Skip to content

Software internationalization, Pseudo Localization Poll Results

October 10, 2012

Alchemy Software Development conducted a number of online polls lately across approximate 300 unique webinar attendees.   One of the most important of these polls was setup up to make a first step towards advancing the tried and tested process of advanced pseudo localization by gauging this process’s utilization across attendees.

The basic finding of our polls were as follows:

41% of attendees currently have a Pseudo Localization process in place 59% of attendees currently do not have a Pseudo Localization process in place

To interpret these figures though we need to look further at the attendee base.

The attendee base was a very broad mixture of professionals with CTO’s, Software Developers, Localization managers, Localization Engineers, Localization Project Managers and Translators all being in attendance.   Additionally some of the above were from a Localization Service background, some freelance translators and freelance consultants and some were from development departments and localization departments within International Software Vendors.

I will follow with additional breakdown of the statistics.

Pseudo Localization Poll

October 3, 2012

An advanced Pseudo Localization Process is a requirement for Agile International Software Development!

September 27, 2012

Advanced Pseudo Localization processes are a requirement of Agile International Software Development. The software localization process will struggle greatly to keep up with the software development team sprints if high level pseudo localization processes are not applied vigorously at early development sprints. The Pseudo Localized builds then pass through a functional QA process in parallel with or indeed in place of source language builds.

Pseudo Localization enables the development of truly world ready software enabling true multilingual software sim-ship.

Pseudo localization is essentially the replacement of source text with readable pseudo-translated text. The pseudo translation language(s) can take a number of formats based mainly on the number of languages to be localised and the UI flexibility available.

The pseudo translated software UI must be readable text and therefore navigable by the software developers themselves the QA team and any others involved in the development process and for this reason a Machine Translation (MT) is not the correct process here.

Typically the minimum constituents of a pseudo language are text expansion and character substitution (“a” converts to “ä”). Text expansion, as English is one of the most concise languages. German for example can expand in areas by up to 50% creating a significant demand on UI layout creating great difficulty in PC application UI design but consider the effects of such significant expansion on a phone UI.

Pseudo languages can be quite complex and designed to include only tests for one language or many languages. The tests can include for example calculations for string length expansion based on the source string length as for example a short string might need a 100% expansion while a longer string may only need a 20% expansion.

Probably the most effective Pseudo localization process is the replacement process whereby a complete transformation to a language neutral position is made by the development team. The early development build QA takes place on Pseudo Localized builds rather than on source text builds.

The Pseudo Localization process checks your software/website for;

  • Text expansion/truncation issues
  • Externalisation of text (hardcoded text)
  • Extraction of too much text
  • Line break issues
  • Some locale specific issues (date, address, numerals)
  • Character encoding and font display issues
  • User interface design problems
  • String concatenation issues

Pseudo Localization