Yesterday night, over an exchange of mails a few friends narrowed down the root cause of the misbehaving OpenOffice.org (the rendering for a particular document was off-the-charts on one distribution, but was pretty much sane on another). All of that led me to think that given that Indic Localization has been the talk-of-the-town for over four years now, and pretty much available through major Linux distributions for the past two years, there has not been a large number of folks actually using localized Linux. That’s kind of sad if that is true, it is really a bad thing if folks actually use it but don’t talk about it.

Let’s take the case which kept a few of us busy last evening. The original poster “uses” localization bits. Being a writer, he’s affected by fonts, input methods, rendering and printing bits. Anything that means:

+ installation or re-installation of the Operating System

+ fiddling around with various fonts

+ tweaking input methods

+ rebuilding/recompiling applications

is a waste of hours which take the writer away from the core task. Somewhere along the way it might be time to start thinking about “users” like our writer to see how localization help them do the work (while also being booted into a local language environment). There would be various types:

o the blue-sky users from the ICT4D group who require the L10n on the user interface to enable to be part of “inclusive computing”

o users of various eGovernance projects who require L10n framework (input-rendering-output) to get the various kinds of transaction down

o the users (like our writers) who use tools and not really the OS and expect the tools to be “L10n aware”

The current mode of Indic Localization aim to meet the requirement of the first two groups (second more than the first perhaps), but what we don’t seem to be getting into the third group into the game. Given that spinning out a custom distribution (and LiveMedia) has become so much easier across distributions it would be a good idea to have feedback loops going by getting the third group involved. Localization of late has attained the unfortunate tag of “string crunching”. Fact remains that L10n does mean that but actually means much more than the very restricted comprehension of the term. Thus, the greater the number of folks using L10n bits the better it is in terms of being enriched.

In a side note, these two were incidents of yesterday and trivially interesting.