Hmm…here we go again. This time it is a rant about why we should attempt to make a localised desktop complete. So why is completeness important ? Or more precisely, why is this important now ?

In the light of the analysis of everyday incidents, and their very remarkable conclusions drawn in The Tipping Point – it is easy to see why a complete Linux desktop is in the offing. For a moment, just think about the wild upswing of Firefox. It started as a trickle and then slowly turned into rolling thunder gathering tremendous momentum along the way. Now that the initial spate has tapered off, a study of Firefox downloads might provide greater insight into the number of people upgrading Firefox. The upgrade part is important. Since each upgrade represents in some ways an incremental feature enhancement, the increasing usage of higher versions (for repeat consumers) would signify the need to have more feature creep into the software development model.

What would be the ideal desktop in any local language ? At any point in time this is a difficult question to answer. Currently more so than ever. Indic L10n shows remarkable maturity skew when it comes to desktop environments and the various development libraries. Choosing either GNOME or KDE immediately limits the number, type and user interface of applications one can have on the desktop. That said, both GNOME and KDE ship with a number of redundant applications, or applications which collectively make a lot of sense but fail individually in terms of feature value. The ideal localised desktop could aim to be modular in terms of applications. The concept is to replicate the stack model made popular by Senpai and Spikesource and extend it to the desktop space. A base set of applications which provide for comprehensive localisation would enable a lot more usage of it than one which provides an unstructured collection of applications.

What could be the basic stack of applications ? To begin with, one needs to look into providing the following: a user manager, a system services administration tool, a web browser (with plug-ins enabled), a ftp client, a mail client, a downloader program, an office suite, a task manager, basic games (card, strategy and the like), sound and video player and graphics tools. In a normal day, these are the applications a normal user would be more likely to use.

To begin with, one could make a list of the L10n status for each of these top-level categories so as to enable one to create a prototype desktop. A tentative plan for the same was unveiled as GRIND. Already GRIND has thrown up some interesting side tracks in the form of Senpai and the Koha stuff. One only has to wait and see how it goes further.

Asa Dotzler on firefox, cats, mars, and more: linux not ready for the desktop

A chosen few of the presentations (by track) are available for download from this link.

Additionally, do check this out – a legal guide for Bloggers.

I subscribed to Newsburst, somedays back and find it a nice service to have in addition to newsfeed readers like Liferea. Haven’t find time to blog about stuff for a long time now. In the meanwhile Google has gone ahead and done a wonderful thing by declaring the Summer of Code open. Please find it here. Another new magazine on Open Source has been released. On the IndLinux mailing lists and some of it off-line, the discussion has been more on the deployment scenario for the localised GNU/Linux distributions. However, nothing concrete has been arrived at in terms of Action Points and milestones. Additionally, perhaps it is time that we did a scorecard on what the various Indic Languages have to offer in terms of a government desktop. The Government of India is by far the largest consumer of desktop PCs and mobile computing platforms. If localised distributions have to make any significant impact on the desktop niche, the application stack have to be identified.

Planet FLOSS India has a few new bloggers, while Planet Gnome has a new look that I rather fancy. It is perhaps time that we did something about the look of Planet FLOSS India.

On a side note, tested out gtkpod some hours back. The utility works fine and seamlessly but /me has some feature enhancements to be put into it. My friend Sanjeev Chauhan proposed a nice idea – how about having the “electronic mosquito repellants” come with a timer and scheduler firmwire built into it ? Would make a nice USP if you think of it…

Indic localisation is a buzz item now. Everyone talks about it and there are projects that attempt to use (and sometimes re-use) various aspects of this hot ticket. The one thing that is missing amidst all this is the need to talk about deployment of the localisation projects and more importantly – discuss about how best these efforts can be leveraged. All the projects, all the efforts to meet translation (and string) freeze deadlines will come to be nullified if we do not manage to deploy them. Sadly, that is what is happening with the L10n effort.

Sometime back (and more often these days) I talked about application driven L10n. This is nothing more than identifying a core stack of applications that are used in eGovernance and other Government transactions and then going ahead to put in place a roadmap for making them available in maximum number of Indian languages (if the application permits). On the face of it, not supporting an Indic desktop, but going after applications might seem to be counterproductive. I agree that it could backfire. But then look at the bright side – you get to introduce L10n in form of interface localisation as well you get to drive FLOSS into the game. The desktop environment will catch up in due course. Already there are DEs Projects in other Indic languages which are attaining a sort of stability wherein they can be put through UI tests. If someone has doubts as to whether such a stack can be identified, do take a look at Spikesource – their stack is wonderfully created to integrate just those elements which enable some developer to go hard on a LAMP course. Additionally, the time is perhaps just right for the L10n groups to consolidate themselves and leverage Intel’s resources in terms of reach as well as its deep involvement in ethnographic studies.

Additionally, what we should be looking towards is involving the user in the development process. I will try and write more about this in the next post.

Here is a nice blog.

A longish thread – but do take time to read through it all.

Take a look at Bug # 267132 for Mozilla, Firefox and Thunderbird L10n CVS request for bn-IN and especially this. This could possibly top the list of the prime reasons FLOSS movements struggle to gain initial momentum.

One wonders how the poster of the Comment # 6 could be in a position to state with unequivocal conviction that The language od (sic) bn-BD and bn-in are absolutely same. .

The funny part in this entire imbroglio is that bn-BD was registered/requested much earlier to the Mozilla Localisation Project and nowhere was bn-IN mentioned to be an Ankur sub project. Unwarranted assimilation of any project related to Bengali (and Linux) had never been the hallmark of Ankur and this is one of those cases where it has been publicly displayed.

There are some times when one just feels the need to shout. Sumit Ranjan Das – Coordinator of Burdwan LUG has been sitting on Mozilla Localisation for bn for a period of over 1 month with no status reports or updates.

Gaaarrrrhhh !!!