Hi,
We're currently looking into solutions for accommodating multiple
wikis. We have cross-login figured out with wgSharedDB, but has there
been any work done to share blocked IP lists and spam blacklists
across wikis? We were hoping to find a solution that would a) allow
admins to maintain these lists without leaving their familiar wiki,
while b) having 1 central version that was free from cross-wiki admin
privileges and synchronization problems (i.e. a Spanish user should be
able to add or remove a blacklist site without leaving the Spanish
wiki, but this change would be immediately available to the English
wiki).
Has there been any thought been put into this? Are there any wikis
doing this right now? I have a few basic ideas about possible
solutions, but would like to find out if anyone has spent any time on
this first.
Thanks,
Travis
Salam
When using SVN revision (now: 21464), the installer will simply break if
you use mysql newer than 4.0.** which is 4.1 or 5.. it will give that
error about blob/text cannot have default stuff and so... I thought that
it was on windows only but I tried it on ubuntu fawn(mysql5 php5) and on
dreamhost account (php5 mysql5) and on windows i tried many things :)
tried php6 (installer didn't understand) php5 and mysql 4.0.26 (the only
one worked) mysql 4.1. (didn't) mysql 5.2 flacon (didn't work)...
note that it only don't work on svn, now i installed 1.9.3 and it was
installed without problems...I guess that there is something wrong got
into the sql stuff, so perhaps you can take a look...
I won't bugzilla it as I already did once and it worked for rob ;)
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
brion(a)svn.wikimedia.org schrieb:
> Revision: 21415
> Author: brion
> Date: 2007-04-20 07:44:55 -0700 (Fri, 20 Apr 2007)
>
> Log Message:
> -----------
> Reverting r21000 for the moment (predefined block messages).
> The defaults aren't very good, and I'm not sure I like the implementation.
> May return after some review and tweaking.
Can you give me please a few hints for the implementation/tweaking? What
do you don't like?
> -// These are examples only. They can be translated but should be adjusted via [[MediaWiki:ipbreason-list]] by the local community
> -// *# defines a reason group in the drow down menu
> -// * defines a reason
> -'ipbreason-list' => '
> -*#Block reasons for IP adresses
> -*vandalism
> -*linkspam
> -*#Block reasons for usernames
> -*insults
> -*sockpuppet',
May I ask some users with a better english than me for good predefined
reasons please? Which are especially for ips not upsetting if they are
reading the block message on their screen. Thanks a lot.
Raymond.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.7 (MingW32)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org
iD8DBQFGKPJkp+qUcbEr56YRArX+AKDEhMM/w6HYDV8AZjmL3h3P0sbS4ACg8Zpz
kOapfv3KelkBjBIQigfEpcI=
=gaPt
-----END PGP SIGNATURE-----
Hi All,
There's a very very rough class map for MediaWiki's includes/ directory located at:
http://files.nickj.org/MediaWiki/mediawiki-includes-rough-class-map.png
It's very much a work in progress, but the reason for this email is to ask for help with
any class descriptions.
Classes in green have a description in their comments, so they are happy; classes in
yellow do not have a description, so they are sad. Roughly 30 to 40% of classes do not
have descriptions. You can help turn a sad yellow class into a happy green class, by
giving it a description all of its own.
Basically, if you see a class in yellow, and you go "hey, I know what that is", or by
looking at the code you work out what it is, can you please either add the description
to the class and check it into subversion, or alternatively can you please email me any
descriptions, and I'll add them?
I'm only looking for one or two or three sentences that give a very brief overview of what
the class does, or what its purpose is. If you want to write more, that's fantastic and
certainly welcome, but for the purposes of this exercise, a short top-level description
is all that's being asked.
Something else that would probably help is if people can maybe include an "@addtogroup" tag
in the class comments, to indicate the general conceptual group that class belong to. The
current @addtogroup tags used in includes/ are:
@addtogroup Cache
@addtogroup DifferenceEngine
@addtogroup Language
@addtogroup Metadata
@addtogroup Parser
@addtogroup Search
@addtogroup Skins
@addtogroup SpecialPage
@addtogroup Spyc
@addtogroup Templates
@addtogroup UtfNormal
There should probably be some extra ones, including at least:
@addtogroup Api
@addtogroup Database
(feel free to call out or create any extra ones that you think are missing)
Also, it may be for some classes that they already have descriptions, but they're not in
a format or a location that a documentation system can pick up. The format that should
always work is like this:
-------------------------------------------
/**
* This is a short description of the class. Anything here will appear
* inside the box, including this second sentence and this
* stuff on the third line.
*
* This is a more detailed description. It is optional, and will NOT
* appear inside the brief description box. It is separated from the brief description
* by a blank line. Some examples of stuff that might go here include design ideas,
* usage examples, thoughts for improving efficiency, technical notes, future plans,
* and so forth.
*
* @addtogroup SpecialPage
*/
class someClass {
// ... implementation ...
}
-------------------------------------------
(i.e. use Javadoc comments, right above the class declaration; any extra tags like
"@author", "@see", "@todo", etc, are fine too).
If we can get a few more class descriptions in, then I'll do another run, and
this time try and arrange them according to their "@addtogroup" tag (so that
SpecialPage classes are grouped together, Api classes are grouped together, Parser-related
classes are grouped together, Database classes are grouped together, etc).
That should get the number of lines crossing over other things down, and should
group conceptually related files together.
As per before, any resulting diagrams will be under a free license (e.g. GFDL + GPL),
and uploaded to commons.
-- All the best,
Nick.
connelm wrote:
>
> Wiktionary uses http://en.wiktionary.org/wiki/Template:audio extensively.
> Please do not break it.
>
Actually the widespread use of that template means that you will likely
suffer *less* problems. The template can be updated in one place, which will
automagically update all places where it is used,
See, this is the *up*side of the template system, which some people seem to
have forgotten (present company hopefully excepted).
HTH HAND
--
Phil
--
View this message in context: http://www.nabble.com/Thoughts-on-Image%3A%2C-Media%3A-and-Download%3A-tf35…
Sent from the Wikipedia Developers mailing list archive at Nabble.com.
On 18/04/07, raymond(a)svn.wikimedia.org <raymond(a)svn.wikimedia.org> wrote:
> Revision: 21342
> Author: raymond
> Date: 2007-04-18 02:14:41 -0700 (Wed, 18 Apr 2007)
>
> Log Message:
> -----------
> * (bug 8746) Support for external links
> $wgImageMapAllowExternLink = false per default -> no risk for WMF live sites
If that's going to be kept, then I suggest changing it to be
$wgImageMapAllowExternalLinks, because it reads better and makes more
sense. Abbreviating arbitrarily for the sake of it is a waste of time.
Rob Church
Wikix has been open sourced under the GNU Public License. This program
will read any XML dump from the foundation and create
paralell processes which will download all images for a given Wikimedia
Project. The program also contains full UTF8 detection
and conversion if enabled.
For all the folks wondering how to get the images, download the code,
build it, and run it from linux or Windows and 36 hours
later, Commons, or any Wikimedia project images will be downloaded.
This one is free. The other tools cost money. :-)
Download link:
ftp://www.wikigadugi.org/wiki/MediaWiki/wikix.tar.gz.bz2
Enjoy.
Jeff
