roots / bedrock Goto Github PK

I can't get wp-cli to work. I see it's in the vendor file and I want to use it but I keep getting an error. I tried to define the path but it didn't work.

I love this feature:
register_theme_directory(ABSPATH . 'wp-content/themes');

Can this be done with the plugin directory as well? That would be really helpful for me, thanks!

The first plugin often installed in many Wordpress setups is W3 Total Cache.

This plugin creates app/cache/ and app/w3tc-config/ directories outside of the app/plugins/ directory in addition to three other files used by the plugin: advanced-cache.php, db.php, and object-cache.php.

Perhaps adding plugin-specific ignores to Bedrock is a bit much, but maybe a line about this is the readme would be a good idea?

From README.md:

plugins, and mu-plugins are Git ignored by default since Composer manages them

As of now, it seems to suggest that all plugins will be placed into the /app/plugins/ directory and be contained there, kept in check via Composer without any additional configuration of Bedrock's files.

Could it be possible, now or in a future version of Bedrock, to require different plugins in composer.json and composer.lock based on the environment?

For instance, I don't want to deal with W3 Total Cache in my development environment, but of course I want it required on production.

I'm new to deploying with Capistrano, but I found that I had to run bundle update before running the deploy check.

Adding this instruction may help others.

Maybe I did something wrong on my side, but I notice I was able to download my .env file by visiting www.example.com/.env

And I don't think that's a good idea. You should add .env(.example)? to Nginx's deny all. Likewise for Apache.

Enlighten me if I'm doing something wrong.

There is a problem with language support (again). In this setup seting WPLANG to something does not do the trick because WP does not expect language directory in wp/wp-content but in app. If I wanted to work, I must move directory manually after composer install and before wordpress installation. Is there a way to tell composer to do this. For some reason themes works from wp/wp-content but language not.

Hi all,
I'm integrating WP Migrate DB Pro in composer.json following this tutorial:
http://codelight.eu/using-private-wordpress-repositories-with-composer/

This plugin is not free so you have to buy it with a license key, this is what we have done.

Now, this key must be specified in composer.json as you can see in this line:

"url": "https://deliciousbrains.com/?wc-api=delicious-brains&request=download&licence_key=[YOUR_LICENSE]&slug=wp-migrate-db-pro-media-files&site_url=[YOUR_SITE_URL]"

I know that composer can handle vars like {$env:name} so maybe I can place the license key in .env, do you think it would be the right place? and would it work? Thanks.

Meybe I'm wrong, but this way I always need to run composer update after instalation.

Hi guys,

I've just gone to setup Capistrano for deploying and after not having much luck I've reset all cap related files back to default and was still having an issue:

Pauls-iMac:project paul$ cap
cap aborted!
cannot load such file -- capistrano/composer
/Users/paul/Sites/project/Capfile:8:in `<top (required)>'
/Users/paul/.rvm/gems/ruby-2.1.0/gems/capistrano-3.1.0/lib/capistrano/application.rb:24:in `load_rakefile'
/Users/paul/.rvm/gems/ruby-2.1.0/gems/capistrano-3.1.0/lib/capistrano/application.rb:15:in `run'
/Users/paul/.rvm/gems/ruby-2.1.0/gems/capistrano-3.1.0/bin/cap:3:in `<top (required)>'
/Users/paul/.rvm/gems/ruby-2.1.0/bin/cap:23:in `load'
/Users/paul/.rvm/gems/ruby-2.1.0/bin/cap:23:in `<main>'
(See full trace by running task with --trace)

Full trace on pastebin.

This happens if I run any cap command, I have already ran composer update

However after installing capistrano/composer (gem install capistrano-composer) I don't get this.

I haven't used capistrano before but I thought I'd point this out.

I got a lot of 503 errors trying to install capistrano/composer, could it have been a 503 was returned when cap read the require in the capfile?

Mac: 10.9.1
Ruby: 2.1.0 - installed with rvm v 1.25.15

My install of Bedrock refused to honor the mu-plugin autoloader setup. I've added the plugins to the installer paths but they continue to be installed as regular plugins. Additionally, the bedrock-autoloader.php file does not exist. Am I doing something wrong? If so, please clarify.

I'm currently testing the deploy command and I'm having some problem. I thought the script would automatically try to symlink my public_html/ -> site/current/web, but it's not!

When I load my site, I simply get the public_html empty directory, am I supposed to configure a new root myself? If I try to add a symlink it is destroyed every time I deploy again.

Would appreciate some help on this!

Thanks!

