Project:Support desk

Jump to: navigation, search

About this board

Edit description
vde   Welcome to's Support desk, where you can ask MediaWiki questions!

There are also other places where to askCommunication: IRCCommunication#Chat, mailing listsMailing lists, Q&A etc.

Before you post

Post a new question

  1. To help us answer your questions, please always indicate which versions you are using (reported by your wiki's Special:Version page):
    • MediaWiki
    • PHP
    • Database
  2. Please include the URL of your wiki unless you absolutely can't. It's often a lot easier for us to identify the source of the problem if we can look for ourselves.
  3. To start a new thread, click "Start a new topic".
By clicking "Add topic", you agree to our Terms of Use and agree to irrevocably release your text under the CC BY-SA 3.0 License and GFDL
This comment was hidden by (history)
Atef81 (talkcontribs)

@Terryrogers1984, did you find a solution for this problem? I have the same problem, but can't find a solution.

I get this error:

Lua error in Module:Authority_control at line 346: attempt to index field 'wikibase' (a nil value).


This comment was hidden by (history)
This comment was hidden by (history)
Tgr (WMF) (talkcontribs)

Probably needs the Wikibase Client extension.

Reply to "Implementing Authority control Template"
Azot944 (talkcontribs)

Hi, I'm using MobileFronted extension for Mediawiki. The search engine also indexes urls with this parameter "index.php?title=Article&mobileaction=toggle_view_mobile". How to exclude url from search engine with this parameter?

Robots.txt is a good way?

'Disallow: /index.php?'

Reply to "Exclude url from search engine"

Hosting MediaWiki on Google Cloud for use in College Courses

Popcrate (talkcontribs)


I am setting up MediaWiki to get more students familiarized/comfortable with the ability to edit and contribute to Wikipedia & MetaWiki projects. (This includes both coding, and general editing)

Google Cloud:

I just set up: <> on Google Cloud Platform.


Main Question(s):

Has anybody here tried Bitnami's Mediawiki, or hosting on GCP? Any recommendations for setting up mediawiki for use in a college setting?


Here is what is included in the Bitnami stack (I think):

  • Ghostscript 9.05
  • Apache 2.4.25
  • ImageMagick 6.7.5
  • MediaWiki 1.28.1
  • MySQL 5.6.35
  • OpenSSL 1.0.2k
  • PHP 5.6.30
  • phpMyAdmin 4.7.0
  • SQLite
Reply to "Hosting MediaWiki on Google Cloud for use in College Courses"
Uvas magicas (talkcontribs)

he creado mi sitio web wikipedia pero me marca que la url no esta disponible y no me deja editar tranquilamente que tengo que hacer

Reply to "Url del sitio"

Extension PDFInclude (not working on android tablet)

Summary by Chantoune

Update the code of PDFInclude;php to check if we are working on an Android browser

and if is is the case : return an HTML rawElement Iframe after an embedded request to google

$myfile = "$myURL&embedded=true";

Chantoune (talkcontribs)

how to perform an update into PDFInclude.php to view the pdf on the android tablet (because nothing is displayed or download)

for memory I solved this problem in PDFEmbed as explain in this post to display PDF on an android tablet

and it will be perfect if you may help me to perform something similar to return a PDFobject as a HTML:rawElement with pdfInclude

Chantoune (talkcontribs)

I solved my problem

I update the code of pdfInclude.php as following :


  $ua = strtolower($_SERVER['HTTP_USER_AGENT']);

     if(preg_match('/android/i',$ua)) {

      $myfile = "$obj&embedded=true";      

      return Html::rawElement('iframe', [

         'src' => $myfile,

        'height' => $height,

        'width'=> $width,

        'frameborder' => "1" ].   );      

      } else {

         return pdfObject( $obj, $width, $height );


 } else {

        return pdfError( "Path is not valid" );


How to authenticate a user by my own Authentication server?

Lsherwinforone (talkcontribs)

If I want to build local wiki by mediawiki, and authenticate a user by my own Authentication server. What should I do? Did I have to write an Extension such like Extension:GoogleLogin? It's so complicated.


Anomie (talkcontribs)

If your own authentication server is LDAP-based, you might be able to use Extension:LdapAuthentication. There might be extensions for other types of authentication servers too.

Lsherwinforone (talkcontribs)

It's OAuth2.0 base.Is Extension:OATHAuth the right one I should choose? I assume oauth extensions let wiki user to login other website/client.what I need to do quite the contrary.


Anomie (talkcontribs)

OATHAuth is for two-factor OATH authentication, not OAuth. You might look at Extension:OAuthAuthentication; that extension seems to be specific to authenticating against another MediaWiki wiki using Extension:OAuth, but might be a starting point.

