On 07/12/12 02:21, Norbert Lindenberg wrote:
The benefit is that the ECMAScript Internationalization API lets
developers create a more consistent localized experience for their
users, with the correct
* date time, and number formats,
* the culturally appropriate calendar,
* correct currency symbols, and
* correct sorting.
Are you able to quantify the relative code/download size impact of these
features? I suspect sorting is the largest by far, but it would be
useful to know if that's correct.
It also helps avoid latency by removing the need to
send lists back to the server for sorting.
Of course, you can add per-column data indexes server-side to avoid this
without needing client-side collation information.
Google has already implemented the Internationalization API and is
shipping it in Chrome (still prefixed in Chrome 23), also by bundling
ICU into their downloads.
Did they use ICU anyway before they implemented this API?
User acquisition is an important goal of course. Has Mozilla studied
how it correlates with download size, e.g., by measuring what
percentage of users cancel out of downloads if the size is
artificially inflated?
I don't think we've ever artificially inflated the size, but as it
varies over time, we may have figures on how that's affected retention
rates. Asa?
Gerv
_______________________________________________
dev-platform mailing list
dev-platform@lists.mozilla.org
https://lists.mozilla.org/listinfo/dev-platform
