Thanks for explaining Alex. Then it should be fine as is. You can see the
tests I added for this.
I thought the typedefs might affect code completion in editors  or
something like that - like I said I have not done anything with that in the
past, so its something I need to learn.

thanks,
Greg




On Fri, Nov 9, 2018 at 8:34 PM Alex Harui <[email protected]> wrote:

> Hi Greg,
>
> I don't think I understand your goals here.  I didn't know there were bugs
> in the current code, but I'm not surprised either.  I was just trying to
> get some other code to work.
>
> Why do we need to change the typedefs?  IMO, the typedefs should represent
> what the browser actually supports and the compiler and Language should
> deal with any ActionScript differences.  The emitters are where I think
> that should happen.  The RoyaleEmitters generate Royale-specific code, but
> other emitters like the GoogEmitter should allow someone to use Google
> Closure Library and write directly at the browser APIs (not the Flash APIs).
>
> My 2 cents,
> -Alex
>
> On 11/8/18, 11:06 PM, "Greg Dove" <[email protected]> wrote:
>
>     I see some discussions about messages not getting through on the list
>     today, so I can't be sure the last one or this one got through.
>     I pushed the commit, but please revert if if you don't want it in the
>     release yet.
>     thanks,
>     Greg
>
>
>     On Fri, Nov 9, 2018 at 3:48 PM Greg Dove <[email protected]> wrote:
>
>     > Hi Alex,
>     >
>     > I checked and only now realized that you made the change in
> Language.as
>     > quite recently to add support for actionscript Array.sort method
> signature
>     > variations to javascript - I looked in Language.as and saw the
> support for
>     > it but did not check when it was added until just now - last month!
> I hope
>     > I did not double up on the compler side which you may have already
> begun as
>     > well... I have a  fix ready for the 2 method signatures that need to
> use
>     > that Language.sort call with the options argument, and new tests in
>     > compiler-jx for that.
>     > As I suspect you are aware, currently it gives wrong, but errorless,
> code
>     > in some places, like mx.colllections.Sort, so I am keen to get this
> in.
>     >
>     > Does anything else need doing for  typedefs - do I need to update an
>     > existing method signature, and if so, how? (I didn't do anything in
>     > typedefs so far)
>     >
>     > ...if so, I *think* it just needs the optional options arg...
>     >
>     > If the typedefs change can wait, or is not needed, please let me
> know and
>     > I will push this fix asap, otherwise please point me in the
> direction of
>     > what I need to do for typedefs.
>     >
>     > thanks,
>     > Greg
>     >
>     >
>     >
>     >
>     >
>     >
>     >
>
>
>

Reply via email to