from:"Piyavkin"


On 06.05.2016 18:36, Pierre Frenkiel wrote:

On Fri, 6 May 2016, Piyavkin wrote:

Well, then Plan B. I don't know which way you use to share your 
printer+scanner for your network. But if you use CUPS, then the pages 
may be interesting for you:


http://xmodulo.com/usb-network-printer-and-scanner-server-debian.html
http://xmodulo.com/configure-network-printer-scanner-ubuntu-desktop.html

You may check if the saned is running both on the server and on the 
client side.
And check: a) access rules in cupsd.conf (server side), b) saned 
hosts in /etc/sane.d/net.conf (client side).


  the printer is directly connected to ethernet, so there is no server 
side.

  All my computers are just clients...

  Anyway, I was right when I said that the problem was with HP software:
  I found that the hplip version provided by Debian/Jessie was 
outdated(3.14...)

  After installing the last version, from http://hplipopensource.com,
  (version 3.16.5), everything worked perfectly.

best regards,


Or because you've just reinstalled it, since everything was fine on 
other machine with the same system, and on the troubled machine before 
the issue had been started.


Cool device you have, by the way. It even can do direct wireless 
printing from mobile devices.
I wonder, how in the configuration (direct connection to network) access 
privileges should be managed: on the router, by settings in the printer, 
or some other way?



Best regards,
Dmitry Piyavkin

Re: network scanner not detected


On 06.05.2016 16:53, Pierre Frenkiel wrote:

On Fri, 6 May 2016, Piyavkin wrote:


Hi, Pierre,


I'm not sure if it is applicable in your case, but may be some 
firmware required or some config-file with proper way to it was 
recently rewritten/moved.
I have an old Epson Perfection 2480 Photo scanner and my usual way to 
install/reinstall it is like such:

. . .

   hi Dmitry
   Thank you for your detailed reply, but in my case I'm quite sure that
   the problem is no with xsane, but at the system level, for the
   following reasons:

   1/ The other scanning programs don't find the scanner either
  (scanimage, simple-scan, hp-scan)

   2/ on my desktop, where the scanner works,  I don't have a snapscan 
directory


   3/ as I said, the scanner is seen and works via the web interface

   4/ I just discovered a more worrying thing:
  the hp utility hp-probe doesn't find any device (printer or 
scanner)

  On my desktop, "hp-probe -bnet" gives:

  hp:/net/Officejet_Pro_6830?ip=192.168.1.9 Officejet_Pro_6830  
HPA48092


  and on the laptop:

  warning: No devices found on the 'net' bus

  As system-config-printer finds the device, I think that the problem
  is related to HP software, but I'm unable to find how...

best regards,


Well, then Plan B. I don't know which way you use to share your 
printer+scanner for your network. But if you use CUPS, then the pages 
may be interesting for you:


http://xmodulo.com/usb-network-printer-and-scanner-server-debian.html
http://xmodulo.com/configure-network-printer-scanner-ubuntu-desktop.html

You may check if the saned is running both on the server and on the 
client side.
And check: a) access rules in cupsd.conf (server side), b) saned hosts 
in /etc/sane.d/net.conf (client side).



Best regards,
Dmitry Piyavkin

Re: network scanner not detected


On 06.05.2016 13:45, Pierre Frenkiel wrote:

hi,
I have an HP Officejet Pro 6830, connected to my local ethernet network,
and the scanner works perfectly on my Jessie desktop. On my Jessie 
(amd64)
laptop, it also used to work, but now it no more detected(printing 
still works)


I tried
   scanimage -L
   xsane
   hp-check
and all fail to find it.
Curiously, it is still seen, and works, when using the Web interface.
I didn't learn anything useful with Google or other search engines.
Has anybody an idea?

best regards,


Also, my it's just access rights issue?

Best regards,
Dmitry Piyavkin

Re: network scanner not detected


On 06.05.2016 13:45, Pierre Frenkiel wrote:

hi,
I have an HP Officejet Pro 6830, connected to my local ethernet network,
and the scanner works perfectly on my Jessie desktop. On my Jessie 
(amd64)
laptop, it also used to work, but now it no more detected(printing 
still works)


I tried
   scanimage -L
   xsane
   hp-check
and all fail to find it.
Curiously, it is still seen, and works, when using the Web interface.
I didn't learn anything useful with Google or other search engines.
Has anybody an idea?

best regards,


Hi, Pierre,


I'm not sure if it is applicable in your case, but may be some firmware 
required or some config-file with proper way to it was recently 
rewritten/moved.


I have an old Epson Perfection 2480 Photo scanner and my usual way to 
install/reinstall it is like such:



1. Installing:

xsane, libsane, libsane-extras (I'm not sure if the last one even needed)


2. Placing firmware file esfw41.bin in the folder:

/usr/share/sane/snapscan

If there is no snapscan directory in /usr/share/sane it should be created.

The firmware file comes from the same directory in the old installation 
(or backup) or from the scanner's Soft CD.


For other scanners there are sites with newer/valid firmware to download.


3. Correcting conf-file for Snapscan /etc/sane.d/snapscan.conf :

the line

firmware /usr/share/sane/snapscan/your-firmwarefile.bin

should be changed to

firmware /usr/share/sane/snapscan/esfw41.bin


4. Testing if it works in XSane.



Best regards,
Dmitry Piyavkin

Re: Multiple live iso's on a single bootable flash drive?


On 03.05.2016 18:37, Richard Owlett wrote:

On 5/3/2016 8:41 AM, Piyavkin wrote:

On 03.05.2016 13:11, Richard Owlett wrote:

On 5/2/2016 2:15 PM, Piyavkin wrote:

On 02.05.2016 20:50, Brian wrote:
[snip]

The OP probably has it up and running by now.


I hop so (about the OP). There is no news from him. I'm worrying
(for him and for his hdd).



Actually I haven't. Life has a tendency to intrude. Even as a
retiree one cannot attend to a hobby 24/7.

Seriously though, the delay itself may prove useful. This
discussion is giving me some insight into another project I've
been working on.

I hope the result will be a document aimed at the beginner but
not the picture book style that seems prevalent at that level.
Much Linux documentation is written by experts for experts.

As to my hdd, I bought a used laptop explicitly for my Linux
experiments ;)




I'm just kidding about hdd, sorry. )


 You have to know some my history. Sometime back I was chided 
for doing multiple installs per day when I was working out what I 
did/didn't want in an installation.


Yeah, that's what I was talking about. )
Multiboot USB-drive (flash) with .iso is a perfect tool for the case.




 If you pay proper attention
to the process it is hardly that there'll be any problem.

Glad to hear that the discussion was in some way useful for you.
May you share with us your resulted document?



That's my intention. But knowing my typical pace of completing 
projects it may be a while ;/





OK. I'm not really fast too. So I'm fine with it. Thanks in advance.


Best regards,
Dmitry Piyavkin

Re: [OT]: May Day


On 03.05.2016 16:34, Lisi Reisz wrote:

On Tuesday 03 May 2016 14:22:26 Piyavkin wrote:

On 03.05.2016 02:21, Lisi Reisz wrote:

…
So it is all a bit circular!!  But I find the concepts of "International
Workers Day" and "the proletariat" intrinsically unpleasant.  I am in
sympathy with Eric Blair (aka George Orwell).  It is as wrong to murder
someone for being an aristo as it is to murder him for being a peasant.

Oh, my!..


I've sent Hello and best wishes on occasion to fellow people. All that
was about expression of solidarity. And intended to be lean and short,
like: a) have a nice day! b) have a nice day too; c) done. In case
someone may be irritated by relatively irrelevant posts or just not
interested the topic was marked as [OT] for filtering purposes (Subject

: contains : [OT] --> Delete from POP Server).

I've received 2 types of responses:
1. People have sent their best wishes to me in return (mostly in
private, because many are already too terrorized to express their
opinion openly).
2. Some others on the basis of «deep» linguistic analysis have detected
here some presence of Nazi and, out of a blue, bloody intention to
murder (by best wishes, I guess).

What's wrong with you, guys?


Best regards,
Dmitry Piyavkin

That was sent off-list and should not have been quoted here.  Presumably Dan
forwarded it to you.


No, the letter have came from list:
https://lists.debian.org/debian-user/2016/05/msg00138.html


Solidarity?  With whom?  Oh, yes, The Workers.


I've explicitly explained with whom. Twice. It was utterly innocent and 
unoffending. And I'm amazed how you guys found pretext to turn that 
small topic into a flame.


So, working people are now kind of terrorists, huh?