Florianschmidtwelzow (talkcontribs)


I think the most important question is: What do you mean with "own authentication server". What technology is it based on, what API does it provide? Like @Anomie already mentioned, there're already a lot of extensions, which provide authentication methods for MediaWiki, so your server may already be supported :)



This comment was hidden by Malyacko (history)
Reply to "How to authenticate a user by my own Authentication server?"
শাহাদাত সায়েম (talkcontribs)

in my wiki site google ocr buttton do not show , how can I solve this problem ? please help . thanks-~~~~

MarkAHershberger (talkcontribs)

I'm not sure why a Google OCR button would show up. I don't even know what one is. The closest relevant thing that I found was a Google Drive feature but it doesn't look like you're talking about that.

শাহাদাত সায়েম (talkcontribs)

Please see : But I don't konw how Can I add this. Please help

Malyacko (talkcontribs)

The page that you link to explains how you can add this. If some specific parts are unclear, please explain why so someone else could provide an explanation.

শাহাদাত সায়েম (talkcontribs)

example: Google OCR button . please tell me this code for ocr button . thanks

MarkAHershberger (talkcontribs)

I'm still confused about what you would want such a button to do. Adding an OCR button seems meaningless.

শাহাদাত সায়েম (talkcontribs)

I want to proofing pdf book

MarkAHershberger (talkcontribs)

It sounds like you want to hire a developer to implement this for you. I cannot help you to accomplish what you want here on the support desk.

শাহাদাত সায়েম (talkcontribs)

In my site , When I put on OCR button Then show me this error

"<error> undefined Image URL must begin with ' " .

Can you tell me how can I solve this error ?

MarkAHershberger (talkcontribs)

I'm sorry, but it looks like you need more help than I can give you here.

শাহাদাত সায়েম (talkcontribs)

but where I can submit my help request / complain ? I need this help emergency .

MarkAHershberger (talkcontribs)

The Google OCR button is not part of MediaWiki, so this is the wrong place to complain and ask for help.

I suggest you find someone who knows about the button. If they can't put it into MediaWiki for you, they might be able to help you ask the right questions.

শাহাদাত সায়েম (talkcontribs)

but you know that there have an extension name ProofreadPage . THis is the error of ProofreadPage extension . So , I submit this in here .

MarkAHershberger (talkcontribs)

Are you sure this is an error in ProofreadPage? Maybe you don't understand it.

শাহাদাত সায়েম (talkcontribs)

Yes , I am Sure .. <error> undefined Image URL must begin with ' " .

Malyacko (talkcontribs)

As long as you do not explain why that is related to ProofreadPage and as long as you just re-post what you have already written without providing any requested information, nobody can help here, sorry. I doubt that ProofreadPage is related at all.

Reply to "OCR button"
Mariana duarte100 (talkcontribs)

the first case(Extension:PdfExport) gives me the link print as formate PDF, but when i click on, gives me an error that export as format pdf- error, means doesnt find any convertion pdf and explain that i need to install PrinceXML, DomPdf o HTMLDoc. but i install DomPDF and Html doc and PrinceXML and doesnt work

and the second case with Extension:ElectronPdfService

gives me the link print PDF but when i click on , told me that the service doesnt find the page

TheDJ (talkcontribs)


ElectronPdfService is probably the best way to go forward, but many of these extensions are quite burdensome to set up indeed. You need to also install Electron service and a restbase endpoint..

This is badly documented I admit. The talk page of the extension has a few more people who have attempted to do the same and ran into similar problems. I encourage you to document your experiences.

This comment was hidden by Malyacko (history)
Reply to "PDF"

Warning and Fatal Error after uploading version 1.28

Wschroedter (talkcontribs)

I've upgraded recently to 1.26 without problems.

Now I felt forced to upgrade again to 1.28.

When I start my wiki I get this message:

Warning: require(/kunden/ failed to open stream: No such file or directory in /kunden/ on line 81

Fatal error: require(): Failed opening required '/kunden/' (include_path='/kunden/') in /kunden/ on line 81

What should I do now? (talkcontribs)

Please make sure that you have updated all files of the installation. Also it is important that you extract the new files into a new folder. Any remaining files of the old MediaWiki version may cause errors. More details are at upgrading!

Wschroedter (talkcontribs)

I did it all again and now I see: I'm getting an error message while depackaging mediawiki-1.28.0.tar. It comes while depackaging the file mediawiki-1.28.0\wiki.phtml which says "Can not create symbolic link : Access denied : C:\ local path...\mediawiki-1.28.0\vendor\pear\net_smtp\README.rst"

And pear is the place where the start-up-time error message refers to.

So what to do about this depackaging?

AhmadF.Cheema (talkcontribs)

I don't think this should be a problem.

From what I understand, the MediaWiki package includes a symbolic link "file" which are for Linux operating systems and as such are unsupported on Windows systems. These are like shortcuts and shouldn't be causing any breaking of the software.

MarkAHershberger (talkcontribs)

"Can not create symbolic link" is what you'll see unless you run the extraction (depackaging) as an adminstrator. It won't cause problems because symlinking a README isn't going to affect the functionality.

Are you still getting the "Failed opening required..." error message?

Wschroedter (talkcontribs)

Yes, unfortunately.

I tried to go back to 1.26 but I damaged the backup.

