Terms in Transifex
Below is a list of common terms you might encounter when localizing or using Transifex.
The Transifex Client is a git-like command-line tool that lets you easily work with large volumes of translation files. Using the client, you're able to sync files between your local device and Transifex without a UI.
Context is information that helps translators understand the meaning behind what they’re translating. This includes providing screenshots and developer notes with each string.
Crowdsourcing is the process of translating with volunteers from your user community. It can be an effective way to translate content if you have a group of passionate users who love your product.
File formats are industry-standard ways for storing and exchanging translation content. Each development framework, programming language, and content platform will generally have its own file format.
The Editor is the environment in Transifex where translators do their work. It lets translators see your source content and submit translations directly from their browser, without having to handle any files directly. The Editor also provides translators with tools such as Translation Memory and Glossary to help them deliver high quality translations.
A glossary is a set of key terms along with their part of speech, definitions, and translation. Glossaries help ensure consistency in translations. And they can be helpful when working with industry- or brand-specific terms. Often times, your Language Service Provider can help you create a glossary before translations begin.
Translating in context is a new approach to translation aiming to give translators better context around what they’re translating. In-context translations usually work through a WYSIWYG-style editor. It first lets translators see (and often interact with) the live version of whatever they’re translating. Then as the translator types a translation, it’s overlaid onto the page, giving the translator immediate feedback into how their translation will look.
Internationalization, often abbreviated as “i18n,” is the process of adapting software so it can be used in different languages. This includes marking strings in code as translatable so it can be extracted into a separate file for translation.
Issues are questions or problems related to a string. Anyone can open an issue for a string. For instance, a translator could open an issue for a typo in a source string.
Machine Translation (MT)
Machine Translation is the use of software to translate content. If you’ve ever used Google Translate, you’ve used machine translation. While MT is generally not recommended for customer-facing content, it can used in conjunction with human review in a process known as “machine translation with post-editing.”
Language Service Provider
A Language Service Providers (LSP) is a translation agency. While LSPs will vary in what services they offer, they all usually provide professional translation and review services to start. Some LSPs specialize in a single language. These are referred to Single Language Vendors (SLVs). And others provide services in multiple languages. These are known as Multi-Language Vendors (MLVs).
A locale refers to a language, or a language spoken in a specific region. Locales are generally represented with two- or four-letter codes. For example, fr refers to French, and fr_CA refers to Canadian French. Transifex uses the language codes specified by the ISO 639-1 standard.
A project is a container for storing your source content and their translations. Projects in Transifex can be private (visible only to those you invite) or public (visible to anyone).
Resources are the content you’re translating along with their corresponding translations. Let’s say you’re translating the file my_file.po. That file together with with the translated versions – my_file_fr.po and my_file_es.po would form a single resource.
The source language is the language you are translating from. To put it another way: your source language is the language your app or website is originally written in.
A string is a sequence of characters. A string can be a word, phrase, or sentence. It can contain numbers as well as variables, and even HTML tags. “Save settings” and “Please look at %(url)s for more detail.” are both examples of strings.
A style guide is a set of instructions provided to translators explaining what kind of capitalization, words, tone, etc. should be used when translating. It’s vital to keeping your brand consistent across languages.
Target languages are the languages you are translating to.
A team in Transifex is a group of people who work together to translate one or more projects. Teams are made of Team Managers, Language Coordinators, Reviewers, and Translators.
A translation check is an automated check of a translation against a set of rules. For example, checks can ensure HTML tags and URLs which are in the source string are present in the translation. Translation checks make sure translations don’t break your app, and the original content is preserved.
Translation Memory Autofill
Translation Memory Autofill lets you automatically use 100% TM matches as translations. This saves you time and money as you won’t have to retranslate those strings or manually copy and save translations inside the Editor.
Translation Memory (TM)
Translation Memory is a database of previously translated strings and their translations. If an untranslated string is similar to a translated string, the translation for the translated string will appear as a suggestion inside the Editor. Each suggestion is marked with a % showing how similar the two source strings are. So a 100% “match” would mean the two strings are exactly the same. With TM, translators able to translate faster while ensuring consistency in their work.
Translation Memory Exchange (TMX)
Translation Memory Exchange (TMX) is an industry standard file format for Translation Memory. It lets you move TM data between systems and tools without losing any of your translations.
Translation Memory Leverage
TM leverage is a breakdown of TM matches for a set of strings. It helps you estimate how much new translation work needs to be (or was) done, and how much you can rely on the TM. For example, a resource with mostly 95-99% TM matches is likely much easier to translate than one with mostly 60-65% matches.