You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Per @vicb this issue is mostly to summarize issues about Intl.API.
Currently our formatting pipes (Number and Date) highly coupled to Intl.Api, and this api have some known issues. From top overview it could be aggregated in following groups:
Compatibility issues between browsers. Some browser have limited support for Intl.Api (IE11, Edge), some browser not support it at all (IE8-10, Safari). Most of this issue can be easily resolved by adding proper pollyphil, but this point is not well informed (possible documentation issue).
Here list of compatibility issues:
date pipe formatter doesn't format minutes and seconds properly in IE 11.0.9600.18350 and Edge #9524 Issue with formatting minutes and second in IE11 and Edge . Chakra implementation of Intl.API little bit strange, so by some reason not allow to developer format date as single Minute or Second, that is required, by current implementation of date Pipe. As far as i know this is cannot be fixed by simply adding polyphil
Missed api. Some of requested feature could not be implemented upon standard Intl.API and require some additional implementation, which could be inconsistence with Intl.Api.
Here is a list of such issues:
DatePipe doesn't allow timezone to be provided as a pipe input #9324 - allow pass timezone as parameter. Current implementation of DatePipe allow to us implement basic time changing, but there will be some troubles with displaying name of specified time zone (current Intl.Api not allow us to do it).
Add milliseconds support on Date Pipe #7590 - add milliseconds support. This is not exist in Intl.Api but rather simple. If we implement just displaying it in A1 way no one harm, but we become inconsistence with Itnl.Api.
Pipes: Allow to specify locale #3429 - this one is not clearly related to Intl.Api, but still need to be mentioned. Right now we not quite flexible with our locale, from this point Intl.Api just fine. This issue is mostly aroung global locale setting for angular app and related to [i18n] future 2.x plan #9104.
[i18n] future 2.x plan #9104 - this one is just aggregation for [i18n] component. Right now your own code internationalization and angular system internationalization is 2 big difference. They not even correlated.
So right now there is a question - should we stand for Intl.Api and be consistence with it, or abondon Intl.Api and start to work with our own implementation for internationalization, so angular [i18n] module become more consistence.
Standing on Intl.Api, will lead to feature loss, since some of them not possible to implement with that api.
Implementing our own implementation of [i18n] is pretty much like A1 way, with main benefit - full control of Api, but yet there is an issues - we will have to think how to integrate this into current app bundles, and more over it requires some additional work. But we could implement this as part of #9104 [i18n] module, so it will fit in current A2 apps nicely.
Per @vicb this issue is mostly to summarize issues about Intl.API.
Currently our formatting pipes (Number and Date) highly coupled to Intl.Api, and this api have some known issues. From top overview it could be aggregated in following groups:
Here list of compatibility issues:
Here is a list of such issues:
So right now there is a question - should we stand for Intl.Api and be consistence with it, or abondon Intl.Api and start to work with our own implementation for internationalization, so angular [i18n] module become more consistence.
Standing on Intl.Api, will lead to feature loss, since some of them not possible to implement with that api.
Implementing our own implementation of [i18n] is pretty much like A1 way, with main benefit - full control of Api, but yet there is an issues - we will have to think how to integrate this into current app bundles, and more over it requires some additional work. But we could implement this as part of #9104 [i18n] module, so it will fit in current A2 apps nicely.