Saturday, January 22, 2011

Troubleshooting the Wikipedia For Schools Offline Edition + Call to Action

DSC02634That big term shall henceforth be referred to as WFSOE (wow, sounds like a cool game! As it should!)
In my pilot drive at Shantabai Ladkat school in Nanapeth, Pune, I came across some issues when installing WFSOE on some older computers: The indexing process never took off. Now, indexing is what kicks in the first time you open a .zim file in the primary form of Kiwix (the software used here.. look, just go with it, ok?) – it’s necessary if you want to be able to Search thru the WFSOE. So that’s a dilemma : a whole encyclopedia and we can’t even search our favorite topics! So, started looking around.. and thanks to a comment on my original post on WFSOE, I started downloading the pre-indexed, ready-to-DVD version of WFSOE that the awesome folks at prepared:
Note: Don’t start downloading it yet – it’s freakin’ huge – 2718MB direct download! And, as in many such ventures, I detected a problem only AFTER fully dloading it over 3 days! I can’t say whether the issue is there in the original file hosted on or if it crept in during my downloadathon, but that’s just it – huge file direct download = Russian Roulette. This blog post is the story of that issue, how I hacked my way past it, and the solutions for anyone who wants to download a good, awesome, readily working version of WFSOE. To make things interesting, we’ve got screenshots galore!
Expand this article to read thru it… long length!
If you’re looking for the wiki page, (short URL:

If you’re just looking for the Call to Action, go to the very end after expanding.
“You must be the change you want to see in the world” – M K Gandhi
So, let’s begin… here’s the downloaded .zip file – all 2718 MBs of it:
zip archive
Opening it in 7-Zip Portable:
zip archive iin 7-zip
I extracted it the first time itself, but just pressing the Test button reveals the error:
file broken
Now that’s a bummer. The schools….rc5.zim file is the MAIN file here – it’s the 2.something GB behemoth that’s a compressed archive of all the htmls, images and everything else that makes up the miracle that is the WFSOE. To put things in perspective : It’s one of the largest ebooks created by mankind. The rest of the stuff in the whole zip archive are just the software for running this baby.
Fortunately, I’d already obtained that .zim file separately from elsewhere! Incidentally, I’ve even set up a mirror website that hosts this behemoth in parts, you can go get it from there: (look for the “schools-…rc5.zim in parts” folder).
So, I now opened the extracted WFSOE folder and made my way to the place where I had to make the switch:
data folder
content folder
It’s the data \ content folder in case you can’t see the images. There, I found a 645mb broken version of our .zim file. Renamed it and copied over my *working* .zim file!
zim file
And.. now for the test run! Go to the main folder of the extracted archive, go into autorun folder and run autorun.exe
autorun folder
autorun exe
And voila…
kiwix launch
Updated : Use the Install option to put it into the computer. You can change the install directory if you don't want it all the way in Program Files etc. It says 42MB but that's a small typo : After installing the software, the whole 2.6GB ZIM files is also copied over to the system. It also creates a shortcut to "Kiwix" in start menu; I recommend you copy that shortcut to the desktop and rename it "Wikipedia for Schools", or even drag it to the startup folder so it starts up every time the computer starts. For now, I just went with the first option, “Launch Kiwix from here".
And IT WORKED!!! WEEEEE!!! Here’s some screenshots.. of the main home page that first loads, then the Subject Index where categories are given, then 2 shots of random articles I searched my way to. Click on them to enlarge (Duh)
WFS home WFS Subject index
WFS article 1 WFS artice 2
Hacking into this pre-indexed business,
data folder
index folder
idx folder
idx properties
Oh. So I’m gonna have to… for this little …
Ohhhhh @#$#$#$@&%^%.
In case you can’t get this thing to run or to search, it may be some drivers missing. From the main folder, go to Install folder and run vcredist_x86.exe . May the Force be with you.
To proliferate this ready-to-run version across the world. I’ll also have to put in something at the top level so people don’t get confused. (The autorun.inf isn’t for humans, and if you’re just copying the dump over a pen drive, you need to know what to run, and many people are too scared to fool around the place like I do).