Then I tried to go back to 1.23 but there were too many differences.

So I'm still most interested in getting 1.28 working.

This, that and the other (talkcontribs)

Would it be possible to try using a different software package to extract the TAR file? The actual unzipping (really un-tarring) seems to be where the problem is.

Wschroedter (talkcontribs)

Thank you all!

All problems are solved now!


MarkAHershberger (talkcontribs)

Could you tell us what worked for you so others will know if they run into a similar problem? (talkcontribs)

I have the same problem and I have used the mediawiki-1.28.0.tar.gz which I unpacked in windows first and then placed on my server. Which file to you Wschroedter get to have it to work? (talkcontribs)

regards HenrikKoch (talkcontribs)

I think this stupid symlink should be reported as a bug! See bugreport for how you can do that! It is a useless link and is only causing trouble!

This, that and the other (talkcontribs)

I filed phab:T156491. Please comment there saying which TAR extraction software you used when you encountered this problem.

HKOCH (talkcontribs)

Does anyone have a solution for this? My MediaWiki is still down and I don't know how to solve this after upgrading to 1.28

Warning: require(/volume1/web/MediaWiki/includes/services/ServiceContainer.php): failed to open stream: No such file or directory in /volume1/web/MediaWiki/includes/AutoLoader.php on line 81 Fatal error: require(): Failed opening required '/volume1/web/MediaWiki/includes/services/ServiceContainer.php' (include_path='/volume1/web/MediaWiki/vendor/pear/pear_exception:/volume1/web/MediaWiki/vendor/pear/console_getopt:/volume1/web/MediaWiki/vendor/pear/pear-core-minimal/src:/volume1/web/MediaWiki/vendor/pear/mail_mime:/volume1/web/MediaWiki/vendor/pear/mail_mime-decode:/volume1/web/MediaWiki/vendor/pear/net_socket:/volume1/web/MediaWiki/vendor/pear/net_smtp:/volume1/web/MediaWiki/vendor/pear/mail:.') in /volume1/web/MediaWiki/includes/AutoLoader.php on line 81

HKOCH (talkcontribs)

Found the problem

Somebody had renamed /includes/services to /includes/Services (note the upper case S ) changing this to lower case made the trick!

How and why does this happen in a release´?

/Henrik (talkcontribs)

This issue is a serious problem. This breaks mediawiki on any system that is case-sensitive.

This, that and the other (talkcontribs)

> This issue is a serious problem.

Do you mean the renaming of /includes/services to /includes/Services that HKOCH mentioned? Could you explain exactly what you did that caused this folder to be renamed? As far as I can see this folder has always been lowercase.

Cavila (talkcontribs)
Re: Somebody had renamed /includes/services to /includes/Services (note the upper case S ) changing this to lower case made the trick!

Exactly the same thing happened to me after relocating files from Windows to a *nix server. Other files had been named, too, such as SpecialUserLogin > SpecialUserlogin (see also: The problem did seem to occur in the process of either zipping/tarring or extracting the files. Not sure if there's a connection with symlinks or PEAR dependencies.

Ciencia Al Poder (talkcontribs)

This is probably caused by Windows being case insensitive and people not following the instructions on Manual:Upgrading stating that you shouldn't unpack new files over the old ones!

On Windows, files SpecialUserLogin.php and SpecialUserlogin are the same. If you unzip a new version over the old one, it sees the file already exists and updates the contents without using the new name. It continues to work on windows, until you move those files on a case-sensitive filesystem.

Cavila (talkcontribs)

Thanks. Mmm, after unsuccessful attempts to upgrade MW in the manner prescribed (I remember it was something specific to MW 1.28.0), I resorted to a third-party installer to upgrade MW, which somehow worked (locally), but that's precisely where things must have gone wrong. A veritable mess, but everything should be in order now.

Moving mediawiki from linux to windows using postgres

Dazzer4020 (talkcontribs)

We are trying to move our wiki from an old linux box to a windows PC.

We have installed the same version of mediawiki on the windows PC (1.15) and have copied the localsettings.php file across and pointed it to the database on the linux box and we just get a http 500 error.

Also restored the database onto the windows pc and changed to use that but we get the same issue

Not sure where to go from here

Ciencia Al Poder (talkcontribs)

You first need to get a descriptive error message instead of a plain 500 error.

See Manual:How to debug. Look at error logs of the webserver, and enable error_reporting in PHP

Osnard (talkcontribs)

Does your PHP on Windows have PostGreSQL drivers enabled?

I'd recommend to set

$wgShowExceptionDetails= true;

$wgShowSQLErrors = true;

in your LocalSettings.php file to get an useful error message.

General hint when migrating from Linux to Windows: Linux has a case sensitive file system. Windows does not. Make sure you don't loose any files from your "images/" directory (e.g. "Some_Image.jpg" could be overwritten by "Some_image.jpg"). Also MediaWiki on Windows does not allow non-ascii-chars (Umlauts and other special characters) in file names.

Reply to "Moving mediawiki from linux to windows using postgres"