I was prompted to update Erlang on my machine and I did. Then when I restarted my server I was prompted to update using mix local.hex.
When i did this it gave me the following error.
mix local.hex --force didn't work either.
I eventually found out that I installed erlang originally without brew originally which is where the prompt came from. After I updated it, it caused an issue with mix local.hex.
I manually removed the folder /usr/local/lib/erlang and all erlang on my machine. I installed erlang using 'brew install erlang' and then ran 'brew link erlang'
This originally didn't work because I didn't have permission but then I ran the following to get permission.
sudo chown -R $(whoami) /usr/local/lib
After that 'brew link erlang' worked and I was able to run 'mix local.hex' like normal.
Related
I've installed direnv (v2.18.2) onto my Ubuntu 16.04 machine using:
sudo snap install direnv
as per the website, and added the line:
eval "$(direnv hook bash)"
to my ~/.bashrc file as per the instructions. When I navigate into the directory with my .envrc file, the following message shown:
direnv: error .envrc is blocked. Run `direnv allow` to approve its content.
Sweet. So I run direnv allow, and I'm immediately hit with exactly the same error. I've also tried using direnv allow . but that doesn't seem to help. Also, completely restarting my laptop hasn't helped either.
All the advice I've seen is for direnv not finding the .envrc file, but here it is finding it, it's just not allowing me to allow it.
I know this is not a propper solution, but I encountered this aftering installing from a snap on Linux Mint.
After I uninstalled the snap and installed it from aptitude I did not have any issues.
While the OP is on Ubuntu I ran into the same problem with the snap installed binary on CentOS 7.7.
I worked around the problem by installing a go binary and then building direnv from source: git clone https://github.com/direnv/direnv.git; cd direnv; make; make install which got me direnv 2.21.2 in /usr/local/bin
Since this morning, on my iMac (macOS High Sierra Version 10.13.6) any time I open an iTerm2 window (with Oh My Zsh installed), I see this:
It seems that it's trying to run xcodebuild and it's stuck. I never get a prompt, even after waiting a long time. I've tried pressing ctrl+C to end the process, and then I see this:
I'm not purposely trying to do anything with Xcode. I just want to use the command line. I've been working for months without this problem, and it just started today. Any suggestion for how I can resolve this and use iTerm like normal?
I had the similar issue.
This resolved the problem:
# sudo xcodebuild -license accept
I got the same issue with macOS Big Sur.
Tried all Google results and nothing works. Eventually, I found it was caused by setting Node.js v8.x as default in nvm, alias default to system will resolve the issue. Although I don't know why nvm need to run xcodebuild for Node.js v8.x.
I have got the same issue after updating my MBP16 with Big Sur today
I have checked my xcode command tools path with xcode-select -p and I found that the path was related to an older version of xcode (10.1 in my case, I do not remember the path, maybe /Applications/Xcode_10.1.app/Contents/Developer/).
I have reinstall xcode command tools (sudo rm -rf /Library/Developer/CommandLineTools and xcode-select --install) and I have switched the xcode command tools path with this command sudo xcode-select -switch /Library/Developer/CommandLineTools
The issue is resolved now
Many answers but at the first start the console is still very slow (15s).
Looking around I found out that somehow nvm needs xcodebuild.
The slowdown was fixed changing the nvm default to system
nvm alias default system
Still not well clear why this is happening, I suggest to figure out trying to change the nvm configuration until you obtain the result.
The hanging I was experiencing happened only once after reboot, and consequent launches of the shell happened instantly.
I've spent hours debugging this issue. I eventually traced the hanging to homebrew's start script. My finding was that /usr/bin/git was slow at start (took 12 secs to load) for some reason and Homebrew used that binary unless another was installed. I installed a newer version of git through Homebrew using brew install git and this fixed the issue.
Investigation
I could see that the hang causes 100% CPU usage in syspolicyd and I could see a -67062 error at the end of the hang through Console.app (this error means code object is not signed at all). Unfortunately, the error log did not point me to a specific file and I had to add echos inside the brew script to isolate the exact location of the hang. It turned out to be an execution of git. Installing another version of git that would overwrite the default one fixed the problem. I used brew install git since homebrew checks its own directory first for a git binary.
I am not exactly sure how/why the code signature of /usr/bin/git is broken. It wasn't broken when I first installed Homebrew. It was somehow corrupted later on.
When executing ZeroBraneStudioEduPack-1.70-linux.sh i get this error: xdg-desktop-menu: No writable system menu directory found.
It could be related to my desktop environment ? currently i'm using sway with xwayland.
According to this answer, seems to be an xdg bug. You can check if running sudo mkdir /usr/share/desktop-directories/ before running the install command fixes the issue. I'll consider adding it to the install script if it helps.
I got an error.
$ valet install
[DomainException]
Unable to determine linked PHP.
install
'which' command returns:
$ which php
/usr/local/Cellar/php70/7.0.6/bin/php
I used Homebrew to instal php7.1 , So I ran this command in terminal
ln -s /usr/local/opt/php71/bin/php /usr/local/bin/php
and everything is done.
if your PHP version 7.x and valet > 2.0.8 update valet
valet stop
valet uninstall
brew install php
composer global require laravel/valet
then valet install
Homebrew creates (or attempts to create) symlinks in /usr/local/bin.
Start by running the following to attempt to create a link.
brew link php70
It'll give you information as to whether or not it was successful. Occasionally permissions are problematic and you can resolve this with
sudo chown -R `whoami`:admin /usr/local/bin
Then try running the brew link command again.
