New answers tagged

0

If you suspect that the program you're installing may be malicious, don't execute it as root or as your usual user. This includes both executing the compiled program and running the supplied build scripts. Once an attacker has run code as you, they own your account (and they'll own your machine as soon as you gain root from that account, at the latest). It ...


2

I did this exact thing, and it is quite easy. I followed the guide for Arch and it works as described. You should follow the instructions for your own distribution of they are provided. Just a few caveats to keep in mind: lowest common denominator hardware: if you want it to run in both 32 and 64 bit computers, install the 32 bit version. while it will ...


0

Modern Linux distributions sport a very minimal kernel, and complete the drivers needed for working by loading modules from a tailored initramfs, loaded with the kernel into RAM by the bootloader. Unless the drivers match, the external drive probably won't boot correctly on the other machine. On the other hand, today's disks drivers are mostly the same, and ...


1

If you want to learn how Linux works, then you would typically install it from source with something like ./configure make sudo make install If you want to do it the Arch way, make a PKGBUILD and share it on the AUR, so others don't have to go through the hassle of installing it from source. Your PKGBUILD will contain the same steps as if you install it ...


0

As user joe, ./configure make sudo make install You don't usually need to be root to configure or compile the code, only to install it. If you don't need to be root, then don't be root. Both configure and make are executing code. Running them as root when you don't need to just introduces another attack vector (someone might be able to compromise the ...


1

If you overuse running programs as root, then you give a potential attacker more spots to attack you than just the "necessary" ones. Attempts to compromise your system mostly won't work (or will be more difficult) without root privilege. Your approach with all commands as root can also lead to simply doing everything as root (listing a directory, opening an ...


2

According to /usr/share/doc/hashdeep/README.md.gz, it's all one executable that acts differently depending on the name of the called program. If the program is called md5deep, it acts like md5deep. I don't use it myself, but if I'm reading the docs right, you should be able to create a symlink to it that will produce the behavior you expect. Do the ...


0

Well, there are a lot of ruby version managers, and tools such as rvm, chruby, rbenv and so on. You may check the whole list of them at https://www.ruby-toolbox.com/categories/ruby_version_management


1

NTP 4.2.8p2 is included if you upgrade to Solaris 11.3, the currently supported version of Solaris 11.


1

Install the full xorg, plus virtualbox support (you might have to reboot after that): pacman -S xorg virtualbox-guest-utils If you really want to use startx you need either to change your ~/.xinitrc to start blackbox or install ugly old twm instead.


2

The other answers on here are great, but I wanted to create a completely automated installation script. I hope this helps someone. #!/bin/bash sudo apt-get install -y cmake yasm libjpeg-dev libsdl-image1.2-dev libsdl1.2-dev pushd /tmp wget -O libpng-1.6.21.tar.xz ...


0

Provides: tags have nothing to do with the ea-apache24-2.4.18-1.1.x86_64 conflicts webserver problem. The package ea-apache24 has a Conflicts: webserver tag which causes this. Which I think is not really clever, as by default nearly all webservers can be installed in parallel. And not sure why you installed apache from some external repo when you can have ...


0

In your situation, I'd build nginx from source. This has the advantage that you can enable only the features you'll need to run as a proxy. There are disadvantages. Keeping up with security issues becomes your responsibility, so you'll have to keep track of nginx errata. Also, you'll have to hand-roll init/systemd start scripts (although you can rip ...


0

It is absolutely not possible to override that. As a note, never attempt --skip-broken. The problem is both ea-apache24 and nginx both "provide" the same thing. Because of this, nginx refuses to install. There is no way around this without doing an rpm --force. But there is a good chance that will also fail. Doing this will create an rpmdb hell. I would ...


0

I was just looking for the same thing and this question is the only thing related I found. Your problem may be solved with LinuxLive_USB_Creator (read as well: https://en.wikipedia.org/wiki/LinuxLive_USB_Creator) Download can be found here: http://www.linuxliveusb.com/en/supported-linuxes You probably have to "install" the livesticks again and this time ...


2

dpkg/apt and rpm/yum use databases that get locked when a package is being installed. This is done on purpose; it is to avoid possible corruption of the databases. The databases are meant for tracking the installed packages on a system.


1

It's fine. Remember, it's your home directory; you can do anything you want in there. GNOME/KDE/whatever might complain if you go around deleting .config, but anything you do in your home directory will, by definition, only affect you. Adding some directories in .local is harmless. Putting something in a dot directory will be mildly inconvenient, ...


0

The common convention (used e.g. with some install-home targets, like for mercurial, known as hg) is to put them directly under $HOME, i.e., in $HOME/bin, $HOME/etc, $HOME/lib, and so on. This is the result of the GNUish configuration dance starting with ./configure --prefix=$HOME.


1

The obvious reasons why not: user's home-directories are often limited in size (quotas on shared systems) predefined profiles often have already added ~/bin to your PATH (making it more convenient to install in that directory) if you have control of the machine, installing into a shareable location works out nicely, e.g., /usr/local/bin. Your ...


0

For *.so search: export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:your_lib_dir For executable search: export PATH=$PATH:your_executable_dir I think this two environment variable is enough.


0

You should do make menuconfig before compile. Find res_rtp_asterisk and check if it need some dependency you not have or just turned off. If it turned off, turn it on and recompile.


1

The reason why you're seeing that error is because the person who maintains the package incorrectly included /usr/share/man and /usr/share/man/man1 in the RPM package's definition of what the package "owns". Because the package claims to own that directory (and will try to delete it if you uninstall it), yum detects that another, already installed package, ...


1

Try installing the previous version: rpm -ivh http://binaries.html-tidy.org/binaries/tidy-5.1.14/tidy-5.1.14-64bit.rpm


1

You need to install the development packages, usually called foo-devel or similar. Or you might be trying to build for the wrong architecture.


1

Yes, it is possible, and a feature Your path may be setup so that your running an older version. This is common if one package is in /usr/bin and the other in /usr/local/bin or /opt/something Try running which pandoc to make sure your running the same one in that is in the deb file. This allows for different users to run different versions of applications ...


0

I decided to try and start the .lnk file directly by opening the DOS Prompt from the wine menu with the com.xilinx.verilog prefix set. After navigating to C:\users\Public\Desktop I ran start "ISE Design Suite 14.7.lnk" and it opened the application window! I'm currently looking for a nicer way to run it, but I'm excited to have it working without needing to ...


0

There is a bug in a Makefile and some man pages for Perl scripts are not generated. Those files ending in .8 are man files. This error seems to occur on CentOS 7 due to the file command version that prints different description for Perl scripts: [root@centos-6.7 xpp]# file xpp_blink xpp_blink: a /usr/bin/perl -w script text executable [root@centos-6.7 ...


0

The errors concerning the cannot stat errors seem only to be related to man pages. /usr/bin/install: cannot stat ‘./dahdi_registration.8’: No such file or directory /usr/bin/install: cannot stat ‘./xpp_sync.8’: No such file or directory /usr/bin/install: cannot stat ‘./lsdahdi.8’: No such file or directory /usr/bin/install: cannot stat ‘./xpp_blink.8’: No ...



Top 50 recent answers are included