Sikher News

Sikher Website Revamp

Sat, 29 Jul 2006 02:55:39 GMT

Our website has undergone a complete update of content. Please visit our all new Support section to promote Sikher on your website or blog using promotional images and text. You can also donate to The Sikher Project either through buying a product from The Sikher Online Donation Shop or directly. The Sikher team are continuing to make vast progress with the project and some critical ideas have filtered through about how to create a truly portable XMLGurbani datastore. A release date has not yet been set, so to avoid any dissapointments with the actual Sikher 1.0 release, however a lot of development is set to occur in September and so we hope we can release something exciting for you then. Developments alongside our marketing campaign roll out should raise the support and awareness we need to move towards an earlier release date. Satnaam.

The Vaisakhi Release

Fri, 14 Apr 2006 19:54:18 GMT

Over the past few months we've been working hard to try and get the Vaisakhi release of Sikher ready. However, even after our last plea of help, our project still is suffering from lack of financial and human resources to complete version 1.0 of Sikher. We've started some advertising by getting people to sign up to our newsletter (this can be found in the panel to the left) and shall be trying to increase our exposure through local marketing. But we still need your help!

Because we could not finish Sikher 1.0 to our agreed specification, in order to avoid any dissapointment to our keen fans, we've decided today to release a 'Look & Feel' Version of Sikher. We are aware of the bugs of this very early Beta version (should really be called an Alpha version) so please do not send us any support requests :). You can download the 'Look & Feel' Version at https://sourceforge.net/project/showfiles.php?group_id=117096&package_id=187437&release_id=409855. It includes:

* Sikher 0.4 BETA ~ The main application, to show you how far we've progressed
* Sample Outputs ~ A sample of the types of outputs expected from Sikher 1.0 release
* Sample XMLGurbani Files ~ A sample of the Guru Granth Sahib which may be searched in Sikher
* Unicode Gurmukhi Fonts ~ The AnmolBaniUni fonts have been included, which needs to be placed in the Java fallback folder as described in the readme.txt

Sikher Version 1.0 Release Delayed

Sat, 31 Dec 2005 16:19:22 GMT

The moment we have all been waiting for has in fact been delayed. But this is an anticipated delay due to the Java re-write. We've now successfully implemented the Search & Display function and are working towards implementing the Akand Paath, Project and Slideshow functions. We've made some improvements to XMLGurbani and volunteers are in the process of creating the Sikher XMLGurbani Convertor which shall aid us in archiving masses of Gurmukhi Unicode text into the XMLGurbani Standard, as well as allowing for the ability for anyone to easily create their own translations and transliterations for any given Gurbani text.

We are making huge progress, but as a result of the delays (and lack of support) we've moved our Version 1.0 release date to the 14th of April 2006 (Vaisakhi - Sikh New Year). No need to worry though, we'll definitely release something for you to see over the next few weeks. Some BETA testers have already had a look at our initial progress and the reaction has been very good! As we approach closer to the new 14th of April 2006 deadline we hope will will have had a number of BETAs out, all the Gurbani archived and added to Sikher and we also hope to roll out a marketing campaign. Please support us in any way you can. God Bless!

The Java Re-write

Sat, 03 Dec 2005 11:00:30 GMT

After much deliberation we have decided to instead switch to writing the application in Java. There are many reasons for this:

(1) The main developers of the project are fluent in Java but not in Python.
(2) The developers have also felt that Java offers much more support in coding and bug fixing than Python.
(3) Also, the libraries used in Java are much more extensive and mature than similar Pytonic projects and less libraries are needed in Java. In Python on the other hand every component is modular based which makes it difficult to track updates to the various libraries. This is not a problem in Java.
(4) Python has been fantastic in creating small projects but Java is more of a favourite for the larger project.
(5) Java programs usually run faster than Python ones.

All in all, Java is the only other viable coding language for Sikher because it is platform independant.

The developers of Sikher are confident that the project shall not be delayed with the Java re-write and so we are continuing in our pursuit to release Sikher by January 1st 2006, and hopefully a BETA version much before (although it is looking unlikely until about mid-December).

We hope you will join us in this pursuit! We're still looking for developers...including XML experts, marketing experts, documentation experts and others who can help recruit and actively promote Sikher in their local and national communities. One thing is certain, the future is looking incredibly bright!

Gadfly and 4Suite Developments

Sun, 16 Oct 2005 08:53:02 GMT

The Gadfly has now indeed been implemented into Sikher and is being used as an indexing tool to allow for lightening fast searching. 4Suite development is also very heavily underway and an XPath querying mechanism should be in place very soon which can query the XMLGurbani files directly! These are exciting times! Once we've finished this, this will be a major milestone for us and we can then go on to develop the other exciting features of the application that utilise this basic search and display function!

The splash screen has been implemented as an added GUI User Interface enhancement. Much of the User Interface has been improved and a drop down added to be able to select a certain Bani (scripture) to search in. There are also other minor changes.

