My OS-Version is AIX 7.2.5.3 (oslevel -s shows "7200-05-03-2148") and i try to install a Mediawiki (1.39.1). I used the RPM-packages from the "AIX Toolbox for Open Source Software " and have installed the versions from there: PHP 7.4.33 MariaDB 10.4.26 Apache 2.4.54 Now, Mediawiki complains on its configuration screen about missing PHP components, namely "intl", the internationalization functions. Alas, the only way to get this extension seems to be to compile PHP from source - a process i'd like to avoid because on AIX this is a PITA. So, my question is: is there a known way to install Mediawiki on AIX without having to compile software? I have searched the net and came up blank, save for "Install Mediawiki on AIX ", an article i wrote myself 10 years ago. Although working, these are versions too old to use any more.

I have a LAMP environment with /var/www/html/x which is a MediaWiki website. I have a few more MediaWiki websites but I'd like to print the version of the x one to terminal. I need to do so I could know what the last version was before I manually update MediaWiki. Inside that dir, there's a file named RELEASE-NOTES-1.32 with the text: > == MediaWiki 1.32 == > > === Changes since MediaWiki 1.32.0-rc.2 === > > MORE_TEXT........... Maybe I should just print line 3 as with awk 'NR>3' /var/www/html/x/RELEASE-NOTES-* but maybe there's a better way to know the full version of a given MediaWiki installation. What would be the best, most stable way to do this outputting?

One can download MediaWiki via a web browser GUI here: https://www.mediawiki.org/wiki/Download The download link there isn't version agnostic (it changes any time a new version comes out), that makes it impossible to download by the same link because the link always changes. How could I still download the latest MediaWiki in a version agnostic way? How to "target" the dynamic download link from command line? Possibly some regex is needed but possibly not.

I wish to lock everything bureaucratic (all actions, all special pages etc.) to all users without a sysop role in MediaWiki. This is how I currently do it semi automatically. First, I install the extension lockdown . Then, before locking all special pages, I lock all actions because it's much easier than locking all special pages (there are about 6 actions but about 150 special pages). To lock all actions, I put this in LocalSettings.php:

$wgActionLockdown['edit'] = array( 'sysop' );
$wgActionLockdown['delete'] = array( 'sysop' );
$wgActionLockdown['move'] = array('sysop');
$wgActionLockdown['protect'] = array( 'sysop' );
$wgActionLockdown['history'] = array( 'sysop' );
$wgActionLockdown['createtalk'] = array('sysop');

After that, locking all special pages is ha bit hard but currently possible somewhat automatically like this: **1)** A list of all special pages is available at https://en.wikipedia.org/wiki/Help:Special_page **2)** From the shell (CLUI) I download the page to stdin, match the names with grep, filter the name with sort (which also removes duplicates) and redirect the output to a txt file:

curl https://en.wikipedia.org/wiki/Help:Special_page  -s | grep -oP 'Special:\K[a-zA-Z0-9]*' | sort -u > special_page_names.txt

**3)** I Use AWK to put each name in the blocking template of Extension:Lockdown:

awk '{ print "$wgSpecialPageLockdown[\47"$0"\47] = array(\47sysop\47);" }' special_page_names.txt > special_page_names_processed.txt

4) I Paste the output in LocalSettings.php --- Do you know an easier way to lock all special webpages? Please note if you lock anything else besides actions and special pages.

I am having trouble understanding this manual about the MediaWiki CMS. > Please note that under POSIX systems (Linux...), parent of a symbolic path refers to the link source, NOT to the target! You should check the env variable MW_INSTALL_PATH in case the extension is not in the default location. What did the author mean?

So I have been trying to setup a "Free" wiki that integrates with my "Free" version of Active Directory (aka. OpenLdap), so I can get a super simple setup together for a dev group, that needs to work and trade spars about their tribal knowledge. I've seen this in action before, so I know it should work... and everything else has been setup pretty fine till now: - https://kifarunix.com/install-and-setup-openldap-on-centos-8/ - https://kifarunix.com/configure-sssd-for-openldap-authentication-on-centos-8/ - https://kifarunix.com/configure-openldap-sssd-client-on-centos-6-7/ - https://www.mediawiki.org/wiki/Manual:Installation_guide While this wasn't simple, it sure was straight forward, and if I followed the CentOS rules, it all just worked (At least on ProxMox). Now here comes the kicker, I tried to integrate MediaWiki with OpenLdap so my user authentication was seemless: - https://www.mediawiki.org/wiki/Extension:LDAP_Authentication That is where I hit a big snag. The documentation doesn't exactly guide you through how to do this, and many times groups seem to be addressing MSFT Active Directory, not necessarily the open source tools I use. For instance the last most recent thread in the discussion: - https://www.mediawiki.org/wiki/Topic:Soz2fta11ij6frc4 when I follow what was done here, my log doesn't show a 'Failed to bind as' message, instead I get this one 'Did not find a matching user in LDAP':