Read the history of the French Revolution.  Or the Stalinist purges.


Lisi, do it yourself, please. French Revolution was bourgeois one. So 
capitalists should pay the bill, I guess.


And what kind of logic is it? So, if colonial Europe had performed all 
those uncountable atrocities all over the world and gave birth to 
Fascism we now should treat Europeans as spawns of evil?




Politics does not belong here,


Yeah, but you do it regularly. Discussing, by example, politcorrect 
forms of addressing audience with proper gender connotations.
I personally have nothing against it or against other touch of OT 
subjects (like movies, personal stories, etc.). There is no human 
interaction without human part in it. The FOSS as itself has significant 
political dimension. Why we shouldn't discuss it?


So, I guess the problem is not in the politics as itself. The problem is 
in intolerance to other's opinion and desire to silence it under any 
ostensible pretexts (politics as a dirty and unmentionable subject). 
Mainstream propaganda encourage the sectarian attitude to neutralize p2p 
communications while they have full control over mass media. So the 
people which opinion differs from PoV of Ministry of Truth should be 
isolated, marginalized, express their views only somewhere in closed 
underground communities and publicly be in shame as some perverts, as if 
that are them who's starting wars, keeping their hands deep in other's 
pockets, buying elections and parliaments, lobbying laws, polluting 
lands, financing military contractors whose hostile actions devastate 
whole regions with results in hundreds of thousands dead and in millions 
displaced as refugees (and doing so they are even pointing fingers at 
century old Stalin, bah!), etc.


The thread was auto-extinguishing by its nature and explicitly marked as 
off-topic. You might easily ignore it if you don't share the opinion, 
you might easily filter it in your mail client. And it would be finished 
in 2 days with a few replies and then be silently removed as outdated 
OT. No, you've found funny linguistic excuses for phony accusations, 
blown the flame about nothing, just to cry in the end «Enough already!» 
Why? Because you fear that other people are… well… not so smart and 
sharp as you are, they may fall for treacherous speeches of some 
tempting snake, and should be protected at any cost. Constant vigilance! 
We're on a mission.


Please, give the people to decide by themselves. People are not stupid. 
If subject rise no interest, it will be ignored,  will die by itself, 
and never be repeated again.



[!] Here I stop and quit all the Nazi, Stalin, innovative politic 
linguistics and all the crap, which I didn't intend to mention at all. 
Please, start your own thread for that. Or follow your own advice and 
stop the insinuations.



but nor do ad hominem attacks.



What attacks? Whom attacking whom? If you feel that I'm attacking you 
(but not your actions), I sincerely apologise and will never do that again.


What you, guys, are even doing in the thread if you believe that wishing 
best each other and expressing solidarity with fellow people means 
divisiveness, call to a murder, and all the horrible things you called 
it? It

Re: Multiple live iso's on a single bootable flash drive?


On 03.05.2016 13:11, Richard Owlett wrote:

On 5/2/2016 2:15 PM, Piyavkin wrote:

On 02.05.2016 20:50, Brian wrote:
[snip]

The OP probably has it up and running by now.


I hop so (about the OP). There is no news from him. I'm worrying
(for him and for his hdd).



Actually I haven't. Life has a tendency to intrude. Even as a retiree 
one cannot attend to a hobby 24/7.


Seriously though, the delay itself may prove useful. This discussion 
is giving me some insight into another project I've been working on.


I hope the result will be a document aimed at the beginner but not the 
picture book style that seems prevalent at that level. Much Linux 
documentation is written by experts for experts.


As to my hdd, I bought a used laptop explicitly for my Linux 
experiments ;)





I'm just kidding about hdd, sorry. ) If you pay proper attention to the 
process it is hardly that there'll be any problem.


Glad to hear that the discussion was in some way useful for you. May you 
share with us your resulted document?



Best regards,
Dmitry Piyavkin

Re: [OT]: May Day


On 03.05.2016 16:05, John Hasler wrote:

Piyavkin writes:

You're right. I've already sent to Pablo Escobar and to local criminal
lords my letters with apologies. Explaining that since using «bad
words» is being a nazi, we shouldn't call them criminals and parasites
anymore.

Then call them criminals.  The word "parasite* usually refers to people
who have committed no crimes but are *in the speaker's opinion*
parasites.


I do. Please, reread.


Best regards,
Dmitry Piyavkin

Re: [OT]: May Day


On 03.05.2016 02:21, Lisi Reisz wrote:


…
So it is all a bit circular!!  But I find the concepts of "International
Workers Day" and "the proletariat" intrinsically unpleasant.  I am in
sympathy with Eric Blair (aka George Orwell).  It is as wrong to murder
someone for being an aristo as it is to murder him for being a peasant.


Oh, my!..


I've sent Hello and best wishes on occasion to fellow people. All that 
was about expression of solidarity. And intended to be lean and short, 
like: a) have a nice day! b) have a nice day too; c) done. In case 
someone may be irritated by relatively irrelevant posts or just not 
interested the topic was marked as [OT] for filtering purposes (Subject 
: contains : [OT] --> Delete from POP Server).


I've received 2 types of responses:
1. People have sent their best wishes to me in return (mostly in 
private, because many are already too terrorized to express their 
opinion openly).
2. Some others on the basis of «deep» linguistic analysis have detected 
here some presence of Nazi and, out of a blue, bloody intention to 
murder (by best wishes, I guess).


What's wrong with you, guys?


Best regards,
Dmitry Piyavkin

Re: [OT]: May Day


On 03.05.2016 01:50, Michael Lange wrote:

On Mon, 02 May 2016 14:52:26 -0500
John Hasler <jhas...@newsguy.com> wrote:


Michael Lange writes:

I would by the way strongly request to dismiss the use of the word
"parasite" when speaking about human beings, since - speaking frankly
- this sounds a lot like nazi-speech

So does any classification of people into "them" and "us".

To some extent, but I feel classifying people into "us humans" and "them
animals" is even a lot worse.

Regards

Michael

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

No one wants war.
-- Kirk, "Errand of Mercy", stardate 3201.7

.


You're right. I've already sent to Pablo Escobar and to local criminal 
lords my letters with apologies. Explaining that since using «bad words» 
is being a nazi, we shouldn't call them criminals and parasites anymore. 
Now they should be named «socially handicapable alternative 
entrepreneurs». Someone of them has sent me back a sample sachet with a 
diacetylmorphine, and now I've got to go. Gonna to be busy for a while.



Michael, let's call Voldemort — Voldemort and do not profane serious 
things with silly word playing games.


https://www.youtube.com/watch?v=vuEQixrBKCc



For your information:

/American Heritage Dictionary of the English Language, Fifth Edition/
parasite —
a. One who habitually takes advantage of the generosity of others 
without making any useful return.

b. One who lives off and flatters the rich; a sycophant.

/Collins English Dictionary – Complete and Unabridged, 12th Edition/
parasite —
2. a person who habitually lives at the expense of others; sponger
3. (formerly) a sycophant

/Random House Kernerman Webster's College Dictionary/
parasite —
2. a person who receives support or advantage from another without 
giving any useful or proper return, as one who lives on the hospitality 
of others.
3. (esp. in ancient Greece) a person receiving free meals in return for 
amusing conversation or flattery.



The meaning «An organism that lives and feeds on or in an organism of a 
different species and causes harm to its host» in most dictionaries 
marked as special term of Biology. Which is probably derivative from 
traditional uses cited above (because biology is a relatively young 
science). And we even didn't discussed any biological subject to do such 
leaps in interpretations, we talked literally about those who «receives 
from others without giving any useful or proper return».



You may write to the language authorities and explain them your theory 
about Nazi-speech and how to use English properly.



Best regards,
Dmitry Piyavkin

Re: Multiple live iso's on a single bootable flash drive?


On 02.05.2016 20:50, Brian wrote:

On Mon 02 May 2016 at 17:54:35 +0300, Piyavkin wrote:


On 01.05.2016 17:55, Brian wrote:

On Sat 30 Apr 2016 at 21:02:56 +0300, Piyavkin wrote:


What's wrong with it?

Nothing, but a Ubuntu live ISO serves a different purpose from the
Debian installer. Which is not to say enhancing a netinst ISO is not
worthwhile and would benefit a few people. Patches to have iso-scan in
these ISOs' initrds were provided a couple of years ago and booting an
ISO with GRUB is on the installer team's list of feature requests.
Meanwhile, there is hd-media and a extra stanza in grub.cfg.

Well, and that's a good part of the story.
As I can see, the all needed functionality is already here. It just waiting
for some reason to be incorporated in distros.

