I'm torn between the little benefit this fix provides, versus the little exposure the feature has. There are not too many folks using the metadata facilities on requestcycle and Application, so probably it will not hurt too many folks.
However to keep it pure, I'd say revert, better safe than sorry, and re-apply in 1.4. Martijn On 2/18/08, Matej Knopp <[EMAIL PROTECTED]> wrote: > Yeah, i was wrong. I wonder why i was thinking that... > > So? Do we stick to the binary compatibility even if requiring > serializable at that particular place makes no sense? > > -Matej > > On Feb 18, 2008 11:25 PM, Johan Compagner <[EMAIL PROTECTED]> wrote: > > Nope doesnt work > > it wants an exact match including return type: > > > > F:\workspace_wicket\Test2\bin>java test2.Test2 > > Exception in thread "main" java.lang.NoSuchMethodError: test.Test1.getObject > > (Lja > > va/io/Serializable;)Ljava/io/Serializable; > > at test2.Test2.main(Test2.java:11) > > > > johan > > > > > > > > On Feb 18, 2008 11:19 PM, Martijn Dashorst <[EMAIL PROTECTED]> > > > > wrote: > > > > > Easy to test... > > > > > > Generate a quickstart on wicket 1.3.0, use the metadata, replace > > > wicket dep with 1.3-snapshot. > > > > > > Martijn > > > > > > On 2/18/08, Matej Knopp <[EMAIL PROTECTED]> wrote: > > > > Actually, i think the getMetaData method should be resolved correctly. > > > > But I might be wrong. > > > > > > > > -Matej > > > > > > > > On Feb 18, 2008 11:14 PM, Matej Knopp <[EMAIL PROTECTED]> wrote: > > > > > Yeah, for setMetaData. But will that work for getMetaData as well? > > > > > > > > > > -Matej > > > > > > > > > > > > > > > On Feb 18, 2008 11:11 PM, Martijn Dashorst <[EMAIL PROTECTED]> > > > wrote: > > > > > > The adapters you have provided should maintain binary compatibility, > > > don't they? > > > > > > > > > > > > Martijn > > > > > > > > > > > > > > > > > > On 2/18/08, Matej Knopp <[EMAIL PROTECTED]> wrote: > > > > > > > Hi, > > > > > > > > > > > > > > I've changed the RequestCycle.set/getMetadata as well as > > > > > > > Application.set/getMetadata methods to take Object instead of > > > > > > > serializable. It doesn't make sense to require serializable > > > arguments > > > > > > > for this data, that's the purpose of storing things in request > > > cycle. > > > > > > > Unfortunately this breaks the binary compatibility as method > > > signature > > > > > > > changes (I didn't realize this when committing). > > > > > > > > > > > > > > Now the question is? Should I revert it? I know we try to maintain > > > > > > > binary compatibility between releases. On the other hand, this is > > > a > > > > > > > rather small change and IMHO won't influence the vast majority of > > > > > > > users. > > > > > > > > > > > > > > Anyway, I'm prepared to revert it, if others think i should. > > > > > > > > > > > > > > -Matej > > > > > > > > > > > > > > > > > > > > > > > > > -- > > > > > > Buy Wicket in Action: http://manning.com/dashorst > > > > > > Apache Wicket 1.3.1 is released > > > > > > Get it now: http://www.apache.org/dyn/closer.cgi/wicket/1.3.1 > > > > > > > > > > > > > > > > > > > > > > > > > > -- > > > > > Resizable and reorderable grid components. > > > > > http://www.inmethod.com > > > > > > > > > > > > > > > > > > > > > -- > > > > Resizable and reorderable grid components. > > > > http://www.inmethod.com > > > > > > > > > > > > > -- > > > Buy Wicket in Action: http://manning.com/dashorst > > > Apache Wicket 1.3.1 is released > > > Get it now: http://www.apache.org/dyn/closer.cgi/wicket/1.3.1 > > > > > > > > > -- > Resizable and reorderable grid components. > http://www.inmethod.com > -- Buy Wicket in Action: http://manning.com/dashorst Apache Wicket 1.3.1 is released Get it now: http://www.apache.org/dyn/closer.cgi/wicket/1.3.1