2020-07-31 14:40:35 master.sinbad.com my_wiki: 2.1.0 Entering getDomain
2020-07-31 14:40:35 master.sinbad.com my_wiki: 2.1.0 Setting domain as: sinbad
2020-07-31 14:40:35 master.sinbad.com my_wiki: 2.1.0 Entering userExistsReal
2020-07-31 14:40:35 master.sinbad.com my_wiki: 2.1.0 Entering Connect
2020-07-31 14:40:35 master.sinbad.com my_wiki: 2.1.0 Entering getDomain
2020-07-31 14:40:35 master.sinbad.com my_wiki: 2.1.0 Using TLS or not using encryption.
2020-07-31 14:40:35 master.sinbad.com my_wiki: 2.1.0 Entering getDomain
2020-07-31 14:40:35 master.sinbad.com my_wiki: 2.1.0 Entering getDomain
2020-07-31 14:40:35 master.sinbad.com my_wiki: 2.1.0 Using non-standard port: 389
2020-07-31 14:40:35 master.sinbad.com my_wiki: 2.1.0 Using servers: ldap://master.sinbad.com:389
2020-07-31 14:40:35 master.sinbad.com my_wiki: 2.1.0 Entering getDomain
2020-07-31 14:40:35 master.sinbad.com my_wiki: 2.1.0 Using TLS
2020-07-31 14:40:35 master.sinbad.com my_wiki: 2.1.0 PHP's LDAP connect method returned true (note, this does not imply it connected to the server).
2020-07-31 14:40:35 master.sinbad.com my_wiki: 2.1.0 Entering getSearchString
2020-07-31 14:40:35 master.sinbad.com my_wiki: 2.1.0 Entering getDomain
2020-07-31 14:40:35 master.sinbad.com my_wiki: 2.1.0 Doing a straight bind
2020-07-31 14:40:35 master.sinbad.com my_wiki: 2.1.0 userdn is: uid=Dsailor,ou=people,dc=master,dc=sinbad,dc=com
2020-07-31 14:40:35 master.sinbad.com my_wiki: 2.1.0 Entering getDomain
2020-07-31 14:40:35 master.sinbad.com my_wiki: 2.1.0 Resource id #454  uid=Dsailor,ou=people,dc=master,dc=sinbad,dc=com
2020-07-31 14:40:35 master.sinbad.com my_wiki: 2.1.0 Did not find a matching user in LDAP
2020-07-31 14:40:35 master.sinbad.com my_wiki: 2.1.0 Entering strict.
2020-07-31 14:40:35 master.sinbad.com my_wiki: 2.1.0 Entering getDomain
2020-07-31 14:40:35 master.sinbad.com my_wiki: 2.1.0 Returning true in strict()

Here is my Ldap configuration:

$wgLDAPDomainNames = array('sinbadcomd');
$wgLDAPServerNames = array('sinbadcomd' => 'master.sinbad.com');
$wgLDAPSearchAttributes = array('sinbadcomd' => 'uid');
$wgLDAPSearchStrings =  array('sinbadcomd' => 'uid=USER-NAME,ou=people,dc=master,dc=sinbad,dc=com');
$wgLDAPBaseDNs = array('sinbadcomd' => 'dc=master,dc=sinbad,dc=com');


# To pull e-mail address from LDAP
$wgLDAPPreferences = array('sinbadcomd' => array( 'email' => 'mail'));

$wgLDAPDebug = 3;
$wgLDAPUseLocal = false;
$wgMinimalPasswordLength = 1;
$wgLDAPRetrievePrefs = array( 'EUROPE.LAN' => true );

$wgLDAPEncryptionType = array('sinbadcomd' => 'tls');
//$wgLDAPEncryptionType = array('sinbadcomd' => 'ssl');
//$wgLDAPPort = array('sinbadcomd' => '636');
$wgLDAPPort = array('sinbadcomd' => '389');
$wgLDAPProxyAgent = array('sinbadcomd' => 'cn=proxyhost,ou=system,dc=master,dc=sinbad,dc=com');
$wgLDAPProxyAgentPassword = array('sinbadcomd' => '***');
$wgLDAPLowerCaseUsername = array('sinbadcomd' => true);