Shall we compare like and like? Not a Ubuntu (or other live image) with
the Debian installer but a Ubuntu live image with a Debian live image.
All six of the Debian live images can be copied to a single partition
and a selected one booted with GRUB. No difference compared with Ubuntu.

Any of the running live images can be installed to hard disk, Again no
difference compared with Ubuntu.

Where's the problem for a new or experienced user? What need is there
for iso-scan to be incorporated into a netinst initrd?


I thought exactly in the same way.
Though I had to have installer images on my USB-drive with distroes, 
because I'd repeatedly failed to do installation from Debian live images 
(I don't remember details now; as a live images they started fine in all 
other respects). So I have two different images of one distro for live 
play/demonstration, and other one for installation. And here with the 
second one the quest started (you already know the rest).


Never had such issue with Ubuntu, because they have the same image as a 
live and as an installer, and everything needed already have been 
bundled in here by default.
From my personal point of view my user experience with Ubuntu install 
was much more simpler.


If in recent years everything was changed in better way, well, OK then.


Not a decent argument. Operations such as partitioning and formatting a
USB stick and installing GRUB to its MBR all require root privilege.

Of course, but the point is: with which «case there are lesser chances» to
shoot in own leg:

Case 1: do sudo work and pay proper attention 1 time and fit for all.

Case 2: do it every time when you update your boot collection on the
USB-drive.

The events are independent. Carrying out an operation more than once
does not change the probability of carrying it out correctly. I can
claim I have used 'ls -l' fifty times in a day without error. :)


Pff! I can easily do 'ls -l' 62 times! )


(Incidentally, a simple udev .rules file avoids having to be root to
deal with a USB stick, so the issue doesn't arise for me).

[...Snip...]


3-rd, I guess, typical Debian user is not a total Linux newbie, who's
problem are an ordinary partitioning process or grub install. I guess, them
more interested in such capabilities as to boot multiple OS/versions from
one partition or to run installers from virtually any commont USB-drive
(properly prepared in advance and easily reconfigurable in use). And
unexplained complication here — that is a problem.

As living illustration to the last point we may look at our real situation:
desirable multiboot from single flash drive (see subj of the thread). One
solution suggests doing partitioning and all the hard stuff just once. Other
— every time when new image added or old one changed and the complexity of
the solution is growing with the changes. OK, I see that mapping .iso to
stick is useful in some common situations (when you, for example, wish to
present ready to use flash drive installer to someone else). But why the
redundant work should be forced upon a user in a slightly more sophisticated
situations? Can it lead to the situation when a skilled user (the target
audience of the Debian) just be reluctant or simply tired repartitioning
their USB-drive with every release and settle down with Linux distribs that
do not demand to much without a reason?

Nothing is forced upon a user. Six Debian live images. One partition.
Add new images when necessary. Install if you wish. Multiboot without
tears. The OP probably has it up and running by now.


I hop so (about the OP). There is no news from him. I'm worrying (for 
him and for his hdd).


  

Resume: I advocate that both alternatives should be presented in Debian
distros. That'll make me happier.

They are! You should be overjoyed.

If you have gained the impression that isco-scan in a netboot initrd
would not be unwelcome to me but I won't lose any sleep over its
omission, then you have interpreted my feelings correctly.



OK, I'm overjoyed then.


Best regards,
Dmitry Piyavkin

Re: [OT]: May Day


On 02.05.2016 20:48, John Hasler wrote:

Piyavkin writes:

But, I wonder, /whom/ exactly did you bear in mind when you triggered
«politics detection» alert at words «professional parasites»? )
...Ouch!

Politicians, most likely.  All of them.  It's not true, of course.  Many
of them are amateur parasites.


No way, John! Why you say such about those noble and worthy gentlemen?
Lisi may not think like that!


Best regards,
Dmitry Piyavkin

Re: [OT]: May Day

On 02.05.2016 19:05, Thomas Schmitt wrote:

Hi,

Haines Brown wrote:

Incidentally, I'm not suggesting that FOSS is crypto-socialist.

Piyavkin wrote:

No, but it is new growing relations of production.

It's in no way crypto, but quite near to the vision of Karl Marx in the
19th century. He expected it to happen for classic economy after
socialism succeeded and evolved into true communism.

Google found me a nice summary in german language:

http://stattkapitalismus.blogsport.de/2008/11/12/665jeder-nach-seinen-faehigkeiten-jedem-nach-seinen-beduerfnissen/
which is obviously by a person not aware of free software.
It quotes Marx' book "Das Kapital", 1867:
"Stellen wir uns endlich, zur Abwechslung, einen Verein freier Menschen
vor, ..."
which i translate as:
"Let us imagine finally, just for a change, a club of free humans, who
work with commonly used means of production and self-confidently spend
their many individual powers of work as a united work effort for society.
[...] The result of the club's work is a product of society.
In part it serves in turn as means of production. Another part is
consumed by the members as subsistence."

He did not imagine highly valuable objects which can be copied at
nearly no cost and can be used by affordable means. He only predicts that
the relation between production and consumption will change but then goes
on to describe a non-communist structure of work merits and right to
consume.
In our world, even Linus Torvalds gets more from GNU/Linux than he gave.

Marx also did not imagine that a (from his view) ideal society would emerge
on top of capitalism rather than replacing it.

Have you said «Karl Marx»?!.. Oh, no!.. Lisi will get you! )

If be serious, such changes are going on not only in software industry.
And I don't think it'll be finished even in the near future which may be
surprising in the end. That is a long and incremental process. But it is
a huge subject to discuss in an OT thread in a mail list. Such subjects
should be debated in form of articles in academic manner, I believe [but
try to rise the subject in an economics community, and their political
police will get you too].
Though it may be interesting for you to look also at the Schumpeter's
theory of development. He's usually considered as an opponent to Marx,
but their theories in fact pretty complimentary.

In real life i do not get my food for free. Housing is expensive.
Even the preconditions for linuxing cost monthly money.
But during my life there were only a few months of what i'd call work.
I earn my keep by selling myself doing my favorite sports.
Nevertheless i am aware that i am swimming as grease drop on a watery soup
of hardship and boredom.

Why don't you consider prof sport activity as a labour? Especially if
you earn by that. What's the sports, by the way?

As to hardship and boredom: first is an objective condition, second is a
subjective attitude and, in my view, should be avoided as deadly sin
whatever situation you happened to be in. Life is too damn short to be
bored.

Have a nice day :)

Thomas

Best wishes to you too!

Best regards,
Dmitry Piyavkin

Re: [OT]: May Day


On 02.05.2016 17:16, Lisi Reisz wrote:

On Monday 02 May 2016 15:07:29 Piyavkin wrote:

professional
parasites

OUCH!  Can we keep politics out of this - be it far left or far right?  This
list is not the place.

Lisi



Hi, Lisi!

Actually, I'm not intended to elaborate the o-topic any further beyond 
the Worker's Day congratulations.


And by mentioned «professional parasites and criminals» I meant in the 
first place exactly those people who live by criminal, fraud, deception, 
and literally at other's expense. Which is by definition a description 
of lumpen-proletariat (declassified) stratum.


But, I wonder, /whom/ exactly did you bear in mind when you triggered 
«politics detection» alert at words «professional parasites»? ) ...Ouch!



Best regards,
Dmitry Piyavkin

Re: Multiple live iso's on a single bootable flash drive?


On 01.05.2016 17:55, Brian wrote:

On Sat 30 Apr 2016 at 21:02:56 +0300, Piyavkin wrote:


On 29.04.2016 22:59, Brian wrote:

On Fri 29 Apr 2016 at 21:57:53 +0300, Piyavkin wrote:

[...Snip...]


It is a source of contention (and a number of bug reports) but it is by

design.

Why such design?

Suppose you want to install Debian; that's the objective after all. An
isohybrid allows mounting the USB stick directly. You dd/cat/cp the
image to a USB stick and boot and there you are - Debian is installed.

What does loop-mounting of an ISO file with GRUB give you?

With Ubuntu distrib you can do the same.
And still you can run it straight from .iso without additional quest and use
of shaman drum.
What's wrong with it?

Nothing, but a Ubuntu live ISO serves a different purpose from the
Debian installer. Which is not to say enhancing a netinst ISO is not
worthwhile and would benefit a few people. Patches to have iso-scan in
these ISOs' initrds were provided a couple of years ago and booting an
ISO with GRUB is on the installer team's list of feature requests.
Meanwhile, there is hd-media and a extra stanza in grub.cfg.


