[Offline-l] Offline blog post

[Manuel Schneider]

Thanks Jessie for the blog post and sharing it with us!

I have put a link on our openzim.org webpage.


/Manuel

Just wanted to draw your attention to the blog post that went out today
http://blog.wikimedia.org/
Feel free to leave comments either there or on this list!
Best,
Jessie
***
*Update on Offline Wikipedia Projects *
The last week was a big week for expanding offline Wikipedia
<http://meta.wikimedia.org/wiki/Offline_Projects> work.
Right now, /offline/ refers to supporting read access to Wikimedia
content without an Internet connection. This increases the reach of the
Wikipedia movement by providing more opportunities for people all over
the world to access the materials. Some of the recent initiatives
surrounding this project were documented in Wikimedia’s tech blog
<http://techblog.wikimedia.org/2011/01/update-on-offline-wikimedia-projects/>
about a month ago (for more detail regarding the purpose for offline
work, see the offline strategy
<http://strategy.wikimedia.org/wiki/Offline> page).
In support of our offline readership work, we're thrilled to announce
the launch of a new feature on Wikipedia developed with our partners
from PediaPress <http://pediapress.com/>. Last week we enabled openZim
<http://openzim.org/Main_Page> (the main file format in which offline
materials are stored) export for the existing PediaPress collections
extension on English Wikipedia and numerous other wikis. This means that
individuals can now use the existing PediaPress Create a book
<http://en.wikipedia.org/wiki/Help:Books> tool and download it in a
format which can be read offline (via an offline reader, such as Kiwix
<http://kiwix.org/index.php/Main_Page>). This is important because it
opens new avenues for the creation of offline materials, for example, an
openZim library hosting different offline “book” options.
Also, the English offline collection Wikipedia 0.8
<http://en.wikipedia.org/wiki/Wikipedia:Version_0.8>was made officially
available
<http://en.wikipedia.org/wiki/Wikipedia:Version_0.8/downloads>, after
much hard work by the Wikipedia 1.0 Editorial Team
<http://en.wikipedia.org/wiki/Wikipedia:Version_1.0_Editorial_Team>.
This collection is an iteration in the process of developing a vetted
collection of offline articles selected based on their quality and
topical importance. The main constraint with an offline product is the
data size restrictions: the entirety of Wikipedia must somehow be
condensed so that it fits on a CD, DVD, or USB stick. Wikipedia 1.0 aims
at creating the highest quality and most valuable subset of Wikipedia to
meet those size requirements, and v0.8 is a precursor. Wikipedia 0.8 is
a general collection of just under 50K articles, It is available for
Mac, PC, or Linux with a Linux or Okawix reader; some mobile phone
versions will be available later this month as well.
More updates are sure to come on this offline front: Wikimedians around
the world are actively assisting in the development of offline
collections as well as distribution. We are excited to support and
document the momentum going forward.
Jessie Wild, Global Development
_______________________________________________
Offline-l mailing list
https://lists.wikimedia.org/mailman/listinfo/offline-l

[Manuel Schneider]

Hi,

thanks SJ for starting this thread, because this is something laying on
my desk. Unfortunately I am currently busy so can't really take care of
it (having internet at home would be a first start to fix this situation
:-) ).

A library / modulestore of thousands of available modules (zim files)
would be nice.

That's what http://openzim.org/ZIM_File_Archive is intended to be.
Currently still a wiki page, which I thought is for start good enough
before we come up with the shiny perfect solution.

My intentions are to combine several things:
* use the experience we got by mirroring / distributing the german
Wikipedia DVD (pre-openZIM), see http://dvd.wikimedia.org/

* give a central place to easily find ZIM files for users - should also
invite more people to use ZIM

* give ZIM publishers an easy way to publish / distribute / mirror their
files


The way it was intended to work ("the final version"):
1) a publisher provides us with a ZIM file (sends us a link, uploads it,
whatever - maybe we could have a neat webinterface where publishers can
register and maintain their files)

2) the file will be put into our archive (filesystem) and categorisation
(database)

3) the whole archive is being mirrored to several sites

4) users can use our webinterface to browse the categories and search
files, downloads will link to random mirrors in order to distribute load

Recently, Emmanuel / Kelson (Kiwix) has introduced a favicon.ico in ZIM
(URL: /-/favicon.ico) to be used in such systems.

Meta Data is also available (Dublin Core) in ZIM files which can be read
from namespace M. A minimal set of attributes have been defined:
http://openzim.org/Metadata

"snapshot type" (wiktionary, abridged wikipedia, wikipedia by
category, wikisource, other/custom ...)
~~
"language[s]"
"articles" (trusted only, by popularity, by wp1.0 score, all)
"article stubs" (yes, no, only popular ones)
"article length" (1st para, lede, summary, full)
"image size" (none, thumbnails, full)
"target size" (<50M, 200M, 1G, 4G, 16G, 64G, any size)
"image % of total" (none, 20%, 50%, 80%)
"templates" (yes, no, never)
~~
"export format" (zim, wikireader, woip, mw-xml, pdf, odt)

This is a very helpful listing which would help us to design the "final
version" of the ZIM Archive.


/Manuel