/*$wgLDAPUserBaseDNs = array('sinbadcomd' => 'ou=people,dc=master,dc=sinbad,dc=com');
//$wgLDAPUserBaseDNs = [];

# Group based restriction
$wgLDAPGroupUseFullDN = array( 'sinbadcomd' => false );
$wgLDAPGroupObjectclass = array( 'sinbadcomd' => "posixgroup" );
$wgLDAPGroupAttribute = array( 'sinbadcomd '=> "gidNumber" );
$wgLDAPGroupSearchNestedGroups = array( 'sinbadcomd' => false );
$wgLDAPGroupNameAttribute = array( 'sinbadcomd' => "cn" );
*/

Face in hands. Can someone give me any pointers here? BTW: I know I can use much newer and likely more relevant systems like [Wiki JS](https://wiki.js.org/modules#auth) it seems I bought into the something simple that used to work for me slogan, and well, seems I need some help (Can't remember when I used PHP last).

I would like to update our stone-old MediaWiki installation (v1.24) to the latest version, running on Debian 10 (not installed as Debian package, just using the archive from the homepage) but I just can't get the database update to work. Neither the CLI update script nor the web variant. For the CLI based script the problem is that I have multiple wikis in the LocatlSettings.php. For the script to run I have to tell it which case it should use. That should work by running it with "-S wiki.xyz.de", but that gives me "Invalid address". It works with adding a port, like :8080, but then it's just telling me that it's running and that I can stop it with CTRL + C. I can forward that port via SSH to my lokal machine and open localhost:8080 in the browser, but that obviously just tells me that the script has to be run in CLI. I can access the web update script by passing "-S wiki.xyz.de:8080" to php and running it with "sudo -u www-data (the whole directory is owned by www-data:www-data like is the old installation). The web update script seems to be running fine, the problem is the port 8080. I recently installed Wordpress and because the nginx config had some errors I tried running the installation script that same way. It installed but then everytime I would try to open the page it get's redirected to url.de:8080. So that's not practical to run the web update script that way. I don't know what else I could do. The only thing possible left could be an error in the nginx config for the page, that worked with the old MediaWiki but not with newer versions. So I uploaded the config for you, maybe someone sees something: Pastebin

In a quite old-versioned CentOS-Bash "shared server" environment, I ran the following commands: current_date="$(date +%F-%T)" db_user_name_and_db_name="db-username_db-name" war="$HOME/public_html" # Web Application Root mkdir -p "${war}/mediawiki_general_backups" My problem is that If I make a quick export (all tables) of the relevant database from PHPMyAdmin GUI, I get a .sql file of **43MBs**.
But, if I export the same database by a CLUI command such as the following, I get a .sql file of **23MBs**: mysqldump \ -u "${db_user_name_and_db_name}" \ -p "${db_user_name_and_db_name}" \ > "${war}/mediawiki_general_backups/${db_user_name_and_db_name}-${current_date}.sql" I don't know how to explain this difference; do you?

I would like to replace the ancient Mediawiki (version 1.21.2,l PHP version 5.3.3, MySQL 5.1.73) used by my company with a brand new Mediawiki site running the latest version (requires php 7.0.13 or newer) and with brand new content. Naturally, I'd like to leave the old wiki intact until the new wiki surpasses it. The old wiki is on http://XX.XX.XX.XX:800 . I'd like the new wiki to be simultaneously available on http://XX.XX.XX.XX:801 . I've tried to upgrading removing php 5.3 and installing php 7.1 (from this site ) but the old wiki was no longer accessible after the install. It appears that the old mediawiki was incompatible with php 7.1, so I had to roll back to php 5.3.

I have a MediaWiki 1.32.0 website which I want to upgrade, hosted on a CentOS "Shared Server" environment.
It is an all-core website with no added extensions, skins and images (besides logo) To upgrade I need to change generally all files in the website's directory to those inside a directory of a newer version's MediaWiki installment (available inside a tartball) by a general overriding operation. To download a latest MediaWiki tarball containing such directory (as of 13/08/19) one could execute: wget https://releases.wikimedia.org/mediawiki/1.33/mediawiki-1.33.0.tar.gz In my existing website directory, there are these files I already edited and shouldn't override: > 1. LocalSettings.php > 1. robots.txt > 1. .htaccess > 1. example.com.png (logo image) > 1. googlec69e044fede13fdc.html (Google search console verification file) How could I download, and extract all files from directory in tarball to override a my current MediaWiki directory to override all files besides listed exceptions (such as the files listed above)?
I do plan to backup the old directory before changes manually as priming part of the script; adding a condition to continue only of the backup was done might be a nice idea; all of this, aside from having automatic daily backups).