Well, and that's a good part of the story.
As I can see, the all needed functionality is already here. It just 
waiting for some reason to be incorporated in distros.



In my view, it is much more convenient to download new .iso files (or
replace old ones) straight to USB-drive and copy+paste one more menuentry in
grub.cfg (working in any OS which supports FAT), than to do the same (.iso
download in some dedicated folder, changing in grub.cfg) plus
partitioning/repartitioning (with calculation of proper partitions' sizes
every time when you want to use more then 2 distros on 1 USB-drive) and
copying (which requires *nix-like OS already running). If I understand the
process correctly.

The "convenience" argument is a decent one, although it does apply to
quite a narrow use-case.
  

And much more safer, I believe. Because in the first case there are lesser
chances that in a stressful & hasty time doing one more

cp debian-hot-new.iso /dev/sda
sync

you may end up like:«Oh, wait… was it sdb?.. wait… and what was sda then?..
Oh… that was my 40+ years long project in astrophysics… THANK YOU, Debian,
for your design!»

Not a decent argument. Operations such as partitioning and formatting a
USB stick and installing GRUB to its MBR all require root privilege.
  


Of course, but the point is: with which «case there are lesser chances» 
to shoot in own leg:


Case 1: do sudo work and pay proper attention 1 time and fit for all.

Case 2: do it every time when you update your boot collection on the 
USB-drive.



I think KISS principle should be applied not only to tools, but to the user
experience too. Which may be more important. Cause, in the end, everything
we do, we do for others. And who ignores it (for some their reason) will
suffer. For soft production one of the rules may sound like: «If you create
unnecessary obstacles in installation process, you hinder distribution and
hence adoption. Good luck!»

Advice to a newcomer for installing Debian
--
cat ISO to stick. Boot.

Alternative advice to a newcomer for installing Debian
--

Clean stick with dd. Partition stick. Format partition. Copy ISO to
partition. Install GRUB. Construct a grub.cfg and copy to partition.
Boot.



No-no, Brian, look at the situation from slightly different angle:

1-st, it is not a XOR choice. Proper design should provide both options. 
Or either part of potential users will be lost, as I said.


2-nd, you are a bit exaggerating the differences between the possible 
installing ways for newcomer. In your 2-nd alternative their already has 
partitioned and formatted stick (which shipped with FAT by default). 
Installation of grub (by one line of command) and copypasting menuentry 
in text file from official wiki (where the information should live) is 
not harder task then performing cat, trying not to kill fs on the PC's 
main hdd in the process. But the real fun starts when our newcomer wanna 
have more than one distro (cause he's a newcomer and wanna try some 
variants first to choose most promising one) but has only one USB-drive. 
And the 2-nd alternative now actually doesn't exist for newcomer at all, 
because it is too complicated for him.


3-rd, I guess, typical Debian user is not a total Linux newbie, who's 
problem are an ordinary partitioning process or grub install. I guess, 
them more interested in such capabilities as to boot multiple 
OS/versions from one partition or to run installers from virtually any 
commont USB-drive (properly prepared in advance and easily 
reconfigurable in use). And unexplained complication here — that is a 
problem.


As living illustration to the last point we may look at our real 
situation: desirable multiboot from single flash drive (see subj of the 
thread). One solution

Re: [OT]: May Day


On 01.05.2016 21:27, Haines Brown wrote:
Incidentally, I'm not suggesting that FOSS is crypto-socialist. Haines 


No, but it is new growing relations of production.


Best regards,
Dmitry Piyavkin

Re: [OT]: May Day


On 01.05.2016 18:39, Nate Bargmann wrote:

* On 2016 01 May 08:23 -0500, John Hasler wrote:

Hans Vogelsberger writes:

Why workers only? There are other humans, too.

Class struggle.

I struggled to pay attention in social studies class way back in grade
school 40+ years ago.  Does that count?

;-)

Happy May Day from the workers paradise of a muddy farm somewhere in
northern Kansas.

- Nate



And special Best Wishes come to — Nate from Kansas!

[a soft but energetic banjo music slowly fade in in the background]


Thanks, Nate! )


Best regards,
Dmitry Piyavkin

Re: [OT]: May Day


On 01.05.2016 15:26, Hans Vogelsberger wrote:

Am Sat, 30 Apr 2016 21:49:06 +0300
schrieb Piyavkin <piyav...@gmail.com>:


Congratulations with International Worker's Day to all the working
(in FOSS industry and at all) people! )
Have a nice day!


Best regards,
Dmitry Piyavkin


Why workers only? There are other humans, too.



Hans, I've actually addressed all good people, since wast majority of 
the humankind (up to 99%, I guess, except classes of professional 
parasites and criminals) lives by working and serving other's in some 
useful social roles. And actually do not counterpose but complement each 
other. I've consciously said «working people», including here all labour 
classes.


Yeah, I understand that it sounds funny and the humankind doesn't even 
hear me, because I have no my own mass media (and the access to the 
carefully cultivated and formed mass consciousness). I've just sent my 
best wishes on the occasion to those who can hear me, can accept it and 
bounce back. End of story.



And, yeah, I have absolutely nothing at all against any goddess of 
flowers, especially if some come my way in a nice sunny May day.




Best regards,
Dmitry Piyavkin

[OT]: May Day

2016-04-30 Thread Piyavkin

Congratulations with International Worker's Day to all the working (in 
FOSS industry and at all) people! )

Have a nice day!


Best regards,
Dmitry Piyavkin

Re: Multiple live iso's on a single bootable flash drive?

2016-04-30 Thread Piyavkin


On 29.04.2016 22:59, Brian wrote:

On Fri 29 Apr 2016 at 21:57:53 +0300, Piyavkin wrote:


Though, I agree, parameters may vary. May be «fromiso» in the last example
is a better choice indeed. I'd just experimented with related instructions
until it started working for me and I'm done.

Look carefully. I didn't recommend 'fromiso'.
OK. ) Then, archlinux.org wiki page which I referenced did. But quite 
opposite: to prefer fromiso over findiso.
I don't remember if I tried it, and if it works, but now I think that 
findiso may be better option, because it's at least more portable (let 
say so; menu records do not tied up to USB-drive UID). As to differences 
in speed — I don't know. Don't think it may be any significant issue.





It is a source of contention (and a number of bug reports) but it is by

design.

Why such design?

Suppose you want to install Debian; that's the objective after all. An
isohybrid allows mounting the USB stick directly. You dd/cat/cp the
image to a USB stick and boot and there you are - Debian is installed.

What does loop-mounting of an ISO file with GRUB give you?



With Ubuntu distrib you can do the same.
And still you can run it straight from .iso without additional quest and 
use of shaman drum.

What's wrong with it?


In my view, it is much more convenient to download new .iso files (or 
replace old ones) straight to USB-drive and copy+paste one more 
menuentry in grub.cfg (working in any OS which supports FAT), than to do 
the same (.iso download in some dedicated folder, changing in grub.cfg) 
plus partitioning/repartitioning (with calculation of proper partitions' 
sizes every time when you want to use more then 2 distros on 1 
USB-drive) and copying (which requires *nix-like OS already running). If 
I understand the process correctly.


And much more safer, I believe. Because in the first case there are 
lesser chances that in a stressful & hasty time doing one more


cp debian-hot-new.iso /dev/sda
sync

you may end up like:«Oh, wait… was it sdb?.. wait… and what was sda 
then?.. Oh… that was my 40+ years long project in astrophysics… THANK 
YOU, Debian, for your design!»



I think KISS principle should be applied not only to tools, but to the 
user experience too. Which may be more important. Cause, in the end, 
everything we do, we do for others. And who ignores it (for some their 
reason) will suffer. For soft production one of the rules may sound 
like: «If you create unnecessary obstacles in installation process, you 
hinder distribution and hence adoption. Good luck!»



(Thank you for attention! I give the floor to the next speaker.)


Best regards,
Dmitry Piyavkin

Re: Multiple live iso's on a single bootable flash drive?

2016-04-29 Thread Piyavkin

On 29.04.2016 16:21, Brian wrote:

On Thu 28 Apr 2016 at 22:26:33 +0300, Piyavkin wrote:

On 28.04.2016 19:31, Curt wrote:

On 2016-04-28, Richard Owlett <rowl...@cloud85.net> wrote:

I've some untried VAGUE ideas on how to accomplish.

Maybe this would work if I'm understanding you correctly:

http://www.pendrivelinux.com/boot-multiple-iso-from-usb-via-grub2-using-linux/#more-5352

Two questions:
Has this been done before?
Any comments?