Hey, just trying to get my head straight on using Dotenv with WordPress as you've done with Bedrock. Bit confused as to how I'm supposed to be setting staging/production environment variables.

From what I've understood, Dotenv is mainly used for allowing developers to setup their own local configuration by creating a .env file and ignoring it within source control.

That's cool, works well. Only, I noticed that the Dotenv::load method throws an exception when the .env file isn't present. That's fine though, because we can use a try/catch.

However, then we've no environment variables set for our staging and production environments, because they're using getenv to get their configuration, only no .env file will exist on the server because its been ignored. So unless we hardcode the values into the staging.php or production.php configuration files, we're going to have no variables set.

We can do that, but still we haven't told our stack that this is the staging or production environment.

Creating .env.staging and .env.development also (I believe) goes against the concept of dotenv. And even still, we'd have to modify config/application.php to load them.

Am I missed something? I have a feeling there is some Capistrano magic occurring here, but doesn't look like any custom tasks have been used.

I guess it boils down to - how are we defining WP_ENV when our .env files are being ignored. In that case, I believe it's always going to fall back to development.

Thanks in advance! Just trying to get this jigsaw completed in my head! :)

When I do a cap deploy on my Bedrock site the WordPress path to the themes folder does not get updated resulting in a white screen.

The only thing I am able to do is wait for WordPress to fix itself whenever or switch the themes then back again.

I have not modified Bedrock in anyway other than deploy files.

Thanks for this wonderful project which really helped me reach new professional heights ! I started with the Roots base theme, then tried Bedrock with Composer only, then I went into Grunt, then Capistrano, and now I'm even managing my Drupal websites with Capistrano and Drush :)

I'd love to see SQL Capistrano tasks in Bedrock, and I know it's your TODO list, such as

• an automatic SQL backup of the old release each time we deploy (not to be included in rollbacks ; just in case, when WP needs to update its database)
• a manually invoked task : SQL sync from an environment to another, which needs also to update the wp_options table with the correct URLs depending on environments

There's a few features of Bedrock that might make some Multisite setups a problem:

• WP installed in subdirectory
• Custom content folder

Looking at some issues regarding this, it seems like MS support is still a bit of a mess. Related issues:

• https://core.trac.wordpress.org/ticket/27287
• https://core.trac.wordpress.org/ticket/23221
• https://core.trac.wordpress.org/ticket/26969

Here's a potential mu-plugin we may want to consider adding in Bedrock by default: https://gist.github.com/danielbachhuber/9379135

/cc @ddebernardy I'm hoping you can speak to/summarize some of the issues you've ran into and potential solutions.

I am not really sure how the .env or config/enviroment/* files are supposed to work? In which do I specify other enviroments like production and so on? And what enviroment data should go into .env?

I am asking because the themes in app/themes are not working because there are not using the correct site URL and are loading every assets from the root, in my case localhost, but I am using a subfolder.

As I am also not fully able to understand dotENV etc. I just tried to upload my project on my staging server. I change site path in the wordpress database after uploading it and the .env-File but nothing shows up, when I try to view the site.
I just uploaded regularly per FTP.
Did I miss something?

I get this error running bundle exec cap staging deploy:

sudo: no tty present and no askpass program specified

I'm not quite sure what's causing since I am quite sure the settings and configuration was all taken care of.

Any help and pointers?

Since Bedrock is supposed to be a starting point and promote modern web and all that fun stuff, I think it's probably a good idea to keep it up to date with PHP versions.

PHP v5.3 is 3 years old and will no longer be supported in 3 months.

PHP v5.4 is 2 years old and will no longer be supported about a year from now.

I think Bedrock shouldn't be afraid of leaving WordPress behind and setting greater PHP requirements. Same with Roots. Your projects are dragging WordPress into 2014, and they shouldn't hold back on actively promoting modern PHP, especially since these are starting points!

If you want to be really bold, both projects could very easily implement the short array syntax (a la javascript, ['value 1', 'value 2']) to intentionally break on older versions of PHP.

Thoughts?

Hey all-

Was curious, with discussions over at roots/bedrock-ansible talking about introducing SSL support on that side, I wanted to get the conversation started on how to introduce SSL support on this side.

Is this something we'd want to include here and activate that functionality using a constant, or just allow the user to enable all that on their own?

Just an FYI.

http://roots.io/wordpress-stack/

Seems a commit message is showing up in the container next to your photo.

Could you make version based on Heroku PHP Buildpack?

