2011/8/11 Tatu Saloranta <[email protected]> > On Tue, Aug 9, 2011 at 5:15 PM, Ivan <[email protected]> wrote: > > Stax TCK should be a bit different comparing others, it is somewhat > > 'public', IIRC. In the past time, Geronimo uses a very old 3.2.7 version, > > Do you know of a link to freely downloadable copy? It would definitely > be nice to have access, it just has seemed to be less than public in > the past for some reason. > And I haven't fully understood what is the official stance; I assume > practices and policies vary between JSR/JCPs, meaning that spec lead > can choose to publicize different things (except they must be shared > for official participants anyway). >
Just confirmed that, it is a bit difficult to get the TCK packages, :-( > > > which seems to be fine. > > Right; Woodstox made the chance to use "" instead of null in a few > places specifically to try to be more compliant, since consensus was > at the time that XML APIs generally expose empty/no-namespace with > empty Strings, not nulls. > Exceptions obviously being cases where we could see explicit wording > stating otherwise, which seems to be what occurs here. > > Can you add a Jira request for this? This probably should be changed > for next minor or major release, as our goal definitely is to be > standards compliant. > Sure, I will open a Jira. > > > Yes, those specification sometimes did not clarify things clearly, and > think > > whether a system property flag could be used to change between the > behavior, > > by default, it might keep the empty string return value. > > That is one possibility, although challenge here is that any changes > to public API (including additions of new configuration property) > should go in new minor release. > > In the past I think most Stax API users have accepted both null and > empty String in most places, given that implementations have differed. > > Thank you for pointing out this issue, I appreciate your help in resolving > this. > I am bit frustrated with the specification itself (it has been a long > journey...), but it is what it is. :-) > > Really be appreciated for following this issue, > -+ Tatu +- > > --------------------------------------------------------------------- > To unsubscribe from this list, please visit: > > http://xircles.codehaus.org/manage_email > > > -- Ivan