Alternately brew link may give you instructions, or you can use brew doctor to get a report on any problems which exist with your Homebrew installation.
When it's set up properly after running ls -la /usr/local/bin | grep bin/php$ you should see something similar to
lrwxr-xr-x 1 YOURUSER admin 29 10 May 21:40 php -> /usr/local/Cellar/php70/7.0.6/bin/php
I was going to update Laravel Valet from version 1.* to 2 and ran same error like this
Unable to determine linked PHP
The solution was to run
composer global update
before I ran
valet install
This happens when your brew php gets Updated.
If you do brew install php this will work but will install php 7.2 by default.
If you want php#7.1 you have to do brew install php#7.1 this will install php7.1 and now when you restart valet valet restart you might get an error. Unable to determine linked PHP.
To make this work with php7.1 you have make brew link
brew link php71
you may have this error
Warning: php#7.1 is keg-only and must be linked with
to fix this
brew link --force php71
now valet should work on php7.1
I tried all of the suggestions here—and several from elsewhere—before I hit on the solution. This applies to PHP 7.2 (no guarantee for other versions). If you are using the original Larvel Valet, open the file:
~/.composer/vendor/laravel/valet/cli/Valet/Brew.php
...or, if you're using Valet+
~/.composer/vendor/weprovide/valet-plus/cli/Valet/Brew.php
Locate the supportedPhpVersions method. In some versions, it will return an array right in the method:
return ['php72', 'php71', 'php70', 'php56'];
Change this to:
return ['7.2', 'php72', 'php71', 'php70', 'php56'];
In other cases, it'll reference const SUPPORTED_PHP_VERSIONS, and the change will be similar.
Save the file, restart valet, and Bob's your uncle!
I'm trying to connect to SQL Server on Ubuntu 9.04 using Ruby. I translated and followed all the steps outlined in getting OSX talking to SQL Server from here:
http://toolmantim.com/articles/getting_rails_talking_to_sqlserver_on_osx_via_odbc
Everything is working on the FreeTDS and unixODBC end. I can see and query the database using tsql.
When I try to access the database from Ruby using IRB I get the following error:
DBI::DatabaseError : INTERN (0) [RubyODBC] Cannot allocate SQLHENV
Has anyone run into this and what can I do to solve this?
I started getting this error when I upgraded to Ubuntu 9.10 (Karmic Koala). Your tip regarding installation order of the Ubuntu packages didn't work for me.
It seems the fix was to manually compile ruby-odbc.
wget http://www.ch-werner.de/rubyodbc/ruby-odbc-0.9997.tar.gz
tar xzvf ruby-odbc-0.9997.tar.gz
cd ruby-odbc-0.9997
ruby extconf.rb --with-dlopen
make
sudo make install
System
Ubuntu 9.10 64 bit
I had to specify the odbc directory in the rubyodbc install
wget http://www.ch-werner.de/rubyodbc/ruby-odbc-0.9997.tar.gz
tar xzvf ruby-odbc-0.9997.tar.gz
cd ruby-odbc-0.9997
ruby extconf.rb --with-odbc-dir=/usr/lib/odbc --disable-dlopen
make
sudo make install
I had the same problem.
But on Centos 5.5 not Ubuntu
Tried many forums/solutions with no joy.
The error message hints at a missing reference to unixODBC.
Which was setting using LD_LIBRARY_PATH variable.
Found another way to set path, by creating
/etc/ld.so.conf.d/odbc.conf.
add unixODBC location to file ie /usr/local/lib.
Run, sudo ldconfig.
Go fig that I actually got this working after submitting my question. What I ended up doing was uninstall libdbd-odbc-ruby and libdbi-ruby and then reinstalling them by installing libdbi-ruby first and then installing libdbd-odbc-ruby. I guess when I installed them before, something must of messed up.
BTW, following the instructions to recompile Ruby-ODBC on Ubuntu 9.10 (Karmic) required installation of either the libiodbc2-dev or the unixodbc-dev package. When using libiodbc2-dev, I got segmentation faults when my Ruby program tried:
connection.select_all('select top 15 * from log_device_healths')
..but no problem when using unixodbc-dev instead.
Tim Morgan's solution didn't work for me. However I was able to get things working by installing an older version of libodbc-ruby (0.9995) from here:
http://mirrors.kernel.org/ubuntu/pool/universe/libo/libodbc-ruby/libodbc-ruby1.8_0.9995-1_i386.deb
Additional details are available from Carsten Gehling's blog:
http://gehling.dk/2010/02/the-woes-of-libodbc-ruby1-8-and-debian-ubuntu/
Be careful though -- Ubuntu's Update Manager will happily "upgrade" this version of libodbc-ruby to the broken 0.9997-2. I accidentally overwrote the older version this way only to end up back here, trying to figure out how I fixed it last time.
Well, it seems my other answer stopped working for me. This thread helped me to solve the issue in another way, and I wanted to share it here.
sudo gem uninstall ruby-odbc
sudo rm /usr/local/lib/site_ruby/1.8/x86_64-linux/odbc.so
cd /tmp
wget http://mirrors.kernel.org/ubuntu/pool/universe/libo/libodbc-ruby/libodbc-ruby1.8_0.9995-1_amd64.deb
sudo dpkg -i libodbc-ruby1.8_0.9995-1_amd64.deb
If you're not on a 64-bit platform, you'll need to download a different Debian package.
Basically, what solves the problem is installing version 0.9995 of the ruby-odbc Ubuntu package.