Comparison of the two models

There are advantages and disadvantages to each localization model. This section describes the pros and cons of each model so that you can select the model that best fits your needs.

Sequential localization model

The sequential localization model is the simpler of the two models; it creates a simpler content structure in the localization folder with less content duplication. It has the following advantages:

  • There is a one-to-one relationship between the files in localization and their authoring sources.
  • The files in localization have the same names as the files in authoring.
  • The localization model does not need to remap links.
  • The files in localization are more independent from each other, which makes it easier to do "localization in advance", that is, localizing some content before the entire document is complete in authoring.
  • The localization process is faster.

However, the sequential localization model does not allow an object to be sent for translation if a previous revision is still being translated. It has to wait for the previous translation to be completed. A new revision can be sent for translation only if:

  • The previous revision has not yet been sent for translation (i.e., its status is at "localization:tb translated").
  • The previous revision has been translated and reviewed (i.e., its status is at "localization:done").

Concurrent localization model

The concurrent localization model is more flexible than the sequential model and has the following advantages:
  • New revisions of the source content can be sent for translation at any point in time.
  • If two maps use the same topic but they are published with different revisions of the topic, both maps can be translated at the same time, since each will use a different file for the topic.

However, this parallel model creates more objects in localization than there are in authoring, so the size of the database keeps growing even if the volume of source content is stable. The growth rate is multiplied by the number of languages in which the content is translated.

Also, if a new revision is sent for translation before the previous one has completed its translation process, the new revision does not pick up the translation from the previous one, which may lead to duplicate translations. Therefore, translation managers must carefully control when content is sent for translation in order to maximize reuse and lower cost.