Upgrading Linux Mint 17.3 KDE edition to 18.2 KDE in situ

Officially, upgrading the KDE version of Linux Mint 17.3 to 18 is not supported, because “Linux Mint 18 KDE will use a new and different desktop called Plasma”. I presume this means that the KDE4 to KDE5 upgrade process is not trivial. The official way to upgrade is to do a clean reinstall.

I use Linux Mint on my main (scientific computing) workstation, and would much rather not do a clean reinstall: following advice, my /home/ is not on a separate partition, I have several different (non-physical!) users installed, and the official installers just want to wipe everything and start again. I’d really much rather not.

What I’d like to do is update all of the apt sources to point at the right place, do an apt-get update && apt-get dist-upgrade and go home for the evening. As it stands at the moment, quite a few things randomly break every time I do an apt-get update (at the last one, nvidia drivers, cuda, and xfce due to a change in dbus!). I like the look of Plasma, and the live usb of 18.2 runs great on my computer. I want to upgrade without wiping everything. This is far from an unreasonable request.

I have’t heard of anyone else doing this successfully for KDE Mint 17.3, and I’d like to know if it’s possible. mintupgrade certainly thinks not. I’d like to try.

What happens if you just do something stupid, like sed -i 's/trusty/rosa/' /etc/apt/sources.list?

How does Linux with pre-installed binaries comply with binary redistribution license clause?

If I download Linux that comes with a pre-installed program in a binary form licensed under, let’s say, BSD 2-clause, how does it comply with the following:

Redistributions in binary form must reproduce the above copyright notice, this list of conditions and the following disclaimer in the documentation and/or other materials provided with the distribution.

Where can I find licenses of installed binaries? Is there a common practice for this.

I’m asking this because I plan to distribute my program that uses some of BSD 2-clause libraries in a binary form via docker images. Should I compile LICENSES.md file with licenses of all libs my program uses and put it to, I don’t know, the root directory of docker image?

Yii2 разные пути при подключение стилей в OpenServer (Windows) и LAMP (Linux)

Приходится работать в разных ОС потому работаю через github, плюс еще тестовый сервер в сети. Не знаю как решить вопрос с синхронизацией.


class LoginAsset extends AssetBundle
    public $basePath = '@webroot';
    public $baseUrl = '@web/web';
    public $css = [


class LoginAsset extends AssetBundle
    public $basePath = '@webroot';
    public $baseUrl = '@web';
    public $css = [

Естественно картинки в лэйаутах тоже надо править в зависимости от сервера.

Issues accessing Windows SMB share from Linux/CIFS, over OpenVPN

Goal: to access a Windows 10 SMB file share, remotely, from Linux clients, over OpenVPN.

Unsure if my current problem is a VPN issue, or a SMB/CIFS issue.

I have an newly operational OpenVPN server, tap/udp/ip4, running on the windows box. Linux clients connect successfully, receive an IP address, can ping the server, and a nmap -PN shows port 135/tcp is open.

But when I try to mount the share with:
sudo mount -t cifs -0 username=myuser,pass=mypass // /mnt/share

I get error message:
mount: /mnt/share: mount(2) system call failed: Connection refused.

Is there a way to get diagnostic information from CIFS, so I can see exactly why the connection is being refused?

Notes/things tried :

  1. is servers normal ip, and is servers VPN ip
  2. “Windows File Sharing” is enabled on the windows tap driver
  3. Windows Firewall is turned off for private & public networks.
  4. in order to reduce the chance of it being a permission issue, temporarily set security for the share and folder to full access for myuser, anonymous login, everyone, & guest
  5. I note that on the windows box, in explorer, if I enter \ or \ then the windows local shares are visible, as expected. But, if I enter \ it times out, error, the shares are not visible…..I feel this may be a clue? The file share on the windows box, is not accessible from itself, when using the VPN IP address…


dev tap
proto udp
remote mydomain.com 1194
resolv-retry infinite
remote-cert-tls server
key-direction 1
cipher AES-256-CBC
verb 3


port 1194
proto udp
dev tap
ifconfig-pool-persist ipp.txt
push “redirect-gateway def1 bypass-dhcp”
push “dhcp-option DNS”
keepalive 10 120
cipher AES-256-CBC
status openvpn-status.log
verb 3
explicit-exit-notify 1

Que herramientas necesito para crear drivers para windows o linux?

Buen día, quisiera saber como se programan los drivers ya sea para linux o windows, tengo entendido que debo conocer la arquitectura del hardware externo y las instrucciones del procesador de la computadora, pero quiero saber que mas necesito y que herramientas necesito para programar un driver.

Grep /var/log/maillog for email to a certain user, based only on his linux username

Level 1, beginner here. Question about grep and maillog.

I have a learning environment, based on Linux CentOS, with Postfix and SquirrelMail running, but my assignment is more in general.

I need to find in the maillog e-mails received by a certain user within a certain time frame, based only on his Linux username.

I see my maillog, but I am not experienced in reading maillog and I have two concerns:

  1. Whether or not these patterns that I see in the log are something reliable, i.e. whether a log for incoming e-mail will always have to= in it.

    Jan 2 20:31:17 tmcent01 postfix/local[27450]: B58C4330038: to=, orig_to=, relay=local, delay=9.7, delays=9.6/0.03/0/0.02, dsn=2.0.0, status=sent (delivered to mailbox)
  2. How does a Linux username correspond to the e-mail name of the user? It is not it always a match (username@domain), is it? We could have alias for it, how can I take this in consideration when composing the Regex for the grep?

My first two attempts were a strike-out.

sudo grep "to=

I would appreciate any help or hints how can I get the proper grep together.

Can a page fault handler be interrupted in Linux?

I am modifying the function like do_anonymous_page in memory.c in the Linux kernel that handles page faults for anonymous memory. I have a data structure that will be updated based on the pages that will be allocated. However, I would need a lock on this data structure so only one process can modify it at a time. I was reading about Linux spinlocks (https://www.kernel.org/doc/Documentation/locking/spinlocks.txt) and they suggest a primitive version to be used in interrupt handlers and a “spin_lock” version that can be used in non-interrupt handlers. My question is whether I can use spin_lock version instead of the primitive version in a page fault handler like do_anonymous_page.