I am getting a "FastCGI sent in stderr: "Unable to open primary script: /var/www/mediawiki/index.php (No such file or directory)"" error when I enter my wiki address in a browser bar. Here is my PHP-FPM www.conf file: [www] user = nginx group = nginx listen = /var/run/php/php-fpm/php-fpm.sock listen.owner = nginx listen.group = nginx listen.mode = 0660 ;chroot = ;chdir = ;env[HOSTNAME] = $HOSTNAME ;env[PATH] = /usr/local/bin:/usr/bin:/bin ;env[TMP] = /tmp ;env[TMPDIR] = /tmp ;env[TEMP] = /tmp php_value[session.save_handler] = files php_value[session.save_path] = /var/lib/php/session php_value[soap.wsdl_cache_dir] = /var/lib/php/wsdlcache php_value[opcache.file_cache] = /var/lib/php/opcache Here is my nginx conf.d file: # HTTP requests will be redirected to HTTPS server { listen 80; listen [::]:80; server_name wiki.example.com; return 301 https://$host$request_uri ; } # HTTPS Configuration server { listen 443 ssl; listen [::]:443; server_name wiki.example.com; root /var/www/mediawiki; index index.php; autoindex off; # SSL Certification Configuration ssl_certificate /etc/letsencrypt/live/wiki.example.com/fullchain.pem; ssl_certificate_key /etc/letsencrypt/live/wiki.example.com/privkey.pem; client_max_body_size 5m; client_body_timeout 60; location / { try_files $uri $uri/ @rewrite; } location @rewrite { rewrite ^/(.*)$ /index.php?title=$1&$args; } location ^~ /maintenance/ { return 403; } #PHP-FPM Configuration NGINX location ~ \.php$ { try_files $uri =404; fastcgi_split_path_info ^(.+\.php)(/.+)$; fastcgi_pass unix:/var/run/php/php-fpm/php-fpm.sock; fastcgi_index index.php; fastcgi_param SCRIPT_FILENAME $document_root$fastcgi_script_name; include fastcgi_params; } location ~* \.(js|css|png|jpg|jpeg|gif|ico)$ { try_files $uri /index.php; expires max; log_not_found off; } location = /_.gif { expires max; empty_gif; } location ^~ ^/(cache|includes|maintenance|languages|serialized|tests|images/deleted)/ { deny all; } location ^~ ^/(bin|docs|extensions|includes|maintenance|mw- config|resources|serialized|tests)/ { internal; } # Security for 'image' directory location ~* images/.*.(html|htm|shtml|php)$ { allow all; types {} default_type text/plain; } # Security for 'image' directory location ^~ /images/ { allow all; try_files $uri /index.php; } } I feel like it is a permissions issue or the php-fpm daemon is looking in a redundant file path or something. I tried passing an absolute path to FPM via the nginx conf.d file by doing: fastcgi_param SCRIPT_FILENAME /var/www/mediawiki/index.php; to no avail. So I know I'm pointing it in the right direction but it still gives me the same error which makes me believe I have a permissions issue. I've also tried: setenforce 0 but this also doesn't work. I've chmod 777 the entire directory up to and including the index.php file. Some background: I wanted to install a wikimedia extension which required a new version of php (7.0+) and I was running 5.4 since it came with the base install of CentOS 7. I wasn't familiar with how to update PHP so I accidentally yum remove php*, installed php73 from remi, removed that, re-installed php 5.4, and finally figured out I could yum update with remi-php71.repo enabled to update my base packages. However, I lost my .conf and php.ini files in this process. Edit: /var/log/nginx/error.log when I go to my website in a browser: 2019/01/22 16:58:19 [error] 10876#0: *1 FastCGI sent in stderr: "Unable to open primary script: /var/www/mediawiki/index.php (No such file or directory)" while reading response header from upstream, client: 10.11.190.1, server: wiki.example.com, request: "GET / HTTP/1.1", upstream: "fastcgi://unix:/var/run/php/php-fpm/php-fpm.sock:", host: "wiki.example.com" /var/log/php-fpm/www-access.log: - 22/Jan/2019:16:58:19 -0700 "GET /index.php" 404 /var/www/mediawiki/index.php /var/www/mediawiki /index.php /index.php

My main goal is to easily backup the page source for a bunch of MediaWiki pages of mine. I have every page I want downloaded, listed on my username page (https://wiki.example.com/index.php?title=User:username) and currently I can: wget -r --level=1 https://wiki.example.com/index.php?title=User:username This downloads MediaWiki's output HTML for each page I linked, but I'd prefer the source of each instead. With Mediawiki, we can add action=raw to the GET request, like https://wiki.example.com/index.php?title=User:username&action=raw will get the source of User:username. As such, I'd like wget's -r to append "&action=raw" to the URL of each other link it finds and downloads, but I'm not sure how. Other posts talk about doing it like wget -A '&action=raw' however this Accept flag doesn't seem to do the job.

