ZIM for entire English WIkipedia
~~ Is there a ZIM file for the entire English Wikipedia? I have the WP1 +30000, yet would prefer the entire version, even if the file is large and the articles haven't been vetted. AnAnthro 06:35, 14 October 2010 (UTC)
- Hi, due to a lack of Hardware ressource we are not able to prepare such a ZIM file currently. Maybe in the future... Kelson 07:46, 14 October 2010 (UTC)
- How about a version without images?
- The future is finally here, we have the whole en.wikipedia text only available for download. Stephenwanjau 20:42, 11 May 2012 (CEST)
I would like to see a Dutch version, covering all the NL entries of Wikipedia. Any chance this might happen?
- Please open a feature request. This can be done. http://requestafeature.kiwix.org Kelson 15:57, 4 November 2010 (UTC)
Entire English Wikipedia
Hi, I see you are not able to create a ZIM file from the entire english wikipedia. Can you provide a tutorial on dumping wikipedia content to ZIM files? This would open the path for enthusiasts to create and share their own dumps and we would see the english dump soon. --Klaufir 06:07, 19 December 2010 (UTC)
Where is the newer arabic wikipedia?
I want newer wikipedia than this in your site now it is old, can you creat newer one soon please>
and I hope your program to be good in search not like okawix which is very very very bad searchin just about 40% of results at least in arabic wikipedia .zeno
is you program better than okawix in searching tell me please,
You may check the search engine online here. http://library.kiwix.org:4205 Your feedback is really welcome.
This is the last ZIM file, do you mean this file is too old? http://tmp.kiwix.org/zim/0.9/wikipedia_ar_all_07_2009_alpha2.zim
Kelson 07:21, 2 February 2011 (UTC)
I have checked the searching although the library is Persian not Arabic ( but both in the sam alphabe)
The Arabic one is 
It seems very good searching like clucene > isnot it?
And yes I means that version is old one year ago 72009 update 32010
When the ar wikipedia is less articles
I want one newer now the articles are 138 000 and much more words pre article> Can you compile one??
Regards, Wish you luck.
- Oooops sorry for the error with Farsi. Great to heard that the search engine works well in arabic. This is done with Xapian... but we are also working to add a clucene backend. I have open a feature request for the new WP ZIM in Arabic here https://sourceforge.net/tracker/?func=detail&aid=3170499&group_id=175508&atid=873518 I do not know how much time I will need (I have also many other ZIM to do)... but I guess this should be done in a few week. Emmanuel
Good news to hear you workin to add clucene it is ultimat searching engine,and to adding new zim files.
THank you .
Hi, I'm trying to install Kiwix on Ubuntu 11.04 Natty Narwhal but it do not find the repositories for this distro. Someone has managed to install it in Ubuntu 11.04?, Thank you. --Karkeixa 13:17, 24 June 2011 (CEST)
Hllo Karkeixa in effect is possible install Kiwix in buntu 11.04, im used one fork of UBuntu - FlavituLinux, he is based on Ubuntu 11.04 and kiwix works, in the past in the first days of ubuntu natty release Kiwix package require some depndencies packages im not remember whats packages dependencies i need to install , but: -in this moment Kiwix install very well in Ubuntu 11.04 and no need package dependencies, in other words the problem is solved! - Portaro -
Wikitaxi Importer-like Tool
Thanks for your dev effort.
How about developing a Wikitaxi-like importer tool ( http://wikitaxi.org/delphi/doku.php/products/wikitaxi/index )? This way, the user would always be able to have/download the latest dump available over at wikipedia and thus convert it for use with Kiwix. Is developing such a tool that much hard to achieve? I guess you could set up an agreement with Wikitaxi's developer for getting programming help from him so as to get know how to do this. Sorry if my suggestion sounds naive - actually it seems his Wikitaxi project is closed source, isn't it? And thus he might not be willing to somehow share pieces of code that could in a way be used for developing a new tool by a third party.
Kind regards from Brazil.
- Hi Mchal. We won't do that. Kiwix renders HTML and do not have an integrated wiki render engine. We think rendering localy the wiki code is not the good approach. You are right, wikitaxi is closed source and also not portable. We work either to provide to download what you need full Wiki ZIM files. This should comes during 2012. Thx for your feedback and patience. Regard Kelson 12:40, 12 November 2011 (CET)
- Hi, Kelson.
- Ok, thanks. In fact, your whole setup is just fine as it is. I really appreciate your dev efforts - you offer an excellent wiki reader to users out there, which, by the way, I'll recommend to them whenever possible. My gratitude. Good luck and all the best.
CHP.EXE Reported as Malware
Windows Defender reports that the file indicated above is malware. Also, please, see this online virus scan report, here:
It's a bit worrisome such reports, you know...
- Hi, chp.exe ist not a malware so this is an error of Windows Defender. This is a normal open source program: http://www.commandline.co.uk/chp/. So you can ignore the warning and allow it running. But this is a problem that Windows Defender reports an error, I'm agree. We are working to avoid the usage of chp.exe in next version which will be released in a few weeks. You may follow this bug report https://sourceforge.net/tracker/?func=detail&aid=3389953&group_id=175508&atid=873515 Kelson 17:20, 14 November 2011 (CET)
- OK, thanks for your reassuring reply.
