use to make it work, Mike.
>
> O'Really? Gene, meet google. Google, meet Gene. There are whole tomes
> of this stuff which are downloadable in slightly frayed editions.
> I've posted references to one or two, generally under Owlett threads.

I'll have to go back and look, might even have one or two marked 
important.

> Cheers,

Take care David.

-- 
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: painted into a corner

[Zenaan Harkness]

On Mon, Aug 20, 2018 at 09:16:22PM -0400, Gene Heskett wrote:
> On Monday 20 August 2018 19:51:36 Zenaan Harkness wrote:
> 
> > On Mon, Aug 20, 2018 at 05:00:10PM -0400, Michael Stone wrote:
> > > On Mon, Aug 20, 2018 at 04:27:18PM -0400, Gene Heskett wrote:
> > > > If you had read that wall of text, the error was that the new
> > > > version of ext4 supported stuff the wheezy version didn't and it
> > > > recommended getting an updated e2fs-utils, which of course is not
> > > > available in a wheezy repo.
> > >
> > > That was the output of fsck, not the output of mount. But at least
> > > you've provided another wall of text.
> >
> > I thought I was the only one allowed to be snarky at others?
> >
> > Feeling cheated of my snark right, time for cheesecake ;(
> 
> Ahh, but that would peak my sugar way up into the ketones range, not 
> good.  Tain't fun, being an 83 yo diabetic.  Everything I like is either 
> illegal, immoral, or fattening.

Ahah! That's where home cooking comes in of course - check the
glycaemic index on various alternative sweeteners (e.g. stevia), and
experiment with some combination, to keep the total for a medium
sized cheesecake below your daily "glycogen" intake requirement -
sour cream, some matured/Euro-style cream cheese, perhaps buttermilk,
touch of fresh lemon juice and BAM, no more than a day's glyco intake
in a full cheesecake!

And if you go full keto ("ketogenic") for a couple months, you can
eat these every day AND lose weight - seriously the ultimate
indulgence diet - it doesn't get better than this.

Re: painted into a corner

[Timothy M Butterworth]

I like network manager when I'm out and about with my notebook it
works great for connecting to wireless hotspots. It does what it was
designed for.

The new udev naming convention does look strange but at least udev no
longer randomly flip flops the nics on my UTM putting the inside
ruleset on the outside interface and outside ruleset on the inside
interface. It would have been better if they could have kept the eth0
style convention and used a file to keep persistence in the naming and
activating.



On 8/20/18, Gene Heskett  wrote:
> On Monday 20 August 2018 20:08:11 David Wright wrote:
>
>> On Mon 20 Aug 2018 at 16:27:18 (-0400), Gene Heskett wrote:
>> > On Monday 20 August 2018 11:13:08 Michael Stone wrote:
>> > > On Mon, Aug 20, 2018 at 10:33:29AM -0400, Gene Heskett wrote:
>> > > >sudo reboot doesn't unmount the system drive cleanly enough? Then
>> > > > I'd call it a bug.
>> > >
>> > > I'm sure you would. But for all of your wall-of-text ranting, you
>> > > failed to ever provide basic information like "what is the result
>> > > of trying to mount the partition" so we're left guessing here.
>> >
>> > If you had read that wall of text, the error was that the new
>> > version of ext4 supported stuff the wheezy version didn't and it
>> > recommended getting an updated e2fs-utils, which of course is not
>> > available in a wheezy repo.
>>
>> Yes, I still don't understand why you install stretch but then run
>> wheezy. It's normal for modern systems to understand older formats,
>> but old systems' authors can't foresee future features. Say that fast.
>>
>> > So I reboot to a different installer supplied by the linuxcnc
>> > people, and installed it. Weird, I had access to the network, and
>> > could download updates, but when I looked in /dev/ no sign of either
>> > eth0 or enp0s8.
>>
>> Eh? in /dev?
>>
> Absolutely no trace of either an eth# or an enp0s# there. At that point I
> looked to see if it was 5 o-clock yet.  Wasn't, dammit.
>
>> > Installed some more stuff with aptitude,logged out and rebooted to
>> > that install, and now my network is dead.
>> > So, working from a terminal, I've tried to configure it manually,
>> > and that fails, so I am now back on wheezy, which Just Works. And
>> > I've no quick and dirty way to copy/paste those errors after a
>> > reboot, isolating that filesystem from wheezy.
>> >
>> > If you want to help, give me a link to a printable tut on how to
>> > make a working static, host based for local lookups, but uses my
>> > router, which in turn will forward the dns requests it cannot answer
>> > from dnsmasq, to real servers on the outside network, like my isp's
>> > network. On stretch.
>> >
>> > I've read the gibberish man page for ip & friends, since ifconfig is
>> > gone, and apparently so it route, but gibberish  is the correct term
>> > when there is not a working example line in the whole man page just
>> > to prove it works.
>>
>> Eh?
>>
>> $ cat /etc/debian_version
>> 9.5
>> $ /sbin/ifconfig
>
> not found
>
>> enp1s0: flags=4099  mtu 1500
>> ether 08:9e:01:c8:67:7e  txqueuelen 1000  (Ethernet)
>> RX packets 0  bytes 0 (0.0 B)
>> RX errors 0  dropped 0  overruns 0  frame 0
>> TX packets 0  bytes 0 (0.0 B)
>> TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0
>>
>> lo: flags=73  mtu 65536
>> inet 127.0.0.1  netmask 255.0.0.0
>> inet6 ::1  prefixlen 128  scopeid 0x10
>> loop  txqueuelen 1  (Local Loopback)
>> RX packets 594  bytes 42487 (41.4 KiB)
>> RX errors 0  dropped 0  overruns 0  frame 0
>> TX packets 594  bytes 42487 (41.4 KiB)
>> TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0
>>
>> wlp2s0: flags=4163  mtu 1500
>> inet 192.168.1.17  netmask 255.255.255.0  broadcast
>> 192.168.1.255 inet6 fe80::e8b:fdff:fe0b:67fb  prefixlen 64  scopeid
>> 0x20 ether 0c:8b:fd:0b:67:fb  txqueuelen 1000  (Ethernet)
>> RX packets 1291291  bytes 238957455 (227.8 MiB)
>> RX errors 0  dropped 0  overruns 0  frame 0
>> TX packets 157560  bytes 65731010 (62.6 MiB)
>> TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0
>>
>> $ /sbin/route
>
> not found, even with a sudo preface for both of them.
>
>> Kernel IP routing table
>> Destination Gateway Genmask Flags Metric Ref   Use
>> Iface default router  0.0.0.0 UG0  0
>> 0 wlp2s0 192.168.1.0 0.0.0.0 255.255.255.0   U 0
>>0   0 wlp2s0 $
>>
>> > So I've zero network troubleshooting tools when booted to stretch.
>> > And you are giving me what for, but zero help.
>> >
>> > So show me a tut which I can use to make it work, Mike.
>>
>> O'Really? Gene, meet google. Google, meet Gene. There are whole tomes
>> of this stuff which are downloadable in slightly frayed editions.
>> I've posted references to one or two, generally under Owlett threads.
>
> I'll have to go back and look, might even have one or t

