On Fri, Sep 5, 2008 at 9:57 AM, Emily Crutcher <[EMAIL PROTECTED]> wrote:
> That's what the weird test in "put" is for, as the current working > hypothesis is that removing all keys that appear in the prototype object > will prevent this problem. I went through and tested a bunch of them, so it > seems solid, but only time will tell if all the browsers/random word > combinations are licked. This is one of the reasons these classes are going > into incubator rather then gwt to begin with. > When HashMap was rewritten to use a colon prefix, there was a long discussion about trying to get "clean" objects and the conclusion was that you could not get reliable behavior on all browsers. If I remember correctly, adding the string "watch" on Firefox was one that couldn't be addressed without it, and there were other problems as well. -- John A. Tamplin Software Engineer (GWT), Google --~--~---------~--~----~------------~-------~--~----~ http://groups.google.com/group/Google-Web-Toolkit-Contributors -~----------~----~----~----~------~----~------~--~---