MediaWiki: How to totally delete a user after creating it (just for practice)? I created a user myself in a one-man-editing MediaWiki and I want to remove it.

I use latest MediaWiki (either on Arch Linux or Debian stable) in a personal project - I'm the only one who works on it, and I use MediaWiki in a core-only way (I don't install any extensions whatsoever). * I'd like the core version to always be upgraded to enjoy all newest features. * I have daily/weekly backups. * I tried to read here but I didn't recognize anything about **automatic** upgrades. What is the correct way to have automatic upgrades for MediaWiki? BTW, This is how I upgrade all my Drupal apps in a given Debian-based environment; maybe a similar approach can be taken for MediaWiki in Arch Linux: #!/bin/bash cat /etc/cron.daily/cron_daily #!/bin/bash for dir in ${drt}/*/; do if pushd "$dir"; then rws composer update drupal/* webflo/drupal-core-require-dev --with-dependencies drush updatedb drush cache:rebuild rws popd fi done 2> $HOME/myErrors EOF cat /etc/cron.weekly/cron_weekly #!/bin/bash find "$drt" -path "*/cache/*" -type f -delete certbot renew -q EOF chmod +x /etc/cron{.daily,.weekly}

When Installing latest stable MediaWiki on A CentOS machine I created a user and a password with my email as part of the LocalSettings.php file that the installer creates. After the installation wizard finished I downloaded the LocalSettings.php file and uploaded it to the site's directory. I then navigated to MediaWiki site in my browser by the website's domain. * When I try to login I get an error message that the credentials for my user are wrong. * When I try to retrieve password to my email my username isn't recognized. * I also tried a capitalized version. Why is it?

Does anybody know which project wiki is that which uses mediawiki and many a times is useful for Debian users ? It is a debian derivative but I'm not remembering either the link or project name. Debian uses moinmoin wiki while the other project uses mediawiki and more often than not is up-to-date as mediawiki offers sectional edits which moinmoin lacks.

I was asking myself how to download * all (.svg) images * from a single wikipedia article without doing it manually for each single image. Wikipedia is offering images in different resolutions and several formats using a more complex looking url-scheme Example url: > https://en.wikipedia.org/wiki/Thirty-six_Views_of_Mount_Fuji

My mediawiki install in httpd server for more than two years，today I get the following error when i restart httpd Unable to verify certificate 'Server-Cert'. Add "NSSEnforceValidCerts off" to nss.conf so the server can start until the problem can be resolved. I solve this by add NSSEnforceValidCerts off in nss.conf in the conf.d folder.But httpd still won't start with following error: [warn] module nss_module is already loaded, skipping [warn] _default_ VirtualHost overlap on port 8443, the first has precedence (98)Address already in use: make_sock: could not bind to address 8443 how to solve it? thanks

I was trying to build a wiki on centos server and I downloaded PHP7 from its official website. When I wanted to remove it, I found that even if I deleted all the PHP7 files(I install it using a tar.gz), I can still get this: [root@xxx ~]# php -v PHP 7.1.0 (cli) (built: Jan 15 2017 19:41:25) ( NTS ) Copyright (c) 1997-2016 The PHP Group Zend Engine v3.1.0-dev, Copyright (c) 1998-2016 Zend Technologies After I installed the PHP5 using yum I found that the version of PHP was still 7. I tried to uninstalled PHP5 and after that, the situation was same. Can anybody help me to remove PHP7 totally?

I had this problem with MediaWiki after updating apache At first, i installed using yum and i had apache 2.4.6, php 5.4 on my Centos7 Then i want to update them to latest version which is apache 2.4.23 and PHP 7.0 Until now, the only way to update apache is using source code. I removed the 2.4.6 and install new 2.4.23 Btw, the old DocumentRoot is /var/www/html After updating, the new DocumentRoot is /usr/local/apache2/htdocs I copy all the MediaWiki file from old documentroot to new documentroot. But when i access the server's IP, i just saw the index of/ and list of MediaWiki file. How should i fix this?