Yep, it can be done and pretty easy. I've done (been doing) it using some
instructions from the pendrivelinux.com.

The idea is simple:

1. Install grub2 on a USB-drive.

The drive should first be partitioned and the partitiion(s) formatted. I
use FAT16 and vfat. If it is an installer image which GRUB is booting
the fat and vfat modules are available from the start of the install.

2. Place there .iso image (in root of the drive or in some directory as you
like).

3. Use in the grub.cfg instructions like such:

menuentry "Xubuntu 14.04.1 Desktop i386 ISO" {
set isofile="/xubuntu-14.04.1-desktop-i386.iso"
loopback loop $isofile
linux (loop)/casper/vmlinuz boot=casper iso-scan/filename=$isofile noeject
noprompt splash --
initrd (loop)/casper/initrd.lz
}

4. Profit!

It is worth stressing that 'boot=casper' is for Ubuntu-based live images
only. It will not work with a Debian live image. I think the same is
true of 'iso-scan/filename=...'.

Images from Ubuntu family works fine. As to the Debian there was some
weirdness (I don't know if it still persists in newer versions), but it is
possible to dodge it. Simple way (from here

You have switched from talking about Debian live images to installer
images. GRUB's loopback facility cannot be used with them because the
iso-scan and load-iso packages are only included in the hd-media initrd.
It is a source of contention (and a number of bug reports) but it is by
design.

There should be no expectation of GRUB's loopback working to install
Debian using just d-i's initrd in the ISO.

https://wiki.archlinux.org/index.php/Multiboot_USB_drive#Debian ) is:

0. Do steps 1 and 2 from list above.

1. Download proper initramfs (which matches your desirable Debian .iso
image) from here:
https://mirrors.kernel.org/debian/dists/stable/main/installer-amd64/current/images/hd-media/initrd.gz
https://mirrors.kernel.org/debian/dists/stable/main/installer-i386/current/images/hd-media/initrd.gz
http://mirrors.kernel.org/debian/dists/stable/main/installer-i386/current/images/netboot/debian-installer/i386/initrd.gz
etc.

2. Place it somewhere on the USB-drive (in /deb/hdd directory, for example).

3. Use such instructions in grub.cfg:

menuentry "Debian 7.8.0 Desktop i386 Gnome CD ISO (CD Install;
/deb/hdd/initrd.gz)" {
set isofile="/debian-7.8.0-i386-CD-1.iso"
set initrdfile='/deb/hdd/initrd.gz'
loopback loop0 $isofile
linux (loop0)/install.386/vmlinuz iso-scan/ask_second_pass=true
iso-scan/filename=$isofile priority=low
# initrd (loop0)/install.386/initrd.gz
initrd $initrdfile
}

4. Profit again!

Debian Live iso works without additional downloads (some kind of
self-sufficient). It is possible to use fromiso param and to provide precise

The live ISO initrds contain sufficient to scan and find the ISO and
provide a loop device.

path to your iso and address your USB-drive by UID like this (theoretically
it should speed up the process):

menuentry "Debian 7.8.0 Desktop i386 Xfce ISO (LIVE from ISO)" {
set isofile="/debian-live-7.8.0-i386-xfce-desktop.iso"
loopback loop0 $isofile
linux (loop0)/live/vmlinuz1 boot=live
fromiso=/dev/disk/by-uuid/19C5-2FB2$isofile live-media-path=/live config
initrd (loop0)/live/initrd1.img
}

vmlinuz1 and initrd1.img are for booting live-586. Files in the isolinux
directory of the ISO should be consulted for other kernels and initrds
to use.

I'd suggest 'findiso=$isofile' as a possible replacement for the
'fromiso=...' directive.

Yes, USB-drive should be partitioned and formatted properly (though,
FAT16 can be a problem with relatively fat images or partitions, I
guess; FAT32 works OK). Just have skipped the preparations as obvious
and thoroughly described part in the pendrivelinux.com instructions (see
the link above). I've intended to dwell on the more interesting part of
the story.

It worth to point out that all the stated menuentries are excerpts from
my current grub.cfg from USB-drive with a pile of .iso images on it. I'm
using both Debian live images and Debian installer images (hd-media,
netboot). All that tested (up to Debian 7.8) and works fine.

Though, I agree, parameters may vary. May be «fromiso» in the last
example is a better choice indeed. I'd just experimented with related
instructions until it started working for me and I'm done.

> It is a source of contention (and a number of bug reports) but it is
by design.

Why such design?

Best regards,
Dmitry Piyavkin

Re: Multiple live iso's on a single bootable flash drive?

2016-04-28 Thread Piyavkin


Hi all!

On 28.04.2016 19:31, Curt wrote:

On 2016-04-28, Richard Owlett <rowl...@cloud85.net> wrote:

I've some untried VAGUE ideas on how to accomplish.

Maybe this would work if I'm understanding you correctly:

http://www.pendrivelinux.com/boot-multiple-iso-from-usb-via-grub2-using-linux/#more-5352


Two questions:
Has this been done before?
Any comments?

TIA







Yep, it can be done and pretty easy. I've done (been doing) it using 
some instructions from the pendrivelinux.com.



The idea is simple:

1. Install grub2 on a USB-drive.

2. Place there .iso image (in root of the drive or in some directory as 
you like).


3. Use in the grub.cfg instructions like such:

menuentry "Xubuntu 14.04.1 Desktop i386 ISO" {
 set isofile="/xubuntu-14.04.1-desktop-i386.iso"
 loopback loop $isofile
 linux (loop)/casper/vmlinuz boot=casper iso-scan/filename=$isofile 
noeject noprompt splash --

 initrd (loop)/casper/initrd.lz
}

4. Profit!


Images from Ubuntu family works fine. As to the Debian there was some 
weirdness (I don't know if it still persists in newer versions), but it 
is possible to dodge it. Simple way (from here 
https://wiki.archlinux.org/index.php/Multiboot_USB_drive#Debian ) is:


0. Do steps 1 and 2 from list above.

1. Download proper initramfs (which matches your desirable Debian .iso 
image) from here:

https://mirrors.kernel.org/debian/dists/stable/main/installer-amd64/current/images/hd-media/initrd.gz
https://mirrors.kernel.org/debian/dists/stable/main/installer-i386/current/images/hd-media/initrd.gz
http://mirrors.kernel.org/debian/dists/stable/main/installer-i386/current/images/netboot/debian-installer/i386/initrd.gz
etc.

2. Place it somewhere on the USB-drive (in /deb/hdd directory, for example).

3. Use such instructions in grub.cfg:

menuentry "Debian 7.8.0 Desktop i386 Gnome CD ISO (CD Install; 
/deb/hdd/initrd.gz)" {

 set isofile="/debian-7.8.0-i386-CD-1.iso"
 set initrdfile='/deb/hdd/initrd.gz'
 loopback loop0 $isofile
 linux (loop0)/install.386/vmlinuz iso-scan/ask_second_pass=true 
iso-scan/filename=$isofile priority=low

# initrd (loop0)/install.386/initrd.gz
 initrd $initrdfile
}

4. Profit again!


Debian Live iso works without additional downloads (some kind of 
self-sufficient). It is possible to use fromiso param and to provide 
precise path to your iso and address your USB-drive by UID like this 
(theoretically it should speed up the process):


menuentry "Debian 7.8.0 Desktop i386 Xfce ISO (LIVE from ISO)" {
 set isofile="/debian-live-7.8.0-i386-xfce-desktop.iso"
 loopback loop0 $isofile
 linux (loop0)/live/vmlinuz1 boot=live 
fromiso=/dev/disk/by-uuid/19C5-2FB2$isofile live-media-path=/live config

 initrd (loop0)/live/initrd1.img
}


Best regards,
Dmitry Piyavkin

Re: Ransomware meets Linux - on the command line!

2015-11-13 Thread Piyavkin

On 12.11.2015 21:14, Ralph Katz wrote:

On 11/11/2015 10:24 PM, Cindy-Sue Causey wrote:

[...]

Brian Krebs of Krebs On Security had
something on ransomware and Linux, just not labeled Ransm-C or
anything:

http://krebsonsecurity.com/2015/11/ransomware-now-gunning-for-your-web-sites/

IF I'm understanding correctly, he appears to have updated that
article with a *potential* way to beat it via a *potential*
vulnerability at least until the perpetrators upgrade their own
tactics, anyway.

I like what Brian's been doing. I can cognitively understand a LOT of
what he writes about. He's caught SlashDot's eye a time or two, too.

