If you want to create a software is used by users from all over world, Localization is important. For 6 months(until July 2006) I try to manage all localization process both technic (I mean technology) and translation (Yes we have translations but they are really understandable by users).
What's Software Localization?
"Software localization is a process of translating software user interfaces from one language to another and adapting it to suit a foreign culture. This process is labor-intensive and often requires significant efforts from development teams. There are tools that can simplify the localization process. A lot of these projects get outsourced to specialized companies to cut costs.
Software localization can also refer to adapting software from one culture to another. For instance, the Nintendo GameCube game Animal Crossing changed various holidays during its localization. During localization, some references to history or culture are replaced with an equivalent native reference which ideally does not change the meaning of the statement."
this is from wikipedia...
In Rainbow 1.6 there is 29 languages' localization but it's not complete. Is it really successful? Of course no. Why was unsecessful? Why am I talking old things? Should I talk about new things?
In order to create complete and successful localization we should analyze our past and our rivals. Today analyzed our past.
Why Rainbow was Unsuccessful technically?Rainbow or Manu :-) created really good component Esperantus(Thanks for Manu and Jes).
But we did not use effectively and support this product with complementary tools. Manu tried to this. He wrote a program this. But result again unsuccessful. Because no standart in keys, it changed time to time. There is really disorganization. Developer creates new keys without paying attention. There is no convention for this. How can a developer give a key name etc.. Lots of double keys... Because we don't have an organization and standarts for localization. These are can be determined at the beginning of project. We tried to arrange keys and put standarts for keys. But it was a bit late. And for translation there is still compilation. People should have compiler to update this. And we added super functionality private assembly without localization support also without theme support. I solved the the theme problem and Jes completed it thanks for Jes.
So we should really had module localization support.Why Rainbow was Unsuccessful in translation of languages?
There is a mail list for localization team. Unfortunately it does not work and anyone managed this. Also localization process a bit complicated. So What happened? Only Developers translated (I was also developer, yes yes I'm a computer engineer if necessary I can deveop again). So some translations a bit comic :-) Why? Because when people translating a key he/she really doesn't know where the key is used. So only translates.
So translations should be tested.
We should increase the number of translator for each language. I liked google's translation system. I translated somethings to turkish language. But now They have many language translators.
When translating we should care about common applications people use, people want to continue their habits. And we should have translators these are not only developers.
Really we don't have complete translations for each language. So communication is important.
For localisation also we should consider RTL languages.
Yes we can create perfect software, if it's not completely localized unfortunately we don't reach all over the world.
And remind you our general vision
"Make Rainbow number one content management solution in the world". (http://www.rainbow.web.tr/portal/alias__Rainbow/lang__en-GB/tabID__3647/DesktopDefault.aspx)
Maybe we can change this "Make Rainbow number one framework".
Today is that's all.
My next entry will be about best and maybe worst case studies for localization.