The automatic transliteration of Gurmukhi Unicode to Latin is almost complete, only one more tweak is needed which should be completed very soon. This will be a milestone, not only The Sikher Project, but also for the Punjabi & Gurmukhi world of language. In my testing of the transliterator I actually found that our pronunciation of many of the sentences we say in Gurbani are wrong. This makes this tool even more important. This is however NOT a function of Sikher yet and will only be available as part of Sukhjinder's release of MetaMorph, in the form of a mapping file.

I can estimate that because the initial searching and indexing has taken so long that the users at home can expect a delay with the BETA release of Sikher, therefore we may be looking at a release in late November/early December to squash those bugs...but we'll do our best to make it soon! G'day world, G'day.

The Data and the Base

Sun, 18 Sep 2005 10:11:59 GMT

In every project there comes a time when one must decide how to store the data of the application. Nowadays there are soooo many databases available to use for any given task. With Sikher the criteria we had was that any database we chose must be able to support XML, Unicode and have Pythonic bindings. The choice was a hard one. There is Sleepy Cat's Berkeley DB XML which is the native XML database engine (which is 20MB!), Metakit (which we don't believe supports either Unicode or XML), 4suite (which we don't know how to use as a database), Gadfly (which cannot work with XML) and some other Open Source one's like DBDOM, dbXML, eXist, MyXMLDB, ozone, XDBM, XDB, Xindice and XpSQL. So you can see the problem. Each has its advantages and disadvantages, therefore of course we only picked a select few to test and attempt to use.

Currently, we are trying to use Sleepy Cat's Berkeley DB XML and will hopefully see some results over the coming week. We've already tried to use Gadfly. After all the database stuff is out of the way then we can start to create the Gurbani, Transliteration and Translation data in the XMLGurbani format and enter it into the database. Potentially this means that for any given scripture or text, there can be an infinite number of translations and transliterations. This should provide us with the flexibility we need! That's all for now, have a good weekend!

Transliteration System Changes

Fri, 09 Sep 2005 21:31:42 GMT

Sukhjinder has been creating new developments with the way MetaMorph handles transliterations. His idea is to use pre-text and post-text processors using javascript to handle all the special and additional conversions required for Gurbani. These will include dropping some vowel sounds, adding special symbols like Ik-Ong-Kaar and refining the phonetics of the transliteration. We'll hopefully finish this within a week or so.

A newer version of Sikher is expected to be committed to the repository over the next few days. This will include the ability to archive any text into the XMLGurbani format, which has been refinied to the T. There's still some work to do but this will be our first major milestone! That's all for now. God Bless.

Transliterator and XMLGurbani

Tue, 06 Sep 2005 20:25:02 GMT

The two major developments for today are that the Latin transliterator for Gurmukhi Unicode is 90% complete and also a visual layout for XMLGurbani has been created to help view the XML data more nicely in a browser. That's all for now. Oh and Sukhjinder from GUCA (http://guca.sf.net) has created a Phonetic Gurmukhi Keyboard Layout which can be found on his website under typography.

Solving the Unicode Plight

Mon, 05 Sep 2005 11:14:15 GMT

After a week of utter misery, the plight of WxWindows not handling Unicode correctly was in fact solved with our proposed solution. Thanks to the mailing lists of WxWindows and to our fantastic Developers! We can now begin to focus on more actual development of Sikher. It feels as if a plague has been eradicated from the software. As a final note: Make sure you stayed tuned and get others tuned in too!

The Plight of WxPython

Thu, 01 Sep 2005 00:26:40 GMT

As we approach into the month of September we've come across a hurdle in the way WxWindows handles Unicode. It would seem in my diagnosis that in fact this inherent flaw with the, wxTextCtrl.GetValue() and Shelve Key, which causes the application to crash is in fact due to the program ending up writing past the wxBuffer array. The size of the wxBuffer, which is about 1k or so gets overwritten with more data which causes it to crash. Therefore, the actual size is much greater. A solution may be to modify the text ctrl python file to create a local buffer greater than the actual size. This should stop the crashes from happening. I have just realised that in fact this News Bulletin that was meant for the average user is becoming more of a Developer's RSS Feed. This is a sham of the entire globe. What a shame, what a grave shame. P.S. I may be forced to create two seperate RSS Feeds if I continue in my ways. May the force be with you.

Sikher gets a facelift

Mon, 29 Aug 2005 19:40:14 GMT

Today Sikher got upgrades to its user interface as well as the optimization of its indexing core. We've also travelled leaps and bounds in creating XMLGurbani, a powerful XML schema designed to archive any of the Sikh scriptures in a common format. XMLGurbani and Sikher are now truly set to revolutionize the world as we don't know it. Stay tuned for other exiciting developments with the Sikher project. Good day, good night and God bless.