On Thu, Mar 21, 2013 at 9:16 PM, Sebastián Gurin <[email protected]>wrote:

> I think this can be really interesting to those looking to find GWT jre
> emulation bugs. I myself just discovered some of them listed on
> http://gwtmauve.googlecode.com/svn/trunk/gwtmauve/readme.txt
>
> (rich conclusions of this project results can be taken from these bugs)
>
> Well, I hope this can be useful to those who want to find and fix GWT bugs
> and also to know which parts of the GWT JRE emulation are "more broken" in
> production mode. Use it as you like. Had to release it under gpl3 for
> respecting mauve sources license. Comments and ideas are most welcome.
> Regards
>
> Thanks, this is useful.  Note that clone isn't supported (it can't be
supported in general in JS), and generally you shouldn't expect the same
exceptions to be thrown where doing so is inefficient -- see
https://developers.google.com/web-toolkit/doc/latest/DevGuideCodingBasicsCompatibility.
 For example, Java String is directly mapped to a JS string, so the only
way to give the same IOOB exceptions would be to add checks around every
index/substring/etc operation.  Likewise for many NPEs.

That explains a large chunk of them.  Some of the others, such as
Character.toTitleCase, are because they can't be implemented properly and
shouldn't really be used anyway (in many cases, the title case of a
character should be multiple characters).  BigInteger is itself a slow
emulation in JS, so I am not surprised that isProbablyPrime is extremely
slow on a large number.

The others, such as the missing Collections.rotate and bugs in Arrays.fill,
Collections.copy, and String.compareTo are good catches and should have
issues filed.

-- 
John A. Tamplin

-- 
-- 
http://groups.google.com/group/Google-Web-Toolkit-Contributors
--- 
You received this message because you are subscribed to the Google Groups 
"Google Web Toolkit Contributors" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
For more options, visit https://groups.google.com/groups/opt_out.


Reply via email to