• https://github.com/heroku/heroku-buildpack-php
• https://devcenter.heroku.com/articles/php-support
• https://blog.heroku.com/archives/2014/4/29/introducing_the_new_php_on_heroku

This PHP Buildpack is included in

• https://github.com/deis/deis
• https://github.com/progrium/dokku
• https://github.com/flynn/flynn
• http://www.centurylinklabs.com/top-10-open-source-docker-developer-tools/

I think it would be great if you updated the Wordpress version to 3.9.1.

Any cons?

In order for twentyfourteen-child theme to work in app/themes, we've had to add twentyfourteen itself to app/themes.

It seems there are differences in the paths set by "wpackagist-theme/twentyfourteen", and the instance of the twentyfourteen theme bundled in wordpress itself, and the child theme will only work if we've included the theme in app/themes.

most users which are not familiar with Ruby found it very hard to deal with Capistrano as a deployment
tool so I would suggest using dg/ftp-deployment php package instead I think this package is very powerful and can deploy throw ftp,sftp,ssh and can extend well

What do you thing guys about that ?

Hi,
For some reason unknown, my bedrock install will not update the WP-core to 4.0 even I have modified my composer.json to the new version 4.0. I did a fresh composer install with the composer.json and it worked OK, but when I run composer update in my project directory, the .lock file still keeps on referencing to 3.9.2.

I have tried to clear composer caches, but no luck. Any suggestions what might be causing this?

Here you can find my composer.json and composer.lock files: https://gist.github.com/traone/e0c045e37ec3c9cee03e

I've installed plugins via the WordPress dashboard and manually dragged the plugins from "plugins" to "mu-plugins" in the app directory, however, the mu-plugins do not show up in WordPress (under "Must-Use" on the plugin page).

Can you help me out?

I'd love to see some information about pros / cons of this method for organization and why it was introduced over just leave it the way it was?

Just curious!

I am using shared hosting to deploy my website.

I configured everything but my hosting has an application folder and a subsites folder with folders of each subsite. I cannot link the folders myself so /subsites/MyNewSite is the web folder of my website and I cant change that. /Applications/MyNewSite is the applications folder.

I deployed to /Applications/MyNewSite and rsynced /* web to /subsites/MyNewSite. I also needed /configure / vendors in subsites because of paths in wp-configure, which I needed to change.

Now WP Admin works but the normal website gives me white screen of death and no errors.

Using PHP 5.3.10-1ubuntu3.9 with Suhosin-Patch (cli) (built: Dec 12 2013 04:24:43) the following error occurs on installation. I couldn't find any other ticket with these error messages, so posting here.

# composer create-project roots/bedrock . dev-master
Installing roots/bedrock (dev-master 88a037bc03a06399eed871380d1707f165c26702)
  - Installing roots/bedrock (dev-master master)
    Cloning master

Created project in .
Generate salts and append to .env file? [Y,n]? Y
PHP Fatal error:  Cannot access self:: when no class scope is active in /var/www/ngh2/web/scripts/Bedrock/Installer.php on line 35

Fatal error: Cannot access self:: when no class scope is active in /var/www/ngh2/web/scripts/Bedrock/Installer.php on line 35

Hi,

I am trying to install Bedrock on my VPS running on Centos and Nginx as reverse proxy, Apache, PHP and MySQL. I am however unable to figure out as to whats wrong in my setup.

nginx vhosts file:

server {
          error_log /var/log/nginx/vhost-error_log warn;
          listen xx.xx.xx.xx:80;
          listen [::]:80;
          server_name example.com www.example.com;
          access_log /usr/local/apache/domlogs/example.com-bytes_log bytes_log;
          access_log /usr/local/apache/domlogs/example.com combined;
          root /home/example/public_html/web/;
          #location / {
          location ~*.*\.(3gp|gif|jpg|jpeg|png|ico|wmv|avi|asf|asx|mpg|mpeg|mp4|pls|mp3|mid|wav|swf|flv|html|htm|txt|js|css|exe|zip|tar|rar|gz|tgz|bz2|uha|7z|doc|docx|xls|xlsx|pdf|iso)$ {
          expires 1M;
          try_files $uri @backend;
          }
          location / {
          error_page 405 = @backend;
          add_header X-Cache "HIT from Backend";
          proxy_pass http://xx.xx.xx.xx:8081;
          include proxy.inc;
          include microcache.inc;
          }
          location @backend {
          internal;
          proxy_pass http://xx.xx.xx.xx:8081;
          include proxy.inc;
          include microcache.inc;
          }
          location ~ .*\.(php|jsp|cgi|pl|py)?$ {
          proxy_pass http://xx.xx.xx.xx:8081;
          include proxy.inc;
          include microcache.inc;
          }
          location ~ /\.ht {
          deny all;
          }
        }

web/ is the directory where wp-config.php is placed by default.

