For this we'd need to run Spring's unit tests though. Our unit tests run fine. :-)
The long term solution is for the Spring integration to be updated to use the concrete implementation class instead of the deprecated interface. Clinton On Mon, Jul 7, 2008 at 3:05 PM, Gwyn Evans <[EMAIL PROTECTED]> wrote: > I suspect that you'll find that just X installations won't get anything - > I'm running with a Beta but haven't come across any of the issues recently > reported, possibly because I'm not doing anything particularly complex. My > suspicion is that it'll come down to just adding as much unit test coverage > as you can... > > /Gwyn > > > On Mon, Jul 7, 2008 at 6:30 PM, Clinton Begin <[EMAIL PROTECTED]> > wrote: > >> I think you fixed getDelegate, but getMappedStatement was the one that >> broke. Who knows if there are more. You did your best to test it against >> your known cases, but we it shouldn't be your responsibility alone. I think >> we need to introduce a more strict Beta period. Basically, until we get X >> number of confirmed installations of the beta, we don't make it GA. Perhaps >> we can include either sending the latest Beta to the Spring team to run >> their unit tests against too... >> >> The code doesn't change from beta to GA anyway, so if people install it, >> they won't have to reinstall when we move from Beta to GA. >> >> Clinton >> >> >> On Mon, Jul 7, 2008 at 11:09 AM, Brandon Goodin <[EMAIL PROTECTED]> >> wrote: >> >>> This is odd. I thought I fixed all the ExtendedSqlMapClient stuff. >>> >>> Brandon >>> >>> >>> On Mon, Jul 7, 2008 at 9:39 AM, nch <[EMAIL PROTECTED]> wrote: >>> >>>> >>>> Well, the only reason why I didn't decide to go for 2.3.2 is that I was >>>> confortable with 2.3.0 untill I found tomcat was not replicating sessions >>>> (I >>>> need this for clustering). So I took a look to 2.3.1 release notes which >>>> explicitly say "# [IBATIS-419] - LazyLoad object is not serializable." >>>> >>>> I tried to add the missing method to ExtendedSqlMapClient so, following >>>> instructions at >>>> http://opensource.atlassian.com/confluence/oss/display/IBATIS/Windows, >>>> I checked out the sources from trunk but they don't compile at this moment >>>> or ... at, least, I was not able to compile them. >>>> >>>> Cheers. >>>> >>>> ----- Original Message ---- >>>> From: Clinton Begin <[EMAIL PROTECTED]> >>>> To: [email protected] >>>> Sent: Monday, July 7, 2008 3:28:31 PM >>>> Subject: Re: spring 2.5.5 and ibatis 2.3.2 >>>> >>>> Darn, I guess not enough people tested it while it was in Beta. :-/ >>>> >>>> Perhaps we need a more formal testing group that can commit to testing >>>> the framework before we release it. I know Brandon uses Spring and did >>>> make >>>> a couple of fixes for Spring, but he's just one person. >>>> >>>> Out of curiosity, why didn't you upgrade or at least test it while in >>>> Beta? >>>> >>>> In a nutshell, the answer to your question is this: >>>> >>>> * We've deprecated a lot of internal interfaces in the core of iBATIS >>>> 2.3.2, in order to clean up the code. One of those interfaces was >>>> ExtendedSqlMapClient. >>>> >>>> * Code that used to cast (or othewise use) ExtendedSqlMapClient should >>>> now simply use SqlMapClientImpl (the default implementation behind the >>>> SqlMapClient interface). >>>> >>>> Cheers, >>>> Clinton >>>> >>>> On Mon, Jul 7, 2008 at 2:41 AM, nch <[EMAIL PROTECTED]> wrote: >>>> >>>>> >>>>> Hi, there. >>>>> >>>>> I'm trying iBatis 2.3.2 with Spring 2.5.5. >>>>> Spring is complaining it can not find method >>>>> com.ibatis.sqlmap.engine.impl.ExtendedSqlMapClient.getMappedStatement >>>>> This method was present in iBatis 2.3.0. >>>>> Is there a workaround I can take to make it work? >>>>> I need to test this cause it seems proxies coming from iBatis are now >>>>> serializable, so I can replicate sessions amont tomcat instances. >>>>> >>>>> Cheers >>>>> >>>>> >>>> >>>> >>> >> >
