Microsoft, you giving us some LIP?

by Michael S. Kaplan, published on 2005/02/10 07:30 -08:00, original URI: http://blogs.msdn.com/michkap/archive/2005/02/10/370460.aspx

Not too long ago, I posted (in Lions and tigers and ~~bears~~ELKs, Oh my!) about ELKs. And about how I would talk more about what is involved in "getting out of the way". This is another post on that theme...

Now ELKs were not written just for the sake of adding locales for the fun of it. They were added to help a specific, important scenario.

You see, no matter how useful additional locale support may be, it still forces you to use Windows localized into a language other than the one in which you may feel most comfortable. And the real scenario here that drove the ELK project was a desire to allow localized versions to be released more quickly, in some cases localizations into languages that were not yet supported on Windows (and the ELKs were in this context a fancy way of making sure that support would be present).

Now remember how I bemoaned the expense of obtaining locale data? Well, the expense of localization blows that number out of the water.

It actually reminds me of an old Dilbert cartoon, which was in newspapers back in May of 1999 (to my knowledge it is the only Dilbert cartoon involving internationalization or localization outside of the funny stories involving the Elbonians. The lines in this strip (sorry, no permissions to repost the original comic!) go like this:

Tina is right -- bad localization is cheap and fast. But you also won't find many people buying it.

Good software localization is hugely expensive, and involves expertise that is hard to find and keep -- knowledge of the source language, the target language, and most importantly the technology and how it is described in both languages. It is not mere translation, because the work to make people feel comfortable with the product also has to be done if you want users to be happy with the product.

So how does the idea of getting localization done faster and for more languages get reconciled with these difficulties?

And this is all very important. Because there is a huge demand for more localizations, and more quickly. So by partnering with third parties allows for cheaper, faster, and culturally accurate localizations, and most importantly by working to localizing "the 20% of strings seen 80% of the time" Microsoft is able to enable cheaper and faster localizations, while not sacrificing quality.

At the time of this post, there are LIPs for 19 languages (when we did the presentation, there were only 4!):

For more information on LIPs, including what is localized and more, see the LIP FAQ, brought to you by Microsoft's own Dr. International!

Looking back at the earlier post about ELKs, the temptation to talk about how some LIPs did not need ELKs (such as Romanian), and other ELKs have LIPs (compare the two lists, you will see many similarities), and there some LIP-less ELKs (again, compare the lists). Think of what those ELKs save on ChapStick!

Of course you never know what might be a LIP tomorrow. Who knows when Microsoft may give you all some LIP?

The goal for both ELKs and LIPs is to support more people with more languages and to constantly work to broaden the number of people who can enjoy Windows.

And yes, there is more that was said that night, and more to say here in general. For another time....

This post brought to you by "æ" (U+00e6, a.k.a. LATIN SMALL LETTER AE)
A letter that itself plans to be the star of a future posting on Sorting It All Out....