I have also set the following values in the .env file:

WP_ENV=production
WP_HOME=http://example.com
WP_SITEURL=http://example.com/wp

With this current setup, I see the directory listing. Can someone point me to the right direction?

Hello everyone, as I am starting to get familiar with Bedrock, I am a little confused about setting the WP URLs in .ENV file.

From ReadMe:

WP_HOME - Full URL to WordPress home (http://example.com)
WP_SITEURL - Full URL to WordPress including subdirectory (http://example.com/wp)

The index.php (with require '/wp/wp-blog-header.php') and wp folder (with "Wordpress stuff) lays under web folder, right? So it should be like this:

WP_HOME - Full URL to WordPress home (http://example.com/web)
WP_SITEURL - Full URL to WordPress including subdirectory (http://example.com/web/wp)

Setting "web" in URLs work for me (does not throw 404 errors on http://example.com/wp/wp-admin and actually displays something on http://example.com) but it's obviously not very pretty.

So... Should I manually edit my .htaccess file? And sorry in advance, guys, I'm sure I'm missing something obvious. :) Thanks, P!

I just upgraded to MAMP v3 and I'm unable to get WP-CLI working. I'm getting this error

Warning: mysql_connect(): No such file or directory in .../web/wp/wp-includes/wp-db.php on line 1372

Took a look at the file and couldn't figure out what went wrong. Thanks again for helping!

Placeholder issue for now but we probably need to improve the default implementation. Look into env vars or maybe env specific ones.

Hello,

I'm wondering what is the correct way to generate salts for other project members?

Currently, when you install using composer create project the salts are generated for the first user (in the .env file). When the user push the project, the env file isn't committed, only the example file. New user clone the project, add his db info / wp urls in the .env.example and rename it to .env, but doesn't have the salts.

What i'm doing currently is have the new user run composer run-script post-root-package-install, regenerating the salts for the user, but I was wondering if there's a better way to do this?

$ cap production deploy
 INFO [e9cd1232] Running /usr/bin/env mkdir -p /tmp/my_app/ on my_app.com
 INFO [e9cd1232] Finished in 0.484 seconds with exit status 0 (successful).
 INFO Uploading /tmp/my_app/git-ssh.sh 100.0%
 INFO [e91dcfcd] Running /usr/bin/env chmod +x /tmp/my_app/git-ssh.sh on my_app.com
 INFO [e91dcfcd] Finished in 0.063 seconds with exit status 0 (successful).

This all looks good, but it doesn't create the directories or anything.

Nice project that you've started!

I'm just wonderig if there's a reason behind not creating a public/ folder?!

Hi there,

I followed the instructions here to integrate WooCommerce into Bedrock and Roots but it seems something is missing on the Bedrock-side.
When viewing my shop sites, I always get the message
Deprecated: preg_replace(): The /e modifier is deprecated, use preg_replace_callback instead in [...]\app\plugins\woocommerce\classes\class-wc-product-factory.php on line 50

Even when I try another theme the same error occurs, so I think something is missing in bedrock for full WooCommerce-support.
Any ideas?

First, thanks for creating Bedrock. It's a great step forward for WordPress developers.

I have configured Capistrano and it's deploying to our remote server. I have the linked resources set up correctly, and I see the assets (current, repo, releases, shared, ..etc) on the server after the deploy. So far so good.

However, Composer install does not run as I expected. I made sure to run Bundler against the Gemfile to get the correct deps and see those installed locally and have Composer installed on the server (I can run composer install on the server manually to get our WP and our other dependencies). But deploy only appears to get the latest source / releases configured on the server and then terminates without errors.

I thought perhaps that I needed to create a task to execute Composer, which I did, and it works just fine. But after re-reading the Bedrock docs, I came across the following in the Capistrano section which led me to believe that writing a task for this may be unnecessary.

"Composer support is built-in so when you run a deploy, composer install is automatically run. Capistrano has a great deploy flow that you can hook into and extend it."

Should Capistrano deploy be running Composer install automatically or were we correct to write a task to run install?

Thanks.

Hello and once again, thanks for an awesome Wordpress starter kit!

