Dear Sugar Developers, Below is a message I sent out to the localizers. Developers can improve their i18n process by checking upstream to see if there is a string that can be reused (in context). Even slight changes in phrasing to match the upstream can be a big help. 1) Developers can enhance i18n by re-using upstream strings (where feasible), helping our localizers find pre-localized strings. 2) In the Sugar UI this produces a long-term (very subtle) vocational training effect, the kids will "just get it" when moving to other FOSS apps / OSes. "Oh, I know what that menu option does. . . " We've all experienced this "training". When you try a new application, you've got a pretty good idea what you will find on the "Edit" menu before you even look. However if it said "Change" or "Modify" or "Alteration", you'd have to spend extra time sorting it out. Please give our users the benefit of seeking to maintain consistency across FOSS applications where it is possible. At the same time, this can give our localizers a chance to just copy the simple strings from upstream and focus on the more complex ones or on doing other PO files. 3) For languages that we have, but upsteam doesn't, we can potentially create an upstream flow of pre-localized strings feeding into the broader FOSS ecosystem, creating opportunities for native language use in "grown-up" tools. All in all, this a virtuous cycle, and all it requires is a little extra effort to check your strings for general consistency with the wider FOSS world, something you probably do unconsciously, but a conscious effort will add to the benefits. Warmest Regards, cjl <div class="gmail_quote">---------- Forwarded message ---------- From: Chris Leonard <<a href="mailto:cjlhomeaddress@gmail.com">cjlhomeaddress@gmail.com</a>> Date: Wed, Oct 12, 2011 at 11:20 AM Subject: Attention all localizers To: OLPC Localization list <<a href="mailto:localization@lists.laptop.org">localization@lists.laptop.org</a>> There may be an excellent opportunity to rapidly pick up some good suggestions for your languages project's strings. <a href="http://open-tran.eu/" target="_blank">http://open-tran.eu/</a> If your language is listed here: <a href="http://open-tran.eu/languages.html" target="_blank">http://open-tran.eu/languages.html</a> you can rapidly search these upstream projects: <a href="http://open-tran.eu/projects.html" target="_blank">http://open-tran.eu/projects.html</a> to see of there is an identical or closely matching string already translated upstream. This works quite well for a lot of the shorter strings and even for some of the very "technical" computer-ese ones, but obviously longer strings that are specific to Sugar will not be matched upstream. This is a translation memory of upstream localized strings, not a machine translation. If you open two browser windows, one open to Poolte and one showing <a href="http://open-trans.eu" target="_blank">open-trans.eu</a>, you can fairly quickly copy/paste a largish number of entries to look for suggestions in a relatively quick timeframe. This sort of analysis can also be done to see of our strings are translated in a similar way to those upstream. As the Open-trans motto says "Consistency matters". Warmest Regards, cjl Sugar Labs Translation Team Coordinator </div>