Question:
This is a followup-question to these posts:
i18n performance: resx vs. database?
Are resx files suitable for Internationalization?
If I would use resx files for internationalization (localization), under this scenario:
- ONE application (one source, one project, one executable) for all customers
- many customers
- multiple languages per customer
- customization of labels possible per language per customer
- application gets replaced from time to time with new version
Is it possible with resx files to maintain customization of field labels ? I mean for one customer and one language, it would certainly be possible.
But for N customers with N languages...
Wouldn't I be required to create N resx files for N customers and every string I add in one file, I must add to every resx files ? Also I would have to take back resx files (as customization happens at the customer).
Is a resx file really suitable for this ?
To me it sounds like it is not (not at all), but on these two posts, people seem to perceive it as the best solution.
This appears doubtful to me.
To me it appears as if resx files just can't deliver the flexibility and maintainability required, and it also appears to me, that under a database-driven i18n, this would be comparatively easily implemented.
Does anybody who works with resx files have actual experience with this kind of scenario, or used and maintained their own product ?