Adding another keyword here, Linux.Decoder.1, which Brian says was a
name dubbed by "Russian antivirus and security firm Dr.Web". It may or
may not be the same as the other, but sounds like it works
similar'ISH.

Next stop is to pop over to a group called BlindWebbers. I'd seen
Brian's email subject line earlier and thought instantly of them, just
didn't get around to opening it then. The guy in Brian's article makes
it sound like it's a little time consuming and still has incidental
glitches afterwards.

That's presumably coming from someone with no visual disabilities. The
difficulty level of getting one's website back would understandably
rise relative to one's ability or lack thereof to actually see what's
going on within the file hierarchy. AND apparently each single
file that reportedly stands to potentially gather random bits AFTER
the files have been decrypted.

As a user, I too, find Krebs informative. Also notable was this recent
Washington Post article about Linus Torvalds and Linux security:

http://www.washingtonpost.com/sf/business/2015/11/05/net-of-insecurity-the-kernel-of-the-argument/

"Fast, flexible and free, Linux is taking over the online world. But
there is growing unease about security weaknesses."

Regards,
Ralph

The ransomware articles from the security companies are pure marketing
efforts to develop customer's «pain» and to exploit it. There is nothing
new in the scheme «pay or suffer». And the companies provide nothing new
as a «cure» either (which haven't been there for decades).

More over, I wonder, what is the difference between the «ransomware»
business model and so called «planned obsolescence» business model,
which, I guess, has become worldwide industrial standard nowaday? And in
what way should differ protective measures for both of them? I mean,
from the end users point of view, there is no much difference if their
data have been stolen/encrypted by one crook or if their data have been
lost because of «sudden» HDD fail planned in advance by another crook.
Except, may be, the fact that in the first case you still have a tiny
chance to get your precious data back (may be, which I doubt).

The WP article seems like a spin. It gives us a spooky filling of great
imminent danger radiating from the Linux, but in the same time it is
surprisingly shallow and inconcrete. Though it uses security thing as a
pretext, I guess, it's not about security.

Of course, I don't think the subject of Linux security does not deserve
attention or discussion. But what the point in such articles as the WP
example, except from not so subtly playing with mass opinion with pretty
obvious commercial intention?

Regards,
Piyavkin

Re: Ransomware meets Linux - on the command line!

2015-11-13 Thread Piyavkin


On 13.11.2015 23:35, Gene Heskett wrote:

On Friday 13 November 2015 14:19:00 Piyavkin wrote:


On 12.11.2015 21:14, Ralph Katz wrote:

On 11/11/2015 10:24 PM, Cindy-Sue Causey wrote:

[...]


Brian Krebs of Krebs On Security had
something on ransomware and Linux, just not labeled Ransm-C or
anything:

http://krebsonsecurity.com/2015/11/ransomware-now-gunning-for-your-
web-sites/

IF I'm understanding correctly, he appears to have updated that
article with a *potential* way to beat it via a *potential*
vulnerability at least until the perpetrators upgrade their own
tactics, anyway.

I like what Brian's been doing. I can cognitively understand a LOT
of what he writes about. He's caught SlashDot's eye a time or two,
too.

Adding another keyword here, Linux.Decoder.1, which Brian says was
a name dubbed by "Russian antivirus and security firm Dr.Web". It
may or may not be the same as the other, but sounds like it works
similar'ISH.

Next stop is to pop over to a group called BlindWebbers. I'd seen
Brian's email subject line earlier and thought instantly of them,
just didn't get around to opening it then. The guy in Brian's
article makes it sound like it's a little time consuming and still
has incidental glitches afterwards.

That's presumably coming from someone with no visual disabilities.
The difficulty level of getting one's website back would
understandably rise relative to one's ability or lack thereof to
actually see what's going on within the file hierarchy. AND
apparently each single file that reportedly stands to potentially
gather random bits AFTER the files have been decrypted.

As a user, I too, find Krebs informative.  Also notable was this
recent Washington Post article about Linus Torvalds and Linux
security:

http://www.washingtonpost.com/sf/business/2015/11/05/net-of-insecuri
ty-the-kernel-of-the-argument/

"Fast, flexible and free, Linux is taking over the online world. But
there is growing unease about security weaknesses."

Regards,
Ralph

The ransomware articles from the security companies are pure marketing
efforts to develop customer's «pain» and to exploit it. There is
nothing new in the scheme «pay or suffer». And the companies provide
nothing new as a «cure» either (which haven't been there for decades).

More over, I wonder, what is the difference between the «ransomware»
business model and so called «planned obsolescence» business model,
which, I guess, has become worldwide industrial standard nowaday? And
in what way should differ protective measures for both of them? I
mean, from the end users point of view, there is no much difference if
their data have been stolen/encrypted by one crook or if their data
have been lost because of «sudden» HDD fail planned in advance by
another crook. Except, may be, the fact that in the first case you
still have a tiny chance to get your precious data back (may be, which
I doubt).

The WP article seems like a spin. It gives us a spooky filling of
great imminent danger radiating from the Linux, but in the same time
it is surprisingly shallow and inconcrete. Though it uses security
thing as a pretext, I guess, it's not about security.

Of course, I don't think the subject of Linux security does not
deserve attention or discussion. But what the point in such articles
as the WP example, except from not so subtly playing with mass opinion
with pretty obvious commercial intention?


Shallow?  Devoid of facts was my impression.


Yeah, may be it's better to name it in that way.

It seems that author don't care much about facts or even about 
explaining any opinions (what exactly were arguments of Spengler or 
Cook, by the way, and what exactly are the mentioned 6 and 12 points 
from them?). What he really seems be interested in is just constructed 
emotional impression (whom we should admire, whom we should condemn, and 
what we should believe in result), cooked with common journalist devices 
and a good chunk of ideological attitude instillation. It stinks for a mile.


Regards,
Piyavkin

Re: An UML editor for Linux


On 09.11.2015 02:02, Lisi Reisz wrote:

On Sunday 08 November 2015 20:23:25 Piyavkin wrote:

and kind women

:-)

Lisi

Sorry, I don't know the answer to your question.




Me too! ;)

Piyavkin

Re: An UML editor for Linux


On 08.11.2015 23:58, Sven Arvidsson wrote:

On Sun, 2015-11-08 at 23:23 +0300, Piyavkin wrote:

— could you please recommend me some decent UML editor for Linux?

Yes, I can see in repositories some, — like dia, umlet, gaphor… — but
which is decent?

I would probably draw them with Inkscape, but that's mostly because I'm
familiar with it, it's not specifically for UML.

You might also want to look at http://blockdiag.com/ (the building
blocks for that is in the archives in the package python-blockdiag)

I'm fine with Inkscape. It's very good tool. But it would be really 
annoying to draw every time everything from scratch. Or there is some 
set of UML primitives for Inkscape, plugin or something?
And then changes in diagrams (especially massive, especially regular) 
will be painful, I guess.


**blockdiag**

Thanks. It looks nice and promising.
Though it seems that it is not cover all UML notations from the box, and 
some tailoring will be needed (if possible).
It is not clear how to install it. Is it some kind of interactive 
on-line constructor?


Thanks for sharing it anyway.

Piyavkin

Re: An UML editor for Linux


On 09.11.2015 02:06, David Christensen wrote:

On 11/08/2015 12:23 PM, Piyavkin wrote:

— could you please recommend me some decent UML editor for Linux?


Umbrello is an impressive product that reminded me of Rational Rose. 
The last time I tried it for a Perl project, I was able to make some 
useful diagrams.  But, there were some Java-isms that I couldn't 
completely work around.  If you're into Java, please tell us what you 
think:


https://umbrello.kde.org/


Now I typically use LibreOffice Draw for diagrams and QCAD for vector/ 
CAD drawings.



David



**Umbrello**

Thanks for the hint. I'll try it to see better.
The problem is that it is from the KDE SDK. I'm on the Gnome. Though KDE 
tools are good looking, I'm trying to use Gnome or independent 
equivalents first.

Now the Umbrello seems very similar to gaphor, by the way.


**LibreOffice**

Has it some UML primitives packs or something?


**QCAD**

Thanks for the hint. May be useful for me (aside from UML). It seems 
that there is LibreCAD instead in Debian repositories.

Can you recommend any proper manual for it?

Piyavkin

Re: An UML editor for Linux


On 09.11.2015 14:08, Chris Edwards wrote:

On 09/11/15 09:23, Piyavkin wrote:

— could you please recommend me some decent UML editor for Linux?


