> Don't forget to make a table (or even a new database?)
> for the "real" language links, while you're at it. This
> will be needed for automatic language link building.
> Magnus
I don't know what this means. Please explain.
This is an area of code I'm planning changes to very soon,
so let me jump in before anyone else explains the way it works
now, which won't help you in a week or two.
The way it's going to work soon is this: the first word
appearing before a : character in a link [[like:this]] will
determine how that link is rendered in the final document.
There are several sets of such words that have different
functions. For example, two-letter ones are reserved for
language links, so that [[pl:xxx]] will create a link to
the article "xxx" on the Polish wiki. Recognized URL methods
are another set: [[http:xxx]], [[ftp:xxx]], etc., will make
direct links to those URLs. Another set is the wiki database
namespaces like "talk", "user", and "wikipedia". Those will
create links to wikipedia pages under those name spaces. Each
of these namespaces has a corresponding talk page--"talk"
links to the talk page of the main article; "user talk" to the
talk page of the User page; "wikipedia talk" to the talk page
of the "wikipedia" space article. This correspondence is used
for the "talk page/subject page" link in the sidebar. Coming
next week sometime, the "file" and "file talk" namespaces will
point to the image description pages and their corresponding
discussion pages (now confused with "image", which causes
problems).
Uploaded files are stored in a database independent of wiki pages.
The special-case prefixes "image:" and "media:" access these.
"Image:xxx" means "look for the file 'xxx' among the uploaded
files, and produce an HTML IMG tag". "media:xxx" means "look
for file 'xxx' among the uploads, and produce a regular A HREF
tag". "File:" and the corresponding "File talk:" will be regular
wiki page namespaces that link to a page with special association
with an uploaded file of the same name (so the page will
automatically show its upload history, what pages use the file,
etc.), but will otherwise be a normal editable wiki page.
Combining interwiki prefixes and article namespaces will work
in that order; so [[de:benutzer:Magnus Manske]] will make a link
to Magnus's user page on the German wiki. Image links will need
to come /before/ any others, so to create an IMG tag referencing
an image on the German wiki, you would use [[image:de:xxx]], and
the German wiki would use [[bild:en:xxx]] to link here.
[[image:http:xxx]] will NOT work; you will no longer be able to
inline-render external images. Alternately, I could make it
work for a specifically allowed subset of external sites.
So you will need to translate the namespaces like "user",
"wikipedia", and "file", and their corresponging talk namespaces,
and also the special prefixes "image" and "media". The two-
letter interwiki codes are international, as are "HTTP" and the
other URL schemes.
Good day to everybody,
I started participating in Wikipedias only 5 weeks ago and lately
I decided to get involved into translation Language.php -> LanguagePl.php
(ie. Polish). I've already got rough translation. Of course many problems
have arisen. Two of them are pointed out beneath. As I studied the code
of the whole Wikipedia for only few hours I'm not aware of many things.
On English Wikipedia, both images and sound files (what else in future?)
are loaded using SpecialUpload.php and appear on 'Special:Imagelist' (both!)
As far as I understand the code (cf. files convertdb.php, OutputPage.php as
well) there is no threat of confusion while using 'media' namespace for other
purposes. Did I understand the code right? Any plans taking into account
this namespace? Can we use 'media' as a namespace name
FOR BOTH IMAGE AND SOUND FILES on Polish Wikipedia without
any harm to the DB?
Another issue is the purpose of the namespace 'Wikipedia_talk'. I've discovered
only one page on English Wiki, sth about bugs of Phase II software.
Are there any other pages in this ns? (of course plain Search does not help
in finding those pages). What is the future of this ns? Any clues would help
in proper translation to Polish.
Thanks in advance for any replies
(User:)Youandme
----------------------------------------------------------------------
Zaufaj swojej intuicji... >>> http://link.interia.pl/f1646
Is there a way for us to sysop people from the web, or is this a database thing. If
it's a database thing, could someone grab this for me?
----- Forwarded message from Steve Callaway <sjc(a)easynet.co.uk> -----
From: "Steve Callaway" <sjc(a)easynet.co.uk>
Date: Wed, 4 Sep 2002 07:47:32 +0100
To: <jwales(a)bomis.com>
Subject: Sysop
Jimmy,
I have long been putting off requesting sysop privileges; I am and always have been, to say the least of it, a little unhappy with certain aspects of sysopdom, as you probably have gathered. However, the SQL querying functionality is beginning to become a "must have" for me. It will enable me to identify shortfall areas particularly in my specialist subjects (e.g. mythology, history, topology of SW England) where we need to get some efforts focussed, spot stubby or quasi-stubby articles that need attention, etc. My SQL btw is very very good (I am a full time DBMS designer) so hopefully I will be able to throw up some useful stuff for the other sysops.
I am also (occasionally) probably the only (sane!) person up at certain times, so if there is any vandalism occurring I might be on hand to deal with it. I certainly won't be deleting articles (except perhaps the patently pointless created by vandals) etc. Hope you think I'm safe enough to be let on the loose with 'em!
rgds
Steve Callaway (user:sjc)
----- End forwarded message -----
> Today around 2 pm (PDT) somebody copied the old version back with
> misspellings intact... I put the latest CVS version back in its
> place; if there was a reason for the revert (such as other changes
> made), please commit the changes to CVS and make sure a current,
> consistent version is put on the live server.
That may have been me--I made changes to that file today to fix
the "Benutzer_Diskussion" bug, and add the ISBN update. I'm
usually pretty good about keeping up with CVS, but I may have
dropped the ball on this one.
Phase III changed the handling of free-standing URLs: if the last
letter of the URL is a period, comma, semicolon or parenthesis, it is
taken to be part of the URL. I think this should be changed back to
the Phase II behavior.
I entered this before as a feature request, and I just learned that
Brion also favors the change. Lee argues that the current behavior is
correct.
I say that the goal is not formal correctness, but user satisfaction.
URLs almost never end in one of these special symbols, but when used
as part of Enlish sentences, they are often followed by one of them.
Numerous links all over Wikipedia have been broken by this change.
Just one example: http://www.wikipedia.org/wiki/BryceHarrington.
Axel