On Feb 6, 2006, at 7:49 AM, Zhong ZHENG wrote: Hi,
I am currently working on the portlet TCK, and trying to make it run against pluto.
Some test portlets have strange behavior... so i have to dive into the code...
I would like to know, if there are bugs in the portlet TCK, what are we supposed to do?
Should we fix the bugs or do something else? If we modify the TCK code, even if we
could pass the tests, we cannot say that pluto passes the *TCK*, right?
correct. There is a challenge process to whereby you can tell Sun you think there is a problem, and sometimes they respond, and sometimes they agree and exclude or fix the test.
So, any suggestions?
For those who have already tried the TCK, did you find any problems?
I found problems only in the web containers, not the portlet TCK. The problems were related to cross context dispatch, which is essentially unspecified anywhere except the portlet tck and to some extent the portlet specification. I believe that that geronimo+jetty has handled x-context dispatch properly for a year, geronimo+tomcat for a couple of days, and probably jetty standalone (5 series) for about a year. I have no information about tomcat. As for me, I found some problems. Should we discuss the TCK problems on the
public dev list, or should we have another place for that?
I suspect there is a NDA governing the contents of the portlet TCK. If so, you should definitely not discuss tck problems on the public dev list. At geronimo, we have a separate private email list for discussing the j2ee TCK. Unless the portlet tck is freely available to anyone, I think a similar email list for the portals group of projects would be a very good idea.
thanks david jencks
|