I have a colleague who swears by PlantUML, which uses textual markup 
and GraphViz for generating diagrams.


http://www.plantuml.com/
--
Chris



Yay! That looks very promising: cover all kinds of UML diagrams; simple 
distribution; integration nearly with everything (from gedit to IDEs); 
well documented (language reference guide and extensive examples); 
customizable…

I just wonder why it is not in repositories yet?

Thanks, Chris!
It seems that your colleague is really savvy about the shrubberies.

Piyavkin

Re: An UML editor for Linux

On 09.11.2015 16:05, Maximiliano Sebastián Castro wrote:
Sorry to all those i hurt the heart recommending a short term solution 
for the original poster.

I think the pragmatism is very good but for some situations is enemy 
of practicality (or how to hell *practicidad* be in english).

Yes, I've tested (because I've worked in software architecture last 
~10 years and it's my core knowledge and the works I do every day): I 
started with Umbrello and Dia years, then I've tested online and new 
resourses like, and during there years I discovered that EA,  FOR ME, 
(***FOR ME, like all people that writes an email writes in their 
opinion not as the supreme true in the universe***) is the soft that 
acomplished my needs, but it does not mean I not support free tools, 
reporting bugs, collaborate and spread in college classes. And not 
mean that I integrate the the other free apps.

I did not know you could get to be a reactive, sarcastic, and 
tasteless response. For it was with good intentions, but it seems slow 
prefer collaboration instead of a rich debate on what can be better. 
The best is relative to each person, and although we are in the 
context of this list, I think you as an advocate of the open, you 
should be more open.

In any case if the response is not helpful to the person who posted, 
he will tell.

Maybe he tell us the first response helps him and ignore others 
responses, and problem resolved or maybe he try EA and resolve his 
problem and then look for similar soft.

I understand what you say, as you should also try to understand what I 
said.

Thank you and I hope you understand the intent of the message.

On Mon, Nov 9, 2015 at 8:56 AM, <to...@tuxteam.de 
<mailto:to...@tuxteam.de>> wrote:

-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

On Mon, Nov 09, 2015 at 09:13:08AM -0300, Maximiliano Sebastián
Castro wrote:
> >> — could you please recommend me some decent UML editor for Linux?
>
> If you don't care about freedom [...]

Who cares, anyway? Freedom, schmeedom.

> Yes, I know it's closed. But the best option for a complete software
   ^^^
How would you know that? Have you tested the alternatives?

Sorry, Maximiliano, for my sarcastic tone. Look -- I'm somewhat
pragmatic,
and don't mind someone recommending a non-free alternative *with a
good
reason* [1] (why is MyFafouriteNonFreeThingie better than
YourNiceFree?).

We could then all learn from it and perhaps help make a free
alternative
better.

And here, *we all care* about freedom: perhaps our approaches and
degrees
of commitment are diverse.

- - - - - - - - - - - - - -

[1] It's with good intentions, anyway: helping the original poster.

regards
- -- tomás
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.12 (GNU/Linux)

iEYEARECAAYFAlZAifgACgkQBcgs9XrR2kYOAgCfbV8ghGD8Y/LtdsShDgnVNskt
MyUAn0ap7hDxy2XLxscxBYWmjQ13uCdE
=sl2t
-END PGP SIGNATURE-

Just for the record: I find all responses helpful, especially sarcastic. )

Piyavkin

Re: An UML editor for Linux


On 09.11.2015 23:01, David Christensen wrote:

On 11/09/2015 11:04 AM, Piyavkin wrote:

**Umbrello**

Thanks for the hint. I'll try it to see better.
The problem is that it is from the KDE SDK. I'm on the Gnome. Though KDE
tools are good looking, I'm trying to use Gnome or independent
equivalents first.


I used Gnome 2 on Squeeze, and now Xfce on Wheezy.  apt-get was able 
to install Umbrello just fine on at least one of them:


I know, I just don't want to muddle Gnome with KDE. I'm sure it will be 
installed in bundle with many dependencies.
I'm going to start LiveCD with KDE (or some VM) and play with it latter 
first.



**QCAD**


Thanks for the hint. May be useful for me (aside from UML). It seems
that there is LibreCAD instead in Debian repositories.


On Wheezy:

2015-11-09 11:53:19 dpchrist@t2250 ~
$ apt-cache search qcad
qcad - Transitional package for QCad to LibreCAD


Can you recommend any proper manual for it?


Perhaps here:

http://qcad.org/



OK, thanks!

Piyavkin

Re: An UML editor for Linux


On 09.11.2015 15:13, Maximiliano Sebastián Castro wrote:



>> — could you please recommend me some decent UML editor for Linux?

If you don't care about freedom, try Enterprise Architect. Runs via 
Wine and in the official page there is a tuto about installing under 
Linux.


Yes, I know it's closed. But the best option for a complete software 
architecture design.



Do you mean ARIS product from SoftwareAG or some other E. Architect?


Piyavkin

Re: An UML editor for Linux


On 09.11.2015 10:40, Johann Spies wrote:



— could you please recommend me some decent UML editor for Linux?


Dia is decent enough. I have used it in the past.

Recently (through the results of postgresql_autodoc) I started using 
Graphviz which produces much beter results than what I can achieve 
with an UML-editor.


Learn the dot-language (part of graphviz).  It is not that complicated.

Regards
Johann
--
Because experiencing your loyal love is better than life itself,
my lips will praise you.  (Psalm 63:3)


**Dia**

Yes Dia seems promising (at first glance). But I have there some little 
but annoying troubles from the start.
For example, I can not resize elements at all, though they have proper 
markers.
Word wrapping in the graphical elements (for example in class nodes) 
doesn't work for me (or I am just too stupid to figure out how to turn 
it on), so boxes with long names looks pretty ugly in result.


**Graphviz**

Well, actually I've used graphviz and dot previously (though it's been a 
while). But before you've mentioned it I had no thought to do with it 
something UMLish.

Now I see that it is a really good idea. I should try it. Thanks!

Piyavkin

An UML editor for Linux

2015-11-08 Thread Piyavkin


Gentlemen,


and kind women,


— could you please recommend me some decent UML editor for Linux?

Yes, I can see in repositories some, — like dia, umlet, gaphor… — but 
which is decent?



Best regards,
Piyavkin

Re: Reporting Bug

2015-10-23 Thread Piyavkin


On 23.10.2015 14:00, Brian wrote:

On Fri 23 Oct 2015 at 00:03:02 +0300, Piyavkin wrote:


On 22.10.2015 17:03, Brian wrote:

On Thu 22 Oct 2015 at 16:15:27 +0300, Piyavkin wrote:


I agree with the topic starter: the Debian's bug reporting process is
horrible and not userfriendly from user's viewpoint. If an organization
cares about usability and effectiveness of bug reporting process, it
shouldn't be like that.

If a user cares about the usability and effectiveness of the bug
reporting process, more detail to back up the complaint would be
welcome.


[...]


As it could look in my view. Bug tracking system is a web-service with such
reporting interfaces as application (~thick client), web-page, mail.

/Web-page/

Web-page is most convenient and most reliable way to report for user. On the
mentioned above web page instead of wall of text there is placed a huge red
button «report a bug» and few links below to pages with short descriptions
of alternative ways (app, mail) and «more details» in those.

There is a standard form to fill. User is asked to fill just needed fields,
and fills only those ones that he can, bug reporting system provides hints
and examples straight in the form (or through «more details»). Selections
from drop-down list are presented whenever possible.

The process of reporting can be partitioned in steps: a) common information
(what reportbug collects from user now, plus contact information); b)
description according to template; c) system specific technical information
(«run such command, copy results here»). When every step is finished,
provided data is recorded in database. Even if the process is not fully
completed and there is no sufficient information to solve the problem, there
is still information for analysis on types of bugs, their intensity, etc. In
some cases user can be contacted through provided contact information and
asked to give more details.

This sounds suspiciously like a Bugzilla type process.


In the part of web-interface: kind of. Why not?


The chances of
Debian going that route are vanishingly small. Also see bug #277744:

   https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=277744

The reason for the 'wontfix' is at

   https://lists.debian.org/debian-devel/2007/03/msg00381.html

Any such frontend must not be worse than reportbug at actually
gathering information and providing real contact information for the
submitter. I personally don't think one for submission will ever be as
useful as reportbug, which is why I've never worked on it myself, and
why I've marked #277744 wontfix.