….thought process….

Ok! I’ll have to post 2 versions :
One – the whole thing, re-packaged. This time I’ll do a 7-zip (.7z) instead of .zip – it’s freeware, got better compression and it can split itself into parts.
Two – Like me, there must be others who have got the .zim file already. For them, Just the supporting file structure (most importantly the pre-indexed stuff) should be enough; they can then just copy the .zim file into the data\content folder the same way I did. So, gotta make a separate zim-less archive for that too. I’ll host both of them on my mirror website :

It’ll take some time over my humble 256kbps broadband connection, but I’ll do the zim-less version first and then go for the longer haul. Give this a day, ok?
Update: Done! Full pre-indexed version available here
So… here’s a stupid question I get asked: Why am I doing this? Am I a trustee of wikipedia or of WFSOE or something? Do I know the guys who made this? Am I a person holding any position? A guy with any resources? Was I hired for this?
DSC02649I haven’t spent a dime in all this nor am I getting anything for it, nor do I want to. I’ve downloaded the stuff off the free web, used 7-zip, a totally free compression utility and my mirror site hosting these large things is 4shared – on a free account! (10GB limit, FYI. Their uploader utility’s awesome and they don’t ask difficult questions or screen out benign exe’s like some other idiots I won’t name here).
I’m just an ordinary guy who wants to do his part in eliminating educational inequity. (ohh yeah, really rubbing it in, aren’t I!) The driving force here is a little bit more intrinsic.
Friends, this is a mission : To take the WFSOE to each and every non-internet-connected computer in the world. So that we can bridge the educational divide and give all those millions of kids access to humanity’s largest knowledge resource – for Freeee, and that too when they need it : in their childhood instead of after they’ve turned adults! We can’t wait for the fancy high speed internet to go everywhere and that too be free to all. That will happen later, but we can do THIS right NOW, from the grassroots. Having a proper ready-to-go version that can be downloaded by one and all is critical here.
DSC02627Wikipedia have done their part in creating an amazing encyclopedia.
SOS Children’s Village have done their part in hand-picking the best articles for school children and creating the Wikipedia For Schools. have done their bit in compiling all of it into a .ZIM archive that we can use without the decompression hassles of some other offline website readers. And they even made a pre-indexed ready-to-DVD version that we can just copy and run anywhere.
Several organizations all over the world are doing their part in getting computers donated to low resource schools where kids from lower economic backgrounds can get a chance to use them. But most of them don’t have the WFSOE.
And now it’s OUR turn.
This cause needs OUR HELP. Get the WFSOE. Go copy it to every computer you can get your sweaty hands on! Pass the baton on - Get your friends in on the action! Spread the word! Here’s the wiki page set up to co-ordinate this effort – enlist yourself as a WFSOE Champion and get to work!
“Never doubt that a small group of thoughtful, committed citizens can change the world. Indeed, it is the only thing that ever has” - Margaret Mead
Nikhil Sheth
Pune, India
Teach For India Fellow, 2011-13 (just a volunteer at the time of writing this)
nikhil.js [at]


Kelson said...

Few remarks:

* The issue should not be on the original file, I have test it. The md5sum is 1ecc048acfbdb96ac7a6f7f4addb9d48

* About the 42mb, this is a bug in the installer. Is already fixed and will be released with Kiwix 0.9 alpha8

* The vcredist_x86.exe should also help you to index ZIM files.

Thank you for spreading Wikipedia. For future versions, if you still have difficulties to get content, I may send you per post the data on DVD.

Ashwin Baindur said...

Way to go Nikhil. Wish we knew about your project before. We could have showcased it here:

BTW a Wikipedia T-shirt to raise your morale awaits you. Contact me through my talk page/email.

Ashwin Baindur said...

User:AshLin on English Wikipedia or

Related Posts with Thumbnails