For my framework-like customization of Wordpress, I need a slightly modified version of Bedrock. According to Bedrock documentation, that is what you encourage us to do. :)

I created a fork of Bedrock and now I need to use it as a source of my create-project command. If I get the idea of Composer, I have a few options:

A. Submit the forked repository to Packagist. This is not recommended neither by Packagist, not Stack Overflow. I've tried that with quite odd results (which I can explain further, if needed).
B. According to Composer documentation, I can specify "repository-url" parameter as well. composer create-project roots/bedrock --repository-url=https://github.com/roots/bedrock Errors again!

The "https://github.com/roots/bedrock/packages.json" file could not be downloaded

C. Place my composer.json to folder and run composer create-project on it. Tried that as well (most elegant sollution IMHO), but receieved an error. Oddly enough, Composer cant find Wordpress as well.

Class Bedrock\Installer is not autoloadable, can not call post-root-package-install script.

The requested package wordpress/wordpress could not be found in any version, there may be a typo in the package name.

Can you tell me what would you see as the best practice in such situation? Would appreciate a quick response, if that's possible, as I am in an extreme hurry to resolve this. Thank you in advance! Petr!

Hi,

I was in the need of using git submodules with capistrano and after some research I've found this:

https://gist.github.com/stevenscg/8176735
http://jjasghar.github.io/blog/2014/06/04/moving-to-capistrano-version-3-from-version-2/

Is there a plan on adding this to the project? I think it might be useful to others.
Also more people are using the same approach: https://github.com/Mixd/wp-deploy

Many plugins use plugins_url() in their enqueue_scripts and enqueue_styles. The plugins urls seems ok when you var_dump it with a filter like this:

function filter_plugins_url( $url ) {
  var_dump($url);
  return $url;
}
add_filter( 'plugins_url', 'filter_plugins_url' );

Example with WP-SEO:
the var dump results in: http://dev.domain.com/app/plugins/wordpress-seo/css/adminbar.css

But weirdly the HTML source contains this:
<link rel="stylesheet" href="/wp/app/plugins/wordpress-seo/css/adminbar.css?ver=1.5.2.8">

Constants from the wp-config:

WP_HOME "http://dev.domain.com"
DOMAIN_CURRENT_SITE "http://dev.domain.com"
WP_SITEURL "http://dev.domain.com/wp

Conclusion:
WordPress is trying to be smart here and it looks like it makes the already absolute urls relative again but uses WP_SITEURL for this and thus breaking the urls for this plugin assets.

I'm not sure what's happening but in my dev environment this is working fine.

After deploying with capistrano in production it's showing this in Firebug:

wp-tin...0131026 (line 1)
"NetworkError: 404 Not Found - /wp/wp-includes/js/tinymce/langs/en.js?ver=359-20131026"
en.js?...0131026
Failed to load: /wp/wp-includes/js/tinymce/langs/en.js
wp-tin...0131026 (line 1)
"NetworkError: 404 Not Found - /wp/wp-includes/js/tinymce/themes/advanced/langs/en.js?ver=359-20131026"
en.js?...0131026
Failed to load: /wp/wp-includes/js/tinymce/themes/advanced/langs/en.js
wp-tin...0131026 (line 1)

I've disabled all plugins.
I connected to the server to see if the wp-lang-en.php file was there and it is, but not sure why it's not producing the proper js files respective to this? Hmm...?
I've tried using the default Twenty Fourteen theme instead of Roots... still doing it...?

Any ideas?

Is there a specific plan you have for the Vagrant portion? I was wondering, because it would be nice to help contribute to this. My main concern is database handling. I have seen some Vagrant boxes that manage it well.

Tried an install using create-project and it installed WP 3.9. Manual fork/clone installs WP 3.9.1 as expected.

This isn't to say there is anything wrong with Ruby or Capistrano, but why require a Ruby instance when we can keep things lean and mean with PHP by using Laravel Envoy.

Our bedrock website have the WordPress 3.9.1 is available message showing in the dashboard. But the current version is already 3.9.1

Anyone else have this problem too?

This is probably a Capistrano issue, but I figure I'd post it here as well just in case.

When I updated to Capistrano 3.1 to bring in some of the new Bedrock changes, on a deployment that previously worked, I'm getting the following:

DEBUG [d6aa3bae] Running /usr/bin/env cd /home/**/deployments/staging/repo && git rev-parse --short HEAD on **.me DEBUG [d6aa3bae] Command: cd /home/**/deployments/staging/repo && git rev-parse --short HEAD DEBUG [d6aa3bae] stdin: is not a tty DEBUG [d6aa3bae] bash: git: command not found

The second I set the gem to 3.0.1 and bundle update capistrano, it works perfectly again. I have paths set, and it's working on previous git commands, but fails out here every time.