The thread with msg00381 starts at

   https://lists.debian.org/debian-devel/2007/03/msg00362.html

It is worth a read to get some feeling for the issues involved. As far
as I can determine the project was not completed, which might be an
indication of the complexity in realising it or a lack of consensus on
whether a web frontend for reporting bugs to the BTS is needed.

[...]


Thanks for the info. I've read the thread, and mentioned bugs.

March 2007… Well, I'm glad that the topic has been already rised by a 
specialist in usability (~10 years ago), many good points added in 
discussion by others too. I support the initiative from user's 
viewpoint. I see no sound arguments against that either (just can guess 
about motives).


Actually the point is wider then just web-interface. It's about 
usability on the user's side. I have nothing against reportbug kind apps 
as itself. And why web-interface should compete with reportbug app, I 
wonder, if it already coexists with mail-channel (manually sent mail 
reports) and web-page based access to the base (I suppose with some 
elaborated CMS functionality).


Anyway, I see that developers are aware about the need. It is good. And 
I'm hardly able to do something more on that beyond just wistfully 
expressing my personal opinion.


Thanks again.

Piyavkin

Re: Reporting Bug

2015-10-22 Thread Piyavkin


On 22.10.2015 17:03, Brian wrote:

On Thu 22 Oct 2015 at 16:15:27 +0300, Piyavkin wrote:


On 22.10.2015 10:18, Ondřej Grover wrote:

Hello Adrian,

could you please be more specific about a few points?
- what installer ISO did you use
- did you use the text or graphical installer
- what was the error message or what failed or happened exactly that
stopped the installation

I wasn't able to reproduce your problem in VirtualBox using the network
installation ISO 
http://cdimage.debian.org/debian-cd/8.2.0/i386/iso-cd/debian-8.2.0-i386-netinst.iso
and with almost all the default choices with the excpetion of these
prompts
- Full name of new user: "Adrian O'Dell"
- user name: "adrian"

I understand that you may not feel like reading a large page of text if
you are in a hurry or feel you may not understand it. However, the Debian
and other open-source projects keep living thanks to individuals like you
that go the extra mile, devote some time to the cause and are willing to
learn new things. Please consider supporting Debian (and through it all
other projects based on it, including Linux Mint) through your extra
effort in filing bug reports when needed.

Kind regards,
Ondřej Grover

On Thu, Oct 22, 2015 at 8:15 AM, Adrian O'Dell <crimsonm...@gmail.com
<mailto:crimsonm...@gmail.com>> wrote:

There appears to be a bug which has Plagued me for years. Oddly I
don't have the bug in Linux Mint. Did they edit this part of the
installer?

My name contains an apostrophe, which causes the Debian installer
to not create my user account. Long time ago when I tried to seek
help via IRC was told I must have done something wrong. Two days
ago I confirmed through multiple installs that the apostrophe is
the culprit.

This is the only attempt I will make at filing a bug report.
Anyone more familiar with filing a bug report, it would be greatly
appreciated if you would make sure it gets filed properly so it
may be resolved. https://www.debian.org/Bugs/Reporting is an ugly
wall of text which immediately discouraged me from wanting to file
a bug report anymore.

Thanks.

I agree with the topic starter: the Debian's bug reporting process is
horrible and not userfriendly from user's viewpoint. If an organization
cares about usability and effectiveness of bug reporting process, it
shouldn't be like that.

If a user cares about the usability and effectiveness of the bug
reporting process, more detail to back up the complaint would be
welcome.

.

«more details» — that's exactly a good point for a good bug reporting 
system (or similar one).

But before that, in the first place, there should be: «less details».

*Problem*

The purpose of a bug reporting system is to gather from end users and 
nodes all relevant information on issues with the real product behaviour 
to be aware of current situation (what will you do with the information 
then — is another question).
It's better if the process is completely automated (invisible for 
users). If it is not, it should be as simple for user as possible, so as 
many involved users as possible could complete it.


The problem with the Debian bug reporting process is that it is not so 
simple for ordinary user. And if you do it first time or rarely you 
should first investigate how to do it, manually (and unnecessarily) 
gather bits of information from different sources to finish the quest 
and just to send report. It involves a good deal of searching, reading, 
studying and tweaking settings, even if it is simplest case. For 
ordinary user it seems very annoying and frustrating. And even if the 
user's willing to report to help developers, he/she probably gives up 
and remains silent. And in result the information in the bug reporting 
system reflects only the part of the real picture: only the reports from 
users who were skilled or stubborn enough to break through the reporting 
process (and it is skewed bit of picture too, because the machines of 
the skilled users maintained better and differently than machines of 
ordinary/average users).


*Example (case)*

I've got some critical issue with kernel updates which prevented me from 
starting my system.

So the only options I had to work through other PC or to start from LiveCD.

First of all, what should I do to report the issue? I've searched and 
found the mentioned above page https://www.debian.org/Bugs/Reporting 
with instructions.
I was even not sure if I should address my report to the list. And I 
first asked here at debian-user. The participants here suggested that I 
probably should report bug.


I've returned to the web-page. And had to plough through the wall of 
text on the page and its links, to understand what's significant here 
and what's not, and what exactly should I do. I couldn't use the 
«strongly recommended» reportbug program in the situation.


And even if I could, there was no difference. Using hints from 
net-search I'v

Re: Reporting Bug

2015-10-22 Thread Piyavkin


On 22.10.2015 10:18, Ondřej Grover wrote:

Hello Adrian,

could you please be more specific about a few points?
- what installer ISO did you use
- did you use the text or graphical installer
- what was the error message or what failed or happened exactly that 
stopped the installation


I wasn't able to reproduce your problem in VirtualBox using the 
network installation ISO 
http://cdimage.debian.org/debian-cd/8.2.0/i386/iso-cd/debian-8.2.0-i386-netinst.iso 
and with almost all the default choices with the excpetion of these 
prompts

- Full name of new user: "Adrian O'Dell"
- user name: "adrian"

I understand that you may not feel like reading a large page of text 
if you are in a hurry or feel you may not understand it. However, the 
Debian and other open-source projects keep living thanks to 
individuals like you that go the extra mile, devote some time to the 
cause and are willing to learn new things. Please consider supporting 
Debian (and through it all other projects based on it, including Linux 
Mint) through your extra effort in filing bug reports when needed.


Kind regards,
Ondřej Grover

On Thu, Oct 22, 2015 at 8:15 AM, Adrian O'Dell <crimsonm...@gmail.com 
<mailto:crimsonm...@gmail.com>> wrote:


There appears to be a bug which has Plagued me for years. Oddly I
don't have the bug in Linux Mint. Did they edit this part of the
installer?

My name contains an apostrophe, which causes the Debian installer
to not create my user account. Long time ago when I tried to seek
help via IRC was told I must have done something wrong. Two days
ago I confirmed through multiple installs that the apostrophe is
the culprit.

This is the only attempt I will make at filing a bug report.
Anyone more familiar with filing a bug report, it would be greatly
appreciated if you would make sure it gets filed properly so it
may be resolved. https://www.debian.org/Bugs/Reporting is an ugly
wall of text which immediately discouraged me from wanting to file
a bug report anymore.

Thanks.




I agree with the topic starter: the Debian's bug reporting process is 
horrible and not userfriendly from user's viewpoint. If an organization 
cares about usability and effectiveness of bug reporting process, it 
shouldn't be like that.


Piyavkin

Re: Re: Machine freezes after kernel update


I've reported the bug here:

https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=801467
#801467

Thanks for help.

Best regards,
Dmitry Piyavkin

Re: Machine freezes after kernel update


On 10.10.2015 22:06, Miroslav Skoric wrote:

On 10/09/2015 08:45 PM, Piyavkin wrote:



I have exactly the same issue with the same kernel-packages. See here:
https://lists.debian.org/debian-user/2015/10/msg00231.html



Yes, my symptom was identical to what you have described there. 
Interestingly you use GRUB and not LILO, but even your GRUB was also 
not keeping the previous kernel(s) too. Very strange.




And, it seems that cleaning out all previous versions of kernel at 
arrival of new ones is another bug. Probably should be reported separately.
I think it's better to have not so aesthetical long list of kernel 
versions in bootloader menu or tweak the menu a bit to show only few 
recent versions than to have been stuck with fully nonfunctional system 
in the situations as the our's one.

Re: Boot process freezes dead after kernel update


Update:

I've reported the bug here:

https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=801467
#801467

See also discussion there (similar case and possible solutions):
https://lists.debian.org/debian-user/2015/10/msg00175.html

Best regards,
Dmitry Piyavkin

Re: Machine freezes after kernel update