Do ResponsePatternCollection patterns need to be translatable?

[jessegreenberg]

In ResponsePatternCollection, we have patterns that arrange the different Voicing responses.

utterance-queue/js/ResponsePatternCollection.ts

Lines 45 to 61 in 5626986

	const DEFAULT_RESPONSE_PATTERNS = {
	nameObjectContextHint: '{{NAME}}, {{OBJECT}}, {{CONTEXT}} {{HINT}}',
	nameObjectContext: '{{NAME}}, {{OBJECT}}, {{CONTEXT}}',
	nameObjectHint: '{{NAME}}, {{OBJECT}}, {{HINT}}',
	nameContextHint: '{{NAME}}, {{CONTEXT}} {{HINT}}',
	nameObject: '{{NAME}}, {{OBJECT}}',
	nameContext: '{{NAME}}, {{CONTEXT}}',
	nameHint: '{{NAME}}, {{HINT}}',
	name: '{{NAME}}',
	objectContextHint: '{{OBJECT}}, {{CONTEXT}} {{HINT}}',
	objectContext: '{{OBJECT}}, {{CONTEXT}}',
	objectHint: '{{OBJECT}}, {{HINT}}',
	contextHint: '{{CONTEXT}} {{HINT}}',
	object: '{{OBJECT}}',
	context: '{{CONTEXT}}',
	hint: '{{HINT}}'
	};

Do these need to be translatable? We do not know how they need to be combined for every language.

[jessegreenberg]

If the design is that each response type can stand independent then they do not need to be translatable. But what about puncuation? They use commans and we probably need a different separator for other languages.

Some languages (Chinese, Arabic, and Hebrew for example) do not use spaces to create boundaries between sentences.