What's the right way to get an unbroken mc package installed in Stretch

[Felix Miata]

Stretch comes with 4.8.18, which for me is unusable due to an upstream bug fixed
in 4.8.19[1], but 4.8.19 is unusable due to another upstream bug, not fixed
until 4.8.20[2]. I configured stretch-backports in sources.list, but apparently
it has no newer mc version available. 4.8.21 is on the mirrors, so I downloaded
it from a mirror and installed it with dpkg. It works fine, but now apt is
broken because it thinks mc is broken and needs to be removed by "apt
--fix-broken install". Apparently apt thinks mc depends on different packages
than dpkg does:

# apt install inxi
...
The following packages have unmet dependencies:
 inxi : Depends: gawk but it is not going to be installed
 mc : Depends: libext2fs2 (>= 1.37)
  Depends: libgpm2 (>= 1.20.7) but 1.20.4-6.2+b1 is to be installed
E: Unmet dependencies. Try 'apt --fix-broken install'

# dpkg-reconfigure mc
/usr/sbin/dpkg-reconfigure: mc is broken or not fully installed

Am I the only one who needs mc to work in Stretch? What do others do?

[1] fubar right pane at startup with saved long file listing
http://midnight-commander.org/ticket/3758

[2] VFS: timestamps not being preserved due to uninitialized struct stat
st_[acm]tim.tv_nsec (#3821)
http://midnight-commander.org/ticket/3821
-- 
"Wisdom is supreme; therefore get wisdom. Whatever else you
get, get wisdom." Proverbs 4:7 (New Living Translation)

 Team OS/2 ** Reg. Linux User #211409 ** a11y rocks!

Felix Miata  ***  http://fm.no-ip.com/

Re: Deep Packet Inspection

[Mimiko]

Thank you all for suggestions.

Yes, I didn't tell my goal. First of course is to limit access to web sites and collect statistics. Yes this could be done with squid and ssl_bump. I 
hope this does not change certificate as internet-banking will not work. The problem for a quick implementation is with need of squid recompile to 
support ssl.


The second goal is intercept packets on other ports for limiting services, like 
skype, teamviewer (especially).

For now I use iptables -m string --algo kmp --to 65535 --string to intercept some strings on conenction and block access to some sites by domain name. 
But this will not allow me to block access to all sites and allow access to only several sites.


I was looking for a quick implementation.

l7filter was interesting for me, but it is not supported anymore. nDPI scares 
me with patching kernel. And OpenDPI is not in repository.

I will try to implement OpenDPI by compiling, also as squid, but this is a long 
process.

As I read for snort, suricata, zorp - it is a self contained firewall. I use a 
standard Debian installation where I run several different services.

Thanks again.

Re: Deep Packet Inspection

[Eero Volotinen]

Well. You can't really open "decipher" ssl without changing certificate,
but you can exclude some sites from ssl bumping.

Eero

On Tue, Aug 21, 2018 at 9:31 AM Mimiko  wrote:

> Thank you all for suggestions.
>
> Yes, I didn't tell my goal. First of course is to limit access to web
> sites and collect statistics. Yes this could be done with squid and
> ssl_bump. I
> hope this does not change certificate as internet-banking will not work.
> The problem for a quick implementation is with need of squid recompile to
> support ssl.
>
> The second goal is intercept packets on other ports for limiting services,
> like skype, teamviewer (especially).
>
> For now I use iptables -m string --algo kmp --to 65535 --string to
> intercept some strings on conenction and block access to some sites by
> domain name.
> But this will not allow me to block access to all sites and allow access
> to only several sites.
>
> I was looking for a quick implementation.
>
> l7filter was interesting for me, but it is not supported anymore. nDPI
> scares me with patching kernel. And OpenDPI is not in repository.
>
> I will try to implement OpenDPI by compiling, also as squid, but this is a
> long process.
>
> As I read for snort, suricata, zorp - it is a self contained firewall. I
> use a standard Debian installation where I run several different services.
>
> Thanks again.
>
>