Open Specific Article on Kiwix Launch
Is it possible to open Kiwix and specify (on the command line) what article to load? We are investigating using Kiwix for an embedded help system, but for that to be context sensitive, the user would click a button in another program and that would cause the other program to launch Kiwix, and also tell Kiwix to load the contextually relevant page. Joeclarkia 18:18, 29 March 2012 (CEST)
- I found the SourceForge Tracker and found ticket 3484628. It has basically the same request. Joeclarkia 21:56, 29 March 2012 (CEST)
- Hi, just wanted to tell you that this is now implemented. More details at:
https://sourceforge.net/tracker/index.php?func=detail&aid=3484628&group_id=175508&atid=873518 Kelson 13:05, 1 April 2012 (CEST)
The part about compiling should be moved elsewhere: it's only suitable for a technical audience. Nemo 18:26, 7 January 2013 (CET)
I think it's better if the pages moved to their original native title. We can imitate Meta-Wiki which is a multilingual project. In there, there is a Main Page (which is in English) and also the same page in other languages with their native title. For instance, "Main Page/fr" would be Accueil. ●Mehran Debate● 10:00, 10 January 2013 (CET)
- Yes, you can make a redirection. Kelson 10:07, 10 January 2013 (CET)
Update of English Wikipedia
As I see at http://download.kiwix.org/portable/, the file of the entire English Wikipedia is over a year old, and the ones with parts of it are even older. Is there, or will there soon be, an updated file? Mikael Häggström (talk) 04:40, 22 July 2013 (CEST)
- We should release soon a new version of the WP1 project. New version of the the whole WPEN without picture is also in the pipe. Kelson (talk) 23:34, 22 July 2013 (CEST)
Day by day update
Hi,Kiwix is a great software but Why there isn't any way to update it every time I want? Can you add it? It is really good espesially for languages like Persian which sometimes don't even update year by year it is so useful. Thanks --Rayeshman (talk) 17:21, 10 December 2014 (CET)
- Hi, We deal with millions of entries (both articles and images). There is no efficient way to update this correctly day-by-day, the amount of data is simply too huge. Our plan is to release a new version one time a month (and this is already running for smaller projects) and to provide incremental update (should be release next year). Kelson (talk) 13:08, 11 December 2014 (CET)
Hi Kiwix Team, First off, thank you for all the work you have done making wikipedia available offline. Offline is greatly needed for archive and for internet poor countries. I am involved with a grant proposal for an extension for easy offline xml generation for a wiki. We won't be using the zim format due to the easy of having xml used within a mobile app. Our goal is to create an extension for use with any mediawiki installation so that offline comparable mobile apps or any application that can parse xml can read the wiki offline. I would appreciate if you could comment and even support this endeavor as we bring offline support to all the mediawiki installations in an easy to use extension and mobile app for iOS and android. https://meta.wikimedia.org/wiki/Grants:PEG/Offline_MediaWiki_search_for_NASA_and_Medicine
This will not be a competitor for kiwix as our scope is for smaller mediawiki installations with support for a highly customizable offline sync file that can be styled, customized and tailored mainly for mobile offline use.
Thank you, Dan Ostermayer
- Dear Dan, thank you for pointing to your request. I'll have a look in detail to it. But, I have been working on offline solution since almost 10 years and I'm sure that using the Mediawiki XML export is not the easiest way to display articles offline. The reason is that this is far more complicated to have an alternative Wikitext parser than just render HTML, even if this HTML is packed in a ZIM file. For the rest I'll ask questions about your requirements on meta... But it looks like that what you try to do is simply the same than Kiwix, so I wonder why we were not in touch before. Re-creating a whole solution ecosystem just because the fulltext search engine is not available on Android or because there is no iOS app. sounds a little bit short as explanation (fulltext search engine might be implemented within a week and an iOS app within a month). Regards. Kelson (talk) 17:45, 21 December 2014 (CET)
Tajiki Zim file?
- I will prepare one, but may you please help us to translate Kiwix user interface in Tajiki on Translatewiki ? Kelson (talk) 19:35, 24 December 2014 (CET)
- I am actually Iranian and my language is Persian(With Persian alphabets) and I'm not fluent in Tajiki which is an accent of Persian but is written in cyrillic script,The Russians did that so nowadays people in Tajikistan use Russain much more than Persian,at least in their online activities!! And I'm going to use this ZIM file and some other things to write an script which can transliterate Tajiki to Persian and vice versa, I'll use my script and Persian translate to make the Tajiki translate available. It will take a while but I will.Thanks a lot for the ZIM file!
- Tajiki is an accent of Persian so there are some little differences also sometimes there are three or two letters in Persian alphabet for one letter in Tajiki alphabet, so I'm thinking about a sqlite database and an script as API. and unfortunately I'm not good at cpp.
Downloaded kiwix-0.9+wikipedia_en_all_07_2014.zip via bittorrent, no problem. Then got these errors when trying to unzip:
UnZip 6.00 of 20 April 2009, by Info-ZIP. [...] file #1: bad zipfile offset (local header sig): 46155327025 [...] UnZip 5.52 of 28 February 2005, by Info-ZIP. warning: 46155327101 extra bytes at beginning or within zipfile (attempting to process anyway) error: start of central directory not found; zipfile corrupt. (please check that you have transferred or created the zipfile in the appropriate BINARY mode and that you have compiled UnZip properly)
Unzip 6.00 keeps going, but throws lots of errors; unzip 5.52 stops after what is shown above.
Bittorrent should of course do checksums itself, but I wanted to make sure. So I tried to look on the site for that md5sum, no luck. So tried the other way, did a local md5sum on the file, then searched that via google. That came up with
So the bottom line is for some reason the zipfile throws errors with what may be the most canonical unzip program. I'm now unziping via the "Keka" program, a GUI front end to 7zip, it seems to be going fine, but that may just be because it's ignoring the errors.
Also I find myself wanting to know the checksums of all of the files inside the zipfile.
I'm planning to put a link to the checksum on the front page. I'll also try unzipping with different tools on different platforms, if that works out all the same I'll post per-file md5sums, although someone who has access to the original files doing that would be better. It's also be good to know which program is used to make the zipfiles.